Educational Outcomes for the Canadian Workplace: New Frameworks for Policy and Research 9781442674295

Table of contents :
Contents
Introduction. Towards a Research Program in Education and Training
PART 1. What Skills Matter in the Economy? Economic Approaches
1. Education, Skills, and Labour Market Outcomes: Exploring the Linkages in Canada
2. Education and Technological Revolutions: The Role of the Social Sciences and the Humanities in the Knowledge-Based Economy
PART 2. Achieving Equity: Three Analyses of Outcomes
3. What Are Canadians Doing after School? An Analysis of Post-school Training Activity
4. The Post-secondary Education of Disadvantaged Adults
5. What Outcomes Matter to You? Exploring Welfare Policy and Programs from the Perspective of Low-Income Women
PART 3. Policy and Practice: Case Studies Linking Education and Work
6. Policy, Rhetoric, and Educational Outcomes: Interpreting Skills Now!
8. Co-op Education: Tensions and Outcomes of Experiential Learning
Conclusion. Learning from Research Networks: The Western Research Network on Education and Training, 1996-2001
Contributors

Citation preview

Educational Outcomes for the Canadian Workplace: New Frameworks for Policy and Research

This page intentionally left blank

Educational Outcomes for the Canadian Workplace New Frameworks for Policy and Research

Edited by JANE GASKELL and KJELL RUBENSON

UNIVERSITY OF TORONTO PRESS Toronto Buffalo London

www.utppublishing.com University of Toronto Press Incorporated 2004 Toronto Buffalo London Printed in Canada ISBN 0-8020-8845-7

Printed on acid-free paper

National Library of Canada Cataloguing in Publication Educational outcomes for the Canadian workplace : new frameworks for policy and research / edited by Jane Gaskell and Kjell Rubenson. Includes bibliographical references. ISBN 0-8020-8845-7 1. Labour supply - Effect of education on - Canada. 2. Education Economic aspects - Canada. 3. Working class - Education - Canada. I. Gaskell, Jane S. (Jane Stobo) II. Rubenson, Kjell, 1944HCC120.E47E38 2004

331.12'0971

C2004-900282-1

University of Toronto Press acknowledges the financial assistance to its publishing program of the Canada Council for the Arts and the Ontario Arts Council. University of Toronto Press acknowledges the financial support for its publishing activities of the Government of Canada through the Book Publishing Industry Development Program (BPIDP).

Contents

Introduction: Towards a Research Program in Education and Training 3 JANE GASKELL AND KJELL RUBENSON

Part 1: What Skills Matter in the Economy? Economic Approaches 19 1 Education, Skills, and Labour Market Outcomes: Exploring the Linkages in Canada 21 W. C R A I G R I D D E L L

2 Education and Technological Revolutions: The Role of the Social Sciences and the Humanities in the Knowledge-Based Economy 56 R O B E R T C. A L L E N

Part 2: Achieving Equity: Three Analyses of Outcomes

87

3 What Are Canadians Doing after School: An Analysis of Post-school Training Activity 89 D E R E K P.J. H U M A N D WAYNE S I M P S O N

4 The Post-secondary Education of Disadvantaged Adults 118 ATLANTA SLOANE-SEALE, LORI WALLACE, AND BEN LEVIN

5 What Outcomes Matter to You? Exploring Welfare Policy and Programs from the Perspective of Low-Income Women 138 SHAUNA BUTTERWICK

vi Contents Part 3: Policy and Practice: Case Studies Linking Education and Work 157 6 Policy, Rhetoric, and Educational Outcomes: Interpreting Skills Now! 159 LARA M. LACKEY

7 Working Outcomes in the Classroom: A Case Study of Applied Academics in British Columbia 186 J I M G A S K E L L , C Y N T H I A N I C O L , A N D L I - L I N G TSAI

8 Co-op Education: Tensions and Outcomes of Experiential Learning 204 GARNET G R O S J E A N

Conclusion: Learning from Research Networks: The Western Research Network on Education and Training, 1996-2001 220 JANE GASKELL AND DEANNA McLEOD

Contributors 237

Educational Outcomes for the Canadian Workplace: New Frameworks for Policy and Research

This page intentionally left blank

Introduction

Towards a Research Program in Education and Training JANE GASKELL AND KJELL RUBENSON

Approaching the year 2000, nations around the world are struggling to position themselves in the skill-developing, knowledge-based global economy. Canada's success in this effort, like that of other developed countries, will depend on the education and skills of its people. Canadians will require extensive education, training and retraining to participate effectively in the new economy. There is a clear need, therefore, to give priority to the quality, effectiveness, responsiveness and resourcefulness of education and training in all facets of the learning continuum in Canada. (Social Sciences and Humanities Research Council of Canada, 1995)

Recognizing the necessity of more interaction among disciplines and researchers in the area of education and training, the Social Sciences and Humanities Research Council of Canada (SSHRC) issued a call for proposals in October 1995, to establish five-year strategic research networks linking researchers and research partners involved in policy and practice. A network dubbed WRNET (the Western Research Network on Education and Training) was assembled at the University of British Columbia (UBC) to respond to this call. This book arose out of WRNET's research to inform the Canadian debate about education and training, and their relationship to work in the 1990s. Debates about the meaning of a new global economy for educational policy and practice were taking place in many countries. Although the Canadian debate contained many elements that were found elsewhere, they had a particularly Canadian twist, reflecting the distinct political and economic context in this country. Moreover, because the federal government has only limited power to shape edu-

4 Jane Gaskell and Kjell Rubenson

cational discourse and policy, the debate in each province and territory across the country differed, reflecting party politics, funding priorities, and dominant cultural traditions. The Canadian debate was constant, but a lot of it was local and not well documented. The discussion rarely was informed by or subject to systematic research, although statistics were often quoted, and beliefs about 'what causes what' were firmly held. The chapters of this book reflect the way a group of western Canadian academics responded to the call for research on education and the economy in the mid-1990s. They begin to document the directions that educational policy took, and they provide some evidence that will allow the reader to examine these policies critically. Canada's educational system ranks high on the indicators of educational quality published by the Organization for Economic Cooperation and Development (OECD). Canada spends more on its public educational system, and has higher levels of attainment, higher levels of literacy, higher scores on science and math tests, and less inequality in outcomes than most other countries (OECD 2001). And our education system is getting better. The OECD noted that Canada's math and science achievement scores improved more than any other country's from 1995 to 1999 (OECD 2001). The rate at which students fail to complete high school declined by one-third in the 1990s - falling from 18 per cent in 1991 to 12 per cent in 1999 (Statistics Canada 2002). While Canada, like other countries, has a substantial number of people whose literacy falls below the level that OECD identifies as necessary to work and adjust to a knowledge economy, and while there are countries that do better than we do, Canada has a relatively good record and thus a comparative advantage in its educational system. Despite Canada's relative educational success, however, the 1990s were replete with complaints about the poor quality of Canadian training and schooling. Policy-makers showed a pronounced tendency to analyse the failures of the educational system, and to proclaim solutions that could address Canada's place in an increasingly competitive global economy. For Canadians may have a stronger educational system than the United States on most indicators, but in the 1990s our economy was not stronger. The American economy grew faster, offered higher wages, and had a lower unemployment rate. The Canadian dollar fell in the face of the American economy's power. The productivity gap between the United States and Canada focused the attention of federal politicians, and this productivity gap was linked to education.

Introduction 5

Although there is some validity to the notion that the skills of workers are developed through educational and training systems, and that skilled workers perform better, the relationship between education and the economy is certainly not a simple one, as the 1990s illustrate. Canada's relatively strong education system did not translate into a strong economy. The questions of what linkages do exist between education and the economy, and how these linkages should be understood, measured, and changed, have been at the forefront of Canada's educational policy debate, yet there is alarmingly little Canadian evidence with which to address them. This book explores education and its links to the Canadian economy, with a focus on western Canada. The educational politics of western Canada do not simply reflect the politics of central Canada, although the federal government and the connections among provincial and territorial ministries of education shape what is seen and believed (Taylor 2001). Schools in the western provinces appear in fact to be doing relatively well compared to schools elsewhere in the country, but in the 1990s, discussion of education in the west reflected the national focus on educational failure, with a distinct twist in each province. British Columbia's polarized political parties came together to focus on making education relevant to an economy moving from the primary industries of forestry, mining, and fishing towards an urban service economy with an emphasis on technology and information industries. Alberta led the move towards cost-cutting and accountability in public services, including education, while Saskatchewan moved to integrate social services with education, and made a concerted attempt to address Aboriginal issues. Manitoba strove to address equity and increase the level of education. These chapters provide some research findings, some documentation of policy, and some exploration of the economic, political, and educational beliefs that inform it in the context of western Canada. Human Capital Theory: The First Wave Preoccupation with the role of education in economic growth is certainly not a new phenomenon: Adam Smith mentioned it in 1776 in The Wealth of Nations. However, it was only after the Second World War that 'human capital' became a central area of economic analysis. What came to be called human capital theory pointed to strong linkages between investments in education and economic growth for both

6 Jane Gaskell and Kjell Rubenson

individuals and countries (Denison 1962; Schultz I960,1961). An individual's income was seen to be dependent on, in fact a sign of, their 'capital/ that is, the knowledge and skills that make them productive. Human capital theory made into a commonplace the view that as technology increased, and the skills demanded at work grew, the necessity of investing in education and training also grew. Human capital theory was widely endorsed by policy-makers in the 1960s, at a time of rapid economic growth and acceptance of Keynesian economics (Lingard, Ritzvi, and Taylor 2001). The first OECD conference on education, Economic Growth and Investment in Education, took place in Washington in 1961. It assembled a group of the world's leading economists and produced a series of policy reports that inspired many countries to develop strategies to expand education in the name of economic prosperity. The influential reports that flowed from this conference include Economic Aspects of Higher Education (OECD 1964), Econometric Models of Education (1965a), The Residual Factor and Economic Growth (1965b), and Financing of Education for Economic Growth (1966). What made the first generation of human capital theory and research so influential was its appeal to all sectors of the political landscape. Increased public investment in educational expansion would not only result in benefits to the gross domestic product (GDP); it would also equalize income structures (Schultz 1960). Market processes were expected both to generate a more equal distribution of earnings and to modernize production through technological and organizational change (OECD 1961). Education was taken, fairly unproblematically, to stand for the positive personal characteristics and knowledge that are imparted through and rewarded during schooling. Research indicated that education was a critical factor in predicting adult economic position, and one that was becoming more important over time (Blau and Duncan 1967). Implicit in this model is a competitive, meritocratic society where individual characteristics are recognized and rewarded by employers. In Canada, federal and provincial policy documents that argued for a rapid expansion of the post-secondary system as a route to economic growth in the 1960s refer to OECD documents, and draw on Canadian research (Economic Council of Canada 1965; Bertram 1966; Porter 1965). Federal policy-makers wanted to expand the existing provision of education to serve a broader segment of the population. A highly selective and differentiated educational system, where only a small

Introduction 7

group - overwhelmingly Protestant, male, and from the upper socioeconomic classes - went beyond secondary education, was seen as a major hindrance to economic development. The problem was defined as the scarcity of education, not its quality. The solution was a more inclusive curriculum, a government-funded expansion of education at all levels, and an increase in student aid. An ambitious program of establishing new post-secondary institutions was designed to maximize the accumulation of human capital while at the same time creating avenues for social mobility. The economic as well as the political context changed dramatically through the 1970s and 1980s. As education expanded, the world economy faltered and the optimism that had driven educational expansion was severely shaken. Despite a vast increase in public expenditure on education, and rapidly increasing levels of educational attainment, economic growth was declining. Some educators mounted a critique of the human capital approach, worried that education was being treated as an economic good and unconvinced that education was reducing inequality. Liberal education would be undermined as more students were enticed to stay in school solely for the economic payoff (Wilson 1972). Vocational classes were a dead end for students, streaming them in a class-based educational system, in preparation for a highly inequitable job market (Bowles and Gintis 1976). Inequality remained very pronounced in Canada, despite the growth in educational opportunity (Fleming 1974). More education could, ironically, stream students more efficiently and completely (Jencks 1972). This disquiet drew on an alternative analysis of the workplace, and of the impact of technology. If technology displaced workers, and the workweek was destined to shrink, education for work was a shortsighted, low-level goal. Analysts began to wonder whether workers were over-skilled for their jobs, and whether technology was decreasing, rather than increasing, the need for skills in the workplace (Berg 1970; Braverman 1974). Writers like Collins (1979) pointed to the historical specificity of the relation between education and labour market outcomes, and provided a cultural analysis of why employers might decide to hire and promote more educated workers. The limited ability of education to combat economic, social, and, more specifically, educational inequalities was clear (Husen 1979; Karabel and Halsey 1977). This work opened up the question of what education does for workers and what employers look for when they hire, pointing out that it is

8 Jane Gaskell and Kjell Rubenson

historically and socially specific and based in how particular labour markets operate (Holzer 1996). Technical attempts to measure skill at work (Spenner 1983) came up against critiques that emphasized the shifting, politically contested, and context-specific nature of what is valued in the workplace (Fudge and McDermott 1991; Brown 1995). Frameworks like signalling theory, or analyses of segmented labour markets, explored how the structure of the economy and the demand for labour shaped returns to education. Debates about the effects of deskilling (Bluestone and Harrison 1989) and about the social organization of particular industries (Osberg, Wien, and Crude 1995; Piore and Sabel 1984) pointed to the importance of understanding the changing nature of the global economy, the impact of technology, and the social structure of particular firms and markets in understanding the economic consequences of education. Analysis of women's experience at work pointed out that education had not paid off for them, and that explanations must take into account the organization of the educational system as well as the characteristics of jobs that men and women were hired into (Blackmore 1992; Gaskell 1992). The accounting literature discussed how much training actually costs and how its results might be captured as assets of a firm (Stewart 1994). This was all informed by the literature on 'communities of practice' (Lave and Wenger 1991), which questioned the ease of transferring knowledge and understanding from one context to another, and stressed instead the importance of relating learning closely to the places and people where it will be practised. The formal characteristics of school-based learning were often seen to be quite different from formal characteristics of work-based learning, and schools were asked to embrace a broader set of contexts for learning (Resnick 1987a, 1987b). Nespor (1994), using the concept of 'actor-networks' (Latour 1988; Gallon 1986, 1987), argued that the outcomes of schooling can only be understood by looking at how its practices prepare students to participate in the networks that constitute fields of practice such as disciplines or industries. From this point of view, it becomes as important to study network structures and the political economy that sustains them, as it is to study students' experiences in specific settings of pedagogy or practice. Questions about what counts as 'skill,' how it is measured, accounted for and developed, as well as which intellectual, social, technical, and personal factors count for whom, opened up a complex and interesting research agenda. There were major academic disagreements about

Introduction 9

what the significant and labour market relevant outcomes of schooling were. But the complexity of this debate lost the attention of policy-makers. Through the 70s and 80s, reflecting both the political context and the new, less definitive research, human capital concerns lost their centrality at OECD. In Canada, educational policy became more firmly provincial, and centred on accountability, funding, and curriculum redesign. Human Capital Theory Revisited: The Second Wave At the end of the 1980s, OECD again turned its attention to education as the generator of economic growth, within a political and economic landscape that had dramatically altered since the 1960s. This was an era of global capitalism, characterized by increased economic competition and rapid advances in information technology. It was also an era of restraints on public spending, and of belief in the private sector. A neo-liberal framework had replaced the Keynesian creed. Educational policy was shaped by severe limitations on public expenditures, a general suspicion of the state and public institutions, and a belief in the greater efficiency of free market forces. In this context, OECD's 1989 report, Education and the economy in a changing society, ushered in what has been labelled the second generation of human capital thinking in educational policy (Marginson 1997; Rubenson 1992). The report noted, as earlier analyses of human capital had, that national differences in economic performance could increasingly be attributed to varying degrees of educational effectiveness and a country's learning capabilities. This report differed from earlier reports, however, in pointing out that '"education" is becoming less clearly distinct from that which is "the economy"' (OECD 1989, 19), and that '"Education and the economy" has become a catch-phrase for a vague but urgent dissatisfaction with the status quo' (17). The report developed a critique of education because of its tenuous links with the economy. This second-generation analysis of human capital grew from research on decision making and the uses of technology at the level of the individual firm. It cited research observing that the introduction of technology generates demand for increased skills, which in turn promotes technological change and ultimately results in productivity gains (Welch 1970; Bartel and Lichtenberg 1987). Thus, 'education must be viewed not only as an investment but also as a factor of production ... If

10 Jane Gaskell and Kjell Rubenson

technology becomes stagnant, this incentive is reduced and may disappear' (Welch 1970,41). Drawing on this line of research, OECD noted, 'the key appears to be the ability of people to cope with changes and to turn them to advantages in the future' (Marginson 1993,48). OECD no longer called for a general expansion of public education but wanted major reforms to the nature of public education and its responsiveness to labour market needs (1987,1989). OECD noted 'the need for a more adequate introduction to jobs, careers, and the world of work in schools and familiarization with and command of information technologies' (OECD 1989, 30). To achieve this, OECD recommended the development of an active exchange between schools, local enterprises, and employers. Improved school-to-work transition was also seen as a way to combat high youth unemployment rates. The establishment of content and performance standards was promoted as another strategy to combat school failure development (OECD 1995). The second generation of human capital theory not only shifted the focus to the quality and vocational relevance of initial education, it also expanded the understanding of education to encompass much more than schooling. The concern for a rapid and continual transformation of working life meant that adult education and training was central (OECD 1989, 1995, 1996). Classic distinctions between firm-specific and general training were blurred as firms were urged to recognize the importance of continual ongoing education for workers. Debates about the governance, financing, and credentialing of training became part of the literature in industrial relations (Streeck 1989) reflecting education's increased prominence as part of the bargaining process between workers and management. This new analysis of human capital also added an analysis of welfare and its relationship to work (OECD 1994,1995). Low-wage jobs could provide employment for a large group of unemployed workers (Cruikshank 2002). Adult education and training was part of an active labour market policy aimed at getting the growing number of unemployed people off welfare and into the labour market. Skills training for this sector should focus on getting the clients 'job ready,' and welfare policy should promote short welfare-to-work programs. This training would be quite distinct from the training for highly skilled jobs, since skilled jobs and unskilled jobs exist in quite separate labour markets. OECD's new arguments were reflected in Canadian policy documents from the early 1990s, as the threats and possibilities of global competition and new technologies focused government attention. In

Introduction 11 1991, the influential Harvard economist Michael Porter wrote in a report to the federal government that 'upgrading human resources will be critical to Canadian firms' ability to become more competitive' (Porter, and Monitor Company 1991). In 1992, the Economic Council of Canada (ECC) released its report, A lot to learn, stressing that 'education is a good investment; after taking account of the direct costs of schooling and of foregone earnings, the additional income from completing secondary school yields a rate of return of some 30 percent for individuals ... the contribution of skill development to economic performance has special significance today when, quite simply, the Canadian economy is under threat' (1). In the mid-1990s, Canada's Prosperity Initiative proclaimed, 'As the pace of global competition quickens and technological complexity intensifies, the fortunes of individuals and of nations turn increasingly on the skills they already possess or are prepared to acquire' (Government of Canada 1994,39). All this stress on skills development did not amount to an argument for increasing existing provision, however. It became a critique of current forms of schooling, a call for reforms that would link education more closely to work, and an argument for lifelong learning. In the early 1990s, the cover of Maclean's magazine proclaimed that Canadian schools were failing. The Conference Board of Canada questioned the appropriateness of traditional liberal arts ideals, promoting a broad version of 'employability skills' as the appropriate goals for education (Conference Board of Canada 1995). The Economic Council argued, 'we have attempted to assess the ability of Canada's education and training systems to meet the challenges that lie ahead. We have found those systems wanting ... for sound investment, we must be more demanding' (58). In response, the federal government developed a new training strategy. The 1985 Canadian Job Strategy (CJS) and its 1989 replacement, the Labour Force Development Strategy (LFDS), were attempts to achieve market sensitivity and private sector involvement in adult education and training. In order to make educational institutions more responsive to the demands of the economy the government introduced more accountability measures, reflecting the rise of what has been termed 'the Evaluative State' (Neave 1998). Welfare policy increasingly stressed training to get people into the workforce. These directions were reflected in provincial government policies across the country. In Ontario, the Premier's Council stated, 'Lifelong learning, therefore, is the key link between our educational and eco-

12 Jane Gaskell and Kjell Rubenson

nomic strategies as the 21st century approaches' (Premier's Council on Economic Renewal 1994, 2). In Quebec, the government noted the importance 'of adapting education to the real requirements of the labour market, of making it more relevant, and of easing the transition into the workplace (Gouvernement du Quebec 1996, 54). Alberta and Saskatchewan emphasized increased accountability, and introduced programs to link schooling more closely to work. British Columbia became a visible example of all the new theories in action. The province's Social Credit government set up a Royal Commission on Education in 1988 (Sullivan 1988) that pointed towards educational reform, including a requirement for work experience in secondary education. Then the newly elected NDP government organized a premier's summit on skills development and training which noted the need for a major restructuring of the learning system to serve the educational and training needs of all British Columbians in the context of lifelong learning: 'Our learning/training system must do three things. It must provide a supply of highly skilled, continuously learning workers who are able to work effectively in the new economy of the 21st Century. It must train and equip for employment workers who have not acquired basic skills or whose skills have become obsolete. And it must provide all British Columbians with equitable and continuing access to relevant learning and training opportunities' (Government of British Columbia 1993,35). A Labour Force Development Board was set up to provide the minister with strategic and tactical advice on labour force development, training, and adjustment. Its first report, Training for What?, offered an analysis of the future of work and how British Columbia could best organize its learning system to meet its skill needs: 'First, the mix of education and training programs does not match labour market needs as well as it might. For example, while there is no gap in the capacity of the province's universities and university colleges to produce the number of university graduates that will be required, there appears to be a relative over-supply of graduates in academic programs and an undersupply of those in applied' (BCLFDB 1995,43). Not surprisingly, the report caused much controversy, and out of it grew a Skills Now! policy that affected all levels of the educational system and is the subject of much debate in this book. The policy encapsulated much of the thinking of second-wave human capital theory, at least in its naive version. It has provided a focus for much of our argument and research.

Introduction

13

This book, organized into three parts, presents some findings from the research program that evolved as academics came into contact with this policy environment. Seven of the nine chapters are written by authors based in British Columbia, and are framed by provincial initiatives, through Skills Now!, to improve educational outcomes by linking education more closely to the workplace. The other two chapters are written by authors in Manitoba, and are framed by provincial debates about equity in educational opportunities and outcomes, particularly in relation to Aboriginal students. The arguments found in all chapters reflect discussions held, over five years, among academics and policy-makers in the western provinces, as we considered economic and educational data from analysts at Statistics Canada and Human Resources Development Canada. The chapters are - perhaps inevitably and predictably - organized by disciplinary approach, showing how economists, policy analysts, and educators approached the questions at hand. Part 1, What Skills Matter in the Economy? Economic Approaches, consists of two chapters written by economists who speak to current concerns about what kind of education and skills, or 'human capital/ matter in the 'new economy/ Part 2, Achieving Equity: Three Analyses of Outcomes, explores differing approaches to understanding how equity is affected by changes in educational policy and structures. Part 3, Policy and Practice: Case Studies Linking Education and Work, looks at specific, contextualized studies of educational policy and programs trying to connect instruction in universities and schools with the world outside the classroom. The final chapter, the book's conclusion, reflects on the process of defining the research and the challenges of carrying it out in a network of policy-makers, economists, educators, and policy analysts. It includes reflection on what can be learned from our experience and how social science research can be made more useful to policymakers and practitioners as well as to the scholarly community. References Bartel, A., and F. Lichtenberg. 1987. The comparative advantage of educated workers in implementing new technology. Review of Economics and Statistics 69(1): 1-11. BCLFDB. 1995. Training for what? Victoria: BC Labour Force Development Board.

14 Jane Gaskell and Kjell Rubenson Berg, Ivor. 1970. Education and jobs: The great training robbery. New York: Praeger. Bertram, G.W. 1966. The contribution of education to economic growth. Ottawa: Economic Council of Canada. Blackmore, J. 1992. The gendering of skill and vocationalism in twentieth-century Australian education. Education Policy 7: 351-358. Blau and Duncan. 1967. The American occupational structure. New York: John Wiley and Sons. Bluestone, B., and B. Harrison. 1989. The great U-turn: Corporate restructuring and the polarizing of America. New York: Basic Books. Bowles, S., and H. Gintis. 1976. Schooling in capitalist America. New York: Basic Books. Braverman, H. 1974. Labor and monopoly capitalism: The degradation of work in the twentieth century. New York: Monthly Review Press. Brown, P. 1995. Cultural capital and social exclusion: some observations on recent trends in education, employment and the labour market. Work, Employment and Society 9: 29-51. Gallon, M. 1986. Some elements of sociology in translation: Domestication of the scallops and the fisherman. In Power, action and belief: A new sociology of knowledge? ed. J. Law, 128-43. London: Routledge and Kegan Paul. - 1987. Society in the making: The study of technology as a tool for sociological analysis. In The social construction of technological systems, ed. W. Cambridge, B.T. Hughes, and T. Pinch, 87-106. Cambridge, Mass.: MIT Press. Collins R. 1979. The credential society. New York: Academic Press. Conference Board of Canada. 1995. Employers skills profile: What are employers looking for? Ottawa: Conference Board of Canada. Cruikshank, J. 2002. Lifelong learning or re-retraining for life: Scapegoating the worker. Proceedings of the 21st annual conference of the Canadian Association for the Study of Adult Education, 30 May-1 June 2002,54-9. Denison, E.F. 1962. The source of economic growth and alternatives before us. New York: Committee for Economic Development. Economic Council of Canada. 1965. Annual review. Ottawa: Queen's Printer. - 1992. A lot to learn. Ottawa: Minister of Supply and Services. Fleming, W.G. 1974. Educational opportunity and the pursuit of equality. Scarborough: Prentice-Hall. Fudge, J., and P. McDermott. 1991. Just wages: A feminist assessment of pay equity. Toronto: University of Toronto Press. Gaskell, J. 1992. Gender matters from school to work. Philadelphia: Open University Press. Gouvernement du Quebec. 1996. The state of education in Quebec, 1995-1996. Quebec: Gouvernement du Quebec.

Introduction 15 Government of British Columbia. 1993. Premier's Summit: Skills development and training. Victoria. Government of Canada. 1991. Learning well, living well. Ottawa: Prosperity Initiative. - 1994. Agenda: Jobs and growth. Improving social security in Canada. Ottawa: Government of Canada. Holzer, H. 1996. What employers want. New York: Russell Sage Foundation. Husen, T. 1979. Schools in question: A comparative study of school and its future in Western societies. Oxford: Oxford University Press. Jencks, C. 1972. Inequality: A reassessment of the effect of family and schooling in America. New York: Basic Books. Karabel, J., and A.H. Halsey. 1977. Power and ideology in education. Oxford: Oxford University Press. Latour, B. 1988. The pasteurization of France. Cambridge: Cambridge University Press. Lave, J., and E. Wenger. 1991. Situated learning: Legitimate peripheral participation. Cambridge: Cambridge University Press. Lingard, H., F. Ritzvi, and S. Taylor. 2001. Globalization and educational policy. Oxford: Pergamon. Marginson, S. 1993. Education and public policy in Australia. Melbourne: Cambridge University Press. - 1997. Markets in education. St Leonards, NSW: Allen and Unwin. Neave, G. 1998. The evaluative state reconsidered. European Journal of Education 33(3): 265-84. Nespor, J. 1994. Knowledge in motion: Space, time and curriculum in undergraduate physics and management. London: Palmer Press. OECD (Organization for Economic Cooperation and Development). 1961. Economic growth and investment in education. Paris: OECD. - 1964. Economic aspects of higher education. Paris: OECD. - 1965a. Economic models of education. Paris: OECD. - 1965b. The residual factor and economic growth. Paris: OECD. - 1966. Financing of education for economic growth. Paris: OECD. - 1987. Structural adjustment and economic performance. Paris: OECD. - 1989. Education and the economy in a changing society. Paris: OECD. - 1994. The OECD jobs study: Facts, analysis, strategies. Paris: OECD. - 1995. The OECD jobs study: Implementing the strategy. Paris: OECD. - 1996. Lifelong learning for all. Paris: OECD. - 2001. Education at a glance. Paris: OECD. Osberg, L., F. Wien, and J. Crude. 1995. Vanishing jobs. Toronto: James Lorimer. Piore, M.J., and C.F. Sabel. 1984. The second industrial divide: Possibilities for prosperity. New York: Basic Books.

16 Jane Gaskell and Kjell Rubenson Porter, J. 1965. The vertical mosaic. Toronto: University of Toronto Press. Porter M., and Monitor Company. 1991. Canada at the crossroads: The reality of a new competitive environment. Ottawa: Business Council on National Issues and the Government of Canada. Premiers Council on Economic Renewal. 1994. Lifelong learning and the new economy. Toronto: Queen's Printer for Ontario. Resnick, L. B. 1987a. Education and learning to think. Washington, DC: National Academy Press. - 1987b. Learning in schools and out. Educational Researcher 16:16-20. Rubenson, K. 1992. Human resource development: A historical perspective. In Developing resourceful humans: Adult education within the economic context, ed. L.E. Burton, 3-31. New York: Routledge. Schultz, T. 1960. Capital formation by education. Journal of Political Economy 64(4): 571-83. - 1961. Investing in human capital. American Economic Review 51:1-17. Social Sciences and Humanities Research Council. 1995. Call for proposals for strategic research networks in education and training. Ottawa. Spenner, K.I. 1983. Deciphering Prometheus: Temporal change in the skill level of work. American Sociological Review 48 (December): 824-37. Statistics Canada. 2001. Measuring up: The performance of Canada's youth in reading, mathematics and science - OECD PISA Study: First results for Canadians aged 15. Ottawa. - 2002. At a crossroads: First results for the 18 to 20 year old cohort of the YITS. Ottawa. Stewart, T. 1994. A company's most valuable asset: Intellectual capital. Fortune (3 October): 68-72. Streeck, W. 1989. Skills and the limits of neo-liberalism: The enterprise of the future as a place of learning. Discussion paper. Berlin: Science Centre-Research Unit on Labour Market and Employment. Sullivan, B. 1988. Royal Commission on Education: A legacy for learners. Victoria, BC: Queen's Printer. Taylor, A. 2001. The politics of educational reform in Alberta. Toronto: University of Toronto Press. Welch, F. 1970. Education in production. Journal of Political Economy 78(1): 3559. Wilson, J.D. 1972. Vocationalism in education: Some comments from Ontario. In The best of times, the worst of times: Contemporary issues in Canadian education, ed. H. Stevenson, R. Stamp, and J.D. Wilson, 274-80. Holt Rinehart and Winston. Wong, L., and S. McBride. 2003. Youth employment programs in British

Introduction 17 Columbia: Taking the high road or the low road? In Training the excluded for work: Access and equity for women, immigrants, First Nations, youth and people with low income, ed. M. Cohen, 230-44. Vancouver: UBC Press. Wozniak, G. 1984. The adoption of interrelated innovations: A human capital approach. Review of Economics and Statistics 66(1): 70-9.

This page intentionally left blank

PART 1 What Skills Matter in the Economy? Economic Approaches

The two chapters in this section speak to questions asked by laypeople, educators, and politicians alike. What kind of education and skills are needed in the new economy? How can we produce them in Canada? Both chapters are written by economists, and both mount arguments that depend on the analysis of large data sets. They make factual claims about what has been happening to education and the economy over time, and they mount arguments about what this means for our social investment in education. Craig Riddell summarizes the research on the relationship between education and the labour market in Canada. He reviews a series of studies that explore how much Canada has invested in education, what skills the population has, and what the causal relationship is between education, skills, and labour market success. He concludes with a series of statements about the current situation in Canada, and the view that our investments in education have paid off, and may in fact be underestimated. Bob Allen suggests that technology has dramatically changed the workplace, but not in the ways one might anticipate. The computer revolution has resulted in an increased demand for general intellectual abilities of the sort that are developed in humanities and social science programs. The demand for technical skills related to computers has been dramatically overestimated. He believes that students should continue to pursue a broad liberal education, and that policy-makers should fund it. Taken together, these chapters display the power of economic analysis, as well as its methods and its voice. The analysis addresses policyrelevant questions, and it is often the starting place for policy and the

20 Part 1: What Skills Matter?

generator for further questions. This analysis deals with education as a set of numbers, generated from surveys, tests, and labour market indicators. Debates about the meaning of these numbers, the samples and questions from which they are generated, and the appropriate ways of analysing them, form a fertile academic field, for the summaries from this kind of work powerfully shape the views of citizens, policy-makers, and educators.

1 Education, Skills, and Labour Market Outcomes: Exploring the Linkages in Canada W. CRAIG RIDDELL

The purpose of this chapter is to assess recent evidence concerning the relationship between the resources devoted to education and skill formation in Canada, and the labour market consequences of those expenditures. The widespread tendency to associate advances in technology and other sources of economic change with the need for greater emphasis on student achievement presumes that there are clear linkages between what students learn in school and their subsequent economic success. This chapter will examine recent research on the nature of these linkages, and will assess the implications of this research for education policy. The first section compares Canadian educational expenditures and educational outcomes with those of other countries. Several educational outcomes are discussed: educational attainment, student achievement, and the literacy skills of the adult population. This comparative examination of educational 'inputs' and 'outcomes' provides a basis for assessing whether Canada obtains good value from its public and private investments in education. I then turn in the second section, to the labour market consequences of education. The question of how best to interpret the strong positive correlation between education and economic success has long been a subject of debate and controversy. Substantial recent progress has been made on this issue. This literature is reviewed here, and the findings of recent Canadian studies are discussed. The third section discusses a new area of research opened up by advances in data collection: the linkages between education and literacy skills, and the labour market consequences of such skills. The final section summarizes the main conclusions and discusses their implications for public policy.

22 W. Craig Riddell

Education Expenditures and Outcomes Education systems vary substantially from country to country. For example, there are important differences across countries in the provision of publicly funded early childhood and pre-elementary schooling, in the extent to which students are streamed into 'academic' and 'vocational' programs, in the ways in which school and work experience can be combined, and in the extent to which the system provides a 'second chance' for those who drop out at some stage. These and other differences make international comparisons of educational inputs and outcomes difficult. Although considerable progress has been made in improving the comparability of educational data across OECD countries, one should keep these institutional differences among education systems in mind when interpreting comparative statistics.1 Investment in Education

Relative to other developed countries, Canada invests a substantial amount on education. Most of this expenditure is publicly financed. Table 1.1 shows a number of measures of educational expenditure in Canada and other G7 countries, as well as the OECD country average. The top panel reports educational expenditure per student (in U.S. dollars), an input-based indicator of the quality of education. At both the elementary/secondary and post-secondary levels, Canadian expenditure per student is second highest (after the U.S.) among the G7 countries and substantially above the OECD average. The gaps between the United States and Canada and other OECD countries are especially large at the post-secondary level. Although not shown in table 1.1, Canadian per-student expenditure also ranks among the highest in the OECD at both the elementary/secondary and post-secondary levels (OECD 2001). The middle panel of Table 1.1 reports expenditure on education as a percentage of GDP. This measure reflects both expenditure per student and the number of students. It indicates the fraction of total output devoted to the consumption of and investment in education. Even among the G7 countries, large differences are evident in the relative share of national resources devoted to formal education. These differences are much more substantial at post-secondary than at elementary and secondary levels. Canada's educational expenditure of 7.0 per cent of gross domestic product (GDP) in 1995 is highest in the G7 countries

TABLE 1.1 Educational expenditures in Canada and G7 Countries, 1995 (a) Expenditure per student from public and private sources by level of education, in U.S. dollars converted using purchasing power parity (PPP) exchange rates*

Elementary & secondary Post-secondary All levels of education

Canada

France

Germany

Italy

Japan

UK

U.S.

OECD Average**

5,401 11,471 6,396

5,041 6,569 5,001

4,690 8,897 6,057

5,099 5,013 5,157

4,282 8,768 4,991

3,810 7,225 4,222

6,281 16,262 7,905

4,162 8,134 4,717

(b) Educational expenditure from public and private sources for educational institutions as a percentage of GDP, by level of education, Canada and G7 countries, 1995

Elementary & secondary Post-secondary All levels of education

Canada

France

Germany

Italy

Japan

UK

U.S.

OECD Average1

4.3 2.5 7.0

4.4 1.1 6.3

3.8 1.1 5.8

3.2 0.8 4.7

3.1 1.0 4.7

1.0 -

3.9 2.4 6.7

3.7 1.3 5.6

(c) Expenditure per student by province in Canadian dollars, 1995

Elementary & secondary Post-secondary All levels of education

Nfld

PEI

NS

NB

Que

Ont

Man

Sask

Alta

BC

Canada

5,516 12,724 7,079

4,761 13,770 6,327

5,394 13,463 6,975

5,120 12,532 6,623

7,063 12,687 8,406

7,617 13,122 7,963

6,660 15,615 7,689

5,507 15,051 7,032

5,520 12,994 6,931

6,399 14,738 8,162

6,677 14,182 7,907

Source: Council of Ministers of Education, Canada and Statistics Canada (2000). * The actual exchange rate may understate or overstate differences in purchasing power. PPP exchange rates equalize the purchasing power of different currencies. They are based on the cost of a fixed basket of goods and services in different countries. ** Unweighted country average, t Includes pre-primary (pre-elementary) and undistributed expenditures.

24 W. Craig Riddell

and (although not shown) among the highest in the OECD.2 In Canada and the United States, the share of GDP devoted to formal post-secondary education is more than double that of all other G7 countries, and substantially higher than the OECD country average. Canada's relatively high percentage of GDP spent on education reflects both the substantial per-student expenditures on education at all levels, as illustrated in Table 1.1 (a), and Canadians' comparatively high participation rates in education, especially at the post-secondary level, which are described in more detail below. The bottom panel of Table 1.1 shows expenditure per student in Canadian provinces. There are moderately large differences in expenditure across jurisdictions, especially at the elementary and secondary level. For example, at this level Ontario - the province with the highest per student expenditure - spends about 50 per cent more than lowspending provinces such as Nova Scotia, New Brunswick, and Prince Edward Island. Canada invests heavily in educating its population. What are the consequences of these substantial expenditures on formal education? The next three subsections summarize the available evidence on this question using several measures of educational outcomes. First to be examined is educational attainment, the dimension we know the most about. The discussion then turns to student achievement and the literacy skills of the adult population - two measures of the skills and knowledge imparted by education as well as by other activities. Educational Attainment

Several measures of the educational attainment of the adult population in Canada and other G7 countries are reported in Table 1.2. The top panel shows the highest level of educational attainment for the population twenty-five to sixty-four years of age. Also shown is the unweighted OECD country average. The middle panel reports average years of schooling. By these measures, Canadian educational attainment is high by international standards, reflecting the substantial expenditure on formal education. Eighty per cent of Canada's adult population has completed upper secondary (referred to as high school in North America) or post-secondary education, much higher than the OECD average of 64 per cent. Canada's proportion is similar to that of Germany, Japan, and the United Kingdom, but substantially below the United States

TABLE 1.2 Educational attainment in Canada and G7 Countries (a) Percentage of the population aged 25-64 years by highest level of educational attainment, 1999

Less than upper secondary Upper secondary graduate Non-university post-secondary University graduate

Canada

France

Germany

Italy

Japan

UK

U.S.

OECD Average

20 28 33 19

38 41 10 11

19 53 15 13

57 30 4 9

19 49 13 18

18 57 8 17

13 51 8 27

36 40 11 14

Source: OECD(2001). (b) Average completed years of schooling of the population aged 25-64 years, 1995

Canada

France

Germany

Italy

Japan

UK

U.S.

OECD Average

13.2

11.2

13.4

10.0

-

12.1

13.5

11.9

Source: OECD(1998). (c) Ratio of upper secondary graduates to population at a typical age of graduation, 1996

Both sexes Males Females

Canada

France

Germany

Italy

Japan

UK

U.S.

75 70 81

85 85 86

86 86 86

79 76 82

99 96 102

-

72 69 76

Source: Council of Ministers of Education, Canada and Statistics Canada (2000).

26 W. Craig Riddell

where 87 per cent of the adult population have at least a high school diploma.3 Average completed years of schooling are also among the highest in the OECD, albeit somewhat below Germany and the United States.4 Canada stands out in terms of the fraction of the adult population with completed post-secondary education. Canada's proportion (52%) is not only more than double the OECD average of 25 per cent, but is also the highest in the OECD countries and substantially higher than the United States, the country ranked second (where 35 per cent have completed post-secondary education). Canada's extremely high ranking on this dimension arises principally because of the very substantial fraction of the population with non-university post-secondary education - at 33 per cent, triple the OECD average and more than double any other G7 country. At the university level, Canada is above the OECD average (19% vs. the OECD average of 14%) and similar to Japan and the U.K., but substantially below the United States, where 27 per cent have graduated from university. Canada's ranking at the top of the OECD, in terms of the fraction of the population with completed post-secondary education, has led several analysts to comment that Canada's population is among the most highly educated in the world - even surpassing the United States, the country traditionally regarded as having the most highly educated population. However, it is important to keep in mind that Canadian educational attainment ranks below the United States in two key dimensions: the fraction of the population with completed secondary education and the proportion with a university degree. Thus at the two extremes of the educational attainment distribution - roughly the bottom 20 per cent and top 20 per cent - Canada ranks significantly below the United States. It is in the middle of the distribution where Canadian educational attainment dominates according to these standard measures. In both countries, approximately 60 per cent of the adult population have completed high school or a non-university post-secondary program. However, the composition of this middle group differs substantially between the two countries: in Canada more than half (33 per cent out of 61 per cent) have completed non-university postsecondary education, whereas less than one-sixth (8% out of 65%) of Americans are in this category. Because of its evident importance in Canada, a closer look at the nonuniversity post-secondary category is warranted. There are two main types of individuals in this group: those with a community college or

Education, Skills, and Labour Market Outcomes 27

CEGEP (College d'enseignement general et professionel) diploma and those with a certificate from a trade school or apprenticeship program. Because high school completion is not necessarily a prerequisite to enter these programs, not all those classified as non-university post-secondary graduates are high school graduates. Furthermore, although many community college programs are two years in length, trade school and some community college programs may be of much shorter duration. For these reasons, one might conjecture that the human capital of some of those in the 'non-university post-secondary' group may not be substantially different from that of the average high school graduate. The monthly Labour Force Survey, the source of the Canadian data in Table 1.2, classifies those who report that they completed a community college, CEGEP, apprenticeship, or trade school program as having a non-university post-secondary certificate or diploma, whether or not they have graduated from high school.5 To obtain some insight into this potentially diverse non-university post-secondary category, Table 1.3 reports data from the 1996 census. The advantage of the census is that the questionnaire asks about all diplomas, certificates, and degrees obtained as well as years of completed schooling. In order to obtain information on individuals' wages - an additional measure of human capital - Table 1.3 reports data for those employed in 1995 rather than the adult population.6 The data on educational attainment of employed Canadians provides a very similar picture to that of the adult population shown in Table 1.2. The non-university post-secondary category constitutes 34 per cent of all workers in 1996 versus 33 per cent of the adult population in 1999. This group is a slightly larger proportion of the female workforce (35%) than the male workforce (33%). On average, those with a non-university certificate or diploma have 1.3 additional years of schooling compared to high school graduates. This differential is also a bit larger for females (1.4 years) than males (1.1 years). However, within the non-university post-secondary category there is a substantial gap of 1.6 years of completed schooling between those with a high school diploma (14.4 years) and those without this credential (12.8 years). Indeed, high school graduates have very similar years of schooling to those with a college diploma or trade certificate but who did not graduate from high school (12.8 vs. 12.6 years). For males, the average high school graduate actually has slightly more years of schooling than his counterpart who completed a trade school or community college program without also completing high school.

TABLE 1.3 Average weekly wages and years of schooling by highest level of educational attainment, Canada, 1996 All workers

No degree High school graduate College/trade without high school College/trade school and high school All college/trade school University degree All levels of education

Females

Males

%

Average wage

Years of schooling

%

Average Years of wage schooling

%

Average Years of wage schooling

23.4 26.2 9.9 23.8 33.7 16.7 100.0

582.4 584.4 712.5 704.1 706.6 941.7 684.9

10.1 12.6 12.8 14.4 13.9 17.5 13.3

19.9 28.0 8.8 26.0 34.7 17.4 100.0

429.3 473.3 537.6 570.4 562.1 786.4 550.3

26.5 24.7 10.9 21.9 32.8 16.0 100.0

681.7 693.5 834.6 841.7 839.3 1,088.3 801.6

10.3 12.6 13.1 14.3 14.0 17.4 13.5

Source: Author's calculations from the 1996 census public use master file (Ottawa: Statistics Canada).

9.9 12.7 12.6 14.4 13.8 17.6 13.1

Education, Skills, and Labour Market Outcomes 29

This evidence based on years of schooling suggests that the human capital of college/trade school graduates who did not complete high school may be very similar to those whose highest educational attainment is a high school diploma, and substantially lower than those with both a high school diploma and a community college diploma or trade school certificate. However, this conclusion does not continue to hold when we use wages rather than years of education as an indicator of human capital. Indeed, for both males and females the average wages of the 'college/trade without high school' group are much closer to their college/trade school counterparts who also completed secondary school than they are to those whose highest educational attainment is a high school diploma. This suggests that it is not unreasonable to group together all those with a college diploma or trade certificate, whether or not they are also high school graduates, and despite the substantial differences in their years of schooling. Relative to other countries, the extent of Canadian non-university post-secondary education may be somewhat overstated because of the Quebec CEGEP system. These institutions provide both 'general' and 'professional' programs. The former constitute a stage between high school and university, providing the equivalent of the final year of high school and the first year of university in most other Canadian provinces. The professional programs provide the equivalent of the final year of secondary school and a two-year community college program in English Canada. Graduates of the 'professional' CEGEP programs are similar to graduates of professional/vocational community college programs in English Canada, and are appropriately classified as 'non-university post-secondary.' Students who pursue the CEGEP 'general stream' and who obtain a university degree will also be appropriately classified in the data as university graduates. However, those who pursue the general stream but who subsequently do not enter or complete university will be measured in Quebec as 'non-university post-secondary graduates/ but would be designated as high school graduates (albeit with some, but incomplete, post-secondary education) in English Canada. Some adjustment to the Canadian data to account for this difference may be appropriate. In summary, according to commonly used measures, Canadian educational attainment is very high by international standards, a finding that is consistent with the country's substantial investment in education. The distribution of the educational attainment of Canadians also

30 W. Craig Riddell

has some unique features. At the bottom and top of the educational attainment distribution - specifically, those with less than completed high school and those with a university degree - Canadian educational attainment is similar to that of several other OECD countries and significantly lower than that of the United States. However, in the middle of the distribution - those who have completed secondary school but not university - the proportion of Canadians with a community college diploma or trade school certificate is unusually high and the proportion of high school graduates relatively low. However, this 'non-university post-secondary' group is heterogeneous. Canada's provincial education systems have 'forgiving' features and provide various routes to a community college diploma or trade school certificate. More than onequarter of the 'non-university post-secondary' group have not graduated from secondary school, and their average years of completed schooling is not much different from those whose highest educational attainment is high school completion. This raises some questions about whether these individuals should be placed in a higher educational attainment category than secondary school graduates. However, these doubts are dispelled to a considerable extent by a comparison of the average earnings of this 'college/trade without high school' group to high school graduates and the 'college/trade with high school' group. In particular, this 'market test' suggests that the human capital of the 'college/trade without high school* is much closer to that of their 'college/trade with high school' counterparts than to secondary school graduates. Accordingly, Canada's high measured educational attainment in the middle of the distribution appears to be real, and not simply due to inappropriate labelling of some of those in the nonuniversity post-secondary category.7 Although the overall educational attainment of Canadians is impressive, high school completion has been a weak spot for many years. For example, based on administrative data on the number of graduates relative to the number of eighteen-year-olds, the recent Canadian secondary school graduation rate is near the bottom of the G7 countries and only marginally above that of the United States, the bottom dweller on this dimension (see Table 1.2(c)). As of the mid-1990s, approximately 25 per cent of 18-year-olds had not graduated from high school. This non-completion rate is much higher among males (30%) than females (20%). Some of these individuals graduate after the 'normal age' of 18; according to Labour Force Survey (LFS) data the high school graduation rate is 81 per cent by age 19-20 and 87 per cent by age 25-29

Education, Skills, and Labour Market Outcomes 31

(Council of Ministers of Education, Canada and Statistics Canada 2000, 91). In addition, as discussed above, a significant number of high school dropouts obtain a college diploma or trade certificate. Nonetheless, Canada's relatively high secondary school non-completion rate is a potential concern. Measures of educational attainment such as years of completed schooling or highest credential received are frequently used to compare the amount of human capital of the population or workforce over time and across regions and countries. Nonetheless, these are indirect measures of human capital, reflecting principally the inputs of time and other resources into the production of skills, knowledge, and competencies. We now turn to measures of the outcomes of human capital formation. Student Achievement

We know a good deal more about student achievement than we did even a decade ago. Canada did not participate in the early rounds of international mathematics and science tests carried out in the 1960s and 1970s. However, some provinces took part in the Second International Mathematics and Science Studies carried out in the 1980s, and all Canadian jurisdictions except Prince Edward Island participated in the third round - the Third International Mathematics and Science Study (TIMSS) - carried out in the 1990s.8 In addition, there was Canadian involvement in some other international studies of student achievement in the 1980s and 1990s, and the decade of the 1990s saw the introduction of the Canadian School Achievement Indicators Program (SAIP), which has now completed several rounds of testing.9 Table 1.4 summarizes some of the key results from the TIMSS tests carried out in the 1990s.10 These data have the advantage of providing information on student performance on a common set of tests administered in numerous countries. For the present purposes, one disadvantage is that the set of countries is very diverse and includes several countries that we do not normally compare ourselves to - both 'high achievers' such as Korea, Singapore, and Hong Kong and 'low achievers' such as Iran, Kuwait, and Portugal. The set of countries participating in each test also varies, so the international average needs to be interpreted cautiously. In order to provide information on student achievement that is comparable to our previous analyses of educational expenditure and

TABLE 1.4 (a) Student achievement in mathematics and science in Canada and G-7 countries, 1995 and 1999 Mean value of per cent correct

Grade 4 math 1995 Grade 8 math 1995 Grade 8 math 1999 Grade 4 science 1995 Grade 8 science 1995 Grade 8 science 1999

Canada

France

Japan

U.S.

England

Germany

Italy

International mean

60 59* 53* 64* 59* 53*

61 54-

74+ 73+ 58+ 70+ 65+ 55+

63 535066+ 58 52-

57535063 61+ 54

5458 -

4849-

59 55 49 59 56 49

+ denotes statistically significantly above the Canadian mean - denotes statistically significantly below the Canadian mean * denotes Canadian average is significantly different than the international average (b) Student achievement in mathematics and science in Canadian jurisdictions, 1995 and 1999 Mean value of per cent correct

Grade 4 math 1995 Grade 8 math 1995 Grade 8 math 1999 Grade 4 science 1995 Grade 8 science 1995 Grade 8 science 1999

Nfld

New Brunswick (English)

Quebec

Ontario

Alberta

BC

Canadian mean

58 56 5062 59 51-

58 5461 57 -

69+ 68+ 57+ 65 59 54

57 5452 62 5652

65 61 53 68+ 65+ 56+

59 63+ 52 64 62 54

60 59 53 64 59 53

Source: Robitaille, Taylor, and Orpwood (1996, 1997); Council of Ministers of Education, Canada and Statistics Canada (2000); and Robitaille and Taylor (2000). + denotes statistically significantly above the Canadian mean - denotes statistically significantly below the Canadian mean

Education, Skills, and Labour Market Outcomes 33

attainment, the top panel of Table 1.4 reports mean scores for the G7 countries that participated in the various TIMSS tests as well as the international average score.11 Canadian achievement is not significantly different than the international average in grade 4 math, but is statistically significantly above the international average in grade 8 math and grade 4 and grade 8 science. Among the G7, Canada tends to rank in the middle of the pack in both math and science. Canadian students' scores are consistently below their Japanese counterparts but superior to those of Italy and equal to or better than those of France and Germany. The ranking between Canada, the United States, and England varies from test to test. Overall, these results indicate that Canadian student achievement is satisfactory but not as good as one might expect given Canada's relatively high expenditure on elementary and secondary schooling. Several Canadian jurisdictions over-sampled their student populations in order to provide meaningful results at the provincial level, and these are reported in the bottom panel of Table 1.4.12 Substantial provincial variation is evident. In mathematics, Quebec student achievement is substantially above the Canadian average and high by international standards, albeit still significantly below the top-ranked countries (Korea, Singapore, Hong Kong, and Japan). Ontario's student performance tends to be below the national average, although the differences are statistically significant only in grade 8 mathematics and science in 1995. In addition, the gap between Ontario and Canada as a whole appears to have narrowed in the last half of the 1990s, perhaps in response to the increased emphasis placed on student achievement in Ontario during that period. Test scores in New Brunswick (English schools) also consistently fall below the Canadian mean, although significantly so only in the case of Grade 8 mathematics. Newfoundland fell from average to below the Canadian average in both mathematics and science during the 1995-99 period. Finally, Alberta student performance is impressive in science; at both grade 4 and grade 8 Alberta's students rank among the best in the world, exceeded only by Korea and Japan. In summary, according to these recent international tests, Canadian student achievement in mathematics is average or somewhat above average among a diverse set of countries. Within the G7, Canada generally ranks in the middle of the participating countries. In science, Canadian student performance is above average among the full set of countries that took the tests but about average among the G7 partici-

34 W. Craig Riddell

pants - above France and Italy but below Japan and similar to England, Germany, and the United States. In contrast to the TIMSS results, the performance of Canadian secondary school students in the recent Program of International Student Assessment (PISA) tests was excellent. These tests assessed the skills of fifteen-year-old students in reading, mathematics, and science in over thirty (mainly OECD) countries. Whereas TIMSS is curriculum-based and assesses students' knowledge of the subject matter, PISA assesses the ability of fifteen-year-olds to apply their knowledge to real world issues and challenges. Table 1.5(a) summarizes the results for the G7 countries, and Table 1.5(b) reports results at the provincial level. Across the three assessed subjects, Canada, Japan, and the United Kingdom are the top three countries, followed by France and the United States in the middle and Germany and Italy at the bottom. Consistent with the TIMSS findings, Canadian student achievement was below that of Japan in math and science. However, in several respects the PISA results paint a more favourable picture than do earlier international assessments. The gap between Canada and the bottom four G7 countries (France, the U.S., Germany, and Italy) is large in all three of the subject areas (in contrast to the TIMSS results in math and science), and in reading Canadian students were the top performers among the G7 countries. Canadian student achievement was equally impressive among the larger group of OECD countries. Variation in student achievement across provinces is similar to that observed in TIMSS, with Alberta, Quebec, and BC students in the top group and students in the four Atlantic provinces at the bottom. Quebec students continue to achieve the highest scores in math, and Alberta students perform best in science (as well as in reading). Ontario, Manitoba, and Saskatchewan tend to have scores close to the national average. Considerable caution is appropriate in interpreting these summary statistics on student performance in reading, mathematics, and science. Many factors, in addition to the resources devoted to the school system, influence student achievement. For example, relative to other G7 countries, Canada has a high proportion of immigrant children (for whom English or French is often a second language) in its schools. Furthermore, countries may differ in the extent to which they aim to raise average performance or to improve achievement principally among those who would otherwise perform poorly. Nonetheless, these measures of student achievement - especially the PISA results - suggest that Canada appears to obtain reasonably good Value for money' from

Education, Skills, and Labour Market Outcomes

35

TABLE 1.5 Achievement of 15-year-olds in reading, mathematics, and science, 2000 (a) PISA results for Canada and G7 countries Reading Canada UK Japan France U.S. Italy Germany

Mathematics 534(3.1) 523(5.1) 522(10.4) 505(5.4) 504(14.0) 487(5.8) 484(4.9)

Japan Canada UK France U.S. Germany Italy

Science 557(10.9) 533(2.8) 529(5.0) 517(5.4) 493(15.2) 490(5.0) 457(5.8)

Japan UK Canada France U.S. Germany Italy

550(10.9) 532(5.3) 529(3.1) 500(6.3) 499(14.6) 487(4.8) 478(6.1)

(b) PISA results for Canadian provinces Reading Alberta BC Quebec Ontario Manitoba Sask NS PEI Nfld NB

Mathematics 550(6.5) 538 (5.7 536 (6.0) 533(6.5) 529(7.0) 529(5.3) 521(4.5) 517(4.8) 517(5.6) 501 (3.5)

Quebec Alberta BC Manitoba Sask Ontario NS PEI Nfld NB

Science 550(5.4) 547 (6.6) 534 (5.6) 533(7.3) 525(5.8) 524(5.8) 513(5.6) 512(7.4) 509(5.9) 506 (4.4)

Alberta Quebec BC Manitoba Ontario Sask Nfld NS PEI NB

546(6.9) 541 (6.7) 533 (6.4) 527(7.1) 522(6.8) 522(5.9) 516(6.7) 516(6.0) 508(5.4) 497 (4.5)

Source: Human Resources Development Canada, Statistics Canada, and Council of Ministers of Education, Canada (2001). Note: Numbers in parentheses are 9% confidence intervals.

the elementary and secondary school system, at least as measured by average student achievement in reading, mathematics, and science. Canada ranks at or near the top of the G7 countries in terms of expenditure per student on elementary and secondary schooling and according to the PISA findings - places in the top three of the G7 countries in terms of student performance. However the TIMSS results lead to a somewhat less favourable assessment. According to these tests, Canada is at the high end internationally in the resources it devotes to elementary and secondary education, but in the middle or upper middle of the pack in student achievement in math and science.13 For example, the PISA tests rank Canadian student performance signifi-

36 W. Craig Riddell

cantly higher than that of the United States in all three subjects, whereas TIMSS ranks U.S. students higher in mathematics and science at the grade 4 level and Canadian students somewhat above their U.S. counterparts at the grade 8 level. Clearly it is important in future research to try to understand the reasons for the differences between the TIMSS and PISA measures of student achievement.14 Although overall national levels of Canadian student performance in mathematics and science are not outstanding, some provinces - such as Quebec in mathematics and Alberta in science - are able to obtain very high levels of achievement within the existing Canadian social, cultural, and fiscal framework. In other provinces, especially the Atlantic provinces, student achievement generally falls below the Canadian average and is relatively low by international standards. The source of these provincial variations is an important subject for future research. Simple inspection of the data reveals that there is a positive relationship between expenditure per student and student achievement as measured by TIMSS and PISA. The provinces that spend the least on elementary and secondary schooling tend to have relatively low student achievement in reading, mathematics, and science, while provinces such as Quebec and British Columbia exhibit both above-average expenditure per student and relatively high test scores. Nonetheless, the relationship between expenditure and test scores is relatively weak, with Ontario having the greatest expenditure per student but student achievement that is typically average or even below the Canadian average. Similarly, Alberta combines frugality in its expenditure on elementary and secondary education with very high levels of student achievement. One factor that appears to account for some of the variation in student achievement across provinces is the presence in some jurisdictions of curriculum-based examinations at the end of secondary school. Bishop (1995, 1997, 2002) finds evidence of beneficial effects of such examinations not only on student achievement but also on school administrator behaviour, teacher behaviour, and home behaviour and attitudes. For example, his estimates indicate that the presence of curriculum-based examinations in a jurisdiction raises achievement in mathematics by approximately four-fifths of a U.S. grade-level equivalent, and achievement in science by approximately three-fifths of a grade-level equivalent (Bishop 1997). These are large estimated impacts and they appear to help account for the relatively strong performance of students in provinces such as Alberta, Quebec, and BC that employ

Education, Skills, and Labour Market Outcomes 37

province-wide curriculum-based exit examinations. This factor may also help explain the relatively poor performance (given the high expenditure per student) of Ontario, where province-wide tests were abandoned in 1967. Literacy Skills of the Adult Population

Data on student achievement provide some information on the skills of those who will be entering the labour force in the future, that is, the flow of new entrants. Until recently, however, no nationally representative measures of the skills and knowledge of the existing stock - the adult population - were available. The International Adult Literacy Survey (IALS), which was carried out in more than twenty countries during the 1994-98 period, represents a breakthrough in international data collection, providing for the first time measures of the literacy and numeracy skills of the adult population that are comparable across countries and language groups.15 The survey provided three measures of literacy: prose, document, and quantitative (i.e., numeracy). Details of the tests used to measure these skills are given in Literacy, Economy and Society, by the Organization for Economic Cooperation and Development and Statistics Canada (1995); the main point here is that these measures correspond to information-processing skills needed to perform everyday tasks at home, at work, and in the community. For each respondent, the survey measures prose, document, and quantitative literacy on a scale from 0 to 500. These numerical literacy scores are also grouped into five main levels of competency, with level 1 being the lowest and level 5 the highest. According to Statistics Canada, individuals with only level 1 or level 2 literacy skills have marginal or quite limited capabilities (Crompton 1996). Table 1.6 summarizes some of the key findings from the IALS. In order to maintain comparability with previous sections of this paper, results are reported for Canada and other participating G7 countries (Germany, the UK, and the U.S.).16 The top panel shows the mean score on each of the three literacy scales and the score at the 25th and 75th percentiles of the literacy distribution. The average scores rank Germany at the top (with the exception of the prose scale, on which Canada ranks first and Germany second), followed by Canada, the United States, and the United Kingdom at the bottom. Although the differences in mean scores among these four countries may not appear large,

TABLE 1.6 Literacy skills in Canada and G7 countries, 1994-1998 (a) Mean scores and scores at the 25th and 75th percentiles of the prose, document, and quantitative literacy scales

U.S.

UK

Germany

Canada Literary scale

25th

Mean

75th

25th

Mean

75th

25th

Mean

75th

25th

Mean

75th

Prose Document Quantitative

243 243 247

279 279 281

322 326 323

245 256 265

276 285 293

308 318 324

233 230 231

267 268 268

311 314 314

237 230 237

274 268 275

320 316 322

(b) Per cent of adults with low literacy skills* Age group

Literacy scale

Canada

Germany

UK

U.S.

16-65 16-65 16-65 16-25 46-55 16-65

Prose Document Quantitative Document Document Document

42 43 43 33 54 43

49 42 33 34 42 42

52 50 51 44 53 50

47 50 46 56 50 50

(c) Mean document literacy score and educational attainment Education

Canada

Germany

UK

U.S.

Less than high school High school graduate Post-secondary graduate

227 288 318

276 295 315

247 286 312

200 266 303

All aril ilto

97Q

9RF,

9RS

9fiR

Sources: OECD (1998); OECD and Statistics Canada (2000). * Low literacy skills are defined as literacy levels 1 or 2 on document literacy. Literacy is measured on a scale from 1 to 5, with levels 1 and 2 being the lowest levels.

Education, Skills, and Labour Market Outcomes 39

they are non-trivial. For example, on the document scale, the mean score in Germany, the top-ranked country, is 285, while that in the United Kingdom, the bottom-ranked country, is 268. An individual with a score of 268 is in the middle of the UK distribution, but would be at approximately the 33rd percentile of the distribution in Germany that is, about two-thirds of the adult population would have superior document literacy skills. Compared with many countries participating in the IALS survey, especially continental European countries, Canada, the United Kingdom, and the United States display substantial variation in the literacy skills of the adult population (Organization for Economic Cooperation and Development and Statistics Canada 1995,2000). This phenomenon is evident in Table 1.6(a) from a comparison of the lower and upper tails of the literacy distributions for Canada and Germany. At the 25th percentile, the German score exceeds that of Canada on all three literacy scales, the differential being especially large for document and quantitative literacy. However, at the 75th percentile the Canadian score exceeds that of Germany in both prose and document literacy and is approximately equal to that of Germany in quantitative literacy. In general, individuals in the top 25 per cent of the Canadian literacy distribution have higher literacy skills than their German counterparts, while individuals in the bottom one-quarter of the Canadian literacy distribution have lower skills than their German counterparts. Table 1.6(b) shows the percentage of adults with low literacy skills (level 1 or level 2) by broad age groups.17 For the adult population as a whole, the ranking is the same as before: Germany has the lowest percentage of adults with low literacy (the exception being the prose scale, on which Canada ranks at the top), followed by Canada, the United States, and the United Kingdom. However, important differences in the country rankings are evident among age groups. Among young adults (16-25 years of age), Canadians rank at the top, followed closely by Germans. In the United Kingdom, and especially in the United States, the incidence of low literacy skills among young adults is much higher. In contrast, Canada has the highest incidence of low literacy skills among those 46 to 55 years of age, exceeding even the UK on this dimension and substantially above Germany. The bottom panel, 1.6(c), provides some insight into the relationship between education and literacy in these countries. Among those with less than a completed secondary school education, literacy skills of Canadians are very poor, substantially below the United Kingdom and

40 W. Craig Riddell Germany but above the United States. However, average literacy scores improve substantially with educational attainment, and this gradient appears to be steepest in Canada. Canadian high school graduates rank second (after Germany) among this group of countries, and post-secondary graduates rank at the top, despite the very large fraction of the Canadian population with completed post-secondary education. These results suggest that Canadian literacy skills are reasonably good by international standards, especially among younger cohorts and post-secondary graduates.18 However, the literacy skills of older Canadians and those with less than a high school education are relatively poor. The fact that, on average, young Canadians display high levels of literacy compared to their counterparts in Germany, the United States, and the United Kingdom may be due in part to the increased quantity of education received by recent cohorts compared to earlier generations. This possibility is reinforced by the result that Canadian post-secondary graduates achieve literacy scores that are relatively high compared to their counterparts in other countries that participated in the IALS survey. In all of these countries, a disturbingly large fraction of the population has low levels of prose, document, and quantitative literacy. Nonetheless, to the extent that these information-processing skills used in daily activities are an outcome of the education system, this simple examination of the IALS data suggests that Canada's education system may be doing a reasonably good job of enhancing literacy skills. This favourable assessment is similar to that resulting from the recent PISA tests of student achievement, which also assessed the ability to apply knowledge to challenges that arise in daily activities. Education and Labour Market Success Schooling may have numerous consequences for individuals and society. For many people, there is some consumption value from the educational process. Human beings are curious creatures and enjoy learning and acquiring new knowledge. Even focusing on the investment aspects, education may enable people to more fully enjoy life, appreciate literature and culture, and be more informed and socially involved citizens. Although these and other potential consequences of schooling are important and should not be ignored, the consequences of education for employability, productivity, and earnings are of substantial importance for both economic and social policy.

Education, Skills, and Labour Market Outcomes 41

As many studies have documented, schooling is one of the best predictors of 'who gets ahead/ Better-educated workers earn higher wages, have greater earnings growth over their lifetimes, experience less unemployment, and work longer. Higher education is also associated with longer life expectancy, better health, and reduced participation in crime (Wolfe and Haveman 2001, Lochner and Moretti 2001, Lleras-Muney 2002). Many studies have analysed the relationship between education and labour market outcomes such as earnings.19 Two principal approaches have been used. The first is illustrated in Robert Allen's contribution to this volume.20 This method estimates life-cycle earnings profiles from data on groups of individuals with different levels of education. Combining these estimated earnings profiles with information on the costs of acquiring additional education - both the direct costs and the opportunity costs associated with the income foregone by not working - all allows the implied rate of return on the investment in additional education to be estimated. For example, the rate of return to a university degree compared to a high school diploma is estimated using the lifecycle earnings profiles for these two groups, together with information on the direct and opportunity costs of attending university compared to entering the labour force after completing high school. The second approach is based on estimation of an earnings function in which the dependent variable, the logarithm of earnings, is regressed on years of completed schooling, years of labour market experience (often entered as a quadratic to capture the non-linear relationship between earnings and experience), and additional variables that control for other influences on earnings. This specification is often referred to as the 'human capital earnings function' because it includes the two main measures of human capital available in most data sets - years of schooling and years of experience.21 The estimated coefficient associated with the 'years of schooling' variable is an estimate of the real rate of return to additional education.22 The earnings function approach is widely used because it provides a good approximation to the rate of return to education, as well as yielding insights into the relative magnitudes of other influences on earnings. Table 1.7 summarizes the results of a recent study of this relationship carried out by Ferrer and Riddell (2002). This study used Canadian census data that provide not only a large representative sample but also information on both years of completed schooling and all degrees, diplomas, and certificates received. This rich information on educational attainment allows one to analyse the influ-

42 W. Craig Riddell TABLE 1.7 Estimates of the private returns to education in Canada, with and without credential effects Males (i) Years of schooling (without credential effects) (ii) Years of schooling (with credential effects) High school graduate College diploma/trade certificate without high school Marginal effect over high school: college/trade with high school bachelor's degree Marginal effect over BA: medicine master's degree Marginal effect over MA: PhD

Females

5.9 3.3 5.2 7.6

8.6 5.5 6.1 8.4

6.6 22.8

5.9 25.2

34.1 4.6 4.2

30.0 7.0 0.8

Source: Ferrer and Riddell (2002).

ence on earnings of both years of schooling and 'sheepskin effects' increases in earnings associated with the receipt of a diploma, certificate, or degree. As indicated in the first row of the table, when 'years of schooling' alone is used to control for the influence of education, each additional year of schooling is estimated to be associated with an increase in female earnings of approximately 9 per cent, and of male earnings of approximately 6 per cent, after controlling for other influences. Females benefit more from higher education than do males, a reflection of the general finding that the gap between male and female earnings is largest at low levels of education and least at high levels. Ferrer and Riddell (2002) find that a more general specification in which both years of schooling and receipt of credentials (i.e., variables that indicate receipt of diplomas, certificates, and degrees) provides a better fit to the data. In this more general specification, the estimated coefficients on the 'years of schooling' variable decline but are nonetheless still substantial (3.3% for males and 5.5% for females). The total return to any specific level of education consists of the 'years of schooling' effect and the cumulative impact of the estimated 'sheepskin effects.' The main point to note for the purposes of this paper is that estimated rates of return to schooling are substantial. Furthermore, there is a significant earnings premium associated with completing an educational program relative to dropping out. Particularly large 'sheepskin effects' are associated with the completion of a university bachelor's

Education, Skills, and Labour Market Outcomes 43

degree and with degrees in medicine, dentistry, optometry, and veterinary medicine.23 Canadian studies using conventional methods to analyse the relationship between education and earnings obtain estimates of the 'return to schooling' that are similar to those obtained in many studies carried out in other developed countries: approximately 8 to 10 per cent rate of return when the analysis is based on annual earnings and 6 to 9 per cent when the analysis is based on weekly earnings.24 Such estimates compare favourably with rates of return on physical capital investments. The strong positive relationship between education and earnings is one of the most well-established relationships in social science. Many social scientists have, however, been reluctant to interpret this correlation as evidence that education exerts a causal effect on earnings. According to human capital theory, schooling raises earnings because it enhances workers' skills, thus making employees more productive and more valuable to employers. However, the positive relationship between earnings and schooling may arise because both education and earnings are correlated with unobserved factors such as ability, perseverance, and ambition (hereafter simply referred to as 'ability'). If there are systematic differences between the less- and well-educated that affect both schooling decisions and labour market success, then the correlation between education and earnings may reflect these other factors as well. According to signalling/screening theory, such differences could arise if employers use education as a signal of unobserved productivity-related factors such as ability or perseverance. In these circumstances, standard estimates of the return to schooling are likely to be biased upward because they do not take into account unobserved 'ability.' This 'omitted ability bias' issue is of fundamental importance not only for the question of how we should interpret the positive relationship between earnings and schooling, but also for the emphasis that should be placed on education in economic and social policy. To the extent that estimates of the return to schooling are biased upward because of unobserved factors, estimated average rates of return to education may substantially over-predict the economic benefits that a less-educated person would receive if he/she acquired additional schooling. The estimated average rates of return in the population reflect both the causal effect of schooling on productivity and earnings and the average return to the unobserved ability of the well-educated.

44 W. Craig Riddell

However, if those with low levels of education are also, on average, those with low ability or ambition, they can only expect to receive from any additional schooling the return associated with the causal effect of schooling on earnings. That is, average rates of return in the population reflect the causal effect of schooling on earnings and the return to unobserved factors. The marginal return - the impact of additional schooling for someone with low levels of education - may be substantially below the average return. In these circumstances, education may not be very effective in improving the employment or earnings prospects of relatively disadvantaged groups. Similarly, investing in additional education may not be an effective way of offsetting pressures for widening income inequality. Unbiased estimates of the causal effect of education on earnings are thus important for current debates about economic and social policy. How can such estimates be obtained? The most reliable method would be to conduct an experiment. Individuals randomly assigned to the treatment group would receive a larger 'dose' of education than those assigned to the control group. By following the two groups through time we could observe their subsequent earnings and obtain an unbiased estimate of the impact of schooling on labour market success. Random assignment would ensure that, on average, treatment and control groups would be equally represented by 'high ability' and 'low ability' individuals. In the absence of such experimental evidence, economists have tried to find quasi-experiments or 'natural experiments' that isolate the influence of education from the possible effects of unobserved ability. A large number of such studies have now been carried out, using data on identical twins or on sources of variation in education such as those implied by compulsory schooling laws or proximity to a college or university. Card (1999, 2000) provides a thorough discussion of the issues in this literature, as well as a review of empirical findings. A consistent result is that conventional ordinary least squares (OLS) estimates of the return to schooling tend, if anything, to underestimate rather than overestimate the causal impact of education on earnings. Why do conventional estimates generally understate the true return to schooling, when the presence of 'omitted ability bias' should cause these estimates to be upward biased? The reason appears to be that there are two additional sources of bias that operate in the opposite direction. First is the presence of measurement error in educational attainment (especially years of completed schooling). Measurement

Education, Skills, and Labour Market Outcomes 45

error in an explanatory variable causes the estimated coefficient to be biased towards zero. The downward bias due to measurement error thus acts in the opposite direction to any upward bias associated with unobserved ability. Second is what is sometimes referred to as 'discount rate bias/ The returns to schooling are not the same for all individuals in the population; rather, there is a distribution of such returns. Consider the case of individuals with high potential returns to education who do not pursue higher education - perhaps because of low family income, limited ability to borrow in order to finance human capital formation, or a family background in which the importance of education is not emphasized. For these 'high potential return' individuals, a policy intervention that results in increased educational attainment would have a substantial payoff. Indeed, the marginal return to the investment may exceed the average return in the population. In these circumstances, the average return from existing investments in education may understate the payoff to incremental investments. That is, policy interventions that focus on increasing education among high potential return individuals may be able to achieve rates of return that exceed those experienced by those who would invest in education even in the absence of any intervention. Two recent Canadian studies have pursued this 'natural experiment' approach. Lemieux and Card (2001) study the impact of the Veterans Rehabilitation Act - the Canadian 'GI Bill.' In order to ease the return of Second World War veterans into the labour market, the federal government provided strong financial incentives for veterans to attend university or other sorts of educational programs. Because many more young men from Ontario than from Quebec had served as soldiers, those from Ontario were significantly more likely to be eligible for these benefits. Lemieux and Card estimate that the VRA increased the education of the veteran cohort of Ontario men by 0.2 to 0.4 years. Further, they estimate the rate of return to schooling to be 14 to 16 percent, substantially higher than the OLS estimate with their data of 7 per cent. Sweetman (1999) investigates the impact on education and earnings of the education policy change in Newfoundland that raised the number of years of schooling required for high school graduation from eleven to twelve. He estimates that this intervention increased educational attainment of affected Newfoundland cohorts by 0.8 to 0.9 years. Estimated rates of return to the additional schooling are substantial: 17.0 per cent for females (versus an OLS estimate of 14.6%) and 11.8 per cent for males (compared to an OLS estimate of 10.8%).

46 W. Craig Riddell

As with this growing body of research, these Canadian studies conclude that conventional OLS estimates of the return to schooling are likely, if anything, to be biased downward, as opposed to being inflated by unobserved ability. Two principal conclusions follow from this body of research. First, rates of return to investments in education are high - and probably higher than has generally been believed on the basis of previous studies of the impact of education on earnings. Second, the payoff to marginal investments in education may exceed the average return in the population. There is no evidence that investments in higher education are experiencing diminishing returns because they require society to 'reach lower into the ability barrel/ Policy interventions that result in additional schooling being acquired by individuals from disadvantaged backgrounds, or those who face other barriers to acquiring human capital, appear likely to yield a substantial return in the form of enhanced employability and earnings, in addition to contributing to equity objectives. Education, Skills, and Labour Market Outcomes Most research on the determinants of earnings and other dimensions of labour market success uses only relatively crude indicators of human capital such as educational attainment and years of labour market experience. Educational attainment is typically measured by years of completed schooling or highest level of education reached. Labour market experience is unobserved in most data sets and is often proxied by 'potential experience/ as described previously (see note 22). However, individuals with the same educational attainment and years of potential experience may have substantially different skills, depending on their family environment, their fields of study, their work experience and onthe-job training, and other factors. More generally, education and labour market experience are 'inputs' into the production of human capital, not direct measures of the 'outcomes' - a set of skills, competencies, and knowledge. Although the relationships between inputs such as education and experience and outcomes such as employment and earnings have been extensively investigated, relatively little is known about the relationship between direct measures of skills and labour market outcomes. Green and Riddell (2001, 2003) use the Canadian component of the International Adult Literacy Survey (IALS) to investigate the relationship between education, skills, and labour market earnings. As dis-

Education, Skills, and Labour Market Outcomes 47

cussed previously in this chapter, the IALS was an innovative survey that married methods of educational testing with sample survey techniques. The survey contains standard questions on demographics and labour force behaviour, but also measures literacy in three domains: prose, document, and quantitative literacy (see Table 1.6). Conventional estimates of the return to schooling and to labour market experience confound two effects. The first is the impact of education and experience on skill production - the relationship between human capital inputs such as education and experience and outputs such as literacy skills or problem-solving skills. The second is the value placed on various skills in the labour market - the relationship between literacy or problem-solving skills and market earnings. When skills are not directly observed, the best researchers can do is to analyse the relationship between human capital inputs and labour market outcomes. However, the availability of directly observed skills in data sets that also contain information on labour force behaviour allows researchers to 'unpack' these two effects to some extent - to obtain estimates of both the skill production effect and the market valuation effect. Green and Riddell find that formal education exerts a substantial effect on the production of literacy skills in Canada, as was suggested by the tabulations in Table 1.6. However, they conclude that labour market experience has essentially no net effect on literacy production. These results suggest that policies aimed at improving cognitive skills such as literacy should focus on formal schooling. Policies designed to increase work experience can lead to earnings growth but they appear unlikely to enhance the cognitive skills of the workforce. Another important result is that the Canadian labour market places a high value on literacy skills. A twenty-point increase in the literacy score - equivalent to one-third of a standard deviation of the literacy score distribution - produces an increase in earnings equal to that associated with an extra year of formal schooling. Together these results imply that a significant amount of the 'return to education' as conventionally measured represents the combined effects of the contribution of schooling to producing literacy skills and the value placed on literacy in the labour market. Indeed, Green and Riddell (2001, 2003) estimate that about one-quarter to one-third of the 'return to education' is associated with these effects. The remainder reflects the impact of education on the production of other (unobserved) skills that are valued in the labour market. We are at an early stage in investigating the linkages between human

48 W. Craig Riddell

capital inputs such as education and experience and the skills and knowledge of workers, as well as the relationship between workers' skills and competencies and their labour market consequences. As new data on the skills and knowledge of the workforce become available, this area promises to be an exciting area of research. Conclusions This chapter has provided a general review of the recent Canadian evidence relating to the linkages between expenditures on education and skill formation, as well as on the relationships between educational attainment, skills, and labour market outcomes. Several conclusions follow from the analysis: • Canada invests heavily in education. Relative to other G7 or OECD countries, Canada ranks near the top in terms of expenditure per student or the fraction of GDP devoted to elementary, secondary, and post-secondary education. Within Canada there are large differences across provinces in expenditure per student, especially at the elementary and secondary level. • One consequence of this substantial expenditure is a population that is well educated by international standards. Canada compares favourably with other G7 and OECD countries in terms of most measures of educational attainment. Compared to the United States, Canada has lower educational attainment at both the bottom (less than completed high school) and top (university degree) of the education distribution. Where Canada stands out is in the middle of the distribution - those who have completed high school or a non-university secondary program. The proportion of Canada's population with a non-university post-secondary education is much higher than that of any other OECD country. However, standard measures may overstate Canadian educational attainment in this dimension to some extent because not all those with a college diploma or trade certificate have completed high school, and because of the unique features of Quebec's CEGEP system. • Recent international data on Canadian student achievement paint a mixed picture. Results from TIMSS, curriculum-based tests of achievement in mathematics and science, indicate that Canadian student achievement is satisfactory but not as good as one might expect given Canada's relatively high expenditure on elementary

Education, Skills, and Labour Market Outcomes 49

•

•

•

•

and secondary schooling. In contrast, the performance of fifteenyear-old Canadian students in the recent PISA tests, which assessed the ability to apply knowledge in reading, mathematics, and science, was excellent. In reading, Canadian students were the top performers in the G7 countries, while in mathematics and science Canada - together with Japan and the United Kingdom - ranked among the top three G7 countries, and substantially above the remaining countries. The PISA results suggest that Canada appears to obtain reasonably good 'value for money' from the elementary and secondary school system, at least as measured by average student achievement in reading, mathematics, and science. Canada ranks at or near the top of the G7 countries in terms of expenditure per student on elementary and secondary schooling, and places in the top three of the G7 countries in terms of student performance. However the TIMSS results lead to a somewhat less favourable assessment. It is important in future research to try to understand the reasons for the differences between the TIMSS and PISA findings. There are substantial variations in student achievement across Canadian provinces. Some provinces - such as Quebec in mathematics and Alberta in science - are able to obtain very high levels of achievement within the existing Canadian social, cultural, and fiscal framework. In other provinces, especially the Atlantic provinces, student achievement generally falls below the Canadian average and is relatively low by international standards. The source of these provincial variations is an important subject for future research. One factor that appears to account for some of the variation across provinces in student achievement is the presence in some jurisdictions of curriculum-based examinations at the end of secondary school. Evidence points to beneficial effects of such examinations not only on student achievement but also on school administrator behaviour, teacher behaviour, and home behaviour and attitudes. The literacy skills of the adult population are above average among the G7 countries that participated in the IALS survey. Canada, like the United States and the United Kingdom, has a high variance in its literacy skills across the population compared to European countries such as Germany. By international standards, older and less welleducated Canadians have relatively poor literacy skills, whereas younger and well-educated Canadians have relatively good literacy skills compared to their counterparts in other G7 countries. The

50 W. Craig Riddell

•

•

•

•

•

IALS data thus suggest that Canada's education system is doing a reasonably good job of enhancing literacy skills. Canadian studies using conventional methods to analyse the relationship between education and earnings obtain estimates of the 'return to schooling' that are similar to those obtained in many studies carried out in other developed countries: approximately 8 to 10 per cent rate of return when the analysis is based on annual earnings and 6 to 9 per cent when the analysis is based on weekly earnings. Such estimates compare favourably with rates of return on physical capital investments. In Canada, women benefit substantially more from additional education than do men. Recent evidence indicates that the rate of return to education is approximately 9 per cent for women and 6 per cent for men. There is a significant earnings premium associated with completing an educational program relative to dropping out. Particularly large 'sheepskin effects' are associated with the completion of a university bachelor's degree and with degrees in medicine, dentistry, optometry, and veterinary medicine. Important recent advances have taken place in our understanding of the relationship between education and labour market success. Conventional estimates of the return to schooling appear, if anything, to be biased downward - so the causal effect of education on earnings appears to be higher than previously believed. Further, the marginal return to incremental investments in education may exceed the average return from previous investments. There is no evidence that investments in schooling are running into diminishing returns. These results suggest that investments in human capital remain an important potential source of growth in earnings and employability. Conventional estimates of the return to schooling and to work experience confound two effects: the impact of education and experience on skill production and the value placed on skills in the labour market. The availability of data on directly observed skills allows researchers to 'unpack' these two effects. Recent evidence indicates that formal education exerts a substantial effect on the production of literacy skills in Canada. However, work experience has no net effect on literacy production. In addition, Canada's labour market places a high value on literacy skills. A twenty-point increase in the literacy score - equivalent to one-third of a standard deviation of the literacy score distribution - produces an increase in earnings equal to that

Education, Skills, and Labour Market Outcomes 51

associated with an extra year of formal schooling. These results imply that a substantial fraction - one-quarter to one-third - of the 'return to education' as conventionally measured represents the combined effects of the contribution of schooling to producing literacy skills, and the value placed on literacy in the labour market. Notes This chapter draws on joint work with Western Research Network on Education and Training (WRNET) colleagues Ana Ferrer, David Green, and Arthur Sweetman. The paper has also benefited from interactions with other WRNET members, including Bob Allen, Jane Gaskell, Ben Levin, Thomas Lemieux, Kjell Rubenson, and Alice Nakamura. I alone am responsible for any errors and omissions. 1 The series of OECD publications Education at a Glance: OECD Indicators make a valuable contribution to international comparisons of educational expenditures and outcomes in OECD countries. 2 The Scandinavian countries and Canada are typically ranked at the top of the OECD in terms of the percentage of GDP devoted to education (OECD, Education at a Glance, various issues). 3 The comparison of Canada and the U.S. with several European countries is quite sensitive to the definition of 'upper secondary education.' For example, France and the UK have both short-duration and long-duration upper secondary schooling, whereas these are rare in North America. If the short upper secondary programs are excluded, the UK's proportion with upper secondary or higher drops from 82 to 62% and the French figure falls from 62 to 34%. See OECD (2001). 4 The measurement of years of completed schooling is problematic in countries like Germany where there are extensive apprenticeship programs that combine work and school. 5 This has been the structure of the educational attainment questions since a major revision to these questions in 1990. Prior to that time, high school completion was required in order to be classified in one of the post-secondary education categories, even in the case of respondents who had completed a trade certificate or community college program. 6 The census, taken in June 1996, asks about income and weeks worked during the previous year. The wage measure used is the weekly wage for those employed in 1995.

52 W. Craig Riddell 7 For some additional Canadian evidence, see Ferrer and Riddell (2002). Nonetheless, further investigation of the comparability of these categories across countries appears warranted. 8 The first round of TIMSS data collection was carried out in 1995, and a further set of tests took place in 1999. A third round is scheduled for 2003. The acronym TIMSS now stands for Trends in Mathematics and Science Study. 9 See Crocker (2002) for an assessment of SAIP. 10 Canada Communication Group (1992) and Riddell (1995) summarize and assess the results of earlier Canadian student achievement tests. 11 Results for England and Germany, as well as the U.S. results for grade 8 math and science, should be treated with some caution because they did not meet the requirements for a nationally representative sample. 12 The results for Canada as a whole are based on a representative sample of schools in all provinces and territories with the exception of PEI, the only province that did not participate in TIMSS. 13 For example, at the grade 4 level five countries (Korea, Japan, Netherlands, United States, and Australia) had significantly higher science scores and eight countries (Korea, Japan, Singapore, Hong Kong, Netherlands, Czech Republic, Austria, and Slovenia) had significantly higher mathematics scores. 14 An interesting question is whether features of the Canadian school system contribute to superior performance on the ability to apply knowledge to daily activities relative to curriculum-based knowledge. Note also that the TIMSS results reported in Table 1.4 assessed student achievement in Grades 4 and 8, whereas PISA tests were administered to students near the end of secondary school. 15 See Organization for Economic Cooperation and Development and Statistics Canada (1995,2000) for further details on this survey. 16 France also participated in IALS, but the French results have not been released publicly. 17 On a scale of 0 to 500, literacy level 1 corresponds to a score from 0 to 225 and level 2 corresponds to a score from 226 to 275. 18 This conclusion continues to hold if a wider group of countries, including Australia and Sweden, is examined (Riddell and Sweetman 2000). 19 Earnings is the most commonly used measure of labour market success because it captures both the wage rate or 'price' of labour services and employ ability (hours, weeks, and years of work). 20 For other recent examples of the use of this method in Canada, see Rathje and Emery (2002) and Vaillancourt and Bourdeau-Primeau (2002). 21 Labour market experience is often not directly observed, and is estimated

Education, Skills, and Labour Market Outcomes 53 by potential experience, defined as Age minus Years of schooling minus 6. This measure assumes that individuals begin school at age 6, enter the labour market immediately after leaving school, and work continually until the date of the survey. 22 Mincer (1974) showed that the estimated coefficient of the 'years of schooling' variable in a log earnings equation equals the rate of return on education if the cost of an additional year of schooling equals the opportunity cost of foregone earnings. Because foregone earnings constitute the main cost of additional years of education, the estimated coefficient on the 'years of schooling' variable is frequently referred to as the estimated 'return to education.' 23 Although this evidence of substantial 'sheepskin' or credential effects may reflect signalling or screening in the labour market, it is also consistent with a human capital perspective if the educational program consists of a package of complementary courses or if program completers learn more than dropouts (Ferrer and Riddell 2002). In addition, in fields such as medicine, the large estimated sheepskin effect may reflect professional licensing requirements and restrictions on entry into the profession. 24 Estimates of the impact of schooling on annual earnings exceed those of the impact on weekly or hourly earnings because those with more education also work more weeks per year. References Bishop, J.H. 1995. The impact of curriculum-based external examinations on school priorities and student learning. International Journal of Educational Research 23(8): 653-752. - 1997. The effect of national standards and curriculum-based external exams on student achievement. American Economic Review 87(2): 260-4. - 2002. School choice, exams and achievement. In Towards evidence-based policy for Canadian education, ed. P. de Broucker and A. Sweetman, 385—417. Kingston: John Deutsch Institute for the Study of Economic Policy and Statistics Canada. Canada Communication Group. 1992. Education and training in Canada. Ottawa: Minister of Supply and Services. Card, D. 1999. The causal effect of education on earnings. In Handbook of labor economics. Volume 3A. Ed. O. Ashenfelter and D. Card. Amsterdam: North Holland. - 2000. Estimating the return to schooling: Progress on some persistent prob-

54 W. Craig Riddell lems. NBER working paper no. 7769 (June). Cambridge, Mass.: National Bureau of Economic Research. Council of Ministers of Education, Canada and Statistics Canada. 2000. Education indicators in Canada: Report of the pan-Canadian education indicators program 1999. Toronto and Ottawa: Council of Ministers of Education, Canada and Statistics Canada. Crocker, R.K. 2002. A decade of SAIP: What have we learned? What do we need to know? In Towards evidence-based policy for Canadian education, ed. P. de Broucker and A. Sweetman, 207-225. Kingston, Ont.: John Deutsch Institute for the Study of Economic Policy and Statistics Canada. Crompton, S. 1996. The marginally literate workforce. Perspectives on Labour and Income 8:14-21. Ferrer, A.M., and W.C. Riddell. 2002. The role of credentials in the Canadian labour market. Canadian Journal of Economics 35 (November): 879-905. Green, D.A., and W. C. Riddell. 2001. Literacy, numeracy and labour market outcomes in Canada. Ottawa: Statistics Canada, Human Resources Development Canada and National Literacy Secretariat, IALS monograph series no. 8. - 2003. Literacy and earnings: An investigation of the interaction of cognitive and unobserved skills in earnings generation. Labour Economics 10 (April): 165-84. Human Resources Development Canada, Statistics Canada, and Council of Ministers of Education, Canada. 2001. Measuring up: The performance of Canada's youth in reading, mathematics and science. OECD PISA study: First results for Canadians aged 15. Ottawa: Statistics Canada. Lleras-Muney, A. 2002. The relationship between education and adult mortality in the United States. NBER working paper no. 8986 (June). Cambridge, Mass.: National Bureau of Economic Research. Lemieux, T., and D. Card. 2001. Education, earnings, and the 'Canadian G.I. Bill.' Canadian Journal of Economics 34 (2): 313-44. Lochner, L., and E. Moretti. 2001. The Effect of Education on Crime: Evidence from prison inmates, arrests and self-reports. NBER working paper no. 8605 (November). Cambridge, Mass.: National Bureau of Economic Research. Mincer, J. 1974. Schooling, experience and earnings. New York: Columbia University Press. OECD (Organization for Economic Cooperation and Development). 1998. Human capital investment: An international comparison. Paris: OECD. - 2001. Education at a glance: OECD indicators 2001. Paris: OECD. Organization for Economic Cooperation and Development and Statistics Canada. 1995. Literacy, economy and society. Paris and Ottawa: OECD and Statistics Canada.

Education, Skills, and Labour Market Outcomes 55 - 2000. Literacy in the information age. Paris and Ottawa: OECD and Statistics Canada. Rathje, K.A., and J.C.H. Emery. 2002. Returns to university education in Canada using new estimates of program costs. In Renovating the ivory tower: Canadian universities and the knowledge economy, ed. D. Laidler, 241-64. Toronto: C.D. Howe Institute. Riddell, W.C. 1995. Human capital formation in Canada: Recent developments and policy responses. In Labour market polarization and social policy reform, ed. K.G. Banting and C.M. Beach, 125-172. Kingston, Ont.: School of Policy Studies, Queen's University. Riddell, W.C., and A. Sweetman. 2000. Human capital formation in a period of rapid change. In Adapting public policy to a labour market in transition, ed. W.C. Riddell and F. St-Hilaire, 85-141. Montreal: Institute for Research on Public Policy. Robitaille, D.F., and A.R. Taylor. 2000. TIMSS Canada Report Volume 5: New findings for a new century. Vancouver: Department of Curriculum Studies, University of British Columbia. Robitaille, D.F., A.R. Taylor, and G. Orpwood. 1996. 77MSS Canada Report Volume 1: Grade 8. Vancouver: Department of Curriculum Studies, University of British Columbia. - 1997. TIMSS Canada Report Volume 2: Grade 4. Vancouver: Department of Curriculum Studies, University of British Columbia. Sweetman, A. 1999. What if high school were a year longer? Evidence from Newfoundland. Western Research Network on Education and Training (WRNET) working paper no. 00-01. Vancouver: Faculty of Education, University of British Columbia. Vaillancourt, F., and S. Bourdeau-Primeau. 2002. The returns to university education in Canada, 1990 and 1995. In Renovating the ivory tower: Canadian universities and the knowledge economy, ed. D. Laidler, 215-^40. Toronto: C.D. Howe Institute. Wolfe, B., and R. Haveman. 2001. Accounting for the social and non-market benefits of education. In The contribution of human and social capital to sustained economic growth and well-being, ed. J. Helliwell, 221-50. Vancouver: OECD/Human Resources Development Canada and UBC.

2 Education and Technological Revolutions: The Role of the Social Sciences and the Humanities in the Knowledge-Based Economy ROBERT C. ALLEN

The modern economy is characterized by a high rate of technological progress involving new 'high-tech' products. This phase of history began in the late nineteenth century with the invention of chemical dyes, the first pharmaceuticals, electrical equipment, the internal combustion engine, and the telephone. Research-oriented universities and the industrial research laboratory date back to this period. The twentieth century saw many new high-tech products ranging from the automobile to the airplane to modern electronics and chemicals. In recent decades, the computer and other information technologies have revolutionized the economy and are continuing to do so. We may be witnessing the birth of even more dramatic changes arising from biotechnology. The question is: what sort of educational programs will help Canada capitalize on the possibilities raised by the modern technological revolutions? A common answer is that, since technological revolutions involve high-tech products, what Canada needs are technologists. I will refer to this view as 'techism.' Its proponents - 'techniks' - maintain that our prosperity requires a redirection of resources towards technical education. Techism comes in two forms. Highbrow techism emphasizes the need for highly educated scientists and engineers to promote the expansion of high-tech manufacturing and related businesses. Lowbrow techism, as exemplified by reports like British Columbia's Training/or What? (BCLFDB 1995) emphasizes the need for the technical skills taught in one- and two-year college programs. Enthusiasts of techism either explicitly argue that too many resources are currently allocated to education, the humanities, and social sciences, or make the

Education and Technological Revolutions 57

same point implicitly by omitting them from their recommendations for future funding. This chapter argues that techism is too narrow to prepare Canada for the new millennium. While techniks are right that the demand for technically trained workers is growing, the same is true for graduates in education, the humanities, and social sciences. These fields are in demand because the widespread utilization of computers and information technology has revolutionized the organization of businesses and government. The new-style organizations put a premium on workers who can relate models to real situations, work well with other members of a management team or with clients, and who can speak and write effectively. These skills are developed in humanities and social sciences programs. Techism, which concentrates on the production of new technologies and on the nuts and bolts of their operation, misses the organizational revolutions that accompany the adoption of new technologies. The Economic Success of Graduates in 1996 A common view that underlies techism is the belief that graduates in education, the humanities, and social sciences cannot find good jobs. Since they do not have the skills that employers require, so the argument goes, arts graduates end up driving taxis or making cappuccinos if they can find a job at all. What is needed for success is technical training, either the sort taught in one- and two-year technical/trade/ vocational/career courses or the sophisticated sort taught in university engineering departments. This common view can be explored using census data. I begin with the 1996 census and add information from the 1991 census for comparison. Both censuses collected information about the employment, earnings, and job characteristics of Canadian workers. All figures in this study are computed from the micro-data files of these censuses. The micro-data files contain coded information on the individual census returns of about 3 per cent of the Canadian population. The Canadian census is a critical source for assessing techism, for it is one of the very few largescale surveys that include the individual's field of study. With census micro-data one can compare the employment success, for instance, of engineering and humanities graduates. The comparisons call into question the common view that the humanities and social sciences are poor preparation for the emerging knowledge-based economy.

58 Robert C. Allen TABLE 2.1 Unemployment rates (%), ages 25-29, by level of education 1996 Level of education

Women

Men

High school non-completer High school graduate Trade certificate College diploma University, less than bachelor Bachelor's degree Graduate certificate or degree

20.6 11.5 10.6 7.9 6.5 4.4 4.8

19.0 10.0 12.6 6.8 7.8 4.8 3.4

Overall

10.1

11.1

Source: Census of Canada, 1996, microdata file. Note: Excludes people in school in the previous year.

Unemployment rates provide a basic test of employability. Table 2.1 shows unemployment rates for various education levels and fields of study.1 Unemployment is a more serious problem for younger people than for older people, so Table 2.1 concentrates on Canadians aged 25 to 29. If arts graduates cannot find work, the problem should show up in that age group. As a general rule, unemployment is worse for the uneducated, and Table 2.1 bears that out. Thus, the highest unemployment rates are those of high school dropouts. They are followed by high school graduates and people with a technical or trade certificate, and then by those with a college diploma. (Generally, trade certificates are awarded for completing training courses of less than a year, but they are also awarded to completers of apprenticeship programs. College diplomas are awarded for completing two-year programs.) University graduates were most likely to be employed. Table 2.1 is not exceptional; this pattern is repeatedly observed in labour force survey data. The mediocre performance of those with college diplomas and the poor performance of those with trade certificates are inconsistent with the lowbrow techism view that specific skills guarantee a job in the new knowledgebased economy. Table 2.2 shows unemployment rates for twenty-five to twenty-nineyear-olds with a bachelor's degree. The results are broken down by

Education and Technological Revolutions 59 TABLE 2.2 Unemployment rate (%) by field, bachelor's degrees, ages 25-29 Field

Women

Men

Education Fine arts Humanities Social sciences Commerce Agriculture/biology Engineering Nursing Other health-related Math/physics

2.9 5.9 7.0 3.4 4.3 7.2 8.8 3.1 0.6 7.8

3.5 6.5 6.7 5.5 4.3 4.6 4.3 0.0 5.7 3.8

Source: Census of Canada, 1996, microdata file. Note: Excludes people in school in the previous year. The Statistics Canada category 'other' has been excluded.

field of study. The experience of women is remarkably inconsistent with highbrow techism, for the worst employment records were those of women in engineering, followed by women in mathematics and in the physical and biological sciences. Graduates in the humanities did better than these, and the records of graduates in the social sciences and education were exceptionally fine. The experience of men was not quite as favourable. Social science and humanities graduates had unemployment rates that were slightly above those of men in engineering or commerce. However, the unemployment rate of men in the humanities was still less than the rate for men who had a college diploma and much below that for men with a technical or trade certificate. The evidence of unemployment rates is, therefore, strongly inconsistent with techism. Graduates in the humanities, social sciences, and education may have been able to get jobs, but were they good jobs? The census can also be used to assess job quality. Table 2.3 shows the proportion of employed people with managerial or professional jobs. Clearly, the probability of having these jobs increased with educational attainment: the probability was only 13.9 per cent for a high school dropout and over 95 per cent for an MD or PhD. University graduates all had much higher probabilities than those with less education.

60 Robert C. Allen TABLE 2.3 Attainment of managerial or professional occupation by level of education, Canadian workers, 1996

Level of education

% with managerial/ professional occupation

High school non-completed High school graduate Trade certificate College diploma University less than bachelor Bachelor's degree University greater than bachelor Medicine Master's PhD

13.9 25.2 25.4 48.5 59.2 71.8 83.3 95.5 88.4 95.1

Total

37.0

Source: Census of Canada, 1996, microdata file. Note: Excludes people in school in the previous year.

This positive outcome extended to graduates in education, the social sciences, and the humanities, as Table 2.4 shows. Graduates of programs targeted to particular professions had the highest chances of being professional or managerial employees, notably graduates in nursing or health (over 90%), followed closely by education (85%). Engineering and mathematics and physical science graduates had somewhat lower probabilities of managerial or professional work (just over 80%). The remaining programs had the lowest probability, but their rankings were almost identical (around 70%). This very broad group includes graduates in fine arts, the humanities, and social sciences - no surprise to proponents of techism - but also commerce and the biological sciences. While commerce is reputed to teach the skills of management, it is significant that commerce is no more successful than the social sciences or fine arts in placing its graduates in managerial and professional work. Also, while graduates in the humanities, social sciences, commerce, and biology had a lower chance of landing a professional or managerial job than did graduates in health, the probability was still much higher than that of a two-year college graduate (49%). Universities claim to prepare people for managerial and professional work, and Tables 2.3 and 2.4 show that the boast is not an idle one.

Education and Technological REvolutions 61 TABLE 2.4 Percentage of Canadian university graduates with managerial or professional occupation, 1996 Field

%

Education Fine arts Humanities Social sciences Commerce Agriculture/biology Engineering Nursing Other health-related Mathematics/physical science

85.1 74.3 69.9 70.9 70.4 70.4 82.6 90.6 93.8 80.7

Total

77.4

Source: Census of Canada, 1996, microdata file. Note: Excludes people in school in the previous year. The Statistics Canada category 'other' is excluded. University graduates include people with a bachelor's degree or higher.

Income is another indicator of job quality. Tables 2.5 and 2.6 summarize the annual earnings of full-time, full-year workers in 1995. (The effects of non-employment and part-time work will be considered later when social rates of return are calculated.) Separate tables are shown for men and women, and earnings are broken down by age. The tables show that graduates in the humanities and social sciences generally earn more than those who did not go to university, and therefore contradict the common view that graduates in these fields do not have the skills that employers require. Tabulations of earnings typically show that earnings increase with educational level and with age due to greater experience. Tables 2.5 and 2.6 show the same patterns, with an important exception: the experience of people with a trade or technical certificate. In the case of women, these people generally earned less than high school graduates. The same was true of men in the older age groups. These results parallel the earlier findings about unemployment rates, and parallels also turn up when employment growth is analysed. These findings contradict lowbrow techism by showing that the completers of one-year

62 Robert C. Allen TABLE 2.5 Annual income of women by level of education and field of study, 1995 Age groups Level/field

20-29

30-39

40-49

50-59

High school non-compleler High school graduate Trade certificate College diploma

19,077 20,778 21,173 24,499

23,697 27,805 26,053 32,169

24,924 29,927 28,284 34,308

24,391 29,746 29,168 34,351

Bachelor's degree education human/fine arts social science commerce agriculture/biology engineering nursing other health-related math/physical science

30,719 29,820 27,474 28,275 30,965 30,892 36,318 36,799 40,125 36,108

41,499 39,296 38,543 40,968 44,473 37,536 46,797 42,453 46,669 44,591

45,699 46,493 42,328 46,756 44,920 41,296 42,969 47,081 45,554 50,687

47,285 46,356 493,439 49,352 42,347 47,434 35,893 49,216 47,071 45,816

Graduate degree education human/fine arts social science commerce agriculture/biology engineering nursing other health-related math/Dhvsical science

38,824 32,094 32,643 33,379 39,437 33,132 40,278 42,500 37,893 32.096

46,009 45,635 39,935 46,233 52,999 40,174 46,142 40,321 50,259 47.232

53,235 53,683 48,661 54,757 59,074 43,917 44,723 53,869 52,895 58.270

55,091 54,866 53,117 55,603 57,617 57,693 60,876 60,290 59,699 51 477

Source: Census of Canada, 1996, microdata file. Note: These are total wages and salaries earned in 1995 by people employed full time for 49 or more weeks in the year. Self-employed are excluded as are people enrolled as students during the academic year before the census or reporting total wages and salaries of less than $4000 for the year. Degrees in medicine and dentistry are excluded. Graduate degree includes PhD, master's, and graduate certificates.

training programs do not readily find jobs or earn high incomes. More generally, the result is an important example of how specific skills training, by itself, has no payoff in the knowledge-based economy. Tables 2.5 and 2.6 show that university degrees have led to higher incomes. The success of university graduates with a bachelor's degree

Education and Technological Revolutions 63 TABLE 2.6 Annual income of men by level of education and field of study, 1995 Age groups Level/field

20-29

30-39

40-49

50-59

High school non-completer High school graduate Trade certificate College diploma

25,063 27,046 30,626 32,241

34,145 37,928 40,725 42,962

38,228 44,449 45,775 49,774

38,735 47,876 46,105 52,944

Bachelor's degree education human/fine arts social science commerce agriculture/biology engineering nursing other health-related math/physical science

35,721 31,321 29,671 33,786 35,562 31,680 40,800 34,065 42,371 38,814

53,107 43,067 39,641 53,518 59,969 48,070 57,162 42,687 53,232 55,278

61,188 52,107 52,175 61,031 65,384 55,860 69,162 42,629 57,235 63,494

65,305 54,004 55,180 68,439 69,416 56,505 76,189 48,055 54,831 67,775

Graduate degree education human/fine arts social science commerce agriculture/biology engineering nursing other health-related math/physical science

38,706 35,682 34,484 36,186 43,365 19,088 42,045 27,200 36,722

56,562 49,923 40,304 56,732 68,490 42,846 56,352

65,868 58,739 51,647 66,931 80,830 60,257 68,256

72,222 63,069 60,442 76,046 87,300 68,133 78,449

60,179 53,982

77,113 65,007

85,851 73,720

Source: Census of Canada, 1996, microdata file. Note: These are total wages and salaries earned in 1995 by people employed full time for 49 or more weeks in the year. Self-employed are excluded as are people enrolled as students during the academic year before the census or reporting total wages and salaries of less than $4000 for the year. Degrees in medicine and dentistry are excluded. Graduate degree includes PhD, master's, and graduate certificates.

included women in all fields, for their average earnings always exceeded those of college graduates, let alone high school graduates. It is true that the earnings of humanities, social science, and education graduates were at the low end of the distribution for women in their twenties, but graduates in these fields also had the highest rate of earn-

64 Robert C. Allen

ings growth and ended up at the top of the distribution for women in their fifties. Arts graduates tend to earn less in their twenties than do people who have completed more specifically focused programs, but the arts graduates often catch up with, and then surpass, people in other fields. Men with bachelor's degrees in the humanities and social sciences also did well, but there are some facts that give prima facie support to techism. In three out of the four age groups, engineers earned the highest income, which supports high-brow techism. In most age groups, however, graduates in social science, commerce, and the physical sciences were not far behind. Graduates in health-related fields, the biological sciences, education, and the humanities had lower incomes. Graduates in education and the humanities earned less than, or about the same as, college graduates in their twenties and thirties, but realized higher incomes after the age of forty. The slower start of humanities graduates provides some support for lowbrow techism. Nevertheless, their earnings exceeded those of high school graduates at all ages, which makes degrees in the humanities a profitable investment for men. The notion that men with degrees in the humanities cannot find good paying jobs is refuted by Table 2.6. Highbrow techism places great emphasis on the need for graduate education in the sciences and engineering, but the labour market places less value on those degrees. The high marketability of the Master's of Business Administration (MBA) means that graduate degrees in commerce netted the highest earnings for most age groups. Engineers and scientists were not far behind, as were social scientists whose earnings were frequently on a par. Graduates in the humanities and education were not at the top of the distribution of graduate earnings, but nonetheless earned more than people with undergraduate degrees in the same area. Graduate education was increasing the economic value of graduates in the humanities and education as well as in the social sciences. The snapshot of the labour market in 1995 and 1996 provides support for some aspects of techism but also shows that it is too narrow a view of what employers demand. It is a big mistake to believe that one-year technical or trade courses lead to a good job with high earnings, for these people experience high unemployment and little or no income gain over high school earnings. The graduates of two-year college programs do better in terms of income and employment. The lowest unemployment rates, highest occupational status, and highest incomes are

Education and Technological Revolutions 65

realized by university graduates. These favourable outcomes are realized by humanities, social science, and education graduates as well as by those in engineering and the natural sciences. Rates of Return While university graduates generally earn more than those with less education, does the difference justify the costs of their educations? This question is particularly important for the humanities, social sciences, and education since they were not at the top of the earnings distribution. Are the earnings of their graduates with bachelor's degrees sufficiently above those of high school graduates to cover the cost of undergraduate education? And are the earnings of those with master's and doctoral degrees high enough to cover the cost of expensive graduate programs? These questions are usually addressed by computing the social rate of return to education. In this approach, education is analysed as an investment.2 The investor is the student who does the studying and the government (federal and provincial), that provides the university. The benefit or gain from this investment is the rise in income that students earn after graduation. This gain is ultimately divided between the student and the government, since some of the increased income is taxed away. That division, however, is irrelevant in a 'social' rate of return calculation, which analyses the combined costs and benefits of all parties. The cost of the investment consists of (1) the earnings that the student loses by studying instead of working; (2) the cost to the student of books and supplies needed for the course of study; and (3) the cost to the government of operating the university. The costs do not include the student's food, clothing, or housing since these expenses would be incurred even if the student were not studying, nor do they include tuition fees. Tuition fees are a cost to the student but a gain to the university and so cancel out when the combined (that is, social) rate of return is calculated. The level of tuition fees as well as taxes on income earned after graduation affect how the profits of education are divided between the student and the government but do not affect the profitability of the investment to the student and the state together. That joint rate of profit is the social rate of return. I will explain how these various components of the social rate of return were measured. Calculating the increase in income that results from additional education requires that a time path be specified. For

66 Robert C. Allen

bachelor's degrees, I assumed that the student attended university from ages eighteen through twenty-one, immediately after high school. The income gain from the investment is measured as the average income earned by university graduates minus the average income earned by high school graduates.3 These increases were computed from tables like Tables 2.5 and 2.6. To compute the rate of return, three modifications were made to the tables. First, average incomes were calculated for more age categories (specifically 20-24,25-29,30-39,40-49, 50-59, and 60-65). Using more categories means that the averages provide a more exact tracking of the sharp increases in income that occur in the twenties and the fall off in income that occurs in the sixties as people retire. Second, the income concept was broadened to include self-employment income rather than just wages and salaries. Third, the averages were taken over all people in the age group whether they were working or not. In that way, the effects of non-employment and part-time work are included. If the educational investment turns out to be profitable, it means that the earnings of the employed university graduates are large enough to cover not only the costs of their own education but also the costs of those not working. The tabulated data on average earnings provide a snapshot of the situation in 1995, which was taken to represent the situation at the time of graduation. If those average earnings were never to change, then the 1995 age-earnings profile would indicate the earnings of an individual over the course of his or her whole working life. However, both inflation and economic growth will cause the age-earnings profile to shift. To project an individual's earnings into the future, the age-earnings profile was increased by 2 per cent per year for inflation and 1 per cent per year for economic growth. The latter represents the long-run growth in productivity and is the assumption made by the Organization for Economic Cooperation and Development (OECD) when it computes rates of return to education (Alsalam and Condy 1995; OECD 1997). To compute the rise in income from university education, separate projections were made for men and women in each field of study and for high school graduates. The same logic was applied to the earnings increase from graduate education, but the procedure had to be modified in detail. Graduate education was modelled as occurring immediately after undergraduate education. Master's degrees were analysed as either one- or twoyear programs. PhD programs were treated as being four years long (two years of courses and two years of thesis supervision) and were

Education and Technological Revolutions 67

analysed as following immediately after the completion of a one-year master's program. Rates of return were calculated for the combined master's/PhD sequence. Tabulating average earnings was not an effective way of computing the earnings gain from graduate degrees because there were not enough observations in each age group to map out a stable age-earnings profile. Instead, regression analysis was used.4 The sample included all of the observations of people aged twenty to sixty-five with a bachelor's, master's, or PhD, whether they worked or not. The dependent variable was wages and salaries plus the net income from self-employment. The explanatory variables were age, age squared, and age to the fourth power - these captured the rise of income with experience for people in their twenties and the fall in income as people retired - and dummy variables for sex, a master's degree, and a PhD.5 The dummy variables measured the effects of sex and graduate degrees on income. The regression results are summarized in Table 2.7. The age variables are generally significant and imply plausible age-earnings profiles. The coefficient of the sex variable measures the economic disadvantage of being a woman. In nursing, it is negligible. In the fine arts and humanities, it is about $8000 per year. The disadvantage increases to $10,000 in education; $14,000 in the sciences; $16,000 in commerce; and $18,000$19,000 in social sciences, engineering, and health-related fields. The coefficients of the master's and PhD variables measure the extra income (compared to a bachelor's degree in the same field) of those graduate degrees. The highest returns for master's degrees are realized in education, commerce, nursing, and other health-related fields. Master's degrees in the humanities and social sciences command income gains that exceed the gains in the sciences and engineering where they are not statistically significant. PhD degrees in education, the humanities, and social sciences realize income gains that are on a par with those in most other fields. The low gain in income from a PhD in engineering raises the question of whether the market values the skills of those highly trained technologists as much as highbrow techniks believe. In the regressions for education, the humanities, and social sciences, the coefficients of the PhD are substantially greater than those of the master's indicating that the PhD adds a lot to the earnings of a master's. This gain largely disappears in the case of commerce for the coefficient of the PhD is only slightly above that of the master's; in other words, the MBA gives so much earning power that a PhD can barely top it. The situation is different again in engineering and the natural sciences. For

TABLE 2.7 Earnings regressions Field

Constant

Age

Age2

Age4

Master's

PhD

Women

Education

50,512 (5.56) 45,876 (1.7) 40,915 (3.2) -17,497 (-1.4) -28,108 (-1.8) 22,023 (1.0) -16,213 (-.86) 1,132 (1.3) -10,745 (-.30) 759 (.04)

-3,121 (-5.2) -2,752 (-1.6) -2,498 (-3.0) 1,820 (2.1) 2,315 (2.2) -658 (-.47) 1,685 (1.4) -1,540 (-1.0) 2,779 (1.2) 679 (.50)

95.5 (9.0) 72.8 (2.30) 77.3 (5.2) 7.0 (.45) 6.2 (.33) 40.5 (1.6) 13.3 (.60) 55.2 (2.0) -29.8 (-.71) 26.1 (1.08)

-.014 (-15.6) -.009 (-3.3) -.011 (-8.5) -.006 (-4.1) -.007 (-4.4) -.007 (-3.4) -.007 (-3.6) -.009 (-3.7) .0003 (.096) -.007 (-3.5)

10,375 (19.0) 2,673 (1.8) 3,449 (4.7) 4,824 (5.9) 12,611 (13.9) 1,416 (1.1) 851 (.78) 10,047 (5.0) 17,431 (9.2) 2,112 (1.7)

16,453 (9.5) 8,045 (1.8) 17,823 (12.6) 12,887 (8.1) 14,540 (3.6) 14,845 (8.5) 9,584 (4.5) 28,925 (3.2) 19,135 (7.6) 11,493 (7.5)

-10,070 (-22.4) -7,642 (-6.4) -8,910 (-14.5) -18,135 (-29.3) -15,686 (-20.3) -13,821 (-13.7) -18,823 (-13.7) -1,868 (-.73) -19,276 (-11.6) -13,974 (-13.7)

Fine art Humanities Social science Commerce Agriculture/biology Engineer Nursing Other health-related Math/physical science

Source: See text. Note: R2 ranged from .08 to .22. T-statistics in parentheses.

Education and Technological Revolutions 69

in those fields the master's adds very little to the earnings of an undergraduate, but the PhD provides a significant return. Nursing shows a significant return to both the master's and the PhD, while other healthrelated fields, which at the graduate level includes many health administrators, exhibits a pattern like commerce - the master's degree substantially raises earnings, while the PhD adds little more. The plausibility of these results gives credence to the regression analysis. In estimating the rates of return to graduate degrees, the coefficients of the master's and PhD variables in Table 2.7 were used as measures of the increased income from the degree.6 They were projected into the future to reflect inflation and economic growth. To compute the rate of return, the projected benefits must be set against the costs, namely the foregone earnings of the student, the costs of books and supplies, and the government's cost of providing the education. The cost of books and supplies was taken to be $1,000 per year. For students in undergraduate programs, the foregone earnings were calculated from the average earnings of someone with a high school diploma of the corresponding age, while the calculations for graduate students used the corresponding earnings of someone with a bachelor's degree. It was assumed that undergraduates worked in the summer and gave up two-thirds of a year's earnings. The same assumption was made for students in two-year master's programs. In computing the cost of a one-year master's program, it was assumed that the student did not work and gave up a full year's earnings. PhD students completing their two years of course work were assumed to work in the summer and hence to give up two-thirds of a year's earnings. When writing their theses, they were assumed to work more and to give up half a year's earnings. The cost of university programs is determined in a series of steps. The total operating costs of Canadian universities in 1995/96 were $9,824,237,000. Interest and depreciation on the buildings and equipment were $1,136,968,000.7 The total cost of university activities was, therefore, $10,961,250,000. These activities included research as well as teaching. Several approaches suggest that about two-thirds of university costs are chargeable to teaching. One is the Hettich (1971) formula, derived from the accounts of Canadian universities, and it implies that teaching costs were $7,462,259,000 in 1995/96 (Dickson, Milne, and Murrell 1996). The total teaching costs are allocated among programs in proportion to enrolment weighted by relative cost. Statistics Canada tabulates full-

70 Robert C. Allen

and part-time enrolment by field of study and degree. Those enrolment figures are converted to full-time equivalents (fte's) on the assumption that a part-time student equals one-third of a full-time student. Various schemes are available showing the relative cost of programs. Typically, if a first-year arts student is rated at 1.0, then a thirdor fourth-year arts, commerce, or education student is 1.5, and a science or health student is rated at 2.0. Graduate students cost more. A master's student in the arts is rated at 3.0 and in the sciences and health at 4.0. Medical students are typically rated at 5.0 or 6.0. I have rated PhD students at the same level as master's students since they often take the same classes, but PhD students have been rated as high as 6.0. These ratings are derived from the Ontario operating grants formula and ostensibly reflect program costs, although the politics of funding may also have played a role in their determination. These program weights imply that there were 1,185,516 weighted full-time equivalent (wfte) students in Canadian universities in 1995/96, so the cost per wfte was $6,295. The cost of a degree can then be computed from the number of wfte's involved. Thus, a four-year undergraduate science degree costs $50,360 (8 x $6295) since it consists of two wfte's per year for four years. Juxtaposing the costs and benefits of university programs implies social rates of return. These are shown in Tables 2.8 and 2.9. The rate of return on long-term government bonds is about 5 per cent, so that is the threshold that educational programs must meet in order to be a good investment. Degrees in the humanities, social sciences, and education surpass that threshold. I begin with the bachelor's degree. In the case of women, the rates of return to SSHRC (Social Sciences and Humanities Research Council of Canada) fields were very close to the average for all women with bachelor's degrees (16.8%). The humanities, social sciences, and education formed part of a group that also included engineering, math and physical sciences, and nursing. Rates of return for all these fields were very similar. The rates of return on these programs are very much above the 5 per cent threshold. The rates of return to undergraduate degrees are more varied for men than they are for women. Men in the humanities (7.6%) and education (13.6%) easily surpass the 5 per cent threshold, while the social sciences (18%) are one of the highest return fields, realizing almost as much profit as engineering (19.5%) and commerce (21.1%). Undergraduate degrees in the humanities, social sciences, and education are good investments for Canada.

Education and Technological Revolutions 71 TABLE 2.8 Social rates of return to university programs tor women Field

Bachelor's

1 -year master's

2-year master's

PI"

Education Fine arts Humanities Social sciences Commerce Agriculture/biology Engineering Other health-related Nursing Math/physical science

18.1 8.8 15.6 17.3 22.0 13.2 16.4 20.8 17.2 17.8

33.8 10.6 13.1 17.3 40.5 5.0 2.5 25.0 28.1 7.4

19.2 6.2 7.9 10.5 22.5 2.0 -0.1 14.4 16.1 3.9

11.3 6.4 9.3 9.4 10.3 9.1 6.3 11.9 14.7 7.4

Source: See text.

TABLE 2.9 Social rates of return to university programs for men Field

Bachelor's

1-year master's

2-year master's

PhD

Education Fine arts Humanities Social sciences Commerce Agriculture/biology Engineering Other health-related Nursing Math/physical science

13.6 1.5 7.6 18.0 21.1 11.0 19.5 17.7 1.1 18.0

31.3 9.8 12.3 16.1 37.4 4.6 2.1 75.0 26.4 6.9

18.3 5.9 7.5 9.9 21.3 1.8 -0.3 37.8 15.4 3.6

10.7 6.0 12.2 8.9 9.8 8.7 5.9 11.4 14.1 7.0

-

Source: See text.

The same is true of graduate programs. Master's degrees are analysed as requiring either one or two years since programs are organized in both ways. Rates of return are always lower under the two-year format since the income gain is the same in both cases, but the costs are higher if the degree is done in two years. Professional programs in commerce, education, nursing, and other health-related fields give the highest rates of return. Next come the social sciences, humanities, and fine arts, which yield rates of return above the 5 per cent threshold.

72 Robert C.Allen TABLE 2.10 Employment in Canada, by level of education, 1991-1996

Level

Employment 1991

Employment 1996

High school non-completer High school graduate Trade certificate College diploma University less than bachelor Bachelor's degree University greater than bachelor Medicine Master's PhD

3,702,500 .3,339,400 1,703,467 1,916,267 309,967 1,302,667 214,900 77,100 326,033 65,000

3,192,732 3,325,536 1,646,028 2,372,688 349,092 1,583,460 246,888 85,572 402,660 84,024

-509,768 -13,864 -57,439 456,421 39,125 280,793 31,988 8,472 76,627 19,024

-13.8 -0.4 -3.4 23.8 12.6 21.6 14.9 11.0 23.5 29.3

Total

12,957,300

13,288,680

331,380

2.6

Change

Per cent change

Source: Census of Canada, 1991 and 1996, microdata files. Note: Total for 1991 excludes 6233 individuals coded as 'degree not available.' All figures exclude people in school in previous year. 'University greater than bachelor1 refers to graduate certificate greater than a bachelor's

Finally, master's degrees in the sciences and engineering give the lowest rates of return. Engineering is never a profitable investment, nor are the sciences if the program is done in a two-year format. PhD programs are profitable for both men and women in all fields of study. As with master's degrees, the return to PhD's in health-related fields and education is always high. The humanities and social sciences are on a par with commerce and generate returns that exceed those in engineering, mathematics, and the physical sciences. The returns to PhDs in agriculture and biology are not much better. Highbrow techniks attribute more importance to these fields than the market does, judging by the rates of return to graduate programs in science and engineering.8 Employment Changes, 1991-1996 The 1996 census provides a snapshot of the labour market taken at one time. While that snapshot has much to teach, it cannot answer all our questions. What we really want to know is how the labour market is changing. Is the demand growing rapidly for people with technical skills and slowly, or even declining, for graduates in education, the

Education and Technological Revolutions 73 TABLE 2.11 Employment of university graduates in Canada, by field of study, 1991-1996

Field of Study Education Fine arts Humanities Social sciences Commerce Agriculture/biology Engineering Nursing Other health-related Math/physical science Total

Employment 1991

Employment 1996

Change

Per cent change

382,467 48,500 238,533 363,467 304.966 93,167 210,966 50,733 138,067 154,833

451,908 59,472 272,304 465,912 385,056 113,508 252,648 59,112 155,160 187,524

69,441 10,972 33,771 102,445 80,090 20,341 41,682 8,379 17,093 32,691

18.2 22.6 14.2 28.2 26.3 21.8 19.8 16.5 12.4 21.1

1,985,700

2,402,604

416,904

21.0

Source: Census of Canada, 1991 and 1996, microdata files. Note: All figures exclude people in school in previous year. Commerce includes secretarial science, engineering includes engineering technician, and humanities includes 'other.' Secretarial science, engineering technician, and 'other' contained tiny numbers of people. University graduates include those holding a bachelor's degree; graduate certificate greater than a bachelor's; medical, dental, etc. degree; master's degree; and PhD.

humanities, and social sciences? Or does the emerging knowledgebased economy require a broad range of education and training? To answer these questions, we need to compare the labour market at two dates, and I do so using the 1991 and 1996 censuses. Two snapshots can offer the perspective of film, and so we now move from a static to a dynamic analysis of education and the labour market. A first step in investigating how the demand for labour has been changing is to examine trends in employment. Table 2.10 shows employment in Canada for various levels of educational attainment according to the 1991 and 1996 censuses. The table also shows the increase in the number of jobs held by people at each educational level and the percentage change of that increase. Table 2.11 breaks down the evidence of university graduates by field of study. Several findings are immediately apparent from the two tables. First, there was a large employment drop for people who had not finished high school. All commentators agree that completing high

74 Robert C. Allen

school is essential for success in the new knowledge-based economy, and Table 2.10 bears that out. Second, employment also declined, but by a smaller extent, for high school graduates and for those holding trade certificates. The decline in employment for those holding trade certificates contradicts lowbrow techism, which maintains that the new economy requires more of those technical skills. This result parallels the earlier findings about the high unemployment and relatively low earnings of people with trade certificates. One-year trade certificates are insufficient preparation for the economy of the twenty-first century. Third, the expansion in employment for those holding college diplomas made the biggest contribution to total employment growth and achieved one of the highest percentage increases between 1991 and 1996. Lowbrow techism is on firmer ground when it maintains that the emerging knowledge-based economy requires the skills taught in twoyear college programs, for employment of the completers of such programs has been increasing very rapidly. However, about half of these completers are women and most of them complete nursing, commerce, or clerical programs rather than technical programs. Among the men, about half complete programs in applied engineering. These technical graduates are the people that lowbrow techniks point to, but they are a minority of college graduates. The experience of the colleges suggests that the emerging new economy demands much more than technical skills. Fourth, university graduates made almost as big a contribution to employment growth as college graduates. The expansion of employment for those receiving university degrees or certificates was 456,029 almost identical to the growth in employment of college graduates. Indeed, the employment category with the largest percentage increase was PhDs. Employment expansion was very strong for those receiving master's degrees and bachelor's degrees. Fifth, the expansion of employment for university graduates extended across all fields. Table 2.11, which focuses on those receiving a bachelor's degree or higher, shows that the field with the largest employment growth - both absolutely and on a percentage basis - was social sciences. It was followed closely by commerce. Engineering and the natural sciences, the fields deemed most important by highbrow techniks, were in the middle of the pack. Once again, the evidence seems to contradict the technik vision of the emerging knowledge-based economy.

Education and Technological Revolutions 75

But the matter cannot be left there, for there is a major problem in interpreting the Canadian employment trends. The question is: did the employment changes reflect changes in the demand for labour or changes in the supply? In a country like Canada, where immigration is small relative to the population, the educational qualifications of the labour force are mainly determined by the output of the education system. Indeed, the technik critique of Canadian education is that the system is not responsive to the needs of the economy and so is producing graduates with a skill mix that is inappropriate for the future. Highbrow techniks, for instance, would say that the low contribution of engineering to employment growth (Table 2.11) shows what is wrong with Canadian universities, not how the economy is evolving. Our problem, therefore, is to determine whether changes in the educational credentials of the workforce reflect a non-economic structure of programs or whether they reflect changes in the demand for labour. I shall approach this problem in steps. From Employment Change to Labour Demand The history of wages and salaries throws light on the growth of demand for labour. The illumination does not occur, however, without a theory of the labour market. I begin with the theory of competitive markets, which is commonly used in economics. According to this theory, wages and employment are determined by supply and demand. In that case, wages will rise if demand grows faster than supply, and wages will fall when supply grows faster than demand. Constant wages indicate that supply and demand are changing in step with each other. If the technik critique is right, then the wages of people with trade certificates and college diplomas should be rising with respect to the salaries of humanities and social science graduates (lowbrow techism). Likewise, the salaries of engineers and scientists should be rising with respect to the salaries of arts graduates (highbrow techism). Findings of this sort would support the technik critique that Canadian colleges and universities are producing a mix of graduates that do not meet the needs of the economy. Such findings would also support the policy recommendation that in order to meet those needs, resources should be shifted from the humanities and social sciences towards technical fields. To investigate changes in earnings, tables like 2.5 and 2.6 in this chapter, were compiled for 1990 from the 1991 census. The consumer

76 Robert C. Allen

price index was used to convert the 1990 wages and salaries into 1995 dollars. Comparison of the 1990 and 1995 tables makes two points. First, there was a decline in real earnings for virtually all educational and age groups, a decline presumably caused by the high unemployment of the period. Second, there was very little change in relative wages between the two periods. To establish this finding, all earnings for each age level were divided by the earnings of high school graduates of that age. Inspection of the tables suggests that there were no changes in relative wages. To pin the matter down, the relative earnings for men in 1995 are divided by the corresponding relative earnings for 1990, and similarly for women. If these ratios equal one, then the wage structure was unchanged. That was, indeed, the case. The conclusion is that the Canadian wage structure was stable. If wages and employment are determined by supply and demand, a stable wage structure means that demand and supply are growing at the same rate. The increases in employment for humanities and social science graduates, in other words, did not occur because there was an oversupply. To repeat, that situation would be indicated if the salaries of humanities and social science graduates were falling relative to other workers, so employers were hiring the arts graduates for lowerskilled work. But their salaries were not falling. The demand for their labour was growing as rapidly as the supply. This interpretation of the data is predicated on the competitive model of the labour market, a theory which presumes that wages move up or down to equate demand and supply. An alternative view of the labour market emphasizes the rigidity of wages; that is, suppose that the wage and salary structure was fixed by collective agreements and bureaucratic policies, so that wages would not adjust if supply and demand were out of balance. Indeed, one might interpret the remarkable constancy of relative earnings as evidence in favour of that view. In that case, the increase in the supply of humanities and social science graduates would result in their unemployment unless the supply growth was matched by demand growth. But in the second section of this chapter, we reviewed evidence on unemployment rates that showed that they were low for all university graduates including those in education, the humanities, and social sciences. Unemployment was higher for people with less education. If we take the fixed wage model of the labour market seriously, the conclusion is that the demand and supply of university graduates in all fields were growing in tandem,

Education and Technological Revolutions 77

while supply growth outstripped demand growth for people with less education. The imbalance was greatest at low educational levels. This conclusion does not support techism but does support the conclusion that the emerging knowledge-based economy requires university graduates in all fields. British Columbia: A Case Study Further evidence that rapid economic growth in the 1990s increased the demand for graduates in all fields is provided by the history of British Columbia from 1991 to 1996. While the economy of central Canada was depressed, BC's economy was booming. Indeed, the province accounted for two-thirds of the job growth in Canada as a whole. The study of growth at the regional level is particularly revealing of trends in labour demand. In Canada as a whole, the educational qualifications of the labour force are mainly determined by the size and character of the education system, so one cannot infer changes in the demand for labour from changes in employment without also examining the stability of the wage structure and unemployment rates as we have done. In a regional economy like that of British Columbia, an increase in labour demand creates jobs that attract migrants from elsewhere in Canada. As a result, the employment pattern is not determined by the provincial educational system but reflects changes in labour demand. So by looking at employment changes in BC in the 1990s, we get another view of how rapid economic growth affects the demand for labour. The important conclusion is that we observe increases in the demand for graduates in the humanities and social sciences that were every bit as big as those for graduates in engineering or commerce. Tables 2.12 and 2.13 summarize the changes in employment in British Columbia from 1991 to 1996. As noted, the total grew much more rapidly in BC than in Canada as a whole. That has some bearing on the absolute magnitude of the changes, but the relative pattern is similar to that of the whole country: high school dropouts had the worst employment experience in that their employment level fell even as the whole economy expanded. The employment of high school graduates and of completers of trades and one-year technical courses grew by a small amount that was considerably less than the growth in total employment. Once again, the employment record of the completers of oneyear technical training programs appears weak. Most of the employment growth in the province was accounted for

78 Robert C. Allen TABLE 2.12 Employment in BC, by level of education, 1991-1996

Level High school non-completer High school graduate Trade certificate College diploma University less than bachelor Bachelor's degree University greater than bachelor Medicine Master's PhD Total

Employment 1991

Employment 1996

407,433 423,133 226,400 238,400 37,267 146,000 22,633 9,533 38,167 7,367

403,812 440,532 242,352 318,636 49,320 204,660 29,880 12,636 54,972 12,240

-3,621 17,399 15,952 80,236 12,053 58,660 7,247 3,103 16,805 4,873

-0.9 4.1 7.0 33.7 32.3 40.2 32.0 32.6 44.0 66.1

1,556,333

1,769,040

212,707

13.7

Change

Per cent change

Source: Census of Canada, 1991 and 1996, microdata files. Note: Total for 1991 excludes 867 individuals coded as 'degree not available.' All figures exclude people in school in previous year.

by the growth in employment of college and university graduates. Indeed, workers with university degrees, certificates, or diplomas accounted for 48 per cent of the growth in employment. PhDs and master's degrees experienced the most rapid growth of all. This is, indeed, the knowledge-based economy. What sort of knowledge was in demand? Social sciences had the highest rate of employment growth. Physical sciences came in second, and engineering was third - an unexpected result for techniks. Equally surprising was the strong showing of humanities, which came in fourth, beating out commerce at five. Health, nursing, education, fine arts, and the biological sciences followed. From the analytical point of view, what is most significant about these increases is that they represent the growth in demand rather than the growth of supply. The increase in employment in all of these fields was considerably larger than the number of graduates of British Columbia's universities. The shortfall was met by international and interprovincial migration. Employment growth in BC was demand determined and not dictated by the output of the province's universities. The history of British Columbia in the 1990s projects a vision of Canada's future that is broader than that imagined by techniks. They are

Education and Technological Revolutions 79 TABLE 2.13 Employment of university graduates in BC, by field of study, 1991-1996

Field of study Education Fine arts Humanities Social sciences Commerce Agriculture/biology Engineering Nursing Other health-related Math/physical science Total

Employment 1991

Employment 1996

Change

Per cem change

46,533 6,867 25,634 38,900 30,200 12,800 23,900 6,200 17,300 16,233

59,508 9,036 36,360 60,660 42,552 17,064 34,416 8,352 22,428 24,012

12,975 2,169 10,726 21,760 12,352 4,264 10,516 2,152 5,128 7,779

27.95 31.6 41.8 55.9 40.9 33.3 44.0 34.7 29.6 47.9

224,567

314,388

89,821

40.0

Source: Census of Canada, 1991 and 1996, microdata files. Note: All figures exclude people in school in previous year. Commerce includes secretarial science, engineering includes engineering technician, and humanities includes 'other.' Secretarial science, engineering technician, and 'other' contained tiny numbers of people. University graduates include those holding a bachelor's degree; graduate certificate greater than a bachelor; medical, dental (etc), degree; master's degree; and PhD.

right that the future economy will be knowledge-based. Even with rapid economic growth there will be little if any employment growth for high school dropouts or for those with only a high school diploma. In contrast to the beliefs of lowbrow techniks, this bleak future extends to graduates of one-year technical and trades programs, who face weak demand growth in the emerging economy. There will be a significant demand for the graduates of two-year college programs, and, to that degree, lowbrow techniks are right. But most of the growth in labour demand will be for university graduates. In contrast to highbrow techniks, the demand will not be limited to engineering or even to other professionally oriented programs. While demand growth will certainly be strong for the graduates of those programs, it will be at least as strong for graduates in the humanities and social sciences. One of the outstanding features of the knowledge-based economy will be the breadth of advanced education and skills it requires.

80 Robert C. Allen

The Computer Revolution and the Demand for Skills Techism concentrates on the production of new technology rather than on it use. To understand the role that technological revolutions play in economic development, we must distinguish these questions. Making this distinction explains why the growth in the demand for educated workers is not limited to technical subjects. Consider, for instance, a genetically modified variety of wheat that increases the yield per hectare. The production of this seed contributes to economic growth by increasing the output of the biotechnology industry. However, the aggregate impact of that output growth may be minor since (a) the share of the biotechnology industry in the Canadian economy is very small; and (b) the output of conventional seed will decline and that fall in output offsets the rise in output of hightech seed. However, the contribution of the utilization of the improved seed could be very large since it will increase agricultural output, and agriculture is a larger fraction of the Canadian economy than is the biotechnology industry. It is likely that many high-tech products contribute more to economic growth by their utilization than by their production. Computers are an example. Even in the United States with its large computer industry, the contribution to GDP growth from the manufacture of computers is probably less than the contribution to growth from using computers in manufacturing, commerce, and administration. This is certainly true of Canada where there is no significant production of computers, but where their utilization is widespread. Indeed, the Canadian economy generates a per capita income that is higher than that in Europe or Japan and almost as high as that in the United States without producing many of the high-tech products of the twentieth century. Canada can do this by using those products, and it gets them through international trade. The possibility of importing high-tech products must be assumed away in order for techism to make sense. In the absence of trade, hightech products cannot be utilized in Canada unless they are produced here, so the contribution of high-tech production to growth includes the utilization contribution as well as the production contribution. This is the implicit assumption of techism, and it is a necessary assumption for that theory. In an era of globalization, it is also an odd and an unwarranted assumption. The widespread utilization of computers in Canada explains why the demand for arts and social sciences graduates has been growing rap-

Education and Technological Revolutions 81

idly. Information technology has revolutionized business organization. The old-style business was organized hierarchically. Legions of clerks and middle managers processed information by hand. This was funneled up the hierarchy to the top where decisions were made. Computers and information technology have rendered this form of organization uncompetitive. The fall in the price of hardware and software has led to the widespread adoption of computers in business and government. Shifting information processing to computers has led to a fall in the demand for clerks and low-level managers, which is manifest in the employment data as falling or slow growth in employment of high school graduates and non-completers. The adoption of computer-based information technology systems has also cut the cost of information. As computerized databases are enlarged, information can be brought to bear on many more business issues. When supermarkets, for instance, adopt checkout scanners and computers to track inventory, they also make it easy to study the effects of advertising and promotional pricing on sales. In the simplest terms, the number of correlations that can be computed increases exponentially, while databases expand linearly. The falling cost of information has made the hierarchical organization of businesses inefficient. Senior management no longer has the time to deal with all the information that can be cheaply and usefully produced. As Herbert Simon (1973), the Nobel Prize-winning economist, noted: 'The scarce resource is not information, it is the processing capacity to attend to information. Attention is the chief bottleneck in organizational activity, and the bottleneck becomes narrower and narrower as we move to the tops of organizations/ The result of the computer revolution has been threefold. First, there is an increased demand for people who can understand the information generated by computer systems, analyse it, relate it to the world, and act on it. These kinds of general intellectual abilities are the sorts that are developed in humanities and social science programs. It is for this reason that the utilization of computers has led to the growth in demand for people with those degrees. Second, organizational structures have become flatter. Instead of multiple layers of managers doing routine information processing, there are fewer layers in the hierarchy, and the employees are charged with analysing and acting on the greater volume of processed information available. There is a much greater demand for people who can make critical and independent judgments, and those capacities are cultivated in social science and humanities programs.

82 Robert C. Allen

Third, the new-style middle managers need greater interpersonal and communication skills both to deal with customers of the firm and to work together in self-directing teams. Humanities and social science programs can make more effective employees in this regard as well. The computer has affected many sectors of the economy, but not all. The affected sectors include the government and those sectors where the employment of university graduates has expanded the most. The lay-offs of clerks and old-style middle managers in manufacturing have been widely reported, since firms in this sector are among the most prominent in the economy. The shift to computer-based, less hierarchical management structures in manufacturing has been widely noted for the same reason. Wholesaling and retailing have also been widely affected. Computerized inventory control has led to the reorganization of many companies and increased possibilities of information analysis, as already noted. New companies have also emerged in this sector based on sales over the Internet. Firms like Amazon.com, without inventories of their own and with highly educated workforces, are setting the pace for the new millennium. Many of the economies of the new management style have been realized by expansion of the business services sector. In accounting firms, for instance, the new-style manager is responsible for a spreadsheet dealing with a particular function such as taxation or inventory control. The spreadsheet does the calculating that used to require many clerks and managers, and it embodies the tax or inventory model. The operator of the spreadsheet has to know the capabilities of the model and how to apply them to the needs of the client. Thus, the new-style managers need interpersonal and communications skills as well as the ability to understand a model and tailor it to a particular application. These skills become critical wherever management has been revolutionized by information technology. Manufacturing, wholesale and retail trade, finance, business services, and the government are the sectors where this revolution has occurred. Aside from the government, where cutbacks have reduced overall employment, these are also the sectors where the demand for university graduates in general - and humanities and social science graduates in particular - have expanded. Computers have led to a technological revolution that has contributed to economic growth in two ways. The production of computers has directly increased GDP, notably in the United States. More generally, the utilization of computers in many sectors of the economy is contributing to growth by raising output per worker. While the pro-

Education and Technological Revolutions 83 duction of computers and software requires people with many technical skills, the utilization of computers requires the general skills taught in social science and humanities programs. The utilization of computers affects skills requirements profoundly. The issue is not whether an employee knows how to operate an Excel spreadsheet, so much as it is whether the employee can apply a model to a problem, deal effectively with customers and other members of a management team, write and speak clearly, and make informed and independent judgments. The reason these skills are in high demand is because business organization has been revolutionized to take advantage of cheap information. As noted earlier in this chapter, that revolution increases the demand for graduates in both social science and the humanities.

Notes 1 The fields are defined similarly to Riddell's work in this volume, although he combines some that are distinguished in this chapter. 2 Riddell, and Hum and Simpson, analyse education with the same human capital framework in their chapters in this volume. 3 One might reasonably ask whether the income difference does not also reflect an ability difference, and consequently whether the income difference overstates the productivity-raising effects of education. Labour economists and other social scientists have explored this issue at length and the consensus is that the rates of return computed without correcting for ability are accurate. Social factors play a major role in determining university attendance. Moreover, 'ability' is multifaceted, and the ability to do well in academic programs is not the same as the ability to earn a high income. Allen (1998b) reviews the literature and presents new Canadian evidence, as does Riddell in his chapter in this volume. 4 Hum and Simpson also analyse training in a regression framework, in this volume, that shows how it can sort out the effects of multiple explanatory variables. 5 Age cubed was entered in all of the regressions but was never significant and is not included in the final models. Models were also estimated using Mincerian potential experience instead of age (i.e., the length of time someone was an adult but not in school, calculated as age minus years of schooling minus five). In other models, potential experience was also multiplied by the master's and PhD dummy variables to allow the age-earnings profiles to vary by degree. These alterations did not lead to systematically different results.

84 Robert C. Allen 6 More complicated specifications were also estimated in which sex was multiplied by the variables representing the master's and PhD degrees to see if the returns to those programs varied between men and women. Statistically significant effects were detected at the doctoral level for graduates in the humanities and at the master's level in other health-related fields. There may also have been effects for fine arts graduates, but they were hard to pin down due to the small number of people in the sample with graduate degrees in this field. These estimated differences in the returns to graduate degrees were incorporated into the rates of return but with reservations, since these effects may reflect old patterns of discrimination against women that may not persist in the future. 7 Interest and depreciation are computed from the capital stock of the universities. Statistics Canada uses four different measures of the capital stock. The capital costs used here are computed from the delayed depreciation stock ($16,242.4 million), which is one of the highest and appears to correspond best to the life expectation of university assets. Interest was computed at 5% and depreciation at 2%. 8 The rates of return reported in this section are higher than the 'before tax' rates of return reported by Riddell in this volume. His 'before tax' returns are comparable to social rates of return. The differences are due to (1) the different assumptions made in projecting cross-sectional earnings into the future, (2) the closer attention paid to the costs of education in this chapter, and (3) Riddell's imputation of some of the returns to 'sheepskin' effects. Despite the differences in procedure, both sets of estimates support similar policy conclusions. References Allen, R.C. 1998a. Standing room only: The case for expanding B.C.'s universities. Vancouver: Canadian Centre for Policy Alternatives. - 1998b. Paid in full: Who pays for university education in B.C? Vancouver: Canadian Centre for Policy Alternatives. - 1998c. The employability of university graduates in the humanities, social sciences, and education: Recent statistical evidence. Ottawa: Social Sciences and Humanities Research Council of Canada. Alsalam, N., and R. Condy. 1995. The rate of return to education: A proposal for an indicator. In Education and Employment, formation et emploi. Paris: Organization for Economic Cooperation and Development. British Columbia Labour Force Development Board (BCLFDB). 1995. Training for what? Victoria: BCLFDB.

Education and Technological Revolutions 85 Dickson, V., WJ. Milne, and D. Murrell. 1996. Who should pay for university education? Canadian Public Policy 22: 315-29. Hettich, W. 1971. Expenditures, output and productivity in Canadian university education. Economic Council of Canada, special study 14. OECD. (Organization for Economic Co-operation and Development). 1977. Education at a glance: OECD indicators, 1997. Paris: OECD. Simon, H. 1973. Applying information technology to organizational design. Public Administration Review 33: 268-78.

This page intentionally left blank

PART 2 Achieving Equity: Three Analyses of Outcomes

The goal of a good deal of educational and economic policy has been the country's aggregate growth and development. But equity, and the social cohesion associated with it, is equally important as a social goal and as an outcome of education and training policy. Equal educational opportunity is a core value in educational institutions, and the impact of human capital development on equality has been a key part of policy discussion in liberal democratic societies like Canada. The easy assumption that more meritocracy and increased access to educational opportunities will bring about more social equality has been challenged theoretically, and by empirical research of many kinds. The three chapters in this section each use different methodologies to focus on how education and training policies affect equity. Simpson and Hum use the traditional tools of economics, analysing national surveys of adult education and training to see what factors predict participation in post-school training. They conclude that adult education is not a force for equity. Disadvantaged groups have less access to training than better-educated workers with professional and managerial jobs in large firms, so training schemes need to pay explicit attention to questions of access. Sloane-Seale, Wallace, and Levin examine a Manitoba policy that does this - providing access to university education for Aboriginal students. By focusing on the students and their experience, the chapter explores the personal and institutional factors that make a difference to whether the program is successful, while pointing out that the problems these youth face go so far beyond education that the struggle remains overwhelming. Butterwick provides a critique of the way notions of human capital are embedded in welfare policies that use training as the way out of

88 Part 2: Achieving Equity

poverty. She also starts from the experience of learners - low-income mothers - and shows how policy affects them and what education means to them. Her conclusion - that politically engaged welfare mothers experienced new regulations as yet another imposition of arbitrary and contradictory requirements that made their lives more difficult - is a cautionary tale for policy. This section illustrates the different voices involved in different kinds of research. Despite their different tones and approaches, all agree that equity matters and that it can be increased as well as decreased by the provision of education. They agree that education helps produce equality if it is carefully structured with the needs of disadvantaged learners in mind; and that the details of how this can be done depends on continual research on and feedback from particular populations, in particular places, and at particular times.

3 What Are Canadians Doing after School? An Analysis of Post-school Training Activity DEREK PJ. HUM AND WAYNE SIMPSON

There is much evidence that those who receive post-school training (PST) do better in the labour market in terms of increased earnings and reduced unemployment. (See, for example, the chapters by Allen and Riddell in this volume.) If PST has such a beneficial effect for individuals, what factors determine which individuals receive it, and which do not? The motivation to seek training involves both economic and noneconomic factors. Some individuals participate in training to gain promotion, higher pay, or perhaps because of a job/career change due to lay-off or termination. Others might obtain training due to family circumstances such as marital dissolution, childbirth, relocation of a spouse, or perhaps a disability onset resulting from accident or functional limitation due to advancing age. From the standpoint of social equity and labour markets, we might ask: do all Canadians have comparable opportunities to participate in PST? And, if not, which groups receive fewer opportunities to train, or less training? What are the personal characteristics of those who receive different amounts of training? What are their employment circumstances? In sum, what are the factors that determine whether an individual undertakes training, for how long, and under what employment circumstances? This essay examines participation by Canadians in training activity beyond formal schooling. Since an important motivation for PST is economic, we begin with an explanation of the core factors for accumulating human capital for the labour market. This is especially necessary when interpreting issues of equity, since observed unequal outcomes may reflect individual differences in experience, skill, or ability rather

90 Derek PJ. Hum and Wayne Simpson

than inequality of access due to discrimination, social barriers, and the like. For this reason, we first take account of those factors giving rise to productivity-related patterns of training. For example, success in past training may be a screening device by employers to give further training. This leads to the result that many who acquire advanced training may already possess extensive training, possibly exacerbating inequality of outcomes. It is economically rational (and safe?) for firms to offer training to operate large heavy-transport trucks only after individuals have learned to drive light vehicles first. It is important for policy response to distinguish between unequal training outcomes resulting from individuals failing to meet the eligibility requirements for training, and unequal training outcomes due to inequitable access or opportunity to take training. We assess the importance of a core economic model to explain PST activity, and examine the influence of other personal and job characteristics as well. We employ multivariate statistical techniques to estimate this model using the 1992,1994, and 1998 Adult Education and Training Surveys (AETS).1 Factors Affecting Education and Training Patterns There have been many empirical investigations of training behaviour. Some studies concentrate on the individuals receiving training (e.g., Lillard and Tan 1986; Altonji and Spletzer 1991; Booth 1991; Green 1991 and 1993; Miller 1994; Jennings 1996; Kapsalis 1996; de Broucker 1997; Wooden and VandenHeuval 1997; Green and Zanchi 1997; Shields 1998). Others examine the training providers, particularly employers (e.g., Barron, Black, and Loewenstein 1987; Holtmann and Idson 1991; Lynch and Black 1998; Holzer and Reaser 1999). Although these studies differ widely with respect to the models and approaches used, the distinction between individual and employer aspects is extremely important for equity issues. Are certain individuals receiving less training because of personal traits such as lack of preparation? Or do certain groups receive less because of employer-related barriers or discrimination? Many studies include gender and family status variables in studying individual participation in adult training. Almost all note that the determinants of training differ between men and women (more recently, Green 1991 and 1993; Miller 1994; Wooden and VandenHeuval 1997; Green and Zanchi 1997). Greenhalgh and Stewart (1987) also introduce the role of marital status. They find that women in the

What Are Canadians Doing after School? 91

United Kingdom receive less training than men early in their careers, and that married women continue to receive less training throughout their careers. Married women at lower occupational positions appear to receive more training, while men and single women receive more training at higher occupational positions, suggesting some interaction between occupational status, marital status, and the incidence of training. Miller (1994) finds a corroborating marital status effect for Australia, but Holtmann and Idson (1991) find no significant marital status effect in the United States. Another potentially important factor is the presence of young children, or their expected presence. Duncan and Hoffman (1979) find that the expectation of having children reduces the chance of training for women. Greenhalgh and Stewart (1987) find that the presence of young children significantly reduces the probability of training for men and women in the United States. On the other hand, Miller (1994) finds that very young children (0-2 years) only reduce women's likelihood of involvement in training. An individual's occupation is another prominent variable in previous studies. Holtmann and Idson (1991) include a dummy variable for the so-called 'white collar' occupations and find it to be significant. Green (1993) and Miller (1994) include variables to capture the influence of high-status occupations on training. Altonji and Spletzer (1991), on the other hand, examine the relationship between high school curriculum, occupation, and adult training. They find that high school curriculum has no significant effect, but that training participation increases with the verbal, math, and clerical skill requirements of an occupation and decreases with manual skill requirements.2 Occupations with large amounts of specific vocational preparation have a lower probability of training. The nature of the employer also affects the patterns of training participation, particularly on-the-job training (OJT). Several studies include firm size as a determinant of OJT. Holtmann and Idson (1991) find that larger firms offer more training programs to workers with risky characteristics, such as low education and short job tenure. Barron, Black, and Loewenstein (1987) find that larger employers provide more training to new workers, interpreting their results as support for the view that these decisions are driven by the costs of monitoring workers, which are higher for larger firms. Other studies that find a positive relationship between training and firm size include Simpson (1984); Booth (1991), for women only; Green (1993); and Miller (1994).

92 Derek P.J. Hum and Wayne Simpson

Training opportunities may also differ for workers in union and non-union jobs. Mincer (1983, 1984) argues that compressed wage schedules and the dominance of seniority rules for wage progression in unionized firms reduce incentives for general training and limit the supply of trained workers from other firms. This conjecture finds support in survey evidence on the incidence of training in the United States (e.g., Mincer 1983), although Simpson (1984) and Betcherman, Leckie, and McMullen (1997) find no significant training effect among unionized workers in Canada. Kennedy et al. (1994) find that unionization per se does not affect training in Australia, although a variable denoting union bargaining activity, or Voice/ appears to be positively correlated with training. Industrial characteristics may also affect training. Lillard and Tan (1986) find that workers in industries undergoing rapid technological change receive more OJT, particularly the more educated workers. Booth (1991), however, finds no significant effect of industrial categories, at least for women. More recently, Lynch and Black (1998) found that manufacturing firms with high capital-labour ratios, and nonmanufacturing firms with high research and development activity, conducted more training. Lillard and Tan (1986) also found that local employment conditions affect on-the-job training participation. Other studies have found training activity to vary among regions of a country; for example, Simpson (1984) and Betcherman, Leckie, and McMullen (1997) for Canada, and Booth (1991) for the United Kingdom. A Human Capital Core Model of Training The research literature suggests that many factors matter to training activity, such as the personal traits of individuals, their family circumstances, and the nature of their employment. These varying characteristics and circumstances of individuals raise important equity questions. Do some groups have less access to training than others? What are the barriers to participation? Can these barriers be corrected by policy changes? In short, do observed differences in participation among individuals reflect differences in economic circumstances, or do they reflect differences in opportunities due to the individual's gender, disability condition, immigration status, employment status, and so on? Since any investigation of the determinants of post-school training must take into account a wide variety of factors, a multivariate statistical approach is needed.

What Are Canadians Doing after School? 93

Economic factors certainly matter in determining the amount of training that is undertaken. Accordingly, we propose a framework that accounts for some 'core' economic factors and then add other relevant considerations to our model. The human capital model remains the preferred economic framework for studying the relationship between labour markets and training. Considerable evidence supports a link between work-related training, earnings growth, productivity growth, and job performance (e.g., Barron, Black, and Loewenstein 1989; Bartel 1995; Lynch 1992). Hence, who undertakes training, how much they train, and who provides this training, are important questions when examining labour market outcomes and equity issues. The human capital model of earnings over the life cycle is a good starting point for studying adult training. In this literature, new earning capacity depends strictly on human capital investment. Human capital investment depends, in turn, on previously accumulated human capital, since many training programs require that individuals have prerequisites prior to undertaking additional training. Human capital investment may often depend on hours worked (e.g., Weiss 1986, 605), not only because much training is delivered on the job, but also because employers often allocate training based upon the hours worked by their employees (part-time vs. full time, seniority status, experience, and so on). Economic theory suggests a pattern of human capital investment that declines over the lifetime of the typical individual. Blinder and Weiss (1976) specify a model of labour supply in which the typical individual passes through four phases: (1) schooling, (2) work and training, (3) work only, and (4) retirement. After the initial schooling period, which we ignore in this chapter, investment in post-school training declines during the working career and disappears at or near retirement. Education and training at any point in the life cycle depends upon past human capital investment, current age, and hours worked (e.g., Weiss 1986). This specification may not account for some training that is job-specific, however, if workers have more than one job in their career. In that case, the model must also account for job mobility as a source of job-specific training. This is due to the fact that on-the-job training declines not only over the worker's career, but also as time elapses within a particular job, since the incentives to invest are greater for both the worker and the firm at the beginning of the job. Hence, observed training participation would be expected to decline with tenure on the job. Moreover, Borjas (1975,1981) notes that the incentive to

94 Derek P.J. Hum and Wayne Simpson

invest will depend upon the expected duration of the current job. Since expected job duration is unobserved, he uses job tenure as a proxy for job duration. Since job tenure is likely to be reasonably associated with expected job duration, the effect of job tenure on training will no longer be unambiguously negative. When a worker begins a job, the prospective returns to investment are highest, but the possibility of the investment being lost is also highest. If the worker stays, returns are high; if the worker leaves, there is no return to the employer. If the worker stays with the firm, however, the prospect of increasing expected job duration will reinforce the incentive to invest early in a worker's career with the firm. As a result, the likelihood of training will initially rise and decline only later in the job.3 Thus, as a starting point, our core economic model includes job tenure, past human capital investment, age, and hours worked. This specification differs from those used in previous empirical work. While many studies include age and previous education, their treatment of past training, hours worked, and job tenure is uneven.4 Since our paper focuses on job-related adult training associated with those who are out of school, we exclude students from our sample, whether full time or part-time. Identifying individuals who are retired is more complex, since there is no uniform retirement age in Canada. Our approach is to include the entire non-student sample, but to account for differences in age and labour market status consistent with the final three stages of the life cycle portrayed by Blinder and Weiss (1976). Methodology and Data The AETS is a cross-sectional data set of individuals and their training activities; it does not capture details concerning work conditions at the firm or establishment level. Consequently, we are somewhat limited in investigating equity questions. We can assess whether or not individuals with similar given traits valued by the market participate in training activity to the same degree. On the other hand, we cannot investigate whether particular firms or establishments have businessrelevant reasons for offering different amounts of training to their workers. Thus, we are confined to examining differences explicable by individual traits. In short, equity questions pertain to individual differences, rather than factors arising from differences in economic organization. Accordingly, we specify an initial encompassing model that

What Are Canadians Doing after School? 95

includes the core economic elements of training decisions. We then expand this model by including additional variables available on different AETS surveys to assess their relative importance and to serve as controls. We estimate two models for training activity: (1) a probit model of participation in training, and (2) a tobit model of training duration.5 The objective of the AETS is 'to measure participation rates' among adults aged seventeen or over in Canada (Statistics Canada n.d., 3). AETS captures information on adult participation in formal training over the previous twelve months. The survey covers job-related training as well as that taken for personal interest, training undertaken full time as well as part-time, and training that is employer-sponsored as well as non-employer-sponsored. The survey covers training given in universities and colleges, private and commercial institutions, and onsite at the workplace, as well as distance learning such as over the Internet. Limited information is gathered about such matters as barriers to participation, the subject matter of courses taken, and respondents' perceptions about the course. The latest AETS survey (1998) is the sixth in a series of similar surveys. Comparable cross-section data from three AETS surveys are available for 1991,1993, and 1997. The AETS is administered as a supplement to the Labour Force Survey (LFS); the sample size for the AETS is large, generally 30,000 to 40,000 individuals. Non-response to AETS is about 15 per cent (Statistics Canada n.d., 25, 36). It is possible to link patterns of participation in adult learning with individual socioeconomic and demographic characteristics.6 AETS gathers no information on informal learning. There is little detail on the costs of training, who paid for the training, and the economic consequences of acquired training for the individual, the individual's family, or the firm. Information about employers is limited to the size of firm, the industrial sector, and whether workers are covered by a collective agreement; there is no information on the conditions of work and the like. Although the AETS surveys span most of the last decade, the individual survey instruments are not identical. Questions on our core variables (age, job tenure, hours worked, and educational attainment) and many basic demographic and employment characteristics (including sex, region of residence, family circumstances, selfemployment, union coverage, firm size, and industry of employment) are consistent across surveys.7 Questions on language, ethnic origin, disability status, professional/managerial status, and employment in

96 Derek P.J. Hum and Wayne Simpson TABLE 3.1 Training incidence and duration, 1992-1998 (weighted estimates; sample size in parentheses) Training

1992*

1994**

1998*

All training: incidence mean duration (hours) Men: Incidence mean duration (hours) Women: Incidence mean duration (hours)

28.4% (42,087) 40.1 (41,081) 28.3% (19,764) 41.8(19,257) 28.6% (22,323) 38.4(21,824)

28.1% (38,671) 37.5 (38,487) 27.4% (17,827) 36.6(17,738) 28.7% (20,844) 38.3(20.749)

26.0% (28,129) 33.6 (27,845) 25.7% (12,423) 34.6(12,295) 26.3% (15,706) 32.6(15,550)

Work-related training: incidence mean duration (hours) Men: Incidence mean duration (hours) Women: Incidence mean duration (hours)

n.a. n.a. n.a. n.a. n.a. n.a.

20.5% (38,671) 27.8(38,473) 22.0% (17,827) 27.9(17,730) 19.1% (20,844) 27.6(20,743)

19.6% (28,129) 27.5(27,886) 20.8% (12,423) 28.6(12,307) 18.5% (15,706) 26.4(15,579)

Source: Adult Education and Training Surveys, 1992,1994,1998. Calculations by the authors. Note: All results are weighted to reflect the Canadian adult population exclusive of fulltime and part-time students * Training question for 1992 differs from 1994,1998. Cannot identify 'career or job related' training duration in 1992. ** 'Training taken for current or future job' used to identify duration in 1994. t Total duration of training for job or career reason taken in 1998. Difference between sample size for incidence and duration reflects missing observations on duration of training; observations with missing hours worked or missing status (age) also deleted to compare with results in Tables 3.2-3.4

the goods sector are available for 1994 and 1998, but not for 1992. Additionally, questions on immigration status, whether or not a respondent is employed at a permanent job, whether or not a respondent changed jobs in the past year, and whether or not employment involved supervisory duties, are available only for the 1998 survey. With this 'instrument instability' a relevant question is the following: How does this additional information affect our understanding of training participation and duration throughout the entire period of the three surveys? Table 3.1 summarizes mean training participation and mean duration of training episodes for the three AETS surveys.

What Are Canadians Doing after School? 97

In 1994, 74 per cent of hours spent training were reported to be 'work-related/ that is, related to a job or career advancement. By 1998 this figure has risen to 82 per cent. Table 3.1 also indicates that adult education and training for those out of school declined slightly between 1992 and 1998.8 Participation in training by men declined from 28.3 per cent in 1992 to 25.7 per cent in 1998, and duration declined from an average of 41.8 hours to 34.6 hours per year. This implies that, for those men who participated in training, their training programs were shorter in 1998 (134.6 hours) than in 1992 (147.7 hours). For women the story is similar. Participation fell from 28.6 per cent to 26.3 per cent while duration fell from 38.4 to 32.6 hours annually, implying that training programs were also shorter (134.3 hours in 1998 compared to 124.0 hours in 1992) for those women who did any training at all. There were, of course, counterbalancing economic incentives to train as labour market conditions improved towards the late 1990s. Training participation is influenced by the business cycle, but the effect is complex. On the one hand, increased hiring and business activity can motivate employers to provide necessary training to new staff. On the other hand, the opportunity costs of training (foregone output) are higher in a robust economy, and this might lead to less training. Accordingly, these counterbalancing influences make any simple causal statement relating the business cycle to training activity impossible. Many factors also changed during the 1990s. There was an increase in education attainment among the population, particularly the proportion of women receiving a university degree. This should increase training according to the human capital model, since previously accumulated human capital (education) is often used as a screening device by firms to select their employees for further training. The upswing in economic activity and employment between 1994 and 1998 is evident, as more workers have less than one year and less than five years job tenure, which should also increase training activity. On the other hand, the population is also aging, as we observe fewer respondents less than thirty-five years of age in 1998 and more respondents over sixty-five. This fact should reduce training activity. Other changes in the socio-economic and demographic structure of Canada's population can also be glimpsed by inspecting the three survey samples. In particular, the population is becoming more urban, less francophone, less likely to be married or to have young children, more likely to be self-employed (especially men), less likely to be unionized, more likely to be in a large firm (500 or more employees),

98 Derek P.J. Hum and Wayne Simpson

and less likely to be 'blue collar' or involved in goods production. Men are less likely but women are more likely to be in the public sector. It is not known exactly how these factors combine to influence training activity. Findings on Training Participation What factors determine whether or not an individual participates in training? What is the duration of training for those who train? Our strategy is as follows. We report estimates of training participation based on the probit model in this section (we consider training duration in the next section), for three subsets constructed from all available variables: the data set of all variables captured in the 1998 AETS (set no. 1); the smaller data set consisting of variables common to the 1994 and 1998 surveys (set no. 2); and the smallest data set consisting of variables common to all three surveys (set no. 3). This approach permits us to examine how the additional variables available in 1998 affect the explanation of training participation and training duration specified by our core model. Table 3.2 presents estimates of training participation for each of these data sets (that is, nos. 1,2, and 3). The regression coefficients estimate the effect of a unit change in the regressor on the probability of training participation. The sign of any coefficient indicates the direction of correlation. That is, a positive coefficient indicates that a unit change in the regressor will increase the probability of training, other factors held constant. Estimates that are statistically significant at the 5 per cent level are in bold text. Our results suggest that educational attainment and age have statistically significant effects on participation for men and women across all data sets. However, hours worked and job tenure do not. Specifically, a university degree raises participation in training for both men and women, although the effect appears to be stronger for women. Other results suggest that training participation falls with education level up to high school graduation. One interpretation is that adult training substitutes for formal education up to completion of high school, but then possibly complements formal education thereafter. Training participation declines uniformly with age, a strong result predicted by our core model. The decline is more rapid for men than women, implying a flatter age-training profile for women, as Green (1991,1993) found. In Figure 3.1, we illustrate this pattern for men and

TABLE 3.2 Estimates of training incidence, 1998 AETS (estimates for variable sets: no. 1:1998 only, no. 2: set common to 1994 and 1998, and no. 3: set common to 1992, 1994, and 1998) Women

Men

Variable/Variable set

#1

Educ: Some high school? high school degree? Post-secondary diploma? university? bachelor's degree? Graduate degree? Age: 20-24 years? 25-34 years? 35-44 years? 45-54 years? 55-64 years? 65 years and over? Hours worked Job tenure: 0.5-1 year? 1-5 years? 5-10 years? 10-20 years? over20 years? Region: Atlantic? Quebec? Prairies? Alberta? British Columbia?

-0.476 (9.3) -0.272(6.8) -0.049(1.1) 0.105(2.1) 0.183(2.1) -0.849(7.1) -1.244(10.7) -1.359 (11.5) -1.457(12.1) -1.830(14.5) -2.154(16.3) -0.002(1.3) -0.019(0.3) -0.065(1.2) -0.076(1.3) -0.012(0.2) 0.054(0.8) -0.133(2.3) -0.142(1.9) -0.119(2.1) -0.067(1.4) 0.063(1.5)

#2

#3

-0.079(2.3) 0.148(3.6) 0.095(2.0)

-0.116(3.4) 0.106(2.6) 0.115(2.4)

-0.773(6.5) -1.144(10.1) -1.282 (11.2) -1.370(11.7) -1.750(14.3) -2.074(16.1) -0.001(0.8) -0.023(0.3) -0.022(0.4) -0.033(0.6) 0.041(0.7) 0.105(1.6) -0.113(2.0) -0.145(1.9) -0.100(1.8) -0.042(0.9) 0.080(1.9)

-0.772(6.5) -1.138(10.0) -1.270 (11.1) -1.367(11.7) -1.754(14.4) -2.055(16.0) -0.000(0.1) -0.012(0.2) -0.005(0.1) -0.016(0.3) 0.053(0.9) 0.128(2.0) -0.112(2.0) -0.338(9.5) -0.075(1.4) -0.033(0.7) 0.108(2.6)

#1 -0.442 (9.6) -0.208(6.0) -0.130(2.8) 0.185(3.9) 0.322(4.2) -0.193(1.7) -0.424(3.8) -0.523 (4.7) -0.642(5.6) -0.860(7.3) -1.125(9.6) 0.003(1.4) 0.026(0.4) -0.149(3.0) -0.053(1.0) -0.031(0.5) -0.041(0.6) -0.136(2.8) -0.133(2.0) -0.012(0.2) 0.012(0.3) 0.152(4.0)

#2

#3

-0.049(1.6) 0.021(0.5) 0.186(4.0)

-0.121(4.0) -0.045(1.0) 0.242(5.3)

-0.133(1.2) -0.353(3.2) -0.484 (4.4) -0.607(5.4) -0.814(7.1) -1.054(9.1) 0.001(1.0) 0.025(0.4) -0.168(3.5) -0.076(1.4) -0.059(1.0) -0.074(1.1) -0.113(2.3) -0.124(1.9) 0.019(0.4) 0.031(0.7) 0.147(4.0)

-0.095(0.8) -0.293(2.7) -0.434 (4.0) -0.574(5.2) -0.787(6.9) -1.015(8.8) 0.003(2.3) 0.032(0.5) -0.130(2.7) -0.039(0.7) -0.018(0.3) -0.016(0.2) -0.140(2.9) -0.298(9.3) 0.000(0.0) 0.025(0.6) 0.141(3.8)

TABLE 3.2

(Continued) Women

Men Variable/Variable set

#1

#2

Urban? French?1 Ethnic origin: Black? Aboriginal? Immigrant? Immigrant after age 18? Disabled? Married or common law? Married previously? Preschool children? Not employed? Full-time job? Permanent job? Changed jobs during yr.? Self-employed? Union coverage? Firm size: 500 or more? 200-499 employees? Probit regression 20-99 employees? Supervisor? Professional/managerial? Blue collar? Public sector?

-0.006(0.2) -0.258(3.4) -0.685(4.4) 0.138(1.1) 0.172(3.2) -0.432 (6.6) -0.245(4.5) 0.171(2.9)

-0.005(0.1) -0.240(3.2) -0.702(4.7) 0.102(0.8)

-0.023 (0.6) -0.010(0.1) 0.024 (0.3) 0.166(2.7) 0.127 (1.8) -0.026(0.5) -0.128(3.3) 0.485(10.9) 0.362 (5.8) 0.226(3.4) 0.119(2.4) 0.206 (5.9) 0.202(4.9) -0.046(1.0) 0.231(4.4)

#3 0.011(0.3)

-0.263(4.9) 0.095(2.7) 0.088 (1.3) -0.018 (0.4) -0.081(0.5)

-0.020 (0.5) -0.367(3.9)

-0.081(1.7) -0.157(4.1) 0.535(12.2) 0.411 (6.6) 0.252(3.8) 0.145(3.0)

-0.084(1.7) -0.187(5.0) 0.542(12.4) 0.429 (7.0) 0.253(3.8) 0.147(3.1)

0.317(8.0) -0.083(1.9) 0.250(4.9)

-0.260(6.8) 0.286(6.0)

0.117(3.4)

#1

#2

#3

-0.061(1.8) -0.208(3.1) -0.158(1.3) -0.073(0.8) -0.075(1.5) -0.127 (2.1) -0.146(3.6) -0.018(0.4) 0.054 (1.1) -0.071 (2.0) -0.297(1.8) -0.055 (1.0) -0.091(1.8) 0.273 (4.5) 0.200(3.8) -0.138(3.6) 0.563(14.0) 0.437 (7.5) 0.383(5.9) 0.235(4.9) 0.245 (7.0) 0.301(8.8) -0.180(2.9) 0.261(6.4)

-0.068(2.0) -0.195(3.0) -0.248(2.1) -0.095(1.0)

-0.060(1.8)

-0.144(3.6) -0.065(2.4)

-0.051(1.9)

-0.095 (2.6) -0.311(2.0)

-0.082 (2.3) -0.341(4.0)

0.164(3.2) -0.166(4.4) 0.576(14.5) 0.439 (7.6) 0.380(5.9) 0.241(5.1)

0.176(3.5) -0.180(4.8) 0.604(15.3) 0.505 (8.8) 0.440(6.8) 0.279(6.0)

0.411(12.6) -0.227(3.7) 0.287(7.1)

-0.406(6.8) 0.411(10.7)

TABLE 3.2

(Concluded) Men

Variable/Variable set Goods sector? Industry: Primary construction transportation trade services Constant Sample size Likelihood ratio test Prediction success

#1 0.003(0.0) 0.112(1.5) 0.130(2.0) 0.211(1.8) 0.015(0.1) 0.158(1.1) 0.497(2.4) 12,423 2,486.5 76.5%

Women #2 0.055(0.4) 0.138(1.9) 0.166(2.5) 0.266(2.3) 0.064(0.5) 0.225(1.6) 0.348(1.9) 12,481 2,321.8 76.0%

#3

#1

#2

#3

0.129(1.8) 0.164(2.5) 0.203(3.8) -0.061(1.2) 0.152(3.1) 0.505(3.6) 12,481 2,194.8 75.8%

-0.085(0.6) 0.183(1.8) -0.198(1.3) 0.080(0.6) -0.304(2.0) -0.109(0.7) 0.125(0.6) 15,706 3,517.7 77.8%

-0.067(0.5) 0.211(2.1) -0.197(1.3) 0.116(0.8) -0.283(1.9) -0.089(0.6) -0.082(0.4) 15,765 3,347.0 76.9%

0.241(2.4) -0.222(1.4) 0.135(1.8) -0.284(4.5) -0.049(0.8) -0.122(0.9) 15,765 3,157.7 76.6%

Source: Adult Education and Training Survey master files. Ottawa: Statistics Canada Surveys Division, 1992, 1994, 1998. Notes: Probit regression; t-values in parentheses. Coefficient estimates b represents the effect of a one-unit change in the independent variable x on the latent index; the effect on the probability of training is f(xb).b where f(.) is the normal destiny function (e.g., see Greene 2000, 815); the categories 'no high school' and 'some high school' are combined where necessary to produce a positive definite matrix. Base comparison group is: no high school, age 17-19, employed in job under 6 months, residence in Ontario, rural community, English-speaking, non-Black, non-Aboriginal, native-born, without disability, never married, agreement, in a firm with fewer than 20 employees, non-supervisory, non-professional/managerial, white collar, without pre-school children, part-time job, temporary employment, paid employment, not covered by a union private sector, non-goods-producing sector, and manufacturing.

102 Derek P.J. Hum and Wayne Simpson Figure 3.1

Age effects on training participation,1998

women in 1998 by age category at the mean of all other characteristics in the sample.

The influence of marital status on training participation is interesting. Married men are more likely to participate in training. On the other hand, the effect of marital status for women becomes insignificant when the full set of variables is included. Unionized workers and workers in smaller firms participate less in training as many other studies have found (Hum and Simpson 1996), while public sector workers participate more even after adjusting for other job and personal characteristics. Other variables, including French language preference, immigration and disability status, and employment in supervisory or professional/managerial positions, are also consistently significant where these variables are available. Moreover, the addition of such variables appears to have an important impact on some coefficient estimates although not for the core variables in general, thereby giving credence to our specification of the core model. For example, the effect of Quebec residence on training participation (relative to Ontario) is much stronger (negative) when language preference is ignored in the third and sixth columns of Table 3.2. Not

What Are Canadians Doing after School? 103

surprisingly, the Quebec variable picks up the reduced participation in training by francophones when language preference is omitted. The results for training participation for 1994 and 1992 in Table 3.3 differ somewhat from those for 1998 discussed above. For the core variables, hours worked (in addition to age and education) is generally significant for both men and women and positively correlated with training participation, as our core model predicts. Education increases training activity as before, and for both men and women the effects are stronger and increase more uniformly than we observe for 1998. Men and women uniformly train less as they age, but there is no evidence here that women have flatter age-training profiles. As before, firm size and jobs in the public sector are consistently important in explaining training participation, although the pattern of increasing participation with increasing firm size is less clear than for 1998. The negative correlation between union coverage and training in 1998 is now observed only for men in 1994. The effect of self-employment on training participation is often significant, as it was for women in 1998, but the results are unstable. Self-employment, which is positively correlated with training for women in 1998, is also positively correlated in 1994 but negatively correlated in 1992. Self-employed men are less likely to train in 1994, although the effect is insignificant in 1998 and 1992. The effect of the omission of language preference on Quebec residence is less clear than in Table 3.2; francophone women are more likely to participate in training in 1994, while the effect of language is insignificant for men. Men and women in professional and managerial positions are more likely to train in all years, while women in so-called blue collar occupations (but not men) are less likely to receive training in all years. Findings for Training Duration Table 3.4 presents tobit regression estimates of training duration for 1998, which correspond, in terms of the data sets analysed, to those for training participation in Table 3.2. Sample sizes are slightly smaller because of missing data on training duration. The regression coefficients estimate the effect of a unit change in the regressor on the latent variable, which now represents training duration (but is censored for those not reporting training), rather than the expected duration of training per se. The sign of the coefficient estimates indicates the direction of correlation. Estimates of the effect on the expected duration of training can be obtained by multiplying the regression coefficient by

TABLE 3.3 Estimates of training incidence, 1994, 1992 AETS (Estimates for 2 variable sets: #1: set common to 1994 and 1998, #2: set common to 1992,1994, and 1998) Men

1994

Women

1994

Men 1992

Women 1992

Variable/Variable Set

#1

#2

#1

#2

#2

#2

Educ:High school degree? post-secondary diploma? university? Age: 20-24 years? 25-34 years? 35-44 years? 45-54years? 55-64 years? 65 years and over? Hours worked Job tenure: 0.5-1 year? 1-5 years? 5-10 years? 10-20 years? over20years? Region: Atlantic? Quebec? Prairies? Alberta? British Columbia? Urban? French?1 Ethnic origin: Black? Aboriginal?

0.276(8.3) 0.612(18.6) 0.663(15.9) -0.349(4.1) -0.769(9.2) -0.792(9.3) -0.965(11.0) -1.232(13.5) -1.436(14.5) 0.003(2.4) -0.170(2.6) -0.123(2.7) -0.005(0.1) -0.067(1.4) 0.141(2.6) -0.052(1.2) -0.098(1.6) 0.103(2.2) 0.052(1.3) 0.207(5.9) -0.020(0.7) -0.025(0.4) 0.028 (0.2) -0.144(1.5)

0.289(8.8) 0.641(19.7) 0.759(19.3) -0.349(4.1) -0.770(9.2) -0.789(9.3) -0.962(11.0) -1.234(13.6) -1.496(15.4) 0.004(2.7) -0.177(2.7) -0.119(2.6) 0.008(0.2) -0.049(1.0) 0.163(3.0) -0.056(1.3) -0.112(3.8) 0.102(2.2) 0.050(1.2) 0.203(5.9) -0.018(0.6)

0.450(14.9) 0.630(20.0) 0.800(20.2) -0.472(5.7) -0.814(10.3) -0.852(10.8) -0.931(11.5) -1.162(13.9) -1.286(14.9) 0.004(3.4) -0.100(1.6) -0.205(4.9) -0.053(1.2) 0.022(0.5) -0.168(2.7) -0.184(4.4) -0.223(4.1) 0.037(0.9) 0.181(4.9) 0.167(5.2) -0.058(2.0) 0.107(2.0) 0.185 (1.7) -0.193(2.1)

0.479(16.0) 0.692(22.4) 0.924(24.6) -0.455(5.5) -0.784(9.9) -0.828(10.5) -0.918(11.4) -1.156(13.9) -1.355(16.0) 0.005(4.5) -0.077(1.3) -0.190(4.5) -0.029(0.6) 0.050(1.1) -0.112(1.8) -0.191(4.6) -0.130(4.9) 0.012(0.3) 0.158(4.3) 0.146(4.6) -0.063(2.2)

0.378(12.8) 0.513(16.8) 0.840(22.7) -0.719(9.0) -0.977(12.6) -1.167(14.6) -1.401(17.1) -1.660(19.6) -1.876(20.6) 0.001(1.1) 0.114(1.8) 0.095(2.2) 0.018(0.4) 0.200(4.3) 0.218(4.1) -0.213(5.0) -0.141(5.1) 0.077(1.8) 0.155(4.2) 0.095(2.8) 0.021(0.7)

0.441(16.1) 0.654(22.5) 0.867(23.4) -0.761(9.9) -0.908(12.3) -0.935(12.6) -1.089(14.4) -1.308(16.7) -1.529(19.1) 0.003(2.5) 0.088(1.5) -0.097(2.5) -0.081(1.8) -0.116(2.6) -0.094(1.6) -0.153(3.8) 0.106(4.1) 0.092(2.3) 0.214(6.0) 0.158(4.9) 0.038(1.3)

TABLE 3.3

(Concluded)

Variable/Variable set

Men

1994

Women

1994

Men 1992

Women 1992

#1

#2

#1

#2

#2

#2

-0.008(0.3) 0.043(1.4) -0.096(1.4) -0.148(4.0) -0.188(6.2) 0.661(19.6) 0.355(6.8) 0.615(11.1) 0.219(5.9)

-0.033(1.4) -0.093(3.1) -0.046(0.8) 0.248(5.9) -0.000(0.0) 0.391(12.7) 0.345(6.7) 0.186(3.3) 0.202(5.1) 0.240 (8.9) -0.191(3.8) 0.213(6.4) 0.360 (6.4) -0.170(2.1) -0.342(2.9) 0.046(0.7) 0.103(2.2) 0.225(5.5) -0.393 (3.9) 20,844 4,754.1 75.7%

-0.024(1.0) -0.086(2.8) -0.070(1.3) 0.232(5.5) 0.001(0.0) 0.418(13.6) 0.378(7.4) 0.247(4.4) 0.254(6.4)

0.128(4.5) -0.100(3.3) -0.072(1.1) -0.079(1.9) -0.048(1.7) 0.265(5.4) 0.585(18.1) 0.453(11.4) 0.228(6.2)

0.039(1.7) -0.115(4.0) -0.196(3.8) -0.111(2.3) 0.073(2.4) 0.055(1.2) 0.443(14.1) 0.317(8.1) 0.161(4.1)

-0.085(1.9) 0.233(7.1)

-0.075(2.7) 0.194(5.5)

-0.233(5.4) 0.153(4.7)

0.026(0.3) -0.083(0.7) -0.051(0.8) -0.019(0.5) 0.147(4.4) -0.362 (3.7) 20,844 4,569.7 75.0%

0.165(3.3) 0.091(2.1) 0.215(5.2) 0.102(2.8) 0.168(5.0) -0.286 (2.8) 19,764 4,392.2 75.7%

0.289(3.7) 0.289(3.3) 0.227(3.9) 0.048(1.3) 0.180(5.7) -0.264 (2.8) 22,323 5,117.1 75.7%

Disabled?

-0.039(1.0)

Married or common law? Pre-school Children? Not employed? Self-employed? Union coverage? Firm size: 500 or more? 200-499 employees? 100-199 employees? 20-99 employees? Professional/managerial? Blue collar? Public sector? Goods sector? Industry: Primary construction transportation trade Services Constant Sample size Likelihood ratio test Prediction success

-0.017(0.6) 0.037(1.2) -0.091(1.3) -0.147(3.9) -0.172(5.7) 0.644(18.9) 0.329(6.3) 0.588(10.6) 0.199(5.3) 0.215 (6.4) 0.013(0.4) 0.187(4.9) 0.063 (1.1) 0.039(0.7) -0.076(1.6) 0.030(0.5) 0.112(1.9) 0.189(3.2) -0.496 (4.3) 17,827 3,744.6 75.9%

-0.193(5.8)

-0.063(2.2) 0.189(5.0) 0.068(1.3) -0.046(1.0) -0.001(0.0) 0.055(1.4) 0.164(4.5) -0.431 (4.0) 17,827 3,693.3 75.5%

Source: Adult Education and Training Survey master files, 1992, and 1994. Notes: See Table 3.2

TABLE 3.4 Estimates of training duration, 1998 AETS (estimates for 3 variable sets: #1:1998 only, #2: set common to 1994 and 1998, and #3: set common to 1992,1994, and 1998)

Men Variable/Variable set

#1

Educ: Some high school? High school degree? Post-secondary diploma? university? bachelor's degree? graduate degree? Age: 20-24 years? 25-34 years? 35-14 years? 45-54 years? 55-64 years? 65 years and over? Hours worked Job tenure: 0.5-1 year? 1-5 years? 5-10 years? 10-20years? over 20 years? Region: Atlantic? Quebec? Prairies? Alberta? British Columbia?

-170.0(9.3) -86.7(6.3) -4.0(0.3) 21.0 (1.2) 31.3 (1.1) -371.8(10.4) -547.4(15.9) -607.1(17.2) -644.6(17.7) -760.3(19.6) -901.4(21.4) -1.5(2.4) -63.0(2.7) -158.7(8.9) -137.6(6.8) -118.9(5.7) -93.1(4.2) -27.9(1.4) -15.5(1.7) -44.1(2.3) -12.6 (0.8) 16.6(1.2)

Women #2

#3

-23.0(1.9) 62.2(4.3) 12.3(0.7)

-35.8(3.0) 47.0(3.3) 20.8(1.3)

-351.6(9.9) -522.9(15.3) -595.2(17.1) -633.1(17.7) -753.3(19.8) -891.6(21.5) -1.3(2.6) -60.5(2.6) -144.6(8.3) -126.7(6.4) -105.4(5.2) -80.3(3.7) -24.5(1.3) -16.6(1.8) -38.9(2.0) -8.6 (0.5) 22.2(1.5)

-351.3(9.8) -521.4(15.2) -592.1(16.9) -634.1(17.7) -756.5(19.8) -882.7(21.3) -0.9(1.9) -57.5(2.5) -138.9(7.9) -120.8(6.1) -101.2(4.9) -70.6(3.2) -22.3(1.2) -93.0(7.4) -32.3(1.7) -6.9 (0.4) 31.4(2.2)

#1 -134.3(9.3) -61.9(5.9) -41.7(2.9) 49.7 (3.5) 116.0 (5.0) -12.5(0.4) -193.4(6.0) -243.5(7.6) -281.9(8.6) -372.9(10.9) -469.6(13.6) -0.1(0.2) -53.1(2.8) -127.6(8.9) -91.7(5.8) -88.0(5.0) -86.8(4.4) -18.3(1.2) -11.4(0.6) 13.3(0.9) 2.1 (0.2) 50.3(4.5)

#2

#3

-19.3(2.0) -1.9(0.1) 52.2(3.7)

-36.0(3.9) -18.3(1.3) 65.4(4.6)

8.9(0.3) -172.1(5.4) -232.1(7.3) -270.0(8.3) -359.0(10.6) -445.4(13.0) -0.9(2.2) -55.5(2.9) -144.8(10.2) -112.5(7.2) -112.6(6.6) -111.7(5.7) -18.5(1.2) -8.4(0.4) 16.2(1.1) 3.2 (0.2) 50.0(4.4)

16.7(0.5) -162.3(5.1) -224.3(7.0) -267.1(8.2) -358.2(10.6) -443.8(13.0) -0.5(1.2) -52.9(2.8) -135.6(9.6) -103.8(6.7) -103.2(6.1) -96.9(5.0) -25.0(1.7) -52.5(5.3) 11.6(0.8) 0.2 (0.0) 48.9(4.4)

TABLE 3.4

(Continued) Women

Men

Variable/Variable set

#1

Urban? 19.5(1.4) French?1 -63.2 (2.4) Ethnic origin: Black? -206.0(3.9) Aboriginal? 4.1(0.1) Immigrant? 41.7(2.3) Immigrant after age 18? -119.7(5.4) Disabled? -99.8(5.2) Married or common law? 39.2(1.9) Married previously? 59.4 (2.5) Preschool children? -12.2(0.9) Not employed? -135.3(2.5) Full-time job? 8.0 (0.3) Permanent job? 64.7(3.0) Changed jobs during yr.? 125.7(5.5) Self-employed? -38.4(2.1) Union coverage? -41.6(3.1) Firm size: 500 or more? 95.5(6.3) 200-499 employees? 74.6(3.5) 100-199 employees? 57.2(2.5) 20-99 employees? -0.7(0.0) Supervisor? 26.6 (2.3) Professional/managerial? 69.0 (4.9) Blue collar? -10.8(0.7) Public sector? 42.6(2.5) Goods sector? -11.3(0.3)

#2 18.1(1.3) -59.5 (2.3) -220.2(4.2) -12.8(0.3)

#3 23.6(1.6)

-108.8(5.6) -6.3(0.5)

0.5(0.0)

-14.2(1.0) -173.8(3.2)

-14.4(1.0) -255.2(7.7)

-38.8(2.3) -46.0(3.5) 113.6(7.5) 87.5(4.1) 63.6(2.7) 8.6(0.5)

-40.6(2.4) -54.4(4.2) 116.9(7.7) 95.8(4.5) 66.0(2.8) 10.0(0.6)

99.2 (7.2) -23.0(1.5) 48.4(2.8) 2.1(0.1)

-79.4(6.0) 58.1(3.6)

#1

#2

1.2(0.1) -46.1 (2.3) -46.4(1.2) 2.5(0.1) 10.9(0.7) -29.7(1-7) -57.0(4.4) -9.1(0.7) 28.9 (2.0) -45.0(4.1) -157.6(3.1) -17.2 (1.0) -80.9(5.6) 58.2(3.4) 49.0(3.0) -36.7(3.2) 135.6(11.3) 116.6(6.7) 104.8(5.3) 67.3(4.7) 48.8 (4.8) 57.1 (5.7) -30.1(1.6) 41.7(3.4) 22.4(0.5)

3.8(0.4) -48.8 (2.4) -50.2(1.3) -10.7(0.4)

#3 5.0(0.5)

-59.8(4.6) -31.3(3.7)

-27.2(3.3)

-52.2(4.8) -149.6(3.1)

-48.8(4.5) -179.1(6.9)

44.9(2.8) -45.7(4.0) 140.8(11.7) 114.6(6.6) 101.1(5.1) 64.4(4.5)

46.2(2.9) -49.7(4.4) 148.2(12.4) 130.8(7.6) 115.0(5.8) 72.3(5.1)

88.0 (8.9) -33.6(1.8) 55.2(4.5) 6.3(0.1)

-73.1(4.0) 83.7(7.2)

TABLE 3.4

(Concluded)

Men Variable/Variable set Industry: primary construction transportation trade services Constant Sample size Percent obs. above limit Mean error Correlation of observed and expected

Women

#1

#2

#3

#1

7.0(0.3) 24.1(1.0) 52.6(1.3) -24.1(0.5) 27.1(0.6) 406.6(6.0) 12,295 24.9% 17.7

12.0(0.5) 28.4(1.2) 68.1(1.7) -11.7(0.3) 49.5(1.0) 404.6(6.6) 12,353 24.8% 17.8

9.1(0.4) 27.7(1.2) 61.5(3.3) -35.3(2.0) 42.3(2.5) 432.9(9.7) 12,353 24.8% 17.8

34.9%

32.0%

31.3%

#2

#3

65.9(2.1) -49.3(1.0) 61.7(1.3) -30.2(0.6) 23.5(0.5) 143.1(2.3) 15,550 25.5% 14.4

56.0(1.8) -57.8(1-2) 46.2(1.0) -50.1(1.1) 7.2(0.2) 82.7(1.4) 15,608 25.5% 14.8

63.7(2.0) -62.0(1.3) 34.2(1.5) -72.5(3.7) -5.2(0.3) 94.9(2.4) 15,608 25.5% 15.0

31.8%

29.2%

29.1%

Source: Adult Education and Training Survey master files, 1992,1994,'1998. Notes: Tobit regression; t-values in parentheses. Coefficient estimate b represents the effect of a one-unit change in the independent variable x on the latent index; the effect on the expected duration of training is F(xb).b where F(.) is the normal distribution function and equals the proportion of the sample above the limit (Moffitt and MacDonald 1980, 319). Base comparison groups as in Table 3.2

What Are Canadians Doing after School? 109 Figure 3.2 Age effects on training duration, 198

the proportion of respondents reporting training (about one-quarter). Among the core variables, age continues to be significant and indicates that training duration declines with age. Researchers need to be aware that this strong effect of age may confound any simple correlation of other job and worker characteristics with training, since those characteristics might be correlated with age rather than training per se. As Figure 3.2 illustrates, when we focus on duration there is no evidence of the flatter age-training profile for women that we observed for incidence. University-educated women undertake more training than women without university education, particularly those with a graduate degree, but the effect of a university education is insignificant for men. As before, there is no uniform effect of increasing education since high school attendance significantly reduces training, other factors considered. More noteworthy, there is a significant pattern of training duration with job tenure that is not apparent in the evidence for training participation only. The pattern is non-linear and U-shaped, with training duration declining up to five years of job tenure and increasing thereafter, particularly for men. This pattern is consistent with the predictions of our core model, in the sense that new workers train more

110 Derek P.J. Hum and Wayne Simpson Figure 3.3 Job tenure effects on duration, 1998

and training declines with subsequent job tenure, but it suggests a more complex relationship between training duration and job tenure than our model predicts. We plot estimated mean training duration for 1998 as a function of job tenure in Figure 3.3. Francophones, black men, adult male immigrants, persons with disabilities, those not employed, women with permanent jobs, selfemployed men, and unionized workers all receive less training than their opposite counterparts. Those previously married but not now married; men with permanent jobs; those who changed jobs in the past year; self-employed women; workers in larger firms; supervisors; and professional, managerial, and public sector workers all received more training than their opposite counterparts. As before, the Quebec residence effect is clearly a predominantly francophone effect. The results for training duration for 1994 and 1992 (not reported here, available from authors) are similar to those for 1998 with a few important differences. The core variables are generally significant. More education increases training uniformly for men and women; the education effect is stronger in 1992 and 1994 than in 1998. The strong decline in

What Are Canadians Doing after School? Ill

training duration with age is observed here as before, but the steeper age-training profile for men is no longer observed. Hours worked are generally positively correlated with training duration, although these were negatively correlated (and insignificant for women) in 1998. Training is significantly greater for new male workers but not for females, and the relationship between job tenure and training is decidedly nonlinear and complex. As we observed for training participation, the negative correlation between union coverage and training duration observed in 1998 is only observed for men in 1994 and not at all in 1992. It would appear that the more limited data set in 1992 cannot account for this result, since the negative correlation between union status and training duration for 1998 holds (with regression coefficients of about the same magnitude) for all the variable sets. As was the case for training participation, the firm-size effects are unstable and do not show the clear pattern apparent in the 1998 results. Self-employment is associated with less training for men but more training for women in 1994 consistent with the 1998 results, but there is no statistically significant pattern in 1992. Francophone women train more in 1994, but there is no effect for francophone men. More training for public sector workers and professional and managerial workers is consistent with the results for 1998. Married women and women with pre-school children train less in all survey years. In work not reported here, we pooled all surveys but included simple dummy variables for the different years, in essence restricting all coefficients except the survey effect dummies to be identical across surveys. Our estimates indicate a statistically significant survey effect for 1998 in every case, suggesting that, other factors considered, training participation and duration were significantly lower in 1998 than in 1994 and 1992. We also compare the pooled results (without the dummy variables for survey years to represent the restricted regression) with results that allow the coefficients to differ across each survey (to represent the unrestricted regression). Comparison of the log likelihood results for these restricted and unrestricted regressions allows us to formally test the stability of the coefficients across the surveys. We find strong formal statistical evidence of parameter instability across the 1994 and 1998 surveys and across the 1992, 1994, and 1998 surveys for both training participation and training duration. Whether this instability results from variations in the survey instrument, inherent instability in training behaviour, or missing crucial variables, is not yet clear (Hum and Simpson 2001).

112 Derek P.J. Hum and Wayne Simpson Conclusion

This study asks the following simple questions: What factors determine who participates in post-school education and training? And for what duration? The life-cycle model argues that education and training are lifelong activities related to the working career. Consequently, postschool education and training by adults is undertaken to augment human capital. Human capital may become obsolete due to new technology. Alternatively, individuals may leave the labour force for an extended period, whether by choice, family responsibilities, or marital disruption. Their human capital may depreciate to the point that additional education and training is necessary to re-enter the workplace. Workers may change employers, locations, or sectors due to lay-offs, quits, and so on, such that human capital must be replaced by new specific capital, again necessitating training. Because human capital must be upgraded for re-entry or advancement, replaced for new employment, or simply updated, adult education and training occurs throughout the entire work life. We examined the master files of the Adult Education and Training Survey (AETS) for 1992,1994, and 1998 to study training participation and training duration for non-student adults. We specified a model of training with age, hours worked, job tenure, and past human capital as core variables. Our empirical results suggest that educational attainment and age have statistically significant effects on participation for men and women across all data sets, but that hours worked and job tenure do not. Training participation and duration both decline sharply with age; education (particularly post-secondary education) raises training activity; and there is a distinct U-shaped relationship between training duration and job tenure. The patterns of training activity for men and women differ; training activity declines more rapidly with age for men, and the effect of post-secondary education on training activity is stronger for women. In addition to the core variables, some employment characteristics are significant. Notably, there is more training in large firms and in the public sector, and workers in professional and managerial occupations train more. For example, there appear to be reinforcing effects of postsecondary education, a large employer, and professional/managerial employment on training activity. We also find some intriguing results among the new variables introduced in the 1998 survey, which suggest that persons with disabilities

What Are Canadians Doing after School?

113

and male immigrants train less. These findings constitute avenues for further study, and have particular relevance to issues of equitable access for women, persons with disabilities, immigrants,9 and workers in particular regions (especially the North), particular economic sectors, and occupations. These are all items for future research. Despite their preliminary nature, our findings have important lessons fpr those interested in equity of access to post-school training. That immigrants or persons with disabilities have less access to training likely comes as no surprise, but programs to improve training for these two groups may require different approaches. Training opportunities are fewer for women in general. However, this result must be interpreted alongside the division of family responsibilities between men and women, women's occupational situation, the tendency for women to be employed in certain sectors or in firms having particular economic characteristics. Thus, it becomes difficult at this point to disentangle whether the training patterns of women (vis-a-vis men) differ because of individual traits or employer circumstances, and whether the distribution of women in the labour force and their access to training opportunities reflect deliberate self-selection by individuals, employer-imposed barriers, or behavioral norms that warrant a deeper and broader investigation than is permitted with our data. Nonetheless, our study confirms the usefulness of certain economic core variables suggesting that age is an important, albeit complex, determinant of training participation. So too is education, and the finding that educational attainment often begets more education opportunities. One encouraging finding is that this effect appears stronger for women than men, suggesting that as women achieve higher educational attainment, their ability to close the training gap with respect to men will improve. There are also a number of factors from the employer-firm side that appear to determine training patterns; these include unionized status, firm size, capital-labour ratio in the firm, and so on. These features are associated with different patterns of training behaviour, and appeared in our model mainly to serve as controls. From our data, however, not too much can be made of these structural characteristics in discussing policy solutions aimed at fostering individual equity of access to training, since we do not know the individual circumstance of employers at the firm level. Our data do not permit this level of enquiry. Further study must therefore await more detailed data.

114 Derek PJ. Hum and Wayne Simpson Notes The research for this chapter was supported in part by the Applied Research Branch, Human Resources Development Canada (HRDC), Ottawa. We are grateful to Statistics Canada for permitting access to the master file of AETS. We thank Gilles Berube and Lucie Gilbert of HRDC, Lisa Shipley of Statistics Canada, and Jane Gaskell and Kjell Rubenson of WRNET and UBC for their comments and assistance. The views expressed are solely those of the authors. 1 The Canadian literature on the individual determinants of training is very sparse. One exception is Betcherman, Leckie, and McMullen (1998), who use the 1994 AETS to report on individual determinants of training. However, the authors do not perform formal regression analysis and do not control for a wide variety of variables. Hence, their findings should be treated with caution. Analyses of the 1994 AETS in a multivariate context include Jennings (1996), Kapsalis (1996), and de Broucker (1999). For a recent analysis of the AETS, including the 1998 survey, see Canada (2001). 2 See Riddell's chapter in this volume for further discussion of the relationship between skills and labour market outcomes. 3 See Allen's chapter in this volume for details on returns to educational investment. 4 One implication of the life-cycle model is that training and hours worked are determined simultaneously, or interdependently. Hours worked will mimic the inverted-U pattern of human capital investment, although the peak will occur earlier (Blinder and Weiss 1976,466). From an empirical standpoint, this raises the possibility that hours of work may be correlated with the error term in the training equation, resulting in inconsistent parameter estimates, but we ignore this complication here as research in this area has traditionally done. One exception is some earlier research we conducted using the Labour Market Activity Survey, in which we were able to reject the hypothesis of parameter inconsistency arising from the inclusion of hours worked (Simpson, Sproule, and Hum 1994). 5 See Greene (2000) for further discussion of probit and tobit models. Briefly, probit models estimate the probability of taking training, and tobit models estimate the duration of the training, taking into account that some individuals take no training at all; i.e., have a training duration of zero. 6 Since the AETS is a supplement to the LFS, its design is closely tied to the LFS. The LFS is a household survey whose sample of individuals is representative of the civilian, non-institutionalized population 15 years of age or over in the ten provinces. Excluded are residents of Nunavut, Yukon, and

What Are Canadians Doing after School? 115 the North-West Territories, Aboriginal people living on reserves, full-time members of the Armed Forces, and inmates of institutions. These groups represent an exclusion of approximately 2% of the population aged 15 or over. The LFS consists of approximately 52,000 occupied dwellings from which information is collected from approximately 102,000 civilians aged 15 or over. For the AETS, the LFS is modified to include all members of the household 17 years of age or older (including those over 70). However, upon completion of the LFS, the AETS is administered to only one randomly selected individual per household. For further details see, Statistics Canada (n.d.) 1998 Adult Education and Training Survey: Microdata User Guide. Past adult education and training is not included in any survey. 7 Age and job tenure are available as categorical data for all three surveys, but actual age and job tenure are not available for 1992. 8 The decline in participation in training from the 1994 to 1998 AETS is consistent with other findings (e.g., Canada 2001,34). Canada (2001) reports a modest increase rather than a decline in the duration of training, however. While we have not explored the reasons for the different pattern of duration in this chapter, there are differences in the sample selection criteria. In particular, our study excludes all students, both full time and part-time, whereas Canada (2001) excludes only full-time students. We also use nonimputed data for training duration in our analysis. 9 We have previously studied the labour market performance of persons with disabilities in Canada. The training pattern of immigrants to Canada is the subject of an ongoing research project currently being studied by the authors. References Altonji, J., and J. Spletzer. 1991. Worker characteristics, job characteristics, and the receipt of on-the-job training. Industrial and Labor Relations Review 45: 5879. Barren, J., D. Black, and M. Loewenstein. 1987. Employer size: The implications for search, training, capital investment, starting wages, and wage growth. Journal of Labor Economics 5: 76-89. Bartel, A. 1995. Training, wage growth, and job performance: Evidence from a company database. Journal of Labor Economics 13(3): 401-25. Betcherman, G., N. Leckie, and K. McMullen. 1997. Developing skills in the Canadian workplace: The results of the Ekos Workplace Training Survey. Ottawa: Canadian Policy Research Networks.

116 Derek P.J. Hum and Wayne Simpson Betcherman, G., K. McMullen, and K. Davidson. 1998. Training for the new economy: A synthesis report. Ottawa: Canadian Policy Research Networks. Blinder, A., and Y. Weiss. 1976. Human capital and labor supply: A synthesis. Journal of Political Economy 84: 449-72. Booth, A. 1991. Job-related formal training: Who receives it and what is it worth? Oxford Bulletin of Economics and Statistics 53: 281-94. Borjas, G. 1975. Job investment, labor mobility and earnings. PhD diss., Columbia University, NY. - 1981. Job mobility and earnings over the life cycle. Industrial and Labor Relations Review 34: 365-76. Canada. 2001. A report on adult education and training in Canada: Learning a living. Ottawa: Statistics Canada & Human Resources Development Canada. Cat. no. 81-586-XPE. de Broucker, P. 1997. Job-related education and training - Who has access? Education Quarterly Review 4(1): 10-31. Duncan, G., and S. Hoffman. 1979. On-the-job training and earnings differences by race and sex. Review of Economics and Statistics 61: 594-603. Green, F. 1991. Sex discrimination in job-related training. British Journal of Industrial Relations 29(2): 295-314. - 1993. The determinants of training of male and female employees in Britain. Oxford Bulletin of Economics and Statistics 55(1): 103-22. Green, F., and L. Zanchi. 1997. Trends in the training of male and female workers in the United Kingdom. British Journal of Industrial Relations 35(4): 635-44. Greene, W. 2000. Econometric analysis. 4th ed.Upper Saddle River, NJ: Prentice Hall. Greenhalgh, C, and M. Stewart. 1987. The effects and determinants of training. Oxford Bulletin of Economics and Statistics 49:171-90. Holtmann, A., and T. Idson. 1991. Employer-size and on-the-job training decisions. Southern Economic Journal 58: 339-55. Holzer, H.J., and J. Reaser. 1999. Firm-level training for newly hired workers: Its determinants and effects. In Research in Labor Economics. Vol. 18. Ed. Solomon Polochek, 377-402. Stamford, Conn.: JAI Press. Hum, D., and W. Simpson. 1996. Maintaining a competitive workforce: Employerbased training in the Canadian economy. Monograph series on education no. 6. Montreal: Institute for Research on Public Policy. - 2001. Participation in adult training by Canadians in the 1990s: A multivariate analysis using AETS data. Study prepared for Applied Research Branch, Human Resources Development Canada, Ottawa. Jennings, P. 1996. Employer-sponsored training in Canada: Evidence from the 1994

What Are Canadians Doing after School? 117 Adult Education and Training Survey. Applied Research Branch Working Paper. Ottawa: Human Resources Development Canada. Kapsalis, C. 1996. Determinants of employer-sponsored training: An analysis of the 1994 Adult Education and Training Survey. Applied Research Branch Working Paper. Ottawa: Human Resources Development Canada. Kennedy, S., R. Drago, J. Sloan, and M. Wooden. 1994. The effect of trade unions on the provision of training: Australian evidence. British Journal of Industrial Relations 32(14): 565-80. Lillard, L., and H. Tan. 1986. Private sector training: Who gets it and what are its effects? Prepared for the U.S. Department of Labor, R-3331-DOL/RC. Lynch, Lisa. 1992. Private sector training and the earnings of young workers. American Economic Review 82(1): 299-312. Lynch, L., and S.E. Black. 1998. Beyond the incidence of employer-provided training. Industrial and Labor Relations Review 52(1): 64-81. Miller, P.W. 1994. Gender discrimination in training: An Australian perspective. British Journal of Industrial Relations 32(4): 539-64. Mincer, J. 1983. Union effects: Wages, turnover, and job training. In Research in Labor Economics, Vol. 5, supp. 2. Ed. Ronald Ehrenberg. Greenwich, Conn.: Jai Press. - 1984. The economics of wage floors. In Research in Labor Economics. Vol. 6, supp. 2. Ed. Ronald Ehrenberg. Greenwich, Conn.: Jai Press. Shields, M. 1998. Changes in the determinants of employer-funded training for full-time employees in Britain, 1984-1994. Oxford Bulletin of Economics and Statistics 60(2): 189-214. Simpson, W. 1984. An econometric analysis of industrial training in Canada. Journal of Human Resources 19(4): 435-51. Simpson, W., R. Sproule, and D. Hum. 1994. Specification of on-the-job training incidence. Unpublished manuscript, University of Manitoba (August). Statistics Canada, n.d. 1998 adult education and training survey: Microdata user guide. Ottawa: Special Surveys Division. Weiss, Y. 1986. The determination of life cycle earnings: A survey. In Handbook of Labor Economics. Vol. 1. Ed. O. Ashenfelter and R. Layard. New York: North-Holland, 603-40. Wooden, M., and A. VandenHeuval. 1997. Gender discrimination in training: A note. British Journal of Industrial Relations 35(4): 627-33.

4 The Post-secondary Education of Disadvantaged Adults ATLANTA SLOANE-SEALE, LORI WALLACE, AND BEN LEVIN

The importance of education to individual and societal success is now an article of faith in public policy. Across the world, average levels of education have been rising, and higher levels of education are related to better individual outcomes in such varied areas as employment, earnings, further learning, health, and longevity. Issues of inequality, however, remain important. In almost every society, some individuals and groups get more education than others, and some benefit more from education than do others. Socio-economic status, place of residence, ethnicity, and disability are among the factors that systematically influence life outcomes. Yet while the importance of these factors in sustaining disadvantage is understood, how best to alleviate them is much less clear. Many societies, including Canada, are struggling with ways to ensure that education and its benefits are available to all segments of the population. In Canada, the educational and economic outcomes for Aboriginal people are a particular concern. Aboriginal people often have poor-quality education, leading to low secondary school completion rates, and they also continue to lack access to programs that would prepare them for entry to post-secondary education. They live in isolated settings, have high rates of individual and community poverty, and may grow up speaking English as a second language. Growing Aboriginal populations also make up a major part of urban poverty in western Canada. In our research, we see that economic and educational outcomes are shaped in complex ways by people's backgrounds and personal characteristics, and by the economic and social structures in which they live. In common with the perspective presented in Butterwick's chapter

Post-secondary Education of Disadvantaged Adults 119

in this volume, we adopt a view that gives attention both to structure and to agency, but we do not see these as equally powerful. Agency does matter, which is why individual outcomes are very difficult to predict. People do make decisions and choices that have important consequences for their lives. Social and economic structures, however, are very powerful influences on outcomes, which is why at the aggregate level we can predict outcomes for particular populations with a reasonable degree of accuracy. Moreover, chance can play an important role as well; unlikely circumstances or events can have powerful effects on the courses of people's lives. So, although some persons will be successful despite facing serious disadvantages in their early lives, the majority of people with such disadvantages will be less successful than those who were more fortunate. The nature of causal relationships in this equation, and especially the place of formal education, is the subject of a great deal of controversy. Some analysts argue that education is a route to mobility in spite of socio-economic background. Others argue that educational attainment is so highly correlated with socio-economic status (SES) that it can safely be regarded as a proxy for it. The same debate occurs over the relationship between education and employment. Increased educational attainment is strongly associated with better employment outcomes, leading to the widespread belief that education is a route to employment. However, another view is that both increased education and better employment are proxies for better socio-economic background. Both views are probably true to some extent - in line with our initial belief in both structure and agency - but it is clear that in Canada, overall, steadily rising levels of average educational attainment over the past thirty years have not diminished economic inequality. This chapter reports on research of a Canadian success story. Based on studies of persons who applied to a special program of university studies for disadvantaged adults, the chapter casts light on the kinds of experiences that can shape positive educational and other life outcomes for such adults. The programs, and hence this chapter, give particular emphasis to Aboriginal people. Related Literature There are few longitudinal studies on the links between Aboriginal peoples' motivations, barriers to participation in post-secondary education and training, and labour market outcomes. Some studies use

120 Atlanta Sloane-Seale, Lori Wallace, and Ben Levin

Statistics Canada data in analyses of the achievement and prospects of Aboriginal students who have completed higher education (Armstrong, Kennedy, and Oberle 1990; Drost 1995; Sarkar and Stallard 1997; Saskatchewan Institute of Applied Science and Technology [SIAST] 1996; Statistics Canada 1985). The available literature, however, has little to say on the ways in which these lives unfold over time. Moreover, much of the school-to-work transition literature tends to stop at around age twenty-five, and the conceptual literature 'on second chance' - the opportunity to return to education some years after leaving (Inbar 1990) - has not led to a significant body of empirical work in Canada. The educational and labour market situation of Aboriginal people is quite well known. Although educational outcomes have been improving, these remain weaker, with lower rates of high school graduation and post-secondary participation than in the general population. Labour force participation rates are lower and unemployment rates are higher for Aboriginal people than for the rest of the population. Urban single-ethnic-origin Aboriginal people are three times more likely to experience unemployment than non-Aboriginals, and twice as likely as Aboriginals with mixed ethnic origins. Over 60 per cent of urban Aboriginal people are concentrated in cities, and, in particular, cities in western Canada. Regina, Saskatoon, and Winnipeg, with high Aboriginal populations, also have particularly low Aboriginal labour force participation rates. In Winnipeg, 44 per cent of single-ethnic-origin Aboriginal people of working age are not in the labour force, compared to 25 per cent of Aboriginal people with mixed ancestries, and 19 per cent of non-Aboriginal people (Drost 1995; Statistics Canada 1993). A number of studies, using data from the 1991 census, show high rates of social assistance among Aboriginal people, and that their overall unemployment rate in major cities (15%) is close to twice that (8%) of non-Aboriginal people (Armstrong, Kennedy, and Oberle 1990; Drost 1995; SIAST 1996; Statistics Canada 1993). Unemployment rates on reserves are anecdotally reported to be much higher. A number of factors may influence labour force participation and employment. Having dependent children is one important factor: since many single mothers have full responsibility for child-rearing, their ability to acquire and keep a job is greatly reduced (also see Butterwick, in this volume). The restrictions on labour market attachment seem to be greater, however, for Aboriginal single mothers than for their non-Aboriginal counterparts (Drost 1995; Statistics Canada 1993).

Post-secondary Education of Disadvantaged Adults 121

The relationship between age and unemployment is U-shaped for both Aboriginal and non-Aboriginal people: the incidence of unemployment declines until prime age is attained and then rises again. The Aboriginal population currently is much younger than Canada's total population; 36 per cent are under the age of fifteen, versus 21 per cent for the total population, with only 7 per cent of the Aboriginal population aged fifty-five years and over, versus 20 per cent of all Canadians. Therefore, higher unemployment rates in the working-age population have an amplified economic impact. A much larger cohort of Aboriginal youth will also be entering the labour force over the next decade. This will create challenges in the provinces and cities of western Canada. If job opportunities for Aboriginal people do not greatly increase, the demographic bulge may lead to a deterioration of the relative wage and employment rates of Aboriginal workers in western cities (Drost 1995; Statistics Canada 1993). Within the labour force, Aboriginal people are under-represented in managerial, teaching, and health-related occupations, and they are over-represented in services, trades, processing, and operative (transportation, equipment-operating) occupations. Consequently, Aboriginal people have lower incomes than other Canadians (Campbell 1986). The average employment earnings of Aboriginal women are also lower than those of other women in Canada and of Aboriginal males (SIAST 1996; Statistics Canada 1993). Aboriginal people also have higher rates of incarceration, accident, morbidity, premature mortality, and other negative life outcomes. Education appears to be especially important for Aboriginal people, as failure to complete secondary schooling is more strongly linked to unemployment than it is for non-Aboriginal people (Canada 1991; Drost 1995; Statistics Canada 1993). Evidence indicates that education and training removes some of the inequities in employment and income levels between the Aboriginal population and the population overall (O'Brien 1992; Poonwassie 1993; Sarkar and Stallard 1997). University graduation is also associated with more stable patterns of employment and higher earnings (Sarkar 1995; SIAST 1995; Smith 1997; Statistics Canada 1993; Yau et al. 1993). At the same time, education is not a guarantee of better outcomes. Lavin and Hyllegard (1996) examined the long-term outcomes of students who benefited from the 1970 open access policy admitting anyone with a high school diploma to the City University of New York. They concluded that open access was beneficial to a large number of

122 Atlanta Sloane-Seale, Lori Wallace, and Ben Levin

students who otherwise would have been excluded from higher education. Their study also found that initial disadvantage continued to affect many people's lives, despite their additional education. Similar patterns are to be expected among disadvantaged adults in Manitoba, many of whom come from Aboriginal backgrounds. While additional education can improve life outcomes, the impacts of early and continued disadvantage cannot be overcome entirely by any one element of the social structure. The Manitoba Access Programs Beginning in the early 1970s, Manitoba's Department of Education developed a series of Access programs that provided post-secondary education for disadvantaged populations. Over fifteen years, a total of seventeen separate programs in different fields of study were implemented at various colleges and universities. Fourteen of these programs are still in operation. Some programs provide support for specific credentials in education, social work, the health professions, and engineering, while others provide general support across a range of fields (Unruh and Levin 1990). Each Access program is delivered by a college or university but is funded specifically and separately by the government of Manitoba. Each program provides support to students in academic, financial, and personal matters so that they can obtain the same qualification as do all other students. Some programs are delivered onsite in northern Manitoba or inner-city Winnipeg, while others are on the main campuses of the institutions. Two fundamental principles guide the Access programs: equality of access and equality of condition. Equality of access refers to changes in the structure of educational delivery so that those who have not had the opportunity to participate successfully may do so (Anisef 1986). It not only means that more people will enrol in post-secondary education; it also means that people with characteristics that have systematically disadvantaged them will have the opportunity to enrol and succeed. Since students require different learning strategies, time frames, and amounts and types of support, these individualized supports must be provided. Equality of condition entails a shift from educational standards based on entrance requirements, to standards based on exit requirements. It implies a rethinking of the role of institutions in curriculum, instruction, support services, and general orientation towards stu-

Post-secondary Education of Disadvantaged Adults 123

dents, so that the institutions are responsible for success as well as entry. It moves institutions from an 'opportunity' model towards a 'success' model. The objectives for the Access programs are to: • increase the number and range of post-secondary education and training opportunities for excluded individuals; • do so under conditions that make it likely for learners to succeed and find employment in their chosen professions; • contribute to the development and self-sufficiency of the community; and • bring about educational change. The University of Manitoba Access Programs (UMAP), Special PreMedical Studies (SPSP) Program, and Professional Health Program (PHP) were inaugurated in the 1970s. These programs provide credentials in education, social work, health professions, engineering, and other fields such as science, arts, and law. Applicants to the programs must meet the University of Manitoba admission criteria in order to be accepted. As well, broader-based admission criteria such as work experience and community leadership are considered. Priority is given to applicants who are also Aboriginal, particularly those living in remote communities who are motivated but lack academic preparation, or financial, social, and personal resources. In addition to the general principles and objectives already noted, UMAP and SPSP embody several other principles. They provide access to the same degrees and diplomas that are available to other students; conduct extensive recruitment; select students who have a reasonable chance to succeed with the supports of the programs; provide an integrated set of academic, personal, and financial supports; redesign curriculum and teaching, where appropriate; and ensure that students are supported in achieving the same high standards as all other students. The students who participate in these programs lack academic preparation, financial, social, and personal resources. They take, on average, an additional eighteen months to complete their degrees. For premedical studies, for instance, a first year chemistry course that is normally offered over one academic year is extended over two years in order to build in the required academic tutoring. Despite these obstacles, a formal review concluded that about 40 per cent of the 2,400 students admitted to the programs had graduated. This is an unparalleled

124 Atlanta Sloane-Seale, Lori Wallace, and Ben Levin

achievement for programs that admit only disadvantaged students (Hikel 1994). The Studies Our team undertook studies of adults who applied to the UMAP and SPSP at the University of Manitoba in order to investigate the factors that help adults overcome disadvantage. Specifically, the studies investigated these applicants' educational and economic outcomes, as well as their decisions and experiences, in order to help educators and government agencies understand how to act more effectively in developing opportunities for Aboriginal students' education and employment. Given the nature of the programs and the application process, the participants in the studies were, by definition, disadvantaged learners. Although we used the University of Manitoba Access Programs as a base for our research, it is important to note that the studies were not an evaluation of these programs. Rather, expressed interest in the programs by potential students provided a way to identify a cohort of disadvantaged Manitoba adults whose long-term economic and educational outcomes would be of interest. Two studies were conducted. The first involved structured telephone surveys with as many former applicants as we could contact, using information from program files. The second involved life-history interviews with twenty-four persons selected from among those who took part in the telephone survey. The life-history approach employed, similar to Butterwick's study reported in this volume, was a qualitative method aimed to give voice to the personal meanings and understandings of participants' lived experiences. We began by identifying persons who had applied to the programs in 1983,1987, or 1991. In choosing these years, we were balancing the need to go back far enough to allow a sense of life course and outcomes to emerge, with the need to locate those involved. Four hundred and seventy-one names were drawn in roughly equal numbers from program participants and non-participants in those years. The telephone survey consisted of four parts: education and training experience, work experience, skill acquisition and use, and demographic information. A number of the questions were the same as those asked in other Western Research Network and Education and Training (WRNET) studies. We also used many questions from the Sta-

Post-secondary Education of Disadvantaged Adults 125

tistics Canada graduates study (1997). The instrument was pre-tested with five persons representative of the sample, and a number of questions were altered to take into account the particular circumstances of Aboriginal students, students from northern Manitoba, and those who had come from disadvantaged circumstances. Obtaining data from files more than fifteen years old, and drawn from a mobile, hard-to-reach population, proved even more difficult than we had anticipated. We invested considerable effort in trying to contact members of the sample, including calls to home communities and efforts to locate people via other program participants and applicants. Despite these efforts, we were only able to complete telephone surveys with 103 students, representing a response rate of only 22 per cent. The life-history interviews were primarily conducted face-to-face, with several completed by telephone in instances where participants were not otherwise available. The intent was to learn in much more detail about participants' life trajectories and to gain access to the personal meanings by which participants construct their world. The first author conducted the interviews over a three-month period. Each of the approximately ninety-minute interviews was audiotaped and later transcribed. Before the data were analysed, the interview transcripts were mailed to participants for further input on accuracy and completeness. We sought to ensure wide variation among the participants with respect to criteria such as employment sector and status, completion versus non-completion of advanced education, and participation versus non-participation in the Access programs. Based on these criteria, twenty-four participants were selected and twenty-three were eventually interviewed. Findings The survey results cast light primarily on demographics, educational and employment outcomes, and skill use. The life-history interviews illuminate aspects of experience such as motivation, barriers encountered, supports required, long-term career goals, and success. Aggregate data are presented for the telephone survey, and the participants' own words are used to present the life-history interviews. The confidentiality of information and anonymity of participants have been protected by the alteration of names and other personal information.

126 Atlanta Sloane-Seale, Lori Wallace, and Ben Levin

Given the relatively small number of responses to the telephone survey, we do not report data by specific years of entry into the program (i.e., 1983, 1987, and 1991). The terms 'accepted' and 'non-accepted' refer to whether respondents were admitted into an Access program in the year in which they applied. Demographics, Education and Employment, and Skill Use1

Fifty-eight per cent of our survey respondents were female; about onethird of males and 40 per cent of females had been accepted into the Access programs. About half the women and 63 per cent of the men were married or had a common-law partner at the time of their application. Women were much more likely to have dependent children. More than 95 per cent of accepted applicants and almost 90 per cent of non-accepted applicants identified themselves as Aboriginal (status or non-status Indian, Metis, or Inuit). Both groups of applicants reported quite low levels of parental education; more than 60 per cent had parents with less than complete secondary schooling. The survey also collected data on the education and training experiences of respondents since the time of application to the Access programs. Data collected included types of programs, completion rates and satisfaction, motivation for taking the program, full-time or parttime status, finances, and continuing education experiences. It is particularly noteworthy that high proportions of non-accepted respondents, as well as those accepted into the programs, had pursued at least some form of post-secondary education since the time of their application. Moreover, both groups gave similar reasons for their interest in postsecondary education, focusing mainly on obtaining work and on being able to help their home communities. It may be that the process of application was itself a stimulus for further education, or that the Access application process focused on people who were already motivated to pursue post-secondary education. At the same time, the survey data support the view that motivation alone, without adequate programs and supports, is insufficient to ensure student success. Though many non-accepted students reported good education and employment outcomes, such as completion of credentials and full-time work in a related field, accepted students had stronger outcomes on all of these measures. The accepted students were more likely to pursue university studies, more likely to complete the program of studies they began, more likely to have been full-time stu-

Post-secondary Education of Disadvantaged Adults 127

dents, and less likely to have been dependent on employment income. They reported higher levels of employment, less unemployment, and fewer job changes, and were also more likely to be in jobs related to their education. Although both groups reported a wide range of occupations and incomes, the occupational and income profile, especially of accepted respondents, was considerably better than what would have been predicted on the basis of their entering characteristics, such as family education and income. For example, current yearly salaries ranged from $12,000 to $96,000, with a median of $34,500 and a standard deviation of almost $14,000. Twenty per cent of respondents were in jobs with annual salaries less than $25,000, and 21 per cent were in positions with annual salaries above $45,000. These data suggest that many but certainly not all of these individuals were able to 'beat the odds' in overcoming, at least to some extent, their initial disadvantages. Respondents in this study, as in all of the Manitoba WRNET studies (Taylor 2000) were asked about the importance of six skills (technical, communication, mathematical, design and building, working with others, analysing and using information) in their current jobs, and the extent to which education, work, life experience, and volunteer work contributed to their developing these skills. Respondents indicated that communication, working with others, and analysing and using information were by far the highest priority rankings in relation to their current work. Nearly all respondents rated these skill areas as important. When asked to identify the single most important skill set, participants indicated that communications skills were by far most important, with more than 50 per cent placing these first. Respondents were also asked about the aspects of their experience that developed the skills rated as most important in their current work. They were asked to comment on education and training programs, on-the-job training, and general life experience. Ninety-seven per cent rated life experience as very important or important, compared with 83 per cent for on-the-job training, and 79 per cent for education and training. This pattern is consistent with other research that shows the high priority people place on direct experience as a source of skill development. Motivation, Barriers, Supports, Long-term Goals, and Success2

The life-history interviews cast light on aspects of experience that alleviate disadvantage. The most compelling motivations for almost all

128 Atlanta Sloane-Seale, Lori Wallace, and Ben Levin

participants were the desire to escape, through education, the impoverishment and lack of opportunity on reservations, and the desire to contribute to their communities. The following statements illuminate these motivations: I feel that if I had not made the effort to leave the community [in the North] I would have ended up like my friends ... who were having children, into alcohol, and not going anywhere. I did not want that kind of life. My parents ... were both alcoholics and I grew up in spite of them, not with their help ... I do not have one friend left from when I was growing up [and that included high school] who is not indigent, in prison, or alcoholic. When you have completed your journey in life, hopefully you would have made improvements on the road itself. Helping others is important because someone has helped you.

Participants experienced many barriers in pursuit of their post-secondary education, including those that were dispositional, situational, and systemic. The problems of lack of academic preparation, lack of selfesteem, lack of child care and social supports, and racism were common. Another frequent problem encountered was the dislocation resulting from the move to a large city from a small, isolated community: I felt alienated from everyone because when you come from a reserve it is a culture shock. The relocation was a terrifying experience because it was hard to leave family and friends. I lacked self-esteem ... I had to learn the [dominant] culture and get accustomed to what is acceptable.

Many participants spoke of the lack of educational and employment opportunities in their home communities, as well as the lack of role models: In my family, education [for the children] was very important to my father and mother. My parents had Grade 6 or 7. They got us off to school. They talked about wanting a better life for us. As a child, I watched my father work hard. He had to get up at five in the morning to do his stuff [trapping] outside in the cold weather. I did not want this life for myself.

Post-secondary Education of Disadvantaged Adults 129 Coming to university was never my plan in life. It was just too far away for me. It was pie in the sky. Nobody I knew or heard of had a university degree. The University of Manitoba was some place that rich kids went to, and we were far from rich.

The programs provided a number of supports, but sexism and racism were nevertheless experienced by a number of participants: The instructor was sexist. I did not feel comfortable approaching him or the other instructors ... We [the women students] felt that we were not taken seriously. We were called 'girls' countless numbers of times. I finally told the instructor that I was a grown woman with children and if he had trouble calling me a woman he could call me 'lady.' He did not call the guys 'boys' - if he had, maybe I would have accepted it. They do not do a good job teaching cross-cultural issues ... whether that would change the racism, I do not know. Racism is systemic. That is an issue that has to be addressed ... In health care, a big portion of the clientele is Aboriginal. There is no understanding of the traditional component ... I was a little dismayed that one of the professors had no clue as to the issues and needs for First Nations people.

These difficulties tested participants and created a need for supports. The programs provided a constellation of necessary supports, including counselling, academic, financial, and social supports. Even with these supports, however, some participants, including the academically prepared participants, experienced emotional difficulties. Virtually all participants reported that, in retrospect, most of the programs' supports were needed but that emotional supports and a sense of community were also important: I ran into a bit of an identity crisis ... Being native was very different... If there was anything that I needed, probably some peer support... Friends and extended family were very important to me.

Participants developed help-seeking strategies that were important elements in their achievements. Some formed their own networks of peer support; others used their traditional spirituality, or drew upon their internal resources. Believing in personal control and a larger purpose in life, however, appeared to be very important to their success:

130 Atlanta Sloane-Seale, Lori Wallace, and Ben Levin I started participating in some spiritual things that helped me to see life in a different light... and allowed me to get back my focus ... I had to make a decision at some point as to what I was going to do with my life. I decided I was here for a reason, a purpose, and that I have to carry on. I received encouragement and understanding from the counsellors. Students in the program were also ... mentors and would give you encouragement. They were role models ... Some strategies were taking pride in myself, acceptance of myself, self-talk ... telling yourself that you can do it.

Participants' formal education greatly enhanced their long-term career and employment prospects by assisting them in developing knowledge and skills in interpersonal relations, technical and procedural areas, organization, self-reliance, and taking a broader perspective. One participant working in Aboriginal health said, 'I [now] have the clinical skills and academic background ... I have a good understanding of not only the individual but also the public health issues within the First Nations communities ... [including] some of the social and economic issues/ In almost all cases, participants' long-term career plans involved continuing their education, becoming role models, further developing their abilities, and improving the lives of families or members of their home communities: I would like to be a principal of a large high school, perhaps to be a superintendent. That would be my ultimate goal. I want to help others, to give something back to my community and family ... I grew up with those values from my parents and grandparents.

Factors of initial disadvantage, such as poorly educated parents, low incomes, difficult family circumstances, or lack of academic preparation, affected the lives of these participants. Many required remedial help or took extra time to complete their educational programs, and most needed financial, emotional, and social supports. Still, these participants' employment and education results have significantly improved since their application to the Access programs. One participant, now working as a teacher, stated, 'My parents ... did not want me to leave because I was babysitting my cousins and three children. They said, "You have a job - you do not have to go." I wanted more out of life.' And another working as an environmental scientist indicated, 'I will

Post-secondary Education of Disadvantaged Adults 131

progress in my field ... I have a good reputation ... I have had opportunities from other organizations... I have been asked to apply for an executive director's position; they feel that I am the right person for the job/ Through her education, one participant escaped the cycle of poverty, pregnancy, and alcohol in which her peers were stuck. She said, 'Of the thirty-four students in my Grade 9 class only five, including myself, graduated from high school/ She overcame enormous challenges to complete her degree. In addition to the financial and emotional challenges she faced, she also fulfilled childcare responsibilities. She indicated, 'You learn to deal with them if you want to be successful ... I wanted to, so I persevered/ In her own eyes, this participant is a success. She beat heavy odds to leave behind a situation that was destroying her community. Most participants reported that they are employed in positions that require the use of the knowledge and skills gained in their education programs, and a substantial number reported having received promotions, salary increases, or recognition. Most are continuing their education and working towards leadership positions, but are also seeking balanced lifestyles. They are contributing members of society and are willing to help others, their people, and their communities. To these participants, success is not measured only by salary, job title, and status, but also by quality of life and commitment to their community. One participant reported: 'I have been head-hunted from several people across Canada ... It means moving ... and a considerable pay increase, but I am not interested ... I have other goals right now/ In summary, the typical respondent in this study is a woman aged twenty-five, living alone with dependent children. She speaks English as her first language and identifies herself as a member of an indigenous Canadian people (Aboriginal, Metis, First Nations, or Inuit), and as a member of a visible minority. She is currently employed at one permanent job, earns an annual salary of $25,000 to $40,000, and depends heavily upon her communication skills in her work. If this woman was accepted to an Access program, she was more likely than a non-accepted applicant to have finished the undergraduate degree program in which she enrolled. She is unlike her nonaccepted counterpart in that her degree was a necessary qualification for her current job. She has also done more than her non-accepted counterpart to continue to upgrade her skills and qualifications through continuing education programs. She is satisfied with her current job but less so with her income.

132 Atlanta Sloane-Seale, Lori Wallace, and Ben Levin

Although this woman has had periods of unemployment in her work life, she (unlike her non-accepted counterpart) has never been out of work for longer than twelve months. The funding that she received in the Access programs allowed her to complete her studies as a full-time student without incurring the debt and depletion of employment earnings experienced by her non-accepted counterparts. Discussion Understanding the lived experience of disadvantaged and Aboriginal learners is a formidable task. Most university researchers, ourselves included, do not come from Aboriginal communities nor have we directly experienced the impoverishment and abuse that are unfortunate characteristics of many economically depressed areas, including many First Nations' communities. The experiences and insights shared by the participants in this study provided us with a humbling view of our own middle-class educational achievements, and admiration for the determination of disadvantaged and Aboriginal learners to overcome the considerable obstacles to their pursuit of post-secondary education. It is unfortunate that there is so little research with respect to the educational and employment histories of Aboriginal people. Although we cannot make generalizations from these studies, a number of observations with respect to demographics, educational experience, and work can be drawn regarding the fit between the data and the picture presented in the literature. The findings in our studies are consistent with the literature with respect to the observation that education, particularly at the post-secondary level, can shrink the gap in employment and income levels and provide the best means for social mobility for Aboriginal people (Gerardi 1996; Sarker 1995; SIAST 1995,1996; Smith 1997; Statistics Canada 1993; and Yau et al. 1993). For many of the participants in these studies, access to advanced education allowed them to earn formal credentials that were fundamental to labour market success. One cannot be a physician, lawyer, teacher, engineer, or nurse in Canada without these qualifications. The study participants had previously been excluded from the possibility of earning these credentials and holding the jobs that go with them. More generally, with respect to the relationship between post-secondary education and job skills, participants reported that their post-secondary education developed in them not only interpersonal,

Post-secondary Education of Disadvantaged Adults 133

technical, management, and research skills but also self-direction, selfconfidence, and the ability to critically assess information. Low socio-economic status is one of the strongest determinants of talent loss, particularly among Aboriginal people (Hanson 1994; Plank and Jordan 1997). In our studies, disadvantaged learners achieved considerable success in their careers - a finding that points to the unrealized potential of many disadvantaged learners. While the impacts of early and continued disadvantage cannot be overcome entirely by any one element of the social structure, additional education and supports can undoubtedly significantly improve employment outcomes. Most of the participants in the study demonstrated what we would consider (but did not measure) a high need for achievement and a low need to avoid failure. Interestingly, educational psychology research (Woolfolk 1995) suggests that persons with a low need to avoid failure tend to choose challenging situations in which they have reasonable chances of success, whereas persons with a high need to avoid failure tend to view risk as threatening, and therefore avoid risk or choose low-risk situations in which they are assured success. Participants saw in the Access programs an opportunity to achieve concrete goals, and persevered against what, in many cases, were tremendous odds. They demonstrated insight into their strengths and weaknesses, and took personal responsibility for their hard work, effort, abilities, and successes as well as for their shortcomings. It is clear that motivation is one compelling factor that contributed to these participants' achievements. In terms of the participants' orientation to learning it appears that, almost without exception, they were goal-oriented. They saw education as the opportunity to escape the problems of poverty and its associated ills that affect many other communities. They enrolled in postsecondary education in order to increase their career and socio-economic prospects, not to achieve social goals or for the joy of learning in and of itself. This finding is consistent with the literature (Gerardi 1996; SIAST 1995). The Access programs moved students beyond disadvantaged family circumstances by building in financial, personal, and academic supports that contributed to students' academic achievements and subsequent career employment. Accepted participants have completed more advanced education because of these supports and therefore are in professional occupations that tend to pay more than other fields. These results support previous findings regarding the efficacy of the

134 Atlanta Sloane-Seale, Lori Wallace, and Ben Levin

programs in compensating for disadvantage (Alcorn and Levin 2000, Hikel 1994). The fact that most participants (even those who were not accepted into the programs) have high completion rates for educational programs and stable employment also likely relates to the participants' high degree of motivation. These learners place a high value upon education as a means to succeed in life and to improve their life chances. With respect to obstacles to success, these participants encountered many situational barriers, usually in the form of financial difficulties, poor academic preparation, social isolation, backgrounds of abuse and poverty, and the conflicting demands of study and family. While lack of confidence was a dispositional barrier for some participants, a number of others displayed considerable confidence in their ability to succeed in education and career. Few mentioned institutional barriers, although some participants identified racism and sexism as systemic problems in the programs. Having achieved a level of education and employment that is rare in their home communities, these participants now see themselves as role models who are in a position to assist their home communities. While this finding is consistent with the literature (Poonwassie 1993), it is ironic in that few participants have actually been able to return home to work. Their absence may then further impoverish their communities, even though they contribute very directly to the larger Aboriginal community. The participants in these studies are employed, contributing members of society. For these individuals, however, an objective measure of success that is based entirely on salary, job classification, and status may be inappropriate. A more realistic and appropriate measure of success may include not only status but also the ability to maintain a balanced lifestyle. This expanded definition of success is more intrinsically based and subjectively defined; less emphasis is placed upon financial reward and more is placed upon how success is experienced in their career and lives. In the aggregate, our data clearly show that access to higher education resulted in better life and employment outcomes, even for persons who were initially considered unqualified for that education. At the same time, our data reinforce the view that the relationship between education, employment, and other life outcomes is not a simple one. For some participants, the same factors that led to their initial exclusion from higher education continued to pose barriers. In other cases,

Post-secondary Education of Disadvantaged Adults 135

participation in advanced education led to a rethinking of life goals and purposes, and changes in life direction, and not always in the direction of a more lucrative career. For some, higher education did not turn out to be the path they were seeking. For disadvantaged groups, greater access to higher education in a supported environment is a benefit. In each life, however, structure, agency, and chance are intertwined in ways that are difficult to disentangle. Notes 1 These data are reported in more detail in Cooke et al. (2000). 2 These data are reported in more detail in Sloane-Seale, Wallace, and Levin (2001).

References Alcorn, W., and B. Levin. 2000. Post-secondary education for indigenous peoples. Adult Learning 11(1): 20-5. Anisef, P. 1986. Accessibility to postsecondary education: A review of the literature. Ottawa: Secretary of State for Canada. Armstrong, R., J. Kennedy, and P.R. Oberle. 1990. University education and economic well being: Indian achievement and prospects. Ottawa: Indian and Northern Affairs. Campbell, P.B. 1986. Outcomes of vocational education for women, minorities, the handicapped, and the poor. Columbus: Ohio State University, National Center for Research in Vocational Education. Canada. 1991. Pathways to success: Aboriginal employment and training strategy. Ottawa: Department of Employment and Immigration. Cooke, K., L. Wallace, A. Sloane-Seale, and B. Levin. 2000. Manitoba Access program: Educational achievement and outcomes. Paper presented at Western Research Network on Education and Training (WRNET) Annual Conference (24-5 March), Vancouver, BC. Cross, P.K. 1981. Adults as learners. San Francisco: Jossey Bass. Drost, H. 1995. The aboriginal-white unemployment gap in Canada's urban labor markets. Social Policy Challenge Series 11:13-57. Gerardi, S. 1996. Factors which influence community college graduation. New York: NYC Technical College, CUNY. Hanson, S. 1994. Lost talent. Sociology of Education 67(3): 159-83.

136 Atlanta Sloane-Seale, Lori Wallace, and Ben Levin Hikel, R. 1994. Review of access programs. Toronto: KPMG Management Consultants, July. Inbar, D., ed. 1990. Second chance in education: An interdisciplinary and international perspective. London: Palmer. Lavin, D., and D. Hyllegard. 1996. Changing the odds: Open admissions and the life chance of the disadvantaged. New Haven: Yale University Press. McClure, K., L. Wallace, A. Sloane-Seale, B. Lavallee, and B. Levin. 1999. Work and education for disadvantaged adults in Manitoba. Paper presented at Western Research Network on Education and Training (WRNET) Annual Conference (26-7 February), Vancouver, BC. O'Brien, E.M. 1992. American Indians in higher education. Washington, DC: American Council on Education, Division of Policy Analysis and Research. Plank, S.B., and W.J. Jordan. 1997. Reducing talent loss: The impact of information, guidance, and actions on post-secondary enrollment, report #9. Baltimore, Md.: Center for Research on the Education of Students Placed at Risk. Poonwassie, D.H. 1993. Higher education for native students in Manitoba universities: The quest for equal access. In Issues in the history of education in Manitoba: From the construction of the common school to the politics of voices, ed. R. del C. Bruno-Jofre, 405-28. Lewiston, N.Y.: Edwin Mellen Press. Sarkar, G. 1995. Employment prospects of Aboriginal graduates from technical programs: A four-year trend study. Regina: Saskatchewan Institute of Applied Science and Technology. Sarkar, G., and C. Stallard. 1997. Do equity groups have a fair chance in our education system? Paper presented at CIRPA Annual Conference (October) Toronto. Saskatchewan Institute of Applied Science and Technology (SIAST). 1995. Aboriginal graduate employment statistics report 1994. Regina: Saskatchewan Institute of Applied Science and Technology. - 1996. Economic Overview 1996. Regina: Saskatchewan Institute of Applied Science and Technology. Sloane-Seale, A., L. Wallace, and B. Levin. 2001. Life paths and educational and employment outcomes of disadvantaged learners. Paper presented at Western Research Network on Education and Training (WRNET) Annual Conference (30-1 March), Vancouver, BC. Smith, T.M. 1997. Minorities in higher education. Washington, DC: National Center for Education Statistics. Statistics Canada. 1985. A demographic overview of the Native populations in Alberta. Background paper no. 1. Edmonton: Alberta Department of Native Affairs. - 1993.1991 Aboriginal peoples survey: Schooling, work, and related activities, income, expenses and mobility. Ottawa: Special Surveys Division.

Post-secondary Education of Disadvantaged Adults 137 - 1997. Survey of graduates. Ottawa: Special Surveys Division. Taylor, L. 2000. Transition and skills development through education, training and work experiences: A follow-up study, Seven Oaks School Division. Paper presented at Western Research Network on Education and Training (WRNET) Annual Conference (24-5 March), Vancouver, BC. Unruh, Dv and B. Levin. 1990. Equality of access and equality of condition: Second-chance programming for success. In Second chance in education: An interdisciplinary and international perspective, ed. D.E. Inbar, 253-66. London: Palmer. Woolfolk, A. 1995. Educational Psychology. 6th ed. Boston: Allyn and Bacon. Yau, M., et al. 1993. The 1991 every secondary student survey, part 3: Program level and student achievement no. 205. Toronto: Toronto Board of Education, Research Department.

5 What Outcomes Matter to You? Exploring Welfare Policy and Programs from the Perspective of Low-Income Women SHAUNA BUTTERWICK

Over the past decade, social welfare policy reform in Canada has focused on two goals: expenditure reduction and moving welfare recipients into the paid workforce. Cost cutting measures have been introduced, as well as measures to assist welfare recipients to enter the paid workforce.1 This chapter looks at the outcomes of such reforms for lowincome single mothers. For this group, the relationship between access to post-secondary education and the acquisition of formal and recognized educational credentials, and the ability to find and keep jobs that pay a living wage and that provide opportunities for work advancement, has been well documented.2 Unlike other chapters in this book, this discussion is not about employment destinations or outcomes of particular educational initiatives; rather, it tells a story of access denied. In this chapter we hear about the struggles of a small group of single mothers on welfare to access education, and to be treated with dignity and valued as citizens. These are views from the margins of the social welfare discourse; they illustrate the frustration and despair as well as the persistence, resistence, and agency of these single mothers. These stories illuminate the problematic outcomes of policies that, for this group of women on low income, are class and gender blind. Examining the material effects of welfare reform from the ground up, from the perspective of the everyday/night lived experience of single mothers on welfare, is an important contribution to the discussion of outcomes. Without such research, a partial and limited view is constructed. The Policy Context: BC Benefits and Quantitative Research Since the 1980s, welfare policy in Canada has been characterized by 'a

Exploring Welfare Policy and Programs 139

crisis in faith in the welfare state' (Guest 1985,235). The Canada Assistance Plan (CAP), which was adopted in the 1960s, provided federal funds to provinces only if they guaranteed that welfare payments would meet basic needs, allowed appeals to welfare decisions, did not insist that welfare recipients work or train while on welfare, and allowed recipients to move from province to province.3 Using reforms in the U.S. welfare system as a model (Swanson 2001; NAPO 1997), governments began to make dramatic changes to welfare policy. In 1996, the federal government replaced CAP with the Canada Health and Social Transfer (CSHT), which removed all of the earlier provisos with the exception of the right of welfare recipients to move from province to province. The government also dramatically reduced transfer payments to the provinces. In British Columbia, welfare reform was one of many other policy initiatives introduced in the mid-1990s by the provincial New Democratic Party, which began with Skills Now! (1994), the labour market and education policy framework which had as its goal the creation of a 'seamless system' of lifelong learning and training. In 1996, the BC Benefits Act4 was introduced, which was designed to help recipients 'develop attachment to the labour force in the shortest time possible' and 'to make work a better deal than welfare and make sure people can get the skills they need to succeed.' Cost-cutting measures dominated the agenda. In an effort to save close to $20 million, payment rates were cut, provincial residency requirements were introduced (and later removed after protest), and new eligibility criteria were put in place restricting welfare access to applicants who were considered to be employable. Welfare recipients were forced into a range of job search and skills training activities. 'Independent Job Search' provided documentation and information sessions outlining effective job search strategies. 'Assisted Job Search' provided for assessment, counselling, and referral. 'Workplace-Based Training' placed recipients with employers who received training credits. 'Employability Skills' determined the gaps in recipients' skills and experience, and directed them to various programs in order to fill those gaps. Job Readiness programs, Adult Basic Education, and Volunteer Incentives (work opportunities through volunteer experience) were also mandated under this rubric. Significant savings and caseload reduction were achieved by limiting access to social assistance; however, there was no reduction in the proportion of long-term cases and no increase in the numbers leaving

140 Shauna Butterwick

social assistance5 (Mansfield and Bell 2001). The dominant concern with the fiscal outcomes of policy initiatives of a federally funded project were also noted. Evaluations of the Self Sufficiency Project (SSP)6 offered in British Columbia and New Brunswick indicated that 'SSP increased employment, earnings and income for recent IA applicants, but it did so without costing the government extra cash transfer payments' (Michalopoulos and Hoy 2001,1). Research on other outcomes of welfare reform is, unfortunately, sparse; there is no data on what happens to those who do leave welfare in BC7 (Klein and Montgomery 2001). Given that the stated intent of these programs is to move people from welfare to work, this 'raises questions about whether the actual motivation behind these policies relates more to cost cutting and political agendas than to assisting people out of poverty' (8). Using a model that examined the relationship between the wage rate and the number of employed workers, these authors examined the impact of cuts to welfare on earnings of lowwaged workers: Low-waged workers find themselves competing with former welfare recipients for jobs. If all those that leave welfare were to become employed, in the absence of other changes to the larger economy and labour market, the result would be to depress wages for lowwage workers. In this way, welfare cuts harm not only those on social assistance, but also the working poor' (4). In contrast to the 'good news' that focused on 'savings/ the Social Planning and Research Council (SPARC) of British Columbia found that BC Benefits rates (basic income assistance) were insufficient to maintain people at Statistics Canada's minimum standard of living (Goldberg and Long 2001). Furthermore, this study argued that the ability of BC Benefits to cover the cost of living had declined significantly for single parents with a five-year-old child: 'This report clearly demonstrates that income assistance in BC is not only inadequate, but actually remains on a downward trend through the end of the 1990s and into the new millennium' (20). This kind of policy research on caseloads, wages, and living standards is important, but partial. It does not examine the lived experience of those on welfare; it is here that the 'outcomes' are of course most dramatic. The rest of this chapter explores a different research approach, one that includes the voices of those who feel a direct impact of changes to welfare policy. Research that brings forth the experiences of those living directly with the results of policy reform is a small move towards democratizing the process of policy-making.

Exploring Welfare Policy and Programs 141 Activist Policy Research

Research that explores the everyday lived reality of single mothers on social assistance adds a crucial dimension to knowledge about the outcomes of welfare reform. Qualitative policy-oriented research '[uncovers] the meaning of social events and processes, based upon understanding the lived experience of human society from the actors' point of view' (Finch 1986,7) and 'bring[s] different skills and strengths to what are admittedly difficult and complex issues' (Rist 2000,1001). Dorothy Smith, a feminist sociologist, calls for research that begins with women's everyday/night experiences, for research that works with women rather than seeing them as objects of investigation. This approach also explores how 'the matrix of our experience is organized by relations tying it into larger processes in the world as well as by locally organized practices' (Smith 1987,10). An important goal of her research methodology is to explicate 'how the organization of social relations has accomplished [women's] exclusion' (78). She uses the term 'social relations' in order to 'direct attention to, and take up analytically, how what people are doing and experiencing in a given local site is hooked into sequences of action implicating and coordinating multiple local sites where others are active' (Smith 1999, 7). In other words, we need research that examines what appears to be the 'choices' of individuals (e.g. welfare recipients) and that traces how such choices or decisions are implicated in the actions of others (policy-makers, bureaucrats). Nancy Fraser (1989) also calls for meaning-oriented kinds of inquiries into welfare policy in order to 'make explicit the social meanings embedded within welfare programs, meanings that tend otherwise simply to go without saying' (146). Politicizing issues, calling for different solutions, and democratizing the decision-making process are key struggles, Fraser asserts, for feminists involved with the 'welfare wars.' Yeatman (1998) extends this approach in policy analysis, calling for policy activism that 'opens it up to the appropriate participation of all those who are involved in the policy process, all the way from points of policy conception to delivery on the ground' (10). Including the perspectives of single mothers on welfare in a discussion about policy outcomes is particularly challenging given the neoconservative orientation that now shapes many governments and policy agendas. As one of the women I spoke with indicated: 'I am an antipoverty activist, a single mother with two pre-school children, somone

142 Shauna Butterwick

on the receiving end of welfare and training policy decisions, who has never had a voice at the policy-making table/ This kind of research aims to bring her voice to public policy discussions and to explicate what Fraser identifies as 'the tacit norms and implicit assumptions' (146) that inform welfare practices. Beginning with the Everyday The stories outlined in this chapter were generated from a day-long focus group meeting, individual interviews, and my involvement with a small group of anti-poverty activists. In 1998, ten single mothers on social assistance met with me, my research assistant, and a community organizer from an anti-poverty agency.8 Participants discussed their experiences of the outcome of welfare policy, and spent considerable time discussing how policies affected their access to further education and training. Additional conversations were held between 1998 and 2001 with a collective of four single mothers on welfare.9 The ages of the participants in these two groups ranged from late twenties to early forties. They had an average of two children whose ages ranged from preschool to young adult. All of these participants were Caucasian except for one First Nations woman and two women of colour; most had some kind of post-secondary education. They lived in various regions of the greater Vancouver area. The collective met regularly in one of the member's homes or at a local coffee shop, and I attended meetings every four to six weeks from 1998 to 2000.10 As I began to work with the collective, a collaborative research relationship emerged. I would use, where possible, my location and academic resources to support their efforts, and I would also document their explorations, their testing of ideas, and their changes in direction. Efforts were made to shift away from a traditional hierarchical researcher-researched relationship towards a more communitybased, collaborative orientation. This was not just a community-based project, but a community-oriented approach which emphasized processes that supported the formation of community - the 'common unity' of all participants - and ways that strengthened the democratic, equitable, liberating, and life-enhancing qualities of social life (Stringer 1999,28). It must be noted, however, that although the players involved in this community-based collaborative project were fully committed to building mutually beneficial relations, principles alone cannot equalize the difference in material locations between a privileged academic and a group of single mothers on low income.

Exploring Welfare Policy and Programs 143

In what follows, the women I spoke with outline their lived experience of the outcomes of welfare policy reform. Central to their stories is their work as mothers and the obstacles they encounter in a system that fails to acknowledge and respect that work. They describe a nowin situation akin to 'close all the exits and then yell fire/ Their stories also illustrate their struggle to live with dignity in the face of increased alientation, self-doubt, and poverty. These women and their children face multiple barriers that block their full participation in civil society. This represents a loss of opportunity for them individually, but the larger community also loses. These experiences and the meanings of policy outcomes are rarely explored, much less highlighted in traditional policy evaluation and policy-oriented research. Women with Children Not Wanted A dominant theme in the conversations I had with these women was the devaluing of children and those who care for them, which puts them in a position to make 'terrible choices/11 On the one hand, they could get 'off the system' by finding paid work and moving outside some of the surveillance that constantly constrains their everyday experiences. But, on the other hand, this work would most likely be at minimum wage, come with no medical or dental benefits, and would not cover their basic household needs, not to mention the childcare expenses made necessary by their working outside the home. As a result, they would put the health and safety of their children, as well as their own health, at risk. As one group member said, 'Who is going to take care of your child while you are in a program or say going to work for that little money? If you don't have extended family or somebody you can really trust, you end up taking some [stranger] who's willing to work for twenty dollars per day. Or you have your kids spend time alone/ Another member of the group described how finding adequate childcare becomes more difficult within the isolation that occurs on welfare: All the people I used to know who were employed, I don't have contact with anymore, so when it comes to childcare, if daycares are full, you ask your neighbour to take care. If they claim the money, it's all deducted so there's no incentive for people helping people. In that instance you end up working with strangers and you don't know where your kids are going, you don't know the people involved and I find that really scary/ The hours of many daycare centres do not correspond with the work schedules of inflexible entry-level shift work. The daycare system isn't

144 Shauna Butterwick

set up for people to work normal work hour jobs or shift work. Daycares close at six o'clock: "Come and get your kid or we're sending them to social services."' These women are aware that their parenting work is not only devalued, but is also under constant surveillance. As one mother noted, 'We're told we're bad parents when we work. We're told we're bad parents when we stay at home on welfare and raise our kids.' Another clearly articulated the contradictory messages given to single mothers: As a single mother on welfare there is a clear message that this caring work that I do is not valued, it does not count. If I were to participate in a training program or find a low-wage job I would have to find inadequate childcare because the costs of such care are not covered by any assistance I would be eligible for. The messages are completely contradictory. On the one hand, I should stay at home with my children. This is what my employment counsellor told me to do. On the other hand, I am told I am a drain on the system and should go out to work, leaving my children with someone else. If I did that, I would be told that I am a bad mother, leaving my children in inadequate care.

The BC Benefits policy did not require women with children under age seven (or those with dependent children with disabilities) to commit to full-time training and/or work.12 Moving into full-time work and/or training once their children were in full-time school, however, was also problematic for these women. School hours do not match the hours of work of most entry-level jobs in the service sector, which rely on 'flexible' workers who can work at any time during the twenty-four hour day. Finding before- and after-school care was a serious problem for these women. Many jobs in the low-waged labour market are parttime - a cost-saving measure for employers as they do not have to pay benefits. This means that many workers must combine multiple jobs a situation that makes finding quality, affordable childcare next to impossible. These women also spoke about the inflexibility of welfare regulations that could not respond to women's individual circumstances and children's needs: 'Kids are individuals and families are individuals; sometimes you need to be home, and sometimes you don't. I think that could be one of the biggest and easiest things they could do, to allow for a more personalized choice about when you're kids are ready. They're not all ready when they're seven.' Childcare advocates have worked for

Exploring Welfare Policy and Programs 145

years to politicize caring for children, succeeding in bringing this issue, at times, to some public policy tables. Current welfare reform, however, pushes childcare back into the private realm, where it is the responsibility of women themselves, making it more difficult to politicize the issues. The single mothers on welfare that I spoke with challenged the narrow vision of work that excludes caring for children, arguing that effects on children and on the quality of parenting should be a greater concern of policy-makers, and on the policy agenda: Parenting has got to be seen as employment and I think that it's gotta be stressed how parenting benefits the community as a whole. If people choose to parent their children, that should be paid. [It should be] considered paid work/ Welfare has been a key element of the social contract, which is being rewritten; even in its older forms, it did not take into account the caring work of women. The new social contract, Pateman (1992) argues, perpetuates inequalities through the unequal exchange of obedience for protection. The party that provides protection has the right to determine how the other party will act to fulfil their side of the exchange. These women are tired of having to express gratitude; there is little space to express their anger, not to mention their expertise or interest in taking a place at the policy-making table. The Labour of Welfare The women also drew attention to how welfare policy affected their everyday routines. The restrictions, rules, regulations, and surveillance that recipients face, together with the labour required to survive while oh social assistance, made it impossible for them to find the time, energy, resources, and support to make a transition to paid employment, even if they could manage to find and pay for adequate childcare. Surviving on welfare, creating a safe home, and providing food and other everyday needs for a family within the web of restrictions shaping these women's lives is a twenty-four-hour per day job. 'If you're surviving in poverty then that's what you're doing, you're just surviving; it's too difficult, it's a full-time job just to do that.' These women spoke about the multiplicity of tasks and responsibilities they face as mothers on welfare, and the constant challenge of meeting their needs given the current welfare rates. They argue that they are the working poor. T find another really big stress is when you don't have enough to make ends meet ... all the shopping you have to do. You can't just go out and do your shopping for a week or two weeks, you

146 Shauna Butterwick

have to wait for the specials or the coupon for this and a coupon for that - wait for a special to crop up at the thrift shop, so you're constantly at the stupid store ... you're all over the place, you can't just go to Safeway and buy your groceries for the month, that's why it's a fulltime job; [it's] not about getting people off welfare to work, they are already working on welfare.' One of the topics that came up frequently in our conversations was the vulnerability of these women to exploitation. The devaluing of the labour of single mothers on welfare has been the focus of much feminist scholarship. Edin and Lein (1997) have explored how single mothers make decisions about staying on welfare or finding paid work. In their study of 214 single mothers on social assistance, they found that contrary to stereotypes these mothers were careful managers of money. But even with the most frugal of budgets, the mothers could manage only by participating in the 'informal economy' by bartering services like childcare in exchange for household necessities. Those who found paid jobs had to face an even larger gap between income and expenses and had to rely on poor-quality childcare. The low-income women I spoke with similarly described how inadequate levels of monthly welfare income forced them to participate in the informal economy in order to bring in extra cash, which made them vulnerable to the exploitation of others. When they were not paid for their work, such as child-minding for other mothers subsisting in the low-waged labour market, they had no recourse, nowhere to turn to support claims for payment. They were cognizant of the contradictions within this situation. Single mothers working in the low-wage labour market rely on them for childcare, but can pay only a small portion of the value of the work. When these 'working' mothers cannot pay, mothers on welfare absorb the loss. The inadequate welfare payments combined with earnings deductions policies force these women to participate in the informal economy and to be silent about their extra income; this silence further contributes to a perception that the current welfare levels are adequate. Thus policy-makers use this as a rationale to further reduce welfare payments in order to make welfare less 'comfortable.' The women were aware that their efforts to survive and care for themselves and their children were either invisible or regarded with suspicion: Tf I were in another location my efforts to care for myself and my children would be considered as entrepreneurial, evidence of my talents, skills, and knowledge. But I am now a category, single par-

Exploring Welfare Policy and Programs 147

ent on welfare, not a human being. And people, especially those in the welfare system, think they know who I am, because all they see is a category.' Access to Education and Training The women I spoke with also pointed out how contradictory welfare is, and how unlikely it is to achieve its goals of getting them education and training and into the paid workforce, given their life circumstances. The system seemed oblivious to the difficulties faced by mothers with small children who have limited time and access to childcare, as well as limited transportation options. One mother described her frustrations with the rigid bureaucratic processes that welfare recipients were required to follow: 'If you phone your training consultant and she tells you, "Before you speak to me, contact this employment assistance program," then you phone the employment assistance program and the receptionist says, "Sorry, we can't give any information over the phone; you have to come to one of our information sessions," that can last from one to two days. Then, at the end of two days you may get to meet with an employment assistance counsellor, who will then tell you what options you have.' At the time of this study, training and education were available only if welfare recipients made a commitment to either full-time training or finding a full-time job. Part-time training and part-time work were not possible for anyone receiving welfare. This 'all or nothing' policy places single mothers in a particularly difficult position. 'With my two children, I couldn't commit to learning or taking a full-time job. You have to be able to, be ready to take a full-time job before you will get any funding or training.' They also faced other hurdles in order for them to access a limited range of training opportunities: 'It's a drag getting shuffled around from office to person, from office to person, all over the place; you never know your ass from the tea kettle. That's a bad thing, to have to talk to so many people. You tell your story over and over, and that in itself is frustrating and becomes a barrier, because if you're telling the same person, or different people over and over, and they're all like shaking their heads and telling you they can't help, and you go on to somebody else, then eventually your dream dies.' At the time of this study, BC Benefits required all recipients who wanted training to meet with a training consultant, who had the power to determine what kind of training each welfare recipient could access:

148 Shauna Butterwick

'They have too much power over you. They can stop your cheque if they want to.' The women were struck by the limited time they would spend with these consultants, who would then make decisions about their educational future: 'You have these people who hardly know you at all telling you what kind of job you should be doing, what kind of program you should be going into, based on a half-hour meeting with this person/ The women found they needed to keep their dreams out of these conversations: 'You tell them what they want to hear, not what is inside of you. If Sadie said she wanted to be an archeologist they would laugh at her, they would tell her to get real/ The women pointed out how other policies also work against single mothers. Welfare recipients are encouraged to leave welfare and take out student loans. Student loans repayment policy creates significant restrictions that limit these women's opportunity to be independent and to care for their families: 'When I went to university [it was] a time when you could get remission [for your student loan] only if you finished in a short period of time, but they counselled single mothers to only take three courses to keep their GPA [grade point average] up and be a good mother, so you missed the remission and there were no grants. There was a period when I didn't eat/ The student loans rules and regulations position single mothers in a no-win situation even when they are able to enter the paid labour market: The whole student loan needs to be totally changed, revamped. I have to pay $122 a month. It comes out of my bank, I don't even have a say in it; before my rent comes out, my student loan comes out. I make a thousand bucks a month, you know. I'm raising a kid of my own, sending him to school and stuff. It's forty bucks a month in bus fare, just to get my kid to school, right? That's not [including] the clothes, the food, the textbooks, because God knows the school won't pay for that. Tuition needs to be dropped so you're not getting such high debt loads, the student loans program needs to have more remission [opportunities]. They need to do something about the interest rates,' cause that's where most people's enormous debt comes in. They need to [change it] so that a person can pay twentyfive bucks a month on their student loan and if that's it, they take it. Don't say, 'We're gonna take you to court and get the debt mongers after you.'

While the oppportunities for accessing formal training and post-secondary education were constrained, these women experienced a great

Exploring Welfare Policy and Programs 149

deal of learning in their day-to-day struggles. This kind of education often goes unrecognized because it is located outside of formal programs; it is devalued because it has not been sanctioned by the welfareto-work system. Their advocacy and activism, their explorations of ideas for creating an income-generating cooperature venture, and their investigations of various policies were opportunities for learning. As one of the mothers pointed out, the narrow vision of learning, education, and training found in welfare regulations ignores the learning they have experienced and restricts their access to further formal education: Since I was turned away from the employment program, which had zero to offer me anyway even if I had been let in, I've become someone who is working outside the system to build structures and resources, which in itself represents revolutionary steps rather than reform. The experience of having a computer in the home has been a rare opportunity for someone as low income as myself. I've been able to be useful to society in ways that fit my values, and learn about the huge network of groups working around the world to combat globalization. I can bring the world into my home. I'm not afraid of borders anymore. I've learned about ideologies that I would be ignorant of it I didn't have access to a computer.

The opportunity to explore diverse income generating ideas in a cooperative, and to learn in the process, was described by some as the element that should be incorporated into current programs: 'What I want to talk about is my experience [of working collectively on income generating project] over the past three years of learning my way, on my own terms, not based on someone else's imposed view of my interests and capacity. This experience has been much more lucrative, more productive and relevant to my economic situation than my trip to the employment program.' What Kind of Democracy Matters to You? Who Counts and Who Doesn't? Finally, these mothers pointed to the stigma, the devaluation, and the marginalizaton that welfare policies produce. Their citizenship and their ability to participate in discussions about policy were limited by their status as welfare recipients. Single mothers on social assistance are rarely regarded as stakeholders; rather, they are constructed as recipients and clients. As the fiscal crisis of the state grows, they are

150 Shauna Butterwick

increasingly regarded as non-citizens who are less deserving of support compared with those who have paid employment. Jean Swanson (2001) explored the practices, impacts, and origins of what she has termed 'poor-bashing/ noting the influence of the 'work ethic' and the ideas put forward in the 1500s by Martin Luther, who positioned work as the only way to live acceptably in God's eyes: 'Luther's plan to help the poor was an expression of the newly developing Protestant ethic that is still embedded in the way many Canadians think nearly five hundred years later, even though they may not be Protestant or religious at all' (32). Feminist scholars also see the work ethic persisting as a key ideological current in current social welfare reforms. Patricia Evans (1997) draws attention to how difficult it is to disrupt the dominant discourse regarding social program - a response that constructs many social welfare programs as 'passive' and too expensive: 'Paid work, the critical nexus for the modern construction of social citizenship, is becoming more difficult to obtain in the global economy, while at the same time, it is increasingly viewed as a "badge" of citizenship' (106). Evans goes on to note how this view of citizenship is particularly problematic for mothers and those whose main work is to care for others. In the current climate, as Ostner (1994, quoted by Evans) notes, 'those who care for dependents during their life course or who are among those groups ascribed such roles can easily become "laggards" in a competitive economy' (127). The past decade of welfare reforms has intensified this devaluation; paid work and only paid work has become the indicator of value and citizenship. One of the most significant outcomes of welfare reform is the pathologizing of the poor - in particular, single mothers on social assistance. As one participant noted, 'Poor bashing is inherent in [welfare] mandates and philosophies. They talk a lot about attitudinal change and motivation, motivating people to get jobs, and that in itself, that philosophy, it's a lie; it disguises the real cause of poverty. The programs are creating the problem that they claim to repair.' As we have seen, maintaining a sense of self-worth is a daily struggle for these single mothers. When there are few jobs to be had, the welfare-to-work programs generated a sense of false hope that led to further despair: 'Letters came from the FAWs [financial aid workers] ... to go to this employment program. Some of the women got so excited, they got this letter saying somebody's gonna get them work and they're going to get out of this situation, and of course that is not what happens. What does this do for people? When you find the

Exploring Welfare Policy and Programs 151

energy to hope, then it's dashed. It's like [the poor] are different human beings completely so we don't have [the same] expectations for them as the rest of society/ Although, at the time of the study, the welfare-to-work component of BC Benefits was not mandatory for those with children under seven years of age, the women felt a strong coercive element in these ostensibly voluntary programs: 'Why do you need a training consultant if it's voluntary? You can find places, go to the university, talk to a counsellor, people who know how to do it/ They also pointed to how the short-term training they can access while on welfare reflects a view that they are not worth spending time on: 'If you've got low selfesteem you're not going to make it very far in the labour market, especially anything beyond a minimum wage, because you don't think you're worth it. Programs like this are [reinforcing] that perspective that I'm not worth two years of therapy or whatever a person needs. "You've got twelve weeks; get it together and get out and get a job/" This current political climate does not go unnoticed by the women I spoke with. They have to fight, and some do not succeed. There is a powerful push to feel guilty about their circumstances - circumstances not of their own making. The emotional outcomes of feeling guilty and ashamed were a huge burden to bear: T had to declare bankruptcy for student loans [and go on welfare]. I felt this huge burden of guilt... But I have worked as a volunteer for four years averaging a forty-hour week, and the people of Canada are paid back. That's a good feeling; I still have a lot of other emotions, but I don't have guilt anymore/ An orientation towards welfare reform that focuses on incentives and disincentives further blames the victims, ignoring empirical evidence that 'labour market conditions (in particular, unemployment) are most reliable predictions of the incidence of welfare usage. Welfare-to-work programs are therefore inappropriately conceived solely in terms of incentives and disincentives to work, instead of acknowledging a broader set of factors that contribute to social assistance usage' (Klein and Montgomery 2001, 8). The current focus and expenditure aimed at those committing 'welfare fraud' is further evidence of the punitive orientation of government, one which reinforces the dominant view that those on social assistance are undeserving: Tt increases the stress and guilt and fear, which traps you even further in your poverty, and it traps you in a particular market, and you lose your connection to something more viable. I know a woman who lives with her boyfriend who is involved with the black market and who beats her, but she cannot go to the Ministry for help/

152 Shauna Butterwick Conclusion

How welfare reforms led to denied access to post-secondary education for the women I spoke with, stands in sharp contrast to the account by Sloane-Seale, Wallace, and Levin in this volume. These researchers talked to Aborignal participants about their experiences and outcomes from participating in Access Programs. In these programs, Aboriginal students could access the same degrees and diplomas available to other students and be provided with the financial, emotional, and social supports needed for their educational success. Programs with similar principles for single mothers on low income - where both access and success are considered the responsibility of educational institutions - would go a long way towards addressing the barriers described by the women in my study. As Schein (1995) notes in the conclusion to her phenomenological study of thirty women living in poverty, responses to the problems identified through such research must be based in a broad perspective that avoids short-term solutions to complex problems, and embraces long-term, comprehensive solutions. We know what works;13 the major hurdle is governments not acting with the political courage necessary to create policy with a long-term, not just a short-term, view. Efforts to address the needs of women and children living in poverty can be threatening, however, to the interests of those in power. As noted by other researchers (Klein and Montgomery 2001, 153), 'the success of the rapidly expanding service sectors is dependent on the low-wage labour of poor women/ This study raises questions about what outcomes should matter. We live in a wealthy nation where those who are affluent can easily deny the existence of poverty. As Rimstead (2001, 7) argues, 'poverty narratives in wealthy countries such as Canada often unfold a national imaginary which locates the poor outside the imagined community on the fringes as fragments of a nation.' She calls for projects that treat the poor as 'speaking subjects, agents of cultural change, and concrete beings who live and feel social boundaries and material constraints. Poor peple often know about possibilities for social change even when they lack the social power or resistant community to effect these changes' (7). The last word is for one of my co-researchers, who clearly outlines the everyday lived reality of welfare reform, and who, in her own words, calls for research that, as Smith (1987) has outlined, explicates the social relations shaping that everyday world. 'Why write about the

Exploring Welfare Policy and Programs 153

miserable details of my life - it would be just too depressing for me to recount. And why should I reveal these details? Whose life is under surveillance here?' Notes 1 To achieve the first goal, welfare payments have been reduced, eligibility requirements narrowed, and the length of time claimants may receive assistance shortened. In relation to the second goal, job search clubs, employment-related training programs, and other initiatives have been established to help welfare recipients make successful transitions to the paid workforce and to help them achieve independence and 'self-sufficiency.' 2 See, for example, V.C. Adair, 2001. Poverty and the broken promise of higher education, Harvard Educational Review 71, no. 2 (2001): 217-39; M. Gittell, M. Schehl, and C. Fareri, From welfare to independence: The college option. (New York: Ford Foundation, 1990); M. Greenberg, J. Stawn, and L. Plimpton, How state welfare laws treat postsecondary education. (Washington, DC: Center for Law and Social Policy, 1999); K. Kahn, Workfare forces single mothers to abandon college education, Sojourner: The Woman's Forum 24, no. 2 (1998): 31-3; T. Karier, Welfare graduates: College and financial independence. In Policy notes, no. 1 (New York: Bard College, Jerome Levy Economics Institute, 1998); L. Lein, and K. Edin, Making ends meet: How single mothers survive welfare and low-wage work (New York: West End Press 1996); P. Loprest, Families who left welfare: Who are they and how are they doing? In New federalism: National survey of America's families, A-012 (Washington, DC: Urban Institute, August 1999); J. Thompson, Women, welfare and college: the impact of higher education on economic well being, Affilia, 8 (1993: 425^41; L. Wolfe, and M. Gittell, College education is a route out of poverty for women in welfare (Washington, DC: Center for Policy Studies, 1997). 3 Regardless of these provisos, provinces systematially reduced their welfare rates, and by the late 1980s some provinces introduced initiatives like workfare that violated CAP principles altogether. In the absence of federal restraint, welfare cutbacks spead across Canada (NAPO 1997). 4 BC Benefits replaced the previous welfare program called GAIN (Guaranteed Available Income for Need). 5 There were 34,000 fewer cases in August 2000 than in April 1997; however, the percentage of continuing cases increased over each seasonal cycle. 6 SSP, funded by Human Resources Development Canada, provided an

154 Shauna Butterwick

7 8

9

10

11

earnings supplement to a small sample of randomly selected single parents in BC and New Brunswick who had been on income assistance for one year and who had left IA for full-time work. There is also no data collected in Quebec, Ontario, and Alberta. The first focus group meeting came about when I was preparing a background paper for a joint committee made up of federal and provincial representatives (see Butterwick, Bonson, and Rogers 1998). The purpose of the background paper was to summarize what was available in evaluation studies of welfare-to-work programs in several countries and to identify those program elements that were effective. Most studies did not reflect the perspective of single mothers on welfare and so additional literature was sought and a focus group organized. I am indebted to End Legislated Poverty, and to my research assistant, Pam Rogers, for her assistance in organizing and helping to document the discussion of this meeting. Copies of the final government report, which included the issues that the women spoke about, was submitted to government and was also provided for the anti-poverty group who helped organize the first focus group, and to the women who attended. Two of the women who attended the first focus group invited me to continue meeting with a smaller group who had formed a collective. This second group met regularly over a two-year period with the goal of investigating cooperative means to generate income so that they could create a path towards economic independence. The membership of the group changed over time, as some women left and others joined. After three years of attempts at different income-generating projects, the collective disbanded. Conversations with the group's initiator, many of them via email, have continued. Meeting at home was challenging given the needs of the children. In order to meet at a local coffee shop, I reimbursed their childcare expenses and paid for the food. Initially, the meetings were tape-recorded, but most often written notes were taken. Audio recordings of meetings proved to be problematic because most meetings were held either in noisy coffee shops or in member's homes where there were often many children. There were also discussions on email between the group leader and the academic researcher. No 'formal' interviews were conducted; rather, the project became a multi-year series of conversations. This is a term they used in our discussions. The notion of choice, however, must be challenged if we consider 'choice' to mean making a decision between two options that in some way have equal, but different, outcomes. Choice is common to the discourse of welfare policy reform; it is often spoken from the perspective of those with resources.

Exploring Welfare Policy and Programs 155 12 This was the policy during the time of this research. Early in 2002, the rules changed so that single mothers must look for work or enter a welfare-towork program once their children are three years of age. Many other changes were introduced, including major cuts to welfare payments. 13 Research has shown that 'mixed strategies' are effective in supporting successful transition for single mothers on welfare (Butterwick, Bonson, and Rogers 1998). A mixed-strategy approach includes programs that are flexible and focused on individual needs of welfare recipients, programs that involve a combination of components such as personal needs assessment, counselling, education and training, job-finding, childcare access assistance, partnerships with employers, and links to other services such as housing, transportation, and financial counselling. Cost benefit analysis of this approach indicates substantial savings to the social assistance system. Mixed strategies have been employed with positive outcomes in Australia; and in Ontario, the OPP (Ottawa-Carleton Opportunity Project) used an investment model that emphasized individualized assessment, support for staff, and strong community partnerships.

References Butterwick, S., A. Bonson, and P. Rogers. 1998. Identifying keys to successful transition from social assistance to paid work: Lessons learned from Canada, the United States, Australia and Europe. Unpublished report prepared for the Ministry of Human Resources Development Canada, BC and Yukon Region. Edin, K., and L. Lein. 1997. Making ends meet: How single mothers survive welfare and low-wage work. New York: Russell Sage Foundation. Evans, P. 1997. Divided citizenship? Gender, income security, and the welfare state. In Women and the Canadian welfare state: Challenges and change, ed. P. Evans and G.R. Wekerle, 91-116. Toronto: University of Toronto Press. Finch, J. 1986. Research and policy: The uses of qualitative methods in social and educational research. London: Palmer Press. Fraser, N. 1989. Unruly practices: Power, discourse and gender in contemporary social theory. Minneapolis: University of Minnesota Press. Goldberg, M., and A. Long. 2001. Falling behind: A comparison of living costs and income assistance rates (BC Benefits) in British Columbia. Vancouver: SPARC BC. Guest, D. 1985. The emergence of social security in Canada. 2d ed. Vancouver: University of British Columbia Press. Klein, S., and B. Montgomery. 2001. Depressing wages: Why welfare cuts hurt both the welfare and working poor. Vancouver: Canadian Centre for Policy Alternatives, BC Office.

156 Shauna Butterwick Mansfield, J., and A. Bell. 2001. Putting research into practice: The BC Benefits longitudinal panel project. Unpublished paper presented at the Western Research Network on Education and Training, 5th Annual Conference (30-31 March 2001), University of British Columbia, Vancouver. Michalopoulos, C, and T. Hoy. 2001. When financial incentives pay for themselves: Interim findings from the Self-Sufficiency Project's applicant study. Social Research and Demonstration Corporation. Ministry of Skills, Training and Labour, British Columbia. 1994. Skills now: Real skills for the real world. Victoria: Queen's Printer. National Anti-Poverty Organization. 1997. Changes to social assistance and services. Ottawa: 21 March 1997. Ostner, 1.1994. Independence and dependency: Options and constraints for women over the life course. Women's Studies International Forum 14:129-39. Pateman, C. 1992. The patriarchal welfare state. In Defining women: Social institutions and gender divisions, ed. L. McDowell and R. Pringle, 223-45. Cambridge: Polity Press. Polakow, V. 1993. Lives on the edge: Single mothers and their children in the other America. Chicago: University of Chicago Press. Rank, M. 1994. Living on the edge: The realities of welfare in America. New York: Columbia University Press. Rimstead, R. 2001. Remnants of nation: On poverty narratives by women. Toronto: University of Toronto Press. Rist, R. 2000. Influencing the policy process with qualitative research. In Handbook of Qualitative Research. 2d ed. Ed. N. Denzin and Y. Cuba Lincoln, 100017. Thousand Oaks, Calif.: Sage. Schein, V. 1995. Working from the margins: The voices of mothers in poverty. Ithaca, N.Y.: Cornell University Press. Smith, D. 1987. The everyday world as problematic. Toronto: University of Toronto Press. - 1999. Writing the social: Critique, theory and investigations. Toronto: University of Toronto Press. Spalter-Roth, R., H. Hartmann, and L. Andrews. 1992. Combining work and welfare. Washington, DC: Institute for Women's Policy Research. Stringer, E. 1999. Action research. 2d ed. Thousand Oaks, Calif.: Sage. Swanson,}. 2001. Poor-bashing - The politics of exclusion. Toronto: Between the Lines. Yeatman, A. 1998. Activism and the policy process. St Leonards, NSW, Australia: Allen and Unwin.

PART 3 Policy and Practice: Case Studies Linking Education and Work

This section examines some specific examples of educational policy and practice that attempt to link education more closely to work. The section includes an analysis of the Skills Now! policy in British Columbia, and two case studies of educational programs designed to connect academic education more tightly to the workplace. All three studies use qualitative methods - interviews, document analysis, and observations - to examine the social processes that lie behind 'outcomes/ All are interested in education as a complex political project, where conflicting goals and interests intersect and economic pressures work indirectly through the key stakeholders in educational institutions: students, teachers, policy makers, administrators. The researchers in this section see their work as part of a political process that incorporates the voices of teachers and students into the technical discussion of outcomes and effectiveness. In the early 1990s, the NDP government in British Columbia translated concern about the fiscal crisis and skills development into a new emphasis on training for the workplace. The government described their Skills Now! policy as oriented towards increasing social equity, developing skills, and reducing unemployment. But, as Lackey shows in her chapter, the leaders of educational institutions found plenty to object to when their existing practices were critiqued as elitist, and their institutional structures were bypassed in favour of new, more responsive ones. Her assessment of the policy takes it up not just as a technical means to bring about particular ends, but as a new set of discourses and understandings that take on legitimacy. Policy is a powerful kind of public talk that, as well as legitimating program changes, affects human lives in all their complexity. Writing about policy this

158 Part 3: Policy and Practice

way makes visible effects that are not intended or announced, adding to a democratic and wide-ranging discussion. The case studies of educational practice included here illustrate what happens in classrooms when educators make the workplace more central to their teaching, as Skills Now! demanded. Gaskell, Nicol, and Tsai, as well as Grosjean, argue that there are significant educational advantages to the work/educational linkages put in place through 'applied academics' and 'co-op programs/ but there are also contradictions and unanticipated outcomes that need to be understood if the reforms are going to be successful. 'Applied academics' was an attempt to challenge the hierarchy of vocational and academic subjects that streamed working-class students into low-level jobs, but it was unable to overthrow that hierarchy. Co-op programs arose to provide opportunities for less advantaged students, but the programs are taking on elite status as they provide privileged access to jobs and are rationed on the basis of academic criteria.

6 Policy, Rhetoric, and Educational Outcomes: Interpreting Skills Now! LARA M. LACKEY

[We were] envisioning where we wanted BC to be going [and] we said, well, we've got to have skilled people if we're going to have a high quality of life, [and a] value-added, high-tech economy - and based on our natural resources but evolving into all other areas, that we had a good chance to grow. And from that we said, well, we've got this imbalance between a shortage of skilled people and a lot of unemployed young people and workers going through adjustments, and how do we deal with that? (Government representative, interview transcript excerpt, 28 October 1998)

In 1992 and 1993, Premier Mike Harcourt of British Columbia orchestrated two large premier's summits, drawing together representatives from the realms of business, education, labour, and government. The purpose was to discuss action to take in light of the apparent rapid and significant shift in the economic terrain, and the simultaneous decline of the province's resource industries. Subsequently, the government developed an educational policy entitled Skills Now! Steeped in language of urgency and impending crisis, the new policy called on the provincial system of education to prepare students and retrain the workforce for a technologically oriented, information-based global economy. As policy, Skills Now! was in step with similar reform movements initiated across the industrialized world in response to changing economic conditions. Gaskell, Nicol, and Tsai in this volume cite numerous commissions and employer groups calling for a newly trained, differently skilled workforce as a remedy to the new circumstances of

160 Lara M. Lackey

production and employment. These authors and Grosjean, also in this volume, note that policy and practice emerging from these times regularly pressed for a conceptual merging of academic and vocational traditions in public school and higher education, as well as strengthened links and transitions between schooling and work. Skills Now! embodied all of these and as such may be viewed as a common solution to a widely shared dilemma. A consideration of selected particular events and player perspectives surrounding the development and implementation of the new policy, however, provides an opportunity to enlighten the larger trend of which it was a part by examining some of its underlying assumptions, uncovering some contradictions that it raised, and presenting some alternative perspectives. It reveals, on one hand, the complex conundrum of a New Democratic government facing worldwide economic changes requiring long-term solutions, while simultaneously hoping to address the idiosyncratic circumstances and immediate problems of the BC economic landscape, maintain electoral power, and remain true to values of equity for a working-class constituency. Necessarily difficult choices were made that ultimately created both possibilities and limitations for educational outcomes, used strategies to gain acceptance and compliance that framed the problem in particular ways, and rewarded some groups over others. Drawing on data from Skills Now! documents and interviews with key policy players, this chapter asks questions critical to this volume and contemporary higher education: How do similar policies frame possibilities for educational outcomes? What mechanisms are used to foster acceptance and compliance? How are 'skills' constructed and who, as a result, is rewarded? Can it be argued that such policies contribute to altered relationships between education and the economy? The purpose of this chapter is not to wholly condemn Skills Now!, but to shake assumptions that it reflects harmonious agreement about a single unquestionable path towards a greater good for all. I conducted open-ended interviews in the fall of 1998 and the winter of 1999. Interview participants were selected on the basis of a capacity to discuss how Skills Now! was received from a particular institutional or sectoral point of view. Participants included a former premier; a former cabinet minister; three deputy ministers; a representative of an association of large businesses; four college presidents; elected and staff representatives of public school, college, and university faculty unions or associations; a labour union official; a private educational

Policy, Rhetoric, and Educational Outcomes 161

consultant; a labour economist; a representative of an association of higher education; and a community-training program coordinator. Many direct interview excerpts are included in this chapter in order to draw attention to the complexity of the issues and the multiple points of view surrounding the policy, as well as to allow readers to engage directly with personal data for purposes of discussion. I begin with information about the content of the policy and the circumstances surrounding its development. I then provide a descriptive account of selected perspectives and responses in the interviews. Finally, I offer an analytic discussion related to the questions posed above, emphasizing Ball's (1993) notion of policy as both text and authoritative rhetoric to frame and delimit possibilities for action, and Cloud's (1998) notion of therapeutic rhetoric. Background to Skills Now! Skills Now! was a large and arguably amorphous policy described in its promotional package as a 'forward-looking skills training plan to make sure students, workers, and the unemployed get the new skills for the new jobs in our changing economy' (British Columbia Ministry of Skills, Training, and Labour 1994). It sought to create a more seamless system of lifelong learning and training in the province by linking levels of formal education together and making education both more accessible to students and more relevant to employment and economic concerns. It featured four overriding goals, referred to as building blocks: 'linking high schools to the workplace so more students graduate with real skills for the real world'; 'Opening more doors - and the right doors - to college and university, helping more young women and men keep ahead in our changing world'; 'Retraining workers closer to home, ensuring new skills for new jobs'; and 'Moving the unemployed from welfare to the workforce by building on people's strengths and abilities.' A preliminary budget, also provided in the promotional materials, gives further clues to the directions and values underlying Skills Now! The policy was initiated with a $200 million start-up budget to be distributed over a period of two years. Of this, $19.9 million (10%) was allotted to 'linking high school to the workplace,' which included mandatory career planning programs for high school students; opportunities to gain credit towards apprenticeship and technical training while still in high school; alternative programs to keep students in school; the

162 Lara M. Lackey

development of Web-based career planning materials; and a focus on updating vocational training in schools to reflect new technologies. A sum of $62.8 million (31%) was allotted to 'opening more doors and the right doors - to college and university/ This involved 'tripling new student spaces' in higher education; opening a new university and giving degree-granting status to six colleges and institutes; developing training programs in new fields; and updating the use of technology in vocational programs. The largest amounts of funding in this category went to the Innovation Fund ($24 million), which allotted grants based on applications from higher education and were overwhelmingly associated with the use of technology; and the Equipment Replacement Fund ($12 million), which provided matching funds for local businesses to update technological equipment. A sum of $38.78 million (20%) was allotted to 'Retraining workers closer to home.' These funds went to the BC Labour Force Development Board, a partnership between business and labour that would advise government about training needs; training for small business; the development of new apprenticeship programs; 'quick response' training programs in communities affected by the sudden loss or gain of employment; and the development of ten Community Skill Centres, specially equipped with computers and interactive video equipment, and designed to provide training for workers in their own communities. Interestingly, the largest allotment of funds in the Skills Now! budget ($78 million, or 39%) went to 'Moving from welfare to the workforce.' This included counselling, training, and career planning (much of which took place in the new Community Skills Centres) for the unemployed and those on welfare, as well as 'training credits' given to BC businesses for training new employees. A large portion of this category ($11.29 million) was also assigned to training for people with disabilities. Skills Now! as Political Strategy As a policy that was also a political strategy, Skills Now! addressed some of the government's immediate electoral concerns as well as the larger social and economic issues facing the province. In the early 1990s, for example, a public rhetoric of fiscal restraint prevailed. Yet the recently ousted Social Credit government - clearly supporters of restraint - had been criticized for the autocratic style of its leader, who stepped down under a veil of financial scandal. As such, members of the

Policy, Rhetoric, and Educational Outcomes 163 newly elected New Democratic government hoped to be perceived as competent money managers capable of strengthening the economy while using a consensual approach to governance (Bockarie, n.d.). At the same time, it was necessary to address a commitment to social justice espoused by the NDP, grapple with the high levels of unemployment at the time, and acknowledge their base of support in the labour union movement. In addition, the newcomers needed to distinguish their own record of action from that of the previous government. Bockarie argues, for example, that in spite of the very different ideological orientations of the NDP and Social Credit, Skills Now! differed little from the policies of the previous Social Credit government, except that it was couched in a language of social equity. Unless steps were taken to frame it otherwise, the new policy risked being interpreted as simply the 'second phase' of a direction initiated by the Social Credit government: I think it was no question that the Harcourt government, including the Premier, really wanted to put its own stamp on what was happening in advanced education at that time. That was the context... Not many years earlier, in 1989, there was a report called Access for All... that put a great deal of emphasis on creating access to post-secondary education. [And then, in 1991, the Deputy Minister of Advanced Education had established the British Columbia Human Resource Development Project] ... as a two-year project to try to get all of the influencing parties together ... to get whatever level of agreement was possible on directions for post-secondary and adult education. Business, labour, government, institutions, faculty associations and so on ... By November of '92 a report was out. Now, that was a time of a change of government, and I think the way I would see it is that the report was seen within government as a report from the previous government rather than the Harcourt government ... And so in that sense it distanced itself from the Human Resource Development Project... Skills Now! put [the] accent on training for skill development, which addresses the interests and needs of different segments of the population. And it tended to give emphasis to the concerns of labour and business. And that's why you have the emphasis on skills rather than the emphasis on post-secondary education. (Private sector consultant, 18 November 1998) I would say that there had always been a - commitment is not the right word - there had always been a strong advocacy within the bureaucracy

164 Lara M. Lackey for the academic version of Access for All. So [Skills Now!] tended to be [the] applied or vocational follow-up to Access for All. Access for All in the late eighties was considered to be a dramatic expansion of the academic part of the post-secondary system. There was always an expectation and a belief [in the bureaucracy] that there should be a significant expansion, enhancement of the applied or vocational side of the post-secondary system. (Education ministry representative, 17 November 1998)

Reorganizing Government One of the ways in which the NDP chose to 'put its stamp on' this educational directive was to initiate the policy directly from the premier's office, passing over and in a sense wresting control of the agenda from the educational bureaucrats in order to assert the views of other stakeholders. As Rubenson (1998/99) notes, this move was unusual and can be seen to hold deep 'symbolic significance ... reflecting a stronger link between state and education' (2). Officially, the policy also grew out of public orchestrations in the form of the two premier's summits referred to previously; the first for the purpose of considering issues related to the BC economy in June 1992, and the second concerned with the 'learning/training system' the following spring (British Columbia Institute of Technology 1993). These sessions drew together representatives of government, relevant ministries, business, labour, and community groups 'to develop a broadly shared understanding of the challenges and opportunities we face; identify partnerships among the key sectors to ensure success; reach a better definition of roles, responsibilities and relationships, and make recommendations for action and follow-up' (BCIT 1993). A sense of urgency to accept increased skills training as the only solution, and a forecast of impending doom if one did not, characterize the message from the premier in the BCIT document summarizing the 1993 summit: Building a skilled workforce is now seen, more than ever, as critical to British Columbia's ability to produce good jobs for ourselves and the next generation of workers in a changing global economy ... We must begin to value skills development much more highly than we have in the past. Our future depends on it. As a province that earns its livelihood through trade, we're powerfully affected by external forces. We cannot design the world economic order to suit the style of life to which we want to be accustomed to living. Like it or not, we have to compete in a changing

Policy, Rhetoric, and Educational Outcomes 165 global economy. Increasingly we can only succeed through the application of our intelligence and skills. We have to dedicate ourselves to upgrading our education skills - and keep on upgrading - or we'll see our standard of living fall and our unemployment and welfare lines grow longer. History, as you know, is full of examples of societies that failed to adapt and change and were simply left behind to be studied by historians. My desire is for British Columbia to be seen by future generations as an example of a province adapting to change and 'getting it right'... We need to train more and train better. And we need to do it in a fiscally responsible way. Following the summits, Harcourt restructured his government, merging components of the Ministries of Social Services; Labour and Consumer Affairs; and Advanced Education, Training, and Technology. A new Ministry of Skills, Training, and Labour was created (Andruske 1998), to which he assigned as minister - again with symbolic impact - former forestry worker Dan Miller. Simultaneously, a series of changes to the K-12 system were underway. In total, the new system not only expanded access to post-secondary education, it lashed parts of the system more closely to one another. This increased articulation across various educational institutions, and, in general, pressed for a greater focus on 'practical/ career-oriented programs. At the same time, initiatives were developed to get more people off welfare and into the workforce: We invested a huge amount into new educational facilities. As a matter of " fact the only three new universities in Canada in the last quarter century are in BC. And we added 25,000 new spots in the post-secondary system, and we brought in the expanded university-college system, brought in the transferability of courses, and now we lead the world in that area. And right down to the high school system, where we said that over the dead bodies of teachers and parents that we were going to have report cards, and they weren't gonna be, you know, just there to measure inane knowledge and...all the randomness that people think are in report cards ... they were there as learning tools. They were there to guide teachers, parents, and students [with respect to] how well they were doing. And we were going to bring in mandatory career counselling, Grades 9 and 10, and mandatory co-op experience, Grades 11 and 12. And apprentice programs, in academic, technical, vocational, and commercial areas. And so it was all linked.

166 Lara M. Lackey And the other set of initiatives we brought in were to make work valuable, to raise the incentives to work. So that's one of the reasons we raised the minimum wage ... That's one of the reasons we removed the disincentives for single parent moms on welfare to get out into the workplace. And the cheating that was going on in welfare with the organized rings ... using different names and collecting welfare and turning it into a scam. And the landlords who were scamming 25 million a year on rental deposits. So there was an integrated series of measures to (a) have a sense of direction for the economy [and] (b) have the skills and training and infrastructure and things that we need, and the other related changes in the education system and the welfare system. (Government representative, 28 October 1998)

The above explanation offers numerous insights into the issues surrounding Skills Now!. First, it suggests that the 'problem' to which the program was posed as a solution was a complex, multifaceted one involving issues of high unemployment, high welfare rolls, and the perception of an untrained or mis-trained labour force - all concerns outside the realm of education per se. Further, in collapsing and blending former ministries into the new Ministry of Skills, Training, and Labour, Harcourt began to structurally solidify an assumption that post-secondary education would no longer be construed as a separate entity in British Columbia. Rather, education would be viewed as intimately interwoven with social, economic, and workplace concerns and issues. They were, as the above speaker attests, 'all linked.' Changing (and Blaming) Education

Without question, Skills Now! and the various changes that accompanied it were undergirded by overt and tacit criticisms of the education system, and, in particular, of what was perceived as an overemphasis on academic approaches and an underemphasis on vocational, workrelated curricula. Skills Now! was viewed by government as an attempt to redress these problems and to lend support to students outside the academic stream: 'We brought in [a] fairly tough-minded [minister of the K-12 education system] to transform the education consciousness of - well, to get rid of the bias against anybody who wasn't in the academic program, which in half a century had written off three-quarters of our high school students' (government representative, 28 October 1998).

Policy, Rhetoric, and Educational Outcomes 167

The directive that the post-secondary system needed to change - to respond to a new world and embrace a new kind of student - was clearly understood by those in the ministry as well: Well, I know government particularly was concerned about the post-secondary system. It had about two to three concerns about it. One [was] ... the wait list for students wanting to get into the post-secondary system [here referring to the fact that qualified students were being turned away due to a lack of spaces] ... a second concern was ... what I would call the framework of the post-secondary system in meeting the needs of the learner, and particularly the concept of a knowledge-based economy and how well the post-secondary system was adapting to deal with the changing nature of the economy itself. And thirdly, there were concerns about people who were really missing out on a post-secondary opportunity. Older workers, people on welfare, etcetera. So there were a number of issues that essentially the government felt it wanted re-examined: its post-secondary policy, particularly in the community college system and that, as it relates to education. (Ministry representative, 7 December 1998)

Although the education system as a whole was under attack, many of the stated goals in Skills Now! were supported by members of the education community: Anyway ... [only] 30 per cent of the kids in BC go to post-secondary [education], and in places outside the lower mainland it's 5 to 7 per cent. And yet... we go on perpetuating this myth that the only place to be [is] in university. So I think that [Skills Now!] helps validate the usefulness of the skills and trades, and other career options, so that is good. I think that is really good, because as a teacher I always found that very frustrating, because we have all got our own biases, because we have all been through the university system ourselves. And we have got our own dreams for our own kids, but we tend, very subtly sometimes, to perpetuate that notion that [university] is where to go. I think [Skills Now!] helped parents and the system ... honour and validate ... career tracks for those 70 per cent of kids. We need to attend to them more, because we are spending all of our energy catering to this 30 per cent. (K-12 teacher representative, 16 November 1998)

And, as one community college representative saw it:

168 Lara M. Lackey 'As I recall, a lot of the philosophical thrust was pretty consistent with ... the fundamental mandate of [our institution] ... So it was positive in the sense that it was to a degree a validation of the kind of programmatic thrust and the priorities we had been preaching over the years.' (23 November 1998)

Precisely because many educators embraced the Skills Now! ideals, however, they were often surprised and puzzled at the government's public chastisement of education, especially since teachers as a group had traditionally supported the NDP. A number of speakers from educational institutions reported feeling attacked or marginalized by the policy and the process. In particular, responses to the government's choice of publicity slogan - Skills Now!: Real Skills for the Real World - highlights how blame was perceived and how the policy was interpreted. Certainly, there were some who viewed this campaign as simply the work of publicists hoping to portray the government as 'doing something' about the education system and the economy, and as appealing to particular sectors and segments of the population for political purposes. For these speakers, the slogan had little impact: Well, I perceived [Skills Now!] when it was announced as yet another in a passing parade of... packages, if you like, that governments tend to enunciate every once in a while as a new thrust and a new emphasis and a new set of priorities. And so what tends to happen is that... a lot of existing programs and existing funding will get scooped into that new area and in some ways it's a repackaging of yesterday's news ... I don't think anybody paid a whole lot of attention to the label. I think they just saw that as something that the political spin doctors have decided was a good thing to do. (Community college representative, 23 November 1998)

For many in the education community, however, the slogan was experienced as an unfair swipe that perpetuated and merged with popular discourse belittling both academic knowledge and the work and expertise of educators. It also implied a simplistic separation between theory and practice and even suggested that the world of the academic was a delusional one: Oh, it was insulting to teachers. Because what it was saying is that you are not teaching 'real skills for the real world.' And our view is that the classroom practitioner has got to do both things, right? You've got to get across

Policy, Rhetoric, and Educational Outcomes 169 the theoretical information as well as the practical ... In fact, our policy position became that all courses should have a nice blend of theory and application. That is how people learn. But again ... it became a very insulting thing in that it said to the system, you are not doing this and we are going to show you how to do it. (K-12 teacher representative, 16 November 1998) In some ways, the message we were getting back is that we were in the way. That we were part of the problem ... I remember the slogan very well. That particular slogan just drove us up the wall. It was like, oh yes, the universities don't exist in the real world. We don't have children. We don't have relationships. We don't have to pay taxes. You know, obviously we don't exist. We are in this imaginary world of unicorns and griffins and things like that... What does that tell people? You know, we have one of the best university systems in the country. (Higher education faculty association representative, 26 October 1998) The negative reaction [to the slogan] from the broadly defined adult ed crew, both teachers and programmers, was, 'What a slap in the face! What do they think we've been doing?' You know, we're 'unreal'? Unreal skills to a non-existent world. So [the slogan] was viewed as a real [speaker makes slapping motion with hand to eye] in the eye ... [On the other hand, recognizing that] government has to be seen to be doing something ... I came up with this kind of cynical formulation [that] it's not an insult to us, it's a way of making them look as though suddenly they're doing something a lot better than anyone else has done before. And appealing ... to that segment of the population in business, industry, labour unions, and the general public that doesn't like schools. So they were tapping into all that. (Community college representative, 25 November 1998) Inserting the Community Skills Centres and Appealing to Labour Even more contentious than the slogan, however, was the creation of the Community Skills Centres, which established an entirely new kind of educational entity in British Columbia. The federal and provincial governments jointly funded the centres, providing upfront costs, along with equipment such as computers and interactive video. The intent was that the centres would not develop educational programming inhouse, but would be educational 'brokers' and would aim to be financially self-sufficient within five years of opening. Centre staff would

170 Lara M. Lackey

work with members of the local community to identify training needs not being met by other institutions, establish sources of funding with special emphasis on encouraging support from employers and businesses, and generally act as facilitators for a whole range of educational needs, from workplace training to adult literacy. Twenty-one centres were established. Most of these were in northern and rural locations, but four were placed in the Lower Mainland, already the hub of British Columbia's college and university system. From the government's perspective, the skills centres grew directly from a perceived lack of responsiveness by the colleges to the training needs of certain groups of workers. A sympathetic reading would depict the centres as cradles for the educationally disaffected working class and those welfare recipients desperately in need of training but for whom the traditional education system was experienced as hostile and elitist: What we were trying to say to the community colleges [was that] you have not been responsive to the communities you serve. Why, go to the pulp mills, for example, and talk to workers, or go to the IWA [Woodworkers Union]. They had absolutely no sense that the community college was in any way an institution there to serve their needs. They felt quite alienated, and yet you can see going back to the kind of transition in the economy there was a huge need to deal with those kinds of workers to give them opportunities to try to better themselves. (Government representative, 27 January 1999)

Certainly, some of the centres were perceived as working very well and seemed to do precisely what they were intended to do. Instances were noted, for example, of innovative collaborations between unions and employers around technological training in the workplace. Staff members cited the value of the influx of technology to the centres and the heavy use of their computer labs, as well as an increased awareness of different ways to deliver educational services. Even where centres' efforts were seen to be successful, however, their appropriateness within the system overall was often questioned: In our case, we worked relatively well with our local skill centre. We tend to use it a lot. We tend to make use of space there. We have a reasonably good relationship with them. But again, it looks funny. Like why did skill centres evolve? I think this is a pretty clear [government minister] type of

Policy, Rhetoric, and Educational Outcomes 171 an issue ... something that he wanted to put his stamp of approval on. (Community college representative, 14 January 1999) Members of the college community were upset at what they thought was an unjust criticism, especially since they had experienced years of funding cutbacks. For some, the government was seen to be giving up on the public system and its potential to meet the needs of students and workers. Certainly skills centres were viewed as competing with a mandate that the colleges felt was theirs: [We perceived] the establishment of community skills centres as a direct affront. A not very disguised reflection on government's perception of the adequacy of college efforts, in other words. And I think this feeling was shared by a lot of continuing and community and adult ed personnel in colleges. In effect they said, 'My God, we have been labouring all these years, trying to do this kind of work, and now suddenly we are found wanting, when we've never really been properly funded to begin with.' And yet now people are saying you haven't done well enough. (Community college representative, 25 November 1998) Another community college representative put it this way: Rather than creating a seamless web of intake and advancement to wherever a person wanted to go, they created an institution, which in fact was by definition an indictment of the public system, and a giving up right at the beginning of the game, the necessity to transform them if they needed to be transformed ... I guess the thing is, it expanded choice in the most important decision that people have to make in their lifetime, and that is, what steps do they take to get their social needs? You know, their means of living. And so, I could never turn my back on a program like this. [But] I was really, really sad about the high profile of community skills centres as an indictment of the public system, without getting onto the public system and shaking it loose ... I mean, if they had had to appoint twenty new CEOs in the college system to accomplish it, that would have been better than giving up. (Community college representative, 30 November 1998) The sense that the centres were indeed there to act as an irritant and a spur - to the colleges was also understood by staff at the skills centres, but the tension in the system made their work difficult:

172 Lara M. Lackey We're set up - and we've heard cabinet ministers, politicians say we were set up to be a thorn in the sides of colleges. To stimulate colleges to deliver training more effectively, in a more innovative way, and cheaply. And that 'set up' a whole group of adult education professionals. I came into the position not knowing what my role was. And it has created no end of grief for me, in that it did not take into consideration what the college response would be ... And then [programs] have been implemented in ways that have allowed the animosity to fester. So you have conflicting government policies ... to the extent that... there's the college system and then they're setting up skills centres to run as private organizations ... to serve the community needs in partnerships with people who don't want to be partners. (Community skills centre representative, 29 October 1998)

Staff members also struggled, however, to carry out a directive to encourage the involvement of the business community, particularly in Centres working with a lot of people on income assistance: My sense is that it was anticipated that industry and business were going to invest significantly more in education and training than they did. We have found it very, very difficult to involve business and industry in education and training, and many of the programs that were in place were really marginally successful. Business only became involved if it was really worth their while. (Community skills centre representative, 29 October 1998)

Lacking support from local business, the speaker quoted above found that he was often bidding for the same government funds that colleges and private educational consultants targeted. As a result, the initiation of skills centres, in combination with the project-based nature of Skills Now!, had the effect of increasing a sense of competition among educational institutions in general. Reducing Welfare Rolls

Although many of the skills centres provided services for people receiving income assistance, the links between Skills Now! and welfare practices are more complex than might at first be apparent. For one thing, the organizational shift that combined parts of the former Ministries of Social Services and Advanced Education added to the complexity of accomplishing the goals of the program in this area. In the

Policy, Rhetoric, and Educational Outcomes 173 following year, as well, the work of Skills Now! merged with the new BC Benefits policy. It was just harder than expected, and it took longer. Like the part around people on welfare. You know, it was quite a challenge. Not only was [Skills Now] introduced as a new program, but part of that whole change was in '93 when the Ministry of Skills, Training, and Labour was created, was taking the training mandate from the welfare ministry and putting it into Skills, Training, and Labour ... So integrating those things. And so that in itself was such a culture change. And an organizational transition ... And so what happened was, through another initiative in '95, BC Benefits became a bigger thing, focusing on people in need, and BC Benefits gave a bigger mandate and push behind the welfare part of what was Skills Now! (Ministry representative, 20 October 1998) One of the practices implemented through BC Benefits was mandatory training for youth on welfare. Programs called Job Clubs, for example, required that people receiving income assistance attend sessions in which they were supervised while creating resumes and applying for work. Centre workers detected a lot of anger in the community around these kinds of practices, citing evidence in local posters that announced 'JOB CLUBS - JOB CRAP/ The sense was that the work of idealistic adult educators became caught up in welfare-to-work programs that were often perceived as both authoritarian and spurious preparing people for 'Mcjobs' that did not sustain families or provide access to health benefits. For people on income assistance, the playing field has changed significantly in the last three years ... and rny sense is that it has become increasingly demanding of individuals [and requires] them to go through a series of programs that really do not prepare them for work and don't have a solid educational footing ... If I can summarize it in a way that might be useful, I think that there was an expectation that business would do more in creating jobs and in training for jobs, and in some way linking the welfare policy to that. And I think that... a lot of it was based on faulty assumptions, both [with respect] to the business end and also with respect to the root causes as to why people are on welfare. (Community skills centre representative, 29 October 1998) In addition, the following speaker notes the impact of Skills Now! in

174 Lara M. Lackey legitimizing negative perceptions about people on income assistance and therefore justifying these more coercive approaches. He raises further the indirect but uncomfortable relationship between a decrease in the income assistance rolls and Skills Now! funding. You can really hurt people too. Like, Skills Now! didn't have a lot of dollars, but it had a lot of ideological opinion that said, well, basically income recipients don't want to work and they should be made to work and they should be bridged to having to work. I mean, nobody ever said that, but it's there ... And that allowed them to do what they did in BC Benefits ... and you know, almost the entirety of BC Benefits was financed by reducing the level of claims to recipients ... Now it didn't wholly do that; you can't say $199 million in over two years went into Skills Now! and $199 million was taken from income recipients, but eventually you get to that conclusion. (Higher education faculty association representative, 30 November 1998) Appealing to Business The language of Skills Now! was consistent with and in many ways was meant to appeal to the values of the business community, who tended to take the need for 'skills' as a straightforward fact: I think because we had a significant mismatch between the skills of the workers who were unemployed and the skills that could make them employable. We had a lot of people with relatively low skills and certain categories of skills, and the kind of people we needed were people with a higher level of skills, more technical skills, more complicated skills, and so on. I remember the skills that people felt were needed were technical skills, computer skills, basically in those areas. (Labour economist, 10 February 1999) I think that what triggered Skills Now! is the fact that the economy was changing, jobs were changing, and the nature of the skills required was changing ... I think that there was perhaps a growing voice from the employer community that at that time was at least being partly heard by some of the decision makers, that the skills the graduates were bringing weren't matching skill requirements of the jobs that were evolving. (Business representative, 25 November 1998) In spite of this appeal, however, it can be argued that representatives of business in the province remained unconvinced by this government

Policy, Rhetoric, and Educational Outcomes 175

to which many were ideologically opposed. They maintained a stance that the NDP had not established an 'economic environment' conducive to business and profit making: I guess the starting point for many people in the business community... is that they don't see government as really doing anything but getting in the way. Secondly, most of the business community would not have been of the same political persuasion as the government of the day, and there was a certain amount of skepticism about that. (Business community representative, 25 November 1998)

In addition, surveys completed among large employers in British Columbia suggest that the particular 'skills' that employers claim to want in employees are related to vaguely attitudinal characteristics such as being 'honest and reliable/ 'flexible and adaptable/ or having a 'good work ethic/ as much as they are to 'technology skills' or any specific knowledge base. 'Communication skills' are consistently at or near the top of the list in every sector (see, for example, Business Council of British Columbia, February 1999). Further, employers admitted that given the choice, they would in any case often hire applicants who had both technical training and traditional academic backgrounds. High unemployment favoured employers in this way, and acted to raise the standard for entry-level positions. Labour representatives charged that employers had been consistently resistant to supporting many workplace-training initiatives, especially with respect to apprentice programs that were long and costly: Well, in some cases, they just don't like apprenticeship. It's a chronic problem that exists today. I am not going to train a worker, I am not going to go to that expense so that they can now go down the road and work for you ... I don't want to train workers because you are going to steal those workers once I've trained them, and the only solution for that would have been for the government to say, 'Look, either you are going to do it on your own, in consultation and cooperation, or we are going to legislate and you are not going to have any choice.' They [the government] never got the political courage necessary to actually do it. Some ministers talked about it openly, but when it came right down to pointing a gun, they caved in. As a result the employers have done next to nothing. (Labour representative, 22 February 1999)

In summary, although Skills Now! documents presented policy

176 Lara M. Lackey

players as acting in concert and harmony, these excerpts reveal multiple competing interests and social relations permeated by issues of power. These tensions in turn signal the contradictions implicit in Skills Now! and portend the very real social consequences of such a policy. In the next section, I discuss contradictions and troubling issues raised by Skills Now! rhetoric, and return to the questions posed at the outset of this chapter. Discussion Skills Now! as Rhetoric

Ball (1993) suggests that policy is too complex to be explained by any single theory and uses at least two notions to describe it. One is that policy is a kind of text to be interpreted within particular structural and ideological circumstances. In this sense, people make policies 'fit' within particular contexts, and settings therefore mould and remodel policy. In their chapter in this volume, Gaskell, Nicol, and Tsai provide an interesting example of how this can occur in practice. A policy is also, however, a kind of discourse or set of truth claims that, delivered with a level of authority, directs (but does not entirely impose or necessarily limit) parameters of thought. Ball argues that this latter effect, which he refers to as the 'second order of policy outcomes/ is concerned with issues of access, social justice, and perceptions of opportunity. Governments, of course, have power, ultimately backed by police and military forces, to impose rule. Increasingly, however, persuasive rhetoric and 'spin' are used in order to entice rather than enforce constituents' cooperation with policies: 'Discourses are about what can be said and thought, but also about who can speak, when, where, and with what authority ... Certain possibilities for thought are constructed. Words are ordered and combined in particular ways and other combinations are displaced or excluded ... We do not speak a discourse, it speaks us' (Ball 1993,14). There are a number of ways in which Skills Now! implicitly or explicitly highlights some ideas while downplaying or obscuring others. In this way, it fosters a perception that some actions and outcomes are not only preferable but within the realm of possibility and reason, while others are not. Assumptions are framed, for example, with respect to the purposes of education, the nature of the economic problems being faced, the kinds of skills that should be valued, and the ways that blame and responsibility for the wider situation are placed.

Policy, Rhetoric, and Educational Outcomes 177 Skills Now! as Educational Philosophy

As an educational philosophy, Skills Now! assumed that the primary purpose of education is to accommodate the demands of employers and the economy. It makes invisible other philosophic orientations to education, such as those that argue the need for critical, analytic thinkers who might solve problems in circumstances yet unimagined; that seek to identify curriculum content that is not merely 'useful' but meaningful, engaging, and compelling in ways that might stimulate an interest in lifelong educational pursuits; and that foster thoughtful, informed citizens capable of making democracy work or of actively challenging inequality (Eisner 1979). Through its emphasis on a new, changed future and its appeals to working-class interests and vocational concerns, the rhetoric surrounding Skills Now! often suggests a philosophic stance of social reconstruction and the building of a better, more equitable world. In many ways, however, such discourse veils an agenda of social adaptation in which education reproduces current social relations. Tensions between a conservative agenda and a progressive rhetoric permeate Skills Now! and highlight ways that education intermingles with economic and social concerns. Skills Now! as Economic Policy

Analysts often think of policies as responses to perceived problems. Because there may be multiple arguments as to what a problem is, however, a policy can be understood as tacitly assuming one problem while discounting or veiling others. In a way that parallels much other discourse surrounding global economic changes, Skills Now! framed the 'problem' faced by British Columbians as a serious gap between the jobs available and the number of skilled workers to fill them. Government perceived its role as fostering changes favouring a vocational slant in education; providing increased access to a wide range of workrelated educational programs; and convincing individuals about the need to take advantage of these opportunities. Those who embraced this orientation would be, it was implied, richly rewarded, while those who rejected it would put secure futures out of reach, risking both personal and provincial economic stability. As Cohen (1996) notes, this is a 'political message [with] moral overtones' (187). In a careful cross-national analysis, however, Crouch, Finegold, and Sako (1999) challenge some of the most fundamental assumptions underlying these claims, which, they suggest, have taken on 'a distinc-

178 Lara M. Lackey

tive prominence and urgency' [for which] 'everywhere the argument is broadly the same, and at a very general level is shared by all shades of political opinion and by business and labour leaders' (1). On one hand, the authors argue that the call for skills rests on a Utopian vision of a world in which all are highly skilled and engaged in fulfilling, meaningful work. In this world, those few who are not capable of attaining high levels of skill can be easily supported by the many, and as a result all benefit. Although admirable in intent, the authors argue that attaining this vision in the long term is complicated by the fact that it is a collective goal for which no single group is responsible, and by the relatively simple proposition that 'skills' and education do not in themselves create employment. Rather, achieving a high skill/high employment society requires that employers take advantage of skilled workers' knowledge by encouraging innovation and the development of new products and services in order to create more jobs. Because businesses must ultimately ensure their own individual survival, and since innovation requires a level of risk and is not the only means to maximize profits, there is no guarantee that employers will take up this challenge. On the other hand, Crouch, Finegold, and Sako (1999) argue that in the short term, a call for increased general levels of skill and education is in the interests of no group except employers, as it serves primarily to increase competition for existing work. For those who already had high skill levels, competition is increased for work that might otherwise have been easily obtained. Although training may provide hope for those with limited skills and allow them to enter the competition, increased education levels overall may not make their search for employment any less difficult. In fostering training and retraining, unions may find some advantage for their own members, but education and the infusion of technology that often accompanies it is often associated with increased productivity. This, in turn, decreases employment while potentially increasing tensions and competition between organized and unorganized workers. For all workers, the call to increase skill levels belies the reality that a significant number of jobs currently being created in the new economy are low-skill, parttime, and temporary. Only for employers are general increases in education a benefit, as this acts to devalue labour costs. All in all, the authors argue that 'Even with substantially increased educational investments, an important minority of the workforce will be unable to participate in the employment provided by the learning society ... For years to come, many members of the workforce will be unable, as a result of both demand and supply factors to gain high-

Policy, Rhetoric, and Educational Outcomes 179

quality places within the labour market. It should not be pretended that, if they only showed adequate initiative and responded to the educational opportunities available, most people would be able to do so. Many will be forced to compete for the low-productivity opportunities that will not require high levels of education' ... (238-9). They argue further that educational policy cannot be substituted for social policy, and they raise the need to balance global economic structures with parallel political ones. A rather more radical critique is provided by Rojek (2000) who argues that we now live in a world glutted by overproduction, making continued goals of increased production untenable and irresponsible. If one accepts these arguments, Skills Now! can be viewed as an initiative that directs individuals and institutions to invest in 'skills' but obscures the fact that the implied rewards cannot come true for many. As Crouch, Finegold, and Sako suggest, such policies have the potential to increase discontent among students when financial resources and time consumed for schooling do not result in employment. This raises questions about the complicity of educational institutions and training agencies that benefit financially from a call for increased skills but may engage in an agenda that misrepresents outcomes. Further, such policies wholly ignore the widening gap between rich and poor that economic globalization has caused, and the serious environmental dilemmas that continual goals of increased production have created. Beyond making assumptions about the purposes of education and the economic issues involved, an analysis of the rhetoric of Skills Now! reveals other significant issues as well. These include the ways in which 'skills' are constructed and who is blamed and held responsible for implied problems. Skills Now! and the Construction of Skills

A prominent part of the Skills Now! campaign was its slogan 'Real skills for the real world,' implying that abilities previously fostered through the education system were in some way deficient, irrelevant, and only valued by an irrational mind. Implicit in the slogan is the notion that worthwhile skills are those that may be applied directly to the workplace and are broadly associated with the ability to negotiate work through technology. Unacknowledged, however, are the contradictions surrounding the association of work with technological skill, and the extent to which what counts as 'skill' has historically been a site of struggle and contention.

180 Lara M. Lackey

Gaskell (1992) argues that a 'skill is not a technical property of a job that can be empirically determined by careful study, but an assessment of value that is rooted in politics: in power, in culture, and in economic position' (113). Definitions of skill, she posits, legitimize some forms of schooling while devaluing others. They are the basis on which students construe life chances, make life decisions, and are used to the justify unequal distribution of employment rewards. Although issues of class emerge around the valuing of skills associated with mind versus those associated with body, women in general have been less able to make a case for the legitimacy of their skills. As such, gender is a useful category through which to examine taken-for-granted assumptions about the value of skills, including those in Skills Now! In the past, for example, by restricting women's access to certain fields, training policy has actually helped to maintain women workers within fields associated with the feminine and with lower-valued skills (Cameron 1996). The interpersonal skills that many women bring to their work are often assumed to be 'natural' rather than learned, and, because a product is not produced, these abilities may be invisible and go unrewarded in the workplace (Poynton 1993). Baron (1992) argues that notions of skill are so embedded within conceptions of gender that redefining skill can become a struggle to define what counts as masculinity or femininity. Probert and Wilson (1993) argue that 'socially acceptable definitions of masculinity continue to hinge on the idea that it involves the ability to do things that women cannot do' (9), and that male job pride and satisfaction often depend on this condition. In these senses, the process of legitimizing skill can be intermingled with constructions of gender. A tension is also created for women to the extent that valued work skills are associated with technology, as is the case in Skills Now! Probert and Wilson (1993) argue that 'technology itself, and the associated social construction of skills, is a site of gender struggle' (9). Sandelowski (2000) notes the overwhelming cultural assumption that links technology to masculinity, yet argues that the relationship between gender and technology is nevertheless a complex one. Although nursing, for example, may commonly be associated with interpersonal care and the human touch, and therefore as the opposite of the technological, nurses have actually been important in ameliorating technology within health care: 'Nurses have served as the primary users of machines and as machine-body tenders in health care ... Nurses have been charged with enlisting patients' acceptance of technology' (2). Similarly, much of what teachers are asked to do is not simply to train

Policy, Rhetoric, and Educational Outcomes 181

people in the use of new technology, but to promote the acceptance of the need for technological skills as normal and natural factors in learning and in the workplace. This role of teachers and other public servants as facilitators of technology is key to what some argue is actually the most significant influence of policies like Skills Now!, that of fostering an acceptance in general of a changed set of rules and field of play related to work and economic survival. In the current conditions created by the shifting economy, issues of equity are easily pushed into the background in light of other crises that appear on the surface to be more impending and urgent. 'In this transition, jobs can be re-classified, skills redefined, and conditions of work re-negotiated' (Jensen 1996, 84). Skills Now! and Social Justice

To the extent that Skills Now! became a tool for reducing welfare rolls, it created an additional conflict for an NDP government espousing values of social justice. Skills Now! did throw support to one relatively disadvantaged group: organized workers, primarily males of the working class, displaced from the resource industries. One can argue, however, that it did so partly at the expense of an even less-advantaged group, those on welfare. As Brodie (1996) writes, a key effect of the discourse surrounding globalization has been to foster a new 'set of assumptions about the role of government and the rights of citizens' (126). In essence, she argues that this has involved redefining what counts as the good citizen, from one who might expect a level of protection from government to one who is 'market oriented [and] values self-reliance, efficiency, and competition' (131). In turn, the welfare recipient is reframed from one who is in need of help to one who is 'dependent' and an irresponsible drain on collective resources. As a result, we 'are encouraged to think of poverty in terms of undeserving, deficient, and wrongly skilled individuals instead of Canada's 'restructured' political economy and its seeming incapacity to provide employment, much less good jobs, for an unacceptable and ever-growing number of Canadians' (138). Arguably, Skills Now! contributed to this discourse and in doing so fostered acceptance of greater social inequity and diverted an understanding of poverty as the result of deliberate social policy (Connelly and MacDonald 1996). The discourse surrounding Skills Now! parallels that which Cloud (1998) refers to as a rhetoric of therapy. She argues that therapeutic language, by which she means references to personal healing, coping,

182 Lara M. Lackey

individualism, and self-help, not only permeates current political and media discourse, it represents a concerted effort to avert the potential for social unrest and collective action towards social change. Cloud is careful, on one hand, to acknowledge the benefits of psychological therapy for people who have experienced trauma. She notes also, however, the latent role of psychology in helping people cope with and adjust to stressful life situations - such as sexual harassment, parenting, stress, or depression - that are often the result of a lack of support from social and political structures. In a like manner, key parts of the Skills Now! policy fostered coping skills such as career planning for students, and life planning, interview, and job behaviour skills for those on welfare, rather than grappling with job creation itself, questioning the overall direction being taken by the business community, or highlighting the role of government in providing social support networks. The implication is that individuals are not only responsible for training themselves, they are to be personally blamed if left in circumstances of hardship or poverty. Cloud argues that the pervasive infusion of therapeutic rhetoric into all levels of political and economic communication is a deliberate political strategy that is now commonly used to divert attention away from political and economic choices that cause instability, and to redirect attention towards the self: "The most important rhetorical feature of the therapeutic is its tendency to encourage citizens to perceive political issues, conflicts, and inequities as personal failures subject to personal amelioration' (3). She traces the history, for example, of the last major moment of social unrest in the United States - the civil rights movement, the feminist movement, and the protests against the Vietnam War - noting how times of political protest were gradually turned to movements of self-involvement and self-reflection. To the extent that Skills Now! blamed individuals for having the wrong skills for work; to the extent that it implied that people on welfare were simply unskilled or unwilling to work; and to the extent that teachers and the education system were chastised for fostering the wrong kinds of knowledge, Skills Now! provided just such a diversionary rhetoric. It drew attention away from the political choices being made to shift work relations and reduce the social safety net provided by former welfare systems. Instead, it directed people to heal themselves through obtaining 'real' skills, to become free agents in a competitive labour market, and to accept that old ways of life were unquestionably gone. While criticizing teachers and the current system of education for being unsympathetic to working-class interests, it

Policy, Rhetoric, and Educational Outcomes 183

simultaneously enlisted teachers in ameliorating an acceptance that all important work and skill must involve new forms of technology. Therefore, although Skills Now! was a short, two-year policy that might otherwise have had minimal impact, it was significant to the degree that it merged with a pervasive rhetoric of therapy that has become standard across North America. Returning to the questions posed at the beginning of this paper, policies like Skills Now! frame possible educational outcomes through selected orientations to the purposes of education, and by presenting policy as solutions to selected social and economic problems, ignoring alternatives. Compliance in this case was fostered through a rhetoric of urgency around economic security, using an attack on the traditional system of education in an attempt to appeal to unionized workers and business, both audiences whose support the NDP was courting. Simultaneously, all facets of the education system were called on to ameliorate an unquestioned acceptance of the need for new technological skills, therapeutic orientations to training, and a new playing field related to employment and citizenship. While many debates in education have centred on the tensions between vocational and academic approaches to education, it seems the more critical question for educators concerns whether or not to cooperate with this more elusive but disturbing therapeutic agenda. In the end, Skills Now! may well have reflected a sincere effort to resolve and prevent perceived problems of the day. Nevertheless, this examination of multiple perspectives and underlying assumptions of the policy suggests that it also used an increasingly universal rhetoric based on a particular vision of the future to justify local changes and address short-term problems, thereby raising important questions about the choices made. References Andruske, C. 1998. Critical analysis of Skills Now!: Real skills for the real world. Program summary (1994). Unpub. paper. Ball, S. 1993. What is policy? Texts, trajectories, and toolboxes. Discourse: The Australian Journal of Educational Studies 13(2): 10-17. Baron, A. 1992. Technology and the crisis of masculinity: The gendering of work and skill in the U.S. printing industry, 1850-1920. In Skill and consent: Contemporary studies in the labour process, ed. A. Sturdy, D. Knights, and H. Willmott, 67-95. London: Routledge.

184 Lara M. Lackey Bockarie, A. n.d. Western Research Network on Education and Training: Skills Now policy analysis. Unpub. paper. British Columbia Institute of Technology. 1993. Premier's summit on skills development and training. Vancouver, BC. British Columbia Ministry of Skills, Training, and Labour. 1994. Skills Now! Real skills for the real world. Program Summary. Victoria: Ministry of Skills, Training, and Labour. Brodie, J. 1996. Restructuring the new citizenship. In Rethinking restructuring: Gender and change in Canada, ed. I. Bakker, 126-40. Toronto: University of Toronto Press. Business Council of British Columbia. 1999. What are British Columbian employers looking for? Skills-Characteristics. Vancouver (February). Cameron, B. 1996. From equal opportunity to symbolic equity: Three decades of federal training policy for women. In Rethinking restructuring: Gendered change in Canada, ed. I. Bakker, 55-77. Toronto: University of Toronto Press. Cloud, D.L. 1998. Control and consolation in American culture and politics: Rhetoric of therapy. Thousand Oaks, Calif.: Sage. Cohen, M.G. 1996. New international trade agreements: Their reactionary role in creating markets and rethinking social welfare. In Rethinking restructuring: Gender and change in Canada, ed. I. Bakker, 187-202. Toronto: University of Toronto Press. Connelly, P., and M. MacDonald. 1996. The labour market, the state, and the reorganization of work: Policy impacts. In Rethinking restructuring: Gendered change in Canada, ed. I. Bakker, 82-91. Toronto: University of Toronto Press. Crouch, C, D. Finegold, and M. Sako. 1999. Are skills the answer? The political economy of skill creation in advanced industrial countries. Oxford: Oxford University Press. Eisner, E.W. 1979. Five basic orientations to the curriculum. In The educational imagination: On the design and evaluation of school programs, ed. E.W. Eisner, 50-73. New York: Macmillan. Gaskell, J. 1992. Constructing skill hierarchies. Gender matters from school to work, 113-33. Toronto: OISE Press. Jensen, J. 1996. Part-time employment and women: A range of strategies. In Rethinking restructuring: Gendered change in Canada, ed. I. Bakker, 92-108. Toronto: University of Toronto Press. Poynton, K. 1993. Naming women's workplace skills: Linguistics and power. In Pink collar blues: Work, gender and technology, ed. B. Probert and B.W. Wilson, 85-100. Carlton, Victoria: University of Melbourne Press. Probert, B., and B. Wilson. 1993. Gendered work. In Pink collar blues: Work, gender and technology, ed. B. Probert and B.W. Wilson, 1-19. Carlton, Victoria: Melbourne University Press.

Policy, Rhetoric, and Educational Outcomes 185 Rojek, C. 2000. Leisure and culture. Houndsmills, UK: Macmillan. Rubenson, K. 1998/99 (winter). Policy studies: Speaking truth to power. Newsletter of WRNET (Western Research Network on Education and Training). Vancouver: Faculty of Education, University of British Columbia. Sandelowski, M. 2000. Devices and desires: Gender, technology, and American nursing. Chapel Hill and London: University of North Carolina Press.

7 Working Outcomes in the Classroom: A Case Study of Applied Academics in British Columbia JIM GASKELL, CYNTHIA NICOL, AND LI-LING TSAI

At the beginning of the 1990s, a variety of commissions and employer groups claimed that schools were not preparing students appropriately for the changing technological requirements and increased worker responsibilities that were the product of industrial restructuring in a global, decentralized, knowledge-based economy (e.g., for example, Commission on the Skills of the American Workforce [CSAW] 1990; Conference Board of Canada 1992; Economic Council of Canada 1992; Secretary's Commission on Achieving Necessary Skills [SCANS] 1991). The low status and low skill levels that were the product of most vocational education programs were compared unfavourably to the German and Scandinavian systems of technical education, which provided routes to high skills and good jobs (Heinz 1999; Deissinger 2000). While the claims of the commissions did not go unchallenged (for example, Bailey 1995; Barlow and Robertson 1994), many jurisdictions in North America responded by developing programs to improve the transition from school to work through focusing on the integration of academic and occupational curricula (Grubb 1995; Stern et al. 1995). This approach was encouraged in the United States by the 1990 amendments to the Perkins Act, which moved away from defining vocational education as narrow skills training for specific occupations and required that every program supported by federal funds integrate academic and vocational education through coherent sequences of courses (Grubb 1995; Steinberg 1998). In British Columbia, the government established a new program called Skills Now! in 1994 under the joint direction of the Ministry of Education and the Ministry of Skills, Training and Labour. The program had the ambitious purpose of restructuring British Columbia's

Working Outcomes in the Classroom 187

educational system to meet the demands of the new workplace (Ministry of Skills Training and Labour 1994,1995). The goals of the program were to be achieved, in part, through the development of high-quality, applied educational programs at secondary and post-secondary institutions, facilitated through the formation of partnerships between school districts, post-secondary institutions, and local business and community members (Province of British Columbia 1994). The program's focus on skills, as Lackey (this volume) notes in her analysis of the development and implementation of Skills Now!, was an attempt to address the perceived lack of emphasis on vocational education and overemphasis on academic education. Business, labour, and community college organizations provided public support for the Skills Now! program, and their representatives sat on advisory committees. As part of Skills Now! a program of 'applied academics' was developed for secondary schools in the areas of mathematics, physics, English, and information technology. The meaning of applied academics was fluid but it generally referred to an increased emphasis on classroom and laboratory activities that connected abstract knowledge to workplace applications (Stern et al. 1995). These educational initiatives are interesting because they challenge one of the fundamental dichotomies of modern schools: the separation of liberal/academic and utilitarian/occupational education found in the academic and vocational streams of most secondary schools today (Lewis 1998). Because they are being advocated mainly by business leaders, they have also come under intense scrutiny from progressive educators who see them in the context of what Apple (1996; 1998) calls the 'conservative restoration.' These educators are also concerned that the connections between business and industry will be very much on business' terms. The perceived focus of these programs on technical skills is challenged by labour market economists such as Allen (1999) who argue that increases in productivity come as much from investment in the social sciences and humanities as from investment in sciences and engineering. There are other progressive educators, both historical and current, who have provided strong educational and political arguments for supporting an effort to integrate academic and vocational education. Dewey (1916), for example, in Democracy and Education, argued for an education through occupations rather than an education for occupations. An education through occupations would be one in which occupations would be used as an organizing principle through which many subjects could be taught and future workers

188 Jim Gaskell, Cynthia Nicol, and Li-Ling Tsai

could be given the power to adapt to, and shape, changing conditions. Education for occupations, in contrast, was narrow, job-specific vocational education that created a second-tier education for working-class youth and subordinated their interests to those of employers. It was this type of education that became prevalent in Canadian and American high schools. The idea that workers should have an education that gives equal value to practical as well as theoretical knowledge, and that gives workers the intellectual tools to participate critically in society, was at the heart of the development of the Soviet polytechnic education system (Beck 1990). Building on these traditions, current progressive curriculum theorists such as Kincheloe (1995), Young (1998), and Lewis (1998) argue that the kinds of technological and economic changes taking place in industry provide an opportunity to challenge the traditional division between mental and manual labour inherent in the current organization of workplaces and schools, and to redirect education towards the preparation of what Kincheloe calls 'smart workers/ Such workers would be skilled in communication, reasoning, and problemsolving; be able to critique production and management processes; and have adaptable technical skills. These arguments for the integration of the vocational and the academic are given additional support from recent developments in learning and situated cognition, which stress the importance of learning in practical contexts in order to promote understanding (Resnick 1987; Berryman 1993; Lave and Wenger 1991). Current educational initiatives designed to integrate academic and vocational education can thus be seen either to be part of a widespread conservative restoration or as having far-reaching progressive possibilities. They can be seen as necessary for maintaining economic competitiveness and productivity, or as a way of blaming the school system once again for failure. But what actually happens to these initiatives in schools? Lewis (1998) claims that the integration of academic and vocational education has remained almost wholly a vocationalist preoccupation, and that it has not been internalized by the academic mainstream or by the public. In British Columbia, however, the applied academics program has been high profile and aimed at the academic mainstream. Despite a history of failure at previous attempts to integrate academic and occupational curricula (Gaskell and Hepburn 1997), there was optimism at the beginning of this program that the current economic, industrial, and technological contexts had created conditions for success. Although the British Columbia program of applied academics began

Working Outcomes in the Classroom 189 with much fanfare and support from powerful business and labour constituencies, it has had a difficult time establishing a stable presence in provincial schools. In those districts where it has survived, the links to the workplace have been drastically curtailed. We trace the development of applied academics through focusing on the various actors who had interests in it, and on the transformations the program went through. We draw on data from two school districts, with a particular focus on the Interior School District - a provincially designated demonstration district for applied academics. Given the high level of support in this district, it provides an interesting context to examine the possibilities and challenges of integrating academic and occupational curricula. Portraits of Applied Academics in British Columbia Applied academics began its official existence in British Columbia in 1994 as a series of pilot programs in eight school districts. Initially, materials were provided in three subject areas: Principles of Technology, Applied Mathematics, and Applied Communications. The materials used for the pilots were those developed by the Center for Occupational Research and Development (CORD) primarily for the U.S. 'tech prep' program (Bragg 1995). Although an evaluation of the pilot program (Gaskell and Hepburn 1995) recommended that applied academics be expanded gradually, the Ministry of Education quickly moved to mandate the program for the entire province and to develop new instructional resource packages (IRPs) in Applications of Physics for Grades 11 and 12, Applications of Mathematics for Grades 9 through 12, and Technical and Professional Communications for Grade 12 (available at www.bced.gov.bc.ca/irp/irlp.htm). New courses in information technology for Grades 11 and 12 were also designated as part of the applied academics program. As applied academics moved from the pilot CORD materials to the provincial IRPs, the curriculum documents, the curriculum in practice, and the meaning of applied academics also changed. We illustrate the changes through three vignettes constructed from our classroom observations and interviews with teachers. Vignette 1: My town School, Northwest Inlet Thirteen Grade 11 students, twelve boys and a girl, sit at benches in Mr Schubert's woodwork shop. The room sits at the dividing line between the industrial education wing of the school and regular classrooms. Instead of the

190 Jim Gaskell, Cynthia Nicol, and Li-Ling Tsai sounds of sawing and planing, however, there is talk about rates of flow and pressure in fluid systems. The class is part of a new program of applied academics and the course is called Principles of Technology. The students are receiving a grade 11 science credit for the course. The teacher makes frequent references to one of the local industries that produces methanol: The pressure in the pipes at the plant is monitored by transducers. The reason you need to monitor the pressure on pipes carrying hydrogen is that when the gas burns it is odourless and invisible. If you are working there, it can cut you in half if there is a leak that is not detected.

Other practical examples are given of compressed air systems in auto shops. References are made to the specialized training required of welders to make high-pressure welds and to the fact that welders make a lot of money - similar to steam pipefitters. As a worksheet is given out on problems related to fluid flow the teacher discusses the pros and cons of two methods of increasing fluid flow in industry: increasing the pressure in pipes versus double piping. One of the worksheet questions reads: '420 Kg of methane gas flow through the main line to the tank farm every 90 seconds. Find the mass-flow rate.' At the end of the class the teacher talks about the importance of getting going on applying to the provincial institute of technology because there can be a twoto three-year wait to get into certain programs. Talk about physics and technology and the jobs associated with them are interwoven throughout the period.

This vignette is from one of the first-year pilot classrooms. It illustrates the effort to link the classroom with realistic occupational contexts. There are explicit references to local industries where the teacher has worked and about which he talks knowledgeably. He uses language that is specific to the industry (for example, tank farms and high pressure welds) and describes the different kinds of jobs that are available, the wages workers can make, and the specialized education required. Students' assignments are related to real issues that are important in the industry: for example, if you need to expand capacity, what is the best way of increasing fluid flow? Workplace safety issues are linked to curriculum content in the school. In this relatively small town, many of the students in the class are likely to get jobs in the plant. The education that is being offered is both an education for particular jobs but also an education through this occupational context in the general principles of physics and technology.

Working Outcomes in the Classroom 191

The concepts and principles of fluid mechanics that are being developed can be applied in many situations, but fluid mechanics is not a significant part of traditional, academic physics programs. The problem of fluid flow, however, is an important part of industry, which frequently has to contend with the resistance and turbulence created when fluids are moved rapidly through pipes. Similarly, transducers are not mentioned in most secondary physics programs although they can be found frequently in industry, transforming one form of energy into another as in the monitor mentioned during the class. This illustrates the way applied academics changes priorities in curriculum content when courses are embedded in occupational contexts. Mr Schubert, the classroom teacher, commented on the content of the course in an interview: "The course covers the basic science/math skills that will be required by either the modern production-line worker or someone who's looking to get into a skilled trade - which is basically not Physics 11 because the content of Physics 11 has very little relevance to anything as far as I'm concerned. There are some useful things in there but if they're not taught how they're used in an applied manner then it just goes in one ear and out the other.' Most of the students in the course were taking other technology education courses but a few were also taking the regular Physics 11 course at the same time. One of them commented: 'One of the reasons why I took the course is because I've heard of people getting job interviews right away and stuff like that, and being preferred because they have taken this course. So I like that aspect and I think it's a little easier than physics. I'm taking physics as well so I've had a chance to compare, and Tech is a lot easier because you know what you're talking about. It's just not a whole bunch of numbers and symbols.' In comparing their experiences in applied academics courses with other regular courses they are taking, students often say an applied course is 'easier' because it is easier to understand. These courses are not necessarily easier in content, but students say that learning content through connections to industry help make content knowledge and skills accessible to them. This can be seen in the following student's comment as he compares the course to the regular math course he was taking: 'If you can think about it in the way you're going to use it, it helps a lot better. It's different than the other regular courses that we take, like math and English, right, because like in math especially, they don't give you situations where you're going to use the stuff.' Students' focus on relevance and meaning when learning the applied

192 Jim Gaskell, Cynthia Nicol, and Li-Ling Tsai academics curriculum is similar to that reported by the university co-op education students in Grosjean's study in this volume. Here, university co-op students considered disciplinary knowledge as more relevant and understandable when learned through and for occupational contexts. In vignette 1, then, the context of instruction and the priorities for instruction are connected to specific local industries although the principles are more generally applicable to industries elsewhere. Vignette 2: Orchard School, Interior City In a Grade 9 mathematics class of 25 students (12 boys and 13 girls), the teacher sets a problem for the students to work on in groups. The atmosphere is relaxed. The teacher says that he is a contractor today and that the students are his workers: I got called out of town at the last minute so you need to do this job. Go to Room 008 of the Orchard Motel [the classroom]. I need to know what it will cost for the following: 1 Two coats of paint on all wall surfaces. Paint costs $14.95/1- 1 litre covers 8.4m2 2 Recarpet floors at $19.95/m2 3 Hardwood trim around windows and doors. $1.65 per linear foot (comes in 8-foot lengths). 4 Tinting of windows (protective film) at $15.95/m2 5 Decorative wallpaper trim around the top of the walls at $1.85/linear foot 6 Baseboards around the floor at $2.17/linear foot (comes in 8-foot lengths).

The students immediately fetch metre sticks and tape measures and begin to spread out around the room measuring doors, windows, and walls. Frequent questions arise about the need to convert between imperial and metric measures. Other questions arise about whether you can buy partial cans of paint or partial lengths of trim and the effect of doing mitres on the corners. The teacher tries to encourage the students to work through these issues on their own, saying that if they need to consult him he is at another work site and they will have to call him on their cellphones, at a cost of $1.50 per call to be taken from their profits. He playfully indicates that he will fire any group whose estimate is too far off. The implication is that they need to learn to estimate quickly and accurately if they are to succeed in a job outside the classroom. In conversations during the class, the students say that the task is more fun than working from a book, as well, and is easier to understand.

Working Outcomes in the Classroom 193

This vignette, collected in the Interior School District three years after Vignette 1, illustrates a transformation of applied academics in terms of its relationship to occupations. No longer are there specific connections to particular local industries. Although there is a generalized occupational discourse, it is clearly fictional. The students are developing an estimate to submit for a job, but there is no sense that outside the schoolroom only one estimate would be accepted and a successful estimate has to trade off cost, quality of service, and profit for the contractor. This mathematics could be part of any math curriculum. The purpose is to have students see that it involves practical judgments: when buying wood trim you can only buy it in specified lengths, and therefore the cost is not simply the length of the distance around a window or floor times the cost per foot. The problem recognizes, too, that in Canada some materials are sold in metric units and some are sold in imperial units, and that students need to be able to translate back and forth in calculating a unified estimate. Here, practical activities are being used to give students practice in using math concepts. The teacher's comments about the applied activities reflected this purpose: 'We want to generate our own numbers. I don't want to use, wherever possible, numbers from the book. We'll do a lab; we'll do some kind of experiments - get some numbers and then we'll work with those to learn what it is we want to learn. So we'll go out and measure the parking lot and then we'll learn how to do volume and area and how to calculate how much concrete [we'll need]. Working with numbers, the kids can see where the numbers are coming from instead of just, you know, opening up a textbook and let's do these questions.' The practical activities lack the immediate relevance that was characteristic of Vignette 1. Their justification relies on the idea that this is a good way of engaging students in mathematics. It is a good way of teaching, but the urgency of preparing students for a restructured economy has faded. Vignette 3: Interior School District The bell rings as students in Applications of Mathematics 10 move to their desks. Students sit in groups of two or three at desks arranged to form two curved rows. The teacher moves to the front of the classroom, stopping to chat with students about the school's placement in the district basketball standings. Smiling, she stands by the overhead projector, a signal that she is ready to begin the lesson.

194 Jim Gaskell, Cynthia Nicol, and Li-Ling Tsai 'The applications activity I'd like us to work on today is called "Let's Make a Deal."' She places a handmade drawing of three doors on the overhead projector. Students are engaged, curious, and ready. 'Kyle/ she shouts, waiting for Kyle's attention, 'Come on up! You've just earned the right to win some major prizes on "Let's Make A Deal."' The teacher, acting as the game show host, waves her arms beckoning Kyle to the front of the class as students cheer Kyle forward. 'You face three doors/ she says, pointing to the projected image. 'Behind each of these doors is a prize. It's a game, of course, because two of the prizes are consolation or disappointment prizes. In this game, the grand prize is a Brand New Car, while the other two prizes are goats. Which door do you choose Kyle?' Students shout out their guesses. Kyle hesitantly decides on door B. Explaining the game rules to Kyle and the class, the teacher displays the contents behind one of the two doors Kyle didn't pick. Behind door A is a goat. 'Now Kyle/ she says, 'before opening your door, door B, you have the option of switching your choice or sticking with it. You know there is a goat behind door A, so the car could be behind door B or C. What do you want to do?' Turning to the class, the teacher invites students to think about Kyle's options. She encourages them to think about how Kyle could increase the probability of winning. 'Does he have a better chance of winning if he switches his original choice or sticks with his choice?' she asks. 'As applied students you might want to try to simulate the game.' Walking around the room handing out paper cups and candies, she continues: 'With a partner, use three cups to simulate doors and different coloured candies to represent the car and goats. One of you will play the contestant and the other the game show host.' She asks the students to perform twenty trials in which the contestant sticks with the original guess, and then twenty more trials in which the contestant switches. The students enthusiastically record their data, report it to the larger group, and discuss the findings. In this vignette, typical of what is happening today, the applied curriculum has little relationship to occupational contexts. The activity is designed to help students develop understanding through active participation (Berryman 1993). Students, however, are involved in learning neither for nor through occupations. In this classroom, the talk about ideas of good teaching does not recognize that 'good' might be related to the occupational purposes for which the students might need the math. Math becomes traditional, essential cultural knowledge dressed up as a game show. In an inter-

Working Outcomes in the Classroom 195

view, the teacher, who was helping to develop resources at the provincial level, described this kind of approach as 'basically working on projects that teachers can facilitate in their classrooms. The contents are meant to be really interesting and fun for the students and we hope that the math kind of falls out of it. So naturally the math comes out of the activity that you're doing, and of course it's a challenge for us teachers to come up with projects that, hopefully, a lot of students will be interested in doing. Some general interest kinds of projects, where the math that falls out of it hopefully has something to do with the curriculum.' While some students perceive the applications as related to future jobs, most see it as just good teaching. As one said, 'A teacher in applied is more like someone who can take the time to help people. He doesn't get frustrated and he gives us a lot of time to do our homework and he will come around and help us. He won't give us the answer but he will go, "What did you do here?" and then "Look at that. Is that right?" and then you'd be, "Oh, wait a minute!" Like he's not giving it to you, he's helping you learn from your mistakes.' How did applied academics become transformed from Vignette 1, where curriculum and instruction are relevant and connected to specific local industries, to Vignette 2, where curriculum is connected to occupational contexts but through more contrived problem settings designed to give students practice applying concepts, to Vignette 3, where there is little relationship to occupational contexts? To understand this journey we must follow a variety of actors performing simultaneously on different, but connected, stages. Enrolling Actors and Shifting Networks The development of new educational programs can be viewed as an effort to construct stable networks of human and non-human actors, just like the construction of new scientific facts and artefacts (Gaskell and Hepburn 1998; Nespor 1994). A new program must build support among a network of actors inside and outside the school system, including groups such as parents, students, teachers, ministry officials, employers, and post-secondary admissions officers. The new program must be linked to non-human actors such as technologies of assessment, international standards documents, school architecture, and available budgets. Applied academics can be seen, initially, as an unstable new artefact that has the potential to become stable as it is

196 Jim Gaskell, Cynthia Nicol, and Li-Ling Tsai

taken up, used, and/or modified by others. The program and the network co-evolve. Although the focus in Vignette 1 is local, the actors involved extend well beyond the local setting. New technologies were changing the industrial processes in Northwest Inlet companies and changing the employment possibilities for the local people. As a result of discussions between representatives from the local school board office, secondary school, industry, unions, and community college, the school agreed to develop a new sequence of courses that would provide students with a better background in mathematics and science set in the context of the increased skill levels needed in the plants. The CORD materials seemed a good basis from which to start. They were designed to make subject content relevant for students 'by showing how the skills are used in the workplace, and by teaching problem solving through hands-on, activity-centered problems arising from workplace environments' (Center for Occupational Research and Development 1995). The applied physics course, Principles of Technology (POT) (Center for Occupational Research and Development 1990), was offered through the school's technology studies program. The support of industry was important in building enrolment for the course. Industry indicated that graduates of the course would be hired preferentially. Labour unions agreed to waive certain seniority privileges to allow graduates to move more quickly up lines of progression, recognizing that existing workers had neither the skills to operate the new technology nor educational background for further training. The local community college accepted the course as an entrance qualification for a two-year diploma in applied technology in industry. Arrangements were subsequently made to ladder this college program into a four-year degree in technology management at a regional university. In 1993 the Conference Board of Canada presented the program the course and its links to college and industry - with its National Award for Excellence in Business Education Partnerships. The Minister of Education visited in 1994 and announced that the course would become a model for the rest of the province. The translation of the program from pilot to provincial curriculum involved new actors who changed the nature of applied academics. The provincial teachers' organization was wary of the close industry connections and the utilitarian focus of the pilot materials. Specialist associations in science, mathematics, and English, with an interest in maintaining the integrity and traditional status associated with sharp

Working Outcomes in the Classroom 197

boundaries, were concerned about the way the new course transgressed traditional subject boundaries. In an effort not to alienate the teacher organizations, the ministry followed established procedures and staffed curriculum development committees with experienced teachers, nominated by their professional organizations. As a ministry spokesperson said with reference to the mathematics committee: 'In order to be able to meet the provincial teacher association's protocols we used teachers who had already been through a selection process and those were exam markers - very traditional teachers. Good teachers but very traditional teachers, and it was a challenge to me to create a shift in their view of what math is and should be.' None of these teachers had been a part of the pilot process. According to one of the pilot teachers, none of the teachers who had been involved in the pilot was even consulted on the development of the provincial curriculum guide. The result of the provincial development process, one pilot teacher felt, was just another traditional course to be taught in an applied way. As he commented; 'There's absolutely no difference [in content] between the applied academics and the [regular] math at the 9 and 10 level. It does start to change at the 11 and 12 [level] in terms of how they [IRPs] are written, but in many cases it's word for word [the same]. That became my real source of frustration because it did not meet the goals of the program as I first envisioned it and how I was led to believe it would be formulated.' A spokesperson from the ministry accepted this characterization of the new provincial applied IRP and justified it not only in terms of enrolling the provincial mathematics teachers' organization, but also in terms of the need to appeal to 70 per cent of the students by keeping their options open for post-secondary education. He listed two requirements for the new applied curriculum: first, all students had to be able to count it for post-secondary admission; and, second, students should be able to cross from applied to regular courses without a bridging course. The ministry began talking about distinguishing between applications that focused on problems from students' lives, where students generated their own data related to a problem and then analysed it; and career development, which focused on making students familiar with the nature of various careers. Thus, problems such as Let's Make a Deal were seen to be applied and relevant although not necessarily related to work or a career. The discourse of applied curriculum changed from content relating to work and occu-

198 Jim Gaskell, Cynthia Nicol, and Li-Ling Tsai

pations, to content relating to issues of consumption such as mortgages, comparison shopping, and interest rates - issues with which teachers, as well as students, were presumed to be more familiar. Applied academics in the Interior School District was also affected by ministry efforts to enrol the universities in the change. During the pilot year, the Grades 11 and 12 applied courses were locally designated as regular math and science courses on the students' transcript, and were thus acceptable for university entrance. The new provincially developed applied courses, however, had distinct names and it was more difficult to make the courses acceptable for university entrance. The ministry developed Grade 12 provincial examinations in Applications of Mathematics and Applications of Physics, in order to make them comparable to other university entrance courses. Again, it chose experienced math teachers who had not been part of the pilot, but who were credible with the provincial teachers' organization, for the examination committee in Applications of Mathematics 12. Again, the paper and pencil, abstract problem-solving questions they produced bore little relationship to the hands-on evaluation procedures that the pilot teachers had been using. Teachers in the Interior School District felt they would have to dramatically shift their teaching strategies in order to prepare their students for the examination, shifting from hands-on applications to content coverage. They worried that students currently enrolled in the courses would not do well and that their school and district would be put in a bad light when results were published in newspapers. The Interior School District administrator most directly responsible for the applied academics program wrote a sharp letter to the ministry complaining about the examination format. He threatened to pull out of the program as a demonstration district if the introduction of the examination was not delayed for a year to give time for reconsideration. He argued: We are in the process of building a comprehensive district-wide Applied Academics program. The impending implementation of the Grade 12 provincial examination for Applications of Mathematics has the potential to derail all of our efforts. Our review of the sample examination has proven the assessment is not appropriate for our applied learners. It is far too theoretical and does not support the contextual approach to learning of this curriculum. We are very reluctant to allow our students to write this type of examination particularly when other districts are intending to use the examination as an enrichment course for their very best grade 12

Working Outcomes in the Classroom 199 students, which was never the intention for Applications of Mathematics. Apparently, this concern is shared by many other schools and districts in the province, all of whom have chosen not to even offer Applications of Math 12 because of the nature of the proposed examination.

The examination was not postponed. The, dilemma of the Grade 12 examination was exacerbated by the decisions of community colleges concerning entrance requirements for their academic programs. University and college mathematics instructors were concerned that the applications sequence did not sufficiently prepare students for taking university-level calculus courses. The instructors were not necessarily against the idea of applications as long as they did not disrupt the basic structure of existing courses. (As one of them said, 'We need to take a conservative approach and see how it works out. We'd like to see some of the ideas of the applications program put into the regular or principles course/) Normally a pass grade in the school-assessed Principles of Mathematics 11 course was sufficient for entrance to general academic programs in colleges and universities. Most community colleges, however, decided that they would now require completion of both Applications of Mathematics 11 and 12 for entrance to their university transfer programs, while continuing to accept Principles of Mathematics 11. While many students in the district were willing to try the two-year sequence of applied courses given the enthusiasm that previous students had shown, this enthusiasm waned when the first draft of the Applications of Mathematics 12 examination was released. The idea of performance-based provincial examinations in Applications of Mathematics and Applications of Physics was briefly considered but ruled out on financial grounds because the resource-based economy of the province was slumping. The poor economy also affected the new curriculum documents, which were developed with an eye to minimizing the need for new equipment. This decision constrained the possibilities of developing a work-based, applied program, particularly in physics, and weakened the links to business and industry. At a time of budget restraints, schools had to increase the minimum number of students enrolling in a course before it could be offered. Reaching the higher minimum class size became increasingly difficult for experimental courses that did not have assured university links. This especially affected the Grades 11 and 12 applications of physics courses. With only about 25 per cent of the age cohort enroll-

200 Jim Gaskell, Cynthia Nicol, and Li-Ling Tsai

ing in Physics 11 and 9 per cent in Physics 12, finding enough students to meet the higher minimum class sizes in a new program proved increasingly difficult. The shift away from workplace applications was also affected, in physics, by the departmentalization of subjects. Both physics and technology teachers were able to teach the original Principal of Technology courses. Some of the teachers who most successfully attracted students were technology teachers, as very few physics teachers have had much first-hand experience with the applications of physics in industry or in other areas of work. It is difficult in secondary schools to have a successful course that is shared between two departments, especially two departments with such large status differences as physics and technology. When the new provincial curriculum guide was created, the course was renamed Applications of Physics with clear indications that it should be seen as part of the science departments' offerings. It became more difficult for non-physics teachers to teach it, but physics teachers, with a few exceptions, were reluctant to take a leadership role. The movement of the applied academics curriculum - from a focus on occupations to a focus on students analysing data on game show and consumer problems - took place as the network of actors tied to the course shifted from the community-based pilot projects to the provincially implemented curriculum. In trying to ensure support for the program, the ministry worked to enrol traditional stakeholders in the curriculum: the provincial generalist and specialist teachers' organizations and the universities. In the process, the focus of the problems shifted from specific occupations to everyone's life. The downturn in the economy discouraged new funding for performance-based examinations, pushed up minimum class-size levels, and limited new equipment purchases. Discussion The story of applied academics in British Columbia illustrates some of the difficulties in integrating vocational and academic education, especially when the change is directed at traditional academic subjects. Applied academics began with the promise of matching education to the needs of a new economy. It arose in particular communities as industrial managers, union leaders, school board officials, teachers, and parents worried about keeping their industries and communities competitive and healthy. Curriculum and teaching practices empha-

Working Outcomes in the Classroom 201

sized local industries and job opportunities. Ways were found to work around provincial and university regulations. When applied academics moved from local pilots to a provincial program, the Ministry of Education worked to incorporate the interests of provincial actors. Chief among these were the teachers' organizations and the universities. Industrial organizations that had successfully lobbied the premier and cabinet to initiate change were not as involved in the process of implementing it. As the direct involvement of industry waned and as the interests of universities and teachers' organizations increased against the backdrop of a struggling economy, applied academics changed dramatically. Looking backward, it is possible to construct a story about the significance of various decisions on the final outcome. Yet each of these decisions might have been different. Officials in the Ministry of Education might have changed the ways they selected curriculum development and examination writing committees. Innovative solutions to the problems of meeting university entrance requirements might have been attempted earlier. Industry might have worked harder to sustain public support for the program and to challenge the rigidity of post-secondary admission requirements. An economic upturn might have provided more money to develop innovative assessment procedures and provide classroom equipment. In this particular time and space, however, the movement from local pilots to provincial programs has been accompanied by a transformation of the original idea. In the end, the sharp boundaries and status differences between vocational and academic versions of subjects have been reinforced. This study highlights the way that the interests of key actors are linked to the development of new educational programs. The role of industry is particularly important. In the case of the Interior School District, parents and students saw no clear reasons to risk alternative academic pathways that were resisted by universities. Industry must ensure that there are clearly defined pathways to higher education and secure, well-paid jobs for graduates of the educational programs that provide the adaptable and adept problem-solvers they want. Academic teachers and their corresponding university departments must also accept an integrated curriculum that is not only an education for work, but is also a critical education through work. Such an education may challenge industry's view of what constitutes a 'smart worker,' but if industry wants workers who are adaptable and adept problemsolvers, then it must also be willing to have workers who will chal-

202 Jim Gaskell, Cynthia Nicol, and Li-Ling Tsai

lenge administrative decisions and working conditions. Therein lie the problems, the promise, and the possibilities of applied academics. References Allen, R.C. 1999. Education and technological revolutions: The role of the social sciences and humanities in the knowledge based economy. Vancouver: Western Research Network on Education and Training, University of British Columbia. Apple, M.W. 1996. Cultural politics and education. New York: Teachers College Press. - 1998. Work, power, and curriculum reform: A response to Theodore Lewis' 'Vocational education as general education.' Curriculum Inquiry 28(3): 339-60. Bailey, T. 1995. The integration of work and school: Education and the changing workplace. In Education through occupations in American high schools. Vol. 1, Approaches to integrating academic and vocational education, ed. W.N. Grubb. New York: Teachers College Press. Barlow, M., and H.-J. Robertson. 1994. Class warfare: The assault on Canada's schools. Toronto: Key Porter Books. Beck, R.H. 1990. Polytechnical education: A step. Berkeley: National Center for Research in Vocational Education, University of California at Berkeley. Berryman, S.E. 1993. Learning for the workplace. Review of Research in Education 19: 343-401. Bragg, D.D. 1995. Linking high schools to post-secondary institutions: The role of tech prep. In Education through occupations in American high schools. Vol. 2, The challenges of implementing curriculum integration, ed. W.N. Grubb. New York: Teachers College Press. Center for Occupational Research and Development (CORD). 1990. Principles of technology. Waco, Tex.: CORD. - 1995. Applied mathematics: A contextual approach to integrated mathematics. Waco, Tex.: CORD. Commission on the Skills of the American Workforce (CSAW). 1990. America's choice: High skills or low wages! Rochester, NY: National Center on Education and the Economy. Conference Board of Canada. 1992. Employability skills profile: What are employers looking for? Ottawa: Conference Board of Canada. Deissinger, T. 2000. The German 'philosophy' of linking academic and workbased learning in higher education: The case of the 'vocational academies.' Journal of Vocational Education and Training 52(4): 605-26.

Working Outcomes in the Classroom 203 Dewey, J. 1916. Democracy and education: An introduction to the philosophy of education. New York: Macmillan. Economic Council of Canada. 1992. A lot to learn: Education and training in Canada. Ottawa: Minister of Supply and Services. Gaskell, P.J., and G. Hepburn. 1995. Applied academics pilot project: Final report. Victoria, BC: Ministry of Education. - 1997. Integration of academic and occupational curricula in science and technology education. Science Education 81:469-81. - 1998. The course as token: A construction of/by networks. Research in Science Education 28(1): 65-76. Grubb, W.N., ed. 1995. Education through occupations in American high schools. Vol. 1, Approaches to integrating academic and vocational education. New York: Teachers College Press. Heinz, W.R. 1999. From education to work: Cross-national perspectives. New York: Cambridge University Press. Kincheloe, J. 1995. Toil and trouble: Good work, smart workers, and the integration of academic and vocational education. New York: Peter Lang. Lave, J., and E. Wenger. 1991. Situated learning: Legitimate peripheral participation. Cambridge: Cambridge University Press. Lewis, T. 1998. Vocational education as general education. Curriculum Inquiry 28(3): 283-309. Ministry of Skills Training and Labour. 1994. Skills Now! Real skills for the real world: Program summary. Victoria: Government of British Columbia. - 1995. Skills Now! Year one report. Victoria: Government of British Columbia. Nespor, J. 1994. Knowledge in motion: Space, time and curriculum in undergraduate physics and management. London: Palmer Press. Province of British Columbia. 1994. Report to parents. Victoria: Ministry of Education. Resnick, L.B. 1987. Learning in school and out. Educational Researcher 16(9): 1320. Secretary's Commission on Achieving Necessary Skills (SCANS). 1991. What work requires of schools. Washington, DC: U.S. Department of Labor. Steinberg, A. 1998. Real learning, real work. New York: Routledge. Stern, D., N. Finkelstein, J. Stone, J. Latting, and C. Dornsife. 1995. School to work: Research on programs in the United States. Washington, DC: Palmer Press. Young, M.F.D. 1998. The curriculum of the future: From the 'new sociology of education' to a critical theory of learning. London: Palmer Press.

8 Co-op Education: Tensions and Outcomes of Experiential Learning GARNET GROSJEAN

The combined forces of globalization and technological innovation are rapidly changing the nature of the labour market and profoundly affecting who has access to employment and how work is carried out. The implementation of advanced technologies in the workplace, restructured management processes, and increased global competition fuel debates about the skills that are needed to meet the spiralling demands of a knowledge-intensive economy. These debates concern relevancy whether our universities equip graduates with knowledge and skills relevant to labour markets in a knowledge-intensive economy. Businesses and workers in industrialized countries are today confronting a dynamic set of forces that influence their ability to compete globally. Wealth is now derived less from physical assets than from increasing numbers and types of services. Producers 'add value' to products and services in the form of customization, shorter delivery time, or improved quality. Attention and effort are invested in design, whether of physical goods or the array of services offered in expanding areas such as medical care, education, entertainment, and travel. A shift thus occurs towards value created through the exploitation of cognitive resources - ideas and concepts - and away from physical resources and manual labour. Products and services are increasingly information-intensive. The growing number of 'breakthroughs' in all areas of technology adds to the expanding list of what are almost wholly conceptual elements in economic output. These developments increase demand in the productive sector for a broad spectrum of sophisticated and specialized knowledge. Similarly, the accelerating pace of technological change affects patterns in the business and finan-

Co-op Education 205

cial services sector, where transactions across global networks are conducted almost instantaneously. In the university, the traditional functions of social critique and the development of civic leaders are increasingly becoming subordinate to the need to prepare young people for work in the global knowledge economy. Critics argue that reorienting the university to meet market demands compromises its ability to deliver intellectual challenge and personal fulfilment, or 'Bildung.' (See also Lackey's chapter in this volume for the response of community college representatives.) There are concerns as well that a market orientation might reduce academic freedom and allow large corporations to influence the nature and content of university education. As universities attempt to meet the demands of government and industry, they also must satisfy the expectations of students who want jobs commensurate with their education. In response, universities are investigating ways of extending the academic curriculum to include experience of the workplace. This has resulted in considerable growth of programs, such as co-op education, which contain a work experience component. This chapter examines the increasing popularity of co-op education programs, and describes the sometime paradoxical outcomes of experiential learning. I will argue that the context in which learning takes place is critical to the way students develop knowledge and skills, and that students in co-op education programs develop knowledge and skills beyond those available to students in regular programs. The chapter begins with an overview of the development of co-op education programs as a strategy that promotes relevant economic and social outcomes. I describe changes in co-ops' role and function that are producing paradoxical effects not previously reported, and examine the way context shapes learning. Co-op Education Programs The origins of co-operative education can be traced to the University of Cincinnati in 1906. Herman Schneider developed the first co-op program as a way to help engineering students connect theory and practice, and thereby 'provide a better preparation, a stronger foundation, for the successful practice of engineering' (Schneider 1907, 356). While employing college students to construct railway bridges, Schneider made two observations: first, that many college students performed

206 Garnet Grosjean

part-time work that bore no relationship to their fields of study or future careers; and second, that certain elements of engineering could not be adequately conveyed in the classroom. In developing a program to correct these problems Schneider also argued that if students could earn money while accumulating experience this 'would enable many worthy young men to attend school who otherwise would be excluded [for lack of financial resources]' (Park 1946, 8). Co-op education combines academic and experiential learning by rotating students between classroom and workplace contexts. Co-op students spend a semester in the classroom developing theoretical knowledge, followed by a semester in a discipline-specific workplace implementing theory and developing skills in practical application; they then return to the classroom to engage in further academic study. Co-op students thus operationalize academic knowledge by undertaking relevant work, then bring their on-the-job learning back to the classroom for further analysis and reflection. This alternating cycle continues for the duration of the undergraduate program. Students are paid 'market rates' while on work placement. Upon successful completion of the requirements for a degree, students graduate with a 'Co-op Designation,' signifying a base of discipline-specific experience. Supporters of traditional liberal education suggest that co-op education is too narrowly focused on vocational issues, and they question whether it is the role of universities to train students for the workplace (see also Gaskell, Nicol, and Tsai, in this volume). Teaching 'employability skills' is seen as the responsibility of the employer, and not the university (Streek 1989). However, supporters view co-op as an investment in human capital that provides a socially beneficial combination of 'relevant' education and skills. The pattern of work and study in university co-op programs is based on a four-month cycle, with most students completing four or five work terms during their undergraduate degree. Co-op programs are offered in most academic disciplines with the exception of professional programs in law,1 medicine, and theology (which have internships and articles as experiential components). Nationally, more than 50 percent of all co-op placements are in the private sector, with various government and social agencies making up the balance (LeBold, Pullin, and Wilson 1990; Grosjean 1999). Canadian post-secondary co-op institutions have formalized exchange programs with Australia, Denmark, Finland, France, Germany, Great Britain, Japan, Mexico, Netherlands, Sweden,

Co-op Education 207

Switzerland, and the United States, which provide students with opportunities for cultural, educational, and technological interaction. Co-op in British Columbia In a recent study of co-op education (Grosjean 2000), I focused on the unique set of social forces and relationships represented in a single BC university, and investigated individualized (classroom) learning and social (workplace) learning in four individual co-op programs (business, chemistry, engineering, and geography). In addition to consulting the historical and documentary record, I conducted a survey to collect data on co-op students' satisfaction with their programs (n = 1012), and interviewed co-op coordinators (n = 7), faculty (n = 27), coop students (n = 45), and university administrators (n = 7). Student interviews were 'in-depth/ exploring methods of recruitment, forms of regulation, effects of learning context, academic implications, and employment outcomes. Who Does Co-op Serve?

Increasing demand for co-op education is a response to a particular social value; one that perceives a need for university graduates who can contribute new ideas that will build prosperity in the global marketplace. In British Columbia, the urgent demand for co-op resulted in a doubling of the number of programs offered in a short space of eight years - from 92 approved programs in 1992-3 to 187 programs in 1999-2000. Co-op student work-term weeks increased 72 per cent during the same period - from 90,000 in 1992-3, to more than 155,000 in 1999-2000. Private sector placements are higher in British Columbia than the national average, accounting for approximately 70 per cent of all co-op placements (TUPC 2000). With 11,450 students enrolled in coop programs in twenty-two provincial public post-secondary institutions, British Columbia has the second largest number of students involved in co-op programs in Canada.2 We know from the research literature that certain benefits accrue to participants in co-op education: co-op students get better jobs (Petryszak and Toby 1989; Wessels and Pumphrey 1995), get them faster (Gardner and Koslowski 1993), and make more money (Somers 1995) than their non-co-op counterparts. In British Columbia, graduates

208 Garnet Grosjean

from university co-op programs also have lower numbers of student loans, command higher starting salaries, are more likely to be in jobs related to their education, and are more likely to find jobs that provide additional training than do conventional students (Grosjean 2000; TUPC 2000). Furthermore, relevant skills are portable. These are important considerations at a time when the workforce is highly mobile, and the organization of capital and production are being fundamentally restructured. An interesting phenomenon that emerged during my study was the practice of over-investing in discipline-specific work experience in order to enhance future employment or career choices. In this scenario, some students extend their undergraduate program in order to accumulate more, or more specific, work experience by undertaking more than the required number of work terms. For these students there is no financial urgency to complete the degree because they are paid market rates while on work terms. As a result, they are not foregoing employment earnings by deferring graduation. By extending their time in the co-op program, however, and applying for prime work placements, they deny other students similar opportunities. Over-investment in work experience, with the aim of obtaining a training-related job upon graduation, provides economic returns. Training-related jobs not only match the graduate's field of study, but are also located in areas within the profession that require specific or specialized training. Co-op students attempt to gain progressive levels of experience (on subsequent work terms) in areas they perceive to be in demand because they know that these positions carry higher salaries. Anecdotal evidence from employers of co-op students involved in the study indicates that when they hire a co-op graduate who has some practical experience of the job, the money saved on training a novice is passed along to the co-op graduate in the form of a higher starting salary. This means that co-op graduates in training-related jobs will receive higher median salaries than non-co-op graduates from the same field of study. This differential suggests that employers value, and reward, training-related experience in the workplace (see also Lackey, this volume). Another interesting finding to emerge from the analysis was the reversal of a long-held belief about the role of co-op. There is a stereotypical assumption in the literature that co-op students graduate into the workforce while other students proceed to further education. I found it surprising, therefore, that more than one-half of the participants in the study expected to attend graduate school. For some, participation in the

Co-op Education 209

co-op program would allow them to graduate debt free, and therefore they could consider undertaking graduate education. Other students had discovered first-hand, from their workplace experience, the academic or professional credentials required for their desired career. Who Has Access to Co-op?

Student interviews and corroborating evidence from the host institution provide evidence that the original philosophy of co-op is being reversed. With restrictive access to co-op and privileged opportunities for 'good' jobs becoming the norm, co-op is showing signs of becoming an elite program. The traditional image is a bootstrap - co-op is a way for students from lower socio-economic levels to access the benefits of higher education. Replacing this is an image of the silver spoon - a program that increases opportunities for those already privileged. Access to co-op programs is becoming increasingly restrictive. Only 'the best' students are admitted. Screening - or 'creaming' - by academic grade point average (GPA) controls entry. The concept of screening is a recurring theme in student interviews. In some co-op programs, high-achieving students are actively recruited. Shortly after results of first-semester exams are known, those with a high GPA receive a letter of invitation from a co-op coordinator. These individuals act as gatekeepers. In voluntary co-op programs the coordinator is instrumental in determining who is admitted. In mandatory co-ops students must still apply to the program through the coordinator. In all co-op programs, coordinators play a pivotal role in the allocation of work-term placements, leading some students to suggest that there is a selective application of the standards when it comes to which students have access to good work terms. A recent survey conducted by the University of Victoria corroborates student perceptions of restrictive access. Results of the survey indicate that one-half of co-op students entered the university with academic averages of more than 85 per cent, compared with only one-third of non-co-op students. The difference becomes even more pronounced at the highest academic level, with more than 21 per cent of co-op students entering university with an academic average above 90 per cent, compared with less than 10 per cent for the general population of nonco-op students (UVic 1999). Because of the recruitment of top achievers and the ratcheting-up effect of screening by GPA, access to co-op programs is thus limited to

210 Garnet Grosjean

those with sufficient levels of academic and cultural capital. After enrolment, ongoing access to the 'best' work terms is restricted to the 'best' students (those with the highest GPA), thereby reinforcing the creaming process. One student - himself a beneficiary of creaming summed it up like this: 'Good work terms go to those students who have good grades. It's as simple as that.' Another disturbing trend affecting access is the rise in supplementary program fees being instituted by some BC co-op programs. Students must pay a 'co-op fee' while on a work term, in order to assure continual registration in their university program and to help cover administrative costs. The co-op fee has usually been based on the cost of a three-credit academic course. In addition, in 1996-7 two BC institutions began charging a $50 application fee for students wishing to enrol in co-op. By 1999-2000, in addition to the traditional co-op fees, seven institutions were extra-billing between $50 and $500 for processing fees, semester fees, program fees, or workshop fees. Charging more than standard tuition further restricts access, favouring those with adequate financial resources. Selective recruitment and restrictive access to opportunity are two of the hallmarks of elite programs. In the next section I provide a brief synopsis of my findings on the outcomes of experiential learning in co-op programs, largely from the student's perspective. Learning and Context Because co-op students alternate between the workplace and the classroom, they draw on their experiences in both contexts to develop their perceptions of learning and work. The workplace is where co-op students develop skills and experience in the practical application of theories they have learned about in the classroom, while at the same time they develop as professionals in their field. Meanwhile, the university combines this work experience and professional training with academic preparation, in an attempt to ensure the relevance of classroom education to real world employment. Learning in the Academic Context

Classrooms are sites where students learn standards of disciplinary practice while being disciplined into the role of university student (Biggs 1987; Entwistle 1996). In the academic classroom co-op student participation is framed by specific faculty/student/coursework rela-

Co-op Education 211

tionships. Students construct their knowledge from material presented largely through traditional modes of curriculum delivery. Rewards are based on academic norms in the form of grades. The discipline-specific coursework that co-op students learn in their academic program is intended to prepare them for increasingly challenging and responsible positions in successive work terms. In each coop program the disciplinary courses provide basic knowledge of the field, and the workplace provides opportunities to develop that knowledge through application. Both co-op faculty and coordinators support an incremental approach to academic training in preparing students to embark on successively more challenging work placements. They also recognize the importance of having co-op students bring their workplace experience back to 'enrich the classroom.' One faculty member sums up the situation as follows: 'When they've got a certain level of skill development then they go out and apply that. Then they're learning more on the work term, and they're bringing that experience and understanding back to the classroom. That's the really important part. They can share their experience, pass it on to other students. So we believe that it enriches the classroom and the campus to have co-op.' Thus, planning appropriate work-term placements is an integral part of ongoing classroom learning. According to a co-op coordinator: 'Proper placements are so important for the student to get the experience that they need, so that when they do come back to the classroom they're motivated. And there's nothing that opens the mind faster than, you know, having a good experience on a work term and then wanting to go to the next stage.' The issue of transferring recently acquired skills from the classroom to the world of work is important for students. In interviews, some described how knowledge and skills acquired in the classroom benefited their subsequent learning in the workplace: 'Having a background of courses provides you with the knowledge that you can learn how to do something. You've got a background and understanding of some things, and when you specifically apply it to the job, you can start to see some relevance of your courses there.' Others demonstrated an understanding of the importance of reflection and praxis when transferring skills from one arena to another: 'Basically it's the problem-solving skills and the stuff you learned how to do in class in a different way, that helped me in the workplace. It's not like stuff you actually learned that you could apply; it was the way you thought about the stuff you learned.' It was not just the possession of skills that make them transferable

212 Garnet Grosjean

but rather an understanding of how those skills can be applied in different situations: Transferable skills are really the key these days and I have skills that I learned on my co-op jobs that can be applied to other jobs. So it's not exactly what you learn in class but it's understanding how certain things, how your job skills and personal skills apply to other things/ The rotation from the workplace to the classroom requires a shift in learning style for some students. The importance of academic grades means that some students will focus on 'tried and true' methods of learning to satisfy the academic demand: 'The girls in engineering are in there because they're good. It's just that we tend to have less practical experience, we don't have as much hands-on experience. When it comes to machinery and things like that, we have basically no experience. So what we do is, we rely on a different strength. We got straight A's in high school, and we know how to study. And so, when we're back in the classroom atmosphere, we just study, study, study.' Because academic rewards are based on grades, personal learning styles tend to favour those known to deliver the high grades required to remain in the co-op program: 'You only take out of a course what you find interesting and pertinent and of value to yourself. But because we are graded, and because of all the things that depend on grades, there is this pressure that you have to learn everything quickly. But you can't, so to get the grades you only learn what you think will be on the exam.' Thus, performance pressures lead some students to adopt learning strategies that meet immediate needs but leave them without a grasp of fundamentals. Lacking in-depth understanding, students must memorize as much material as possible to satisfy questions on upcoming exams: 'A lot of students are straight memorizers. They try to memorize as much of the course material as they can, and retain it long enough for the exam. They finish, they get good marks, but they don't understand it.' A majority of co-op students view the academic assessment and reward system as favouring those with well-developed memorization skills: 'Some subjects - [the] sciences in particular - it's all memorization. And it does favour the students who memorize well. And it disfavours those who understand the concepts, but don't remember the formulas/ They perceive a trade-off between high academic grades and 'real' knowledge. A learning style structured around short-term memorization produces high marks, but often at the expense of understanding.

Co-op Education 213 Learning in the Workplace Context The work term is the distinguishing feature of the co-op program. It is what sets co-op students apart from their conventional counterparts, and is the axis on which co-op turns. Co-op students state clearly that the work term provides opportunities for learning and skill development different from those available in the classroom. To understand how co-op students develop skills and expertise in the workplace, we must first acknowledge that workplace cultures embed particular values. There is increasing evidence that learning, and motivation for learning, are mediated by activities embedded in a context that makes sense or matters to the learner (Billet 1999; Engestrom 1994; Grosjean 1996). Co-op students describe entering the workplace with little or no specific knowledge and gradually becoming more expert as they become familiar with, and actively use, the work setting. The specific social, symbolic, technical, and material resources available in the workplace enable co-op students to complete assigned tasks with increasing success (Scribner 1984; Grosjean 2000). Therefore, the environment is not just the context in which a problem is embedded, but is an active component of problem-solving. (See also Gaskell, Nicol, and Tsai, in this volume, for a discussion of 'smart workers/) When co-op students arrive at a workplace to begin a work term, they enter a community of practice (Lave and Wenger 1991; Wenger 1998). Membership in a community of practice requires co-op students to undergo a process of enculturation into the professional environment. Learning and skill development take place during this process. Co-op students learn not only content knowledge, but also disciplinary norms, expectations, and standards in a particular area. It is through interactions in the workplace, and participation in the activities that make up a profession, that students begin to adopt the characteristics of its members, and start to develop a 'professional persona/ Many students claim that the work term's practical experience provides a foundation that later allows conversion of theoretical knowledge into understanding. A third-year female engineering student suggested that her academic courses did not help her on the job, but 'skills learned on the job helped me with coursework when I returned to the classroom/ More than one-half of students interviewed spoke of the impact of workplace learning on their academic performance, while less that a quarter mentioned the importance of academic learning to workplace practice.

214 Garnet Grosjean

'If you learn it in school first, and then do it on the job, you don't learn it as well. For me it's better the other way around. I'd rather see my hand learn it first and then my head. So, learning the application is more important first and then I can understand the theoretical part when I come back into the classroom.' In other words, practical knowledge comes before theoretical understanding rather than the other way around: 'You learn a lot more on the job because you can see how it ties in, in so many different ways, to what you are learning. And that is better. In class we might learn a particular concept A, whereas in the workplace we learned B, C, and D. But they tie into A! But we never knew that they tied in until we did it. So yeah, I think you learn more in the workplace. And, you come back with skills that you wouldn't have learned otherwise.' This emphasis appears to put learning flows in reverse, with workplace learning providing students with a better understanding of their academic coursework. A majority of students also suggested that learning skills in the milieu of workplace practice leads to a deeper understanding than the classroom provides: "There are a lot of times where we will look at a concept in one of my classes and I think I understand it at the time, but I can't remember it later. Whereas on my work term, if we work on that concept, and I can see how it applies, it sticks. I will remember that forever.' Workplace learning also seems to possess a durable, lasting quality: 'When I learn something in the workplace I remember it. Once I have a chance to use it, it just makes sense. I could go back to the same job that I did four years ago, and still remember how to do all the basics. I may be a bit rusty, but I will remember it, whereas if I was to go back and try to recite a formula I learned in first year, no way!' Thus, the activity involved in practical application affects how students remember and subsequently recall the procedure when required. While some students were enthusiastic about linking the classroom with the workplace, a greater number expressed reservations. Some described the feeling of dissatisfaction that arises when they are not given an opportunity to integrate workplace and academic learning on their return to the classroom: T came back and I felt I'd really done some growing over the time I was doing co-op - I really had a great time on co-op, and I thought I learned a lot. It was disappointing that there was so little follow-up. I just felt that here I was back in the classroom again, and I was doing the same old, same old. Everything that I'd learned on co-op had sort of been forgotten. I just felt that, yeah, they've forgotten about what I've just done; they weren't interested in

Co-op Education 215

my experiences. I was back in school, and I was just another student expected to do the same old things in class again/ Rather than having their work-term experience reinforced, students are socialized back into a subordinate status of 'student' where they are taught about a discipline rather than learning it through application and being recognized as professionals-in-training. Students use their time in the workplace to determine where the market is going and which skills are in demand. Returning to campus to continue with coursework is an opportunity to develop their marketability. They can seek out the courses that will allow them to develop the skills or qualifications that will situate themselves to best advantage when they complete their programs. However, the perceived benefits of experiential learning are such that some will abandon an academic term, even after securing enrolment in key courses, to participate in a 'good work term' (one that provides training in a particular area, or allows students to develop a network of contacts for the future). Discussion Despite its significant expansion in size and scope, co-op has never been fully accepted into the academic mainstream. There is a tension between the vocational aspects of co-op and the traditional values of liberal education that emphasize the acquisition of disciplinary knowledge and higher-order thinking skills. But now, like the Trojan Horse in Greek mythology, co-op has breached the walls of the academy and appears set to become an elite program. Historically, supporters have argued that co-op provides a route for economically disadvantaged students to access higher education. But recent research suggests this may no longer be the case. Evidence suggests that increasing demand for co-op places leads to restrictive screening (by GPA) for admission to co-op programs, thus favouring the already advantaged. Once admitted, co-op students are able to access the collectively owned social capital embedded in the co-op credential, and to develop experiential capital in the workplace. Social capital (in the form of networks of employer contacts) and experiential capital (in the form of industry-specific skills) establish forms of credit on which co-op students can draw. These enhance the cultural capital they already command as university students. The combination of effects permits those students already endowed with certain levels of capital to accumulate even more. Therefore, depending on how those

216 Garnet Grosjean

in power use co-op it becomes either a form of relevant education, or a means of social reproduction. Co-op students frame classroom activities as learning for the workplace, and the co-op work term as learning in the workplace. They distinguish between learning from books and learning through hands-on application. What I call the 'co-op effect' is the perception that learning takes place as a result of the activities of practical application in the workplace, rather than through the activities of the classroom. What goes on in the classroom, students suggest, is not learning but study, or learning about a discipline. For co-op students, classroom learning does not take on full meaning until there is an opportunity for practical application. This leads to student concerns about the potential for transferring classroom learning to another, more practical context. Although they learn disciplinary skills in the academic context, these skills are usually transitional to workplace applications, not reflective of them. In other words, students consider disciplinary skills learned in the academic context as skills developed 'for' the workplace, while skills learned 'in' the workplace are grounded in practice and therefore provide a way to strengthen understanding of the range of potential applications. While some students were enthusiastic about linking classroom and workplace learning, others expressed reservations about the academic component of their program. Course instructors failed to acknowledge the disciplinary learning students had gained on the work term. Rather than reinforcing workplace experience, instructors taught 'content.' Students 'received' a discipline rather than learned disciplinary skills and understanding for themselves. In contrast to their autonomy in the workplace, they were made to feel subordinate. Without the opportunity to integrate workplace and academic learning on their return to the classroom, students felt the academic component of co-op lacked meaning. The key to understanding this reaction is the non-conventional flow of learning in co-op. The co-op work term provides a foundation of practical experience that allows subsequent conversion of theoretical knowledge into understanding, rather than the other way around, as is more usually the case. Rotation between the academic and workplace contexts creates a particular learning style that helps to structure the learning required for a chosen profession. To develop an understanding of the academic course material, and to be motivated to learn it, some students need to relate classroom learning to future use. Subsequent application on the work term then reinforces their understanding. There is a feeling of artificiality attached to learning workplace proce-

Co-op Education 217

dures in the classroom. One student described the difference between learning in the two contexts as the difference between 'the textbook world and the real world.' For many, classroom learning does not take on meaning until there is an opportunity for practical application. The work term appears to have a strong enough impact on students' perceptions of learning that the pedagogical activities of the co-op classroom are diminished and those of the workplace enhanced. During the study it became obvious that the co-op work term appeared to students to be synonymous with learning, and effort was required to focus on the concept of learning in relation to the classroom. One student explained that learning 'takes place on the work term when we get to see how something actually works'; that in class they 'study' rather than 'learn.' In discussions it became clear that this perception of learning is predominant among co-op students in the study. They consider the richly contextualized environment of the workplace a dynamic and robust site of learning, while learning in the classroom is disembedded from the context of professional practice. In other words, in experiential education programs the context in which learning activities take place determines how students develop knowledge and skills. In summary, a number of the outcomes of co-op education appear paradoxical. First, the structured rotation of co-op education appears to put learning flows in reverse - with workplace learning not only providing scaffolding for classroom learning but also making it come alive. Second, long stigmatized for the inclusion of a vocational component, co-op appears on the verge of developing into an elite program, thereby reversing Schneider's aim of making it a social equalizer. This reversal of intent creates tension between educational quality and standards on the one hand, and equality and access to the benefits of co-op on the other. Third, the preconceived notion that co-op prepares students for the workplace is also called into question when results indicate that more than half of co-op students include graduate education as part of their future career path. Results indicate that while some coop students learn firsthand the importance of graduate credentials to professional career progression, others take advantage of the financial opportunities offered by co-op to further their individual objectives for advanced academic credentials, and to expand career opportunities beyond the discipline-specific workplace. Finally, co-op education programs allow universities to address the problem of relevance, by adequately preparing students academically to contribute new ideas to the knowledge-based economy and society.

218 Garnet Grosjean Notes 1 However, the University of Victoria established a co-op program in Law in the fall of 1996. 2 The Province of Ontario leads Canada with 35,121 students enrolled in postsecondary co-op education programs. References Biggs, J.B. 1987. Student approaches to learning and studying. Melbourne: Australian Council for Educational Research. Billett, S. 1999. Guided learning at work. In Understanding learning at work, ed. D. Boud and J. Garrick, 151-64. London: Routledge. Engestrom, Y. 1994. Training for change: New approach to instruction and learning in working life. Geneva: International Labour Office. Entwistle, H. 1996. Ideologies in adult education. In International Encyclopedia of Adult Education and Training. 2d ed. Ed. A.C. Tuijnman, 182-7. Oxford: Pergamon. Foucault, M. 1977. Discipline and punish: The birth of the prison. 2d ed. Trans. A. Sheridan. New York: Random House. Gardner, P., and S. Koslowski. 1993. Learning the ropes: Co-ops do it faster. Journal of Cooperative Education 28(3): 30-41. Grosjean, G. 1996. Generalised specialists-specialised generalists: Education and the changing workplace. Paper presented at the Combined Conference, Canadian Association for the Study of Adult Education/Canadian Association of Adult Education, University of Manitoba (6-8 June 1996). - 1998. Education and work in a market-driven economy: Policy implications. Learning Quarterly 2(1): 21-5. - 1999. Experiential learning: Who benefits - who loses? Paper presented at the Canadian Association for the Study of Adult Education Conference, Sherbrooke, Quebec (9-12 June 1999). - 2000. 'Doing co-op': Student perceptions of learning and work. PhD diss. University of British Columbia. Lave, J., and E. Wenger. 1991. Situated learning: Legitimate peripheral participation. Cambridge: Cambridge University Press. LeBold, I.A., R.A. Pullin, and J.C. Wilson. 1990. Cooperative education in Canada. Journal of Cooperative Education 26(2): 7-13. Park, C.W. 1946. Origin and development of cooperative courses at the University of Cincinnati. Journal of Engineering Education 36 (March): 420-3.

Co-op Education 219 Petryszak, N., and A. Toby. 1989. A comparative analysis of cooperative education and non-cooperative education graduates of Simon Fraser University. Unpub. report. Vancouver: Simon Fraser University. Schneider, H. 1907. The cooperative course in engineering at the University of Cincinnati. Society for the Promotion of Engineering Education Proceedings, 15: 391-8. Scribner, S. 1984. Studying working intelligence. In Everyday cognition: Its development in social context, ed. J. Lave, 9-40. Cambridge: Harvard University Press. Somers, G. 1995. The post-graduate pecuniary benefits of co-op participation: A review of the literature. Journal of Cooperative Education 31(1): 25^41. Streek, W. 1989. Skills and the limits of neo-liberalism: The enterprise of the future as a place of learning. Work, Employment and Society 3: 89-104. TUPC. 2000.2000 BC University Baccalaureate Graduate Survey: Report of findings. Victoria, University Presidents' Council. UVic. 1999. University of Victoria, 1998 Undergraduate Student Survey. Victoria, Office of Institutional Analysis. Wenger, E. 1998. Communities of practice: Learning, meaning, and identity. Cambridge: Cambridge University Press. Wessels, W., and G. Pumphrey. 1995. The effects of cooperative education on job search time, quality of job placement and advancement. Journal of Cooperative Education 31(1): 42-52.

Conclusion Learning from Research Networks: The Western Research Network on Education and Training, 1996-2001 JANE GASKELL AND DEANNA MCLEOD

No single piece of research and no single disciplinary framework is powerful enough to communicate new understandings of the links between the economy and the educational system in a way that will impact future research, policy, and practice. Individual researchers produce important findings; disciplines produce systematic and powerful ways of understanding the social processes at work; policy-makers communicate some of these findings and understandings widely. But only in the interaction among researchers and policy-makers with different perspectives can new shared and well-founded understanding be developed. And only shared understanding will provide the cumulative knowledge base for moving research and policy forward towards an approach that is more sensitive to context, and more open to nuance. Making social science research useful, accessible, and multifaceted is a huge challenge, but an important one if we are to realize its potential for contributing to societal change. Notwithstanding the claim that 'disciplinary interests, boundaries, and constraints are dissolving, disciplines are merging in areas where their overlap forms a new field' (Weingart and Stehr 2000, xiii), the university community is very ambivalent about multidisciplinary, multi-sectoral research, especially when it is instigated by outside agencies. It seems to take scarce dollars away from basic disciplinary and curiosity-driven research. It threatens to tie researchers too closely, perhaps dangerously, to an agenda set by governments and industry. It challenges the primacy of academic journals as a source for communicating findings. Education is a field where the argument for a multidisciplinary and multi-sectoral approach to research is particularly persuasive. Perspec-

Conclusion 221

tives from psychology, linguistics, sociology, history, philosophy, and management are all routinely part of the curriculum and research in faculties or schools of education. Economics is less often present, although Weiss (1995) points out that governments have called upon economic theory and economic analysis more frequently, and for a longer time, than any other discipline. As the introduction to this volume pointed out, the Social Sciences and Humanities Research Council of Canada (SSHRC) issued a call for proposals in October 1995 to establish five-year strategic research networks linking researchers and research partners involved in policy and practice. The preamble to the call for proposals declared that 'problemoriented, applied research is needed for Canada to successfully address the challenges facing our education and training systems/ The program's objectives were 'to bring together individuals from the academic, public, private, and non-profit sectors to pool knowledge and best practices in the areas of education and training, to integrate research findings with policy-making and socio-economic development, to encourage multidisciplinary and multisectoral interaction and communication, and to transfer the research findings into practice.' A network dubbed WRNET (the Western Research Network on Education and Training) was assembled at the University of British Columbia (UBC) to respond to this call. The WRNET proposal brought together five university-based researchers from economics, commerce, adult education, educational administration, and the sociology of education. Three knew each other slightly because they worked at UBC. The other two were recruited because they were well known to and respected by a UBC researcher, and because they represented other academic approaches, institutions, and provinces (Alberta and Manitoba). Importantly, all possessed the established research records necessary to secure funding. These five in turn recruited four teams of researchers, including non-academic partners, to study educational policy, educational practice, and economic trends and relationships. Some of the results of this networking have been presented in the preceding chapters. In this concluding chapter, we explore the relationships from which this research sprang over five years. We draw on interviews with participants and observations of network activities in order to highlight the issues that arose as researchers struggled across significant differences - to generate policy-relevant educational research. Our roles as principal investigator and administrator of the network provide insights and knowledge that are not easily accessible

222 Jane Gaskell and Deanna McLeod

to an outsider, but these roles also colour our views. We think networks are important and want them to succeed; we experienced personal frustrations and moments of delight; we are trained in sociology, education, law, and psychology, but not economics. This chapter starts with a discussion of the network's structure. We then describe the challenges of communication across disciplinary and professional boundaries, and suggest individual and institutional factors that affect the ability of such networks to share understandings and produce new ways of finding and communicating knowledge. The Structure of the Network: Fragile Relations and Accountability Frames WRNET's proposal for funding promised to 'explore what educational outcomes matter and to whom, how these outcomes can best be measured, and what social processes account for the outcomes that are observed in different contexts.' The proposal framed these common questions at a high level of abstraction, pointing to theoretical debates about the importance of individual and contextual variables that exist in slightly different ways in the sociological, economic, education, and management literatures on educational outcomes. Like the introduction to this book, it referenced the interesting work that had developed since the first wave of human capital theory in the 1960s, opening up the question of skill, contextualizing what counts in the workplace, and debating the impact of technology. Following the proposal's elaboration of overarching themes, each of four teams presented specific research questions and plans that reflected their particular disciplinary and sectoral orientations. The Manitoba team of economists, educators, and policy-making partners proposed to study a number of educational programs in that province. The team of economists proposed quantitative analysis of data sets in partnership with government economists. A third team proposed studying policy with a critical, phenomenological approach in collaboration with policy analysts. The fourth team concentrated on case studies of educational practice in partnership with community groups and school boards. Each team had a defined budget for research over five years. WRNET's resources, institutional setting, and problem characteristics were not what Klein and Porter (1990) describe as ideal for interdisciplinary research. The research teams did not work on similar sources

Conclusion 223

of data, educational policies, or sites of practice. Participants were not located in a single place, and were not paid salaries by the network. The people involved remained accountable to their primary work sites, and were only loosely accountable to the network. WRNET represented an example of what has been called instrumental or problem-driven interdisciplinarity (Salter and Hearn 1996), where researchers from different disciplines come together to work on a common topic but not to interweave methodologies or to transcend any discipline. The network was designed from the outset to enable discrete groups to do their own work while the network executive and staff attempted to facilitate dialogue. We did not have the grand goal of developing a common language, or a common agreement on where policy should go. Our more modest goal was to increase communication across the different groups, trying to build shared understanding of the issues and the problems. The network's executive consisted of the five initiating academics, five partners chosen from among collaborating organizations, and one (rotating) representative of the international academic community. The staff included one full-time administrator and a half-time secretary. Accountability involved annual reports to SSHRC, and an external review after three years. While traditional academic publications took pride of place in these reports, SSHRC also wanted to know about the dissemination of research in accessible and innovative formats to a variety of audiences. WRNET involved an extended group of about 35 academic researchers, over 20 students and postdoctoral fellows, and 30 non-university partners. It had a mailing list of over 100 academic and non-academic 'interested parties,' many of whom participated in network activities and expressed a sense of membership in the network. When the network was funded in December 1996, WRNET's academic researchers were involved in the usual round of duties: teaching, administration, and research projects. The senior academics - the mainstay of the network - had several pre-existing research commitments. Junior academics were equally pressed for time, were not as centrally involved in WRNET, and were in need of prestigious publications to secure tenure and promotion. The demands of the new network - spending time with 'odd' folks outside normal patterns and habits of disciplinary communication and socialization - were not easy to accommodate. As one of our academic interview participants commented, 'What's going on is this kind of weird pull where on the one hand people are getting money so they're being asked to do things:

224 Jane Gaskell and Deanna McLeod

"You must do research that fits in line with this." On the other hand, they're thinking they're not getting that much money. So do you really have that much call on their time? You end up in this kind of weird inbetween world, where it's not enough money to really capture people's research, but it's not complete freedom either.' The network executive and staff organized meetings, an annual conference, seminars, a web site, a quarterly newsletter, regular electronic updates, and a working paper series. The executive, however, left the four research teams, and their budget allocations, alone to carry out the research as they saw fit. WRNET remained, as a result, an extremely decentralized network without the strong central leadership that some call for (Klein and Porter 1990), and that others, like the following member of our network, decry: 'Now to the extent [networks are] just mandated to happen, it's unlikely to lead anywhere. People, research creativity, and production ... it has to be a free choice or it doesn't work' (academic). WRNET's decentralized structure meant that sustained interaction between researchers depended to a large extent on individual enthusiasm for encounters with people who did different kinds of research and had different understandings of its role. The conferences provided an environment where those encounters were forced, but creating more opportunities for interaction within the network was an uphill effort. When it did occur, this interaction revealed markedly different assumptions about knowledge, truth, rewards, and motivations. Words such as 'camps,' 'tribes/ 'communities,' 'worlds,' 'territories,' 'inside and outside,' 'ends of the continuum/ and 'cultures' were used to designate WRNET's various groupings. The kinds of research being carried on in the four teams were 'profoundly different/ 'dramatically different/ or 'like mixing oil and water/ and they reflected 'different language/ 'different values/ 'different prisms.' And, 'part of people having pride in what they do tends to be rivalry towards people who do something else. It's human nature. It goes with the territory.' The next two sections will explore the ways these differences were experienced and described, while the third section will look at the dialogue that the network enabled. Economists and Educators: The Difference a Discipline Makes The academic world has been organized through disciplines and their administrative units into 'tribes' (Becher 1989) that police their own

Conclusion 225

boundaries in a myriad of ways. Becher writes that 'being a member of a disciplinary community involves a sense of identity and personal commitment, a "way of being in the world"' (24). As such, disciplines affect interaction in profound and unexpected ways, from small talk to formal academic communication. 'Disciplines display fundamental differences not only between types of evidence and procedures for proof, but also in the ways in which others' work is evaluated and in the modes in which arguments are generated, developed, expressed, and reported' (273). There is no neutral terminology. Disciplines differ in their ability to engage with and accommodate difference, a characteristic variously described as 'permeability/ 'boundedness/ and 'coherence' (Salter and Hearn 1996; Becher 1989). They vary in the range of topics considered part of the discipline, the controls exercised over members, and the extent to which members are allowed to deviate from the accepted norms. Economics is an example of a relatively tightly bound, cohesive discipline, while education is much more loosely bound. While education is a field full of ongoing debates about fundamental methodological issues (Hammersley 1998; Labaree 1998; Miller, Nelson, and Moore 1998), economists have a much more consistent and shared sense of what counts as good research. Economics works with numbers and offers definitive findings. Education, on the other hand, 'emerges as the softest of the soft fields of inquiry. Problems of teaching and learning, curriculum and governance, educational organization and educational reform - all of these resist efforts by researchers to establish causal claims that are verifiable, definitive, and cumulative in anything like the way that researchers in hard-knowledge disciplines can accomplish these things' (Labaree 1998,5). Education also has lower status than economics in the university community. It is relatively new on many university campuses, and its traditions are linked to teachers colleges. Educational research is 'marked by a variety of stigmas that undermine its ability to provide credentials with high exchange value - for example, an association with women, the lower classes, public employment and a "semi-profession," along with its weak academic standards and modest institutional origins' (Labaree 1998, 6). Economics, on the other hand, has high prestige, reflecting its tighter academic boundaries and entrance requirements; its masculine, mathematical traditions; its more mainstream university history; and the higher salaries its graduates expect. These differences in boundedness and status were quickly displayed in WRNET. Status differences made it challenging for some econo-

226 Jane Gaskell and Deanna McLeod

mists to even participate in a network hosted by an education faculty, and reporting in educational outlets: 'My dean turned around, and basically said ... his interest in having me involved in a proposal where the principal investigator was in an education faculty was negative and he told me to drop off.' Language differed between the disciplines as well. The complex numerical presentations made by economists during early conference sessions were criticized, even when they tried hard to explain what they were doing: '[economists should] make their overheads more presentable; they put up all of those damned numbers in such small print. I mean, why don't you just show me some patterns. You don't have to show me every line ... I'll know you did the work, you don't have to prove to me that you did the work. All I want to know is what you found' (practitioner). The language of educators was misunderstood by those who knew the everyday meaning of the words (e.g., curriculum, accountability frames), but not the nuances added by years of exposure to the academic literature. Economists avoided sessions where educational researchers 'use too much jargon.' Although WRNET was studying 'outcomes,' participants did not share a common understanding of what counted as an outcome. Educators protested the way economists limited the discussion to numerical outcomes, arguing that social processes and personal experiences can be both the mechanisms for producing outcomes, and important outcomes in their own right. As one educator put it: '[I've been] very critical about some of the economically oriented studies ... I'm interested in the everyday lived reality of policy and talking to human beings and doing ... more qualitatively oriented work ... The language of economics is not about people, or at least that's not what I've heard/ The practice of entering a variable called 'education' into an equation was also very hard for many educators to understand and accept: 'They look at education as one thing that you can measure the outcomes of, whereas I'm very interested in all the different shades and variations within that thing. I found it very difficult to understand how people could say, "If you take a certain class of program, the outcomes of this program are X, Y, and Z" ... I find it difficult when I see other people lumping that together and saying it's one more or less homogeneous program.' Or, as another stated, "They're running all these numbers but the categories that they're using ... don't make any sense ... Anybody who's actually been in a school would know that this doesn't make any sense.' Researchers accustomed to working with large quantitative data sets

Conclusion 227

not surprisingly asked questions that could be answered by large data sets. The gap between quantitative and qualitative researchers overlapped with, though it was not the same as, the gap between economists and educators: "There's this real conflict... In some ways it's hard for one group to say that what the other group is doing is valuable. On the economist's model, storytelling is one observation and we know what we would do with one observation, which is nothing. So if you see someone emphasizing one observation, it's not just that it's a different way of doing things, within the context of that kind of a model, it's an absolutely wrong way of doing things! And my guess is that the same thing would be true if what you view as relevant is understanding individual experience and understanding' (economist). There was also disagreement about what research could achieve and how definitive it could be. An educational researcher who conducts qualitative research commented, '[in my research] you're not looking for the solution to something, you're trying to find out how things work and maybe look at it from a different point of view.' Economists, on the other hand, are looking for a more precise 'test' of a hypothesis: Tor economists, good work looks a lot like the scientific model... [a] clear hypothesis that is refutable and [you] essentially go to the data and find data that will allow you to test that hypothesis.' In all, the cultures of the two main academic groups that came together in WRNET were dramatically different. The people had rarely interacted with each other before. Differences in status and language, as well as in political and epistemological assumptions, separated researchers along largely disciplinary lines. While individuals had complex commitments that crossed these boundaries, drawing the differences starkly illustrates how substantial were the academic chasms the network had to bridge. Research and Policy: The Difference an Academic Setting Makes Partnerships with non-university sectors were a condition of SSHRC's funding for our network, and members of WRNET enjoyed interacting across the university divide. During our second conference, an academic exuberantly advised other academics to 'hug a government person today; they're nice and they know more than you do!' However, practitioners and academics also differ in the kinds of knowledge they value, the ways they communicate, and the working conditions and reward structures they confront. As Levin (1993) describes another collaborative project in Manitoba: 'Our research was never at the top of

228 Jane Gaskell and Deanna McLeod

any of our partners' agendas. Indeed, it would never have appeared on their agendas at all had we not put it there' (334). Our partners were as different from each other as the economists and the educators were different from each other. Some were primarily researchers, albeit situated in government, while others were policy makers and practitioners. Some worked in provincial governments, some with the federal government, and some in school boards. Some were in education, some in labour market policy, some in teaching. Academics tend to complain that non-academics ignore their research (Weiss 1978; Gaskell 1988; Salter and Hearn 1996), yet they rarely make communicating outside academia a priority. Academic publications remain critically important for careers in academia. The academic researchers felt caught between the expectation to 'bury' results in top disciplinary journals, and the network's mandate to communicate to and with other disciplines, public policy partners, practitioners, and the lay public: The reality is [academics are] paid based on academic articles. They're not paid based on public lectures. (Academic) Joint authorship and collaboration...although it's getting better, it's still not as valued. (Academic) My impression of academia is that interdisciplinary journals tend not to be as well regarded as ones that are solely within a discipline. (Partner)

Partner groups were accustomed to modes of communication that were accessible and oriented towards their own goals. Practitioners and policy-makers were heard to complain about the expense, longer timelines, irrelevance, and inaccessibility of much of the academic research they heard at WRNET. As one government employee explained, 'I think to a certain extent there is a view that academics don't deal with real issues ... You could get a pool of people whose specialty was navel-gazing, and you could have 500 of them and they would have the navel gazer's journal. They would all publish in it and only the best ones would get published and they would cite each other's work and they'd each know what was the best work. But people on the outside, we kind of look at them and say, why the heck are you doing that? ... they may be doing very difficult stuff and doing an excellent job of it, but it's just not relevant for anything.' Individuals from partner groups were not all strongly supported by

Conclusion 229

their supervisor or department to join WRNET, and even those who were had their own competing demands and scarce resources to deal with. Whether they were working in federal or provincial human resources or education ministries, in Statistics Canada or in labour, the private sector or community organizations, WRNET was just one of many commitments and demands on their time - and their workplace timeline was more pressing than the academic timeline. As one of our government partners observed, with some frustration, 'The world of public policy ... turns in 100 days and academe turns in 1,000 days. So by the time the academic has done his or her work, the public policy has changed ten times.' Since our non-university partners did not receive direct research funding, their commitment to WRNET was dependent on personal interest and the connections that the network offered. When government participants were asked what they got out of their participation, they pointed to indirect, personal, and long-term effects, not to any immediate impact on policy. Among the things they valued: 'psychic rewards'; 'helps to develop my own research agenda'; 'there's an underutilization of our data so we're happy to feed studies'; 'nice to talk things out and get feedback; if I have to share information and data I'll be more careful'; 'working with academics, who are custodians of knowledge, brings new ideas.' They enjoyed taking the time to sit and discuss educational issues in a reflective way, and learned from the intellectual stimulation this offered. They also appreciated the opportunity to meet with fellow policy-makers from other governments or different branches of the same government. There were different kinds of partnerships in the network, for different purposes. One educator had a very close relationship with a community group, shaping her research as they suggested, and presenting it with them in both academic and policy settings. She commented: 'I like to help people who don't usually have access to academic resources ... My sense of good research is my commitment to the establishment of long-standing networks and bridges with community, to have my work be meaningful to the community ... and it also is a commitment at a political level.' An economist described quite a different position, where distance from the community was necessary for his research, although he wanted them to know what he found: 'I don't see why you'd want to be that close to generate real research. It becomes socially hard to look someone in the face and say, "what I just found is that regulation, in spite of the fact that you guys have been yelling that it's awful, is actu-

230 Jane Gaskell and Deanna McLeod

ally not awful." It's hard to do that... it has to be, to the largest extent possible, a situation where you maintain academic freedom, where you can stand up and say what you think you've really found without worrying that someone's paid the piper.' One of the easiest partner relationships was between researchers in and outside of government. They described a mutually recognized complementarity: 'Academics are up to date on the techniques, widely read, and have a broader perspective. We're a gain to them in terms of access to data and policy relevance'; 'People in the civil service have just much more and better data/ But even here, there are differences in approach and standards: To have something be deemed to be good from the academic perspective, it has to have the up-to-date statistical techniques and it has to make reference to other academic articles in a similar area ... From our perspective, elements of 'good' would involve good data and address a relevant policy issue. And take into account the appropriate policy parameters. Those things are definitely of secondary importance to academics' (government employee). On the other hand, some partners were concerned about their low status in a university research network: T go into this room where it's Dr so and so and Dr so and so and Dr so and so. I experience a sense of, what am I doing here, what can I contribute? ... The tendency in our society is to discount the practical, the hands on' (practitioner). Educational researchers were frustrated as well by their low status in the eyes of policy-makers. An educational researcher complained, 'Government won't even look at research unless there's an economist involved.' Another commented: 'government officials like numbers. Numbers fit very well in a sixty-second sound bite, and if you can say that this shows that this is ten times more successful than this, that's what politicians want.' The many different kinds of partnerships that are possible between academic researchers and those working in other worlds were illustrated in WRNET. A tight fit between research and practice would involve a more selected group, on both sides. What did occur were some selective connections that were very productive, a continual pressure to widen the audience for research, and a discussion about the nature of partnerships that revealed the varying epistemologies assumed by the four different research teams. Communicating across Differences Encountering those who communicate in ways one is not used to can

Conclusion 231

be stimulating, empowering, and enlightening; it can also be infuriating, boring, and embarrassing. As Miller, Nelson, and Moore (1998) write, since 'researchers all have blind spots born of different research languages, processes, and epistemological histories, we need productive dialogue to help all of us understand the gaps in our knowledge and locate our expertise on the broader map of the field' (408). Our hope, like theirs, was that 'a dialectic of respectful, reciprocal critique can highlight the assumptions and values that shape different traditions' ways of grounding interpretations' (411) and produce tolerance, reflexivity, and intellectual self-assessment. It was relatively easy to enable research groups to have their own conversations. A summer institute for quantitative researchers was popular, partly because participants didn't 'have to worry about being empirical or putting up tables.' It was more difficult to force the confrontation of different kinds of knowledge and research, expanding these conversations and making them interesting and accessible to a wider audience. Newsletters digested academic work to provide a taste of the different disciplinary and methodological approaches for a general audience. As one academic said, 'somebody like me can pick up the newsletters from the different areas and kind of pick up the flavour of how they live in their world.' Newsletters that integrated different kinds of work were much more challenging and time-consuming than newsletters produced by a single research team, because they required a commitment to co-authorship across disciplines and sectors. An annual conference, held in the same few rooms, and with more or less the same core group of people attending, became the primary site of communication across research groups. As one academic observed, it was 'not a big circus but a smaller group of people where you can actually interact with other individuals.' Participants had an opportunity to chat over sandwiches or coffee, and then over wine and dinner. This interaction played a critical role in developing a community of interest and respect for the other participants (Robinson 1996): I think that comfort and trust in other people is very important. And a lot of that is accomplished through social activities, getting to know people. So when someone comes in, it's not just 'the economist/ it's a certain person who's an economist and who's a nice person and you trust them and you're willing to engage in a conversation with that person. I think it really changes the level of interaction and the comfort of interaction. (Academic) People are getting a little more comfortable with each other and learning

232 Jane Gaskell and Deanna McLeod more about each other ... Once that trust is established, you get the sharing of a lot of information. We find out a lot more about what the situation is within government now, and what the constraints are on them. So we needn't get frustrated because they're not responding to something as quickly as they should, or in a specific way. (Academic)

Over the years, conference organizers became increasingly prescriptive about how presentations should be handled. Plenary sessions were constructed to force interaction. Papers were distributed ahead of time. Specialized terminology was negotiated as people came to realize what words were and were not widely understood. The economists' language and style of presentation remained a source of irritation to some, but it also became amusing. One economist declared during a conference session that he was not 'ashamed' to call himself a labour economist, and 'confessed' that he understood all the graphs and numbers. The teasing and humour became a means to redress status differences as well as approaches to knowledge. But different research groups held onto different assumptions about what should be distributed in advance, what should be explained, and what could be assumed. Participants were sometimes reluctant to attend sessions beyond their own group, because they felt loyalty to a team member who was presenting, or because they expected to get more useful feedback from like-minded people: T'm kind of specialized on the quantitative side and so for me to learn about qualitative research doesn't help me that much, cause I've got just a tremendous amount that I have to learn on the quantitative side. And my brain isn't big enough to be a specialist in two such diverse areas' (partner). Time constraints, rather than lack of interest, were most of ten blamed for missed opportunities for interaction: "There's only so much time people can devote and if it's not in your specific group interests, it's hard to make the time'; T'm definitely interested but it falls a little bit down the list'; 'that's not out of non-interest, it's just a ranking.' Given that the length of the day is not likely to expand beyond twenty-four hours, those funding and administering networks need to think deeply about what structures can overcome the centrifugal pull of home departments and ingrained habits. Creating the structures and relationships to sustain dialogue is costly in terms of infrastructure. Administering the network, building partnerships, making Web sites and newsletters, and keeping people engaged with each other, require time and money that cannot be spent on direct research: Tf you don't spend

Conclusion

233

money on the research itself, you wouldn't have any claim on the time of the people who come together at the conferences' (academic). That personal relationships are so critical for professional interaction offers possibilities for movement forward; it also means research networks have difficulty incorporating people whose personal style is different. One female participant commented that there were a lot of 'suits/ and that the network could be a bit of a 'boys club' in which she sometimes felt her research was marginalized. Some of the trust and comfort established in this network can be attributed to a generally shared political view that is critical and on the left side of the political spectrum. This book reflects the dynamics of WRNET, where the conversation among approaches is present, if muted. The voice, the evidence, and the argumentation of each chapter reflects the disciplinary training of the authors. Chapters that are co-authored are written by like-minded authors. The studies of educational practice use qualitative methods to explore what happens in particular contexts when policy is implemented, and they make their theoretical premises part of the argument. The economic chapters use quantitative analysis, make factual claims about what is happening, and avoid theoretical discussion. The economists largely reference the work of other economists, the educators largely reference the work of other educators, and the policy critics largely reference a literature on policy analysis. However, the chapters are written in a way that is comprehensible to a general policy audience, and the authors are well aware of the policy debates in which their work is located. A debate about the value of vocationalism raises its head in one way or another in every chapter. There are asides in most chapters that reference the debates of the network. Riddell provides caveats about overgeneralizations from OECD data and recognizes that the structural issues that distinguish Canada's educational system remain to be accounted for. Gaskell, Nicol, and Tsai wonder about the economic analysis that would underpin applied academics, and reference the critique by Allen. Lackey's policy analysis, though it concentrates on discourse, has taken into account concrete studies of practice by Grosjean, and the economic critique mounted by Allen. For, despite all the differences, over time WRNET began to function as its own culture with its own 'inside' and 'outside.' Participants started to use phrases such as 'a member of the family' or 'a WRNET person.' Importantly, these discussions continue in various forms now

234 Jane Gaskell and Deanna McLeod

that the network has concluded its formal existence. WRNET created some 'trading zones of common interest' (Klein 1996,70), where different constructions of knowledge and of acceptable inquiry interacted. While the zone started out quite small, and was buffeted around and challenged by disciplinary chauvinism, disciplinary loyalties, and the distrust absorbed through many years of inhabiting academic or government culture, it gradually increased as a shared history developed. Research networks simply become 'less dysfunctional' (as one participant observed) as the vocabulary and research approaches of others are understood better - and as the participants simply grow familiar with each other. The exposure does not necessarily change the research, but it opens up avenues for exploring other perspectives more deeply, both in the immediate and the longer term. As we have heard repeatedly, those participating with WRNET had opportunities to think about ideas, and to speak with people, which would have been impossible otherwise. The interaction did not create dramatically new paradigms or a shared approach to understanding educational outcomes. But it modestly increased understandings and questions about other worlds of research and practice. Participants found themselves challenged about methods, values, and practices they had simply taken for granted. They were forced to think about, and answer to, issues and ideas that like minds would never have provoked. The resulting confrontation provided the opportunity for 'productive conflict' (Bromine 2000; Miller, Nelson, and Moore 1998). Sometimes I go, 'oh, my god, that's incredible!' I mean, they've just opened up a world that I had not even thought about before. (Academic) They teach me in another language. So we're learning how to listen to each other. (Partner) Learning more about who they are as individuals gives me a better idea of how their research is focused and allows me to understand, I think, a lot more about their research than just from getting their paper or listening to them at a conference. (Academic) I think we each have troubles with each other's work, but I hope we can also see value in each other's work in terms of some broader perspectives. There's always a conversation back and forth. (Educator)

This chapter has been written in the hope that more tale-telling can

Conclusion 235

increase understanding of how the organization of research shapes its practice and its results. By bringing together economists and educators, WRNET tried to bridge one of the largest structural and epistemological differences in the social sciences. Our experiences suggest several conclusions. Expectations for collaboration must be modest if networks are developed within existing university structures. Achieving a common approach is unlikely, but greater understanding of difference is possible. The range of difference that can be corralled into a network is limited and should be carefully targeted. Economists interested in the organization of educational institutions are most likely to participate in networks with educators. Educators who have some aptitude for numbers will be more likely to engage with economists. Bringing together private sector partners and community activists who do not share any political starting point is more likely to bring heat than enlightenment. At the same time, discussion of research is a good way to bring together people who would not normally sit down together. Much of the really meaningful interaction is underpinned by good personal relationships. We found it is important to feed participants well and entice them into opportunities to chat informally, face-to-face. Cassell (1977), on the other hand, concludes from his experience in an interdisciplinary network of physicians and philosophers that 'good food is not necessary, while alcohol seems quite useful' (361). Accountability schemes and budgeting practices for research networks should take into account the importance of infrastructure, including food, travel, and scarce time. Finally, a network that succeeds depends on participants who are committed to interacting, and interested in the findings of others, and this will take time. Our network involved educators, economists, policy analysts, women in poverty, Human Resources Development Canada, the BC Ministry of Education, and the Kamloops school board, among others. Not all of these groups will continue to work together, but some will. WRNET was a start in figuring out who wants to continue with whom. The academy needs more of this dynamic of opportunity. References Becher, A. 1987. Disciplinary Discourse. Studies in Higher Education 12(3): 261-74. - 1989. Academic tribes and territories: Intellectual enquiry and the culture of disciplines. Buckingham, UK: Open University Press.

236 Jane Gaskell and Deanna McLeod Birnbaum-More, P.H., F.A. Rossini, and D.R. Baldwin, eds. 1990. International research management: Studies in interdisciplinary methods from business, government, and academia. New York and Oxford: Oxford University Press. Bromme, R. 2000. Beyond one's own perspective: The psychology of cognitive interdisciplinarity. In Practising interdisciplinarity, ed. P. Weingart and N. Stehr, 115-33. Toronto: University of Toronto Press. Cassell, E. J. 1977. How does interdisciplinary work get done? In Knowledge, Value and Belief, ed. H.T. Engelhardt Jr. and D. Callahan, 355-61. The Hastings Center. Institute of Society, Ethics, and the Life Sciences. Gaskell, J. 1988. Policy research and politics. Alberta Journal of Educational Research 34(4): 403-17. Hammersley, M. 1998. Telling tales about educational research. Educational Researcher 27(7): 18-21. Klein, J.T. 1996. Crossing boundaries: Knowledge, disciplinarities, and interdisciplinarities. Charlottesville and London: University Press of Virginia. Klein, J.T., and A.L. Porter. 1990. Preconditions for interdisciplinary research. In International research management: Studies in interdisciplinary methods from business, government, and academia, ed. P.H. Birnbaum-More, F.A. Rossini, and D.R. Baldwin, 11-19. New York: Oxford University Press. Labaree, D. 1998. Educational researchers: Living with a lesser form of knowledge. Educational Researcher 27(8): 4-12. Levin, B. 1993. Collaborative research in and with organizations. Qualitative Studies in Education 6(4): 331^0. Miller, S.M., M.W. Nelson, and M.T. Moore. 1998. Caught in the paradigm gap: Qualitative researchers' lived experience and the politics of epistemology. American Educational Research Journal 35(3): 377-416. Robinson, J. 1996. Falling between schools: Some thoughts on the theory and practice of interdisciplinarity. In Outside the lines: Issues in interdisciplinary research, ed. L. Salter and A. Hearn, 85-92. Montreal and Kingston: McGillQueen's University Press. Salter, L., and A. Hearn. 1996. Outside the lines: Issues in interdisciplinary research. Montreal and Kingston: McGill-Queen's University Press. Weingart, P., and N. Stehr, eds. 2000. Practising interdisciplinarity. Toronto: University of Toronto Press. Weiss, C.H. 1978. Improving the linkage between social research and public policy. In Knowledge and policy: The uncertain connection, ed. L.E. Lynn Jr., 2381. Washington: National Academy of Sciences. - 1995. The haphazard connection: social science and public policy. International Journal of Education Research 23(2): 137-50.

Contributors

Robert C. Allen is professor of economic history at Oxford University. His research interests include economic history, economics of technological change, and public policy. He has written extensively on agriculture and economic development and has recently completed an economic history of the Soviet Union. His work related to the Western Research Network on Education and training (WRNET) includes the economic benefits of post-secondary education and training. Shauna Butterwick is an assistant professor in the Educational Studies Department, Adult Education Program, at the University of British Columbia. Her research interests include examination of how women are served through training policy and programs from the perspective of feminist approaches to social justice. She was responsible for two WRNET BC case studies: one involved working collaboratively with a group of low-income women to examine welfare-to-work policy from their perspective and to document their self-directed learning process; the other involved an examination of life skills as a theme that permeates policy and programs for marginalized groups. She is also interested in how popular education and popular theatre (theatre and education for social change) can be used as an action research methodology. Jane Gaskell (EdD Harvard, 1974) is a professor and dean of the Ontario Institute for Studies in Education/University of Toronto. She was the principal investigator of the Western Research Network on Education and Training from 1996 to 2002 and has written widely on educational policy in Canada. Her books include Secondary Schools in Canada: The Report of the Exemplary Schools Project (Canadian Education

238 Contributors

Association 1996); Debating Dropouts: Critical Policy and Research Perspectives on School Leaving (Teachers College Press 1997) and Gender Informs Curriculum: From Enrichment to Transformation (Teachers College Press 1995). Jim Gaskell is a professor of curriculum studies in the Faculty of Education, at the University of British Columbia. His research interests are in the social context of science education. In recent years, he has focused on issues of gender and the integration of academic and occupational education. He was the principal investigator of the Applied Academics case study for WRNET. Garnet Grosjean holds a PhD in adult education. He is a research fellow at the Centre for Policy Studies in Higher Education and Training, and academic coordinator for the Doctor of Education in Education Leadership and Policy at the University of British Columbia. He was a member of the WRNET policy team and conducted the research on coop education. His research interests include experiential education, work-based learning and learning cultures, higher education and the economy, and higher education and the labour market. Derek PJ. Hum is a professor of economics at the University of Manitoba. He is a graduate of Mount Allison University, Oxford University, and the University of Toronto. During the late 1970s, he was research director of Mincome, an experimental guaranteed income program in Canada. Dr Hum has written extensively on social policy, income security programs, and work incentives. He has written on federalism and the poor, and on income maintenance programs in Canada and the United States. Lara M. Lackey is assistant professor of art education in the Department of Curriculum and Instruction at Indiana University. In 1998-9 and 1999-2000 she was a postdoctoral fellow at the Centre for Policy Studies in Higher Education at the University of British Columbia. There she participated in research related to educational policy analysis, with special emphasis on the multiple perspectives of policy players, policy as discourse, and issues of gender equity and bias. Ben Levin is a professor in the faculty of education at the University of Manitoba. He was deputy minister of both Manitoba's Department of

Contributors 239

Education, Training and Youth, and the Department of Advanced Education. He was previously dean of continuing education at the University of Manitoba and was team leader of WRNET's Manitoba Case Studies. His research interests include education policy, politics, and economics, with a longstanding interest in the educational implications of labour market change. Deanna McLeod was the administrative coordinator of WRNET from 1997-2001. Her research interests include the theory and practice of interdisciplinarity. She works in research, administration, and communications, and holds degrees in biological psychology and law. Cynthia Nicol is assistant professor with the Department of Curriculum Studies at UBC. Her work as a teacher-educator and researcher is situated in practice, using her own teaching practice as a site for investigating teaching and learning to teach. She studies efforts to improve teaching through curriculum change, reform initiatives, and teacher education. Her teaching and research interests include helping beginning teachers make cross-disciplinary and workplace connections to develop their understandings of mathematics and pedagogy. She has written chapters, articles, and papers on teacher education; the processes involved in learning to teach; the linking of mathematics, school, and work; and the ethical issues involved in researching teaching. W. Craig Riddell is professor of economics at the University of British Columbia, and led WRNET's labour market analysis team. His professional specializations include labour economics, labour relations, evaluation of labour market and social programs, education and training, unemployment and labour market dynamics, and public policy. Professor Riddell is an associate of the Canadian Institute for Advanced Research, former head of the Department of Economics at UBC, former academic co-chair of the Canadian Employment Research Forum, and past-president of the Canadian Economics Association. He currently holds a Royal Bank faculty research professorship at UBC. Kjell Rubenson is a professor in the Educational Studies Department at the University of British Columbia and director of the Centre for Policy Studies in Higher Education and Training. He leads WRNET's policy analysis team. His research interests include policy studies; adult and higher education; and education, learning and work.

240 Contributors

Wayne Simpson is professor of economics, University of Manitoba. He is a graduate of the University of Saskatchewan and the London School of Economics. He is a specialist in labour economics and quantitative methods, and has worked for the Bank of Canada and the Economic Council of Canada. He has written on urban structure and the labour market; income maintenance, work effort, and the Canadian Mincome experiment; and maintaining a competitive workforce. Atlanta Sloane-Seale is assistant professor and program director of the Continuing Education Division, University of Manitoba. Dr SloaneSeale's areas of interest are accessibility, issues of equity, and program planning. Specifically, Atlanta's work includes studies on accessibility and the Aboriginal learner; women managers and barriers to their career progess; and program planning, including quality instruction, needs assessment, curriculum development, and program evaluation. Li-Ling Tsai is a doctoral candidate in the Department of Curriculum Studies at the University of British Columbia. She was a research assistant of the Applied Academics case study for WRNET from 1999-2001. Her research interests focus on gender issues in science and science education, such as how gender disparity in the participation of science is understood, and why the development of a gender-reflexive science curriculum is necessary. She is currently conducting her doctoral research on identity formation and agency demonstration in the making of gender in an institutional physics community. Lori Wallace has been a faculty member at the University of Manitoba since 1985. She has held appointments in the Access Program and in the Distance Education Program, where she is currently serving as director. Dr Wallace has also worked in the private sector in human resource development, and has considerable international experience in distance education in developing countries. Her research and publications are in the area of distance education, and include course development and tutor training materials, development of in-service specialists, and trends in student demographics and motivations.