191 35 2MB
English Pages 365 Year 2007
FACULTY OF PHYSICAL EDUCATION AND PHYSIOTHERAPY Department Sport Policy and Management
Sports Policy factors Leading to International Sporting Success Veerle De Bosscher
Advisor:
Prof. Dr. Paul De Knop, Vrije Universiteit Brussel Co-advisor: Prof. Dr. Maarten van Bottenburg, Utrecht University
Dissertation presented in partial fulfilment of the requirements for the degree of Doctor in Physical Education Year: 2006-2007
Cover design: Frisco, Oostende Print: Grafikon, Oostkamp Credits photographs: Belga © 2007 Uitgeverij VUBPRESS VUBPRESS is an imprint of ASP nv (Academic and Scientific Publishers nv) Ravensteingalerij 28 B-1000 Brussels Tel. ++32 (0)2 289 26 50 Fax ++32 (0)2 289 26 59 E-mail: [email protected] www.vubpress.be ISBN 978 90 5487 421 8 NUR 843 / 744 Legal deposit D/2007/11.161/011 All rights reserved. No parts of this book may be reproduced or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior written permission of the publisher.
Preface Just as athletes that are preparing for one ultimate moment of Olympic performance over more than four years, I have been looking forward to this supreme moment in my career, which is the end of an Olympic cycle, but also the beginning of many new challenges. It has often been a tough process, with ups and downs, hard training, immense demands on time, total dedication, enthusiasm and personal drive. But at any moment I have thoroughly enjoyed the process. I have had the opportunity to benefit from this tremendously instructive experience and to work with a formidable team of experts from whom I have learned an immense amount. This study could not have been completed without an extraordinary amount of help from a wide variety of individuals and institutions that I would like to thank. It appeared from this study that the personal coach is one of the most important factors contributing to the success of an athlete. dissertation.
This was definitely also the case in my
My coaches and supervisors, Paul De Knop and Maarten van Bottenburg
have given me the confidence to realise this study and to find my own way in this complex subject.
They have opened doors and enabled me to increase the social,
political and scientific relevance of this work. I have the greatest appreciation for their high levels of expertise, know-how and professionalism.
I want to thank them in
particular for their support and advice during this whole process. This study has been a global one involving a breadth of other researchers and policy makers in different nations seeking to find the keys to success in international sport. My special thanks go to the consortium group that was set up to conduct the pilot study of this dissertation. Apart from Paul and Maarten I am indebted to Simon Shibli (UK) and Jerry Bingham (UK).
The endless discussions were a conditio sine qua non for the
completion of this work and I certainly do not regret the sleepless nights that followed. It has been a captivating process to co-operate with very competent and professional researchers abroad and to exchange ideas and information. I also want to thank them for their highly valuable inputs in this dissertation and their constructive comments. And I sincerely hope to collaborate with them in future research projects. My greatest thanks also go to the dedicated researchers who joined this study from the other participating nations: David Legg (Canada); Berit Skirstad and Torkild Veraas (Norway); Alberto Madella and Lorenzo Di Bello (Italy); Bas Rijnen (the Netherlands); Chris Gratton (UK); Luc van de Putte and Thierry Zintz (Wallony). This study was time-
consuming for them and in spite of their limited financial resources they have always searched for the answers to my everlasting questions. The pilot study could not have been carried out without the help of the athletes, coaches and performance directors in different nations who provided information on their personal circumstances and filled in lengthy questionnaires.
Many other policy makers, high
performance co-ordinators, and sports officials also gave freely of their time and provided valuable insights into their policies and programmes. I am most appreciative of all their help. I would also like to acknowledge the contribution of the sport organisations and governments in different nations who provided funding to carry out the research: Olympiatoppen (Norway), CONI (Italy), UKSport (UK), NOC*NSF (the Netherlands), Sport Canada (Canada), the ministers for sport from Flanders and Wallony and the BOIC (Belgium).
We hope their contribution will arouse the interest of other nations and
researchers to be involved in further development of this study. Furthermore there are a range of individuals who deserve my greatest thanks; Sophie Daniëls, Sylvie Leblicq and Bruno Heyndels (Vrije Universiteit Brussel) who have been involved in separate parts of this study; Marc Jegers for his preparedness to read and comment on this manuscript; finally Tracey Shibli for the English corrections and improving the clarity of the text. They have all been of a great help. In expressing words of thanks there is always the danger that some people are forgotten. The people whose names have not been mentioned in this list will definitely know I appreciate their support throughout the years. But my most intense appreciation goes to my family for shepherding this long project from start to finish.
Bruno, my husband,
would laugh at me as if it was obvious that he should take a great deal of the responsibility for the housework so that I could disappear into my study. It was certainly not obvious and I am very grateful for his encouragement and for his practical and moral support. I simply could not have done it without the affection and love of my husband and children. Finishing this work feels like winning a medal. After several years of engagement and practice, time is ripe to contest my game. But it would be unfair to compare a PhD with an athletic career. Maybe there is a parallel, but there is also one substantial distinction. A medal you either win or loose. In a PhD, eventually you always win.
Veerle De Bosscher, February 2007
Table of Contents List of Tables and Figures ...................................................................................... v List of appendices ………………………………………………………………………………………………………………..ix Preface .............................................................................................................. xi I. Introduction ..................................................................................................... 1 Elite sports policies: An international comparative perspective .................................... 1 1.1.
Rationale and problem definition ................................................................ 1
1.2.
Objectives of the research......................................................................... 3
1.3.
Research structure ................................................................................... 4
1.4.
Definitions .............................................................................................. 5
2. Measuring the success of nations in elite sport...................................................... 7 2.1.
Introduction ............................................................................................ 7
2.2.
Evidence of increasing competition?............................................................ 7
2.3.
Measuring absolute national success of nations .......................................... 14
2.4.
Measuring relative national success of nations............................................ 20
2.5.
Comparing countries on equal grounds applied to the Summer Olympic Games in Athens 2004 ...................................................................................... 23
2.6.
Other measurements of Performance: the Olympic Winter Games ................. 33
2.7.
An overall Performance measurement: the World Sporting Index .................. 38
2.8.
Conclusion ............................................................................................ 41
Chapter 3: Theoretical model of factors determining international sporting success...... 43 3.1.
Introduction .......................................................................................... 43
3.2.
Classification of factors leading to international success in top level sports ..... 44
3.3.
Factors leading to international sporting success: the Macro-level ................. 46
3.4.
Factors leading to international sporting success: the Meso-level .................. 54
3.5.
Preliminary research on the factors leading to success according to major stakeholders in elite sport ....................................................................... 58 3.5.1. Flemish athletes’, coaches’ and performance directors’ views on the prerequisites for success ............................................................... 58 3.5.2. International expert tennis coaches’ view on the factors determining international tennis success ........................................................... 70
3.6.
A theoretical model for analysing sports policy factors leading to international sporting success .................................................................................... 79
3.7.
The environment of the sport system as a resource of world class performance in sport ................................................................................................ 95
i
3.8.
Conclusion ............................................................................................ 97
Chapter 4: An empirical application of the theoretical model..................................... 99 4.1.
Introduction .......................................................................................... 99
4.2.
Organisation of the pilot study ............................................................... 101
4.3.
Cross-national comparisons ................................................................... 105 4.3.1
Why compare nations? ………………………………………………………………………… 105
4.3.2
Competitive advantage of nations ……………………………………………………....106
4.3.3
Methodology to measure competitive advantage of nations ……………….107
4.3.4
The complexity of comparing nations: searching for a common "fruitology" …………………………………………………………………………………………….108
4.3.5
Points of interest in cross national research ………………………………………..109
4.3.6 The selection of the sample nations ……………………………………………………..111 4.4.
Measuring success of the sample nations................................................. 114 4.4.1 Market share of the sample nations for Olympic Summer Games ...... 114 4.4.2 Market share of the sample nations for Olympic Winter Games ......... 118 4.4.3 Market share of the sample nations for the UK World Sporting Index for Olympic Winter and Summer sports .............................................. 119 4.4.4 Relative success of the sample nations for Olympic Summer and Winter Games...................................................................................... 120 4.4.5 Summary.................................................................................. 122
4.5
Methodology of data collection and response............................................ 125
4.6
Methodology of data analysis: development of a competitive analysis of elite sports systems .................................................................................... 142
4.7
A comparative analysis of elite sports systems in six nations ...................... 152 Pillar 1: financial support………………………………………………………………………..…153 Pillar 2: Elite sport structures and policies……………………………………………….160 Pillar 3: sport participation………………………………………………………………………..168 Pillar 4: Talent identification and development systems………………………….173 Pillar 5: athletic and post athletic career………………………………………………….182 Pillar 6: training facilities…………………………………………………………………………..194 Pillar 7: coach development and coach provision…………………………………….200 Pillar 8: national and international competition ……………………………………….209 Pillar 9: scientific research………………………………………………………………………..215
4.8
ii
Summary: performance against the nine pillars........................................ 219
Chapter 5: Discussion and conclusions ................................................................ 225 5.1.
Introduction ........................................................................................ 225
5.2.
Theory development on sports policy factors leading to international sporting success .............................................................................................. 226
5.3.
Methodological development in international comparative sport research ..... 236
5.4.
Policy development in elite sport systems ................................................ 236
5.5.
Recommendations for further research.................................................... 247
REFERENCES ................................................................................................... 251 Appendices...................................................................................................... 265
iii
iv
List of Figures Page number Figure 2.1: The number of events divided by the number of NOCs (National Olympic Committees) taking part
8
Figure 2.2: The number of nations winning medals against the number of nations winning no medals in the Olympic Games since 1948
9
Figure 2.3: The total number of medals that could be won in each sport in Athens 2004
13
Figure 2.4: Market share of ten best performing countries in Athens during the Olympic Games between 1988-2004
18
Figure 2.5a: Distribution of the population across the five continents Figure 2.5b: Number of medals in Athens per continent Figure 2.5c: Medals per head of population for each continent
20
Figure 2.6: Boxplot of regression residuals with independent variables: ln(POP), ln(GDP/cap), ln(DENS), Muslim, Protestant and communism and weighted medals as dependent variable, illustrating outliers in Mahalanobis distance
26
Figure 2.7: Histogram and normal probability plot of the regression standardised residuals for weighted number of medals (gold=3, silver=2, bronze=1) in Athens 2004
29
Figure 2.8: Standardized residuals versus standardized predictors. Please note that for the clarity of the figure, only a fraction of the medal winning nations is presented by name
30
Figure 2.9: Graphic representation of a linear regression and the policy as part of the residual
31
Figure 2.10: Market share in Winter Olympics 1992-2006 of the ten best performing nations (in Turin 2006)
34
Figure 3.1: Model showing the relationship between factors determining individual and national success
45
Figure 3.2: Olympic Summer Games, 1896-2004: participation of men versus women
53
Figure 3.3: The relative importance of eight policy dimensions that have most influenced the international success of athletes
67
Figure 3.4: A theoretical model of 9 pillars of sports policy factors influencing international success
87
Figure 4.4.1: The six sample nations 1988-2004 – Market Share for Olympic Summer Games
116
Figure 4.4.2: The six sample nations market share in Winter Olympics 19922006
118
v
Figure 4.5.1. Number of athletes in education and in other employment alongside their sport
137
Figure 4.6.2: An illustration of pillar two regarding the methodology used for development of a scoring system.
145
Figure 4.6.3: An illustration of the points attributed to an open ended (qualitative) question from the overall sport policy questionnaire
147
Figure 4.6.4: An illustration of a dichotomous question for pillar 2: Does your NGB have an Athletes Commission? (according to athletes and taken from the athletes' elite sport climate survey)
148
Figure 4.6.5: An illustration of a net rating question for pillar 2: How do you assess the supply of information from your governing body? (according to athletes and taken from the athletes' elite sport climate survey)
148
Figure 4.6.6: Traffic light scores
149
Figure 4.6.7: An illustration of a traffic light for pillar 2
149
Figure 4.7.4.1: Percentage of athletes, national governing bodies and coaches indicating that sufficient support is offered to talented young athletes
180
Figure 4.7.5.1: Income from sport activities for athletes
190
Figure 4.7.6.1: The proportion of athletes rating quality and availability of training facilities satisfactory/good
199
Figure 4.7.7.1: Number of coaches trained by federation and coaches who practiced sport at an international level as an athlete
204
Figure 4.7.7.2: Total annual income of coaches from coaching activities
207
Figure 4.7.8.1: Sufficiency of the number of international events organised in the
213
own nation (according to athletes, coaches and coordinators)
Figure 4.7.8.2: Sufficiency of the number of international competitions athletes can
214
take part in (according to athletes, coaches, coordinators)
Figure 4.8.1: Radar graph for the UK
223
Figure 4.8.2: Radar graph for Flanders
224
Figure 5.1: Change in expenditure on sport from 1999 – 2003
237
Figure 5.2: Change in market share between Sydney 2000 and Athens 2004
238
vi
Pillar figures Page numbers Figure 1A: Comparative analysis of pillar 1: Financial support Figure 1B: Selected and available criteria for pillar 1.1: Financial support Figure 1C: Selected and available criteria for pillar 1.2: Financial support for National Governing Bodies (NGBs)
153
Figure 2A: Comparative analysis of pillar 2: An integrated approach to policy development Figure 2B: Selected and available criteria for pillar 2 Figure 2C: Selected and available criteria for pillar 2 - Assessment by athletes and coaches
160
Figure 3A: Comparative analysis of pillar 3: Participation in sport Figure 3B: Selected and available criteria for pillar 3
168
Figure 4A: Comparative analysis of pillar 4: Talent identification and development system Figure 4B: Selected and available criteria for pillar 4 Figure 4C: Selected and available criteria for pillar 4 - Assessment by athletes and coaches
173
Figure 5A: Comparative analysis of pillar 5: Athletic and post career support Figure 5B: Selected and available criteria for pillar 5 Figure 5C: Selected and available criteria for pillar 5 - Assessment by athletes and coaches
182
Figure 6A: Comparative analysis of pillar 6: Training facilities Figure 6B: Selected and available criteria for pillar 6 Figure 6C: Selected and available criteria for pillar 6 - Assessment by athletes and coaches
194
Figure 7A: Comparative analysis of pillar 7: Coaching provision and coach development Figure 7B: Selected and available criteria for pillar 7 Figure 7C: Selected and available criteria for pillar 7 - Assessment by coaches
200
Figure 8A: Comparative analysis of pillar 8: International competition. Figure 8B: Selected and available criteria for pillar 8
209
Figure 9A: Comparative analysis of pillar 9: Scientific research Figure 9B: Selected and available criteria for pillar 9
215
vii
List of Tables Page numbers Table 2.1: Absolute success of the 10 best performing countries in Athens 2004, ranked according to weighted medals
15
Table 2.2: Correlations (Pearson) of various indicators of absolute success during the Olympic Games in Athens 2004: total number of medals and gold medals, weighted medals, first 8 places and number of participants (N = 201)
16
Table 2.3: Ten best performing nations according to the points for weighted medals per million inhabitants in Athens 2004
21
Table 2.4: Ten best performing nations according to GDP per head in Athens 2004
22
Table 2.5: Linear regression analysis with all selected independent variables (ENTER); explained variable: weighted number of medals (gold=3, silver=2, bronze=1) in Athens 2004
28
Table 2.6: Stepwise Ordinary Least Squares for weighted number of medals (gold=3, silver=2, bronze=1) in Athens 2004
28
Table 2.7: Ten best performing nations during the Olympic Games in Athens according to OLS method (relative success), with Log (points of medals) as the dependent variable and Log (pop), Log (GDP/head and communism as independent variables
32
Table 2.8: Ten best performing nations during the Olympic Winter Games in Salt Lake City according to OLS method (relative success), with “points of medals” as the dependent variable and “GDP/head and communism” as independent variables
37
Table 2.9. Measurement of performance (market share) using the World Sporting Index for the Olympic Summer and Winter sports in 2004
39
Table 2.10: Ten best performing nations using the UK World Sporting Index (mix of Olympic Summer and Winter sports – different events) according to OLS method (relative success), with “ln (index points)” as the dependent variable and “ln (GDP/head), ln(pop) and communism” as independent variables
40
Table 3.1: Overview of important studies on the factors leading to international success at the macro-level, showing the independent variables and the event. Please note that the independent variables that were found to be significant are written in italics.
46
Table 3.2: Fraction of the inductive content analysis with regard to the five external factors that have had the greatest influence on the success of athletes, according to Flemish athletes, coaches and performance directors from federations.
61
Table 3.3: Overview of important external factors that have had the greatest influence on the success of athletes according to Flemish athletes (N=114), coaches (N=99) and performance directors (N=23) from federations.
62
viii
Table 3.4: Differences among the sample groups (athletes, coaches and performance directors) for each grouped dimension of factors that influence the international success of athletes
66
Table 3.5: Main factors of success according to elite athletes in three nations (Ireland, USA, Flanders**), using an open ended question.
68
Table 3.5: Overview of inductive procedures to cluster quotes from international tennis experts (N=21) on an open ended question regarding important sports policy factors leading to international sporting success according
73
Table 3.6: Rating on a 9-point Likert scale of important factors for international success in tennis according to tennis experts: frequency of most and least important ratings, modus and media
77
Table 3.6: A literature overview of success determinants at the meso-level, (including 2 preliminary studies) clustered in nine policy areas
80
Table 4.3.1: Overview of socio-economic data and achievements during the three editions of the Olympic Summer and Winter Games for the sample nations in this research
112
Table 4.4.1: Sample nations in the Olympic Games Athens 2004 using different measurements of absolute success.
115
Table 4.4.2: Relative rankings of sample nations 1988 – 200’
117
Table 4.4.3: Relative rankings of sample nations 1992-2006
118
Table 4.4.4: Measurement of performance (market share) of the six sample nations using the UK World Sporting Index for the Olympic Summer and Winter sports in 2004
120
Table 4.4.5: Relative success of the SPLISS nations during the Olympic Games in Athens according to OLS method, with Log(points of medals) as the dependent variable and Log (pop), Log(GDP/head and communism as independent variables
121
Table 4.4.6: Relative success of the SPLISS nations during the Olympic Games in Salt Lake City according to OLS method, with Log(points of medals) as the dependent variable and Log (pop), Log(GDP/head and communism as independent variables
121
Table 4.4.7: Relative success of the SPLISS nations during the Olympic Games in Athens according to OLS method, with Log(points of medals) as the dependent variable and Log (pop), Log(GDP/head and communism as independent variables
122
Table 4.4.8: Relative ranking of the sample nations according to different performance measures in Olympic sports: absolute success measurements
123
Table 4.4.9: Relative ranking of the sample nations according to different performance measures in Olympic sports: relative success measurements
123
Table 4.5.1: Athletes' response rate by nation in descending order of number of respondents
132
ix
Table 4.5.2: Athlete profiles in alphabetical order
132
Table 4.5.3: Coaches' response rate by nation in descending order of number of respondents
133
Table 4.5.4: Coaches' profile in alphabetical order
134
Table 4.5.5: Performance Directors' profile
135
Table 4.5.6: The representativeness of the samples in alphabetical order
135
Table 4.5.7: Data summary in alphabetical order
141
Table 4.6.1: Number of Critical Success Factors (CSF) for each pillar
144
Table 4.7.4.1: Characteristics of elite sport and study systems in secondary education in the sample nations
178
Table 4.7.5.1: Number of world top eight and top three athletes in the sample nations
185
Table 4.7.5.2 : Financial support for athletes
187
Table 4.7.5.3 Kind of facilities athletes can make use of (according to athletes)
190
Table 6.7.7.1: Number of qualified coaches in the sample nations.
203
Table 4.8.1: The key to pillar rating charts
219
Table 4.8.2: Evaluation of sports policy factors
220
Table 4.8.3: Assessment by athletes and coaches
221
Table 5.1: Overview of criteria for international comparison of elite sports policies according to the eight important policy pillars for international success
227
x
Appendices APPENDIX 1: Responses by nation and by sport for athletes, coaches and
267
coordinators APPENDIX 2: Pillar 1 APPENDIX 2.1: TOTAL NATIONAL EXPENDITURE ON SPORT (2003)
271
APPENDIX 2.2: TOTAL NATIONAL EXPENDITURE ON ELITE SPORT (2003)
272
APPENDIX 2.3: FINANCIAL SUPPORT FOR NATIONAL GOVERNING BODIES
273
(2003) APPENDIX 3: COUNTRY INFORMATION
274
Background information on sports policy and structures in the participating countries APPENDIX 4: RADAR GRAPHS FOR OTHER SAMPLE NATIONS BASED ON
287
TABLE 4.8.2 APPENDIX 5: list of indicators, standards and scores for each nation in the
291
pilot study
xi
xii
1. Introduction: An international comparative perspective
I. Introduction Elite sports policies: An international comparative perspective
1.1. Rationale and problem definition Over the last century the power struggle between nations to win medals in major international competitions has intensified.
The achievement of international and
especially Olympic sporting success is increasingly important to a growing number of countries.
As a consequence of the continuous escalating standards in international
sport, competition has become a competition, not only between ‘individuals’ but also between ‘systems’ (Heinilä, 1982). At the same time, increased (media) attention and popularity of major sports events have given sports the status of an issue of ‘national importance’.
Medal-counting has been used by politicians and the media to compare
international success despite the International Olympic Committee's protestation that the Olympic medal table is not an order of merit. This has lead to an increasing awareness among governments of the value of elite sporting success.
More often than not elite
sporting success has been seen as a resource valuable for its malleability and its capacity to help achieve a wide range of non-sporting objectives (Green & Houlihan, 2005). As a result governments have become more willing to intervene directly in elite sport development by investing large amounts of money. Several nations, such as the former communist countries have indeed shown that accelerated funding in elite sport can lead to an increase in medals won at the Olympics.
These traditional sport powers have
increasingly sought to maintain their relative advantage by adopting a more strategic approach to elite athlete development and by copying best practices from other competitors. Consequently, in their quest for international success in a globalizing world, the elite sports systems of leading nations have become increasingly homogeneous. More than ever before, they are based around a single model of elite sports development with only slight variations (Clumpner 1994; Krüger 1989; Oakley & Green 2001a, 2001b). This is the fundamental principle of what Oakley and Green (2001b) describe as ‘a global sporting arms race’ that international sporting success can be produced by investing strategically in elite sport. From this power struggle emerged an interest in sport systems and an urge by researchers to explain (mainly) Olympic successes.
Questions have been posed as to
why some nations succeed and others fail in international competition. The answers to
1
1. Introduction: An international comparative perspective
these questions are obviously of central concern to policy makers wishing to understand why nations gain competitive advantage in sport. Making transnational comparisons with the best competitors is very common in the economic sector, but in sport it is rather new.
Yet, just as in economies, ‘competitiveness’ of nations is unusually ill defined
(Porter,
1990).
As
Porter
(1990)
points
out
“instead
of
seeking
to
explain
‘competitiveness’ at the national level, we must understand the determinants of productivity” (1990, p9).
Notably, there are large gaps in sports literature on the
determinants of elite sporting success. Moreover, the influence of elite sport policies on international success is under-researched. From this dearth of literature and a lack of an accepted theory to explain the international success of nations, three research issues emerged that lie at the root of this dissertation, namely theoretical, methodological and political issues. A theoretical issue appeared because of a lack of an empirically grounded, coherent theory on the factors determining international sporting success.
There are various
studies trying to explain differences in the Olympic success of nations through socioeconomic determinants such as wealth, population, land mass, religion and politics. Clearly, none of these studies focuses on the contextual factors that can be fashioned by sport policies and their relationship to success. Moreover, empirical research on the topic is still rather fragmented and often restricted to a descriptive level. From this omission emerged the need for a theoretical model that defines the components of an effective and efficient elite sport policy and how they are related to each other. A methodological issue was perceived because of the nature of international sporting success. Elite sports are international by definition. International comparative research between nations is the only means by which one can detect the variables which explain international success and the strategies nations use to develop medal winning capability. However, making transnational comparisons is a complex process. This is caused by the difficulty of comparing nations on a like for like basis, due to the cultural differences and the uniqueness of each sport system.
It was noted by Houlihan (1997) and Henry,
Amara and Al-Tauqi (2005) that relatively few comparative studies in sport have been conducted.
As a result of the lack of an appropriate methodology to operationalise
success determinants in order to initiate comparative research, the need emerged to create a framework of key indicators to enable international comparison. A third issue addresses policy makers.
The success of an athlete or team depends
increasingly on the performance capacity of the national system and its effectiveness in using all relevant resources for the benefit of elite sport (Green & Houlihan, 2005).
2
1. Introduction: An international comparative perspective
Governmental authorities worldwide spend large sums of money to compete against other countries for superior sport performances. After all, amidst increasing international sporting competition, standing still could mean going backwards if those countries taking a strategic approach develop a competitive advantage over those countries which do not plan for success (SIRC, 2002). But, it is not known precisely how sports policies can influence these achievements. How can nations sustain their competitive position amidst increasing competition? How can the efficiency and effectiveness of elite sport investments be enhanced? Without a clear model of the influence of various factors on elite sport success, it is difficult for sport policy makers to understand the problem and to make rational allocations and long-term planning decisions about their sport delivery system. In order to bridge this identified gap in the research, this dissertation aims to develop and test a theory on sports policy factors leading to international sporting success. Using data collected from literature and preliminary sources and through an international comparison of elite sport policies in six nations this study presents what might be learnt from theoretical, methodological and sports policy perspectives. The theory presented in this study attempts not to capture the full complexity of elite sporting success.
In
developing the theory this study sought to integrate the many elements which influence international success. For the focus of this dissertation, we do not presume to claim a comprehensive understanding of all the complex factors that influence success but the aim was to abstract from this holistic theory and the main focus of this dissertation concerns the determinants that can be influenced by sport policies and are important for the success of nations rather than the individual success of athletes.
1.2. Objectives of the research The objectives of this dissertation are threefold: 1. To improve our theoretical understanding and methodological approach regarding the identification of determinants that are important for international sporting success 2. To construct a theoretical model, tested in an empirical environment, of key sports policy factors and operational indicators as key drivers of elite sport systems 3. To develop a research methodology for the measurement of success and for the evaluation and international comparison of elite sport policies
3
1. Introduction: An international comparative perspective
1.3. Research structure As in any scientific research, this dissertation follows a logically composed structure.
In
chapter two the dependent variable, international success, will be defined. We examine various methods of performance measurement, resulting in the presentation of an elite sport index based on mainly two success indicators: absolute success using 'market share' analysis and relative success using ‘Ordinary Least Square’ methods.
Both
methods are applied to summer and winter Olympic sports and also to other than Olympic events. The purpose of these indices is to evaluate and compare the success of different
nations
objectively
and
to
understand
the
complexities
of
success
measurements. In chapter three we develop a theoretical model of sports policy factors leading to international success.
We present a classification of factors leading to international
sporting success. A comprehensive literature review and preliminary research result in the introduction and inclusion of a range of success indicators under nine key sport policy dimensions, or “pillars”. These pillars are defined and operationalised. This theoretical model is tested in an empirical environment in chapter four. This chapter provides a pilot study in six nations: Belgium (Flanders and Wallony), Canada, Italy, the Netherlands, Norway, and the United Kingdom. The aim of this pilot study is to: (1) test the theoretical model of key sport policy indicators (2) operationalise critical success indicators into measurable criteria (3) achieve insight into the definition of standards for evaluation (4) develop a scoring system for the evaluation of elite sport policies (5) broaden our knowledge on elite sport systems in different nations in order to apply this knowledge to the adjustment of our theoretical model This chapter has led to the production of an 'at a glance' comparative analysis of each nation against each pillar and a measurement system for international comparison. Chapter four is structured as follows. Following an introductory part on the organisation of the research in section 4.1 and 4.2 and on cross-national comparisons and problems accompanied with this kind of research in section 4.3, section 4.4 is concerned with a discussion of the international success of the sample nations. In section 4.5 we present the methodology of data collection and the responses of the sample nations. In part 6 of this chapter the methodology of the scoring system is outlined, by exploring one pillar in
4
1. Introduction: An international comparative perspective
depth. In part 7, the largest section, we present, compare and discuss results collected for the six sample nations on the nine identified pillars, namely: 1. financial support; 2. organisation and structure of (elite) sport policies: an integrated approach to policy development; 3. participation in sport; 4. talent identification and development system; 5. athletic and post career support; 6. training facilities; 7. coaching provision and coach development; 8. international competition; 9. scientific research. In the last chapter we round off with methodological, theoretical and political conclusions. We will end up with an overview of sport policy criteria in a model that can be used for further research. From a political viewpoint we capture the main characteristics of the competitive analysis of the six sample nations and seek to explore the relationship between our assessment of the relative success of the sample nations and our analysis of their respective sports policy frameworks. The study concludes with a series of strategic, operational and methodological recommendations designed to inform current practice and future research.
1.4. Definitions Prior to starting the main body of the text it is worth clarifying a few key terms that are frequently used in this dissertation. MACRO-level factors influencing international success Factors that determine international success of a nation but that cannot be influenced by sport policies. It concerns the social and cultural context in which people live: economic welfare, population, geographic and climatic variation, degree of urbanisation, political system, and cultural system. MESO-level factors influencing international success Factors where well-considered sports policies may influence the long-term performance of a nation.
5
1. Introduction: An international comparative perspective
MICRO-level factors influencing international success Factors that may influence the international success of an individual athlete. It concerns intrinsic factors (such as genetic characteristics of athletes) and the close environment of the athlete (such as parents, friends, coaches, sport club). At the micro-level some factors can be controlled (such as training techniques or tactics) and others cannot be controlled (such as genetics). Absolute success: international success of a nation that can be measured as: -
total number of medals
-
total number of gold medals
-
medal points: weighted number of medals (e.g. 3 for gold, 2 for silver, 1 for bronze); or first six places (allocating for example 6-5-4-3-2-1 points) or first eight places (allocating for example 10-8-6-5-4-3-2-1 points)
-
Market share: medal points won as a proportion of points available to win during international events (SIRC, 2002)
Relative success
Measuring absolute success controlling for exogenous macro-influences. Efficiency The cost/benefit ratio incurred in the pursuit of those goals (Steers & Black, 1994, cited by Chelladurai, 2001) Effectiveness The extent to which operative goals can be attained (Steers & Black, 1994, cited by Chelladurai, 2001) Elite sports climate
The social and organisational environment that provides the circumstances in which athletes can develop into elite sports athletes and can continue to achieve at the highest levels in their branch of sport (Van Bottenburg, 2000). Elite athlete
In this survey an elite athlete is defined as an athlete who, as an individual or as part of a team, has participated in an elite sports discipline in a European Championship, World Championship, Olympic Games or other competitions that are comparable to these championships or games in the last twelve months.
6
2. Measuring the success of nations in elite sport
2. Measuring the success of nations in elite sport
2.1. Introduction The definition of sporting success is a theme which has been widely discussed in both the literature and media with respect to national comparisons. Success is typically expressed in absolute terms, such as the total number of medals that a country wins during the Olympic Games or other championships. Such comparisons of countries, however, give a biased view of how successful countries actually are in using their available resources. These accounts do not take into consideration a number of socio-economic and political variables that play an important part in determining each country’s success. Since these variables cannot be influenced by (elite sport) policy they should thus be taken into consideration as exogenous parameters when examining the efficacy of sport policy. Top level sports continue to reflect inequalities at the world level very closely. The Olympic Games are still dominated by a small number of capitalist core and (formerly) socialist countries (Green & Houlihan, 2005; Stamm & Lamprecht, 2001).
One should hardly be
surprised that large and rich countries like the United States tend to win more medals than Zimbabwe or Ecuador. Yet there are exceptions, such as Cuba, Ethiopia and Kenya. The question is how external factors should be taken into account when defining success. The aim of this chapter is to examine various methods by which the outputs of an athlete production system can be measured. The definition of absolute national success using the medal tables from the 2004 Athens Games will be taken as our point of departure. We shall then proceed to examine relative success, or rather success controlling for
economic, sociological and political determinants.
Furthermore we also examine the
limitations of the analysis and propose some alternative measures using other events. However, before examining performance measurement methods in more detail and in order to estimate the competitive position of nations we will first focus on the market of supply and demand in elite sport. Because defining success of nations clearly depends on the evolution of this market we first present how competition has increased and try to provide an explanation for this phenomenon.
2.2. Evidence of increasing competition? Along with the FIFA Football World Cup, the Summer Olympic Games is the most high profile event in the global sporting calendar. In recent Games, the International Olympic
7
2. Measuring the success of nations in elite sport
Committee (IOC) has sought to make the event truly global. This point can in part be appreciated by looking at the number of nations taking part in the summer Olympic Games: an increase from 14 in 1896, over 59 in 1948 to 201 in 2004. The number of athletes competing in the Games started with 241 in Athens 1896 and has more than doubled in the period 1980 (5,217 competitors) to 2004 (11,099 competitors).
This
makes one assume that competition has increased. On the other hand the number of events contested at the Games also increased: from 43 events in 1896 to 301 events in 2004, mainly because an increased number of sports (from 9 sports in 1896 to 28 in 2004) and the increased number of events contested by women.
The figure below
presents the number of events divided by the number of nations taking part in the Olympic Games. Figure 2.1: The number of events divided by the number of NOCs (National Olympic Committees) taking part
9,00 8,00 7,00 6,00 5,00 4,00 3,00 2,00 1,00 0,00 96 00 04 08 12 20 24 28 32 36 48 52 56 60 64 68 72 76 80 84 88 92 96 00 04 18 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 20 20 number of events divided by the number of NOCs taking part
The decreasing line until 1972 in Figure 2.1 shows that since the start of the Games, the number of nations taking part has grown more than has the number of events. suggests that, until 1972, the competition for medals had been increasing.
This
After the
boycotted Games in 1976 (by 26 African countries), in 1980 (62 countries, led by US) and in 1984 (14 countries, led by USSR), a stable trend can be noticed. This indicates that if, during the last 20 years, all nations would have equal grounds to win medals, the number of medals that can be won by each nation has equalised.
8
However, more in
2. Measuring the success of nations in elite sport
depth analysis of the nations who won medals was done by Shibli (2003). The author showed that competition also increased during the last two decades. More nations are indeed winning medals, but the number of nations winning no medals has increased even more, as is illustrated in Figure 2.2. Figure 2.2: The number of nations winning medals against the number of nations winning no medals in the Olympic Games since 1948.
250
200
150
100
50
0 1948
1952
1956
1960
1964
1968
1972
1976
1980
1984
1988
1992
1996
2000
2004
number of nations winning medals in the Olympic Games number of nations winning no medal in the Olympic Games (source data: Shibli, 2003)
In the London Games in 1948, 37 (63%) nations won at least one medal and 22 did not. In Athens 75 nations (only 37%) won at least one medal and 124 did not. These figures suggest that proportionally mainly the same nations are winning more medals instead of more nations winning at least one medal.
Interestingly, however, since 1984 the
increase in the number of non-winning countries is smaller than the increase in the number of winning countries.
As a result, the gap between the medal-winning countries
and those without medals has decreased.
On the other hand, according to Shibli and
Bingham (2005) further analysis of the Olympic medal table reveals that (1) more nations are becoming successful and (2) the most successful nations are maintaining their medal table rankings despite winning proportionately fewer of the medals available. In 1988 the top ten nations won 74.5% of the medal points1.
This fell to 56.9% in
Athens 2004. According to the authors the reason for this fall is attributable to the share
1
medal points: 3 points for gold, 2 for silver, 1 for bronze
9
2. Measuring the success of nations in elite sport
of success being distributed more widely, that is, more nations are developing medal winning capability.
If more nations have developed medal winning capability, then it
follows that medals must be even harder to win and competition to win them must therefore have increased. The view that the Olympic Games have become increasingly competitive is an important contextual point to be analysed further. There are a number of factors that are attributed to the increasing number of NOCs that have developed medal winning capability. According to Shibli and Bingham (2005) these can be divided into two distinct categories: first, factors attributable to an increase in demand for success in elite sport; and, second, factors attributable to manipulating the supply of success.
The demand side factors are in part covered by Oakley and Green
(2001b) who describe the increased demand for sporting success as a 'global sporting arms race'.
The term 'global sporting arms race' is an appropriate description of
international sport at elite level because it conveys an image of a battle for sporting supremacy with no absolute goal, only the relative goal of staying ahead of the competition. Shibli and Bingham (2005) identified three demand side factors. First, new countries, for example those arising from the break up of the former Union of Soviet Socialist Republics (USSR), have used sporting success as a means of establishing their national identity. At the 2004 Athens Olympics 10 of the 15 NOC’s that previously comprised the Soviet Union won at least one medal.
On a smaller scale the former
Yugoslavia (one NOC) is now made up of five sovereign states which all have a recognised NOC. In the four summer Olympic Games from 1992 to 2004 a minimum of two and a maximum of three nations that were formerly part of Yugoslavia won at least one medal. Of the 30 new nations that have been recognised since 19902, 22 have won at least one medal in the four summer Olympic Games held since 1992. Second, as a direct response to increased competition, some nations (notably those with a tradition of international sporting success) have seen their share of international sporting success reduced. In some nations this has led to massive investment in elite sport, often as a means of overcoming perceived failures.
Chalip (1995) states that
changes in policy follow what he calls “focusing events” which are “nationally traumatic events that can symbolise an issue and focus policy makers' attention on proposals to redress the issue” (1995, p.5). Some ‘focusing events’ in national sporting performance have led to massive investment in elite sport often as a means of overcoming perceived failures.
2
Oakley and Green (2001a) illustrated this phenomenon of increased policy
http://geography.about.com/cs/countries/a/newcountries.htm
10
2. Measuring the success of nations in elite sport
interest after weak performances at significant events in the United States (after 1972 Olympics), Canada (after 1956 and hosting 1976 and 1988), France (after 1960), UK (after 1996) and Spain (hosting Olympics 1992). Oakley and Green (2001b) point out that one of the factors that contribute to an understanding of this, is the length of time different nations have been committed to fully integrated elite sport development. Thus, in addition to more nations increasing the competition of elite sport success, the response of nations seeking to preserve or enhance their status has led to a further escalation of the ‘global sporting arms race’. A third argument according to Shibli and Bingham (2005) to show the increased number of nations with medal winning capability is that politicians have realised and sought to capitalise on the potential of sporting success via the media platform that elite sport provides internally and externally.
Internally, success in sport can help to develop
national pride and to create a national 'feel good' factor. By contrast, failure can have the opposite effect. Thus politicians may invest in elite sport to achieve the former and to avoid the latter.
Externally, there is a view that international sporting success can
enhance a nation's reputation abroad which in turn may lead to improved political and trading opportunities.
An increased number of nations have realised and sought to
benefit from the internal and external opportunities of international sporting success, thereby contributing further to the 'global sporting arms race'. From the supply side there is evidence that qualification rules, the selection of sports and rule changes, have had the effect of limiting the opportunities for some nations to dominate the Olympic Games and simultaneously increased the opportunity for other (arguably less dominant) nations to win medals. In addition to its desire to create 'a real universality', the IOC is also actively engaged with trying to 'bridge the North/South gap' (Rogge 2002). The term 'North/South gap' is a reference to the dominance of Northern Hemisphere nations in both the summer and winter Olympic Games. According to Shibli and Bingham (2005) it can be argued that the IOC has manipulated the supply of success in order to achieve its broader aims. To substantiate this claim some examples are discussed below. First, qualification rules ration the number of athletes from any NOC taking part in a particular event. For example in the case of athletics the qualification rules state (Shibli & Bingham, 2005): "Individual events - An NOC may enter a maximum of 3 qualified athletes in each individual event if all the entered athletes meet the A qualification standard for
11
2. Measuring the success of nations in elite sport
the respective event, or 1 athlete per event if they have met the B standard only."3 Hence, nations can not send an unlimited number of athletes to the Games.
Renson
(2004) notes that “if the United States and China could send as many participants in relation to their population as Naru did in 2004 (3 participants from a population of 12.809 = 1 participant per 4.270 inhabitants), then 68.618 American athletes would have ended up in Athens and 303.981 Chinese. Indeed, American athletes have a saying that it is easier to win an Olympic medal, than it is to get on the US team” (2004, p11). The net effect of rationing the number of athletes per NOC who can take part in an event reduces in part the effect of population. Second, the IOC also ration the number of countries from any given continent taking part in certain events by using a continental quota system whereby a specified number of athletes from a given continent are allowed to qualify for the Olympic Games. Third, in addition to the continental quotas, rule changes in certain sports have created conditions whereby a greater number of nations have been able to win medals (Shibli & Bingham, 2005). A notable example is judo which was introduced as an Olympic sport in Tokyo, 1964, where 12 medals were awarded.
In 2004 there were 14 Judo events and 56
medals awarded. An identical situation takes place in boxing and taekwondo. A further IOC rule in these sports is that each NOC is permitted to enter only one athlete per weight category.
For example when there are four medals awarded in each weight
category, then it follows that four different NOCs will win a medal and therefore the supply of success is distributed more widely. A fourth argument concerns the growth in the number of different sports contested in the summer Olympic Games that can be attributed to attempts to achieve an improved ‘cultural balance’ in the sports contested. For example, the re-introduction of archery in 1972 has enabled nations such as Korea, China, Indonesia and Taipei to win medals. Similarly, the introduction of table tennis in 1988 and badminton in 1992 has made the Olympic Games relevant to huge viewing audiences (or markets) in the Far East (Shibli & Bingham, 2005). These examples are further evidence of a policy designed to distribute success widely rather than to permit dominance by a few nations. On the other hand, Oakley and Green (2001b) state that there is a distinct 'western' bias in the portfolio of sports contested at the Olympic Games. Indeed, new trendy sports that have been introduced during the last three Olympic editions, like beach volleyball, curling, trampoline and softball are sports that are more practiced in Western (or read:
3
http://www.athens2004.com/en/AthleticsQualifications
12
2. Measuring the success of nations in elite sport
wealthier) nations.
This ignores the sporting traditions of some nations and therefore
puts such nations at a disadvantage. The following figure illustrates more evidence of sporting traditions as a determinant of the total number of medals that nations are able to win. This figure presents the total number of medals that could be won by sport in Athens 2004. Figure 2.3: The total number of medals that could be won in each sport in Athens 2004
140
120
100
80
60
40
20
men
women
softball
triathlon
Baseball
hockey
moden penthatlon
football
handball
Basketball
tennis
volleyball
table tennis
archery
Badminton
equestrian
sailing
taekwondo
boxing
rowing
weighlifting
fencing
shooting
canoe/Kayak
wrestling
gymnastics
judo
cycling
athletics
aquatics
0
mixed
It is clear from Figure 2.3 that the total number of medals to be won in each sport is unequally distributed. 29% of all medals, are contested in only two sports, athletics and aquatics.
Consequently, it may be much more efficient for nations to invest in a few
good swimmers than in a wide range of sports where only two or three events are contested.
For example 8 of the 22 medals won by the Netherlands in Athens 2004,
were won by just two athletes: Inge de Bruijn and Pieter van den Hoogenband. Australia won 15 medals in swimming and the USA 28.
Nations with strong traditions in team
sports, such as Denmark or Brazil are disadvantaged even more. These nations not only have fewer events where they can gain medals, they also need more players to create a team for just one medal. In addition to the factors relating to the demand for and supply of success in the Olympic Games there is another factor which has led to an increase in the number of nations developing medal winning capability, namely the increased ease of nationality change
13
2. Measuring the success of nations in elite sport
(Shibli & Bingham, 2005). Recently there has been a 'talent drain' from Kenya via the change of nationalities of Stephen Cherono (Qatar), Wilson Kipketer (Denmark) and Bernard Lagatt (USA) (Bale & Sang, 1996).
Nationality change is likely to remain an
influential factor in determining the number of NOCs with medal winning capability. The maximum qualification period for an athlete changing nationality is three years and this period can be reduced to one year if both the current and prospective NOCs support such a change. For some nations, particularly those wishing to derive the potential benefits of international sporting success discussed above, 'purchasing' talent via nationality change may well be a more cost effective and time efficient approach than setting up their own talent confirmation and development programmes. From all the arguments given above, it is reasonable to conclude that the summer Olympic Games have become increasingly competitive and thus medals have become relatively harder to win. As the supply of medals (success) remains essentially fixed (the IOC has indicated that it would like the number of events to be capped at around 300), and the demand for success is increasing (more nations taking part and more nations winning medals), the 'market' adjusts by raising the 'price of success' (Shibli, 2003). In practice, an increase in the 'price of success' means that even more resources need to be invested in order for a nation to retain its medal winning capability. This point has two important implications for assessing performance. First, if the view that there is a 'global sporting arms race' in elite sport is accepted, then it follows that if a nation is able to maintain its level of performance in an environment of increased competition, then the nation concerned must have improved simply to keep pace with the competition. Second, if at any point a nation disengages from the 'global sporting arms race', then all things being equal, it will expect to see a deterioration in performance as it loses ground to competitor nations that have continued to invest in the development of medal winning elites. We now proceed to explain how performance can be measured. The chapter will present different methods to measure ‘absolute’ and ‘relative’ success with the Athens Olympic Games as a point of departure.
Finally this chapter ends with an application of the
methodology on Winter Games and the UK World Sporting index.
2.3. Measuring absolute success of nations Although the IOC does not recognise the Olympic medal table as an order of merit, it is widely accepted outside of the IOC that the final medal table for each games is an order of merit. This finding is perhaps best demonstrated by the fact that many nations invest
14
2. Measuring the success of nations in elite sport
heavily in sport precisely to climb the medal ranking.
Several authors used different
methods of measuring absolute success, i.e. without taking into account socio-economic variables. Some methods will be further explored in the next part. In its most simple form, the total number of medals that a country wins serves as the standard measure for absolute success (see for example Hoffmann, Ging & Ramasamy, 2002a; Gärtner, 1989; Grimes, Kelly & Rubin, 1974; Kiviaho & Mäkelä, 1978; Levine, 1974; Novikov & Maximenko, 1972). The argument for using absolute scores is based on the fact that modern competitive sport at an elite level is absolute in nature. Victories and point scores are what counts (Kiviaho & Mäkelä, 1978). A key weakness with this absolute medal table is that it does not account for the ‘quality’ of medals won (more gold, more silver and less bronze) nor other dimensions of ‘quality’ (for example the value of a medal in athletics may be different from a medal in archery). A method which allows for the relative values of gold, silver and bronze medals is a ‘points’ system which makes use of a weighting system to convert a nation’s medal haul into a points equivalent. This methodology is standard practice when analysing Olympic results (see for example Ball, 1972; De Koning & Olieman, 1996; Den Butter & Van der Tak, 1995; Van Bottenburg, 2000).
These are usually weighted as follows: a gold medal gets 3
‘points’ (or 4), silver receives 2 and a bronze just 1. Applying this point system to the Athens 2004 medal table for the top 10 nations means that the United States was the most successful country, followed by Russia and China (see Table 2.1).
Table 2.1: Absolute success of the 10 best performing countries in Athens 2004, ranked according to weighted medals
Rank 1
Country United States
Points
Total
3-2-1
Medals
212
103
2
Russia
173
92
3
China
144
63
4
Australia
99
49
5
Germany
92
48
6
Japan
78
37
7
France
64
33
8
Italy
63
32
9
Korea (South)
60
30
Great Britain
57
30
10
Source: IOC, 2004
15
2. Measuring the success of nations in elite sport
Table 2.1 shows that the ranking positions for the top ten nations is the same for total medals or weighted points.
Exceptions can be found if one looks exclusively at the
number of gold medals that have been won. For example, ranked on the basis of gold medals, China with 32 gold medals (total medals 63) would be ranked ahead of Russia which won 27 gold medals (total medals 92) (Shibli & Bingham, 2005).
Whilst medal
based measures of performance are easily understood measures of success, they still ignore the totality of achievement of an elite sport programme.
As has already been
demonstrated competition for medals is increasing as more nations take part in the Olympic Games.
It is quite possible for Performance Directors in individual sports to
make considerable progress in developing a sport without this progress being translated by medals in elite competition.
Therefore, a number of studies examine the first six
(Shaw & Pooley, 1976; Gillis, 1980) or eight places (Condon, Golden & Wasil, 1999; Kiviaho & Mäkelä, 1978; Stamm & Lamprecht, 2001), using a weighting or 6-5-4-3-2-1 or 10-8-6-5-4-3-2-1 respectively.
Kuper and Sterken (2003 a & b) also found the
number of participants in the Games per country as the indicator for determining success. When these methods are compared, it may be concluded that – as a general rule – it does not make much of a difference if and how the number of medals are weighted or not, nor whether the first three or the first eight places are taken into account (see Table 2.2).
Table 2.2: Correlations (Pearson) of various indicators of absolute success during the Olympic Games in Athens 2004: total number of medals and gold medals, weighted medals, first 8 places and number of participants (N = 201) Med total Total medals
1,000
Gold medals
Medal, weighted First 8, weighted
Med gold
Med weighted First 8 (10–8–6–5–4-3-2-1)
0,971**
0,997**
0,985**
0,886**
1,000
0,984**
0,964**
0,849**
**
0,984
0,882**
1,000
0,901**
1,000
Total participants **
Participants
(3–2–1)
1,000
correlation is significant at 0,01 level
In Table 2.2, high (Pearson) correlations can be observed for all methods. This means that, in general, a country that gets a high score using one method will also have a high score using another. Even the number of participants in the Games from each country is significantly correlated to the number of medals won (r>0,8).
In other words, the
selection for achieving success has already taken place, that is before participation in the
16
2. Measuring the success of nations in elite sport
Games with norms that are, on the one hand, set by the IOC and, on the other, by each National Olympic Committee (NOC). Nevertheless, these correlations are only statistical methods that never give a correlation of 100%, which means that there will always be exceptions to this generalisation of methods, like in the example of China and Russia given above.
Great Britain is more
successful than Italy in terms of top eight places whereas Italy won two more medals (of which one was gold). Using one Olympic Games may result in coincidental rankings.
For example the
Netherlands won only 4 gold medals (22 in total) in Athens but they won 12 gold (25 in total) in Sydney 2000. To overcome the problem of success by coincidence, countries must be evaluated over time.
Whilst the points system is a more useful measure of
performance than position in the medal table or total medals won, it has one major limitation. As the number of events contested at each Games has varied considerably over time and to a lesser extent the number of points per event has also varied (for example two or more nations 'tying' for the same medal), the number of points available at each Olympic Games has also varied.
In order to convert points won into a
standardised measure, SIRC (2002) offers the principle of computing ‘market share’, that is, points won as a proportion of points available to win.
Using market share it is
possible to make a more accurate time series diagnosis in standardised terms. Figure 2.4 illustrates the temporal evolution in the share of medals for a selection of the ten best performing countries (in medal points in Athens 2004) over the period 1988-2004. There is a logic to starting with Seoul in 1988 as this was the first games since 1972 that had not been contaminated by some form of boycott.
The market share shows for
example that in the 301 events contested in Athens, a total of 1,832 medal points was awarded, of which the United States won 212. This is a market share of 11.6%, which in turn is slightly more than the 203 points won out of 1,829 (300 events) in Sydney 2000, a market share of 11.1%.
17
2. Measuring the success of nations in elite sport
Figure 2.4: Market share of ten best performing countries in Athens during the Olympic Games between 1988-2004 18,0% 16.3%
16,0% 15.6% 15.1%
14,0%
13.6%
12,0%
13.4%
11.2%
11.6% 10.4%
11.1%
10,0%
10.0% 9.4% 8.2%
8,0%
7.9% 6.8%
6.4%
7.5% 5.8%
6,0%
5.8% 5.4% 5.0%
5.6%
4,0%
4.1%
3.6% 3.0%
2,0%
1.5%
0,0% 1988
1992 United States
1996 Russia (USSR)
2000 China
2004
Germany
Australia
5,0% 4.5%
4,5%
4.3% 4.3% 4,0%
3.8% 3.7%
3.7% 3,5%
3.4%
3.3%
3.1% 3,0%
4.3%
3.0%
3.0%
3.5% 3.4% 3.3% 3.1%
2,5% 2.3% 2,0%
2.1%
2.0%
2.1% 1.8%
1,5%
1.6% 1.5%
1.4%
1,0%
0,5%
0,0% 1988
1992 Japan
18
1996 Korea (south)
Great-Britain
2000 France
2004 Italy
2. Measuring the success of nations in elite sport
In terms of market share, Figure 2.4 reveals that only four of the top ten nations (in Athens) have increased their performances from Sydney to Athens: USA (increase of 0.5%), China (a remarkable increase of 2.1%), South Korea (0.3%) and Japan (a remarkable increase of 2.3%).
Over a longer period a steadily increasing trend was
noticed in France and Italy until 1996 and in Australia until 2000 with slightly decreasing performances thereafter.
A decreasing trend can be observed for Great Britain until
1996 and Korea until 2000; both nations increased their performances thereafter. Although the UK won fewer medals in Sydney (28 medals) than they did in Athens (30 medals), their market share in Sydney was higher (3.28%). After the unification of East and West Germany in 1989 and the break up of the former Union of Soviet Republics (USSR), the United States is constantly outperforming other nations with a market share above 11%. Another point that can be observed from Figure 2.4 is the narrowing gap in market share between the ten best performing nations in Athens 2004 compared to 1988. This finding may again be indicating that the gap between the top ten nations is decreasing due to increasing competition.
The market share relates the success of the individual country to the total ‘amount’ of success (e.g. the total number of medals). In terms of absolute success, market share is the only standardised measure of the performance indicators shown in this chapter thus far. The reason for this is that market share can be increased in one of three ways, or a combination of all (Shibli & Bingham, 2005): •
an increase in the number of medals won (when the number of events has
•
an increase in the quality of medals won (when the number of events has
•
maintaining absolute performance when the number of events has declined.
remained the same); or remained the same); or
A key question arising from the data presented in Figure 2.4 and Table 2.1, is how might nations be expected to perform given the resources at their disposal? In the previous methods, success continues to be defined in ‘absolute terms’; external (socio-economic) influences are not taken into account. The medal success (output) is not presented in relation to its determinants (input). Bernard and Busse (2004) raise the question of how many medals does a country need in order to be classified as successful. In other words, what can a country expect in terms of medal outputs, given its inputs (resources). is the definition of relative success used in this chapter.
This
We contend that when
establishing indicators for national sporting success, one must take the socio-economic and political differences between countries into account.
19
2. Measuring the success of nations in elite sport
2.4. Measuring relative success of nations It is clear that not all countries in the world start on an equal footing to win medals. The ten best performing countries in Athens 2004, accounted for 56% of all the medals; 21% of these were won by Russia and the United States combined. The unequal distribution of medals is illustrated in Figures 2.5a-c, where the medals are divided across the various continents.
Figure 2.5a: Distribution of the population across the five continents Oceania 1%
Europe 12%
Africa 13%
Asia 60% America 14%
Figure 2.5b: Number of medals in Athens per continent* Oceania 6%
Africa 4% America 19%
Europe 52% Asia 19%
*Source data: IOC, 2004
Figure 2.5c: Medals per head of population for each continent Africa 2%
America 8%
Asia 2%
Europe 24% Oceania 64%
20
2. Measuring the success of nations in elite sport
European countries won more than half of all medals, while just 12% of the world’s population actually lives in Europe.
In contrast, Asia, which is home to 60% of the
world’s population, only won 19% of the medals (half of which were won by China and Korea).
If we take the number of medals per head of the population, then Oceania and
Europe, the two smallest continents, are the most successful. It is implicit from these figures that there are other variables playing a role in the determination of success. Numerous empirical studies show that population and wealth are the most important socio-economic determinants of success (see, for example, Bernard & Busse, 2004; De Bosscher, De Knop & Heyndels, 2003 a & b; Jokl, 1964; Johnson & Ali, 2002; Kiviaho & Mäkelä, 1978; Levine, 1974; Morton, 2002; Novikov & Maximenko, 1972; Suen, 1992; Van Bottenburg, 2000). These two variables frequently explain over 50% of total medals or medal points. More details on these studies and explanations of the variables will be given in chapter 3. For this reason, success has also sometimes been expressed in terms of medals per head of population or in terms of per capita GDP.
A breakdown of
performance by population and wealth is shown in Tables 2.3 and 2.4. It is obvious that the size of a country’s population will be a determining factor for sporting success. The bigger the population, the larger the pool from which talent may be recruited and the greater the opportunities to organise training and competitions. The ten best performing nations in Athens, when medals are divided by population, are presented in Table 2.3. Table 2.3: Ten best performing nations according to the points for weighted medals per million inhabitants in Athens 2004 Rank
Country
Population X 1 million
Medal points
Medal points per million inhabitants
1 Bahamas
0,300
4
2 Australia
19,913
99
13,35
4,97
3 Cuba
11,309
52
4,60
4 Hungary
10,032
39
3,89 3,76
5 New Zealand
3,994
15
6 Jamaica
2,713
10
3,69
7 Norway
4,575
16
3,50
8 Latvia
2,306
8
3,47
9 Greece
10,648
34
3,19
10 Estonia
1,342
4
2,98
[Source: http://www.cia.gov/cia/publications/factbook/ 2004]
21
2. Measuring the success of nations in elite sport
This table shows that, in relative terms, the Bahamas was the most successful country in Athens. While it won only two medals (one gold and one bronze), this can be considered a big success taking into account that the Bahamas has a mere 300.000 inhabitants. Both medals were won in athletics (by Debbie Ferguson (200m) and Tonique WilliamsDarling (400m)). Of course, it should be noted that this analysis may be biased by the increased ease of nationality change as was mentioned earlier and/or by the fact that athletes train in other nations; both are likely to remain an influential factor in determining the number of NOC’s with medal winning capability. The figures throughout this chapter assume that athletes developed in and performed for their home nation. In this table Australia takes the second place, and Cuba third. In such calculations, highly populated countries, such as the United States (40th) and China (70th) have little chance of achieving a top of the world ranking. Nations can not send athletes to the Games in proportion to their population size. The good performance of Europe and the poor position of African and Asian countries in Figures 2.5a-c, suggest that international success is partly determined by a nation’s affluence.
Richer countries can invest more in sport and elite sport, individuals may
participate in a broader number of sports and a higher living standard may improve their general fitness and ability to perform at top level. Den Butter and Van der Tak (1995) have found that the number of medals won correlates strongly with income (GDP) as well as with more general welfare indicators, such as the human development index or the quality of life index.
We shall thus look at the success in Athens per Gross Domestic
Product per head of the population for the best 10 performing countries according to this method (see Table 2.4). Table 2.4: Ten best performing nations according to GDP per head in Athens 2004 Rank
Nation
1
China
2
Ethiopia
3 4
GDP/ head €
Medal points
Medal points per GDP/head (x 1000)
4,130
144
578
14
28,80 20,00
Russia
7,351
173
19,44
Cuba
2,395
52
17,93
5
Kenya
826
13
13,00
6
Ukraine
4,460
46
8,52
7
North Korea
1,074
9
6,92
8
Uzbekistan
1,404
10
5,88
9
Romania
5,782
40
5,71
31,223
212
5,61
10
United States
[Source: http://www.cia.gov/cia/publications/factbook/ 2004 - currency modified from $US to € using a
currency conversion of 1$ US = 0.826 €]
22
2. Measuring the success of nations in elite sport
It is immediately apparent from the table that China, which completely lags behind in terms of success per inhabitant, now does very well given the wealth of the country (GDP/head), as does Russia. The position of Ethiopia and Kenya, which both won a total of 7 medals, is also unsurprising.
Per head of the population, their wealth is around
13,000 times smaller than that of the United States. Nevertheless, the US remains in the top ten.
There is only one country that has a top ranking irrespective of whether
success is expressed in per capita terms (Table 2.3) or relative to the countries’ wealth: Cuba.
2.5. Comparing countries on equal grounds applied to the Summer Olympic Games in Athens 2004 The techniques described above clearly show that it is possible to assess the sporting success of a country whilst taking the specific context of that country into consideration. The aim of the present chapter is to present a method whereby some of the differences between countries on the macro-level, which are of major significance in international success and cannot be influenced by policies, are controlled. This allows us to construct indicators for relative success, i.e. success-indicators that control for exogenous macroinfluences.
This approach looks at the efficiency of sports systems when the national
characteristics of Olympic success are isolated. Taking into account just one determinant, population size or wealth, as was done in section 2.4, is rudimentary in two respects. important determinants.
Firstly, it disregards other potentially
Secondly, it assumes an implicit linear relationship between
these two factors and success. By dividing medal points by population or wealth (criteria that are sometimes used by the media) the degree to which these factors can influence success is not taken into account.
This creates a potentially biased view.
After all, a
country that has twice as many inhabitants cannot win twice as many Olympic medals. This is the principle of decreasing returns to scale (Glejser 2002).
This derives from,
among others, institutional characteristics mentioned above such as the fact that countries are allowed to send to the Games only a limited number of athletes who have met the IOC criteria.
On the other hand, the competitive disadvantage for small
countries with respect to team sports is greater, since it is more difficult to bring together more champions to form a team.
In the first instance, therefore, large countries are
disadvantaged and, in the second, they have an advantage.
23
2. Measuring the success of nations in elite sport
To assess whether a particular country or group of countries does ‘well’ at the Olympic Games, the literature offers a number of methods.
These take both of the
aforementioned criticisms into account: (a) several determinants of success at the same time and (b) possible non-linear effects.
Several studies explored the relationship
between international sporting success and the (macro) economic, sociological and political context within which sporting talent thrives.
Most studies used simple
correlations and regression analysis. During the last decade, some authors have tried to improve the methodology of these studies (see for example Baimbridge, 1998; Bernard & Busse, 2004; De Bosscher, De Knop & Heyndels, 2003 a & b; De Koning & Olieman, 1996; Den Butter & Van der Tak, 1995; Johnson & Ali, 2002; Tcha & Perchin, 2003). Chapter three explores these studies in depth and provides a detailed overview of the independent variables that were used and ‘why’ they influence success. The theoretical assumption is that talent is equally distributed throughout the world and that each country has a priori just as great a chance of producing good athletes (Grimes, Kelly & Rubin, 1974; Levine, 1974; Kiviaho & Mäkelä, 1978).
Environmental and sport policy
factors thus influence the identification and development of talent.
These studies
revealed that aside from the ‘key’ environmental factors for success (wealth and population), which have already been addressed above, the main explanatory factors for success are: area, degree of urbanisation, religion and political system. These variables are the inputs in the production of sporting success that cannot be controlled by sports policies. The starting point for our empirical work is a simple OLS (Ordinary Least Squares) estimation of a reduced form model that captures the main macro-determinants of absolute Olympic success.
The ‘outputs’ are the weighted number of Olympic medals
that a country has won.
A. Regression With the Athens 2004 Olympic Games as a point of departure a linear regression analysis (OLS) of socio-economic factors that influence (absolute) sporting success was carried out using the Statistical Package for the Social Sciences (SPSS). Here the logarithm of the weighted medals (gold=3, silver=2, bronze=1) for the Olympic Games in Athens in 2004 are taken as the absolute success measurement.
24
2. Measuring the success of nations in elite sport
Taking into account the aforementioned factors, the functional form to be estimated is as follows: Ln MED (weighted points) = β0 + β1 Ln (POP) + β2 Ln (GDPCAP) + β3 Ln (DENS) + β4 MUSL + β5 PROT + β6 COMM + ε
In the above4, Ln (POP) is the (logarithm of the) number of inhabitants, that were recorded in millions. Ln (GDPCAP) is the (logarithm of the) Gross Domestic Product per head (recorded in US dollars). Ln (DENS) corresponds with (logarithm of the) population density (population/ area). Given that we control for the number of inhabitants, this variable takes the possible influence of the area of a country into account. These three independent variables and the dependent variable have been transformed into logarithms, to account for non-linear effects in the specification.
MUSL and PROT are
related to the ‘religious’ structure of a country and includes the percentage of Muslims, and Protestants respectively.
COMM is a dummy for (former) communist countries5.
This dummy is equal to 1 for (former) communist countries and 0 for other countries. β1 to β6 are the regression coefficients to be estimated.
ε is the error term for the
regression model, which is the unknown variation (the vertical deviation from the unknown true regression line)6. Initial analysis examined the fitting of the model described by the equation above to the entire data set. Significant independent variables were population, wealth, communism and Protestantism resulting in the following regression model (with only significant variables):
4 As a point of departure, a correlation-matrix was taken for the selection of these variables and to avoid problems of multicollinearity. In this respect the percentage of Catholic inhabitants was omitted, due to the high correlation with percentage of Muslims (r=-0.435; sign. 0.000) and moreover only a low correlation with medal points. Urbanisation, which was highly correlated with GDP/head (r=0.605; sign. 0.000), was left out for the same reason; area, which correlates highly with population (r= 0.558; sign. 0.000) was replaced by density (=population/ area); the number of protestant inhabitants in a nation correlated significant at the 0.05 level with GDP/capita (r=0,281) but was included because of the significant correlation with medal points (r=0,317; sign. 0.000). 5 Methodologically speaking, assessing ‘political’ influence is not self-evident. The reason for examining particular variables is that these variables are exogenous with respect to the perspective of the responsibility for sport policy. The economic and sociological situation of a country are most probably so. The political context of a country is, however, a choice of variable [given that countries can choose their own political system]. In the following text, when we speak of the political context as an ‘external factor’, we refer to the characteristics of the political context that cannot be influenced by sport policy. In all the studies that will be discussed in chapter three, political variables are included because of the dominant role of former communist nations, an effect that is still currently noticeable (Hoffmann et al., 2002a; Stamm & Lamprecht, 2001). In concrete terms, although a country may have ‘chosen’ a communist regime, this occurs at another (political) level than the one in which sport policy is developed. 6 The source of our data is the World Factbook 2004 (http://www.cIa.gov/cia/publications/factbook/index.html) for the variables of population, wealth, religion and density. The data relating to the former communist countries derives from Encarta (http://encarta.msn.com/related_761572241/Communism.html). The information on the medals comes from the IOC official report 2004, available at: http://www.olympic.org/uk/games/index_uk.asp
25
2. Measuring the success of nations in elite sport
Ln MED = -5.158 + 0.496 Ln (POP) + 0.598 Ln (GDPCAP) + 0.016 PROT + 1.036 COMM + ε t-values (-4.861)
(6.993)
(5.690)
(1.976)
(4.506)
Adjusted R²: 52.8%
Diagnostic tests were done to see if there were any outlying values (with respect to their Y-values and/or their X values) that could influence the appropriateness of the fitted regression function. These are usually carried out indirectly through an examination of the residuals for example from a boxplot (Kutner, Nachtsheim, Neter et al., 2005).
A
boxplot with the Mahalanobis distance identifies a case as having extreme values on one or more of the independent variables. Figure 2.6 below shows that, in this model, three nations could be identified as outliers: Jamaica, Iran and USA7. Figure 2.6: Boxplot of regression residuals with independent variables: ln(POP), ln(GDP/cap), ln(DENS), Muslim, Protestant and communism and weighted medals as dependent variable,
Mahalanobis Distance
illustrating outliers in Mahalanobis distance
25,00000
S
Jamaica
S
Iran
20,00000
A United States
15,00000
10,00000
5,00000
A major reason for discarding an outlier is that ‘under the least squares method, a fitted line may be pulled disproportionately toward an outlying observation because the sum of squared deviations is minimized’ (Kutner et al., 2005, p23).
Further analysis showed
that this was the case with Jamaica, mainly because of its Protestant nature. Indeed,
7 Also the Centered Leverage Values were too high for the USA and Jamaica. We therefore need to ascertain how influential these cases are in the fitting of the regression function.
26
2. Measuring the success of nations in elite sport
when Jamaica was omitted from the regression model, Protestantism was not a significant variable anymore.
According to Kutner et al. (2005) a safe rule frequently
suggested is to discard an outlier only if there is direct evidence that it represents an error in recording, a miscalculation, a malfunctioning of equipment or a similar type of circumstance.
Refined measures, to ascertain whether nations are influential, showed
that for Jamaica the ‘DfFITS’ (= 0.475) (which measures its influence on the fitted value) was too high and so was the DFBETA value (= 0.4) for the influence on the intercept8. Although these measures are not problematic it was decided to omit Jamaica from the sample mainly because of its influence on the regression parameters (Protestantism) and thus on the other residuals. The influence measurements for both Iran and the United States did not take extreme values.
Furthermore there were no straight indicators to
omit Iran or the USA after repeating the OLS with and without these nations. Moreover the (adjusted) R square increased slightly when Jamaica was excluded which was not the case (or rather the opposite) with the other nations. After omitting Jamaica, 74 cases remained in the sample. Table 2.5 below shows the regression results of all the independent variables for the final model.
In general, we
note that the model provides a reasonable explanation for the variation in the dependent variables: 53.8% of the variation in medal success is explained by the model. Furthermore multicollinearity diagnostics revealed that the independent variables were not correlated too highly (VIF9 did not exceed 1.2).
8 Three measures that are widely used in practice were tested in order to analyse whether the outlying cases are influential, each based on the omission of a single case to measure its influence: DfFITS, Cook’s Distance and DFBETAs. Cook’s distance was only 0.117 for Jamaica, the largest but one (after India) but within the required critical values. 9 According to Kutner et al. (2005) a maximum Variance Inflation Factor (VIF) in excess of 10 is frequently taken as an indicator that multicollinearity may be unduly influencing the least squares estimates.
27
2. Measuring the success of nations in elite sport
Table 2.5: Linear regression analysis with all selected independent variables (ENTER); explained variable: weighted number of medals (gold=3, silver=2, bronze=1) in Athens 2004 Coefficients Constant
t-values
-5.559**
-5.027
0.519**
7.104
LOG Population LOG GDP/CAP
0.632**
5.837
LOG Density
-0.125
-1.809
Muslim
-0.001
-0.349
0.010
0.992 4.628
Protestant Communist
1.065**
Number of observations
74
‘Adjusted’ determination
0.538
coefficient R² Level of significance
1% (**)
The results in table 2.5 corroborate our earlier findings, namely that the Olympic success of countries is to a large extent determined by socio-economic factors.
Three
determinants appear to be highly significant: population size, per capita GDP and the (former) communist character of a country. Finally, a more parsimonious estimation was done, that is a regression where the nonsignificant coefficients are (stepwise) omitted.
Interestingly, the stepwise estimation
showed exactly the same results as a normal ‘enter’ method. For the final determination of the coefficients, the regression was repeated taking only the significant independent variables into account: Ln (population), Ln (GDPCAP) and communism (dummy). These results are showed in Table 2.6. Table 2.6: Stepwise Ordinary Least Squares for weighted number of medals (gold=3, silver=2, bronze=1) in Athens 2004 Dependent variable
Independent
Coefficients
t-value
variables Ln(Population)
.193
Ln(Population) Athens 2004 Ln (Medal points)
.388
Ln(GDP/cap) (Constant)
-5.532
-5.405
Ln(Population)
.511
7.215
Ln(GDP/cap)
.669
6.534
1.068
4.620
Comm
28
Adjusted R²
.524
2. Measuring the success of nations in elite sport
A stepwise regression (whereby significant explanatory variables are added one by one and deleted one by one), as shown in Table 2.6 indicates that the population size is responsible for 19% of the international success. Wealth adds another 38.8% and together with the political system for (former) communist countries, we end up with a model where 52.4% of the international success is explained. This model was used for further analysis of the residuals.
B. Analysis of residuals Whether a linear regression function is appropriate for the data being analysed can be studied from a residual plot against the predictor variable and to examine whether the variance of the error terms is constant (Kutner et. al., 2005). Furthermore the residual plot can easily be used for examining other facets of the aptness of the model, like heteroscedasticity.
The standard assumption is that the residuals are normally
distributed and constant, or that no systematic pattern is present. In this respect the Figures below show a histogram and normal scores plot of the residuals. Figure 2.7: Histogram (left) and normal probability plot (right) of the regression standardised residuals for weighted number of medals (gold=3, silver=2, bronze=1) in Athens 2004
1,0
0,8
Expected Cum Prob
Frequency
15
10
5
0,6
0,4
0,2
0 -3
-2
-1
0
1
2
Regression Standardized Residual
3
Mean =1,62E-15 Std. Dev. =0,979 N =74
0,0
0,0
0,2
0,4
0,6
0,8
1,0
Observed Cum Prob
From the Figure above it can be seen that there is no serious skew and that the residual reasonably approximates a normalcy.
This was confirmed with a Kolmogorov Smirnov
Test for normality, which was not significant (sign. = 0.200) and a Shapiro Wilk (sign. = 0.488).
29
2. Measuring the success of nations in elite sport
Figure 2.8 plots the residuals against the predictor variable to examine for independence and homoscedasticity. Figure 2.8: Standardized residuals versus standardized predictors. Please note that for the clarity of the figure, only a fraction of the medal winning nations is presented by name
3
Cuba
Regression Standardized Residual
Australia
Kenya
2
Ethiopia
Greece
Bahamas
1
Korea (South)
Netherlands Italy
Zimbabwe
Russia
France
United States
United Kingdom
Eritrea Cameroon
China
Canada
0 Paraguay
Brazil
Trinidad & Tobago Nigeria
-1
Germany
Poland
Mongolia Belgium
Syria
Chile
Hong Kong
-2 Colombia
India
-3
-3
-2
-1
0
1
2
3
Regression Standardized Predicted Value
As can be seen in Figure 2.8, no systematic pattern is evident, there are thus equal error variances, which is called homoscedasticity (Kutner et al., 2005). Tests for outliers on this final model as were done in the figure above and a boxplot of the residuals indicate that India may be an outlying case (in the Y direction).
To test this we use the
Bonferroni simultaneaous test procedure with a family significance level of alpha = .10. The semistudentized residual for India is -3.43 and the required Bonferroni value is l3.46l, which suggest that India is at the acceptable limit of being an outlier. Another outlying cases (in X-direction) is China which was identified in a boxplot of the Mahalanobis distance (mah. China = 11.487) and Centred Leverage Values (Lev. China = 0.157).
Further statistical analysis revealed that none of these nations are influential,
nor on the single fitted value (DfFits), on all fitted values (Cook’s distance) or on the regression coefficients (DFBeta).
30
2. Measuring the success of nations in elite sport
Figure 2.8 also shows the nations that appear to be more successful and have a positive standardised residual. The following section covers in greater depth this residual analysis as a way to determine success of nations controlling for macro-level determinants.
C. Identifying relative success of nations A regression analysis serves two purposes. First it identifies the determinants for international success on the macro-level. Second, under ceteris paribus conditions, an analysis of residuals allows the comparison of countries and thus also the determination of their relative success.
The analysis of residuals compares this ‘prediction’ with the
weighted number of medals, actually won.
Using a ‘case-by-case’ analysis, we can
answer the question: ‘which countries are successful, if one accounts for socio-economic variables?’ Figure 2.9 illustrates, in a two-dimensional space, (that is for a situation in which we only analyse one explanatory variable), the regression line that is the best fitting line of a topographical point system (Ottoy, Van Vooren & Hughe, 1993).
The
points in the figure show the positions of the respective countries. The regression line divides the points (countries) into two groups. A successful country is one (above the regression line) that performs better than one would expect on the basis of macroeconomic determinants. The degree to which the country performs “better” is reflected in the size of the residual.
Figure 2.9: Graphic representation of a linear regression and the policy as part of the residual
Absolute Success
Residual: variation that cannot be explained by socio-economic variables = relative success of countries
Population, GDP/head or communism
31
2. Measuring the success of nations in elite sport
Applying this statistical technique to our study, the starting point of our analysis is now that the residual represents (partly) the effects of elite sport policies. In other words, the positive residual of nations is among others the result of effective elite sport policies, which is part of the unexplained variance. It should be noted that some other factors will, to an unknown extent, also influence the size of the residual. Examples are the elite sport culture, the tradition of sport and sporting success in a nation and maybe just ‘coincidence’. However, these factors cannot be fashioned by policies. Although it is not known to what extent the residual can be explained by elite sport policies, these are the only factors that can be fashioned. Table 2.7 offers an overview of the ten countries with the greatest relative success in Athens, taking the differences in population size, wealth and, where relevant, the (former) communist character of the country into consideration. This ranking reflects the regression
residuals
[residual
=
absolute
success
–
predicted
success]
in
the
multidimensional model with only the significant variables.
Table 2.7: Ten best performing nations during the Olympic Games in Athens according to OLS method (relative success), with Log (points of medals) as the dependent variable and Log (pop), Log (GDP/head and communism as independent variables
Rank 1
Country Cuba
Predicted
Residual
2.11
1.84
2
Australia
2.87
1.73
3
Kenya
0.86
1.70
4
Ethiopia
1.01
1.63
5
Greece
2.30
1.22
6
Korea (South)
3.00
1.09
7
Bahamas
0.36
1.03
8
Russia
4.16
0.99
9
Zimbabwe
0.82
0.98
United States
4.42
0.93
10
Table 2.7 shows that Cuba was the most successful country in Athens in terms of relative success, followed by Australia, Kenya, Ethiopia and Greece. Kenya’s third and Bahamas’ seventh position is noteworthy.
On the basis of the socio-economic context it was
anticipated that these nations would win respectively only 0.86 and 0.36 medal points. In practice, things turned out differently. Moreover, the results showed that the United
32
2. Measuring the success of nations in elite sport
States is ranked 10th, still with a positive residual (0.71). It should be noted that many (former) communist nations, would be ranked higher, when communism would not be included in the model.
In this case, R² would decrease to only 38.8% and Hungary,
Romania, Belarus, Georgia, Ukraine would all be in the top ten. Russia would then be second.
For the reason mentioned earlier, we decided that communism should be
included as communism is a factor that can not be fashioned by policies and more over may be a cultural indicator. When Table 2.7. is compared to the top ten of nations in terms of absolute success (Table 2.1), it can be seen that only three nations are mentioned in both tables: the USA (ranked first in absolute success), Russia (ranked second) and Australia (ranked fourth). Hence,
these
nations
are
highly
successful
when
both
absolute
and
relative
measurements are used.
2.6. Other measurements of Performance: the Olympic Winter Games The preceding analysis has focused primarily on the Summer Olympic Games, illustrating different methods using the Athens Games as a case study. The Olympic Games are a truly global event containing a portfolio of different sports that are popular and recognised in a number of nations.
Moreover, Olympic performances are seen as the
ultimate performance in many sports. In this regard the Olympic Summer Games were indicated to be the best measure of the overall sporting performances of nations. However, some nations may prefer to invest in Winter sports because their geographic environment makes this likely.
This was not included in the preceding analysis.
To
illustrate the point, Canada is particularly successful in speed skating and in Short track whilst Norway is successful in cross country skiing and alpine skiing. These nations do not belong to the top 20 in summer Olympic sports. Thus there is the danger that the success of nations who particularly value success in the Winter Olympic Games is underestimated when only summer Olympic sports are taken into account.
We will
therefore apply the same analysis for Winter Games as that presented earlier in this chapter, using market share as a measurement of success in absolute terms (i.e. without controlling for socio- and macro-economic determinants) and residual analysis as a measurement of relative success. In terms of efficiency of elite sport policies an analysis for winter sports, comparable with the preceding analysis, may help to understand a broader interpretation of success with regard to nations’ elite sport investments. However, the Winter Games are less globalized and have different characteristics
33
2. Measuring the success of nations in elite sport
compared to the Summer Games and their size is a little less than 20 percent of the Summer Games in terms of participants and events (Kuper & Sterken, 2003b). Only 77 National Olympic Committees participated in the Games in Salt Lake City and of these 25 won medals.
Market share during the Winter Olympic Games For the last five Winter Olympic Games, we have replicated the market share figures for the ten best performing nations in Turin 2006 and these are shown in Figure 2.10. Figure 2.10: Market share in Winter Olympics 1992-2006 of the ten best performing nations (in Turin 2006)
18,0%
16,0%
16,3% 15,1% 14,5%
14,1%
14,0% 13,7%
16,0%
13,4% 12,5%
12,0%
11,7% 10,3%
10,2%
10,0%
9,5% 8,7%
8,2% 8,0%
7,8%
7,3% 6,8%
7,2%
6,8%
6,0%
5,7% 4,0%
4,1%
4,4%
2,0%
0,0% 1992 Albertville
1994 Lillehammer
GER
34
1998 Nagano USA
CAN
AUT
2002 Salt Lake City RUS (EUN in 1992)
2006 Turin
2. Measuring the success of nations in elite sport
18,0%
16,0%
15,6%
14,0%
13,4% 12,8%
12,0%
11,9% 10,7%
10,0%
8,2%
8,0%
6,0% 4,9%
5,2%
5,3%
4,0%
4,0%
4,2%
4,1%
2,6%
2,1%
3,2% 2,0%
6,2% 6,0% 5,6% 5,2%
1,7%
1,9%
2,2%
1,5%
1,2%
0,0% 1992 Albertville
1994 Lillehammer NOR
1998 Nagano SWE
SUI
2002 Salt Lake City KOR
2006 Turin
ITA
From Figure 2.10 three general conclusions can be drawn.
First, the market share of
Winter Games for the top ten nations is higher than for Summer Games. This finding may indicate that competition is less in Winter Games and for participating nations it is easier to win medals compared to Summer Games. Second, in Turin the medals have been divided more equally among the top ten nations than in Salt Lake City, resulting in a range of 8.3% between the first and tenth nation in Turin versus 14.1% in Salt Lake City. The market share figures for Korea (5.2%), Switzerland (5.6%), Sweden (6.0%) and Norway (6.2%) are very close to each other in Turin. Third, five nations in Table 2.10 are not in the top ten list of summer sports: Norway, Sweden, Switzerland, Austria and Canada. Figure 2.10 also shows that Germany has always played a leading role in winter sports, with the exception of Lillehammer 1994 where the Russian federation heads the other nations, despite the break up of the former USSR. Russian performances decreased until Salt Lake City 2002 with a remarkable increase in 2006 (Turin).
Germany’s market
share decreased with 3.5% in Turin. Austria begins the time series with a third place in Albertville 1992 and then has a remarkable decrease of 7.3% in the run up to Lillehammer (1994). Afterwards a steady increasing trend can be noticed with in Turin an equal score to Canada who also improved its market share from 4.1% in Albertville (1992) to 9.5% in Turin.
The USA was the host nation in Salt Lake City where it
improved its market share from a fourth in Albertville to a second in the last two Olympic
35
2. Measuring the success of nations in elite sport
Games, despite falling market share in Turin. Norway also gains its highest market share in Lillehammer when it hosted the Olympic Games.
Relative success in Winter Olympic Games Contrary to the Summer Games only a few authors have estimated success on Winter Games (see Balmer, Nevill & Williams, 2001; Kuper & Sterken, 2003b). Making an OLS estimation on one edition of the Winter Games is less plausible than for Summer Games because of the small data set. Without going into as much detail as with Summer Games, an OLS estimation was made for the 25 medal winning nations in Salt Lake City 2002.
The same independent parameters as for the Summer Games were
entered with one additional variable: mountain elevation for Alpine Skiing (MOUNT). The variable was inserted as a dummy equal to ‘1’ when mountains with a minimum height are available in the country and ‘0’ when they were not10. Initial analysis of the entire data set was done with the functional form thus to be estimated as follows:
MED (weighted points) = β0 + β1 POP + β2 GDPCAP + β3 DENS + β4 MUSL + β5 PROT + β6 COMM + β7 MOUNT
11
+ε
Interestingly, the results revealed that only wealth and communism are significant variables in this estimation for Winter Games. Population is not significant (t= 1.134), nor is mountain elevation (t=1.094).
The latter corresponds with earlier findings of
Kuper and Sterken (2003b) who estimated time series models for each Olympic Games from 1924 – 1998.
Apparently, contrary to the Summer Games, a large basis for
recruitment of young talent is less important in Winter Games. A logical explanation for the wealth of nations as the main precondition for success may be found in the fact that Winter Sports are generally more expensive sports and require more expensive equipment. Poor nations can send fewer athletes to the Games.
The poorest medal
winning nations in Salt Lake City, in terms of GDP/CAP (60%
3 4 5
0.04
0.02
14.92 19.90 >
9.95
4.97
2.49
score 1
standards
2 3 4 5
score 1
average/5
3 4 5
2
score 1
average/5
>
Population (source: CIA World factbook)
352.40 5
Flanders: government (sport + employment contracts) + lottery money to Bloso (m62.7€ + m5.63€) + department youth and Sport (m1.7€) and lottery to the BOIC (m1.86€)
88.10
44.05
score 1
average/5
264.30
1
5
220.25
average
4
40.8
610
WAL
Canada: excl. provincial government budget on sport (idem other countries - but in Canada it is probably more important)
2
130
UK
176.20
2
127.2
NOR
3
4
273.7
NED
upper limit
2
1
1
score
71.9
67.9
Total national expenditure on sport (cash terms)
ITA*
standards
Italy: includes expenditure on sports facilities (€1.5m) and payments made to nine military clubs (€1.2m) that employ top level athletes and coaches
FLA*
CAN*
x million Euro
There is sufficient financial support for sport
Standards for pillar 1: "financial support"
score
FL 8.4 1
CAN* 38.1 3
score
Increase / decrease in national expenditure on elite sport since 1999-2003 score
1999
National expenditure on elite sport (as a proportion of total national expenditure on sport) score
National expenditure on elite sport (per head of population)
Population (source: CIA World factbook)
115.38 5
NA NA
3.9
FLA
CAN NA
1
5
FLA
CAN 11.68
3
2
56.11
1.34
1.17
FLA 6.269
CAN 32.508
Canada: estimation; budget does not include provincial government budget on elite sport
Italy: estimation of what goes to elite sport from NOC and lottery budget; incl. money for 8 military clubs
National expenditure on elite sport (cash terms)
There is sufficient financial support for elite sport
NA
NA
NA
ITA
4
45.67
ITA
4
2.15
58.057
ITA
5
125
IT*
4
54.07
27
NED
3
32.70
NED
4
2.55
16.318
NED
3
41.6
Nl
3
27.27
5.5
NOR
1
5.38
NOR
3
1.53
4.575
NOR
1
7
NOR
5
125.98
39.96
UK
2
14.80
UK
3
1.50
60.271
UK
5
90.3
UK
5
182.53
1.345
WAL
1
9.31
WAL
2
0.93
4.077
WAL
1
3.8
WALL
126.31
average
29.28
average
1.60
average
44.89
average
3 4 5
2
score 1
standards
2 3 4 5
score 1
average/5
3 4 5
2
score 1
average/5
2 3 4 5
score 1
average/5
standards
0-30% 31-60% >60%
(-1)-(-20)%
11.71
5.86
1.92 2.55 >
1.28
0.64
0.32
35.91 53.86 71.82 >
17.95
8.98
upper limit
score
22.8 1
27.5 2
score
68 2
55 2
FLA
CAN 0.34
3
1
0.50
3.64
0.85
FLA 6.269
CAN 32.508
77 5
1.82
ITA
2
2.41
58.057
ITA
5
140
ITA
total financial support for NGBs for elite sport score
FLA 5.7 1
CAN 18.5 3
3
25.1
ITA
National Governing Bodies (NGBs) receive sufficient financial support for elite sport
number of recognised & funded NGBs
total financial support for NGBs: average funding per NGB
score
total financial support for NGBs for sport per head of population
Population (source: CIA World factbook)
UK: all sports councils; excluding awards by sports councils for WCP or international events
Flanders: only funding from Bloso to NGBs
Canada: excluding provincial government subsidisations to NGBs
total financial support for NGBs for sport (cash terms)
FLA*
CAN*
National Governing Bodies (NGBs) receive sufficient financial support for sport
4
31
NED
72 3
0.91
NED
4
4.03
16.318
NED
3
65.7
Nl
1
4
NOR
55 2
0.35
NOR
4
4.15
4.575
NOR
1
19
NOR
5
50
UK
120 3
0.79
UK
2
1.58
60.271
UK
4
95
UK
1
3.81
WAL
64 1
0.14
WAL
2
2.21
4.077
WAL
1
9
WALL
23.02
average
0.81
average
3.14
average
63.17
average
3 4 5
2
score 1
average/5
2 3 4 5
score 1
average/5
3 4 5
2
score 1
average/5
2 3 4 5
score 1
average/5
27.62 36.83 >
18.41
9.21
4.60
0.65 0.97 1.29 >
0.32
0.16
3.77 5.03 >
2.51
1.26
0.63
50.53 75.80 101.07 >
25.27
12.63
FLA 0.22 26 2
CAN 0.39 47 2
average national subsidisation for elite sports per NGB
number of recognised and funded NGBs for elite sport score
41 3
0.61
ITA
63 3
0.49
NED
30 1
0.13
NOR
40 5
1.25
UK
36 1
0.11
WAL
0.53
average
2 3 4 5
score 1
average/5
0.43 0.64 0.86 >
0.21
0.11
2 1
16-20
21-23
5
4
3
2
1
There is a cabinet minister for sport and a ministry of sport (both with sport in the name)
There is a ministry, with sport in the name, but no cabinet minister for sport
There is a ministry of sport, but no minister for sport
Sport is part of another ministry and sport is delegated to another minister(s)
There is no ministry of sport, and no minister responsible for sport
Sport is recognised as a valuable political task: there is a ministry and (cabinet) minister for sport
There is a strong coordination of all the agencies in elite sport
4 3
11-15
5
6-10
1-5
ranking of 23 countries
criteria for public sector efficiency
Ranking score on the research of Public sector efficiency (European Central Bank, 2003)
Simplicity of administration: Public sector efficiency (European Central Bank, 2003)
Norway: There is no Ministry of sport. Sport is not mentioned in the name of the ministry, but forms part of the ‘Kultur- og Kirkedepartementet’ (Ministry of Culture and Church Affairs). Within this ministry, there is a Department of Sport Policy Italy: There is no Ministry of sport. Sport is not mentioned in the name of the ministry, but forms part of the ‘Ministero per i Beni e le Attività Culturali’ (Ministry of Cultural Heritage and Activities). Within this ministry, there is a Secretariat of Sport. The ministerial functions for sport have been delegated by the Minister to the Junior Minister (not being a Cabinet Minister); Canada: Stephen Owen is the Minister of Western Economic Diversification and Minister of State (Sport) but he is not a Cabinet Minister; sport belongs to the department of Canadian Heritage; note that in February 06, Michael D. Chong is Minister of Intergovernmental Affairs and Minister for Sport (this would change the score from Canada in a 5)
The Netherlands: VWS is the Ministry of Sport; Hans Hoogervorst is Minister of Sport, but he is not a cabinet minister; UK: Richard Caborn MP is a Minister in the Department for Culture, Media and Sport (which is headed by the Secretary of State, Tessa Jowell), but he is not a Cabinet Minister.
Flanders: Anciaux (2006) is a (cabinet) Minister of Sport youth, culture and Brussels Wallony: Claude eerdekens is a (cabinet) Minister of public functions and sport
Italy: 23rd
Belgium (Flanders+Wallony): 20th; the Netherlands: 16th
Canada: 12th;
Norway: 7th; UK: 8th
Legend: O (overall sports policy questionnaire), ESC (Elite Sport Climate survey), ATL (athletes), COA (coaches), PD (Performance Directors); NGB: national governing body
Standards for pillar 2: "organisation and structure of sport policies: an Integrated approach to policy development"
3
1
Elite sport is coordinated by an organisation that takes most important decisions (like expenditure, subsidy, etc), but this organisation is also responsible for national sport for all; there is, however, an elite sport department.
There is no organisation with sole responsibility for elite sport and no department for elite sport
5
3
1
High level of coordination: there is one main organisation at national level which makes decisions about the majority of expenditures and activities in elite sport; eventually the Olympic Committee is merged with the national sport administration
Reasonable level of national coordination: there is more than one organisation at national level spending money (independently) on elite sport, but there is a coordination between these agencies so that expenditure and activity is delineated transparently to avoid duplication
Low level of national coordination: there is more than one organisation at national level spending money on elite sport; this is not recorded and not centralised; it is not known at national level how much money NGBs and athletes in the country receive.
National coordination of financial inputs and activities ( horizontal direction): expenditure on elite sport at national level are centrally recorded and coordinated, so that no overlap takes place
5
Elite sport is coordinated by an organisation with responsibility only for elite sport and not sport for all; most important decisions like expenditure, subsidies, etc, are taken by this organisation
There is an organisation at national level with specific responsibility for elite sport only (not sport for all)
In Wallony there is no coordinating agency for the activities of Adeps and BOIC; Adeps has no elite sport department; expenditures on elite sport from different organisations are not nationally coordinated or recorded.
Flanders: BOIC, Bloso and Minister all spend money on elite sport; these have been coordinated by a steering group since 2003; in Canada elite sport is mainly the responsibility of Sport Canada and also the COC (Olympic Committee), they have meetings to delineate responsibilities and expenditures.
Italy: CONI coordinates all money flows at national level; Norway - Olympiatoppen: idem (except for the NGB's in football and skiing, which have enough money on their own). Furthermore, the current NOC, was formed in 1996 following an amalgamation of the Norwegian Olympic Committee and Confederation of Sports; the Netherlands: NOC*NSF is a fusion of the sport administration and Olympic Committee; NOC*NSF coordinates expenditures from SNS (lottery funding) and VWS (Ministry); responsibilities are delineated transparently to ensure clarity of purpose and to avoid duplication; UK Sport is the most influential agency in the UK regarding expenditure on elite sport; The role of the British Olympic Comittee is limited just to sending athletes to the Olympic Games
the Netherlands: NOC*NSF is the main coordinator of elite sports policies, but also has responsibilities for sport for all; idem for Italy (CONI): combining powers of both a Ministry of Sport and a Confederation of Sport NGBs; In Flanders, Bloso is responsible for elite sport and sport for all, but there is a department for elite sport (however, this department does not decide on funding for NGBs); idem for Wallony, ADEPS has a department "Elites et cadres"; although the BOIC in Flanders and Wallony is only responsible for elite sport, this is not the main (funding) body. Idem for Sport Canada as a main body in elite sport; however their focus is on elite sport; it is divided into five groups including Policy, Programs, Games and Hosting, Management and Strategic Planning and Accountability
Olympiatoppen (Norway), a division of the NOC, and UKSport are mainly responsible for elite sport and take the most important decisions.
3
1
Reasonable level of national coordination: there is major expenditure and activity in elite sport at regional/district level; this is nationally recorded and initiatives are taken to gear out this activity and avoid duplication of effort and payments;
Low level of coordination/analysis of expenditure at regional/provincial level: there are important initiatives in elite sport at regional level; these are not nationally recorded and coordinated, consequently there is an overlap of funding to athletes and/or NGBs
26 55 40 30 47 41 36
Norms for number of funded NGBs funding for elite sport 50.0%
3
2
1
points
20.1-50.0%
0.1 - 20.0%
0 tot (-19.9)