Incentivizing Collaborative BIM-Enabled Projects 9781628256246, 9781628256239

The use of digital representations to aid in projects--Building Information Modeling (BIM)--is gaining traction worldwid

214 98 3MB

English Pages 208 Year 2018

Report DMCA / Copyright

DOWNLOAD FILE

Polecaj historie

Incentivizing Collaborative BIM-Enabled Projects
 9781628256246, 9781628256239

Citation preview

INCENTIVIZING COLLABORATIVE

BIM-ENABLED PROJECTS A Synthesis of Agency and Behavioral Approaches

CHEN-YU CHANG, P h D (LON)

Incentivizing Collaborative BIM-Enabled Projects A Synthesis of Agency and Behavioral Approaches Chen-Yu Chang, PhD (Lon) Director, Bartlett Infrastructure Center Bartlett School of Construction and Project Management University College London, UK

Library of Congress Cataloging-in-Publication Data has been applied for. ISBN: 978-1-62825-623-9 Published by: Project Management Institute, Inc. 14 Campus Boulevard Newtown Square, Pennsylvania 19073-3299 USA Phone: +610-356-4600 Fax: +610-356-4647 Email: [email protected] Internet: www.PMI.org ©2018 Project Management Institute, Inc. All rights reserved. Our copyright content is protected by U.S. intellectual property law that is recognized by most countries. To republish or reproduce our content, you must obtain our permission. Please go to http://www.pmi.org/permissions for details. PMI, the PMI logo, PMBOK, OPM3, PMP, CAPM, PgMP, PfMP, PMI-RMP, PMI-SP, PMI-ACP, PMI-PBA, PROJECT MANAGEMENT JOURNAL, PM NETWORK, PMI TODAY, PULSE OF THE PROFESSION and the slogan MAKING PROJECT MANAGEMENT INDISPENSABLE FOR BUSINESS RESULTS are all marks of Project Management Institute, Inc. For a comprehensive list of PMI trademarks, contact the PMI Legal Department. All other trademarks, service marks, trade names, trade dress, product names and logos appearing herein are the property of their respective owners.  Any rights not expressly granted herein are reserved. To place a Trade Order or for pricing information, please contact Independent Publishers Group: Independent Publishers Group Order Department 814 North Franklin Street Chicago, IL 60610 USA Phone: +1 800-888-4741 Fax: +1 312-337-5985 Email: [email protected] (For orders only) For all other inquiries, please contact the PMI Book Service Center. PMI Book Service Center P.O. Box 932683, Atlanta, GA 31193-2683 USA Phone: 1-866-276-4764 (within the U.S. or Canada) or +1-770-280-4129 (globally) Fax: +1-770-280-4113 Email: [email protected] Printed in the United States of America. No part of this work may be reproduced or transmitted in any form or by any means, electronic, manual, photocopying, recording, or by any information storage and retrieval system, without prior written permission of the publisher. The paper used in this book complies with the Permanent Paper Standard issued by the National Information Standards Organization (Z39.48—1984). 10 9 8 7 6 5 4 3 2 1

To my parents, whose love is the everlasting power of my life.

Acknowledgments It was not possible to complete this report without the excellent assistance of several of my doctoral and master’s degree students, including Robert Howard, Mingyu Zhu, Xiao Sun, Anna Wang, and Vinayak Arakere. The data collected and the analysis conducted by these students enabled me to construct an evidence-based theory broad enough to understand BIM proliferation in three distinct national environments. It is hoped that the attempt to instill behavioral insights into the principal-agent theory could evolve into a robust theoretical approach that is conducive to the design of incentive structures for BIM projects.

v

Contents Chapter 1 – Executive Summary...............................................................1 The Research Questions ................................................................................. 1 Context of the Research................................................................................. 2 Brief Overview of Methodology.................................................................... 2 Clearly Stated Project Findings .....................................................................  3 Applications to Practice................................................................................. 6 Chapter 2 – Introduction........................................................................ 9 Research Background and Significance........................................................ 9 Aims and Objectives ..................................................................................... 12 Sketch of Research Process........................................................................... 13 Framing of Research Questions ................................................................... 14 Chapter 3 – Literature Review............................................................... 17 Studies of Collaborative Governance in Construction (IPD/Partnering).................................................................... 17 IPD and Incentivization................................................................................ 19 Studies of BIM Diffusion.............................................................................. 27 Behavioral Approaches..................................................................................32 Chapter 4 – Analytical Framework........................................................35 Overview of the Life Cycle Theory of BIM Diffusion..................................35 What Problems Can Be Solved by BIM?..................................................... 36 Behavioral Influences................................................................................... 44 Chapter 5 – Research Methodology...................................................... 47 Research Steps.............................................................................................. 47 Research Method.......................................................................................... 47

vii

viii

Incentivizing Collaborative BIM-Enabled Projects

Chapter 6 – Case Studies – China .......................................................... 51 Overview........................................................................................................ 51 Project Background Information ................................................................ 52 Analysis......................................................................................................... 57 The Desired Delivery Environment for BIM Implementation .................. 69 Chapter Conclusions.................................................................................... 76 Chapter 7 – Empirical Investigation – China (1)................................... 79 Problem Identification................................................................................. 79 The Model and Hypotheses.........................................................................80 Questionnaire Development.......................................................................80 Empirical Analysis........................................................................................ 84 Discussion.....................................................................................................90 Chapter Conclusions.................................................................................... 93 Chapter 8 – Empirical Investigation – China (2).................................. 95 Introduction.................................................................................................. 95 The Model..................................................................................................... 97 Result ...........................................................................................................104 Discussions....................................................................................................111 Chapter Conclusions....................................................................................114 Chapter 9 – Case Studies – United Kingdom........................................115 Problem Identification.................................................................................115 Method..........................................................................................................116 Cross-Case Analysis.....................................................................................120 Chapter 10 – Empirical Investigation – United Kingdom................... 135 Hypotheses .................................................................................................. 135 Research Method.........................................................................................142 SEM Analysis ............................................................................................... 145 Discussions..................................................................................................148 Chapter 11 – Case Studies – United States ............................................151 Introduction .................................................................................................151 Project Background..................................................................................... 152 Interview Result .......................................................................................... 153

Contents

ix

Chapter 12 – Empirical Investigation – United States ........................ 159 Development of Hypotheses ...................................................................... 159 Research Method .........................................................................................161 Discussions..................................................................................................164 Chapter 13 – Conclusions .................................................................... 167 China Survey (1): BIM and IPD ................................................................... 167 China Survey (2): BIM and Incentivization Under the Influence of Behavioral Biases....................................................................168 UK Survey.....................................................................................................168 U.S. Survey...................................................................................................169 Appendix...............................................................................................171 Budget...........................................................................................................171 Summary of Proposal Objectives vs. Research Accomplishments.........................................................................171 References ........................................................................................... 175 About the Author................................................................................. 193 Index.................................................................................................... 195

C H A P T E R

1

Executive Summary The Research Questions While the concept of Building Information Modeling (BIM) has been around for nearly two decades, the large-scale rollout of this technique is a recent phenomenon. In this project, the focus was initially placed on the effect of BIM incentivization measures on project performance in China, the United Kingdom, and the United States. However, through several pilot studies, it soon became evident that, while “explicit” monetary incentivization is uncommon in practice, there are “implicit” nonmonetary motivators that could have driven BIM participation. Through contrasting BIM practices observed in three countries, it is manifest that the best framework for understanding BIM diffusion is based on a life cycle perspective; in other words, the development of BIM in different national contexts can be best portrayed as an S-curve: starting sluggishly, accelerating in the middle, and slowing down in the end. This view has led to a shift of research focus to the evolution of BIM practices in the process of BIM proliferation with the aim of exploring the effect of explicit and implicit incentivization on the effectiveness of BIM. BIM incentivization could arise out of two motives: intentional and spontaneous. In the current research, the former refers to the actions imposed by governments through BIM mandates, while the latter represents the actions chosen according to economic calculations. The central question of interest is: How do spontaneous and intentional incentivization drivers interact with each other over different stages of BIM diffusion? Given the different level of BIM maturity in three countries, the questions addressed for each country are slightly different: 1. China: In what way has mandate-driven BIM adoption in China propelled the evolution of a more integrated and better-­ incentivized delivery environment? 1

2

Incentivizing Collaborative BIM-Enabled Projects

2. United Kingdom: What are the critical factors that affect the impacts of BIM use on project performance? 3. United States: To what extent have incentivization measures led to an improvement in project performance through the lens of user-related factors?

Context of the Research The proliferation of BIM on a national scale has only emerged in recent years. The United States led the way, the United Kingdom came next, and China was catching up. The development of BIM in these three countries is at different levels of maturity. The environment in which BIM is implemented varies widely in these three countries. The greater maturity of BIM development, the stronger awareness of what governance could be desirable for implementing BIM; and hence, the more likely BIM is to be implemented within the right delivery environment—specifically: 1. Practitioners are still experimenting with different incentivization arrangements for BIM-enabled projects. The issue of interest is whether different methods have resulted in a systematic difference in project performance. 2. The drivers for BIM participation may emanate from tangible and intangible factors. As for the tangible effects, incentivization could be derived by two levels of forces: procurement system selection (first-order economizing effect) and incentive measures (second-order economizing effect). The effectiveness of these drivers can be scrutinized in an econometric way by drawing upon a complementary theoretical framework. 3. The effectiveness of incentivization measures could be affected by behavioral biases. The issue of interest is: To what extent have behavioral biases blunted the effectiveness of incentivization measures?

Brief Overview of Methodology Since this research attempts to capture a broad range of motivators that could incentivize BIM participation from the perspective of the

Executive Summary

3

principal-agent theory and behavioral theory, the research method should be exploratory in nature with the aim of discovering the current BIM incentivization practices through case studies and exploring their effectiveness using rigorous structural equation modeling (SEM) techniques.

Clearly Stated Project Findings Theory The benefit of BIM could vary significantly with the breadth, depth, and level of its employment on a project. The proliferation of BIM is most likely to take an S-curve trajectory. Initially, BIM is considered an add-on to existing project information systems. However, ill-fit delivery environments could militate against the potential of BIM for efficiency improvement. Following the tenet of transaction cost economics, the formation of the desirable delivery environment should be understood as a process of seeking alignment at two levels: The first-order economizing is achieved by getting the delivery system right, while further refinements can be made by getting incentives right. As the latter can be implemented in a piecemeal way, it could evoke much lower cost in transition than the former. The life cycle theory of BIM diffusion aims to capture the drivers discovered in the case studies and empirical investigations of three countries with equal emphasis on the internal benefit of BIM mandates in improving project performance and its external benefit in raising the awareness and thus acceptance of advanced incentivization measures.

Empirical Findings: China In the case studies of four advanced BIM applications in China, it is found that BIM employment is primarily driven by the practical need of dealing with complex projects. Owing to China’s rigid procurement law, the mandatory use of design-bid-build has forced project stages to be separated, making it infeasible to apply BIM coherently across the project life cycle. The case evidence reveals that an ill-fit delivery environment has led to transaction costs in various forms, especially about the duplication of effort in building BIM models. Using the data from 145 Chinese BIM-enabled projects, the first survey result of China reveals the channels through which BIM application could have impacted Integrated Project Delivery (IPD) acceptability: First, the firsthand experience of working in a BIM-enabled environment

4

Incentivizing Collaborative BIM-Enabled Projects

can make practitioners better appreciate the importance of incentivization, and that perception can drive the acceptability of IPD; second, the positive impact of BIM on communication quality can translate into another drive to support IPD. It is hoped that these robust statistical relationships can spark follow-on research to investigate the benefits of BIM in a wider context. In the second survey of China containing returns from 223 Chinese BIM-enabled projects, it is shown that the benefit of BIM is sensitive to the way it is implemented in the project, and its potential cannot be fully reaped in a delivery environment where project parties are not well incentivized to harness BIM in improving communication and coordination across project stages. However, the employment of integrated systems could present a great challenge in some countries as it may require new legislation and heavy training. It is easier to achieve the second-order incentive alignment through a piecemeal implementation of incentivization measures. The data reveal that the perception of actual incentivization created by the project delivery environment has a significant impact on the perceived usefulness of advanced incen­ tivization systems. The establishment of this awareness could herald a smoother process when introducing these systems into BIM-enabled projects. Embedding BIM in a better-incentivized environment provides a self-sustaining driver for the proliferation of BIM, which will eventually pave the way for the acceptance of integrated delivery systems. This research also explores the influence of behavioral biases on the effectiveness of incentivization.

Empirical Findings: United Kingdom Considering Britain as a country striving to catch up to become the world’s BIM leader and with a greater acceptance of advanced delivery practices, the focus of analysis is slightly different. United Kingdom (UK) case studies are designed to explore whether the incentivization structure embedded in two renowned integrated systems can facilitate the deployment of BIM. Interviews span the supply chains of two systems. The main conclusions are twofold: First, while both systems have an established incentivization scheme, its influence is only limited to first-tier contractors. Lack of system-wide incentivization has inhibited the participation of lower-tier contractors in the production of BIM information. Second, there is agreement among interviewees

Executive Summary

5

that it could be more efficient to incentivize BIM participation through a project-wide incentivization scheme instead of an incentivization scheme geared for BIM only. The primary objective of the UK survey is to explore the web of relationships among project attributes, incentive measures, behavioral biases, and the effectiveness of BIM. The causation runs as follows: In the execution of a complex project, it is more likely for the owner to engage with the contractors at early stages and motivate them with an incentive pool tied to the joint performance of the project team. Early involvement and the use of a group incentive scheme could increase the likelihood of project goals being set jointly. The performance of incentivization could benefit directly from the use of common project goal setting, and in­ directly from the experience the contractor has previously had working under the incentive scheme, but the contractor’s aversion toward risk could militate against it. How well incentivization works is a crucial factor in the effectiveness of BIM utilization.

Empirical Findings: The United States The United States (U.S.) case studies are based on two projects at the San Francisco Airport (Air Traffic Control Tower completed in October 2016 and the Terminal 1 Redevelopment Project that commenced in October 2016) that involved the use of an advanced form of integrated project delivery and the implementation of 6D BIM, which includes the functions of program management, cost management, and life cycle facility management in addition to traditional 3D geometric information. The Terminal 1 Redevelopment Project is seen as a more collaborative and technologically advanced construction environment than was utilized for the control tower. In these two projects, few monetary incentives were employed, and BIM participation was mainly driven by instilling trust into the partnering relationship. The U.S. empirical study is positioned differently: Explore the extent to which an organization’s external support for BIM implementation could be affected by BIM incentivization and user resistance, and how the joint effect of these factors could impact project performance. By drawing upon the insights from the technology acceptance model, this model aims to probe the paths through which the aforementioned cause-effect relation could transmit by changing the user’s perception of BIM usefulness and ease of use and, as a result, changing the intention

6

Incentivizing Collaborative BIM-Enabled Projects

to use and actual system use. The empirical findings reveal that the impediment of user resistance to the effectiveness of external support can be lessened by the proper use of incentivization, which could eventually lead to project performance improvement.

Applications to Practice As an enabling tool, BIM’s full potential depends on the readiness of all parties concerned. To secure BIM readiness, the architecture engineering construction (AEC) industry needs to make a lump sum investment in hardware, software, and training at the outset. The worthiness of this investment bears upon how frequently the acquired capability can be reused. In the early stage (Stage I in Figure 18), inhibited by lack of sufficient evidence in support of its benefit, the employment of BIM is limited to the small group of early adopters. In cash flow terms, the additional cost arising from BIM is high, as most AEC companies have to build in-house capability from scratch, which will naturally constrain the possible scope of BIM application in the project. In the environment of projects featured by a web of independent parties (designers, constructors, and suppliers), the benefit of BIM can grow exponentially as its application grows broader (more life cycle stages), deeper (levels of BIM), and more diverse (variety of analysis supported by BIM). As a result, fragmented application of BIM can only realize a small fraction of its potential. The gap in financial feasibility (D in Figure 2) is a fundamental problem hindering the voluntary adoption of BIM. In economic terms, it can be regarded as a case of market failure under which coordination mediated by the price signal cannot occur spontaneously, and that gives a rationale for government intervention. This reasoning explains why mandating BIM deployment in public projects is widely embraced as a kick-start strategy by governments. The nature of a government mandate is not much different from regulation as both serve to restrict the range of legal actions for public interests. In recent decades, the pendulum of regulatory philosophies in Europe has swung to risk-based assessment in which the cost of regulation is explicitly evaluated against its benefit. In the design of BIM mandates, the benefit is significantly harder to evaluate than the cost because the latter involves a direct cash expenditure, while the former involves a delayed receipt of benefit. During the development stages, the cost and benefit of BIM deployment will tend to converge as more companies upgrade to

Executive Summary

7

“BIM-ready” (see Figure 2). To the left of the point where those two trajectories intersect, the promotion of BIM is primarily driven by “push” forces, such as BIM mandates. After the benefit can cover the cost (to the right of the intersection point), then “pull” forces will dominate. It is useful to understand this conversion from the perspective of the Nobel Prize–awarded principal-agent theory. In designing an optimal contract, the principal should first ensure that compensation could more than cover the agent’s opportunity cost. This so-called participatory condition can persuade the agent to take part, but cannot induce him to exert the best effort. This theory suggests that efficiency can be improved by holding the agent accountable for the outcome of his action via risk-sharing arrangements. In the promotion of BIM, mandating can “push” some owners to embark on experimentation with the hope of driving industry BIM capability toward greater maturity through a “learning by doing” process. The push force could only make BIM nominally deployed as an enhanced 3D visualization tool, instead of giving participants strong incentives to explore the potential of BIM. For this reason, after BIM deployment becomes financially viable, the “pull” forces should be considered by way of various incentivization measures. With the data collected from countries covering a BIM-leading country (i.e., U.S.), a BIM-following country (i.e., UK), and a BIM-lagging country (i.e., China), this research can build a robust evidence base to support the broadening of BIM mandates.

C H A P T E R

2

Introduction Research Background and Significance In recent years, the construction industry has undergone a transformation from a traditional system emphasizing competition and contract enforcement to a collaborative system that stresses early supplier engagement and collective sharing of risk and reward. The key lesson learned from price-driven procurement practices is that the squeezing of contractors’ markups by competitive tendering could translate into massive costs for contract enforcement and intensify the contractor’s proclivity to withhold information and hold up the client during change order negotiations (Chang, 2013b; Chang & Ive, 2007). In response to these problems, construction owners are encouraged to adopt integrated project delivery (IPD) (or partnering) to govern and facilitate cross-party and across time coordination (American Institute of Architects, 2010a). In the meantime, we have seen an upsurge in the application of building information modeling (BIM) in the built environment worldwide. The benefit of BIM does not only vary widely with the depth, breadth, and diversity of its utilization in the project, but also with the delivery environment where it is employed. As observed by the authoritative BIM Handbook (Eastman, Teicholz, Sacks, & Liston, 2011): These requirements, however, are often difficult to meet without some modifications to the fee structure and relationships between project participants or without the use of incentive plans that define the workflow and digital hand-offs between disciplines. Often, these are more difficult to define in a workflow centered on a digital model, as opposed to files and documents. Additionally, approval agencies still require 2D project documentation as do a majority of professional 9

10

Incentivizing Collaborative BIM-Enabled Projects

contracts. Consequently, many owners maintain the traditional document and file-based deliverables; and they insert digital 3D workflows and deliverables into the same process. That is, each discipline works independently on their scope and BIM applications and hands-off the 3D digital model at specified times. Clearly, this is not a desirable approach to using BIM to its maximum advantage. (p. 183) While our understanding of the characteristics of desirable BIM delivery environments is still at the nascent stage, there is a view that BIM and integrated project delivery (IPD) could reinforce each other (Cohen, 2010; National Association of State Facilities Administrators, 2010; Thomsen, Darrington, Dunne, & Lichtig, 2009). IPD is a type of collaborative governance that emphasizes “best-for-project thinking” and “shared risk and reward” (American Institute of Architects, 2011). This research focuses on the governance and incentivization measures used to drive BIM participation from designers and all tiers of contractors (henceforth, IPD parties). The primary hypothesis is that differential performance across BIM-enabled projects can be attributed to “characteristics” of incentives provided to IPD parties. At the heart of integrated delivery systems lies a risk-sharing mechanism (Australian Government, 2011). An efficient incentivization system is not only essential for integrated delivery, but also for construction procurement in general (HM Treasury, 2013). In the current research, BIM incentivization involves two levels of motivation. At the “as-is” level, incentivization is aimed to capture the joint effect of all tangible and intangible factors that could have affected participation in the production of BIM information in the projects under study. For instance, BIM could be employed in circumstances where adopters find the project too complicated (so BIM can improve efficiency), hope to secure a market position as an advanced BIM user, or are only compelled to do so owing to the owner’s requirements. At the advanced level, the term represents the explicit act of employing measures to align the divergent interests of BIM participants (Baddeley & Chang, 2015). Provision of the right incentives requires a system approach to addressing a series of interrelated issues in an integrated way (Chang & Howard, 2016), including target-cost setting, incentive pool funding, risk sharing, and performance measurement. Unlike in the principal-agent theory where strong incentives can be neatly defined in mathematical terms (Chang, 2014d, 2015), incentivization involves

Introduction

11

leveraging a system of measures, so during the early stages of BIM development, explicit incentivization measures (e.g., performance-­linked payment) are rarely used and it can take an organization an extended period to ultimately settle on an (optimal) mechanism suitable for its own projects. The pace of evolution toward greater BIM maturity varies with the efficiency of learning, and the feedback loop of the learning process is a decisive mechanism for the way a system could evolve (Sterman, 2000). In introducing new information technology (IT) systems, as asserted by the technology acceptance models, the major hindrance stems from user resistance (Howard, Restrepo, & Chang, 2016; Venkatesh, Morris, Davis, & Davis, 2003). The central proposition of this research is that the outcome of current incentivization practice in BIM-enabled projects could propel the acceptance of advanced incentivization practice in the future. Government mandates have been a common strategy employed worldwide to expedite BIM applications. In the evolution process of best BIM practices, it is important for policymakers to build a deeper understanding of the impacts BIM could generate for the industry as a whole and harness it as a momentum for further development. These mandates should be better justified by drawing upon evidence. The analysis of this research is built on two theories. The first one is agency theory from economics (Hart & Holmstrom, 1987; Holmstrom, 1979), whose influence on the study of organizations within management literature is profound (Harris, Johnson, & Souder, 2013). However, this mainstream model has been challenged by applied psychologists and behavioral economists who have observed a range of anomalous decisions (Kahneman, Knetsch, & Thaler, 1991; Simon, 1955; Thaler & Sunstein, 2008). Human behavior is not only shaped by the underlying incentives, but also by the heuristics that drive decision making. Heuristics are fast decision rules, and it is often very sensible to use them. Only a “rational fool” would spend a lot of time and effort on a complex mathematical calculation when a quick rule of thumb would deliver a satisfactory outcome most of the time. The problem with heuristic decision rules is that sometimes they generate systematic biases in decision making, and the contexts of decisions can affect the outcomes. For example, the presentation and framing of choices as well as the setting of default options can change the decisions that people make (Baddeley, 2013; Kamenica, 2012). The complementarity of the two approaches promises to enrich the understanding of BIM’s efficacy in construction collaborative governance. Project management

12

Incentivizing Collaborative BIM-Enabled Projects

researchers have realized that some project failures might be ascribed to systematic behavioral biases (Shore, 2008). Along with this line, this research attempts to build statistical evidence to demonstrate the effect of behavioral biases on the effectiveness of incentivization measures in a systematic way. The value of this research can be seen from three aspects. For the owners, how to incentivize BIM participants remains a mundane decision. This research will provide a timely analysis of the effectiveness of incentivization measures implemented in prior projects, thereby informing practitioners of the sound principles that can improve the design of BIM incentive systems. For the academic community, this research can contribute to two research themes. First, the theory developed for BIM incentivization mechanisms under IPD can advance the analysis of project governance in general and IPD/partnering in particular. Second, this research can improve the application of behavioral perspectives to the analysis of project organizations, which is increasingly recognized as a promising line of inquiry (Ansar, Flyvbjerg, Budzier, & Lunn, 2014). This research represents a first attempt to explore the effect of behavioral biases on incentivization embedded in the delivery environment of BIM-enabled projects. The identified biases jointly prove to have a significant impact on the effectiveness of incentivization. This finding also provides an empirical support to the policy initiative aiming at changing behavior by design (e.g., UK’s Policy Lab). For the government, the empirical finding of this research demonstrates that there should be a synergy between BIM strategy and delivery system reform. In promoting Level 3 BIM, the government should attend to the issue of how to get the delivery system right.

Aims and Objectives The objectives of this research are twofold: First, discover the best practices for BIM incentivization, and second, develop a theory drawing on modeling techniques from organizational and behavioral approaches to understanding the practices found. Specifically, there are three objectives: 1. Conduct three case studies covering the participants in supply chains of a typical BIM-enabled project in three national contexts (U.S., UK, China).

Introduction

13

2. Build a theory for IPD parties’ response to a given incentive system by drawing on the methods of agency approach and behavioral approach. 3. Develop hypotheses and test them via econometric analysis to form an evidence base to guide the design of incentivization systems for BIM-enabled collaborative governance.

Sketch of Research Process Whereas the concept of BIM has been in development for two decades, the large-scale rollout of this technique worldwide is a recent phenomenon. From the pilot case studies in the first year of research, there is a clear sign that China, the United Kingdom, and the United States possess distinctly different levels of BIM maturity, which will affect the nature of problems experienced and faced by these countries. Instead of developing a universal theory for BIM incentivization, this research expounds a life cycle theory to make the focus of analysis better aligned with the issues discovered. In 20 months, this research went through several stages, which is not as expected in the original plan: Stage I: Review literature to identify the relevant factors that could change the effectiveness of BIM and develop a conceptual framework for analyzing BIM incentivization problems. The output of this stage led to a journal manuscript titled “Incentivizing BIM Participation: Conceptual Framework and Empirical Evidence.” Stage II: Pilot case studies were conducted in three countries through the lens of the framework above. It was found that the extent of BIM deployment and the delivery environment for BIM exhibit a wide variation, and the framework is only suited to analyze advanced cases found in the U.S. rather than those in the UK and China. To accommodate diverse practices, this research deemed it necessary to modify the analytical framework. Stage III: The framework considers the national environment as a factor to reflect the general level of BIM maturity. Case studies were conducted in three countries to gain insights into the factors that influence BIM effectiveness and project performance. Stage IV: A large-scale survey was conducted in three countries to discover the causation of BIM incentivization. Stage V: Revise the theoretical framework in response to the empirical findings.

14

Incentivizing Collaborative BIM-Enabled Projects

The journey started with a deductive approach and then changed to an inductive approach in identifying the regularities of BIM incentivization practices, and finally drew upon empirical evidence to modify the original theory. As a whole, this research method is in line with the spirit of a grounded approach.

Framing of Research Questions In this project, BIM is treated as an “enabler” with the aim of facilitating collaboration among project parties. To gain a full understanding of how project parties can be incentivized to participate in BIM production, there are three levels of factors to consider (Chang, 2014a; see Figure 1): 1. Level 1: National environment for BIM implementation Whereas applications of information technology in the AEC industry have a long history, the acceptance of BIM on the national scale is a recent phenomenon. As the originating country of BIM, the U.S. is leading the way. The UK tries to catch up by enforcing a mandate that requires all publicly procured projects to adopt Level 2 BIM by 2016. China is behind regarding readiness to deploy high-level BIM. Similar to other technologies, BIM is expected to go through several stages of its life cycle. The barriers at the national level are mostly put up by legal constraints, which in turn could restrict the use of BIM across project stages and hence its impact on project performance. 2. Level 2: Project governance When it comes to organizational design, as suggested by transaction cost economics (TCE), governance alignment is of first-order significance in efficiency terms. In other words, the choice of delivery system is the most fundamental factor in achieving efficiency, so the owner will choose the delivery system by aligning project attributes with procurement systems (Chang & Ive, 2007). 3. Level 3: Incentivization system The core issue of interest in this project is to seek the most effective measures to power the participation of project parties to make consummate contributions to achieving the

15

Introduction

owner’s objectives. In TCE, this is a second-order economizing strategy—namely, that incentives are aligned for BIM-­ enabled projects under the national business environment and the project governance chosen. These three levels of factors jointly shape the project delivery environment. In principle, the more extensively BIM is deployed as an enabler, the greater influence BIM could have on project performance. The incentives embedded in the delivery environment can motivate project parties to harness the technological power of BIM to its potential through three mechanisms (i.e., improving coordination, increasing collaboration, and strengthening communication).

Level 1 National Construction Business Environment Level 2

Alignment Delivery System

Level 3 Incentivization System

Project Attributes Risks

Teamwork

Use of BIM in individual projects as an enabler

Owner MEP suppliers Designer

Focus of this research

Constructors

Project Performance

Figure 1.  Conceptual framework for BIM incentivization.

The shaded oval at the center of Figure 1 indicates the focus of research in this project. Table 1 gives an overview of the key research questions to be answered for three national contexts and the corresponding evidence base.

16

Incentivizing Collaborative BIM-Enabled Projects

Table 1.  Research questions and evidence bases.

China

Research Questions

Evidence Base

In what way has mandate-driven BIM adoption in China propelled the evolution of a more integrated and well-incentivized delivery environment?

Pilot case studies (1)  Maotai Distillery extension project (2)  Guian Foxconn (3)  Xixian bonded area service center (4)  Shanghai Tongji University Sports Center (5)  Guangzhou 21 Line Rail Transit Project Case studies (1)  Shanghai World Center Project (2)  Shanghai Leisure Park project Survey: 223 returns

UK

What are the critical factors that determine the impact of BIM on project performance?

Pilot case studies Network Rail’s East Coast Main Line Project Case studies 1.  Anglian Water (@one alliance) 2.  Environment Agency (WEM) Survey: 175 returns

U.S.

To what extent have incentivization measures led to an improvement in project performance through the lens of user-related factors?

Pilot case studies (1)  Sutter Medical Center, Castro Valley (2)  Autodesk Waltham project (3)  Sutter Health Fairfield Medical Office Building (4)  Cardinal Glennon Children’s Hospital Expansion (5)  Encircle Health Ambulatory Care Center Case studies San Francisco International Airport (Integrated Project Delivery) Survey: 232 returns

C H A P T E R

3

Literature Review Studies of Collaborative Governance in Construction (IPD/Partnering) Collaborative governance took root in construction in the form of partnering. Since the 1990s, partnering has been extensively studied (Bygballe, Jahre, & Swärd, 2010; Hong, Chan, Chan, & Yeung, 2011). The purpose of this section is to provide a snapshot of the current state of research on this topic. Bresnen and Marshall (2000) make an early attempt to analyze the use of incentives in partnering/alliancing projects through seven case studies. They look at incentive systems, gain/pain share, other conditions (e.g., liquidated damages), input into pricing (i.e., target-cost setting), as well as procedures for variations and changes. Constrained by the qualitative method employed, the authors are not able to rigorously evaluate the effectiveness of incentive measures. Black, Akintoye, and Fitzgerald (2000) aim to investigate 19 success factors (e.g., mutual trust, effective communication) and benefits of partnering using data collected from a postal survey. The respondents are asked to rate the importance of a list of factors for the success of partnering projects. As the returns are based on the respondents’ free-of-context perceptions, it seems difficult to interpret the meaning of these returns. In principle, the potential significance of a success factor could vary with some other factors. With no attempt to build a theoretical framework, the research findings are no more useful than informing upon which areas deserve attention in follow-on studies. Wong and Cheung (2005) investigate the effect of trust upon project performance using structural equation modeling (SEM). The variable “trust” is measured by the respondents’ assessment of the importance of 14 trust attributes. Similarly, project performance is measured by 17

18

Incentivizing Collaborative BIM-Enabled Projects

five groups of attributes with respect to time, completion cost, quality, communication, and management. The SEM approach enables identification of statistical relationships between the latent factors regarding explained and unexplained variances. It is found that the trust level is positively correlated with project performance at a high level of significance. However, this result should be received with caution. Correlation does not equal causation, and it could be that other factors explain both trust level and project performance, and so the relationship between trust and performance is spurious. Also, what the variable “trust” actually captures includes the partner’s capability, attitude, communication style, efficiency in resolving disputes, and reputation. Having all of these heterogeneous dimensions buried under the vague heading of “trust” could impede the development of a more fundamental understanding of partnering projects. Tang, Duffield, and Young (2006) seek to identify the critical success factors of partnering agreements by asking postal survey respondents to rate on a 5-point Likert scale the extent to which they agree with a set of statements associated with critical success factors (CSFs) identified in a conceptual framework. The CSFs are grouped according to the “distance” (a measure of statistical correlation) between variables using hierarchical cluster analysis. The significance of individual CSFs for project performance and the interconnection between CSFs (e.g., incentives and risk management) are also evaluated on the same basis. The inadequacies of this research seem similar to those in previous studies. The respondents are not asked to answer questions for a particular project context, so the returns are simply reflective of the respondents’ general perceptions of the problem. Also, with no underlying theory behind the framework, the interpretation of the statistical relationships found in the analysis appears tricky. For example, it is found that there is a high correlation between the use of incentives and the choice of risk management practices. In the study, “incentives” are measured by the frequency of incentives being given for good performance in quality control, schedule control, completion, safety record, cost savings, and budget control. With this measure alone, one cannot evaluate how well incentives could have performed and thus has no way to assess what impact incentives could have had on behavioral uncertainty. El-adaway (2010) presents a case study of the way IPD has been introduced in a multinational contracting firm. The method involves three steps: First, the author conducted a survey of the company’s supply chain to collect general views on the contractual requirements needed for IPD

Literature Review

19

to work; second, the feedback gathered from the survey was drawn upon to develop the guidelines for crafting IPD multiparty agreements; third, the firm followed the guidelines to implement IPD on a US�5 million project. The issues addressed in the guidelines include project environment, project manager, partnering advisor, design process, project schedule, suppliers and manufacturers, prices and profits, incentives, changes, problems, and disputes. While the views from IPD participants are useful, building the entire credibility upon the survey result makes the rationale appear feeble.

IPD and Incentivization The concept of IPD is primarily practice-driven. As opposed to partnering arrangements that seek to improve collaboration by fostering trust between parties through informal mechanisms (e.g., workshops), IPD emphasizes the role of formal mechanisms (e.g., contracts). For instance, a formal risk-sharing arrangement is a key to differentiating Level 1 and Level 2 IPD (National Association of State Facilities Administrators, 2010) and a binding multiparty agreement is a prerequisite condition for Level 3 IPD. Since IPD has been heavily advocated by American construction professional bodies and deemed the most suitable vehicle for BIM-enabled projects (American Institute of Architects, 2007; Cohen, 2010; Lahdenperä, 2012; National Association of State Facilities Administrators, 2010; Thomsen et al., 2009), it is increasingly necessary to expedite the advancement of the understanding of IPD—an area where this research can make a contribution. Incentivization is a crucial issue in governing construction projects (HM Treasury, 2013; Victoria Department of Treasury and Finance, 2010). All projects involve incentives. As a result, the choice is not between giving incentives or not, but about which incentive to provide (Darrington & Howell, 2011). Monetary incentives play a dominant role in building up collaborative behavior and reinforcing trust (Badenfelt, 2010), and could have a positive impact on project performance (Hart & Holmstrom, 1987; Jensen & Murphy, 1990). In practice, risks are allocated through a gain/pain sharing scheme (Bresnen & Marshall, 2000; Meng, 2012; Walker & Hampson, 2008). There are several components: The first is to control the target cost (Boukendour & Hughes, 2014), which BIM leads to a more precise cost estimation than traditional method (Sunil, Pathirage, & Underwood, 2015). The second is to set up

20

Incentivizing Collaborative BIM-Enabled Projects

an incentive pool, which can be funded by various sources (e.g., profit, contingency fund, fees) (Baddeley & Chang, 2015). Darrington (2010) suggests that the “at-risk fee pool” can cover cost overruns or generate profits in the event of cost underruns, so it can align stakeholders’ performance toward the common goal and reinforce intrinsic motivation (e.g., reciprocity, fairness, and social reputation). The third is associated with the risk-sharing arrangement. Fourth, it is imperative to probe the basis on which incentive pays should be awarded (i.e., to individuals or groups [Baddeley & Chang, 2015]). Holmstrom (1982) suggested that applying monitoring systems in group work could resolve problems caused by risk-averse agents. Besides, Kadefors (2004) proposes a reward strategy based on both individual and group performance, finding that the group incentives perform relatively better in various scenarios. Fifth, another important decision is about the choice of remuneration plan. Simple linear plans (e.g., share the saving costs by a certain ratio) are more pragmatic than complex, nonlinear plans in practice (Chang & Howard, 2016). A suitable cap on risk sharing could be necessary, particularly when a project involves innovative design solutions (Chen, Zhang, Xie, & Jin, 2012). Economic theory also suggests that there might be a threshold for an incentive pay to yield a motivational effect (Gneezy & Rustichini, 2000). Last, performance measurement holds the key to the success of an incentivization system. There is a debate concerning how to assign weight on the objective and subjective measurement (Gibbs, Merchant, Van der Stede, & Vargus, 2004). Owners (as the principals) may prefer subjective judgment, while contractors (as the agents) may favor objectivity. It requires ingenuity to find a right balance of these two measurements. While the analysis of the optimal incentivization is at the nascent stage, the issue of economic incentives has been extensively studied. In the process of building a conceptual framework for analyzing BIM incentivization problems, the initial challenges lie in the identification of key issues. Whereas the application of collaborative BIM is still in the formative stage, some advanced users have embarked on experimentation. The practices employed in successful projects can serve as an evidence base for discerning the building blocks required for a BIM incentivization system. Apart from current practices, the second area where insights can be sought is academic literature. Since the provision of incentives is a central decision in the design of organizations, it has naturally attracted attention from across social science disciplines. A systematic analysis of academic

Literature Review

21

incentive issues can shed light on the future design of BIM incentivization systems. In the selection of literature, quality is embraced as the paramount priority. Instead of conducting a comprehensive keyword search, this research takes an alternative tack by collating the works of prominent researchers and major literature review papers. Since the essays collected by the authoritative Handbook of Organizational Economics (Gibbons & Roberts, 2013) bear strong relevance for construction organization design (Chang, 2014b), the search is initially guided by the references cited in these essays and continues until no new reference sources and insights can be found. In the end, around 200 journal articles are included in the initial list. These articles are then classified into six themes according to their research questions. The output of this step includes a list of six theoretical decisions relevant to the design of BIM incentivization systems. In the last step, an online survey was designed to measure BIM practitioners’ preferences for these six decisions against the context of their most recent BIM-enabled project. Since BIM incentivization systems may take years to mature, it is useful to gain insights from practitioners into the essential “characteristics” of an ideal BIM incentivization system. This result can serve as a default setting in the design of an optimal mechanism for incentivizing BIM participation.

Do Monetary Rewards Lead to the Desired Result? In the agency approach, financial rewards are assumed to be a positive driver of human behavior (Bolton & Dewatripont, 2005; Hart & Holmstrom, 1987), but literature is full of examples that incentives may actually motivate the wrong behavior (Kerr, 1975; Kohn, 1993; Lawler, 1990). In recent years, the behavioral perspective has grown from obscurity into prominence within mainstream economics and management literature (Levinthal, 2011). As a result, it is useful to pull together the current views on whether monetary rewards are more likely than not to yield the desired results. In psychology, it has been acknowledged for decades that the use of extrinsic rewards may have a detrimental effect on intrinsic motives (Deci, 1971; Lepper, Greene, & Nisbett, 1973). The harmful effects of extrinsic rewards stem from their tampering with people’s right of self-determination (autonomy) (Ryan & Deci, 2000). The possibility that monetary rewards would result in weaker intrinsic motives is known as the crowding-out effect in economics, which is then theorized as the motivation for crowding theory (Frey & Jegen,

22

Incentivizing Collaborative BIM-Enabled Projects

2001; Frey & Oberholzer-Gee, 1997). The reception of this theory in economics is mixed. Provided it holds, monetary rewards will result in a decrease of supply rather than an increase (Frey & Jegen, 2001), which will work against the fundamental assumption behind the law of demand and supply. In the face of this anomaly, economists take three different stances: ignore, wait and see, and explain why. An example of the first stance is Prendergast (1999). Whereas he agrees that the crowding effect has an intuitive appeal, “there is little conclusive empirical evidence (particularly in workplace settings) of these influences.” This is because the effects reported in psychology experiments can also be ascribed to an alternative explanation (footnote 12 in Prendergast, 1999, p. 18). Gibbons (1998) seems to take a more sympathetic stance. He acknowledges the possibility that use of monetary incentives may harm intrinsic motivation and social relations, but demands more field experiments to demonstrate the significance of this effect in practical settings. By contrast, Kreps (1997) is more interested in supplying an explanation for why the effect is observed. Kreps argues that “[j]obs high in intrinsic motivation often involve a great deal of task ambiguity” (p. 361). As a result, the job undertaker’s performance becomes multifaceted. Use of high-­powered incentives in a multitasking environment would lead the agent to focus on reward-linked aspects of the job at the expense of those issues that could potentially add value but do not affect reward (B. Holmstrom & Milgrom, 1991). Attracting people with a natural desire for the job (e.g., passion) is a low-cost solution for incentive provision. Nonetheless, the introduction of a performance-linked reward scheme to the job would obscure the nature of the employment relationship as it signals that the relationship is a market exchange and reacts accordingly. When the adherence to norms is replaced with the mentality of taking advantage of opportunities as they arise, the agent will divert his effort away from the valuable activities that are hard to monitor (Rebitzer & Taylor, 2011). Whereas the debate over the effectiveness of monetary compensation looks set to continue, this research subscribes to the thoughtful observation of Baker, Jensen, and Murphy (1987): We believe that careful examination of the criticisms of monetary pay-for-performance systems indicates not that they are ineffective but rather that they are too effective: strong pay-for-performance motivates people to do exactly

Literature Review

23

what they are told to do. Large monetary incentives generate unintended and sometimes counterproductive results because it is difficult to adequately specify exactly what people should do and therefore how their performance should be measured. (p. 597) This position leads to an immediate question: What is the best way to harness incentives? Some guidance can be found in the experiment results. Through the review of 74 experimental papers, Camerer and Hogarth (1999) believe the answer lies in what type of task/decision incentives have been applied to: Incentives improve performance in easy tasks that are effort-­ responsive, like judgment, prediction, problem-­solving, recalling items from memory or clerical tasks. Incentives sometimes hurt when problems are too difficult or when simple intuition or habit provides an optimal answer and thinking harder makes thing worse. (p. 34)

Group-Based vs. Individual Reward As pointed out by the American Institute of Architects (2010c), “[c]ompartmentalized compensation can lead to divisive behaviors, with each team member doing what’s best for their firm instead of what’s best for the project” (p. 4). This begs the question as to what basis rewards should be allocated upon. In a modern economy driven by a trend toward specialization, a division of labor becomes prevalent. With few individuals owning all inputs necessary for nontrivial tasks, team production can be more efficient if the output of team production is expected to be greater than the sum of separate outputs of individual team members (Alchian & Demsetz, 1972). Organizing into teams would present the problem of shirking (Alchian & Demsetz, 1972). Monitoring can help alleviate this problem but will also result in metering costs. This is where the role of the firm sets in, where supervisors should be made the residual claimant to reduce their shirking. The main problem of team incentivization rests with the difficulty in reliably measuring individual output. When the joint output is the only basis of compensation, it is in one’s interest to free ride on others’ hard work. The presence of the tendency to free ride in team production is independent of whether there is uncertainty in output. In an

24

Incentivizing Collaborative BIM-Enabled Projects

influential paper sparking another milestone contribution to the study of teamwork, Holmstrom (1982) develops a principal-agent model to prove that group incentives can serve as an alternative solution to free-riding problems. In the model, the principal seeks to maximize the total surplus of an agency relationship under the budget-­balancing constraint. This constraint requires the reward pool to be totally financed by the joint output of the team (i.e., the total of individual payoffs equal to the group output). Under this constraint, team workers can obtain a share of the joint output by predetermined sharing rules regardless of their efforts. An effective way to curb free riding is through a joint contract between team workers and a third party (the principal) who has no input to the production, whereby group penalties can be imposed should the output fall below the Pareto-optimal level. However, the effectiveness of “budget-­breaking” (Holmstrom’s term) in neutralizing externalities from team production depends upon the credibility of this scheme (Eswaran & Kotwal, 1984). Other solutions have been explored for resolving free-rider problems, including the threat of discontinuing the relationship (Radner, 1986) and mutual monitoring by way of peer pressure (Kandel & Lazear, 1992). However, compared to monetary incentives, there exists little understanding of how well they would work empirically. Regarding the relative efficacy of team-based and individual compensation in the workplace setting, Van Dijk, Sonnemans, and Van Winden (2001) report an interesting experimental result, showing that a piecerate scheme (individual payoff equal to individual output) and team payment (individual payoff equal to average share of the team output) elicit the same level of effort for the employer. This result echoes Jeffrey Pfeffer’s (1998) observation that individual merit pay is not necessarily superior to group-based reward schemes. However, in team production, free-riding problems could become an issue. Theoretically, there are three solutions: a budget-breaking mechanism, reputation, or peer-mutual monitoring. In the future, the extent to which these solutions have been employed and how efficaciously they have worked deserve attention.

Objective vs. Subjective Performance Measurement A performance award is a significant means of motivating non-owner IPD members to work toward nonfinancial goals. In practice, qualitative criteria, such as design quality and client satisfaction, might be considered for the award of incentive payment (American Institute of

Literature Review

25

Architects, 2010c). In choosing between subjective and objective criteria, there is a need to go beyond the factors of transparency and fairness identified by Thomsen et al. (2009). The efficacy of the reward scheme suggested by the agency approach rests on two preconditions: (1) The performance indicator is reflective of the agent’s performance, and (2) payments are closely linked to the chosen indicator. Either of these two conditions failing to hold will lead to the dysfunction of the performance system. The choice of performance metrics is therefore crucial for the effectiveness of incentives in inducing the agent to work toward the owner’s objectives. From a principal-agent perspective, the first-best contract is feasible only when the agent’s effort is observable, and there is no uncertainty in output (Lambert, 2001). When asymmetric information is present, incentives should be built upon the “informativeness” of the metrics (Holmstrom, 1979). Most of the time, the effort is not observable, so a proxy such as revenue is used to capture the agent’s effort. The drawback is that a positive sales record may have more to do with luck than with actual sales aptitude. For this reason, it is sometimes useful to include auxiliary metrics in the compensation scheme, so long as it adds more information about the agent’s effort and the agent has no control over it. An example would be the average sales of other salespersons in the same area, which is valuable because it can help remove some of the random effects. Equally useful is a principle developed by Baker (1992) that a viable performance measure should satisfy the condition that “the marginal product of the agent’s actions on the performance measure is highly correlated with the marginal product of these actions on the principal’s objective” (p. 612). Feltham and Xie (1994) put forward a set of useful principles for the selection of performance measures: congruence and precision. A performance measure is said to be congruent if the impact of the agent’s action on the principal’s objective can be mostly captured by the impact of the agent’s action on the performance measure. Also, a performance measure containing less randomness is preferred (higher precision). The ideal measure is noiseless (so the principal does not have to pay a high-risk premium to the agent), and can direct the agent to work for the principal’s interest. Where effective objective measures are not available, subjective performance evaluation then has a role to play (Murphy & Cleveland, 1995; Prendergast & Topel, 1993). For example, stock price performance may be reflective of the CEO’s contribution to

26

Incentivizing Collaborative BIM-Enabled Projects

firm value, but it contains too much noise for the evaluation of lower-­ level employees. The working of a subjective performance system should build trust between the employer and employee. In a one-shot game, the principal would skip the bonus payment even when the agent performs satisfactorily. However, this opportunistic move would undermine the credibility of the incentive scheme in the long run. A remedy for this problem is to foster a reputation as an implicit “self-enforcing” mechanism for the scheme (Bull, 1987). Under certain circumstances, it could be more efficient when objective and subjective measures are used together rather than singularly (Baker, Gibbons, & Murphy, 1994). An empirical investigation of the effects of human resource practice on productivity for steel finishing lines lends support to the superiority of the joint use of an explicit profit-sharing scheme and discretionary bonuses (Ichniowski, Shaw, & Prennushi, 1997). However, subjective assessments are prone to supervisorial biases such as favoritism. Excessive personal preferences have two implications for compensation design (Prendergast & Topel, 1996): (1) Additional noises caused by favoritism will make incentive contracts less desirable, leading to underuse of low-powered incentives; and (2) it could result in the implementation of bureaucratic rules in place of supervisor appraisals for performance assessment. Establishing an optimal mix of subjective and objective measures is a topic worthy of further exploration.

Weightings of Performance Metrics In theory, objective and subjective performance measures could complement each other. The relative weights placed on performance payment and gain/pain share would have a decisive effect on the allocation of the effort of IPD non-owner members. The successive logical question is how to balance the influence of different measures of compensation. A formula-based plan would induce the agent to “game” the system to maximize measured performance instead of intended performance. Baker, Gibbons, and Murphy (1994) prove that distortions caused by objective performance measures can be mitigated by the principal reserving discretion over the weightings of these measures. The wide use of discretionary bonus schemes is an example (Baiman & Rajan, 1995; Hayes & Schaefer, 2000). Objective measures are used less in the evaluation of complex jobs (MacLeod & Parent, 1999). In an empirical study of compensation practices in car dealerships, Gibbs, Merchant,

Literature Review

27

Stede, and Vargus (2004) find that the evaluator’s subjectivity plays a greater role in the determination of bonuses under four conditions: first, if the job entails the agent’s greater long-term investment in intangible assets; second, if the job requires significant input from coworkers; third, if the target is set high and subject to recalibration in the event of environmental uncertainty; fourth, if loss occurs to make subjective judgment the only way to filter out “bad luck.” With regard to the allocation of weights to measures, the principle of “informativeness” is influential in decisions on the assignment of weightings (Antle & Demski, 1988; Bushman, Indjejikian, & Smith, 1995). The inclusion of a performance measure should be justified by whether it can add more information content about the agent’s actions than other included measures. However, the informativeness of a measure (relative strength of one measure over others) would vary with intra-firm interdependencies (e.g., product line or geographic diversification) (Bushman et al., 1995). Principal-agent theory suggests that the weight given to a performance metric should be lower if it contains high variations, and higher if the measure can induce the responsive reaction from the agent (Banker & Datar, 1989; Holmstrom, 1979). The relative weights of measures should consider both the precision and sensibility of the metrics (Banker & Datar, 1989). In statistical terms, the former indicates the reciprocal of the variance of the measure, while the latter is concerned with how much the signal would change owing to a unit change in the agent’s effort. Bushman and Indjejikian (1993) show “the ratio of the optimal incentive weights for a linear combination of two performance measures is the ratio of the sensitivity times precision of the individual performance measures” (Bushman, Indjejikian, & Smith, 1995, p. 107).

Studies of BIM Diffusion BIM has the potential to be a game-changing factor in the industry for three reasons (Eastman et al., 2011): First, it is a unique way of integrating information into design schematics. Second, BIM can be easily standardized. Third, by accommodating all information into virtual models, BIM provides an opportunity to improve quality assurance through the formalization of model specifications. As a result, BIM can be perceived as both a “technology” and a “process.” In pursuit of these benefits, several countries (e.g., Singapore, South Korea, the United Kingdom, and

28

Incentivizing Collaborative BIM-Enabled Projects

the United States) have mandated the compulsory use of BIM in public projects (Cao et al., 2015). However, BIM is just beginning to register significant awareness and adoption within the industry at large. Eadie, Browne, Odeyinka, McKeown, and McNiff’s (2013) recent investigation shows that contractors are less involved in BIM use than designers, and many BIM practices are limited to the design stage. While one can derive benefits from BIM in separate applications, only when BIM is embedded in the process to generate the interoperable and interactive workflow around it can the full potential of BIM be unlocked (Monteiro, Meda, & Pocas Martins, 2014). This requires a new form of delivery system that supports collaborative procurement processes (Australasia, 2012). It is widely recognized that IPD could be an organizational solution (Australasia, 2012; McGraw-Hill Construction, 2014). As argued by Succar (2009), BIM development may go through three stages (object-based modeling, model-based collaboration, and network-based integration) before it reaches the long-term goal of embedding BIM in an IPD environment. Behind this evolution, there are three interlocking, driving forces at work, which are associated with the policy, technology, and process. Along a similar line, Succar and Kassem (2015) develop five models for the assessment and management of BIM diffusion (diffusion areas model, macro maturity components model, macro diffusion dynamics model, policy actions model, and macro diffusion responsibilities model). There is ample evidence from the U.S., UK, and China that project delivery systems with a higher level of integration could lead to better project outcomes (American Institute of Architects, 2007; Chen & Jiao, 2011). There is also a view that a BIM-enabled collaborative environment could facilitate the implementation of IPD (Cohen, 2010). While IPD principles have been promoted for over a decade, IPD projects remain uncommon. Ill-devised legal frameworks, inadequate competencies, and lack of experience have all impeded the adoption of IPD (Autodesk White Paper, 2008). Most existing IPD contracts include elements that are designed to encourage teamwork for the success of the entire project rather than any particular team member. Unlike traditional projects where all parties pursue their own risk minimization, IPD combines the risks and rewards of all team members and correlates them with common project goals (Kent et al., 2010). Interest alignment holds the key to the success of integration. As defined in Baddeley and Chang (2015), “incentivization” refers to the act of employing measures that help align the divergent interests of BIM participants. Chang (2014)

Literature Review

29

and Chang and Howard (2016) identified seven fundamental questions involved in the design of a BIM incentivization system and their theoretical foundations: 1. How to manage the co-evolution of design and target cost? 2. How to fund the incentive pool? 3. On what basis to award compensation? 4. What weightings to assign to objective and subjective evaluation? 5. How to allocate risk through the choice of risk-sharing ratio? 6. How to choose the right compensation from between linear and nonlinear plans? 7. How to set the threshold value for each incentive award band? The current research adopts these BIM incentivization questions and previous research results as the theoretical frame of reference. Within the project environment, BIM’s greatest effects relate to communication (Mourshed, 2006). Trust and communication are critical to effective supply chain relationships (Baddeley & Chang, 2015). The processes for the extraction, interpretation, and communication of design information from drawings and documents are frequently time-­consuming and arduous (Sebastian, 2010). However, BIM protocols can help facilitate this process. For example, during the construction process, BIM can support communication among parties and locations (e.g., the building site, the factory, and the design office), which is crucial for efficient prefabrication and assembly, as well as prevention of unexpected errors. As maintained by Brennan (2011), effective communication, trust, and respect are among the most important critical success factors (CSFs) for team collaboration under an IPD approach. Adding communication into the IPD acceptability model begs the fundamental question of how to measure the quality of communication. As cited in Mohr and Spekman (1994), communication quality is a critical aspect of information transmission, including issues such as the accuracy, timeliness, adequacy, and credibility of the information exchanged. In a recent study of trust in Chinese IPD teamwork, Wu (2012) identified communication as one of the major indicators of project performance and measured it using three dimensions, including communication effectiveness, accuracy, and degree of involvement. By also referencing Freeman, Weil, and

30

Incentivizing Collaborative BIM-Enabled Projects

Hess (2006) and Pocock, Hyun, Liu, and Kim (1996), the current research takes a broader view by defining communication quality as consisting of accuracy, timeliness, transparency, initiative, and frequency. Large construction projects mostly span several years in which the interaction between owner and contractor could be intense (Kadefors, 2004). BIM projects are aimed to enhance collaboration by improving information sharing across business boundaries and interdisciplinary teams. In recent years, practitioners have become increasingly aware that efforts should be made toward removing the barriers to collaboration within the construction supply chain. Ertel, Jeff, and Laura (2001) explored the function of collaboration in multiparty agreements, finding that poor collaboration is the most significant factor leading to the failure of project alliances. Respondents in a recent investigation of BIM practices also observed mistrust and collaboration issues among participants in their projects (Cao et al., 2015). IPD is an emerging delivery system in which members’ success depends on cooperation and teamwork among main parties. Although research has demonstrated that collaboration is a critical requirement for IPD, it is not solidly grounded in empirical evidence. Only a few studies have focused on collaboration assessment and improvement. An example is Abdirad and Pishdad-Bozorgi (2014), where the authors developed a framework of metrics for measuring collaboration within IPD, including colocation (Brewer & Mendelson, 2003), multidisciplinary work (Brewer & Mendelson, 2003), team productivity (Brewer & Mendelson, 2003), cost impact of collaboration (El Asmar, 2012), training (Thompson & Ozbek, 2012), immediate feedback (Brewer & Mendelson, 2003), real-­ time sharing of data (Moore, Woodward, & Grogono-Thomas, 2005), methods of communication (Thompson & Ozbek, 2012), degree of interaction (Pocock et al., 1996), individual human aspects (i.e., turnover), and BIM technology (Cohen, 2010). This comprehensive list provides a sound basis for the selection of metrics used in the measurement of collaboration in the current research. Compared to the literature, the value of the current research can be seen in three aspects: First, the focus of analysis is placed on the extent to which mandated BIM implementation could change the perception of the desirability of IPD features for BIM-enabled projects. This provides a new angle for scrutinizing the benefits of BIM. The finding demonstrates that the spillover effect of using BIM, voluntarily or not, could facilitate the acceptance of IPD. The second distinguishing point

Literature Review

31

lies in the empirical method used. For example, both Succar (2009) and Succar and Kassem (2015) are prescriptive and conceptual in nature. While the framework of Succar (2009) is validated by a standard qualitative approach called “triangulation,” he also calls for researchers to use different methods in testing his framework. By contrast, through the SEM technique, the current research can rigorously demonstrate that the more extensively BIM is deployed in the project, the stronger the perception of the necessity of advanced IPD features for BIM-enabled projects. This cause-effect relation suggests that BIM mandates could propel a more desirable delivery environment for high-level BIM. The model also reveals that the momentum is generated by the awareness of incentivization measures and the improvement in communication quality enabled by BIM. While the effect of BIM on the transformation of the construction management work process is increasingly acknowledged (Hartmann, Van Meerveld, Vossebeld, & Adriaanse, 2012; Monteiro et al., 2014), the underlying forces remain under-studied. This research furnishes timely evidence to fill this knowledge gap. Third, as elaborated in Succar and Kassem (2015), BIM diffusion could be portrayed in various ways. In the development of a parsimonious life cycle theory of BIM diffusion, the two statistically significant constructs (incentivization and communication) found in the SEM analysis can effectively sharpen the research focus. Successful BIM implementation depends on two groups of critical factors: enablers (e.g., inoperability) and inducements (e.g., incentives). Enablers make BIM a feasible IT solution to coordination failure, but, without proper incentives, its effectiveness would be undermined. As shown in Whyte (2012), the current study of BIM is mainly approached from the perspective of IT. The majority of research efforts have been directed toward the technical issues of BIM, leaving the fundamental problem of how to incentivize BIM participation unaddressed (Chang, 2014c). The first relevant body of literature is associated with BIM itself. Despite expanding at an extraordinary rate, BIM research was mostly focused on the technical issues. Conceptually, BIM can be seen as the 3D graphic components embedded with the valuable information of construction projects, accessible by different participants (clients, designers, contractors, operators) over the project life cycle. With a threelayer parametric model, supported by data-exchange standards and controlled by information interchange protocols (Crotty, 2011), BIM

32

Incentivizing Collaborative BIM-Enabled Projects

could help “ensure that accurate and current project information is always available at the right time in the right format to the right person” (Winch, 2012, p. 378). From a different angle, BIM is an innovation in the construction process (Demian & Walters, 2014; Eastman et al., 2011; Liu, Van Nederveen, & Hertogh, 2016; Succar, 2009; Zuppa, Issa, & Suermann, 2009), because it can serve as a data-sharing platform to incorporate data from various systems such as Geographic Information System (GIS) (Bansal, 2011), Energy Management System (EMS) (Becerik-Gerber, Jazizadeh, Li, & Calis, 2012), and Construction Operation Building Information Exchange (COBie) (East & Carrasquillo-Mangual, 2012; Love, Matthews, Simpson, Hill, & Olatunji, 2014). For this reason, BIM provides a “technology pull” to align its functionality with the construction management work process (Hartmann, Van Meerveld, Vossebeld, & Adriaanse, 2012; Monteiro, Meda, & Pocas Martins, 2014). In some developed countries, including the UK, the U.S., and Singapore, BIM has been mandated as compulsory in public projects (Cao et al., 2015), which was also adopted by the Chinese government. However, BIM maturity in China is rather low (Cao, Li, Wang, & Huang, 2017). The slow proliferation of BIM is attributed to a wide range of barriers, including data interoperability (Bernstein & Pittman, 2004), contractual issues (Ashraf, 2008), and personnel issues (Ku & Taiebat, 2011). In other words, barriers may stem from product, process, and individuals (Lindblad, 2013), and a lack of integration in the working process has been pinpointed as the primary reason (Bernstein & Pittman, 2004; Eastman, Teicholz, Sacks, & Liston, 2011; Hartmann et al., 2012). There is a debate concerning how to promote BIM: Hartmann et al. (2012) and Eadie, Odeyinka, Browne, McKeown, and Yohanis (2014) maintain that the top-down approach is better because project management process is organized in breakdown structures. On the other hand, Arayici et al. (2011) contend that bottom-up can be more effective in dealing with user behaviors, such as the resistance to change.

Behavioral Approaches Since the seminal contribution of Tversky and Kahneman (1974), social scientists have started to study various types of heuristics employed in decision making, and explore a range of biases that can emerge when

Literature Review

33

heuristics (fast decision-making rules) are misapplied (Baddeley, 2013). The influence of the behavioral perspective can be seen in its application to a wide range of topics, including law and economics (Jolls, Sunstein, & Thaler, 1998), management (Wiseman & Gomez-Mejia, 1998), health decisions (Schweitzer, 1995), energy consumption behavior (Grubb, Brophy Haney, & Wilde, 2009), and finance (Barberis & Thaler, 2003). As shown in Shore (2008), project failure can also be attributed to behavioral biases, including data unavailability, conservatism, groupthink, optimism bias, overconfidence, disproportionate emphasis on the latest information, and the illusion of control. Status quo bias and risk aversion caused by wealth constraints appear to be pivotal for BIM participants’ responses to incentives, and therefore for the effectiveness of BIM incentivization measures. The presence of behavioral biases could be a crucial factor for the efficiency of incentivization. Researchers have employed various types of heuristics in decision making and explored a range of biases that can emerge when heuristics (fast decision-making rules) are misapplied (Baddeley, 2013). Richard Thaler’s review articles for the Journal of Economic Perspectives from 1987 to 1991 brought research attention to the circumstances where actual behavior could deviate from rational prediction. The influence of the behavioral perspective can be seen in its application to a wide range of topics, including law and economics (Jolls, Sunstein, & Thaler, 1998), and management (Wiseman & Gomez-­Mejia, 1998). The explosion of research on behavioral biases has reshaped our understanding of the mechanisms underlying human’s decisions under uncertainty. Given the ample evidence for the existence of biases in controlled experiments, it is interesting to examine the significance of these biases for high-stakes business decisions. There are three prominent theories in the literature. In classical decision theory, risks are modeled as a distribution of possible outcomes (Arrow, 1965) and the relative desirability of alternatives is judged by the expected return (Arrow, 1965; Schoemaker, 1982). This rational view of risk was challenged by two streams of literature. The behavioral theory of the firm maintains that the decision theoretic conceptions of risk are not a good description of the actual decision-­making process (March & Shapira, 1987). For instance, decision makers could only look at a limited number of possible outcomes rather than the whole distribution (Alderfer & Bierman, 1970), and tend to focus on the amount expected to lose (March & Shapira, 1987).

34

Incentivizing Collaborative BIM-Enabled Projects

The second challenge comes from the psychological view of risk. Kahneman and Lovallo (1993) contend that decision makers should be described as “subject to the conflicting biases of unjustified optimism and unreasonable risk aversion” (p. 17), instead of a rational actor who always seeks the optimal option through forecast updating. The applications of behavioral approaches in project contexts were just to emerge. As shown in Shore (2008), project failure can be attributed to behavioral biases, including data unavailability, conservatism, groupthink, optimism bias, overconfidence, excessive emphasis on the latest information, and the illusion of control. Using case study evidence, this author illustrates the instances where biases were present. This perspective also holds promise to yield new insights in enhancing our understanding of the impact of optimism bias on megaprojects (Flyvbjerg, 2014). Ansar et al. (2014) assert that, by drawing upon the “outside view” first suggest by Kahneman and Lovallo (1993), optimism bias could be alleviated (i.e., “debiased”) to arrive at empirically grounded, rather than optimistic, judgments. The establishment of the external reference points involves three steps: identify reference class, collate data from a set of reference projects for producing the distribution of parameters needed in forecasting the outcomes of the project under study, and compare the project under study against the reference distribution to make “reference class predictions.” In nature, this method was developed as a prescriptive approach, but further empirical tests are required to demonstrate that this approach can result in better cost estimation. Saad and Hegazy (2014) represent another attempt to incorporate the principle of the Prospect Theory (Kahneman & Tversky, 1979) into the fund allocation decision for road rehabilitation projects. The authors employed optimization techniques to understand the implications of a different form of objective functions formulated around the gains (e.g., improvement in road conditions) and losses (e.g., deterioration in vehicle operating costs) of the rehabilitation investment. While the authors expressly discussed behavioral economics in the second section as the motivating theory, little effort was made in the end to interpret the optimization results in light of the Prospect Theory. Compared to the existing literature, the current research makes the first attempt to build empirical evidence concerning the impact of behavioral biases on incentivization effectiveness in the context of BIM-­enabled projects.

C H A P T E R

4

Analytical Framework Overview of the Life Cycle Theory of BIM Diffusion BIM is a “scalable” technology for projects because the benefits of BIM increase considerably with the level of its application, the scope of project parties involved, the range of stages it is deployed to, and the number of functions it supports. The realization of BIM’s full potential requires a desirable delivery environment. Conceptually, BIM is a technological enabler, so the deployment strategy of BIM needs to take into account high-level factors as constraints. The transaction cost theory of construction procurement suggests that procurement systems are chosen in accordance with transaction attributes (Chang & Ive, 2007; Ive & Chang, 2007). The first-order delivery system alignment should be supported by the second-order incentive refinement to achieve the best project outcome (Chang, 2015). Achieving two-level alignments involves a natural selection process (Williamson, 1996) by which misaligned governance structures can be gradually transformed into a more efficient form. What role could BIM play in the evolution of delivery systems? Given the current trend, a host of integrated procurement systems have been promoted in advanced economies (e.g., integrated project delivery in the U.S., two-stage open book in the UK, and alliancing model in Australia). In the meanwhile, BIM is also heavily advocated by various government digital programs (e.g., UK’s Digital Built Britain). This research evinces that these two driving forces for construction productivity improvement are intertwined and can reinforce each other. The value of this empirical finding can be appreciated from the perspective of a life cycle theory of BIM diffusion. As argued in Chang, Pan, and Howard (2017), initially the proliferation of BIM could be held back by the lack of a strong financial case (indicated by D in Figure 2) because the high setup cost of BIM (e.g., expenses on hardware, software, training) cannot be justified 35

Incentivizing Collaborative BIM-Enabled Projects

Cost/Benefit

36

Intentional Institution

Spontaneous Institution Benefit

Δ: financial gap

Proportion of BIM-Enabled Projects

Cost

BIM Mandate

Delivery environment alignment +

+

Full benefits achieved

Incentive alignment –



Stages of BIM implementation I. Early adopter

II. Early majority

III. Late majority

VI. Laggard

Figure 2.  Life cycle trajectory of BIM development.

by the limited additional benefit that BIM can bring, particularly in an ill-fit delivery environment. The shortfall in net benefit makes it unlikely to see a large-scale spontaneous adoption of BIM, which explains why government mandates are upheld worldwide as the main strategy to accelerate BIM uptake. A mandated implementation could face two levels of resistance: interest misalignment among parties and the ill-fit delivery environment. The latter presents a much higher barrier than the former in implementation. As found in Chang, Pan, and Howard (2017), mandatory BIM adoption could facilitate the acceptance of integrated delivery systems in the long run as a result of BIM users’ direct experiences with the practical need for incentivization measures and the BIM-enabled improvement in communication quality. However, ushering in integrated delivery systems may take more time in some countries as it may require new legislation. Comparatively, the implementation of incentivization measures commands greater flexibility. However, it may encounter user resistance as a result of low perceived usefulness (Davis, 1989).

What Problems Can Be Solved by BIM? A project involves the organizing of parties with divergent interests to work toward the goals set out by the owner. BIM, as a coordination

Analytical Framework

37

platform, is alleged to have two effects on contracting behavior. First, digitizing design into a set of parametric 3D objects could reduce the incidence of misinterpretation of design information arising from human error in the transfer process between parties. This inherently reduces the frequency of change orders, leaving the owner less exposed to holdup payments (Chang, 2013a; Chang & Ive, 2007; Chang & Qian, 2015). Second, the use of BIM could make it possible to incorporate subcontractors’ design inputs at an earlier stage and facilitate the detection of clashes, which would result in rework during construction. While BIM can provide an avenue for the owner to reap these benefits, its application also creates a new source of conflicting interest: Why should contractors exert their best effort toward consummate collaboration that mainly benefits the owner? This is a fundamental issue that should be resolved by a BIM incentive plan.

Cooperation and Coordination The core of the model is the workflow involved in the production of design and construction in the project. The inputs from the design team and construction supply chain are brought together by a delivery system in which project tasks can be done efficiently—“by the right people, in the right way, and at the right time and place” (Roberts, 2007, p. 75). The choice of delivery system makes two major differences: First, it changes the extent to which activities can be synchronically executed and thus expedite project delivery; second, it affects the ability to achieve early engagement and deploy lean production techniques, which in turn could change the quality of design solutions in terms of value for money and fit for purpose. As argued in BuildLACCD (2016), “The principal objective of using BIM is to improve the quality of design solutions and the exchange of information between the parties.” Conceptually, cooperation and coordination are different problems. The former indicates misaligned interests, while the latter refers to suboptimal actions (Table 2). In the organizing of project parties, the challenge lies in how to make parties work toward a set of predefined goals by aligning their divergent interests. As eloquently argued by Eastman et al. (2011): With regard to the use of BIM, the general issues that either enhance or diminish the positive changes that this technology offers depends on how well and at what stage the project team works collaboratively on one or more digital models. (p. 9)

38

Incentivizing Collaborative BIM-Enabled Projects

Table 2.  Difference in the nature of cooperation and coordination. Cooperation

Coordination

Origin of problem

Motivation

Cognitive limitation

Meaning

Ensure that project parties pursue “best for all” instead of “best for self” options

Ensure that workforce, materials, and machinery are in the right place, at the right time, and with the right quantity and quality, so works can be done right the first time

Purpose

Aligning interests

Aligning actions

Solution

Incentives

Shared understanding, common knowledge

Problems

Suboptimal effort (not working hard enough)

Goods/services not delivered as planned, to the right quality and at reasonable prices

Key measure

Contractual form (incentive alignment) and financial protection

Governance

Incentivization The importance of incentives for BIM collaboration is well acknowledged in practical reports (American Institute of Architects, 2010c; Thomsen et al., 2009). BIM-enabled collaboration is driven by a reward system. To facilitate early engagement, contractors are mostly compensated on a cost-reimbursable basis. In theory, cost-reimbursable contracts provide the weakest incentives for cost reduction (Chang, 2014d; Smith, Merna, & Jobling, 2006), and hence, should be used alongside some auxiliary mechanisms. A complete BIM incentivization system consists of three subsystems (see Figure 3). 1. Target-cost setting

The first mechanism regulates the co-evolution of design and construction costs. In practice, the owner can either set a cost target at the outset to ensure that the design develops under this constraint, or can allow the IPD team to produce best value design against which the target cost is formed with a guaranteed maximum price (GMP). In setting the target cost, timing is a crucial decision. Setting a rigid budget too early could serve only to create a perverse incentive for the IPD team to reduce the scope of the project. Conversely, giving nonowner IPD members adequate freedom to search for optimal design solutions seems more likely to result in the best value outcome. A quote from Cohen (2010) provides a vivid description as to the nature of this collaborative process: “the first step should be a scoping exercise taken to the level of conceptual design, in which everyone works at cost until a deep understanding of the project and a level of comfort around the

Analytical Framework

39

Owner’s sponsor decisions I. Provision of monetary rewards?

I

1

2

II Group-based vs. individual reward II

Fees Contingency fund

Co-evolution of design and cost

Target cost

Incentive pool Underruns/ Overruns

3 Performance evaluation

III

III-1. Objective vs. subjective performance measurement III-2. Weightings of performance metrics

Gain/pain share rules

Risk impacts VI

VI-1. Linear vs. nonlinear contracts VI-2. Threshold

Incentive payment

Figure 3.  Conceptual framework for the design of a BIM incentivization system.

program and budget is achieved by all parties” (p. 15). Designs based on a rigid budget and target value design may not be easily demarcated in practice. For example, in the Sutter Health Fairfield Medical Office Building project, the owner establishes an internal budget by reference to a generic project, against which to negotiate a GMP with the IPD team. As a result, what is essential is a deeper understanding of the dynamism of design and target-cost formation, as well as its effect on design innovativeness and incentive effectiveness. 2. Incentive pool

The second mechanism concerns how to fund the incentive pool. An effective strategy to align interests is through the creation of a pool of funds in which every IPD participant has a stake. The fund may come from various sources. In a cost-reimbursable contract, IPD participants can recoup all the production costs plus a fee as profit. It is common that instead of paying out when due, the payable fees are credited to a fund jointly managed by IPD members. Each member’s entitlement for

40

Incentivizing Collaborative BIM-Enabled Projects

the proportion of the fund at the end of the project is in proportion to his contribution. When the project underruns, IPD members can recover more than their contribution. Conversely, unexpected costs fall proportionally on each member. A critical complication lies in who pays for mistakes. The owner can set aside a contingency fund earmarked for this purpose or leave this downside risk hidden in the IPD member’s offering price. In the five projects studied, the incentive pool is funded by various sources, including the owner’s surplus fund, non-owner members’ profit, contingency funds, and cost savings. According to a view from the contractor in the Sutter Health Fairfield Medical Office Building project, the incentive pool should “put profit in a separate bucket from fee” (Cohen, 2010, p. 22). From the contractor’s point of view, this position is understandable. However, incentive design should be more concerned with effects on the allocation of IPD non-owner members’ efforts. As suggested by agency theory, subjecting profits to the risk of collective performance could help induce consummate performance from IPD non-owner members, while, on the downside, it could cause the project to be perceived as riskier, and thus cause a change in the dynamism of design and cost formation. Comparatively, using a contingency fund as the source of the incentive pool could reduce the risk exposure of IPD non-owner members. However, this will limit the effect of incentivization to mistake avoidance (so as to save funds for final sharing) and thus may not effectively strengthen the incentive to innovate. 3. Gain/pain share

The third mechanism specifies the award of incentive payment. It consists of three elements: The first element relates to the percentage of savings/losses applied to each party in apportioning the incentive fund. The second element is associated with the choice of award criteria. Criteria can be linked to wider project objectives, and are subject to the owner’s subjective discretion. The third element stipulates how performance measurement is tied to fee awards. The design of a performance measurement system involves several dimensions, including who sits on the judging panel, when to pay performance awards, and how much to pay. Risk-sharing mechanisms are the engine of collaboration for IPD projects. In some of the five projects, the incentive plan employed was rather sophisticated. For example, in the project of Sutter Medical Center

Analytical Framework

41

in Castro Valley, California, the reward structure can be described in general notations as follows: 1. The owner sets out a budget envelope as the target cost (C). 2. The payment to any of the non-owner members (ci) should be agreed upon with the owner so that the total payments can be controlled not to exceed C, i.e.: 11

C

c 5C i

T

i51

3. Eleven project members signed into an IPD agreement with a clause stipulating how a contingency fund resulting from cost savings (i.e., C 2 CT) is shared among them. The provision of this fund aims to motivate non-owner members to work toward cost reduction. For an individual agent, the payment is dependent upon both the forecast cost of his work and his share of cost savings from the entire project. The sharing ratio changes with the savings made. In the first tier (C  C  C1), the savings are equally split between the owner and non-owner members. An individual member can only take a share proportional to his cost in the total project cost (xi). When the savings rise further, the non-owner members’ shares increase to 75%. The total bonus is capped if the outturn cost falls below C2:

Pi 5 ci 5 ci 1 0.5xi (C 2 C) 5 ci 1 0.5xi (C 2 C1) 1 0.75xi (C2 2 C) 5 ci 1 0.5xi (C 2 C1) 1 0.75xi (C2 2 C1)

if C  C if C  C  C1 if C1  C  C2 (1) if C  C2

This incentive fee is funded by the owner’s unspent budget (original internal budget net of the target cost) and the non-owner members’ profits. Performance in cost control is the only determinant of incentive fee. No subjective criteria are employed in performance evaluation. A similar formula is also used in the Encircle Health Ambulatory Care Center project. The incentive plan utilized by Autodesk in modernizing a 55,000-square foot, three-story interior space in a new office building near Boston is slightly different. It consists of two limbs, one of which

42

Incentivizing Collaborative BIM-Enabled Projects

is linked to cost performance. Sixty percent of cost savings can be added to the first incentive compensation layer (ICL0) for sharing. Conversely, cost overruns will be first borne by ICL0 until it is fully exhausted. The other is tied to program performance. When the completion time (t) exceeds the target time (T), a penalty will be imposed by a constant day rate d:

ICLT 5 ICL0 1 (C 2 C) 2 d(T 2 t) 5 ICL0 1 0.6(C 2 C)

if 0  C 2 C  ICL0 and T  t if C  C and T  t

(2)

By contrast, the incentives provided to the Cardinal Glennon Children’s Hospital Expansion project comes from the sharing of an unspent contingency fund with a simple, pre-agreed-upon rule (40% to the owner, 20% to design team, 40% to the builder and lean partners). Performance-linked payments hold the key to broadening the scope of objectives that non-owner IPD members would be motivated to achieve. However, this is not widely used. As seen in the case studies, only one project has considered nonfinancial factors for the award of incentive payment. In the Autodesk Waltham project, an independent third party was brought in to evaluate whether the goals of sustainability, quality of craftsmanship, functionality, and design quality have been satisfied. Nevertheless, this plan does not fully utilize subjective metrics in performance evaluation since the outcomes of evaluation were only used for determining whether, instead of how much, the incentive payment should be awarded. Based on the limited samples, the benefit of giving relative weightings to different metrics seems not well acknowledged in current practice. Apart from the size of the incentive pool, equally important is how incentive payments should be awarded. In the five projects, the incentive payments are all linked to team performance. 4. Contract forms and thresholds

A central element of the gain/pain share plan is how to factor into account the deviation of outturn cost from expected cost in the incentive payment. In the agency model, Holmstrom (1979) assumes that the agent’s effort can yield a higher profit for the principal. However, without imposing the monotone likelihood ratio property (MLRP), this is not necessarily true (Milgrom, 1981). It is also criticized for placing a singular focus on the first-order condition in solving the agent’s best

Analytical Framework

43

action as it cannot guarantee that the answer is a global maximum (Grossman & Hart, 1983). In a famous study of executive pay, Jensen and Murphy (1990) find that there is no low sensitivity between incentives and firm value (pay-for-performance slope 5 0.03), and resort to nonlinear incentives as an explanation. Theoretical concerns aside, the complicated contract predicted by the principal-agent model appears at odds with the simplicity of real-world contracts. Also, some evidence from the field seems concerning to nonlinear contracts (Gibbons, 1997). As a result, it is imperative to examine the plausibility of linear contracts. A strong justification can be found in Holmstrom and Milgrom (1987), which provides proof to show that a linear contract is coincidently the best compensation scheme for agents who can adjust their effort levels over time. This “as if” explanation provides a valid justification for restricting research focus on linear contracts in compensation design. The assumptions used in the model, including linear contracts, exponential utility function, and normal disturbance, form the tripod of the influential Linear-Exponential-Normal (LEN) framework. Banker and Datar (1989) work under this framework to demonstrate that it is possible to find a linear combination of performance measures upon which an optimal contract can be made conditional. Basu and Kalyanaram (1990) evaluate the relative performance of compensation plans taking the form of (A 1 Bx)a (A, B, a: parameters) using numerical simulations. Their finding shows that nonlinear compensation plans would perform better than linear plans in low-uncertainty environments. Linear incentive plans appear dominant in the five cases studied, which gives empirical credence to the employment of the Linear-­ExponentialNormal model (Feltham & Xie, 1994; Holmstrom & Milgrom, 1987; Lambert, 2001) as a convenient modeling technique in the study of BIM incentivization problems. Target cost plays a critical role in BIM incentivization systems. In practice, it is used as the point of reference against which incentive payments are evaluated. A question of theoretical significance is how much reward is needed to generate motivational effects. The monotonic effect of money on motivation in the principal-agent model may not be as straight as assumed (Gneezy & Rustichini, 2000). First, small compensation may have no effect because of the agent’s unwillingness to “work for peanuts.” Second, the social norm would pose a hindrance to the acceptance of monetary reward. For example, the practice of paying small rewards for recycling bottles works less well than

44

Incentivizing Collaborative BIM-Enabled Projects

the norm of not recycling being deemed bad behavior by the society. Drawing on their experiment results, Gneezy and Rustichini (2000) call for a rethinking of the rule that “a small payment is better than nothing.” Setting a threshold is a common practice in executive compensation. A significant element of executive pay comes from bonuses awarded for above-target performance (e.g., annual bonus, long-term incentive plan) (Lambert & Larcker, 1991; Murphy, 1999). A general form of the one-kinked, performance-threshold-based linear contracts can be depicted as follows (Gjesdal, 1988):

{

(b2 2 b1)Q0 1 b1Q Q  Q0 W(Q) 5 b Q Q  Q 2 0 Where the line is kinked is the threshold performance. Zhou and Swan (2003) find that when risk is high (low), the optimal compensation plan will specify a low (high) threshold. Raju and Srinivasan (1996) explore the best sales compensation plan by comparing the efficiency of a threshold-­based linear plan and a curvilinear agency-theory-based plan using numerical simulations. Regarding the total profit generated by different compensation plans, the piecewise linear plan is superior to the linear plan by 7% for the parametric scenarios. In a similar vein, regarding the ratio of the principal’s utilities obtained from piecewise-linear-threshold contracts to those from linear contracts, Chen and Miller (2009) use simulation methods to demonstrate that relative efficiency is sensitive to the agent’s utility function. For the agent with exponential utility, piecewise-linear-threshold contracts can perform no better than linear contracts in most scenarios. This offers a justification to restrict focus on the single-period LEN model in the design of compensation plans. As shown in the five cases under study, kinked performance-threshold-based incentive plans are also used in BIM-enabled projects.

Behavioral Influences The impact of behavioral biases on the effectiveness of incentivization in the context of BIM-enabled projects is not explored in the literature yet. At the outset of the research, it is not certain which biases could be significant and to what extent. Individual behavior could be affected by motivational biases and cognitive biases (Baddeley, Curtis,

Analytical Framework

45

& Wood, 2004). The former can be controlled using incentives, while the latter are not under conscious control (Baddeley, 2013). The current research takes account of both types of biases. As has been discussed previously, monetary incentivization systems are not common in practice. For this reason, the role of behavioral biases does not feature as prominently in the final report as planned in the proposal. In the second empirical study of China survey data, an attempt is made to cover a comprehensive list of behavioral biases. Through a twostage filtering process, 10 existing bias sources are investigated in the survey. The combined effect of these biases is captured as a latent construct in the SEM model. In the UK empirical analysis, the focus is placed on the influence of status quo biases, which are identified to be pivotal in pilot interviews. In the U.S. study, behavioral influences are broadened to cover the factors addressed by the technology acceptance model (TAM). The purpose of this change is to understand whether TAM variables are significant in BIM contexts. The TAM-based model provides another angle to probe the reasons why BIM could lead to a better project outcome.

C H A P T E R

5

Research Methodology Research Steps This project takes a positivist’s view to understanding the causality between the extent of BIM use and project performance, thereby inferring under what delivery environment BIM could reach its full potential in improving project performance. In the development of the proposal, the conceptual framework is built around the existing economic theories, including governance structures, incentivization, behavioral biases and TAM, and case knowledge garnered from the literature. To sharpen research focus, this research took the following steps: 1. Review literature and conduct pilot studies to investigate the key drivers behind BIM adoption in different national contexts, thereby identifying the barriers and analyzing its sources. 2. Formulate a conceptual framework comprehensive enough to capture the stylized facts discovered in Step 1. 3. Collect large data sets through four surveys and conduct a SEM analysis of the data collected through a customized questionnaire from three countries.

Research Method In the spirit of grounded theory building (Glaser & Strauss, 1967), initially, we take no theoretical position on why IPD parties would have responded differently to a given incentivization system. This research employs both case studies and statistical analysis in data analysis. Scientists understand the real world chiefly in two ways: story and metaphor (McCloskey, 1990). A phenomenon can be understood through the lens 47

48

Incentivizing Collaborative BIM-Enabled Projects

of an abstract model (metaphor) or by way of successive events (story). We have seen that the combined use of two methods can yield fascinating intellectual achievements. An example is the famous Fisher Body story scrutinized in Klein, Crawford, and Alchian (1978). The happenings within the long-term supply contract between General Motors and Fisher Body has inspired a generation of organizational economists to probe into the dynamic process of contracting that is exposed to holdup problems. The effect of appropriable quasi rent on transaction costs discovered in the case has been upheld as the primary causal link in the generalization of theory. The theory (Williamson, 1985, 1996) and empirical findings (Macher & Richman, 2008; Shelanski & Klein, 1995) of transaction cost economics have far-reaching impacts on our understanding of the nature of contracting behaviors in a wide range of contexts. For economists, the clinical study is mostly used as a tool to explore ill-­ defined issues (Baker & Gil, 2013), while management researchers use it as a standard method in both research and teaching (Eisenhardt, 1989). Given the very limited understanding of BIM incentivization problems, this research will conduct three case studies to collate the perceptions/ views of participants of a typical BIM-enabled project in three countries. As the research methods will be further elaborated in individual chapters, the following only provides a synopsis of the methods.

Case Studies As the nationwide rollout of BIM is still a recent phenomenon, and BIM incentivization is fairly under-studied, it is pivotal to conduct exploratory case studies to deepen our understanding of the interlocking relationship between delivery environment and BIM implementation. The case study provides an appropriate method to enable researchers to probe the multiple facets of the problem in depth (Baxter & Jack, 2008; Eisenhardt, 1989; Stake, 1995; Yin, 2013). A multiple case study methodology was employed to scrutinize two distinct types of large-scale collaborative infrastructure delivery frameworks and enable the interviewees to effectively flesh out their perceptions, experiences, and views of reality through their stories (Baxter & Jack, 2008).

Empirical Strategy In recent years, SEM has risen as an alternative to multiple regressions in investigating causal hypotheses (Pearl, 2009), with the great strength

Research Methodology

49

of integrating confirmatory factor analysis (CFA) (Jöreskog, 1963) and path analysis (Wright, 1934), which allows a latent construct measured by multiple observed variables. This method is particularly fitting to the current research because of three of its benefits (Bagozzi & Yi, 2012): (1) It helps researchers be more precise in their specification of hypotheses and operationalizations of constructs; (2) it guides exploratory and confirmatory research in a manner combining self-insight and modeling skills with theory, working well under the philosophy of discovery or the philosophy of confirmation; (3) it often suggests novel hypotheses originally not considered and opens up new avenues for research. Since the causality among the seven issues explored in the theorizing stage is yet to be discovered, SEM provides a flexible framework for accommodating the “unforeseen” relationships between constructs, which could arise when incorporating the effect of heuristics and biases. Moreover, several constructs (e.g., communication, collaboration, and perceived need for incentivization considered in Figure 1) contain multifaceted dimensions. These reasons all make SEM a suitable method. In implementation, the analysis follows a two-stage procedure suggested by Anderson and Gerbing (1998): Build a measurement model first for specifying the relationships among measured variables that underlie the latent variables and then a structural model for the relationships among the latent variables.

C H A P T E R

6

Case Studies – China Overview China lies in the “laggard” stage in terms of BIM use. The potential value of BIM has been acknowledged, but work practices have not been modified to create a desirable delivery environment for BIM. The barriers to adopting IPD features in China mainly result from a rigid legal system that forces procurement, tendering, and implementation to be divided. This fragmented system will inevitably impede the realization of BIM’s full potential. The four Chinese case studies discussed here are representative of the most sophisticated applications of BIM in China as of 2016. As shown in Table 3, all of them are located in the Shanghai area. Projects 2–4 are from the same owner. Owing to confidentiality, the second owner is anonymized throughout the report. The sophistication of BIM applications in these four projects can be seen from three dimensions: First, BIM appears to have been applied across all stages in all of the four projects, spanning feasibility, concept Table 3.  Summaries of the case studies. No.

Project

Location

BIM applications

1

Shanghai Tower

The Lujiazui Finance and BIM used to achieve better design, construction, and Trade Zone of Pudong District, operations management. Shanghai, China

2

The Research-Based IPD Project

The center of East Shanghai’s Pudong District, China

BIM used within an innovative IPD environment meant to involve the general contractor earlier.

3

The Core Project

The center of East Shanghai’s Pudong District, China

Won the architectural practice of technology award of American Institute of Architects (AIA) for its successful application of BIM to the digital production and construction management of complex projects.

4

The Management Center

The center of East Shanghai’s Pudong District, China

The general contractor engaged early to reduce construction alterations through BIM.

51

52

Incentivizing Collaborative BIM-Enabled Projects

Table 4.  The project phases covered by the BIM application in each case study. Feasibility

Concept Design

Detailed Design

Procurement

Construction

Operations & Maintenance

Shanghai Tower The Research-Based IPD Project The Core Project The Management Center

design, detailed design, procurement, construction, and operations and maintenance (Table 4). Second, the primary functions assisted by BIM include visualization, collaborative design, space validation, and clash detection (Table 5). There is evidence that BIM has also been employed to support construction simulation, regulatory compliance checks, and asset management. As shown in Table 6, interviews for each case included participants from across the supply chain. This provided a balanced view of BIM applications and effectiveness.

Project Background Information Case 1: Shanghai Tower Currently, Shanghai Tower is the tallest building in China. The developer turned a former golf course into a 128-story (632-meter) skyscraper with Table 5.  The functions assisted by BIM in each case study. Shanghai Tower

The Research-Based IPD Project

The Core Project

The Management Center

Visualization

Y

Y

Y

Y

Collaborative design

N

Y

Y

Y

Space validation

Y

Y

Y

Y

Environmental analysis

P

N

P

N

Model-based estimation

N

N

N

N

Digital fabrication

P

N

P

N

Clash detection

Y

Y

Y

Y

Construction simulation

P

P

P

P

Regulatory compliance

P

P

P

N

Asset management

P

P

P

N

Functions

Note: Y: fully applied; N: not applied; P: partially applied

Case Studies – China

53

Table 6.  Participants in the case studies. Participants

Owner

Designer

General contractor

Subcontractor

Supplier

User

Shanghai Tower













Pilot IPD Project













Core Project













Management Center











a cost of US�24 billion. Since completion, it has become the landmark of the whole Shanghai city, and even the whole nation of China. The building has nine functional districts, each of which has a sky lobby and atrium sandwiched between the inside and outside glass walls (Figure 4). This project took eight years to complete, from 2008 to 2016. Its designer, U.S. architecture practice Gensler, established the use of BIM on

Viewing Platform & Dining

Zone 9: L119–L121

Hotel & Boutique Office

Zone 8: L101–L115

Hotel

Zone 7: L84–L98

Sky Lobby & Office

Zone 6: L69–L80

Sky Lobby & Office

Zone 5: L52–L65

Sky Lobby & Office

Zone 4: L37–L49

Sky Lobby & Office

Zone 3: L22–L34

Sky Lobby & Office

Zone 2: L08–L19

Multifunction Space & Retail

Zone 1: L01–L05

Commerce & Passages Parking & Equipment Room

B1–B2 B3–B5

Figure 4.  Shanghai Tower and its floor functions.

54

Incentivizing Collaborative BIM-Enabled Projects

the project from its outset. However, the owner, Shanghai Tower Construction and Development Company, only became aware of the potential benefit of BIM when they saw the detailed design models created by the local architect, the Tongji Architectural Design Group. While the owner agreed to employ BIM in construction and operating stages, the original design contract was not revised accordingly. The construction tender document contains some technical requirements that the contractor needed to comply with. The general contractor, Shanghai Construction Group (SCG), was responsible for the production of the entire construction BIM model and integration of the models produced by subcontractors and suppliers. As BIM in China was not promoted until 2008, there was nearly no local expertise available for implementing BIM in such a megaproject. Many BIM applications were permeating into local design practices from international practices such as Gensler.

Cases 2–4: Shanghai International Tourism and Resorts Zone The other three case projects were commissioned by the Shanghai International Tourism and Resorts Zone. This zone covers an area of 210 square kilometers for a theme park and 23 square kilometers for support facilities. This project commenced in 2010 with a total investment of US�41 billion. The theme park, as the core of the resort area, was managed by an American team, while the support area was managed by the Chinese Shendi Group. Thus, both the experimental pilot IPD project and the core project were implemented by the American team, while the management center was executed by the Shendi Group. Over 70% of the building was built with the aid of BIM in the program. Different strategies were taken by the American and Chinese owners, who proved critical factors in the ways in which BIM was implemented on the projects. The American owner organized an integrated team, including a project integration manager and a BIM manager, to lead the integration of BIM models for the portfolio projects. The contractor worked closely with the owner’s integration team determining how construction BIM models would fit into the overall model. To ensure the integrity of the design and process optimization, the construction team built a clear understanding of the work scope and work plan through participation in the integration process, and took charge of shop drawings refinement

Case Studies – China

55

and process simulation analysis (4D). Through this, they produced a comprehensive work plan and as-built models. The Chinese owner, Shendi Group, required BIM to be used from the project’s inception. However, constrained by the traditional delivery system, BIM was employed separately by individual parties without a mechanism enabling coordination between parties. In this case, BIM can only serve as a tool to enhance visualization.

Case Study 2: IPD Pilot Project The IPD pilot project involves building a three-story experience pavilion with a floor area of 2,260 square meters. Initiated by the American owner, this was the first project in the theme park to pilot IPD. According to the eighth commandment of the Tendering and Bidding Law of the People’s Republic of China, a competitive tendering process must be followed if the public authority is the major fund provider. To avoid breaching the law, the contractor, Shanghai International Tourism and Resorts Zone Construction Company (SITRC), was engaged via a professional service contract for the feasibility study of BIM. As shown in Figure 5, this mechanism enabled the general contractor to provide input to the owner and design team to deal with the works of construction stage, such as shop drawing deepening, model building, and materials review in advance. This approach helped resolve a lot of potential design and materials conflicts that may have adversely affected construction implementation.

1 Jan 2014

IPD Consultant Phase I • Establish and use BIM model for the feasibility analysis

1 Jul 2014

Contract bid notice 24 Nov

IPD Consultant Phase II • Construction simulation for the key and complex parts

General Contract • Contract for the construction

Figure 5.  The process of the general contractor involved in the project.

56

Incentivizing Collaborative BIM-Enabled Projects

Case Study 3: Core Project This project in the theme park involved building a fairytale castle and featured a wide variety of complex decoration components. The total construction area is approximately 10,510 square meters, including a 3,000 square-meter basement and a 21-meter building. To achieve the best effects, the owner organized a design team of top talent and imposed the strictest workmanship standards. For a project of this nature, where design ideas emerged gradually over the process, design changes occur quite often, leading to frequent claims in the construction process. All these problems have contributed to delaying the project and delivering it over budget. However, in terms of outcomes, this project won an award from the American Institute of Architects (AIA) for its creative use of BIM technology.

Case Study 4: Management Center The management center building commenced in 2014. It is a typical Chinese BIM-enabled project. The traditional management model and competitive procurement were adopted in the project implementation process. Therefore, the contractors did not engage in the early design stages, and the BIM application of this project was fragmented and suffered from a lack of synergy throughout (as shown in Figure 6).

Designer provides drawings and BIM models

General contractor provides principles, construction models, and site information

Subcontractors develop their own BIM models

General contractor integrates subcontractors’ BIM models into the master model

Building a common practice BIM platform

Figure 6.  The workflow for the BIM model.

Case Studies – China

57

However, this project was delivered within the agreed-upon scope, budget, and timescales, in part because there were only three design changes during the whole execution process. This was owing to various factors. First, it is a traditional office building and relatively easy to construct. Second, the general contractor worked for the owner previously on other projects and had a future expectation of securing work with them. They were, therefore, willing to participate in the design stage to provide their suggestions even though there was no contract and fees. Third, both the designers and general contractor had BIM application experience, and both the parties had worked together on several projects, such as the Shanghai Tower. The designers, therefore, understood the contractors’ requirements well and vice versa.

Analysis Through exploring the case studies, two main findings can be found that address how BIM was introduced in the projects and why BIM was adopted.

How Was BIM Introduced in the Projects? In the Commercial Tower Project, the designers initially introduced BIM. Through demonstration, the owner was persuaded to try BIM as a way to improve project delivery efficiency. A quote from the technical director of the local design institute is pertinent: The owner of the Commercial Tower Project did not require using BIM in the beginning. We used this technology on our own because it was difficult to carry out work for such a megaproject without BIM. Based on the BIM models provided by the U.S. architectural company, we established our model and used it to solve lots of problems that cannot be worked out with traditional 2D drawings. After that, the owner considered BIM helpful and decided to employ it in the subsequent construction and operational phases. The process by which BIM was introduced in the Commercial Tower Project is described in Figure 7.

58

Vague requirements Step 2

Step 4

No requirements

........................

Supplier 4

Supplier 2

Step 3 Clear requirements

Supplier 1

Subcontractor 4

Subcontractor 3

Subcontractor 2

Property Management Company

General Contractor

........................

Design Institute

Some technical support

Subcontractor 1

Clear requirements

Supplier 3

Owner Step 1

Designer-initiated

Incentivizing Collaborative BIM-Enabled Projects

Figure 7.  BIM introduction process in the Commercial Tower Project.

In contrast, in the Theme Park Project, BIM was introduced by the owner, who recognized the potential value of BIM. This was credited to the American experts on the team who had prior BIM experience. The Theme Park Project was executed after the Commercial Tower Project, so some Chinese members of the owner team had experience with BIM. In the IPD Pilot Project, in keeping with the spirit of IPD, immediately after concept design approval, the owner’s management team started to involve both designers and contractors during the early design phase. As the pre-construction consultant, the contractor team participated in design and core team meetings and gave feedback on the constructability of models and design. To keep the original teams in the delivery of the whole project, the owner applied for exemption from competitive tendering. The application was not approved, so the whole team had to work under this constraint. This project is branded as “IPD-ish” because key project parties championed the implementation of IPD and BIM with intention rather than through the contract. The BIM introduction process in the Theme Park Project is described in Figure 8. As an innovative technology in China, BIM proliferates through two routes: One is the local basic practice (the Commercial Tower Project).

Case Studies – China

IPD Core Members

Owner-Initiated

Owner

Owner

Owner-Initiated Design Institute

59

Design Institute Consultant Team of General Contractor General Contractor

Property Management Company

Construction Team of General Contractor

Subcontractors Suppliers

Subcontractors Suppliers

Property Management Company

The IPD Pilot Project

The Core Project The Management Center

Figure 8.  The BIM introduction process in the Theme Park Project.

Another is the basic international practice (the theme park). According to Rogers’ (1983) Diffusion of Innovations theory, BIM diffusion can be broken into five different segments: innovators, early adopters, early majorities, late majorities, and laggards. Each group has its propensity to adopt a particular innovation (Rogers, 1983; Turnbull & Meenaghan, 1980). Both of the two routes follow innovation diffusion theory. The main difference is that the early adopter of the Commercial Tower Project is its designers, while in the Theme Park Project, it is the owners.

Why Did the Participants Adopt BIM? The objectives of all the Commercial Tower Project members for using BIM are summarized in Tables 7–10. In the Commercial Tower Project, the owner was attracted to BIM for a simple reason: The project was too complex to be managed using traditional methods. With BIM, the design workflow and information can be more efficiently managed. For the contractors, their motivation in using BIM was simply to fulfill their contractual obligations. Contracts for all other participants also stipulated

60

Incentivizing Collaborative BIM-Enabled Projects

Table 7.  Why participants adopted BIM in the Commercial Tower Project. Participants

For this project

After this project

Owner

Efficiency-driven

Efficiency-driven

BIM was helpful in improving the efficiency after the designer had used BIM.

We find that BIM helps achieve a better design, construction, and operation of the late management.

Solution-driven

Solution- or contract-driven

In 2009, the owners did not require BIM use, but it was difficult for us to carry out the work without BIM.

If the contract does not require the use of BIM, and the project is not complicated, we have no aspiration to apply BIM.

Contract-driven

Contract- or efficiency-driven

Before this project, many of us did not know BIM, and we were required to use BIM by the contract.

The current environment is not mature for promoting BIM on a large scale because the benefits of BIM deployment cannot cover its costs. We will adopt BIM only when it is helpful to improve construction efficiency, or there are contract requirements.

Contract-driven

Contract- or solution-driven

Before this project, many of us did not know BIM, and we were required to use BIM by the contract.

We will use BIM to carry out work efficiently (e.g., the building glass surface is curved with lots of changes). Otherwise, we will not adopt it unless the owners require BIM.

Contract-driven

Contract- or financial compensation–driven

Before this project, many of us did not know BIM, but we were required to use BIM by the contract.

We are not the direct beneficiaries of BIM presently and shortly, so we are not enthusiastic for BIM use unless required by the contract.

Contract-driven

Contract-driven

Before this project, many of us did not know BIM, but we were required to use BIM by the contract.

We are not willing to use BIM, as we cannot engage in the early design stage. The conversion from design and construction models to operating models involves a lot of modification and reconstruction, and thus heavy workload and cost.

Designer

General Contractor

Subcontractor

Supplier

User

the use of BIM. Only contract conditions or difficulty drove designers to adopt BIM in this project because BIM application involves huge additional costs in hardware, software, and training. Additionally, some pipeline collision problems, which usually appear in construction phase with the traditional design methods, are brought forward to the design stage. This leads to a 30% increase in design time according to the interviewees’ response. After this project, the BIM Research Department deputy director of the general contractor, the subcontractor of the glass curtain wall, the

Case Studies – China

61

Table 8.  Why participants adopted BIM in the IPD Pilot Project. Participants

For this project

After this project

Owner

Efficiency-driven

Efficiency-driven

BIM use can improve design coordination and integration among disciplines, and support the general contractor to use model-enabled tools such as 3D layout, scanning, and 4D planning.

The ex ante view was held.

Contract-driven

Solution- or contract-driven

BIM could increase lots of workload for the designer. We use BIM only to comply with the owner’s requirements if the project is not complex.

If there are technical problems that cannot be solved without BIM, we will use it in its right even though there is no requirement for the owner. Otherwise, we are driven by contract.

Contract-driven

Efficiency- and trend-driven

We did not have the experience and ability to satisfy the BIM requirements of the owner before we completed some of their projects. We learned and accumulated experience through the project.

Our enterprise has identified BIM as our most important niche area and established a BIM-enabled management platform. We will be an active BIM user even without the owner’s requirements. Besides, using BIM is a trend in the construction industry.

Contract-driven

Contract- or efficiency-driven

We have a long-term relationship with the general contractor, and there has been an ongoing separate BIM service contract.

The use of BIM should consider the actual needs of the project because of the high running costs of adopting BIM.

Contract-driven

Contract-driven

We are driven by contract because using BIM requires massive inputs (e.g., human resources) that are scarce and quite expensive under the current environment.

We are not the direct beneficiaries of BIM in the present and the near future, so we are enthusiastic to use BIM without contractual requirements.

Contract-driven

Contract- or efficiency-driven

Presently, few local property management companies use BIM in operations. I think the user is heavily influenced by the owner’s requirements.

We are not willing to use BIM because the conversion process from the design and construction models to operation models involves a lot of modification and reconstruction.

Designer

General Contractor

Subcontractor

Supplier

User

elevator supplier, and the property management company all expressed the same view that they will not adopt BIM voluntarily unless (1) it is financially rewarded, (2) it could lead to efficiency savings or better solutions, or (3) it is required by contract. For the Theme Park Project, the owner required all participants to use BIM, so all of the parties were contract-driven except the designers of the core project who adopted BIM for better problem solving. After these projects, all the designers chose to use BIM according to the

62

Incentivizing Collaborative BIM-Enabled Projects

Table 9.  Why participants adopted BIM in the core project. Participants

For this project

After this project

Owner

Efficiency-driven

Efficiency-driven

Our primary objectives for using BIM are to shorten the construction period, save costs, and improve the effectiveness and design accuracy.

BIM can help improve the efficiency of construction management and solve many difficult problems with 2D drawings.

Solution- and contract-driven

Solution- or contract-driven

First, the owner required us to use BIM. Second, the project is too complex. Some of the problems cannot be solved without BIM.

If there are technical difficulties that cannot be solved without BIM, we will use it in its own right even though there is no requirement for the owner. Otherwise, we are driven by the contract.

Contract-driven

Efficiency- and trend-driven

We did not have the experience of using BIM in projects before. We were required to use it by the owner, but we learned and accumulated experience through their projects.

Our enterprise has associated BIM with our main business activities, and established a management platform based on BIM. So, we will be more active to use BIM, even if the owners do not have the requirements. Besides, using BIM is a trend in the construction industry.

Contract-driven

Contract- or efficiency-driven

The general contractor required us to help improve the construction process and to solve the installation space problems by using BIM.

We are willing to use BIM in large projects. Although using BIM needs more time to deepen the design model in the initial stage, it can help reduce problems for the latter construction stage and save time for the whole process.

Contract-driven

Contract- or trend-driven

We are driven by contract, and the owner’s requirements are very strict. There are some punitive measures in the contract, so our thought is very simple: getting the project completed by the requirements.

Although using BIM could increase workload, we will not resist. To improve our adaptability to changes, we are very willing to make new attempts, although there will be some extra cost.

Contract-driven

Contract-driven

At present, few local property management companies use BIM in the operational phase. I think the user is mainly affected by the owner’s requirements.

We are not willing to use BIM because the conversion process from the design and construction models to operation models involves a lot of modification and reconstruction workload.

Designer

General Contractor

Subcontractor

Supplier

User

difficulty degree of project implementation or owners’ requirements. After the “learning by doing” phase in these projects, the general contractor emphasized the trend and efficiency of BIM use. Their chief engineers said they would continue to apply BIM even without contractual requirements. The steel structure subcontractor who had a long-term partnering relationship with the general contractor was driven by the contract and efficiency. However, all the suppliers and

Case Studies – China

63

Table 10.  Why participants adopted BIM in the Management Center Project. Participants

For this project

After this project

Owner

Efficiency-driven

Efficiency-driven

The primary objectives of BIM application are to assist decision making, program optimization, site construction management, and schedule and cost control.

Although BIM application will increase the workload, it can bring management benefits for the projects.

Contract-driven

Solution- or contract-driven

We are mainly affected by the project requirements.

At present, BIM application in projects will increase our workload, but we will make decisions based on the difficulty and characteristics of the project.

Contract-driven

Efficiency- and trend-driven

We were required to use BIM for this project.

At present, our company stipulates that all housing projects should use BIM, even if the owners do not have a requirement because the use of BIM can improve construction quality. Besides, using BIM is a trend in the construction industry.

Contract-driven

Contract- or efficiency-driven

At present, we are mainly affected by the requirements of the owners.

Depending upon the nature of the project, the potential of BIM in reducing construction costs will be considered.

Contract-driven

Contract- and trend-driven

The owner of the project required using BIM. If the owner does not require it, we will not take the initiative to use BIM. It involves huge input costs.

We will use BIM if required by the owner, and we are also happy to develop our expertise in the area further because it is a significant trend in the future.

N/A

N/A

N/A

N/A

Designer

General Contractor

Subcontractor

Supplier

User

users (Property Management Company) were driven by the BIM contract requirements. All findings are summarized in Tables 7–10.

The Effects of Delivery System (Traditional) and (Lack of) Incentivization on the Effectiveness of BIM Through the investigation of the four projects, all these projects were delivered using the traditional delivery system except the IPD Pilot Project. The IPD Pilot Project was conducted within a conventional construction contract and a pre-construction consultancy package. There was virtually no incentivization for using BIM, though some project contracts contain fees for BIM services. This status quo presents impediments to effective BIM applications, as it could result in hidden costs and extra time.

64

Incentivizing Collaborative BIM-Enabled Projects

“Hidden” Costs of Employing BIM Under the Traditional Delivery Environment With the traditional procurement system, contractors cannot engage in the project at early stages. For example, the contractors of the Commercial Tower Project were not engaged in the project until they won the bidding and, at that time, the design was nearly finished. This led to the low utilization of the design institution’s BIM model in the construction phase. As argued by the owner’s BIM technical director, Since the designer mostly concerns building systems and performance rather than constructability and installation procedures, the contractor thought the design models were not operable for construction. Modifying and checking the design models was going to be more time-consuming than redoing the model, so the contractor prefers to re-build their BIM models. To rebuild their models based on 2D drawings, the contractor needed to retain a BIM specialist company, which led to additional transaction costs for searching information search and contract enforcement. These costs can be regarded as the first type of transaction costs (TCI) (Chang & Ive, 2000). If contractors choose to reuse design models, they have to check these models and undertake design mistakes together with design institutions. Once some design errors are not discovered before construction and cause losses during the execution process, it will result in disputes over defects, the consequence of which is that both parties could incur unnecessary legal costs and opportunity costs of delay. In more severe situations, welfare loss due to the adoption of inefficient technology or extra payments caused by the second round of information problems (Chang & Ive, 2000). Therefore, both the remodeling and reusing of the design models will attract extra costs and consume more time. For example, the general contractor for the Core Project was not engaged in the design stage and was asked to reuse design models. With limited tendering time, the general contractor did not fully understand design models and underestimated the difficulty involved in converting design BIM models into construction BIM models. This resulted in an unrealistic low bidding price, so additional compensation claims were

Case Studies – China

65

asked during and after the construction stage. This increased the second type of transaction costs (TCII) significantly according to TCE. Besides, the lagged effect of a bitter experience could impact the next transaction. To preempt the contractors’ claims, the owner’s requirements provide detailed requirements. As vividly described by a designer: “We were asked to provide even the models and parameters for screws and nails.” Even though these requirements are conducive to improving accuracy in construction, more resources are needed. Furthermore, a value loss may happen as a result of less productive technology being employed for avoiding holdup (Chang & Ive, 2000). If the general contractor can participate in initial discussions about BIM protocol and give inputs to model setup, the contractor’s BIM team will be more helpful in supporting the designer to start modeling details during the construction phase. This helps the general contractor become more familiar with the design and reduce the designers’ workload. This observation was seen in the IPD Pilot Project. As the integration leader of the project explained: We wanted to set up this project delivery system as IPD at the beginning but were forced to use the traditional method because of the public tendering law. Under a real IPD contract providing appropriate incentives, the parties (especially contractors) would be more incentivized from the start to give better inputs and feedback. This is not only true for BIM implementation but for overall performances. While in the end, the project was procured by the traditional construction contract, the pre-construction consultancy package can help engage the general contractor as a consultant in design and at core team meetings, and give feedback about the model and constructability of the design. This helps to increase the reuse rate of design BIM models, and reduce remodeling. Nevertheless, the two contracts signed between the general contractor and owner not only resulted in double bargaining and decision costs during the contracting stages, but also gave rise to commu­nication costs due to the contractor team that participated in pre-construction consultancy not being the same group that was involved in the later construction stage. The owner and designers had to explain their ideas and

66

Incentivizing Collaborative BIM-Enabled Projects

regulations to the new group repeatedly. Likewise, the general contractor side had to take time on making an internal work handover. In the Management Center Project, while the general contractor was not engaged in the design stage, the owner was able to reuse design models that resulted in fewer reworks than is typical. This was mainly attributed to the low complexity of this project and that the general contractor and design institute are familiar with each other since they have collaborated previously. All the actions of general contractors adopted and the effects of employing BIM under the traditional delivery environment are summarized and shown in Table 11. Table 11.  Hidden transaction costs of employing BIM under the traditional delivery environment. Projects

Actions

Effects

Extra Costs

The Commercial Tower Project

1. Without early engagement 2. Remodeling

1.  Researching information 2.  Remodeling and re-measuring 3.  Repolicing modeling process

1.  More time-consuming 2.  Resource-incurring TC

The Core Project

1. Without early engagement 2. Reuse of design models

1.  Check design models 2. Undertake hidden design mistakes 3. Not fully understand design models 4.  Underestimate difficulty 5.  Too-low quoted price 6.  Massive claims for compensation 7.  Huge disputes over trifles 8. Lagged effect of a bitter experience 9. Devote more resources to next project to avoid vulnerable position

1. The legal fee for damages claims and opportunity costs of delay (TCII) 2. More resource-incurring in next project (TCI) 3. Welfare loss due to adoption of inefficient technology or extra payments caused by inferior quality (TCII)

The IPD Pilot Project

Engage early with a traditional construction contract and a pre-­ construction consultancy package

1. Double bargaining and decision during the contracting stages 2. Pre-construction consultancy team was not the construction group 3. Repeat communication with general contractor’s two groups 4. Internal work handover of the two groups in general contractor

1. Cost of bargaining and re-measuring (TCI) 2. Cost of repeat communications (TCI)

The Management Center

1. Without early engagement 2. Reuse of design models 3. With good experiences of working together

1.  Check design models 2. Undertake hidden design mistakes 3.  Fewer reworks

Based on the low project complexity and experiences of working together with the general contractor and the design, less rework arose, and low extra costs were incurred.

Case Studies – China

67

Overall, remodeling and reusing (design models) are two possible directions for contractors to take when BIM is employed under the traditional delivery system, but both choices could incur extra costs, lengthen the time, and reduce BIM value. The choice between these two options should be based on the project context.

Effects of Employing BIM Without Incentivization None of the four projects provides incentivization for BIM deployment. The only exception is that the Commercial Tower Project contains separate financial compensation for BIM. The Theme Park Project had a BIM contract with agreed-upon service prices. Lack of incentivization will lead to lower BIM effectiveness, ineffective communication, and transaction costs. According to the owner’s BIM technical director of the Commercial Tower Project, there are three BIM application models in China. First, in the basic model where the design team and the BIM team work separately, the BIM work cannot start until 2D drawings are completed by the design team, which will inevitably lengthen the time. In the second model, while the project team and the BIM team are still separated, their working relationships are drawn closer through coordination, which may be able to shorten the time in BIM implementation and lead to greater efficiency. The weakness is that the coordination between parties has limitations. In the third and more advanced model, project engineers with professional backgrounds carry out the production of BIM models directly, which can thus achieve greater efficiency. A summary of the BIM models employed in four case studies is shown below (Table 12). Table 12.  BIM practice models of the four case study projects. No.

Project

Phases

BIM Model

1

The Commercial Tower Project

All Phases

Secondary Model

2

The IPD Pilot Project

Design Phase Other Phases

Advanced Model Secondary Model

3

The Core Project

Design Phase Other Phases

Advanced Model Secondary Model

4

The Management Center

All Phases

Secondary Model

Without incentives in the BIM contract, contractors are naturally concerned as to how to get the job done as required, rather than how to realize its value. With the basic or secondary BIM model, the project

68

Incentivizing Collaborative BIM-Enabled Projects

team and the BIM team tend to go about their own business. The BIM group focuses mainly on the implementation of contract requirements while the project team prefers to execute construction works in a traditional way. This may result in the misalignment of BIM employment and the project implementation process, with contractors still relying on two-dimensional drawings to carry out construction. This sort of phenomenon exists in all four projects. Therefore, the lack of an incentive mechanism for motivating participants to collaborate and improve project delivery efficiency is adverse for achieving the full value of BIM. If owners want to reap more potential benefits of BIM, they have to add costs on detailing the BIM requirements and increase effort on enforcements and policing the whole delivery process. Without incentive mechanisms, many contractors and suppliers buy in BIM services that meet owners’ requirements, rather than developing their BIM expertise. Thus, transaction costs are incurred between the buyers and BIM consulting firms. As a result, BIM consulting firms become intermediary agents for communication. This increases the communication costs of the whole project team and conflicts the original aim of adopting BIM technology to increase value. Incentives are important mechanisms in advancing BIM implementation and improving project delivery. For example, in the IPD Pilot Project, the owner attempted to discuss incentives with the general contractor at the onset of the project. The project manager of general contractor did not give a positive response. Later, the contractor team realized the potential long-term business opportunity, changed their attitude, and put efforts into BIM implementation. Thus, if there is no possible future work opportunity as an incentive, general contractors will not put their best efforts into collaborating for BIM implementation. So more costs are required to enforce and police the performance of participant parties. The effects of (lack of ) incentivization on the application of BIM are recapitulated in Table 13. Table 13.  Effects of (lack of) incentivization on the application of BIM. No.

Participants’ behavior without incentivization

Effects and extra costs

1

The BIM group and the project team are separated, and two teams tend to go about their own business.

•  Discounted BIM application value •  Higher communication costs •  Higher enforcements costs

2

Contractors and suppliers tend to source BIM services from outside instead of training employees to be BIM experts.

•  Discounted BIM application value •  Greater communication costs •  Higher transaction costs

Case Studies – China

69

The Desired Delivery Environment for BIM Implementation As BIM in China is in its infancy, suitable incentive mechanisms on projects have not yet been established. The four investigated projects were delivered without incentive mechanisms. However, the participating parties put forward some views about the incentivization arrangements. This research collected the participants’ perception of the main elements of this mechanism, including financial or nonfinancial incentivization, incentive funding sources, subjective or objective evaluation weighting, and linear or nonlinear compensation contracts. This section also discusses the differences found in answers according to the professional features of interviewees and the characteristics of projects.

Contrast the Answers to Financial Incentivization Answers to the efficacy of financial incentivization from the participants of the four projects are summarized in Tables 14–15. From these summaries, it appears that the key parties, such as owner and designer, have more positive responses on the efficacy of financial incentivization. Three of the four owners agreed that financial incentivization was needed to enforce participants to pay more attention to BIM applications. The owner of the core project held a neutral attitude because she thought the company management layer was more interested in financial compensation, while the grassroots staffs, such as designers and construction workers, prefer nonfinancial incentives such as more autonomy. This was due to the financial incentives that were beneficial to enterprises rather than distributed to specific executors or bounded directly with their performance. This implies that the efficacy of financial incentivization is likely to be affected by the salary system within the participating enterprises. The designers of the Commercial Tower Project and the Core Project both advocated additional financial compensation as BIM implementation involved significant extra costs. While the designers of the IPD Pilot Project and the Management Center expressed neutral voices, factors such as design market situation and project and participant characteristics were considered. For the designer, when the market is booming, nonfinancial incentives such as advertisement benefit, reputation, and learning opportunities that help improve core competitiveness are valued. When the market is shrinking, financial incentives are then emphasized. Also, if the project has no high public attention, the design firm will need more financial incentives. Data suggest that senior designers

70

Incentivizing Collaborative BIM-Enabled Projects

Table 14.  Answers to financial incentivization from the participants of the four projects. Participants

The Commercial Tower Project

The IPD Pilot Project

The Core Project

The Management Center

Owner

Yes

Yes

Neutral

Yes

Financial incentivization is more effective.

Financial incentives certainty can help project managers find a general contractor to pay more attention.

Top decision makers prefer financial incentives, while grassroots staff prefer more autonomy.

Sharing gains is more efficient to stimulate the enthusiasm of the participating parties.

Yes

Neutral

Yes

Neutral

BIM implementation involves enormous costs for design institution.

The market situation determines this.

Massive workforce and This is dependent upon resources were required the characters of projfor BIM implementation. ects and participants.

General Contractor

Neutral

Neutral

Neutral

Neutral

The reputation of the project decides this.

This is decided by the reputation of projects.

This is decided by the reputation of projects.

This is decided by the reputation of projects.

Subcontractor

No

No

Neutral

Yes

The values of BIM application can exceed its costs.

Project management is the core security condition for BIM implementation.

The financial incentivization is effective at the beginning, but its efficacy will decline with time and is affected by the workload.

The financial incentivization is direct, and the nonfinancial award mechanism is not perfect enough.

No

No

No

No

Reputation is more efficient for state-owned holding enterprises.

Contract requirements are more effective than financial incentivization.

Nonfinancial incentivization like reputation is more effective than financial incentivization.

Financial incentivization is few in Chinese projects, and the penalty is more common, so brand promotion and honor are more attractive.

Neutral

Neutral

Neutral

N/A

Financial and nonfinancial incentivization is equally important.

Early engagement opportunities are more important for users.

Early engagement opportunities are more important for users.

N/A

Designer

Supplier

User

care more about financial compensation, while younger participants tend to be more interested in nonfinancial policies. All the general contractors adopted a neutral attitude to BIM implementation. They viewed project reputation as a nonfinancial incentivization that has a strong effect on the general contractors’ attitudes. All four projects have a relatively high public profile, so they had enough enthusiasm even without financial incentives.

Case Studies – China

71

Table 15.  Answers to incentive funding from saving costs. Participants

The Commercial Tower Project

The IPD Pilot Project

The Core Project

The Management Center

Owner

Yes

Yes

Yes

Yes

Adopting saving costs is more effective to motivate participants.

Adopting saving costs helps to give a scale to the incentives.

Adopting saving costs is more effective for cost savings.

Sharing cost savings is more effective to motivate participants.

No

Yes

Yes

Yes

It is hard to judge the amount of savings and to allocate it fairly.

If all parties can join and share the cost savings, it could help stimulate everyone’s enthusiasm.

If all parties can join and share the cost savings, it could help boost everyone’s enthusiasm.

Adopting cost savings is more effective, but it is important to set up an allocation mechanism.

General Contractor

No

Yes

Yes

Yes

It is difficult to realize cost saving under current unit price accounting system.

If all parties can join and share the cost savings, it could help stimulate everyone’s enthusiasm.

Sharing savings is easier to convert the adversarial relationship into a cooperative relationship.

Sharing saving is easier to turn the adversarial relationship to a collaborative relationship.

Subcontractor

No

Yes

No

Yes

Tying financial awards to individual parties is more effective than sharing savings.

Cost savings can be allocated according to the performance of participants.

It is hard to measure the amount of savings, and funding from the owner is safe for participants and good for quality.

By sharing savings, it is easier to convert the adversarial relationship to a cooperative one.

No

No

Yes

No

It ’s hard to measure the amount of savings and participants’ performance.

When all parties join in one contract, project accounting is complex, and someone will tend to shirk on the job.

Sharing cost savings can help improve project effectiveness and enhance team collaboration.

There is no such experience, and most suppliers just want to get payment sooner rather than later.

Yes

No

No

N/A

Sharing savings can motivate all parties and owners do not need to pay more.

It’s hard to measure the amount of savings.

It’s hard to measure the amount of savings.

N/A

Designer

Supplier

User

Both the glass curtain subcontractor of the Commercial Tower Project and the steel structure subcontractor thought nonfinancial incentivization was more effective than financial incentives. The glass curtain subcontractor reaped significant benefits from BIM application for digital fabrication in the Commercial Tower Project via saving labor input and decreasing errors and rework. Therefore, the realization of BIM value is more attractive and persistent. The steel structure subcontractor considered financial

72

Incentivizing Collaborative BIM-Enabled Projects

incentivization as only an additional measure for accelerating BIM application, while proper management was the determining factor for ensuring its implementation. The mechanical, electrical, and plumbing (MEP) subcontractor of the Core Project thought the financial incentives were effective at the beginning, and its influence was easy to decline with time and workload growth. The subcontractor of the Management Center preferred financial incentivization because the nonfinancial award system was not established entirely in China. Currently, BIM is only used for space validation and clash detection in most Chinese projects. This is far from the full value of BIM. Most participating parties have not reaped its benefits, so they need financial incentives. However, as pioneers in the use of BIM, the glass curtain walls and steel structure subcontractors have received the value of sophisticated practice systems and experience. Whether there are financial incentives has no significant influence on them. As interviewees are from state-owned holding enterprises and small supplier companies that value reputation and advertising effectiveness, all suppliers considered nonfinancial incentives such as reputation, brand promotion, and honor as the most compelling. All users held a neutral attitude to BIM incentives, as some of them had not participated in the early stages and had limited experience using BIM. To summarize, the efficacy of financial incentivization could be affected by many factors, which can be classified into two categories, endogenous factors (enterprise characteristics, project characteristics, costs of BIM implementation, the realization of BIM value) and exogenous factors (market situation, nonfinancial award systems). Using 1 to represent positive effect and 2 for adverse effect, the relationships between the efficacy of financial incentives and factors are described in Figure 9.

Contrast the Answers to Incentive Funding Sources If there are situations where financial incentives are offered, it is essential to identify the source of incentive funding carefully. Based on 23 interviewees from four projects, more than half of the respondents (14) preferred linking incentive payments to cost savings, and more key participating parties (owners, designers, general contractors) chose to share cost saving. Three-fourths of subcontractors and suppliers and two-thirds of users did not accept taking project cost savings as incentive funding. All four owners agreed that adopting savings was more effective in motivating participants and was helpful to give a scale to incentives.

Case Studies – China

Shrinking market (+)

Exogenous factors

73

Booming market (–)

Endogenous factors Positive characters: • Normal size • Salary related with incentives • Elder executors

Value realization of BIM application (–)

Positive characters: • Normal shape • Easy to implement • Low reputation

Poor nonfinancial award system (+)

Characters of participating parties (+/–)

Efficacy of financial incentivization

Negative characters: • Small size • State-owned • Salary not related with incentives • Younger executors

Cost of BIM implementation (+)

Endogenous factors

Negative characters: • Special shape • Difficult to implement • High reputation • Good management • Long duration • Heavy workload

Exogenous factors

Good nonfinancial award system (–)

Characters of project (+/–)

Figure 9.  The model of factors influencing the efficacy of financial incentives.

They also pointed out that they do not need to prepare extra funding to motivate participants and can get some reward if there is a savings. Three-fourths of designers and general contractors believed sharing savings was more effective in stimulating everyone’s enthusiasm and establishing improved cooperative relationships. While it is complicated to define the savings and responsibilities resulting from faultless design and suitable construction methods, increased costs caused by design alterations occur frequently. The designer of the Commercial Tower Project worried it would be difficult to measure the amount of savings and to allocate it fairly. The general contractor of this project thought it would be hard to realize cost savings under the current unit price accounting system because participants were likely to increase further alterations to reap more benefits.

74

Incentivizing Collaborative BIM-Enabled Projects

Two subcontractors, one supplier and one user, agreed that incentive funding from cost savings was effective and helpful for improving project effectiveness and team collaboration. However, two subcontractors, three suppliers, and two users disapproved of using cost savings as incentive funding, as cost savings are hard to measure. Besides, when all parties join in one contract, the project accounting is complex, and all participants’ interests cannot be guaranteed. Additionally, lots of the subcontractors and suppliers are only responsible for one or two parts of the project, so their financial award proportions are not high according to their turnovers. From this perspective, they would prefer receiving all potential fees sooner rather than waiting for a portion of incentives until the end of the project. Moreover, some participants tend to use cheaper, inferior substitutes to save costs so that project quality will be affected. As explained before, the advantages and disadvantages of taking savings as incentive funding can be summarized in Table 16. Table 16.  The advantages and disadvantages of taking savings as incentive funding. Advantages

Disadvantages

1. More effective to motivate participants’ enthusiasm

1. It’s hard to measure savings amounts and participants’ performance

2.  Helpful to give a scale to incentives

2.  It is difficult to allocate the savings fairly

3.  More efficient for cost saving

3.  The project accounting is complex

4. Helpful to convert adversarial relationship to cooperative relationship

4. Some parties are likely to loaf on their jobs when all parties share gains and pains

5. Owner does not need to pay more money for incentives

5. Some parties would rather receive all potential fees sooner rather than waiting for a portion of the incentives until the end of the project

6. The owner can get some reward if there is a savings

6. The project quality would be affected when some parties use deliberately cheaper, inferior materials to save costs

Contrast the Answers to Objective or Subjective Evaluation Weightings Nearly all interviewees agreed that objective standards need to play a leading role when evaluating the performances of participants. However, they were not sure about the weighting of each sort of standard. Given that some of the projects are incredibly complex, the participants’ preferences for objective measures indicate that such measurements can be used less in the evaluation of complex jobs. The answers are summed up in Table 17.

Case Studies – China

75

Table 17.  Answers to weightings of objective and subjective evaluation. Participants

The Commercial Tower Project

The IPD Pilot Project

The Core Project

The Management Center

Owner

No

Yes

Yes

Yes

The majority of the standards are subjective, but there is no specific proportion.

Objective metrics are set up for outcomes. In addition, behavior and task-oriented requirements are set up.

Objective are greater than subjective ones, and subjective standards have no effect on payments.

The objective standards are measurable.

Yes

Yes

Yes

Yes

There was no BIM contract and evaluation for this project. For design, more than 80% of standards were objective.

It is difficult to set the proportion of objective standards, but now the objective measures outnumber subjective ones.

The objective standards outnumber subjective ones.

The objective standards are around 70% of all measuring criteria.

General Contractor

Yes

Yes

Yes

Yes

At the beginning, the requirements were subjective because of no experience.

The standards the owner provided were mainly objective.

The requirements of the contract are mostly objective.

The conditions of the contract are mostly objective.

Subcontractor

Yes

Yes

Yes

Yes

70% of the requirements from the contracts for BIM are objective.

Measurement standards should be quantified as far as possible.

The majority of requirements are objective.

The majority of requirements are objective.

Yes

Yes

Yes

Yes

The majority of requirements are objective.

Except for the requirements for behavior and service attitude, the majority of requirements are objective.

The contract agreement for BIM is very detailed, and the majority of requirements are objective.

80% of the measurements are objective and most of the subjective clauses are for service attitude.

Yes

Yes

Yes

N/A

The majority of criteria are objective.

The majority of requirements are objective.

The majority of requirements are objective.

N/A

Designer

Supplier

User

Only the owner of the Commercial Tower Project said the majority of its performance indicators for BIM are subjective. This was verified by the responses from the designer and general contractor. At the beginning of the project, there was no BIM requirement and contract for design. After the design stage, the owner required contractors to use BIM in this project and provided some general demands like a completed model and satisfying the requirements of operation. These requirements are extremely vague, subjective, and immeasurable. With collaboration

76

Incentivizing Collaborative BIM-Enabled Projects

during the process of construction, contractors assisted the owner in setting up a suitable and more objective measurement system, which is also followed by subsequent projects. This bottom-up process of establishing standards is described in Figure 10.

No requirements • Design stage • Owner did not require BIM use

Vague and subjective requirements • Early construction stage • Owner required BIM use and provided primary requirements

Objective and subjective requirements • Late construction stage • Owner and contractors built a more effective measurement system

Figure 10.  The evaluation standards forming process in the Commercial Tower Project.

Informativeness theory (Antle & Demski, 1988) provides a useful angle for the development of an evaluation system in the Shanghai Project. At the outset, few participants had experience with BIM, so using objective measurements may not add much informational content about their actions, and this is the reason why the evaluation is mainly subjective. With the project progresses, participants knew more specific and measurable requirements about BIM implementation, so objective standards play an increasing role in the evaluation of agents’ performances.

Chapter Conclusions Despite BIM being enthusiastically promoted worldwide, the realization of its value through advanced implementation is still rather elementary. To explore the reasons for this phenomenon and to provide solutions, this chapter conducts an overview of current BIM practices in China and extracts the desired delivery environment for BIM implementation through examining four Chinese BIM-enabled projects. The results indicate that, except for owners and some pioneers, most practitioners adopt BIM owing to contractual obligations rather than voluntarily. The reasons why participants take BIM before and after these projects suggest that how the net benefit of BIM is realized has a substantial effect on the promotion of BIM. When benefits cannot offset

Case Studies – China

77

the costs, BIM use is mainly driven by push forces such as contractual requirements or financial compensation. When benefits can cover costs, pull forces such as efficiency and problem solving dominate the reasons for BIM adoption. Traditional procurement method (e.g., design-bid-build) dominates the current delivery system and there is no formal incentivization in China, so the analysis has identified many adverse effects regarding extra costs, time, and reduced utilization value in BIM applications. In dealing with BIM contracts without incentive structure, contractors tend to focus on contract completion rather than BIM value realization. If owners want to reap more benefits from BIM, they have to add costs on detailing BIM requirements and increase their efforts in enforcements and policing the whole delivery process. Therefore, the lack of incentivization is likely to cause extra costs, and is also adverse for achieving the full value of BIM. Overall, the separated project delivery process, together with lacking incentivization, has severely impeded project participants to establish an integrated team to collaboratively use BIM throughout the project life cycle. Therefore, it is important to take the delivery system and incentive mechanisms into account when creating the delivery environment for BIM implementation.

C H A P T E R

7

Empirical Investigation – China (1) Problem Identification In May 2015, the Chinese government published a series of national standards for utilizing BIM and regulations related to BIM implementation. In July 2014, the Department of Housing Construction issued the Suggestion for Advancing Construction Reform and Development (as cited in Ni & Wang, 2015), which requires promoting the use of information technology in the whole life cycle, from design, construction, and maintenance of a project. This document also indicated that by the end of 2020, the ratio of projects using BIM in medium- and large-sized buildings financed by state-owned assets, public green buildings, and green ecological demonstration housing projects must achieve 90%. The overall adoption rate of BIM in China remains considerably lower than that of developed countries (as cited in Cao et al., 2015). The use of BIM to date is still limited principally to visualization in China. With the current high drive from Chinese central government, it can be anticipated that BIM will proliferate quickly in the Chinese architecture engineering construction (AEC) industry. Given the predominance of the traditional design-bid-build delivery system in China, Chinese BIM users will come to realize that BIM cannot reach its full potential in improving project coordination without introducing collaborative delivery systems. In this research, integrated project delivery (IPD) is chosen as the exemplary collaborative project governance owing to its widespread influence in the U.S. construction industry. Based on the survey returns of 145 Chinese BIM-enabled projects, this research demonstrates statistically that the acceptance of IPD features increases with the use of BIM applications through two channels: one via the improved awareness of incentivization 79

80

Incentivizing Collaborative BIM-Enabled Projects

being a crucial element in governing BIM-enabled projects, and the other by improved communication quality effected by BIM. This finding can be drawn upon by the Chinese government as an additional rationale in support of the compulsory implementation of BIM.

The Model and Hypotheses Based on the literature review, the core model (see Figure 11) contains five latent variables: degree of BIM application, perceived importance of BIM incentivization, communication quality, collaboration quality, and the extent of IPD acceptability. Each of these variables is comprised of several observable variables. As the scope of BIM application in a project is determined at the outset, it is treated as the independent and only exogenous variable. Additionally, other influencing factors (i.e., respondent’s experience, project attributes) may be of importance to the prediction power of this model. In summary, the model consists of six hypotheses: Hypothesis 1 (H1) The degree of BIM application can raise the perceived importance of BIM incentivization. Hypothesis 2 (H2) Perceived importance of BIM incentivization will have a positive effect on IPD acceptability. Hypothesis 3 (H3) The degree of BIM application can improve the quality of communication. Hypothesis 4 (H4) Better communication quality will lead to greater IPD acceptability. Hypothesis 5 (H5) The degree of BIM application can improve the quality of collaboration. Hypothesis 6 (H6) Better collaboration outcomes can increase IPD acceptability.

Questionnaire Development This research designed a questionnaire to elicit experts’ assessments of the five constructs in Figure 11. Data were initially recorded by SPSS 19

Empirical Investigation – China (1)

H1 Degree of BIM Application

H3 H5

Incentivization Communication Collaboration

81

H2 H4

IPD Acceptability H6

Figure 11.  SEM model of IPD acceptability.

and then entered into a structural equation model using AMOS 17. Since the quantitative approach was considered appropriate to analyze individuals’ attitudes, main questions were measured on a 7-point Likert scale. The first construct is concerned with the extent to which BIM was used in the project, which can be measured by three dimensions (see Table 18): how advanced was the BIM model (Level 0, 1, 2, 3) in which project phases the model was used, and what functions BIM has assisted in serving. The four-level BIM maturity model originally developed by Bew and Richards (2008) has been employed in this research. This should ensure clear articulation of the standard classifications and help respondents understand the processes, tools, and techniques involved in each of the BIM levels defined in this model (BIS, 2011).

Table 18.  Measurement of degree of BIM application. BIM level

Level 0 Level 1 Level 2 Level 3

In which project phases was BIM used?

What functions has BIM assisted to serve?

– Feasibility –  Concept design –  Detailed design

– Visualization –  Collaborative design –  Space validation –  Environmental analysis

Implementation document procurement

–  Model-based estimation

– Construction

–  Digital fabrication –  Clash detection –  Construction simulation –  Code checking –  Facility management

–  Operations and maintenance Normalized score  number of level/4

Normalized score  Number of phases assisted by BIM/7

Normalized score  Number of functions served by BIM/10

82

Incentivizing Collaborative BIM-Enabled Projects

As the three dimensions are nesting to one another, they cannot be used as parallel constructs to form the variable. Considering the cross-­ effect of three dimensions, this construct is calculated by taking the multiplication of the normalized score of each dimension (see Table 18 for details). The variable to be explained in the model is to what extent the acceptance of IPD features could change in response to the differing degree of BIM application in the project. While most of the respondents were familiar with BIM, they were less familiar with IPD and its relevant concepts. Given that there is no existing measurement of IPD acceptability, this research first identified the standard features of IPD based on the literature (Cohen, 2010), and second, developed the questions that can effectively elicit the respondent’s view on the necessity of IPD futures for BIM-enabled projects in the future. All the features adopted were originated from IPD case studies reported in Cohen (2010). For ease of referencing, the 15 features and their measurements are grouped into three categories: contractual, managerial, and technological (see Table 19). The second construct aims to assess the quality of collaboration. This construct is measured using several metrics discussed in the literature for measuring IPD collaboration (Abdirad & Pishdad-Bozorgi, 2014; Brewer & Mendelson, 2003; Moore et al., 2005; Pocock et al., 1996; Thompson & Ozbek, 2012): aligned goals, centralized working place, multidisciplinary knowledge, and real-time information sharing. The third construct is to evaluate the quality of communication. Aside from the traditional measures of communication quality by virtue of accuracy and timeliness (Mohr & Spekman, 1994), three additional criteria are also included here: First, transparency reveals another aspect of communication quality as information flow within the project may be impeded by asymmetric information (Kadefors, 2004; Zaheer, McEvily, & Perrone, 1998). Second, an initiative in participation is concerned with the degree of keenness in contributing to decisions and goal formulation within the project (Mohr & Spekman, 1994). Third, communication frequency is meant to capture how actively parties have interacted with each other in exchanging information (Freeman et al., 2006; Mohr & Spekman, 1994; Pocock et al., 1996). The detail of three constructs can be found in Table 20, including a brief explanation for each construct, constituent elements of each construct, their measures, and notations in the model.

Empirical Investigation – China (1)

83

Table 19.  Measurement of IPD acceptability. Notation in the model

Categories

Dimensions

Representative case

Measurement

Contractual

Multiparty contract

Cathedral Hill Hospital

A new type of contract should be signed between the main project stakeholders to realize co-management and promote multilateral collaboration.

VA1

Incentive tied to goals

Edith Green-Wendell Wyatt Federal Building

Financial incentives tied to goals (e.g., setting target cost) should be specified in legal forms that could incentivize collaboration on the specific projects.

VA2

Liability waiver

SpawGlass Austin Regional Office

Appropriate liability waivers can positively affect the relationship between contracting parties and help resolve the dispute.

VA3

Integrated project insurance

Cathedral Hill Hospital

Integrated project insurance specific to the project should be used in the case of unbearable project loss that the relevant participants are not able to cover.

VA4

Financial transparency

MERCY & Schiller Remodel

Fiscal transparency (no hidden profits, contingencies, or allowance) can be accepted and should be achieved by open-book documentation and reporting.

VA5

Early involvement

Autodesk, Inc.

Key project stakeholders should become involved in the project early on—even without the contract in place—for achieving collaborative attitudes and improving accuracy in estimating.

VB1

Full-time staffing

Edith Green Wendell Wyatt Federal Building Modernization

To increase the efficiency of problem VB2 solving, investment should be made to support full-time staffing.

Intensified planning

Sutter Health Fairfield Office Building

The time-consuming process of intensified planning and team building to reach the aligned goals is worthwhile.

VB3

Integrated group building

Cardinal Glennon Children’s Hospital Expansion

A layered interdisciplinary team (e.g., cluster group) with openminded members should be built up to ensure cross-collaboration and coordination between groups.

VB4

Collaborative decision making

Walter Cronkite School of Journalism

Increased number and frequency of meetings is necessary to deal with problems and assist collaborative decision making.

VB5

Managerial

(continued)

84

Incentivizing Collaborative BIM-Enabled Projects

Table 19.  Measurement of IPD acceptability. (continued) Notation in the model

Categories

Dimensions

Representative case

Measurement

Technological

Colocation working

UCSF Mission Bay Medical Center

Colocation working has a positive effect on the BIM-enabled project in general.

VC1

Necessity of BIM

St. Clare Health Center

BIM is a necessary tool for effective sharing of information in an integrated project team.

VC2

Lean construction

Sutter Health Fairfield Office Building

More lean construction techniques (e.g., last planner system and target value design) should be applied in project implementation.

VC3

Standardized documentation

Cathedral Hill Hospital

Project documents should be standardized to facilitate sharing/ transferring between project parties.

VC4

Information-sharing platform

UCSF Mission Bay Medical Center

An IT platform (e.g., SMART board) should be used to enable information/document sharing in real time between project parties.

VC5

To fully understand the potential impact of BIM utilization on the prospect of IPD, it is essential to include all three constructs in the model. The constructs “collaboration” and “communication” both concern the actual impact of BIM on one of the 145 projects under study in these two aspects, while “incentivization” is evaluated via the respondent’s perception of the need for such an incentivization system against his experience in a BIM-enabled project. This is because, while incentivization measures are not widely adopted in practice yet, their significance for efficiency improvement is well acknowledged in recent procurement reform (e.g., HM Treasury, 2013), and thus the demand for incentivization is expected to be a crucial driver for ushering in integrated delivery systems in the future.

Empirical Analysis Data Collection The data used to test the hypotheses were collected via three primary methods: sending the survey link set up on Sojump (a pay-out service similar to SurveyMonkey) direct to 170 BIM professionals (12%); posting

Empirical Investigation – China (1)

85

Table 20.  Measurement of the constructs of “collaboration,” “communication,” and “incentivization.” Notation in the model

Key constructs

Dimensions

Measurement

Collaboration Quality of collaboration

Aligned goals

Team members have reached an agreement on the project goal and cooperate with one another throughout the life cycle.

VL1

Centralized working place

Each project party has worked in a relatively centralized place and organizes regular meetings.

VL2

Multidisciplinary knowledge

Project members have possessed a certain degree of multidisciplinary knowledge and are ready to collaborate with the professionals from different parties.

VL3

Real-time information sharing

The project data were shared in real time among all relevant project parties.

VL4

Accuracy

In the process of transferring information, there was no distortion or incomplete messages that would cause misunderstanding.

VM1

Timeliness

Project-related information could be transmitted in a timely manner through suitable communication platforms.

VM2

Transparency

Team members were fully informed about issues that affect their work, and information was not hidden by any individual or small group of people.

VM3

Initiative in participation

Team members proactively participated in goal-setting activities, and were open to providing/receiving any information or suggestions that might help the other party.

VM4

Frequency of communication

The frequency of communication was high enough to support the daily exchange of working information.

VM5

Monetary reward

Financial rewards can improve the effectiveness of BIM considerably better than nonmonetary rewards.

VI1

Group-based reward

Group-based rewards will work considerably better than personal rewards in incentivizing contractor participation in a BIM system.

VI2

Objective metrics

Objective metrics are considered better than subjective ones as the basis for determining incentive rewards for BIM participants.

VI3

Differentiated weightings to performance

It is necessary to assign different weightings to performance metrics in the determination of incentive rewards for BIM participants.

VI4

Communication Quality of real-time information sharing

Incentivization Strength of motivation for pursuing the interest of the whole project

Linear reward-­sharing A simple, linear reward-sharing rule (e.g., rewards rule linked to a fixed percentage of cost savings) will work considerably better than a more complicated nonlinear reward-sharing rule in incentivizing contractors to contribute to BIM.

VI5

Minimum amount of incentive

VI6

There is a minimum amount of incentive reward that can motivate contractors’ full participation in BIM.

86

Incentivizing Collaborative BIM-Enabled Projects

the online survey link on a social media interest group on Sojump and Wechat (50%); and distributing 30 questionnaires in person (28%). In total, 163 returns were received, 145 of which were complete and able to be used in the analysis. The background of the respondents spans six professions (owner, architect, engineer, general contractor, subcontractor, and consultant) which are representative of the composition of BIM participants in China. The result shows that the vast majority of projects have reached Level 1 (42.8%) and Level 2 (43.4%) with similar proportions, meaning that a managed 2D and 3D environment has been built up using BIM, but Level 3 BIM features (e.g., 4D construction sequencing, 5D [cost information], and even 6D [life cycle information]) are not utilized yet. As revealed in Figure 12, BIM has been applied to various functions in the surveyed projects, more than 80% of which have seen BIM used to assist in design and construction.

Feasibility Concept Standard design Detailed design Implementation documents Procurement Construction Operation 0%

20%

40%

60%

80%

100%

120%

Figure 12.  Project phases supported by BIM.

Reliability and Validity Test First, the Cronbach’s alpha is used as the reliability indicator to check the internal consistency of three constructs. The results show that all possess a score over 0.8 (BIM incentivization perception: 0.80; communication quality: 0.83; collaboration quality: 0.82), indicating excellent reliability. The next step is to examine the validity of these constructs. Figures 13–15 show that the loadings (standard coefficient) of the observable items on the latent variable are all above the acceptable value of 0.5, and the model fit is well achieved compared to the thresholds suggested by Wu (2009).

Empirical Investigation – China (1)

VI1

e1

VI2

e2

VI3

e3

0.69

VI4

e4

0.79

VI5

e5

0.63

VI6

e6

0.60

0.59 Incentivization

Default model Criteria of good fit

X²/Df 1.186 Not significant

87

0.65

P 0.303 P>0.05

RMSEA 0.036 0.5

GFI 0.978 >0.90

CFI 0.994

Figure 13.  The construct validity test—BIM incentivization perception.

Last, the explained variable IPD acceptability passes all the tests excepting the loading of VA5 (open-book finance) on the sub-dimension contractual (Figure 16). Given its importance in the IPD model, VA5 is still kept in the analysis. As for reliability, the alpha scores of three sub-dimensions are all close to the acceptable level (0.79, 0.80, and 0.80, respectively), so no further action was taken.

Communication

Default model Criteria of good fit

X²/Df 2.168 Not significant

0.75

VM1

e1

0.71

VM2

e2

VM3

e3

0.67

VM4

e4

0.74

VM5

e5

P 0.055 P>0.05

RMSEA 0.090 0.5

GFI 0.972 >0.90

Figure 14.  The construct validity of communication quality.

CFI 0.976

88

Incentivizing Collaborative BIM-Enabled Projects

0.68

VL1

e1

0.70

VL2

e2

0.73

VL3

e3

0.82

VL4

e4

Collaboration

Default model Criteria of good fit

X²/Df 1.003 Not significant

P 0.367 P>0.05

RMSEA 0.005 0.5

GFI 0.993 >0.90

CFI 1.000

Figure 15.  The construct validity of collaboration.

VA1

e1

VA2

e2

VA3

e3

0.71

VA4

e4

0.41

VA5

e5

0.59

VB1

e6

VB2

e7

VB3

e8

0.79

VB4

e9

0.56

VB5

e10

0.70

VC1

e11

VC2

e12

VC3

e13

0.71

VC4

e14

0.61

VC5

e15

0.68 0.71

Contractual

0.75

0.86

0.95

0.68

Managerial

0.72

0.95

0.65

Technological

Default model Criteria of good fit

0.68

X²/Df

P

RMSEA

PGFI

GFI

CFI

1.059 Not significant

0.299 P>0.05

0.020 0.5

0.904 >0.90

0.993

Figure 16.  The construct validity of IPD acceptability.

89

Empirical Investigation – China (1)

Path Analysis The purpose of path analysis in SEM models is to test the statistical significance of the effect of explanatory variables (BIM degree, incentivization, communication, collaboration) on the independent variable (IPD acceptability). The first step is to ensure that the chi-square result is not significant through some modifications, including building correlations between the errors of VL1 and IPD management, VL2 and IPD acceptability, VM5 and collaboration, VM1 and VM2, as well as VM3 and IPD management. By way of this process, chi-square to the degree of freedom ratio is improved, indicating that the conceptual model is a good fit to the real data. This is also confirmed in other indicators of the model fit suggested by Browne and Cudeck (1993) (see Table 21). Table 21.  Model fit summary. X2/Df Default model

1.059

Criteria of good fit

Not significant

P

RMSEA

0.299 P . 0.05

PGFI

0.020 ,0.08

0.671 .0.5

GFI

CFI

0.904

0.993

.0.90

After estimation, it was found that the coefficient on each path, except for the one between collaboration and IPD acceptability, is significant as hypothesized (see Table 22). Specifically, a greater extent of BIM application in the project can lead to a stronger appreciation for the significance of incentivization in strengthening BIM participation (H1), and that will eventually translate into support for IPD (H2). If construction professionals recognize the importance of having well-functioning incentive mechanisms in place, it will be more likely for them to accept IPD contracts and their gain/pain sharing arrangements in the future. Table 22.  Path analysis of six hypotheses. Hypothesis

Dependent variable

Independent variable

Estimate

S.E.

H1 accepted

Incentivization

BIM degree

0.124*

0.045

C.R. 2.730

0.006

P

H2 accepted

IPD acceptability

Incentivization

4.284***

1.030

4.160

,0.001

H3 accepted

Communication

BIM degree

0.095**

0.034

2.809

0.005

H4 accepted

IPD acceptability

Communication

2.207**

0.780

2.683

0.005

H5 accepted

Collaboration

BIM degree

0.192*

0.072

2.677

0.007

H6 rejected

IPD acceptability

Collaboration

21.867

1.283

21.455

0.146

Note: ***p , .001,** p , .005, *p , .05. S.E. 5 standard error; C.R. 5 critical ratio; P 5 p-value

90

Incentivizing Collaborative BIM-Enabled Projects

Also, the greater use of BIM in a project can lead to improvements in the quality of both collaboration (H3) and communication (H5). The effect of the BIM degree on communication can work its way to increase IPD acceptability (H4), while this is not the case for the impact of BIM on collaboration (H6). The reason can be investigated through a mediation model (Figure 17). When modeled without including communication, collaboration has a statistically positive effect on IPD acceptability (Wc 5 3.570, p , 0.001). A possible reason why H6 fails is that the two variables are completely mediated by communication (Baron & Kenny, 1986). This conjecture is corroborated by the significance of the coefficient on the paths of collaboration to communication (Wa 5 0.907, p , 0.001) and communication to IPD acceptability (Wb 5 3.193, p , 0.001). This result means that collaboration positively affected IPD acceptability through changing communication rather than affecting it directly.

Communication Wa

Collaboration

Wb

Wc

IPD Acceptability

Figure 17.  The mediation model of communication.

Discussion Technically, BIM can provide a flexible modeling technique to visualize a design idea and store it digitally as parametric objects, which could then be fed into other analyses within the design (e.g., building services simulation) and facilitate collaborative working among project parties throughout the project life cycle. Dissimilar to other information technologies, BIM adoption is an investment decision. From a business perspective, the cost of BIM deployment must be justified by the benefits accrued from it. The sources of benefit discussed in the literature primarily concern the cost savings from early clash detection without .

Empirical Investigation – China (1)

91

paying much attention to the qualitative changes BIM could bring about to the construction industry in the long run. The current study represents the first attempt to take a forward-looking view on the long-term benefit of BIM. It is found that the increasing use of BIM can considerably raise practitioners’ acceptance of key IPD features, which should then translate into support for implementing this system in the future. This finding has a strong implication for BIM implementation policy. As an enabling tool, the realization of BIM’s full potential depends upon the readiness of all parties concerned. To secure BIM readiness, the AEC industry needs to make a lump sum investment in hardware, software, and training at the outset. The worthiness of this investment bears upon how frequently the acquired capability can be reused. In the early stage (Stage I in Figure 18), inhibited by a lack of sufficient evidence in support of its benefit, the employment of BIM is limited to the small group of early adopters. In cash flow terms, the additional cost arising from BIM is high as most AEC companies have to build in-house capability from scratch, which will naturally constrain the possible scope of BIM application in the project. In the environment of projects featured by a web of independent parties (designers, constructors, and suppliers), the benefit of BIM can grow exponentially as its application grows broader (more life cycle stages), deeper (levels of BIM), and more diverse (variety of analysis supported by BIM). As a result, fragmented application of BIM can only realize a small fraction of its potential. The gap in financial feasibility (D in Figure 18) is a fundamental problem hindering the voluntary adoption of BIM. In economic terms, it can be regarded as a case of market failure under which coordination mediated by the price signal cannot occur spontaneously and that gives a rationale for government intervention (Williamson, 1991). This could be the main reason why mandating BIM deployment in public projects is widely embraced as a kick-start strategy by governments. The nature of a government mandate is not much different from regulation as both serve to restrict the range of legal actions for public interests. In recent decades, the pendulum of regulatory philosophies in Europe has swung to risk-based assessment in which the costs of regulation are explicitly evaluated against their benefits (Löfstedt, 2004; Organization for Economic Cooperation and Development, 1997). When applying the same philosophy to the design of BIM mandates, the benefit is significantly harder to evaluate than the cost because the latter involves a direct cash expenditure while the former involves a

92

Incentivizing Collaborative BIM-Enabled Projects

“Pull”

Benefit/Cost of BIM

“Push”

Benefit

Cost I. Early adopter

II. Early majority III. Late majority Development stages

VI. Laggard

Figure 18.  Trajectories of the costs and benefits of BIM deployment.

delayed receipt of benefit. During the development stages, the cost and benefit of BIM deployment will tend to converge as more companies upgrade to “BIM-ready” (see Figure 18). To the left of the point where those two trajectories intersect, the promotion of BIM is primarily driven by “push” forces, such as BIM mandates. After the cost can be covered by the benefit (to the right of the intersection point), then “pull” forces will dominate. It is useful to understand this conversion from the perspective of the famous principal-agent theory (Holmstrom, 1982). In designing an optimal contract, the principal should first ensure that compensation could more than cover the agent’s opportunity cost. This so-called participatory condition can persuade the agent to take part but cannot induce him to exert the best effort. This theory suggests that efficiency can be improved by holding agents accountable for the outcomes of their actions via risk-sharing arrangements. In the promotion of BIM, mandating can “push” some owners to embark on experimentation with the hope of driving industry BIM capability toward greater maturity through a “learning by doing” process. The push force could only make BIM nominally deployed as an enhanced 3D-visualization tool, instead of giving participants strong incentives to explore the potential of BIM. For this reason, after BIM deployment becomes financially viable, the “pull” forces should be considered by way of various incentivization measures (Chang & Howard, 2016). To set the virtuous cycle of BIM implementation in motion, the initial push force is essential. In a BIM mandate, the government generally sets out requirements without providing much information about

Empirical Investigation – China (1)

93

its rationale. A good example is from the UK Government Construction Strategy (Cabinet Office, 2011): The government will require fully collaborative 3D BIM (with all project and asset information, documentation and data being electronic) as a minimum by 2016. (p. 14) In a follow-up report, several benefits were identified for BIM, including reduced life cycle cost, the potential for higher whole-life value, expanded services to clients to raise the quality of their outcomes, enhanced international competitiveness, increased offsite construction, and growing ICT services in construction (Saxon, 2013). This research demonstrates that utilizing BIM could have an additional benefit in raising practitioners’ awareness of the importance of IPD features and helps increase the likelihood of these features being accepted for the same project in the future.

Chapter Conclusions In recent years, BIM has been feverishly promoted by governments throughout the world by issuing mandates to force the adoption of BIM. The justification for these mandates is restricted to current rather than long-term benefits. In addition to BIM, promoting IPD has also attracted considerable government effort (e.g., Cabinet Office, 2014). While IPD is not yet piloted in China, the awareness of its importance has emerged. For instance, more than half of the respondents in Ni and Wang (2015) agreed that there should be a suitable delivery system to support BIM. The statistical analysis of this research shows that potential cost savings aside, BIM could also propel procurement reform in the long run. This finding not only lends empirical support to the BIM mandate in China, but also predicts that the wider application of BIM can facilitate the implementation of integrated delivery in the country. Using the data from 145 Chinese BIM-enabled projects, this research can further probe the channels through which BIM application could have impacted IPD acceptability: First, the firsthand experience of working in a BIM-enabled environment can make practitioners better appreciate the importance of incentivization and that perception can drive the acceptability of IPD; second, observing the positive impact of BIM on communication quality can translate into another drive to support IPD. It is hoped that these robust statistical relationships can spark follow-on research to investigate the benefits of BIM in a wider context.

C H A P T E R

8

Empirical Investigation – China (2) Introduction In recent years, Building Information Modeling (BIM) has risen sharply worldwide as a promising technology for improving efficiency in the AEC industry. Similar to other technologies portrayed by the theory of innovations, the proliferation of BIM could follow an S-curve trajectory with a slow start, an accelerating middle, and a slowdown in the end. Although the notion of BIM has existed for over two decades, the largescale rollout through BIM mandates has happened only recently. In the initial stage, the attention of governments and academia was mainly placed on resolving technical barriers lying in the way of BIM implementation. However, awareness soon emerged that equally important is to get the delivery environment right. For instance, governance issues are featured much more prominently in the UK’s strategy for Level 3 BIM (HM Government, 2015). While our understanding of the characteristics of desirable BIM delivery environments is still at the nascent stage, there is a view that BIM and integrated project delivery (IPD) could reinforce each other (Eastman et al., 2011). At the heart of integrated delivery systems lies a risk-sharing mechanism (Australian Government, 2011). An efficient incentivization system is not only essential for integrated delivery, but also for construction procurement in general (HM Treasury, 2013). In broad terms, incentivization is referred to as the act of employing measures to align the divergent interests of BIM participants (Baddeley & Chang, 2015). Provision of the right incentives requires a system approach to addressing a series of interrelated issues in an integrated way (Chang & Howard, 2016), including target-cost setting, incentive pool funding, risk sharing, and performance measurement. 95

96

Incentivizing Collaborative BIM-Enabled Projects

It could take a long journey for an organization to settle on an optimal mechanism most suitable for its projects. The pace of evolution toward greater BIM maturity varies with the efficiency of learning. The feedback loop is a decisive mechanism for the way a system could evolve (Sterman, 2000). In introducing new information technology (IT) systems, as asserted by the technology acceptance models, the major hindrance stems from user resistance (Howard et al., 2016; Venkatesh et al., 2003). The central proposition of this research is that the outcome of current incentivization practice in BIM-enabled projects could propel the acceptance of advanced incentivization practice in the future while allowing for the influence of behavioral biases. Through a structural equation modeling (SEM) analysis of 223 BIM-enabled projects from China, it is found that this positive feedback loop is achieved through two channels: If incentivization turns out to function well in the current project, this performance could either result in a direct impact on the acceptance of advanced incentivization measures in the future or work its way to strengthen the impact of BIM utilization on project performance, which will then translate into greater acceptance of advanced incentivization measures. This empirical finding has general implications for both governments and academia. Government mandates have been a common strategy employed worldwide to expedite BIM applications. In the evolution process of best BIM practices, it is essential for policymakers to build a deeper understanding of the impacts BIM could generate for the industry as a whole and harness it as a momentum for further development. The justification for these mandates is not as strong as desirable in terms of evidence strength. This research makes two contributions to this issue: First, it affirms that greater BIM utilization in the project can significantly improve project performance; second, the effect of incentivization after accounting for the influence of behavioral biases is conducive to the acceptance of advanced practices needed for driving an integrated system. This finding resonates with early empirical evidence that the benefit of BIM could spill over into the long run by lowering the impediment to usher in IPD (Chang, Pan, & Howard, 2017). As further elaborated in the discussion section, two findings together can furnish important pieces of evidence for the life cycle theory of BIM diffusion. For academia, this research represents a first attempt to explore the effect of behavioral biases on incentivization embedded in the delivery environment of BIM-enabled projects. The identified 10 biases jointly prove to have a significant impact on the effectiveness of incentivization.

Empirical Investigation – China (2)

97

This finding also provides an empirical support to the policy initiative aiming at changing behavior by design (e.g., UK’s Policy Lab).

The Model Conceptually, an incentivization system to the project is similar to an engine in a car. The effectiveness of the incentivization system is a decisive factor for project performance. However, how well incentivization could work does not only depend on the measures taken, but also on the delivery environment in which it is embedded. Actually, from a theoretical point of view, there are two dominant factors (Chang, 2015): The first-­ order alignment should be secured by choosing the appropriate delivery system and the second-order alignment by selecting right incentivization measures (e.g., the risk-sharing ratio; see Chang, 2014d). According to an empirical investigation of UK infrastructure projects, these two factors together can account for 75% of the variations in project outcome (Chang & Mills, 2016) and thus they could be the most important pillars upon which the framework can rest. Instead of directly modeling them, the strategy chosen was to measure the result of these two factors. In other words, the delivery system and incentive system in the project are treated as exogenous (Figure 19) and take the result as input to the model. The reason for choosing this strategy was based on the 10 case studies conducted for Chinese BIM-enabled projects prior to the survey. It is found that few explicit financial rewards were used, but a broad range of intangible factors could translate into long-term financial benefits to motivate BIM participation. China has a unique business environment with a high diversity of local practices, so discovering the comprehensive list of motivators and their effects on BIM participation could entail another research. To make the analysis manageable, this research chooses to focus on the effect of incentivization created by a project’s delivery environment instead of the measures taken. The exogeneity of these factors is indicated by dotted lines in Figure 19. This effect could be affected by behavioral biases (H1). The questions concerning behavioral biases in the questionnaire are all phrased in a way to make the high score representative of a bias source’s strong positive impact on the effectiveness of incentivization, so the presence of (positive) biases could strengthen the effectiveness of incentivization (H2). The effect of these factors could work forward to influence the effectiveness of BIM on project performance (H3), and the acceptance of advanced incentivization features (H4). In the meantime, the experience with incentivization in

98

Incentivizing Collaborative BIM-Enabled Projects

Delivery Systems

Incentivization Systems

Effectiveness of Incentivization in the Project

H1

Behavioral Biases

H2 H5

BIM Maturity H3 H4

Project Performance

Default model Criteria of good fit

X2/Df

P

2.279

0.000

Not significant

P.0.05

Perception of Advanced Incentivization Features

RMSEA 0.076 ,0.08

PGFI

GFI

CFI

0.528

0.937

0.979

.0.90

.0.90

Figure 19.  The model.

the project could have a direct impact on BIM experts’ perceptions of the essential characteristics of an advanced incentivization system.

Influence of Behavioral Biases The impact of behavioral biases on the effectiveness of incentivization in the context of BIM-enabled projects is not explored in the literature yet. At the outset of the research, it was not certain which biases could be significant and to what extent. Individual behavior could be affected by motivational biases and cognitive biases (Baddeley et al., 2004). The former can be controlled using incentives, while the latter are not under conscious control (Baddeley, 2013). The following analysis takes account of both types of biases. In the face of the lengthy list of biases, a two-stage process was designed to sift out potential behavioral biases to sharpen research focus. First, a pilot survey was conducted to solicit 10 experts’ assessments of the relative importance of the biases identified in the literature, the result of which make up the initial list of biases used in the survey. Through the confirmatory factor analysis, 14 biases were trimmed down to 10. The list of biases considered at each stage is shown in Table 23.

Empirical Investigation – China (2)

99

Table 23.  Selection of behavioral biases. Item No.

Initial behavioral biases included in pilot survey

Selected by professionals in pilot survey

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 Total Items

Ambiguity effect Anchoring effect Attentional bias Availability cascade Backfire effect Bandwagon effect Belief bias Choice-supportive bias Comfort zone effect Curse of knowledge Framing effect Present bias Sunk cost fallacy Loss aversion Neglect of probability Omission bias Optimism bias Overconfidence effect Planning fallacy Projection bias Status quo bias Opportunistic behavior Zero-sum heuristic Pseudocertainty effect Egocentric bias Self-serving bias Extrinsic incentives bias Free-rider problems Paying too little Paying too much 30

Ambiguity effect

Confirmed by the main survey in SEM analysis

Availability cascade

Availability cascade

Bandwagon effect

Bandwagon effect

Comfort zone effect

Comfort zone effect

Present-bias

Present-bias

Loss aversion

Loss aversion

Optimism bias

Optimism bias

Status quo bias Opportunistic behavior

Status quo bias Opportunistic behavior

Egocentric bias Self-serving bias Extrinsic incentives bias Free-rider problems Paying too little Paying too much 14 (25 and 26 combined into a single item in the main survey)

Egocentric bias & self-serving bias

Free-rider problems

10

Present bias results from the differential preference between the benefits immediate to the present and those realized later (O’Donoghue & Rabin, 1999). As an emergent technology, the evidence in support of BIM application is relatively scant. The potential advantage of possessing BIM capacity in gaining competitive edge may not be as highly valued as it should be by practitioners. The tendency to underestimate the longterm benefits of investing in BIM could negatively impact the incentivization effect.

100

Incentivizing Collaborative BIM-Enabled Projects

The next bias stems from availability cascade, which is defined as “a self-reinforcing process of collective belief formation by which an expressed perception triggers a chain reaction that gives the perception increasing plausibility through its rising availability in public discourse” (Kuran & Sunstein, 1999). In the early stage, BIM is not well known. Governments could expedite its adoption by sponsoring promotional activities, so the positive image of BIM can cascade down the industry with the effect of raising its publicity and endorsement among practitioners. When BIM becomes an often ­discussed topic, people can be easily influenced by informational cascade (Anderson & Holt, 1997). In the meantime, the increasing recognition of BIM benefits in the industry could lead to reputational cascade (Bikhchandani, Hirshleifer, & Welch, 1998). Both of them could drive one to follow the industry view on BIM with the effect of increasing BIM participation. The third bias results from the differential mental impact of financial gain versus financial loss. In expected utility theory, the same amount of gain and loss has an equal impact on a decision maker’s utility and thus the course of action taken. Yet, from the perspective of managers’ views of their own behavior, the size of stake involved by far outweighs the probability distribution of outcomes in making risky decisions (March & Shapira, 1987), implying that, in the face of risk, loss aversion can better capture risky decision-making behavior than risk aversion (Kahneman & Tversky, 1979). The preference for loss avoidance may militate against the functioning of risk-sharing arrangements. The fourth bias arises from one’s tendency to maintain the status quo (Samuelson & Zeckhauser, 1988). It means that in making risky decisions, one tends to compare alternatives with a default option. The cost of switching to an alternative could be overemphasized and thus hold back the momentum to change (Just, 2014). The next bias is associated with one’s prior experience with working under an incentive contract. The familiarity with risk-sharing arrangements can create a psychological comfort zone effect, which can strengthen the effectiveness of financial incentives. Bias may also arise from over-optimism in cost estimation (Flyvbjerg, Garbuio, & Lovallo, 2009). Lack of a reliable target cost could weaken the effectiveness of incentivization. The advanced level of BIM is meant to serve as a platform to allow coordination to go across parties. The working of an incentivization mechanism should build upon the appropriate recognition of each party’s credit. However, one could tend to over-claim the contribution he deserves

Empirical Investigation – China (2)

101

(Ross & Sicoly, 1979). This so-called egocentric bias may arise out of four reasons: selective encoding and storage, differential retrieval, informational disparities, and motivational influences (Ross & Sicoly, 1979). The presence of this bias could reduce the credibility of an incentivization system and thus one’s aspiration to make BIM work to its full potential. Relatedly, another inhibitor to group-based compensation is one’s tendency to free ride on another party’s effort (Alchian & Demsetz, 1972). Different from egocentric bias, free riders occur mainly out of economic motives to save on cost. This tendency could hamper the motivational effect of incentivization on BIM participation. The last bias considered is the bandwagon effect. While conceptually this psychological effect can be interpreted as a consequence of information cascade, in the current research, two terms are meant to capture different effects. The major impact of availability cascade lies in lowering BIM users’ resistance to employing new tools when BIM is pushed through from the top down. By contrast, the bandwagon effect tries to capture the momentum created by the increasing capability of the industry as a whole in harnessing BIM functions. Similar to other technologies, BIM could proliferate following an S-curve. The speed could pick up after BIM gradually rises as a standard tool for practitioners. The initial “push” drive from government mandates will turn into a “pull” drive by which voluntary adoption of BIM becomes prevalent as a result of its perceived usefulness. The increasing strength of this pull drive can improve incentivization effectiveness. In the model, these 10 biases are treated as indicator variables to a latent construct, behavioral biases. By reference to theory, the way these variables are measured in the questionnaire is to make high-response scores reflective of the positive contribution of these indicator variables to incentivization. Considering the effects of possible bias sources, we hypothesize: Hypothesis 1 Behavioral biases can positively impact the effectiveness of incentivization in the BIM-enabled project.

Linkage Between Incentivization, BIM Maturity, and Project Performance and Acceptance of Advanced Incentivization Features In the current research, incentivization is a broad term that refers to the motivational effect of the delivery environment on BIM participation, so it includes both financial and nonfinancial, as well as short-term and long-term, benefits for deploying BIM. The majority of the countries are

102

Incentivizing Collaborative BIM-Enabled Projects

in transition to create an environment favorable to the implementation of BIM. In principle, the greatest potential of BIM is most likely to be realized under an integrated delivery environment (e.g., integrated project delivery [IPD]) where BIM-enabled coordination can go across organizational boundaries driven by a reward structure that ties individual interests to common project goals. The advanced features portrayed in Chang and Howard (2016) entail a system engineering, and it could be more feasible to implement them on a piecemeal basis. In the evolutionary process, it is imperative for the government to understand what factors could be significant in propelling the transformation of the delivery environment toward an ideal BIM delivery environment. The benefits of BIM vary widely with the way BIM is deployed in the project, which can be measured in four dimensions: the number of project parties involved, the number of project stages covered, the level of BIM application, and the range of functions assisted by BIM. The maturity of BIM deployment in the project can then be evaluated against these dimensions. In other words, the greatest maturity is achieved when all project parties are enabled across all project stages by Level 3 BIM that provides support to all essential functions required in the design and construction of projects. Metaphorically, incentivization serves as an engine for driving BIM participation, so it is expected to increase the effect of BIM on project performance. For these reasons, we hypothesize: Hypothesis 2 The strength of incentivization created by the delivery environment can positively impact the utilization of BIM in the project. Hypothesis 3 The greater utilization of BIM in the project can improve project performance. In the implementation of new technologies like BIM, there are several barriers. According to the famous technology acceptance model (Davis, Bagozzi, & Warshaw, 1989; Howard et al., 2016), users’ perceptions of a new IT system (e.g., usefulness and ease of use) play a pivotal role in the determination of its implementation result. A central proposition of this research is that the greater the extent of BIM application in the project, the greater BIM potential can be reaped. An integrated delivery system that can facilitate coordination across parties and time

Empirical Investigation – China (2)

103

is the most desirable environment for BIM. Integrated delivery systems can serve this purpose well. The working of these systems is primarily driven by an incentivization system to link up the interests of participating parties with the project objectives. Whereas the significance of incentivization is acknowledged in procurement guides, there is no guidance concerning its optimal design. IPD is one of the most implemented integrated systems. In a prior study, Chang and Howard (2016) developed a conceptual framework for guiding the design of a BIM incentivization system based on the practices reported in IPD case studies (American Institute of Architects, 2010b; Cohen, 2010) and incentive theory. Under an integrated delivery system, incentivization is provided through a gain/pain mechanism (Love, Davis, Chevis, & Edwards, 2011), which involves a series of questions: (1) On what basis is compensation awarded? (2) What weightings are assigned to objective and subjective evaluation awarded? (3) How does one choose the reward-sharing rule (linear versus nonlinear)? (4) How should the threshold value for each incentive award band be set? In a survey of 368 BIM experts from advanced economies (mostly from the U.S. [36%], the UK [18%], European Union [12%], and Australia [4%]), there is general agreement in statistical terms concerning the answers to these questions: 1. Group-based rewards will work considerably better than individual rewards in incentivizing participation in BIM systems. 2. Objective metrics are considerably better than subjective ones as the basis for determining incentive rewards for BIM participants. 3. A simple, linear reward-sharing rule (e.g., reward linked to a fixed percentage of cost savings) will work considerably better than a more complicated nonlinear reward-sharing rule in incentivizing contractors to contribute to BIM. 4. An incentive reward should exceed a threshold level to have a motivational effect on contractors’ BIM participation. In the model of Figure 19, the dependent variable is to what extent Chinese BIM practitioners have come to realize the desirable features of BIM incentivization systems for future BIM-enabled projects in China. By taking as a reference point the characteristics of BIM incentivization systems preferred by international BIM experts, it is possible to measure the awareness of Chinese BIM practitioners in this regard and use

104

Incentivizing Collaborative BIM-Enabled Projects

it as a predictor for the level of user acceptance for working under a well-incentivized BIM delivery environment. As a result, we hypothesize: Hypothesis 4 The performance of a BIM-enabled project can positively affect the awareness of advanced incentivization measures needed for propelling effective BIM participation. Hypothesis 5 The effect of incentivization in the current project has a direct positive impact on the awareness of advanced incentivization measures needed for propelling effective BIM participation.

Result Summary Statistics The data were collected through several channels in August 2016: First, the survey link set up on Sojump (a pay-out service like SurveyMonkey) was distributed to 182 professionals directly by WeChat and email; second, the online survey link was also posted on the professional BIM interest group on Sojump, Tencent, and Wechat; third, 28 paper questionnaires were distributed directly to BIM experts in China. In total, 258 responses were received, 223 of which were complete and used in the analysis. As shown in Table 24, the respondents were requested to answer survey questions on a 1–7 Likert scale about a BIM-enabled project in which they were recently involved. The respondents have a good mix of work experience and professions. The sample projects also span a broad range of capital sizes (see Table 24).

The SEM Model In recent decades, SEM has been popularly employed in the study of human behavior in social sciences as it allows a latent variable to be measured by multiple observable variables. As seen in Figure 19, several constructs (e.g., behavioral biases, project performance) considered in the current research contain multifaceted dimensions, so SEM is a suitable analytical tool. The implementation of SEM typically takes two steps: Specify the relationships among observable variables underlying each of the latent variables and then a structural model for the relations between

Empirical Investigation – China (2)

105

Table 24.  Composition of the data. Variables

Category

Work experience (years)

Roles of project participants

Response Count

Percentage

1–2

40

17.94%

3–5

82

36.77%

6–10

56

25.11%

11–20

31

13.90%

.21

14

Owners

62

27.8%

Architects/designers

109

48.9%

Engineers/specialists

21

9.5%

Contractors

26

11.6%

Others Project size (contract value) (in million RMB)

6.28%

5

2.3%

,30m RMB

43

19.28%

30m RMB–100m RMB

52

23.32%

100m RMB–1000m RMB

68

30.49%

1000m RMB–5000m RMB

36

16.14%

.5000b RMB

24

10.76%

RMB 5 Renminbi

latent variables. These two steps have to call on a different set of statistical tools (i.e., confirmatory factor analysis [CFA]) (Jöreskog, 1963) and path analysis (Wright, 1934). Validity and Model Fit

The robustness of a statistical analysis must build on the validity and reliability of the observable variables. The purpose of the CFA is to check if the sample variance-covariance data are well fit to the specified model. Each fit index has its limitation, so it is a common practice to report a set of complementary indices whereby the audience can evaluate the result from a different angle. There are three types of fit indices (Hair, Ringle, & Sarstedt, 2011): absolute fit, incremental fit, and parsimonious fit. First, absolute fit indices show how much improvement can be made by the theoretical model in comparison to no model at all. Through the x2 and its p-value, researchers can check whether the null hypothesis can be accepted that the sample covariance matrix is equal to the fitted one. A good fit model can lead to a p-value more than 0.05 (i.e., accepting the null hypothesis), so this is why the x2 statistic is also known as a “badness of it” measure (Kline, 2016). Since x2 increases with the sample size, it is usually reported as a ratio to the degree of freedom (x2/df), which should

106

Incentivizing Collaborative BIM-Enabled Projects

not exceed 3 (Kline, 2016). x2 has another limitation: The assumption of multivariate normality behind this index could result in the rejection of a well-specified model (McIntosh, 2007). As a result, RMSEA (root mean square error of approximation) and GFI (goodness of fit index) are also reported. The former can reveal whether the theoretical model is parsimonious enough (RMSEA,0.08) (MacCallum, Browne, & Sugawara, 1996), while the latter measures the extent of variance that can be explained by the model (GFI.0.9) (Shevlin & Miles, 1998). The second category of indices helps researchers compare a model’s fit against a baseline model, which assumes that all variables are uncorrelated. According to a common choice, comparative fit index (CFI), a value of less than 0.9 can detect a poorly specified model (Hu & Bentler, 1999). The third type of index examines whether a model is accepted as a result of including unnecessary variables. The parsimony goodness-offit index (PGFI) is calculated based on the GFI by adjusting for the loss of degrees of freedom, so it penalizes model complexity (Mulaik et al., 1989). There is no consensual threshold level for this statistic, so this index should be read in conjunction with other indices. The constructs considered in the model (see Figure 19) are measured by indicator variables (see Table 25 for details). The analysis of these constructs is explained below: 1. Behavioral biases In the second-stage selection process, four biases were excluded because their parameters were not significantly different from zero. The measurement model of behavioral biases is modified by linking the residuals between (1) present bias and availability cascade, (2) loss aversion and status quo bias, (3) status quo bias and comfort zone effect, and (4) optimism bias and opportunistic behavior. These links were connected mainly because these biases are conceptually related. After these modifications, all fit indices are satisfactory except the p-value (see Figure 20). As discussed previously, x2 could blow up when the sample size is over 200, so this problem is discounted. 2. BIM maturity The first independent latent variable is to measure the extent of BIM being deployed in a project. This research considers four dimensions in measuring BIM maturity: the number of parties involved in BIM, the number of project stages where BIM is

Empirical Investigation – China (2)

107

Table 25.  Description of variables. Variable name

Notation Representation

Description

Incentivization effect

INBI

Incentivization effect

Project members were strongly incentivized to participate in BIM implementation in the project.

BIM maturity

BIPA

BIM participant parties

Number of parties involved in BIM

BIST

BIM implementation stages

Number of project stages where BIM was deployed

BIFC

BIM-realized function

Number of functions BIM is used to support

BILVS

BIM level

Level of BIM

INIG

Group-based reward

Group-based rewards will work considerably better than individual rewards in incentivizing contractor participation in BIM systems.

INOS

Objective standard

Objective metrics are considerably better than subjective ones as the basis for determining incentive rewards for BIM participants.

INLN

Linear reward method

A simple, linear reward-sharing rule (e.g., reward linked to a fixed percentage of cost savings) will work considerably better than a more complicated nonlinear reward sharing rule in incentivizing contractors to contribute to BIM.

INTH

Threshold for minimum incentive pay

There is a minimum amount of incentive reward that can motivate contractors’ full participation in BIM.

BEPB

Present bias

Lack of knowledge concerning BIM’s long-term benefits does not make “education and training” critical in the incentivization systems.

BEAC

Availability cascade

The public promotion of BIM benefits can significantly facilitate the employment of BIM.

BLLA

Loss aversion

The project parties do not have a stronger preference for avoiding losses than acquiring the opportunity to obtain the same amount of gains.

BRSQ

Status quo bias

The contractor’s preference for maintaining the status quo does not reduce the motivational effect of risk sharing.

BRCZ

Comfort zone effect

The contractor/designer’s prior experience with risk-­sharing arrangements can make them feel more comfortable in working under an incentivization system.

BTOB

Optimism bias

The client’s tendency to underestimate the target cost will not hurt the working of incentivization systems.

BROB

Opportunistic behavior

One’s tendency to leverage the advantage from asymmetric information will not hurt the operation of incentivization systems.

BRES

Egocentric bias/self-serving bias

One’s tendency to overclaim credit and evade blame does not hurt the working of incentivization systems.

BFR

Free-rider problems

When compensation is tied to group performance, the free-riding behavior will not hurt the operation of incentivization systems.

BBE

Bandwagon effect

When BIM implementation reaches a certain scale in the AEC industry, the “bandwagon effect” will prompt the voluntary adoption of BIM.

Perceptions of incentivization measures

Behavior bias

(continued)

108

Incentivizing Collaborative BIM-Enabled Projects

Table 25.  Description of variables. (continued) Variable name

Notation

Representation

Description

Project performance

OKPI1

Time

The client was highly satisfied with the performance of this project in terms of schedule management.

OKPI2

Cost

The client was highly satisfied with the performance of this project in terms of construction cost.

OKPI3

Financial performance

The client was highly satisfied with the financial performance of this project in terms of the project NPV (IRR).

OKPI4

Safety control

The client was highly satisfied with the safety record in this project.

OKPI5

Environmental performance

The client was highly satisfied with the environmental performance of this project in terms of the environment impact assessment (EIA) scores.

SKPI1

Quality

The client was highly satisfied with the performance of this project in terms of construction quality.

SKPI2

Innovation

The client was highly satisfied with the innovativeness of this project embodied in the design scheme and construction technology employed.

SKPI3

Participant satisfaction

Stakeholders (clients/end users/design team/ construction team) were highly satisfied with their involvement in the project.

SKPI4

Long-term relationship

Participants (stakeholders) have built up reputations from the project and enhanced the relationships with the client, which may lead to repeat business or long-term partnerships.

SKPI5

Participant collaboration

With the assistance of BIM, participants have perceived a strong tendency among project members to work toward shared goals rather than individual interest in the design and construction process.

deployed, the number of functions BIM is used to support, and the level of BIM. With all factor loadings above 0.5, the measurement model has a good fit. Other model fit indices appear strong, so all these indicator variables are kept to capture different aspects of BIM maturity in the project (Figure 21). 3. Project performance The original construct of project performance does not have a good model fit. With the guidance of modification indices, covariances are added onto the model between four pairs of the constructs: cost and time, time and project financial performance, time and project financial performance, as well as long-term relationships and collaboration. Conceptually, these performance metrics are related and thus can be linked up.

Empirical Investigation – China (2)

BEPB

e2

BEAC

e3

.50

BLLA

e4

.50

BRSQ

e5

BRCZ

e6

BTOB

e8

.78

BROB

e9

.72

BRES

e10

BFR

e11

BBE

e12

.54

.58

.62

.23 .22

.72 Behavioral Biases .78 .71

.69

Default model Criteria of good fit

X²/Df 2.279 Not significant

P 0.000 P>0.05

.36

RMSEA 0.076 0.90

CFI 0.979 >0.90

Figure 20.  Construct validity and reliability test for behavioral biases.

.61

BIPA

e1

.76

BIST

e2

.89

BIFC

e3

.65

BILVS

e4

BIM Maturity

Default model Criteria of good fit

X²/Df 0.094 Not significant

P 0.759 P>0.05

RMSEA 0.000 0.90

Figure 21.  Construct validity and reliability test for BIM maturity.

CFI 1.000 >0.90

109

110

Incentivizing Collaborative BIM-Enabled Projects

.75

OKPI1

e1

OKPI2

e2

.35 .29 .35

.82

OKPI3

e3

OKPI4

e4

OKPI5

e5

SKPI1

e6

.85

SKPI2

e7

.82

SKPI3

e8

SKPI4

e9

SKPI5

e10

.79 .84 Project Performance

.85 .88

.76 .68

Default model Criteria of good fit

X²/Df 2.279 Not significant

P 0.000 P>0.05

.31

RMSEA 0.076 0.90

CFI 0.979 >0.90

Figure 22.  The construct validity of project performance.

After modifications, both the model fit and loadings turn out to be good (Figure 22), so this is a valid construct for analysis. 4. Perception of advanced incentivization features The most favorable incentivization system could share some common characteristics. What these constructs capture is the extent to which Chinese BIM experts agree to the four key features that an effective BIM incentivization system should possess (Figure 23). Path Analysis

The estimation result of the structural model is presented in Figure 24 with standardized regression weight and key statistics. This so-called path analysis is aimed to demonstrate the statistical links between a latent variable identified by theory. The result (reported in Table 26)

Empirical Investigation – China (2)

.64

INIG

e1

.62

INOS

e2

.77

INLN

e3

.79

INTH

e4

Perception of advanced incentivization features

Default model Criteria of good fit

X²/Df 1.987 Not significant

P 0.137 P>0.05

RMSEA 0.067 0.90

111

CFI 0.993 >0.90

Figure 23.  Construct validity and reliability test for perception of incentivization.

is based on maximum likelihood estimates. All paths have a positive loading with a level of significance higher than 95%, which means all hypotheses are accepted.

Discussions BIM is a “scalable” technology for projects because the benefit of BIM increases considerably with the level of its application, the scope of project parties involved, the range of stages it is deployed to, and the number of functions it supports. The realization of BIM’s full potential requires a desirable delivery environment. Conceptually, BIM is a technological enabler, so the deployment strategy of BIM needs to take into account high-level factors such as constraints. Transaction cost theory of construction procurement suggests that procurement systems are chosen in accordance with transaction attributes (Chang & Ive, 2007; Ive & Chang, 2007). The first-order delivery system alignment should be supported by the second-order incentive refinement to achieve the best project outcome (Chang, 2015). Achieving two-level alignments involves a natural selection process (Williamson, 1996) by which misaligned governance structures can be gradually transformed into a more efficient form. What role could BIM play in the evolution of delivery systems? Given the current trend, a host of integrated procurement systems have been promoted in advanced economies (e.g., integrated project delivery in the U.S., two-stage open book in the UK, and the alliancing model in Australia). In the meanwhile, BIM is also heavily advocated by various government

112

Incentivizing Collaborative BIM-Enabled Projects

Chi square=777.857 p-value=.000 degree of freedom=365 Chi square/degree of freedom=2.131 e11 PGFI=.689 GFI=.822 AGFI=.787 CFI=.899 RMSEA=.071 Effectiveness of incentivization in the project

e13

.35 .29

.35

BIST

e15

BIFC

.91

e16

BILVS

.65

e17

OKPI1

e18

OKPI2

e19

OKPI3

e21

OKPI4

.84

OKPI5

.85

e23

SKPI1

e24

SKPI2

e25

SKPI3

e26

SKPI4

e27

SKPI5

Default model Criteria of good fit

Figure 24.  Path model.

.74

.75

e1

BEAC

e2

BLLA

e3

BRSQ

e4

BRCZ

e5

.77

BTOB

e6

.70

BROB

e7

BRES

e8

BFR

e9

BBE

e10

INIG

e30

INOS

e31

.69

INLN

e32

.80

INTH

e33

.52 .50 .59

.57

Behavioral Biases

.19

.74

.78

.60

e14

e22

.31

BIPA

e12

BEPB .56

.71 BIM Maturity

.68 .67

.33

.81 .79

.88

.63 Project Performance

.36

.85 .82 .78

e28

Perception of advanced incentivization features

.58

e29

.69

X²/Df 2.131 Not significant

P 0.000 P>0.05

RMSEA 0.071 0.90

CFI 0.899 >0.90

.61

.23 .19

.37

Empirical Investigation – China (2)

113

Table 26.  Estimates of the path model. Path

Estimate

S.E.

C.R.

P

1.053

.156

6.752

***

.020

.008

2.625

*

2.262

.536

4.223

***

Incentivization effects

← Behavior bias

BIM maturity

← Incentivization effect

Project performances

← BIM maturity

Future incentivization perceptions

← Incentivization effects

.387

.045

8.508

***

Future incentivization perceptions

← Project performances

.293

.054

5.407

***

Note: *** p,.001, ** p,.005, * p,.05. All the paths are significant.

digital programs (e.g., UK’s Digital Built Britain). This research evinces that these two driving forces for construction productivity improvement are intertwined and can reinforce each other. The value of this empirical finding can be appreciated from the perspective of a life cycle theory of BIM diffusion. Initially, the proliferation of BIM could be held back by lack of a strong financial case (indicated by Δ in Figure 18) because the high setup cost of BIM (e.g., expenses on hardware, software, training) cannot be justified by the limited additional benefit that BIM can bring, particularly in an ill-fit delivery environment. The shortfall in net benefit makes it unlikely to see a large-scale spontaneous adoption of BIM, which explains why government mandates are upheld worldwide as the primary strategy to accelerate BIM uptake. A mandated implementation could face two levels of resistance: interest misalignment among parties and the ill-fit delivery environment. The latter presents a much higher barrier than the former in implementation. Mandatory BIM adoption could facilitate the acceptance of integrated delivery systems in the long run as a result of BIM users’ direct experience with the practical need for incentivization measures and the BIM-enabled improvement in communication quality. However, ushering in integrated delivery systems may take more time in some countries as it may require new legislation. Comparatively, the implementation of incentivization measures commands greater flexibility. However, it may encounter user resistance as a result of low perceived usefulness (Davis, 1989). The current research demonstrates that the effect of a project’s delivery environment on BIM incentivization can work its way through two channels to change the perceived usefulness of advanced BIM incentivization measures: First, the strength of tangible and intangible incentives embedded in the project delivery environment has a significant effect on the maturity of BIM

114

Incentivizing Collaborative BIM-Enabled Projects

implementation in the project, which can then translate into an impetus to improve project performance. The effect of this chain effect on lowering the barriers to advanced incentivization measures provides a self-sustaining driver for BIM proliferation (indicated by the smaller circle loop in Figure 2). The improved incentive alignment can grow into a momentum to propel high-level delivery environment reform (indicated by the larger circle loop in Figure 2). In current practice, the government should take a risk-based approach to evaluating the pros and cons of a mandate (Löfstedt, 2004; Organization For Economic Cooperation and Development, 1997). The finding of this research can be drawn upon as post-implementation evidence in support of these mandates.

Chapter Conclusions The benefit of BIM is sensitive to the way it is implemented in the project, and its potential cannot be fully reaped in a delivery environment where project parties are not well incentivized to harness BIM in improving communication and coordination across project stages. However, the employment of integrated systems could present a great challenge in some countries as it may require new legislation and heavy training. It is easier to achieve the second-order incentive alignment through a piecemeal implementation of incentivization measures. The data collected from 223 Chinese BIM-enabled projects reveal that the perception of actual incentivization created by the project delivery environment has a significant impact on the perceived usefulness of advanced incentivization systems. The establishment of this awareness could herald a smoother process when introducing these systems into BIM-enabled projects. Embedding BIM in a better-incentivized environment provides a self-sustaining driver for the proliferation of BIM, which will eventually pave the way for the acceptance of integrated delivery systems.

C H A P T E R

9

Case Studies – United Kingdom Problem Identification A central tenet of this chapter is that integrated delivery environments can facilitate coordination across project parties without being constrained by legal or organizational barriers. Procurement integration can take place at either the program level, project level, or both. Since the Latham and Egan reports in the 1990s, the UK government has endeavored to promote the ethos of partnering in construction projects. The successful experience of partnering gradually permeated from the project level to organizational level and evolved from informal arrangements to a formal framework. This evolution has led to the emergence of dual governance. In the current research, dual governance is a term created to describe the organizational arrangement in which the owner enters into a framework contract with its supply chains for governing long-term relationships, while employing one of existing delivery systems (e.g., traditional design-bid-build, design-build, management system) in the execution of individual projects. Integrated delivery systems (e.g., integrated project delivery, alliancing model) can fit into this definition if they are applied to a multiyear, multiproject program. Several delivery systems of this nature are in operation in the UK (e.g., Transport for London’s STAKE, Anglican Water’s @one alliance, and Environment Agency’s Water and Environment Management [WEM]). Driven by the UK central government to reduce the excessively high cost involved in the delivery of UK infrastructure projects (HM Treasury, 2010), these systems are representative of recent procurement innovation in the UK infrastructure sector, so in principle, they could also provide the most 115

116

Incentivizing Collaborative BIM-Enabled Projects

appropriate environment for implementing BIM. It is of interest to explore to what extent these delivery system arrangements could have supported the performance of BIM. In the design of interviews, we are particularly concerned with the way an integrated delivery environment could have affected BIM implementation.

Method As the nationwide rollout of BIM in the UK was still a recent phenomenon, the nature of this chapter’s central question is an exploratory study with the aim of obtaining evidence to deepen our understanding of the interlocking relationship between delivery environment and BIM implementation. This case study is an appropriate method to enable this research to probe the multiple facets of the problem in depth (Baxter & Jack, 2008; Eisenhardt, 1989; Stake, 1995; Yin, 2013). A multiple-­case study methodology was employed to scrutinize two distinct types of large-scale collaborative infrastructure delivery frameworks and enable the interviewees to effectively substantiate their perceptions, experiences, and views of reality through their stories (Baxter & Jack, 2008). Given that problems regarding BIM participation are not well identified, an exploratory case study approach is adopted. Moreover, the main rationale of such an approach is that the effect of economic incentives on the use of BIM is a novel phenomenon. Because previous studies are either conceptual or hypothetical, BIM incentivization remains uncommon, particularly in highly collaborative and complex projects. Therefore, it is reasonable to explore the wider economic aspects, aspiring to wield incentives as an enabler to enhance efficacy in the use of BIM. The nonexistence of such case studies also convincingly makes case study methodology the judicious choice to allow for a comprehensive study of the available cases in one of the leading BIM-implementing countries.

Data Collection A qualitative data-collection procedure was applied in this research. Data were gathered through semi-structured interviews via email and telephone. For both case studies, substantial preliminary data were collected, further online documents supported the delivery frameworks,

Case Studies – United Kingdom

117

and interview participant characteristics assisted in shaping the interviews. The interviews were aimed at: • Chronicling how the collaborative frameworks were established; • Exploring the BIM utilization, collaboration, and incentivization elements; • Examining the structures put in place; and • Understanding stakeholders’ views and perspectives. Because each stakeholder had a distinct experience of BIM from varying perspectives on their undertakings, this lent itself to semi-structured interviews to capture their distinctiveness and commonalities in a given BIM experience within a project environment. All interviews had the same objective, which was to recognize their perceptions, impressions, and opinions for the adoption and implementation of BIM. These interviews enabled the perception of every organization’s experience about the use and implementation of BIM from their practical experience not only in-depth, but also in a broad manner. Data collection commenced with establishing the openly available documents about the case studies (such a general framework information). Open-ended interview questions were prepared and issued to the interviewees in advance of the interviews. The record of all these interviews was examined through thematic analysis. The data collected in the case studies were first analyzed using a qualitative content analysis approach, which involved classifying the data gathered through semi-structured interviews to identify the key motivation drivers. Each case study is an independent study subject to cross-case analysis. Identification and refinement of driver categories were achieved by inductive coding manually. The interviewees were provided with information regarding the context, process, and perceived outcome on the evolution of BIM within the collaborative framework.

Data Analysis Case study 1: Collaborative delivery framework A

Established in 2005 around an alliancing model, framework A is a cooperative organization set up between the owner (one of the largest water and sewerage companies in the UK) and seven key partner organizations (see Figure 25) to deliver the owner’s asset renewal and maintenance

118

Incentivizing Collaborative BIM-Enabled Projects

Framework partners

Partner 1

Partner 2

Partner 3

Owner Partner 4

Partner 5

Partner 6

Partner 7

Extended framework suppliers

Figure 25.  Structure of framework A.

programs collectively. Following the two previous programs, the current five-year program comprises about 800 vast network upgrade and improvement schemes of various levels of complexity worth approximately �1.2 billion, taking place between April 2015 and March 2020. In addition to the design and construction of the wastewater treatment facilities in eastern and northeastern England, the schemes also involve maintenance and improvement of the water mains and sewerage network across the water company’s region. Asset delivery projects under framework A are allocated through a group negotiation process that includes all framework suppliers. Five stakeholders involved in framework A were interviewed. The main characteristics of the interviewees are summarized in Table 27. Case study 2: Collaborative delivery framework B

Following two preceding generations of frameworks, framework B was established by a nondepartmental public water company in 2013 as its third-generation main asset delivery vehicle. The four-year framework is a commercial agreement among the owner, consultants, and contractors (or suppliers) with agreed-upon terms for the award of individual contracts to deliver projects for flood and coastal risk management across the UK. Governed by a partnering approach, framework B is structured into four lots (see Figure 26), with each comprising between three and six consortiums made up of either individual organizations or joint ventures. The framework was structured in such a way that suppliers could be selected from within a single framework to deliver a range of

Case Studies – United Kingdom

119

Table 27.  Interviewee characteristics for framework A. Interviewee 1 Stakeholders Job title Engineering manager Company size Type of firm Tier 1 Location England— international Turnover £92.3 million Number of 1,032 employees Specialization Type of Water resources business engineering consultants

Business relationship Role Partner Years of 11½ years relationship

Supplier 1

Interviewee 2

Interviewee 3

Digital asset creation manager

Engineering manager

Sales manager

Principal design engineer

Tier 1 England— international £92.3 million

Tier 2 Ireland—national

Tier 2 England—regional

Tier 2 England—regional

£43.8 million

£14.4 million

1,032

291

96

 £5 million 55

Water resources engineering consultants

Design, manufacture supply, and install machinery and equipment

Design, supply, and install and commission mechanical and electrical services

Design, manufacture, supply, and install electrical, mechanical, and fabrication engineering services

Partner 11½ years

Supplier 5 years

Supplier 11½ years

Supplier 11½ years

Supplier 2

Supplier 3

Interviewee 4

Supplier 1

Lot 2

Lot 1

Supplier 4

Supplier 5

Interviewee 5

Supplier 6

Supplier 2

Supplier 3

Owner Supplier 1

Supplier 2

Supplier 1

Lot 3

Supplier 3

Supplier 4

Supplier 2

Supplier 3

Lot 4

Supplier 5

Figure 26.  Structure of framework B.

Supplier 4

Supplier 5

Supplier 6

120

Incentivizing Collaborative BIM-Enabled Projects

Table 28.  Interviewee characteristics for framework B. Interviewee 1

Interviewee 2

Interviewee 3

Interviewee 4

Flood and coastal risk management manager

Commercial strategy manager

Director

Water global technical leader—BIM

Type of firm

Water company

Water company

Tier 1

Tier 1

Location

England

England

England—international

England—international

Turnover

Not applicable

Not applicable

£190.1 million

£137.1 million

Number of employees

Not applicable

Not applicable

780

279

Not applicable

Not applicable

Engineering, procurement, Environmental and and construction engineering consulting

Role

Owner

Owner

Supplier (JV)

Supplier

Years of relationship

Not applicable

Not applicable

3 years

3 years

Stakeholders Job title Company size

Specialization Type of business Business relationship

services while giving maximum flexibility regarding how suppliers were engaged. It also allowed flexibility for use in a wide range of contractual approaches for different types of projects. Contrasting with the predecessor frameworks, major asset delivery projects under framework B were allocated through a competitive bidding process that was open to all framework suppliers on the selected lot. Taking inputs from three other lots, the asset delivery projects were commissioned through a single contract approach such as design-build, where each supplier had an integrated supply chain that provided both design and construction services. This contracting approach consolidated responsibilities for design and construction into a single contracting entity in such a way that streamlined task management for the owner. Four stakeholders involved in framework B were interviewed. The main characteristics of the interviewees are summarized in Table 28.

Cross-Case Analysis This section examines the key elements that impact supply chain participation, effort, and performance on BIM across collaborative delivery frameworks A and B. The behavior of the suppliers is investigated in the

Case Studies – United Kingdom

121

Table 29.  Summary of the characteristics of frameworks A and B. Framework A

Framework B

Description

A commercial model that is driven by performance against baselines and enables early integration to maximize input from framework partners and suppliers, selected by cost, capability, and culture

A commercial agreement between the owner, consultants, and contractors (“suppliers”) with agreed-upon terms for the award of individual contracts to deliver projects

Origin

Incremental improvement from the two previous Replacement of the two frameworks program periods of the preceding generations as the owner’s primary asset delivery vehicle

Capital budget (current)

£1.2 billion

£1.0 billion

Contract duration

15 years (since 2015)

4 years (since 2013)

Ownership

Privately held

Publicly held

Collaborative approach

Alliancing

Partnering

Governance structure

Management system

Design-build

Operational location

Colocated

Distributed

Contractual arrangement

Incentive contracting

Incentive contracting

Contract

NEC3 suite

NEC3 suite

Incentive type

Gain/pain share Increased work allocation through outperformance

Gain/pain share Increased work assignment through competition

Supplier selection

Qualifications-based

Price- and qualifications-based

Work allocation

Group negotiation

Mini-competition

Performance evaluation

Integrated performance evaluation

Integrated performance evaluation

BIM initiative

Owner

Owner

context of delivery environments, and the common characteristics across the two frameworks are identified using a cross-case lens, guided by research propositions. The following consists of the cross-case analyses, resulting from discovering the identified case features that are significant contributors to the uptake of BIM among the supply chain. Table 29 presents a summary of the characteristics of frameworks A and B.

Similarities and Differences Regarding similarities, both case studies are selected from the UK water industry and are large-scale collaborative infrastructure delivery frameworks owned by water companies. Both are driven by incentivized commercial models to deliver services across the UK. Both frameworks adopted NEC contracts in contracting with suppliers and functioned at two interdependent levels: the “program” and “project” levels.

122

Incentivizing Collaborative BIM-Enabled Projects

The program level primarily dealt with high-level requirements and procedures of BIM implementation, suppliers’ performance measurements, and incentive plans, whereas the project level dealt with the project-specific requirements, which varied depending on the types of projects. While initiated by owners, the adoption and implementation of BIM, which was of high importance to both framework organizations in achieving digital asset management, was majorly driven by partners and suppliers with strong BIM capabilities. Frameworks A and B varied by the supplier-selection approach. For framework A, partners and suppliers were first selected for the framework through qualifications-based selection, under which suppliers were assessed based on their organizational objectives and cultures in alignment with the owner’s objectives, and also their capabilities and abilities to add value to the framework. However, for framework  B, suppliers were selected through price- and qualifications-based selections, under which suppliers were assessed based on their rates for staff and major construction operations that were used as a cap for each commission of the projects, and also their capabilities and abilities to add value to the framework. In other words, frameworks A and B included exponential measurements based on added value and the percentage of cost savings below a target cost. One of the remarkably significant differences was observed between frameworks A and B with regard to work allocation. For framework A, projects were allocated through negotiations between partners, whereas suppliers were assigned to different projects based on their outperformance and capabilities. For framework B, projects were awarded through competitive bidding among suppliers for each lot. The overall commercial framework A allowed outperformance to be incentivized as part of the multiparty incentive mechanisms, which required overall asset delivery to be managed on a program approach, rather than purely on a project-by-project basis. Furthermore, all partners for framework A operated colocation at the owner’s office, whereas suppliers for framework B operated dispersedly at an individual’s office. According to Dossick et al. (2009), “where delivery methods alone will not address the inter-organisational challenges, strategies such as colocation support a stronger team orientation to the project through informal communication.” While these findings have provided general similarities and differences between two frameworks, they also reveal that framework B was

Case Studies – United Kingdom

123

more traditionally price-focused by the captured competitive element, as compared to framework A.

BIM Utilization As the UK infrastructure industry advanced, the priorities for both framework organizations had started to shift toward achieving a total expenditure, or “totex” approach, under which regulated companies are driven to focus on the total value of an asset over the whole life cycle rather than just over the course of project delivery in formulating corporate strategies. Successfully reduced asset totex was perceived to deliver better life cycle value ultimately. Due to its integrative and interactive capabilities to provide more complete and accurate data, and thus help achieve an efficient asset life cycle management, a fully integrated and interoperable BIM was seen as a solution for facilitating both organizations to develop an advanced data and asset management system, which would then move their businesses into the digital age. Although perceived by tier 2 suppliers to be predominantly benefitting the owner, such an initiative could better align suppliers with the owner’s objectives in BIM utilization, which in turn could benefit their organizations. As one interviewee explained: The collaborative framework has enabled us to see the benefits of BIM, and also the benefits of offsite construction. We have managed to provide the offsite solutions to many of our clients. I see that both owner and we have benefited from the collaborative framework. With regard to integrating BIM into a collaborative environment, the stakeholders of both frameworks adopted the processes and procedures as outlined in PAS 1192 (see BSI [2013, 2014] for definition) produced by BSI Group. Having to put forth more effort into learning how to operate more efficiently with the new processes and techniques in place, it took extra time and effort, and thus was perceived to have created additional work not only by tier 2 suppliers, but also partners and tier 1 suppliers across both frameworks.

Collaborative Working Having to share information across multiple organizations, a consensus between interviewees was that BIM projects were necessarily

124

Incentivizing Collaborative BIM-Enabled Projects

collaborative. However, as stated by one of the tier 1 suppliers from framework B: BIM will only be truly efficient and effective when all levels of supply chain, top and bottom, are involved. At the moment, as a company, and also as the industry, we have not engaged enough with the lower level suppliers or the SMEs. We need to do that a lot more to make sure we are getting information in a digital form in a way that we want it in terms of the level of details and the way we want it. So, we can use it effectively and efficiently. The client also needs to do that as well. This finding reveals that although BIM serves as a collaborative platform for sharing information, supply chain collaboration still appears relatively disjointed and distributed in the industry. It also implies that BIM collaboration requires all stakeholders to recognize the high level of interdependence in achieving positive outcomes, which are the benefits that can result from supply chain collaboration actions. As a goal, interdependence was high in both collaborative organizations, and trust was perceived to be essential, especially with regard to sharing information openly and transparently. In this regard, suppliers for framework A stated that they were willing to share information early with others openly. Nonetheless, there was a perceived risk of adverse outcomes in the long run by one supplier, in which he felt that their ideas would be used by other suppliers. Each framework displayed differing collaborative development solutions and processes toward the collaborative commercial model, which characteristically involved multiple organizations, although both had in common a relationship that sought to stimulate collaboration using the best collaborative tools and techniques such as BIM. Both frameworks A and B demonstrated an approach based on collaboration with not only the top-tier suppliers but all members of the supply chain. Efficiency was perceived to be achieved by aligning organizational behaviors and culture, incentivizing the supply chain against business objectives, and jointly managing program risk over the duration of the program period. Since lower-tier suppliers still regarded BIM as 3D modeling, partners for framework A stated that the organization dealt with such problems through collaborative engagement approaches, such as working

Case Studies – United Kingdom

125

alongside suppliers and providing training, to improve their knowledge and capabilities. This collaborative quality, in turn, enhanced project delivery efficiency. In framework A, for example, collaborative working resulted in their recent high level of efficiency of 38%, which far exceeded the efficiency target of 22.5%. An interviewee explained: When it comes to collaboration, it is all about trust. It is important to encourage right behaviors. We have to make sure that no one takes free rides. A consensus among tier 2 suppliers in framework A was that collaborative working had been beneficial to the development of their organizations and resources. High trust, positive behavior, and right culture were prioritized following identification as the major determinants of collaboration. Moving toward a digital data-exchange platform, suppliers were particularly relied upon by owners for their specialist services and expertise. The findings, however, reveal that the behavior of suppliers was influenced by the establishment of trust and trustworthiness in ongoing and long-term social and economic exchanges between the owner and suppliers. Therefore, a successful BIM project relies highly on effective collaboration across the supply chain. However, these findings imply that without strong relational quality, the impact of collaboration on BIM participation may be decreased.

Early Engagement Associated with NEC3 contracts, both frameworks A and B were prompted to the “early contractor involvement” (ECI) clauses, in which suppliers took part from the outset of project planning and design development processes. Not only did ECI form part of the contractual agreement between owners and suppliers, but it was also comprehensively recognized across frameworks A and B as one of the key approaches to benefit improving the efficiency of delivery, quality of design, and life cycle value of infrastructure assets. Top-tier suppliers observed that early involvement from the lower-tier suppliers was the key fundamental in improving the delivery across the board, and it created a project delivery environment that allowed for working collaboratively and helping each other through all project stages. The majority of the tier 2 suppliers observed that the owner had been very open in encouraging them to contribute in every way as early as possible, which had been very successful

126

Incentivizing Collaborative BIM-Enabled Projects

regarding commitment to robust design. The collaborative framework had also enabled them to see the benefits of BIM. Therefore, collaboration was perceived to have benefited both the owner and suppliers. Given the large-scale, complex nature of projects delivered under frameworks A and B, a consensus of the partners and tier 1 suppliers was that providers, particularly those in lower tiers, should be engaged no later than the design development stage, and at best during the initial design period. The early involvement of suppliers in the design development stage was seen to improve the quality of engineering design and manage the integration of design and construction by providing advice related to buildability and life cycle costing. Apparently, suppliers were selected not only for their lowest-priced deals in the first instance but correspondingly for the value of their specialist expertise and capabilities needed to meet owners’ objectives. Suppliers’ expertise was regarded as necessary for virtual design and construction, which the BIM concept portrays. It was pointed out by partners and tier 1 suppliers across two frameworks that ample innovation potential and added value gained from the involvement of suppliers at an early stage was derived from the lower tiers of extended supply chains. For example, in framework A, such value-adding options included the alternative work package designs coupled with carbon reduction targets that could increase delivery efficiency without increasing delivery time and costs, life cycle costs, or the take-up of industrialization in construction techniques such as offsite construction. This practice evidently implies that the early involvement of lower-tier suppliers raised effectiveness in developing design options. Moreover, the early involvement of lower-tier suppliers was recognized to have allowed early collection of data and generation of information that helped generate detailed solutions for BIM projects. The enrichment of BIM models based on the solutions generated dynamically enabled all stakeholders to visualize and simulate the processes and the built assets and environments, and hence provided owners with a good basis for decision making. Although both frameworks were focused only on asset delivery, the total costs of asset ownership and subsequent operations and maintenance were taken into the context of decision making in response to the totex reduction initiative. These findings, therefore, indicate that the early involvement of suppliers in the design process not only enhances the delivery efficiency of the projects but also optimizes the life cycle value of the assets.

Case Studies – United Kingdom

127

Since BIM’s greatest potential lies in generating information necessary to capture asset life cycle value, it was pointed out by one supplier from framework A that the level of detail ought to be included in a BIM model at an early stage. This is because the late involvement of lower-tier specialist suppliers could potentially lead to unnecessary constraints for suppliers concerning variations caused by information disparity during the onsite installation of required products. The peculiar situation of early supplier engagement, however, occurred in framework B. Due to the project delivery strategy of competitive bidding, lower-tier suppliers in framework B were seen guarding information at an early stage. Such behavior was attributed to the uncertainty of winning the contracts. The resulting situation is contrary to what was perceived by suppliers for framework A as evident below. One interviewee explained: We are happy to share information with others as the projects are allocated to us without going through the competitive tendering process. So, we do not have to worry about sharing designs because we are not competing against each other. These findings imply that competition, although only in a small group of framework suppliers, is much more likely to influence suppliers’ levels of effort and engagement, as well as their willingness to share information. This behavior bias would be associated with a change in project procurement and delivery strategy. It is likely to be more apparent in the case where the owner has a strong focus on lower prices, which in turn leads to strong competition among suppliers. Accordingly, the findings suggest that project delivery systems can affect suppliers’ behavior and effort directed toward early involvement and engagement, which in turn affects the quality of information in BIM models. While competition, direct or indirect, is likely to have been an effective motivator to reduce costs, it clearly does not provide for an efficacious collaborative BIM. Taken together, these research findings, with regard to the early engagement of suppliers, provide evidence that mutual trust, relationships, and supporting processes play a vital role in promoting motivation toward utilizing BIM effectively. The findings also reveal that the greater the level of supply chain capability developed through ongoing training, the greater the impact on BIM participation among lower-tier suppliers.

128

Incentivizing Collaborative BIM-Enabled Projects

Nonetheless, while the early involvement of suppliers was evidently deemed necessary and beneficial for collaborative working, such as in the BIM environment, one of the tier 2 suppliers for framework A stated: . . . it is great and brilliant to do this way, but the only problem is that we did not get paid for it. The above statement raises the issue concerning compensation among lower-tier suppliers.

Shared Learnings, Culture, and Capabilities In framework A, partners were relatively satisfied with other partners’ performances, due to having become familiar with their organizational capabilities and culture, and preferring to have a long-term relationship and collaborate with them within the framework. Partners for framework A stated that the adoption of BIM technologies within an organization should accommodate its impact on people, culture, and the organization’s use of technologies. Every organization had its capabilities and culture based on the skills and competencies of their people and technological advancements used for standard processes. For the implementation of BIM to be successful and make an effective contribution to the implementation of business, it would require that the culture and capabilities of different organizations be aligned with the owner’s objectives. Lower-tier suppliers were satisfied working in a collaborative environment, in which they could share the learnings from the top-tier suppliers.

Incentive Contracting Arrangements The incentivized commercial models with gain/pain share arrangements were adopted in both frameworks A and B. Owing to the characteristics of the gain/pain share arrangements, suppliers were incentivized to deliver measurable savings (gains) and not losses (pains), and thus they were propelled to work more efficiently. One primary difference between frameworks A and B was that the use of financial incentives in framework A was only applied to the seven partners, but not to the extended supply chain. This means the financial incentives were only used with tier 1 suppliers, but not those beyond tier 1. By contrast, the incentive pool for framework B was to be shared by all suppliers on the framework, both top and lower tiers.

Case Studies – United Kingdom

129

In framework A, partners and suppliers were committed to the owner’s 10 key outcomes and worked on a total outperformance basis. They could only be incentivized when they outperformed business plan efficiency targets of 22.5%. While partners were incentivized financially, the incentivization for lower-tier suppliers was made on a project-by-project basis, in which suppliers were awarded for more projects following their outperformance. However, moving toward an integrated approach for alliancing, the organization realized the importance of supplier incentivization. Therefore, tier 2 and 3 suppliers had recently signed the incentive agreements to be involved in the financial incentive-sharing model based on efficiency. As efficiencies made by suppliers would be assessed based on a reasonable baseline matrix, it was necessary for the owner and partners to be willing to allow for full transparency. According to the partners, these arrangements were beneficial to the organization. All suppliers could be standing back to back with the owner’s objectives in achieving the organizationally desired outcome. More important, it was able to facilitate the long-run alignment of suppliers’ behaviors. As the performance evaluation was largely driven by innovation primarily due to increasing organizational commitment to carbon reductions, partners also perceived that the adoption of incentive mechanisms would induce more innovative effort from suppliers. However, such outcomes would rely heavily on strong collaborative relationships, as well as the suppliers’ levels of innovation. One partner further stated that the effectiveness of supplier incentivization would depend on trust, particularly among the new suppliers, whereas such effectiveness would depend on behavioral change among the existing framework suppliers. Target-cost contracts were used in frameworks A and B as they created greater alignment of objectives to achieve efficiency through high performance and innovation. As innovation was associated with cost efficiency, it essentially formed part of the evaluation measures. This, therefore, compelled suppliers for frameworks A and B to produce new ideas and introduce them in a BIM context, which would give suppliers a chance to be allocated more projects in the case of framework A, and to be awarded a contract in the case of framework B. While target-cost contracts were used in collaborative framework arrangements, as stated by one of the suppliers for framework B, there lied the risk of inaccurate cost estimation. While this is seen as the most equitable because both parties share the risk equally, which helps develop partnering behaviors, it was seen to be less likely to encourage

130

Incentivizing Collaborative BIM-Enabled Projects

the contractor to drive up the target cost value or chase compensation events. On the downside, the suppliers may seek to increase their target cost or maximize changes to avoid hitting the pain share cap. Another issue brought up by him was that the suppliers might not be motivated to make savings below the cost target as they would potentially get a reduced return. Clearly, the choice of incentive model will have to take into account the potential behaviors of suppliers that it will drive in the supply chains. The lack of an accurate estimate of project outturn cost could result in difficulties.

Financial Incentives The consensus among partners for framework A and tier 1 suppliers for framework B was that financial incentives would induce suppliers’ effort exertion in developing design solutions, which in turn would improve performance outcomes. One partner considered that financial compensation was one of the ways to objectively measure the success of a project. When asked if their organizations were under strong incentives to support the development of optimal design solutions for the project, a consensus from the suppliers was that they were never incentivized. Therefore, it is not surprising that suppliers were susceptible to the positive effects of financial incentives, as one of the suppliers stated: Monetary compensation will probably encourage us to contribute even more, and this will lead to a better result. It will be in our interest to [use] expertise and to improve designs so that we can benefit from the monetary compensation. Apart from the lack of compensation, such tendencies were due to the higher resources required in terms of time and cost, resulting from their early involvement in design which was often accompanied by uncertainty. Coupled with the inherent demands and complexity of BIM projects, suppliers often needed to spend more time. For example, one supplier had an experience in which he spent additional 400 man-hours during the initial design phase to develop a design solution, for which he was not compensated. Therefore, it is not surprising that the lack of additional incentives for early supplier involvement, which often required BIM, inclined lower-tier suppliers to seek more financial returns. From a behavioral perspective, financial incentives were seen as mechanisms to induce suppliers to exert more effort in exchange for commercial return.

Case Studies – United Kingdom

131

Due to the nature of the procurement strategy employed, the strength of financial incentives in inducing supplier effort was also perceived as necessary in framework B. One supplier considered determinants such as low project overhead and low contingency allowance for such tendencies, particularly with the price-driven contracts. Submitting the lowest bid with the minimum requirements being met would help suppliers win a contract. He also felt that the target-cost contracts made extra effort useless, as they were bound to meet the target cost, which was often set relatively low based on the rates submitted by suppliers. Moreover, there would be no compensation made for any extra work on the design solutions and development of a BIM model. Nonetheless, target-cost contracts could allow suppliers to be more effective in the use of BIM in highly complex projects, in which the outturn cost could be reduced to promote lower bidding. Therefore, cost- or time-reimbursable contracts would be more beneficial to suppliers, for which time and cost spent for added value, such as visualization and animation from the BIM perspective, could be reimbursed.

Relational Incentives One supplier for framework A considered financial incentives relatively minor determinants of effort because the focus of the organization was on developing long-term business relationships with the owner, which was regarded as a prioritized incentive. Correspondingly, although suppliers for framework A were uncertain about the effects of incentive contracting, they perceived such implementation as less significant. They placed a high importance of their intrinsic motivation on building strong relationships with the owner. The strength of relational incentives is evidently revealed by one supplier: Effort contributed is equal because all our customers are important. The strength of relational incentives was seen as less significant for framework B since the focus was largely on winning the contract. Taken together, the findings suggest that a strong relationship-improvement strategy will increase supplier commitment to the owner’s business goals and objectives, enhancing the level of effort induced through financial incentives. These findings indicate that relational incentives play a dominating role in inducing more effort from suppliers, particularly

132

Incentivizing Collaborative BIM-Enabled Projects

in a less competitive environment. Evidence also suggests that the impact of financial incentives as a means of inducing effort exertion may be decreased without strong relational incentives. Subsequently, these findings provide further evidence that financial incentives and relational incentives can co-occur rather than preclude each other. Taken together, the results reveal that the potential for a continual relationship with a major client is a very strong driver. This driver can also be related to the desire of suppliers to uphold and improve their reputation as to increase future commercial opportunities. This driver is particularly relevant to the research population, as both framework owners are large agencies that are likely to be key repeat clients. The desire to strengthen their reputation with repeat clients is likely to be stronger here than with clients who are less likely to provide further work opportunities.

BIM Incentivization With regard to BIM incentivization, partners and tier 1 suppliers across frameworks A and B stated that suppliers should not be incentivized explicitly for the use of BIM. Although incentives were able to drive the motivational context of behaviors effectively, incentivization of suppliers should be based primarily on their performance and efficiency in alignment with the owners’ objectives. It was, however, pointed out that BIM could act as a catalyst for supplier incentivization. In other words, suppliers’ performance efficiency could be achieved through the effective use of BIM, which hence enabled them to be incentivized. This line of thought was further strengthened by a summarized statement below made by one tier 1 supplier for framework B: . . . we are doing it because not only will it deliver better information by using the proper BIM processes, techniques, and technologies, and in the right culture, we will also deliver efficiency. This will benefit our organization in the way we undertake our work, and hence make a better profit through winning extra work. So, this is our incentive. Suppliers were seen to benefit themselves from using BIM in the way they undertook their works. BIM not only delivered better information to them using the proper BIM processes, techniques, and technologies, but also delivered efficiency, which improved their business profit.

Case Studies – United Kingdom

133

For this reason, top-tier suppliers across two frameworks perceived BIM incentivization as unnecessary. Contrary to these insights, apart from the relational incentive-driven supplier who also emphasized the development of their organization’s technical capabilities through collaboration with large and mature organizations, tier 2 suppliers for framework A felt that they should be incentivized for the use of BIM due to the additional costs associated with BIM adoption. These findings imply that an increase in suppliers’ financial incentives does not entail a corresponding decrease in suppliers’ desires to maintain their business relationships with well-established organizations, and also to develop their BIM capabilities for their commercial benefits. These results also provide further evidence that BIM incentivization and relational incentives can co-occur rather than inevitably precluding each other, and thus have stronger incentives to gain immediate benefits from BIM adoption. The findings also provide further evidence that suppliers’ BIM capabilities have impacts on their financial motivations to adopt BIM.

Performance Evaluation Both frameworks A and B adopted a series of key performance indicators (KPI) or supplier performance indicators (SPI) that covered the success measures of suppliers’ performance and efficiency. One supplier for framework A stated that performance evaluation should be based on both quantitative and qualitative measures. It was further revealed that precluding one form or another could potentially lead to lower-tier suppliers providing the products, which was not in accordance with the product specifications in the contracts. Therefore, while achieving cost savings, the quality of goods supplied by suppliers should be assessed. Although the constraint on the overall cost of delivery was seen to be useful, one lower-tier supplier perceived that it would be difficult to establish a system to identify any time and cost savings made and how these savings could be related to an individual supplier. Moreover, the target-cost contracts would not be able to take account of suppliers’ performance in all key priority areas. The perception of injustice in the evaluation of performance implies that performance evaluation may lead to the existence of biases. Moreover, in a highly interdependent organization, the difficulty in evaluating performance lies in the challenge that individual output may be indistinguishable from group output.

134

Incentivizing Collaborative BIM-Enabled Projects

Suppliers were asked the standard questions: “Do you think the collaborative arrangements support the working of an incentive scheme?” Overall, all suppliers agreed that collaborative frameworks would support the working of an incentive scheme. However, one of the tier 2 suppliers stated: . . . the collaborative framework certainly supports the working of [an] incentive scheme. However, the alliance has to be broadened. Currently, the framework only involves the tier 1 contractors and specialist contractors like us. So, what they probably have to do is to broaden their alliance to include and help the lower tiers. This finding implies that while incentive-contracting arrangements are commonly adopted in the collaborative practices, such incentives are seen to have precluded the lower-tier suppliers. This can potentially affect the levels of effort from and engagement with suppliers, which in turn affect the levels of use of BIM, and hence affects overall project performance.

C H A P T E R

10

Empirical Investigation – United Kingdom Hypotheses Project Characteristics, Early Involvement, and Use of Group-Based Incentive Contracts Construction project characteristics have been studied from different angles in academic literature. Winch (2001) addresses the interrelation between information, project uncertainty, and project life cycle, arguing that projects experience a progressive reduction in uncertainty through the project life cycle. From the perspective of transaction cost economics, Ive and Chang (2007) categorize project characteristics and relate them to procurement systems selection. From the perspective of systems thinking, early involvement of contractors is regarded as conducive to other group processes such as joint decision making and team incentives. Through eight case studies, Mosey (2009) suggests that contractors’ early inputs could significantly improve design, costs, and risk management in projects. There are three useful case studies that help build the connection between project complexity, early contractor involvement, and use of incentive mechanisms in propelling contractors’ input to design development. First, Potts and Ankrah (2014) investigate how the 18 main projects of Heathrow Terminal 5 (T5) were managed through 140 sub-projects. They point out that many contractors were involved during early planning stages to solve potential problems using 3D models before design was finalized. The T5 agreement also provided a ring-fenced profit to parties involved by retaining project risks and incentivizing parties right from the early project life cycle. The role of joint decision making as a 135

136

Incentivizing Collaborative BIM-Enabled Projects

process to facilitate better coordination and as an effective mode of governance is also discussed. Second, the American Institute of Architect’s (AIA) report (2010a) presents six case studies involving BIM applications. Common factors in these cases are the early involvement of project participants with the support of incentive contracts. The use of well-designed compensation layers as an early incentive to project participants is found in several cases, and the use of contingency funds along with the use of “The Integrated Form of Agreement” (a relational contract) is also reported. Third, Daniel, Chan, Lam, Chan, and Wong (2010) study the use of incentive contracts under two-stage tendering in Hong Kong subway station projects. The two-stage tendering process allowed early design inputs from contractors. The competition in the second stage of tendering provided incentives to contractors by providing a more reliable target cost. The case studies above all involve substantial project complexity in terms of interface management and require early design inputs from specialist subcontractors as well as cost data from main contractors. Project developers in these cases are allocated potential monetary rewards through an incentive pool at the outset of the project as an incentivization mechanism. Therefore: Hypothesis 1 Greater project complexity will lead to early involvement of project contractors. Hypothesis 2 Greater project complexity will result in the greater use of incentive pools as an incentivization mechanism. Hypothesis 3 The employment of an incentive pool increases the use of group-based incentives. Hypothesis 4 The employment of group-based incentives has a direct impact on the joint development of project goals.

Performance Measurement and Subjective Evaluation of Performance A critical condition for the effective use of incentives is that the output of the agent is measurable. Performance measurement depends on

Empirical Investigation – United Kingdom

137

either subjective or objective evaluation. In theory, each has its pros and cons and should be given different weightings (Chang & Howard, 2016). In practice, owners tend to evaluate design quality and satisfaction through subjective criteria when awarding incentive payments to consultants and contractors (American Institute of Architects, 2010b). As the agent’s effort is not readily observable in most interesting cases and output is subject to randomness, an effective strategy is to link up the agent’s output directly with their pay (Baker, 1992). However, the owner’s discretion over incentive payments would impede the contractor to relinquish part of their decision rights to a joint decision making mechanism. Therefore: Hypothesis 5 Subjective evaluation of project performance has a negative effect on the extent of decisions being made jointly in the project.

Joint Decision Making Gibbons and Roberts (2013) identify that solutions to motivation problems in organizational design are not limited to contracts and incentives, but may also include aspects of personal satisfaction and other social factors. Baron and Kreps (2013) subscribe to the argument that explicit extrinsic rewards would drive out intrinsic motivation and thus provision of incentives should consider social and psychological elements (e.g., social identity, equity, and reciprocity). In construction, lack of cooperation and teamwork has been well documented in the government reports (e.g., Latham, 1994). In response to this problem, improving contract involvement through partnering is embraced as a solution (Bresnen & Marshall, 2000). The importance of more delicate issues (e.g., trust and equity) (Potts & Ankrah, 2014; Yeung et al., 2007) and joint project goals (Eriksson & Westerberg, 2010) is regarded as essential in the partnering relationship. Through reviews of the early use of partnering and IPD as a response to project complexity in oil and gas as well as U.S. Army projects, Lahdenperä (2012) provides insight into the issues of authority assignment in joint decision making from a process perspective. The American Institute of Architects’ (2010a) report provides a rich elaboration of practices regarding collaborative decision making and control process, showing that decision-making issues at

138

Incentivizing Collaborative BIM-Enabled Projects

different levels are solved by the consensus of representatives from project parties. Therefore, we hypothesize: Hypothesis 6 Early involvement of project participants positively impacts joint decision making. Hypothesis 7 Joint decision-making processes have a positive impact on the effectiveness of incentivization.

Behavioral Theory Heuristics and biases were crutical in explaining the maps of bounded rationality (Baddeley, 2013). Heuristics are the rules of thumb derived from common sense, which would lead to a systematic bias in decision making, known as behavioral biases. One form of these biases is called motivational biases that are attributed to an individual’s circumstances and interests. Since motivational biases (e.g., loss aversion and endowment effect), unlike other forms of bias, are controlled by rationality, these may be manipulated by incentives (Baddeley, 2013). Kahneman and Tversky (1979) put forward Prospect Theory as an alternative to expected utility theory, where they replace the objective probability of outcomes with subjective weightings and introduce concave utility for gains and convex utility for losses, with the latter being much steeper than the former. They conclude that people weigh losses more than gains (Baddeley, 2013). Therefore, the central idea of Prospect Theory is that losses are valued higher than corresponding gains, and that gains and losses are differentiated by a reference point (Kahneman, 2011). Thaler (1980) submits that the shape of Kahneman and Tversky’s (1979) value function indicates decision makers are subject to inertia in the face of loss events, which would result in underweighting of the opportunity costs of gains (endowment effect). Tversky and Kahneman (1991) refer to endowment effects in the context of reference dependence, arguing that decisions on utility are made on perceptions of advantages or disadvantages about the reference point (Baddeley, 2013). To study endowment effects (Thaler, 1980), Knetsch (1989) conducted three experiments in which two groups of students are offered varying amounts of goods and money (hypothetical) in order to study the indifference curves of willingness to pay (WTP) and willingness to

Empirical Investigation – United Kingdom

139

accept (WTA). List (2003) follows a similar experiment design to Knetsch (1989) by employing both field experiments and questionnaires for exit interviews to study the relationship between market experience and endowment. He concludes that individual behaviors edge closer to the neoclassical model in a market environment as market experience increases. For these reasons, this study examines the influence of loss aversion arising from the endowment effect by measuring the contractor’s current workload and current financial conditions as their points of reference. As a result, we hypothesize: Hypothesis 8 Experienced contractors do not exhibit an endowment effect, and thus their experience positively influences the use of group-based incentive mechanisms in construction projects. Hypothesis 9 Loss aversion arising from the endowment effect has a negative impact on the effective use of financial incentives.

Incentives for Building Information Modeling In the discussion of BIM benefits, Eastman et al. (2011) analyze various procurement systems that would facilitate the uptake of BIM and advocate the use of IPD as an ideal system of procurement for the implementation of BIM. The American Institute of Architects (2010a) provides rich evidence for BIM-driven innovations in design and technology. Chang (2014a,c) addresses the barriers to BIM adoption by asking why project contractors should exert their best efforts to use BIM collaboratively since only the owners benefit. He pegs a set of practical and theoretical questions through a comprehensive literature review and an array of case studies. The effectiveness of BIM utilization can be measured in terms of its effect on quality and innovativeness. Therefore: Hypothesis 10 Group-based incentives have a positive impact on the effectiveness of BIM in terms of project outcomes, quality, innovativeness of design, and technology. Hypothesis 11 Effective incentives have a positive influence on the effectiveness of BIM.

140

Incentivizing Collaborative BIM-Enabled Projects

Financial Arrangements Affecting Extrinsic Motivation In organizational design, a central issue lies in how to motivate the agent through provision of appropriate incentives so that the tendency of shirking can be effectively curbed (i.e., moral hazard problem). The inf luential principal-agent theory posits that, subject to the verifiability of effort, the most effective solution is to hold the riskaverse agent accountable for the financial consequence of their actions (Gibbons & Roberts, 2013; Milgrom & Roberts, 1992). Scherer (1964) proposes a mechanism for contractors to decide whether to take up contracts based on the expected value of income in defense procurement contracts and suggests that risk-averse contractors prefer low-sharing ratios. The essence of efficient contracts (based on the efficiency principle) therefore lies in a balance between the cost of risk bearing and the efficiency gains from stronger incentives (Milgrom & Roberts, 1992). Construction projects are invariably fraught with risks, so allocation of risks between the owners and contractors is a central issue in contract design. In practice, each contract type is a point on the continuum of two polar forms—namely, lump sum (strongest incentives for cost reduction) and cost plus contracts (weakest incentives for cost reduction) (Chang, 2014d), and project costs can be reduced by allowing agents to benefit from a share of cost savings through the choice of contract types. Broome and Perry (2002) develop eight templates for sharing ratios based on varying degrees of project risk and confidence of achieving the target cost in motivating contractors. Perry and Barnes (2000) examine the interrelation between contractors’ fees and target cost, concluding that a low fee does not motivate contractors to reduce target costs and vice versa. Badenfelt (2008) captures the clients’ and contractors’ views on sharing ratios and highlights instances where clients suggest a ratio of 0.5 to 0.7 to the contractors. He concludes that ratios should be selected on previous experiences and with a view of long-term relationships to overcome barriers of asymmetric information. Daniel et al. (2010) examine the use of target-cost contracting (TCC) using gain/pain sharing ratios under a two-stage tendering process. As reviewed by Potts and Ankrah (2013), the Heathrow Terminal 5 agreement contains a no-risk transfer policy that included a single insurance policy for the entire project. For the first time in construction, the client, BAA, assumed the cost of errors committed by contractors

Empirical Investigation – United Kingdom

141

to improve the collaborative working environment. In doing so, the mechanism served as an incentive to contractors in increasing performance and creating a cooperative working climate. The NEC family of contracts, options C and D in particular, also places an emphasis on the cooperative- and team-based approaches, and presents challenges for cost control in target-cost contracts in the event of change orders (Potts & Ankrah, 2008). Table 30 provides a summary of the relationship between observed and latent variables in the model. The hypotheses discussed before can be visualized in Figure 27. In addition to the latent variables, this study considered market experience (EXP), loss aversion (LA), and subjective evaluation of project performance (SE) as exogenous variables. The interrelationship among these exogenous variables with the other latent variables can also be seen in Figure 27.

Table 30.  Relationship between observed and latent variables. Latent or endogenous variable

Exogenous or observed variable

References

Project complexity

Interface management

AIA, 2010a; Broome, 2002; Chang & Ive, 2007; Ive & Chang, 2007; Potts & Ankrah, 2008

Change orders Availability of contractors/asset specificity Early involvement of project contractors

Early involvement of project participants

Mosey, 2009

Joint decision making and commitment to a common goal

Joint decision making Jointly developed project goals

AIA, 2010a; Anvuur & Kumaraswamy, 2007; Potts & Ankrah, 2013

Effective use of BIM in projects

Construction quality

AIA, 2010a; Eastman et al., 2011

Design innovativeness Innovativeness in technology Effectiveness of incentives

Collaborative work environment Cost control Schedule management

Incentive pools

Early incentives to main contractors Early incentives to trade contractors

Group-based incentives

Multiparty contract Explicit risk-sharing arrangement Open-book accounting

Endowment

Contractor’s current workload Contractor’s current financial position

AIA, 2010a; AIA, 2010b; Anvuur & Kumaraswamy, 2007; Badenfelt, 2008; Broome, 2002; Potts & Ankrah, 2008 AIA, 2010b; Chang, 2014d; Daniel et al., 2010; Mosey, 2009; Potts & Ankrah, 2013 AIA, 2010a; Chan et al., 2010; Chang, 2014c; Mosey, 2009; Potts & Ankrah, 2013 Kahneman & Tversky, 1979; Thaler, 1980; Tversky & Kahneman, 1991

142

Incentivizing Collaborative BIM-Enabled Projects

0,

0,

e28 0,

e202 0,

e201 0,

e200

1

0, 1

e100

INT

0

CO

Complexity

1 1

0,

1

0, 0,

1

e26

MCI

1 1

e9

e83

TCI

1

1 0

1

1

JPP

JTPP

1

Incentive Pool

0,

0,

e18

e19

e20

1

1

1

TECH

QLTY

INVTN

0

e28

0

1

EffBIM

e400 0,

e300

1 1

EIOPP

1

0

0,

EarlyInv

e555

1

0,

0,

JtDec

0,

COMX

0,

e81

e550

0

GBI 1

SE

MPC

ERSA 1

1

e29

1

0,

0,

0,

e222

e444

e112

EXP

0,

OBA

1

0

EffIncentives LA

0,

1

e110 1

0

Endowment 1

CFC 1

CWE

TIME

COST

1

1

1

0,

CCWL

e23

0,

e22

0,

e21

1 0,

e103

0,

e101

Figure 27.  SEM model for the UK data.

Research Method Sample A survey questionnaire was designed as the research instrument. The respondents surveyed were mainly senior-level directors and partners of firms involved in high-value construction projects (with project costs over �10 million) in the UK during the period 2011–2014. The median years of experience for respondents in the construction industry is 25 years. Information about project participants was obtained from the Glenigan database. The survey was sent out to 682 prospective respondents. Of those, 168 people responded to the survey with a response rate of 25%. The areas of specialization of the respondents are shown

Empirical Investigation – United Kingdom

143

Table 31.  Area of specialization of survey respondents. Q4. In which area of the construction industry does your organization specialize? Answer options

Response percent

Response count

Public sector construction client

11.5%

20

Private sector construction client

6.9%

12

12.6%

22

Tier 1 construction contractor (main contractor) Tier 2 construction contractor (subcontractor)

9.8%

17

Engineering consultants

16.7%

29

Architecture and design consultants

19.0%

33

Cost consultants

5.7%

10

Project management consultants

5.2%

9

Property development

4.6%

8

Logistics

0.6%

1

Legal

2.9%

5

Other (please specify)

4.6%

8

in Table 31. Hypotheses are tested by an SEM model using SPSS AMOS. To ensure the external validity of the questionnaire, one telephone interview was conducted to obtain early feedback from the BIM director of a leading UK construction company.

Measurement The questionnaire consists of six sections, each of which contains five to eight positive statements. Respondents were asked to rate the questions on a 1–7 scale (1—Very strongly disagree; 7—Very strongly agree), based on the context of the project they were most familiar with. Section 1 is made up of eight statements to explore the collaborative measures adopted. Section 2 contains five statements about incentive measures taken. Section 3 aims to explore the influence of behavioral biases based on the perceptions of the respondents. In Section 4, there are five statements to evaluate the influence of collaborative mechanisms on the effectiveness of BIM in the project. Six key project characteristics of the project are covered in Section 5. The purpose of Section 6 is to collate evaluation of project performance with respect to six criteria.

Confirmatory Factor Model Structural equation models support some statistical approaches, such as confirmatory factor analysis (CFA), exploratory factor analysis (EFA),

144

Incentivizing Collaborative BIM-Enabled Projects

path or causal analysis, and multivariate regression analysis. This study adopted CFA for the analysis of observed and latent variables. SEM analysis, also known as covariance structure analysis, was carried out using SPSS 24’s maximum likelihood function to analyze the interrelationship between latent variables. Each ordered categorical variable was analyzed as a stand-alone variable. There are eight latent or endogenous variables measured in the model (code name in brackets), including effectiveness of incentive measures used (Effincentives), effectiveness of BIM (EffBIM), project complexity (Complexity), monetary rewards based on joint output (Incentive Pool), endowment (Endowment), early involvement (EarlyInv), joint decision making (JTDec), and group-based incentive mechanism (GBI). The survey data show that 71 of the 166 projects adopted Level 2 BIM as per PAS 1192-2:2013 (see Table 32), and there is no evidence of any construction project in the UK utilizing Level 3 BIM.

Table 32.  Project life cycle where BIM was used. In which stage(s) was building information modeling (BIM) used on the project? Response percent

Response count

Not used

Answer options

30.7%

51

Feasibility

13.9%

23

Concept development

30.1%

50

Design

57.8%

96

Construction

41.0%

68

Operation

15.1%

Answered question

25 166

LEVEL 0 (Unmanaged CAD, in 2D, with paper or electronic paper-data exchanges)

24.1%

40

LEVEL 1 (Managed CAD in 2D- or 3D-format with a collaborative tool providing a common data environment and standardized approach to data structure and format. Commercial data managed by stand-alone finance and cost management packages with no integration)

42.8%

71

LEVEL 2 (A managed 3D environment held in separate discipline BIM tools with data attached. Commercial data managed by enterprise resource planning software and integrated by proprietary interfaces or bespoke middle-ware. This level of BIM may utilize 4D construction sequencing and 5D cost information)

33.1%

55

LEVEL 3 (Characterized by a fully integrated and collaborative process enabled by web services, and incorporating 4D construction sequencing, 5D cost information, and 6D project life cycle management information)

0.0%

0

Empirical Investigation – United Kingdom

145

The survey results revealed that 8.4% of the respondents reported that a formal partnering arrangement was adopted for their projects (Table 33). These statistics also reveal that during the period 2011–2014, 53% of the major projects were procured through design-build, and 13.6% of the projects adopted an informal partnering arrangement.

Table 33.  Procurement methods employed in the sample projects. Which type of procurement system was used in the project? Answer options

Response percent

Response count

Traditional method (design-bid-construct)

21.7%

36

Design and build

53.0%

88

Traditional method plus informal partnering arrangement

5.4%

9

Design and build plus informal partnering arrangement

7.8%

13

Management system plus informal partnering arrangement

0.6%

1

Formal partnering contract

8.4%

14

Other (please specify)

7.2%

12

Answered question

166

During the period 2011–2014, construction clients’ policies toward BIM adoption varied. While some clients were BIM-neutral, other clients strived to implement BIM by mandating it through contacts. However, all construction clients wanted to tap into the benefits of BIM, such as cost benefits, single source of data, improved decision making, and accuracy of design outcomes.

SEM Analysis This section details the steps adopted in the development of the structural equation model. SEM has two components—namely, measurement model and structural model. The measurement model analyzes the relation between exogenous variables and the corresponding latent variable using confirmatory factor analysis. The structural model measures the relation between the endogenous variables. Since SEM is considered an efficient way of describing the latent variables underlying the set of observed data (Kline, 2011), this section aims to set out the structure of the model and test how well the observed data can fit the model.

146

Incentivizing Collaborative BIM-Enabled Projects

Based on the literature review, a confirmatory factor model is constructed, and the interrelationships between variables are defined (see Table 34). In addition to these variables, three theoretical constructs—namely, subjective evaluation of project performance, market experience, and loss aversion—are treated as exogenous variables in the model (see Figure 28). Model identification refers to verifying whether or not all parameters in the model are uniquely estimated. Each parameter in the model is specified as either a free parameter or a fixed parameter. The model must satisfy the general requirement for model identification—namely, that the model’s degrees of freedom must at least be equal to zero. The model presented in this study has 196 degrees of freedom, therefore satisfying the condition required (Kline, 2011). 0, .27

0, .64

e28 0, 1.65 1

e202

0, 1.94 1

e201

0, 1.63 1

e200

0, 1.19 1

e300

4.51

CO

e100

3.76 .65

1.06

EIOPP

1

Complexity 0, 1.10

COMX

e400

e26 1.00

.82

0

SE

Incentive Pool

0, 1.61

1

.38

–.19

1

0, .26

e555

MPC

ERSA

OBA

1

1 0, 1.49

e444

0, 1.12

e29 0, 1.46

1

EffIncentives 0, 1.10

.84

e110 1

0

1 2.27

2.44

CFC 1

CCWL 1

0, .81

e103

4.10

CWE 1.00

0, .57

e101

1.00

.95

–.43

Endowment .75

0

e112

.28

.41

Figure 28.  The full SEM model.

.86

2.42

1

LA

.90

0

1

.42

1.61

0, 2.09

4.11, 2.08

QLTY

.97

1.46

1.63

e222

TECH 1.00

.10

0

2.44

.27

1 4.15

EffBIM

GBI 1.00

4.11, 1.44

.77

e20

1 4.02

INVTN 0, 30

e28

0, .49

e19

1 4.14

JTPP .90

0, .43

e18

4.36

0

.77

0, .51

e81

4.35

JPP

1.00

EXP 1.04

1

1

1 .83 0

3.49, 2.18

.14

e83

TCI

.71

EarlyInv

.25

0, .72 0, –.12 e550

3.57

JtDec

1

1.00

0, 1.83 1

MCI 0

1.00 4.07

4.86

1

0, 1.01

INT

e9

4.06

TIME

1 0, 1.32

e23

4.10

COST 1

0, .37

e22

0, .27

e21

147

Empirical Investigation – United Kingdom

The estimation of coefficients for the paths of the SEM model in Figure 28 is reported in Table 34. Estimation involves the use of a fitting function. In this study, this is the “maximum likelihood” between the observed variables and the model’s latent variables (i.e., variance and covariance matrices). Table 34.  Maximum likelihood estimates of the SEM model. Estimate

S.E.

C.R.

P

Incentive pool



Complexity

.818

.246

3.324

***

par_11

GBI



EXP

.273

.095

2.872

.004

par_20

EarlyInv



Complexity

.713

.239

2.978

.003

par_21

GBI



Incentive pool

.383

.091

4.213

***

par_23

Endowment



LA

.411

.129

3.187

.001

par_13

JtDec



EarlyInv

.772

.223

3.466

***

par_17

JtDec



GBI

.770

.210

3.664

***

par_24

JtDec



SE

2.185

.065

22.840

.005

par_27

Effincentives



Endowment

2.430

.210

22.043

.041

par_14

Effincentives



JtDec

.417

.112

3.728

***

par_16

Effincentives



LA

.283

.141

2.009

.045

par_25

EffBIM



Effincentives

.859

.089

9.649

***

par_18

EffBIM



GBI

.100

.111

.898

.369

par_26

MCI



Incentive pool

1.000

TCI



Incentive pool

.834

.086

9.702

***

par_1

INVTN



EffBIM

1.000

TECH



EffBIM

.969

.086

11.273

***

par_2

QLTY



EffBIM

.901

.085

10.630

***

par_3

COST



Effincentives

1.000

TIME



Effincentives

.952

.072

13.247

***

par_4

CWE



Effincentives

.843

.105

8.003

***

par_5

COMX



Complexity

1.000

CO



Complexity

.646

.216

2.990

.003

par_9

INT



Complexity

1.058

.280

3.777

***

par_10

CCWL



Endowment

1.000

CFC



Endowment

.220

3.420

***

par_12

EIOPP



EarlyInv

1.000

JPP



JtDec

1.000

JTPP



JtDec

.901

.136

6.649

***

par_15

MPC



GBI

1.000

ERSA



GBI

1.633

.366

4.456

***

par_19

OBA



GBI

1.457

.333

4.372

***

par_22

.754

Label

148

Incentivizing Collaborative BIM-Enabled Projects

As suggested in the literature (Kline, 2011), the criteria used to assess the model under study are of two types: root mean square error of approximation (RMSEA) and goodness of fit estimates. The results exhibit that all indices are within the recommended ranges (Barron & Kenny, 1989; Kline, 2011), indicating that the survey data can satisfactorily fit the model. Table 35.  Goodness of fit measures for the hypothetical model. Goodness of fit measure Root mean square error of approximation (RMSEA)

Recommended levels

Model’s score

0.05 to 0.1

0.085

Comparative fit index (CFI)

0 to 1

0.791

Tucker Lewis index (TLI)

0 to 1

0.730

Normal fit index (NFI)

0 to 1

0.685

Incremental fit index (IFI)

0 to 1

0.801

Table 36 indicates some noteworthy findings, as 10 of the 11 hypotheses developed by this study were statistically significant. These results are further analyzed in the discussion section. Table 36.  Standard coefficient estimates for hypothesized model. Hypothetical paths with direction

Estimates or beta value

Significance (p-value)

Standard error

Interpretation

H:1 Complexity to EarlyInv (1)

0.713

0.003

0.239

Significant

H:2 Complexity to Incentive pool (1)

0.818

0.000

0.246

Significant

H:3 Incentive pool to GBI (1)

0.383

0.000

0.091

Significant

H:4 GBI to JtDec (1)

0.770

0.000

0.210

Significant

20.185

0.005

0.065

Significant

H:6 EarlyInv to JtDec (1)

0.772

0.000

0.223

Significant

H:7 JtDec to Effincentives (1)

0.417

0.000

0.112

Significant

H:8 EXP to GBI (1)

0.273

0.004

0.095

Significant

H:9 Endowment to Effincentives (2)

0.430

0.041

0.210

Significant

H:10 GBI to EffBIM (1)

0.100

0.369

0.111

Not Significant

H:11 Effincentives to EffBIM (1)

0.859

0.000

0.089

Significant

H:5 SE to JtDec (2)

Discussions Project Complexity and Early Involvement of Contractors H1 and H2 maintain that project complexity would increase early involvement of contractors and use of incentive mechanisms (e.g., incentive

Empirical Investigation – United Kingdom

149

compensation layer or ring-fenced profit arrangements) (Potts & Ankrah, 2013). Evidence reveals that both H1 and H2 are statistically significant and the total effects of complexity on the observed variables are significant for interface management (INT), change orders (CO), and availability of project contractors (COMX). In other words, project complexity increases the use of incentive mechanisms and early contractor involvement mechanisms.

Effects of Financial Rewards on Group-Based Incentive Mechanisms and Joint Decision Making H3 hypothesizes that a positive relation exists between financial rewards to contractors and use of group-based incentives. This statistically significant result indicates that financial rewards are normally awarded on a group basis. Together with H4, group-based incentives (GBI) make it essential to introduce joint decision making (JtDec).

Effects of Subjective Evaluation on Joint Decision Making The finding indicates that inputs from various contracting parties are blurred during the early design stages, and thus it is difficult to observe inputs from each contracting party. Baker (1992) suggests that when the output of each agent is hard to observe, the marginal product of output leading to marginal gains is a good measure of performance. To capture the effects of subjective evaluation on the use of group-based incentives, Hypothesis 5 tests that subjective evaluation negatively impacts commitment toward joint project goals (JPP) and use of joint decision-​ making process (JTPP), both of which are measured by the latent variable, JtDec. The negative effect of subjective evaluation (SE) on the use of group-based incentives (i.e., JtDec) proves statistically significant.

Effects of Early Involvement with JtDec and the Effectiveness of Incentives From a practical or theoretical perspective, a host of prior studies submit that early involvement of project contractors would engender a positive effect on the presence of joint-committed project goals (JPP) and joint decision-making process (JTPP) (AIA, 2010; Baron & Kreps, 2013; Daniel et al., 2010; Potts & Ankrah, 2013). The test of Hypothesis 6 shows that early contractor involvement does increase the use of joint decision making (JtDec) in terms of JPP and JTPP. JtDec is also found to have a positive impact upon Effincentives. The total effects of Effincentives

150

Incentivizing Collaborative BIM-Enabled Projects

on time, cost, and collaborative work environments are 0.952, 1.00, and 0.843, respectively. The R-squared value of Effincentives indicates that 31.7% of the variance in the latent variable can be accounted for by the observed variables of Effincentives. Therefore, the observed variables used to measure the effectiveness of incentives are significant. All said, early contractor involvement has led to a greater use of joint decision making and strengthened the effectiveness of incentives.

Effects of Market Experience of Group-Based Incentives In experimental settings, List (2013) finds that the effects of endowment would disappear as market experience increases. A similar effect is tested in construction settings by Hypothesis 8. It is found that, with EXP treated as an exogenous variable, a contractor’s experiences have translated into a greater use of group-based incentive mechanisms (GBI). A plausible explanation for this finding is that experience could enable contractors to build in-house capability in monitoring contract performance, which could help overcome the barriers of adopting GBI.

Loss Aversion Due to Endowment Effect on Effectiveness of Incentives Behavioral economics suggests that ownership itself would change one’s preference for the course of actions. It is because one tends to attach a higher value to retain the good he has already owned than otherwise. An implication of the endowment effect for construction projects is that loss aversion may have a negative impact on the effectiveness of incentives. The testing result of Hypothesis 9 demonstrates that this effect is also present in BIM-enabled projects.

Effects of Effectiveness of Incentives on Effective Utilization of BIM Hypothesis 10 is concerned with the extent to which group-based incentives would have affected the potential of BIM for project performance improvement in terms of construction quality, innovativeness in design, and technology deployed. Given that data do not lend support to this hypothesis, it implies that the benefit of BIM does not result from the use of group-based incentives alone. By contrast, the working of incentives proves to be essential, as predicted by theory.

C H A P T E R

11

Case Studies – United States Introduction This chapter pulls together three strands of literature to examine the interdependence of project delivery environments on BIM implementation: incentive theory, technology acceptance, and status quo biases. Through three separate pilot studies and a comprehensive primary survey, the resulting quantitative data have revealed several key trends and perceptions that inform how project delivery environments could impact BIM adoption. To better frame the results in a real-world construction environment, a case study subject was sought. Since both integrated project delivery (IPD) and advanced levels of BIM adoption are rare in practice, a great deal of effort was expended to find a viable subject. Ultimately, two projects at the San Francisco Airport were chosen due to their use of an advanced form of IPD and their implementation of 6D-BIM. To the writer’s knowledge, this represents the most advanced collaborative environment for a major construction project in the United States. Specifically, two projects are reviewed: the Air Traffic Control Tower completed in October 2016, and the Terminal 1 Redevelopment Project, which commenced in October 2016 with a planned completion in 2024. The Terminal 1 Redevelopment Project is seen as a more collaborative and technologically advanced construction environment than was utilized for the Control Tower, so while still in the engineering phase, its approach is key to understanding the results of the quantitative analysis.

151

152

Incentivizing Collaborative BIM-Enabled Projects

Project Background Contracting Approach The San Francisco International Airport (SFO) has developed a unique approach to delivering projects known as the exceptional project delivery paradigm. This approach guides the planning, design, and execution of airport projects through structured collaborative partnering (SCP) and a stakeholder engagement process. SCP aims to create an integrated, high-performing project team from multiple parties with commitments to teamwork, communication, trust, transparency, respect, and fairness. The stated purpose of SCP is to cultivate solid working relationships before problems and issues arise. The goal is to redefine expectations of how all parties work together, thereby minimizing negative results such as financial losses, damaged relationships, and unresolved claims. This is accomplished through a multiparty agreement sometimes referred to as “super IPD” by the SFO team. The contracts are based on a guaranteed maximum price format with the parties sharing in any cost savings from budget underruns. This base approach is akin to AIA integrated project delivery methodologies, but SFO takes this a step further with each stakeholder rating and commenting on the collaboration throughout the project.

Air Traffic Control Tower Project As a joint venture between SFO and the Federal Aviation Administration (FAA), the airport built a new air traffic control tower with the latest in technology and design. Standing 221 feet tall, the new tower features a modern, flared design clad in curved metal paneling. The project also includes a new, three-story, integrated facility building for the FAA and other personnel, two connector walkways, and improvements to the Terminal 1 boarding area C entrance. Construction began in the summer of 2012 and became fully operational in October 2016. The new tower was built to satisfy specific technical and site requirements as well as stringent seismic, safety, and security design standards while also achieving LEED Gold status. SFO provided full access to the researcher of all contract documents, budgetary information, and tracking metrics for the project, which are being reviewed against the results of the quantitative data.

Terminal 1 Redevelopment Project SFO is redeveloping Terminal 1, one of its oldest terminals built in the 1960s, to meet the needs of modern travelers and revolutionize the

Case Studies – United States

153

guest experience. When fully completed in 2024, the terminal is expected to meet or exceed the award-winning standards of Terminal 2 and, at minimum, receive LEED Gold certification. This US�2.4 billion project includes design and construction of north, south, and central areas; a new boarding area; new passenger loading bridges; concessions; a refreshed space for assenger check-in; a consolidated security checkpoint; a re-composure area; baggage handling system and baggage claims; and a new mezzanine with connections to the air train and central parking garage. As with the control tower, SFO provided full access to the re­ searcher of all contract documents and budgetary information to be reviewed against the results of the quantitative data.

Interview Results Design Following extensive preliminary discussions via email and over the phone, in-person interviews were conducted with the SFO’s chief development officer and director of infrastructure. The below questions were utilized, but only as a starting point. Since this case study is being used to reflect upon and understand qualitative data, the questions were meant as starting points for further discussion rather than standardized for comparative purposes. The following questions were posed to stakeholders: 1. How great an impact does the project delivery method have on BIM implementation? 2. Did SFO encounter employee resistance when implementing BIM? 3. What about IPD? 4. How was IPD implemented? 5. Is implementing full-scale BIM possible without IPD? 6. How does SFO’s approach incentivize BIM participation specifically? 7. Regarding BIM and cultivating a collaborative environment, what lessons were learned in the Control Tower Project that impacted the approach for the terminal expansion? 8. Literature suggests that group-based rewards would likely result in one party free riding on another’s effort. Do you agree?

154

Incentivizing Collaborative BIM-Enabled Projects

Results Three interviews with the owner, first-tier contractor, and second-tier contractor enabled the story of how IPD/BIM evolved at SFO to be revealed. The adoption of IPD was initiated by the airport’s art director (who was a founding member of the National Partner Institute), and was furthered by the director of construction, who promoted a qualifications-­ based contracting methodology rather than a standard bid cycle. The goal was to facilitate early involvement, one of the central goals of integrated project delivery and breed trust. While partnering had a long history at SFO, it was not until 2008 that they began to formalize the approach and developed what they call a “structured collaborative process”—a structured way to build a high-performing team through issue identification, commitments, and then measuring how successful they were in meeting those commitments. They then began to refine the progressive design-build model, specifically how the stakeholders work together through a programming phase, complete the basis of design, and negotiate a target budget that everybody works toward. SFO believes that the clear majority of issues can be traced back to a misaligned expectation of scope, cost, and schedule. In response, their approach seeks to co-create these items from the outset and achieve a common purpose. In so doing, partnering at SFO seeks to achieve both integration and collaboration. Integration is the ability of collaboration to be shared with a larger group who keeps working toward the same common purpose. It’s important to recognize those two things as separate things. Partnering is focused on the collaboration and the conversation. Integration is where the tools come into play because it allows the information to be recorded in a way that everybody can see it in real time and that is the value of BIM. Integration is achieved not only through BIM-enabled communications, but also through a matrix-type organization where a large team is built around common purposes. Traditionally, project teams have been built around discipline (i.e., civil, structure, landscape, architectural, mechanical, electrical, and plumbing); or worse than that, they are structured around site and builder. The question is how to organize a project into an approach that is a matrix-type organization. For example, if there are 100 tasks that need to be done this way, the owner can select those who should be part of that discussion; who is appropriate; and who brings insight, interest, passion, and knowledge through a stakeholder engagement process. It works like a virtual team that exists over here within this organization.

Case Studies – United States

155

The organization structure remains on execution. That is only to execute on the linear process that has already been co-created through this collective wisdom. All the innovation is made in the virtual team. Once it is agreed upon, it goes across the schedule or is aligned over here. The change is then put back into the organization of the individual party. As vividly described by an interviewee: You have the designer design what it is, bring it back, review back to this group. Yep, that is where we intend it to be; then it goes back to the builder, and now the builder can just build it because he has the right amount of money, he created the money over here, everything is good, and you keep moving along. This is where the application of BIM and technology lies because the stakeholders can come into play by accessing a three-dimensional instead of two-dimensional plan. In the meantime, the data produced can be transformed into information that is useful to the people who have to own and operate. The push for BIM specifically started in February 2015. While BIM has been applied on T2 to detect clashes, it was the builder leading the effort. It was not done from the perspective of the owner. Its interest was not so much in the full life of the building; it was more about just getting the thing done because there was always the pressure on schedule and budget. One interviewee explained: In a traditional adversarial contracting environment, I am riding around with another hammer hitting the subcontractor on the head, and he is hitting his suppliers on the head, and nothing is actually getting resolved. We are all furiously writing letters to each other telling each other how terrible they are and how we are going to send them to claims hell and things like that. That is a very interesting concept to get together and objectively score each other right. Partnering could instill some trust into the leadership where one can be a leader right now but might eventually become a follower at a later stage of the project, and the roles keep changing. The project governance is not organized as a hierarchy. While economic theory predicts that one partner may free ride on the efforts of another in a shared risk and

156

Incentivizing Collaborative BIM-Enabled Projects

reward environment, SFO overcomes this problem with the mantra of “soft on the people, hard on the problem,” as one interviewee explained: It is so easy to get on the people first, and when you can never get to the problem and also, he’ll break it down for us. There are three levels: there are issues, there are problems, and then there’s a discipline. So traditionally, we are looking for arbitration, looking for attorneys, writing letters. But it’s really not working with a structured partner; collaborations to start with because when there are some issues that you are trying to resolve. There are no feelings hurt, everybody’s still friends. The great thing is that IPD allows pricing to be negotiated at each stage. With BIM, the accuracy of that estimation can be refined, which will make IPD work better. BIM and IPD are interdependent in this sense, especially in an incredibly complicated project involving an intellectual thought process: The biggest powerful tool we all have in our team is our brain and it’s all of our brains and the collective wisdom. That has nothing to do with the tools. It has to do with the way we interact as human beings. It has to do with: Are we working toward the same common purpose? You can have the best rowers in the same boat, but if they row out of phase, you still go west. You can have the best equipment in the world and the best of everything, it still doesn’t matter. If you don’t have a common purpose and you’re not all working together, you’re not a team, forget it. Owing to the Spearin doctrine, a contractor who is bidding on a low bid has every right to assume the drawings are perfect. This is the law, and also what is upheld by U.S. courts. In reality, the drawings may contain errors in omission in anywhere from low end (maybe 3%) to high end on a complex project (maybe 7–8%). The price of getting perfection is exorbitant. Even worse, one could get non-perfect drawings with the assumption that he has perfect drawings. To protect his own interest, the owner kept writing one more term and one more condition. However, this cannot stop confrontation from escalation. This explains why, since 1990, the

Case Studies – United States

157

U.S. construction industry has been infested with disputes. When it comes to solutions, one respondent reckons that: As long as you recognize a lot of the businesses are going on and losing money so you get something for free, but once you’re fair and everything, you’re going to get better pricing from people anyway. I’m going to always give you a better price to do whatever I’m going to do if I trust you. In the collaborative environment, a concern is free riding. When all project members are working together, maybe one is doing a little bit more of the work, but the others are still getting paid. To preempt this problem, SFO implemented a scorecard system, where all members can grade their score on one to five on the common objectives of the project, so each party can be held accountable. Grading takes place in the party workshops against the commitments that project members make initially. An interviewee said: It goes back to this public humiliation because we can’t be successful and get a higher score once I get you to come in with me. So, I’ve got to play nice with you and you have to play nice, but we have to work together. So, I’m going to be nice to you but we’re going get this problem solved and you’re not going home until it’s done. We are going to have fun doing it. From the interviews, it is evident that, whereas SFO ushered in a collaborative delivery environment, the drivers for project parties to work toward common purposes are built upon traditional partnering techniques (e.g., neutral facilitator, colocation, and review workshops). The pricing of individual work packages was through direct negotiations with contractors. Monetary incentives were not explicitly employed.

C H A P T E R

12

Empirical Investigation – United States Development of Hypotheses The United States is a country where we have seen the most advanced application of both BIM and integrated delivery. The central question is: Under a more BIM-friendly delivery environment, to what extent could BIM incentivization have affected project performance with consideration of BIM-user behavioral factors? Similar to other information technologies, BIM may not receive immediate acceptance by users. According to the influential technology acceptance model (Davis, 1989; Davis et al., 1989; Venkatesh et al., 2003) and its voluminous empirical studies, user reactions to the usefulness and ease of use of the information technology (IT) system are critical to the success of IT system implementation (see Figure 29). In this chapter, Figure 29 is expanded to consider behavioral factors and BIM incentivization effects in order to build causal links between BIM use and project performance. When implementing a new technology like BIM, the amount of support received by individual implementers from the organization (such as training, leadership) is a crucial factor.

Perceived usefulness Behavioral intention to use

Actual system use

Perceived ease of use

Figure 29.  Standard setup of the technology acceptance model. 159

160

Incentivizing Collaborative BIM-Enabled Projects

Perceived usefulness

Resistance

H7

H4

H1 External support H2 BIM incentivization

H6 H5

H3

Behavioral intention to use

H9

Actual system use

H10

Project performance

H8 Perceived ease of use

Figure 30.  The SEM model for the U.S. data.

The support given could be offset by user resistance, but strengthened by the incentivization measures taken (see Figure 30). Incentivization measures can provide BIM users with stronger incentives to deploy it with the effect of increasing the perceived ease of using BIM. Therefore, we hypothesize: Hypothesis 1 Resistance to use of new technology will have negative impacts on the effectiveness of external support. Hypothesis 2 BIM incentivization can reinforce the external support. Hypothesis 3 BIM incentivization will increase the perceived ease of use. The support an organization gives in implementing BIM could fundamentally change users’ perceptions of how useful BIM is to their jobs and how easily BIM can assist them in their jobs. We hypothesize: Hypothesis 4 External support (organizational) will have positive impacts on perceived usefulness. Hypothesis 5 External support (organizational) will have positive impacts on perceived ease of use.

Empirical Investigation – United States

161

When BIM is perceived as easy to use, it is more likely for users to think of BIM as a useful tool, and therefore raises the users’ intention to use BIM. Ease of use can impinge directly upon the intention to use BIM. We hypothesize: Hypothesis 6 Perceived ease of use will have positive impacts on perceived usefulness. Hypothesis 7 Perceived usefulness will have positive impacts on behavioral intention to use. Hypothesis 8 Perceived ease of use will have positive impacts on behavioral intention to use. Finally, tendency to use BIM will translate into actual use, which in turn could result in a positive influence on project performance. We hypothesize: Hypothesis 9 Behavioral intention to use will have positive impacts on actual system use. Hypothesis 10 Actual system use will have positive impacts on project performance.

Research Method Measurement and Confirmatory Factor Model The eight constructs are all modeled as latent variables. The measurements of each construct’s indicator variables are listed in Table 37. As indicated by Cronbach’s alpha, all constructs pass the reliability test, so these constructs are internally consistent. The loadings (standard coefficient) of the observable items on the latent variable are all above the acceptable value of 0.5. It is also important to scrutinize the validity of these constructs. The CFA can help determine which set of observed variables shares common variance-covariance characteristics that define latent variables. It is a norm to report the results of complementary fit indices. The first category, absolute fit index, is to check if the model is well fit to the data. x2/df should not exceed 3 (Kline, 2016). Two complementary indices are

162

Incentivizing Collaborative BIM-Enabled Projects

Table 37.  Measurement of constructs in the SEM model for the U.S. data. Resistance to use

Loadings

RESI_RePart

My company or I may come to regret the decision to participate in a BIM-enabled project.

.896

RESI_ReInve

My company or I may come to regret investing the time and money necessary to become proficient in the use of BIM.

.870

RESI_DeconIndi Participating in BIM decreases the control I have to complete my job.

.896

RESI_DeConCo

BIM decreases the control my company has in the success of the project.

.832

RESI_WaInv

Investing in BIM technology would waste my investment in standard non-BIM construction practices (AutoCad, etc.).

.805

Reliability Cronbach’s alpha 5 .908

Perception of incentivization IncenMon

Monetary rewards can improve the effectiveness of BIM considerably better than nonmonetary rewards.

.825

IncenGr

Group-based rewards will work considerably better than individual rewards in incentivizing stakeholder participation in BIM systems.

.791

IncenObj

Objective metrics are considerably better than subjective ones as the basis for determining incentive rewards for BIM participants.

.808

IncenWei

It is necessary to assign different weightings to performance metrics in the determination of incentive rewards for BIM participants.

.744

IncenMin

There is a minimum amount of incentive reward that can motivate contractors’ full participation in BIM.

.579

EX_Assist

A specific person (or group) is available for assistance with BIM difficulties.

.857

EX_Res

I have the resources necessary to work with BIM.

.837

EX_Clear

My interaction with BIM is clear and understandable.

.814

Cronbach’s alpha 5 .807

External support Cronbach’s alpha 5 .792

Perceived usefulness (U) U_IndiInten

I intend to use BIM if given the opportunity.

.826

U_IndiProd

I want to be among the first to adopt emergent BIM processes.

.813

U_IndiUseful

I would find BIM useful in my job.

.810

U_IndiAdo

Working with BIM increases my productivity.

.799

U_BeCoorW

BIM tools improve coordination, so that workforce, materials, and machinery are in the right place, at the right time, and in the correct quantity.

.796

U_BeEnv

BIM tools are fit for purpose within the project environment.

.768

U_BeAll

BIM tools ensure that project parties pursue “best for all” instead of “best for self” options.

.736

U_BeMoney

BIM provides the most value for money when weighed against other coordination methodologies.

.579

Cronbach’s alpha 5 .898

(continued)

Empirical Investigation – United States

163

Table 37.  Measurement of constructs in the SEM model for the U.S. data. (continued) Resistance to use

Loadings

Reliability

Behavioral intention to use (BI) BI_IndiRaise

If I work with BIM, I will increase my chances of getting a raise.

.868

BI_Skillful

It would be easy for me to become skillful at working with BIM.

.836

BI_Inter

BIM makes work more interesting.

.852

BI_Fun

Working with BIM is fun.

.889

BI_Like

I like working with BIM.

.879

BI_Should

People who are important to me think that I should use BIM.

.709

Cronbach’s alpha 5 .912

Actual system use (ASU) ASU_IndiInvo

What is your involvement with BIM within your job?

.743

ASU_CoInvo

What is your company’s involvement with BIM?

.794

ASU_IndiVolu

To what extent is the decision (on a personal basis) to work with BIM voluntary within your company?

.695

ASU_BeCoor

To what extent did BIM increase coordination on the project in question?

.904

ASU_BeColla

To what extent did BIM increase collaboration?

.873

ASU_BeInno

To what extent did BIM encourage innovation from the project team?

.853

ASU_OrHelp

The senior management of my organization has been helpful in the use of BIM.

.763

ASU_CoSup

In general, the organization has supported the use of BIM.

.776

Cronbach’s alpha 5 .919

Project performance ProSatis

The BIM project in question satisfied the owner’s needs.

.829

ProSpeci

The project met specifications.

.807

ProCost

The project was completed on budget (making appropriate allowances for any major scope changes).

.858

ProTime

The project was completed on time (making appropriate allowances for any major scope changes).

.862

Cronbach’s alpha 5 .854

also reported. RMSEA is an index sensitive to the number of parameters estimated in the model, so it can help choose a parsimonious model. An RMSEA below 0.08 shows a good fit (MacCallum et al., 1996). Another index is GFI, which measures the proportion of variance that can be accounted for by the model. The second category, incremental fit indices, allows researchers to compare a model’s fit against a baseline model that assumes that all variables are uncorrelated. Comparative fit index (CFI) is a common choice. This index is in the range of 0 to 1. A value of greater than 0.9 can ensure a poorly specified model is detected (Hu & Bentler, 1999). Finally, it is useful to examine whether a model is accepted as a result of including unnecessary variables. The parsimony goodness-of-fit

164

Incentivizing Collaborative BIM-Enabled Projects

Table 38.  Diagnostic statistics of key constructs. X2/Df

p

RMSEA

PGFI

0.28

0.07

0.20

P . 0.05

,0.08

GFI

CFI

Resistance Default model Criteria of good fit

2.53 Not significant

0.98

1.00

.0.90

.0.90

Incentivization 0.00 Not significant

0.72

0.00

P . 0.05

,0.08

0.10

1.00

1.00

.0.90

.0.90

0.89

0.97

.0.90

.0.90

0.96

1.00

.0.90

.0.90

Perceived usefulness 26.55 Not significant

0.15

0.076

P . 0.05

,0.08

0.53

0.00

P . 0.05

,0.08

0.494

Behavioral intention to use 7.10 Not significant

0.37

Actual system use 9.66 Not significant

0.29

0.06

P . 0.05

,0.08

0.36

0.94

0.99

.0.90

.0.90

Project performance 19.46 Not significant

0.00

0.04

P . 0.05

,0.08

0.17

0.90

0.91

.0.90

.0.90

index (PGFI) is also reported, so it penalizes model complexity. With no consensus threshold level for this statistic, it should be interpreted in conjunction with other indices. As reported in Table 38, the model fit is well achieved compared to the threshold value of each indicator suggested in the literature. The corroboration of the validity of three constructs lays a solid foundation for the credibility of the statistical analysis. Based on the same set of diagnostic statistics, Figure 31 shows the whole model is reasonably well fit. As reported in Table 39, the estimated path coefficients are all statistically significant, which means that the 10 hypotheses under study all prove to be valid.

Discussions As the most advanced country in BIM applications, the United States experiences provide a fertile ground for testing the interlocking relationships between user behavior and the BIM implementation environment. The acid test of BIM lies in its effect on project performance.

165

Empirical Investigation – United States

e33

RESI_DeConc RESI_DeConIn

e32

e37

.83 .90

Resistance

.78

.81

RESI_ReInve

e31

-.69

e11

EX_Clear

e10

EX_Assist

e9

EX_Res

.53

.41

.31

External Support

IncenObj

e22

IncenGr

e21

IncenMon

e5

U_BeAll

e8

.36

.66

Behavioural Incention to Use .83

e41

e40

e39

Actual System Use

.60

.90

.84

.67

Project Performance

.56

.55

.87

.92

.43

.75 .83

e4

.61

.76

.52

Perceived .35 Ease of Use

.31

e23

U_IndiProd U_BeMoney

e35

.95 .58

.53

Perceived Usefulness

BI_Should BI_Like BI_IndiRaise ASU_OrHelp ASU_BeColla ASU_IndiVolu ProTime ProCost

Incentivization

.74

e38

e30

X²/Df Default model 1.309 Criteria of good fit Not significant Model fit is relatively acceptable.

e29

e25

P 0.003 P>0.05

e17

RMSEA 0.074 0.90

e20

e19

CFI 0.910 >0.90

Figure 31.  Estimation of the SEM model for the U.S. data.

The corroboration of 10 hypotheses enables us to paint an overall picture of what BIM can do. While the supporting mechanisms provided to facilitate BIM adoption by an organization can be undermined by user resistance (H1), this negative effect is offset by good incentivization measures (H2). The net effect of external support has a positive impact on the Table 39.  Estimated path coefficients of the U.S. SEM model. Path

S.E.

C.R.

H2

EX



I

Estimate .312

.145

2.156

.031

par_19

H1

EX



R

2.536

.165

23.244

.001

par_20

H5

E_Easy



I

.601

.189

3.180

.001

par_18

H3

E_Easy



EX

.494

.194

2.543

.011

par_22

H6

U



E_Easy

.468

.133

3.514

***

par_23

H4

U



EX

.379

.189

2.002

.045

par_24

H7

BI



U

.906

.232

3.897

***

par_14

H8

BI



E_Easy

.371

.156

2.386

.017

par_17

H9

ASU



BI

.450

.142

3.159

.002

par_15

H10

P



ASU

.444

.146

3.044

.002

par_16

All paths are significant, which means all the hypotheses are accepted.

P

Label

166

Incentivizing Collaborative BIM-Enabled Projects

perception of usefulness (H4) and ease of use (H3). The practical value of BIM in the eyes of users may also be strengthened via the chain effect of incentivization on the perceived ease of use (H3) and then on the perception of usefulness (H6). The joint effect of perceived usefulness and perceived ease of use is found to be a decisive factor in the determination of intention to use (H7 1 H8), which in turn will boost the actual use of BIM (H9) and eventually lead to better project performance (H10).

C H A P T E R

13

Conclusions This research makes considerable efforts to explore the influence of implicit and explicit incentivization measures on project performance in three national contexts. As the three countries’ studies are in different stages of BIM maturity, the central questions explored for each country are different. The main conclusions are summarized as follows:

China Survey (1): BIM and IPD In recent years, BIM has been feverishly promoted by governments throughout the world by issuing mandates to force the adoption of BIM. The justification for these mandates is restricted to current rather than long-term benefits. In addition to BIM, promoting IPD has also attracted considerable government effort (e.g., Cabinet Office, 2014). While IPD is not yet piloted in China, the awareness of its importance has emerged. For instance, more than half of the respondents in Ni and Wang (2015) agreed that there should be a suitable delivery system to support BIM. The statistical analysis of this research shows that potential cost savings aside, BIM could also propel procurement reform in the long run. This finding not only lends empirical support to the BIM mandate in China, but also predicts that the wider application of BIM can facilitate the implementation of integrated delivery in the country. This evidence can also be drawn upon by governments when considering enacting a new BIM mandate or extending an existing one. Using the data from 145 Chinse BIM-enabled projects, this research can further probe the channels through which BIM application could have impacted IPD acceptability: First, the firsthand experience of working in a BIM-enabled environment can make practitioners better appreciate the importance of incentivization, and that perception can drive the acceptability of IPD; second, observing the positive impact of 167

168

Incentivizing Collaborative BIM-Enabled Projects

BIM on communication quality can translate into another drive to support IPD. It is hoped that these robust statistical relationships can spark follow-on research to investigate the benefits of BIM in a wider context.

 hina Survey (2): BIM and Incentivization Under the Influence C of Behavioral Biases The benefit of BIM is sensitive to the way it is implemented in the project, and its potential cannot be fully reaped in a delivery environment where project parties are not well incentivized to harness BIM in improving communication and coordination across project stages. However, the employment of integrated systems could present a great challenge in some countries, as it may require new legislation and heavy training. It is easier to achieve the second-order incentive alignment through a piecemeal implementation of incentivization measures. The data collected from 223 Chinese BIM-enabled projects reveal that the perception of actual incentivization created by the project delivery environment has a signification impact on the perceived usefulness of advanced incentivization systems. The establishment of this awareness could herald a smoother process when introducing these systems into BIM-enabled projects. Embedding BIM in a better-incentivized environment provides a self-sustaining driver for the proliferation of BIM, which will eventually pave the way for the acceptance of integrated delivery systems.

UK Survey The UK empirical study aims to explore the chain effect of project attributes, delivery systems, BIM incentivization, behavioral biases, and project performance. The main conclusions include the following: When the project is complex, the owner tends to involve contractors earlier and is more likely to set up incentive pools to motivate project parties. The use of incentive pools increases the employment of group-based incentives and can yield a direct impact on the joint development of project goals. In projects where project parties are involved early, decisions are more likely to be made jointly. Objective evaluation of project performance can lead to a better quality of joint decision making, which can then strengthen the effectiveness of incentive mechanisms. Experience can improve the working of group-based incentive mechanisms, but the

Conclusions

169

contractor’s appetite for avoiding significant losses would work the opposite. The effective functioning of incentives can eventually lead to an improvement in BIM effectiveness.

U.S. Survey The fourth survey conducted in this research explores the factors that could change the effectiveness of external support given to BIM users, and the pathways through which external support could eventually impact project performance. As predicted by theory, user resistance is an inhibitor, while incentivization is a booster to the efficacy of the support provided by the BIM-implementation organization. The effect of the user support is transmitted through two channels: perceived usefulness of BIM and perceived ease of use.

Appendix Budget a. Budget—current budget with an explanation of the deviations from the planned one at the proposal. The budget allocated to the project by University College London is �28,312. There could be an underspent amount of around �8,000 (not all expenses have been claimed yet). Most of the savings are from assistant costs. Originally, this project planned to hire part-time postgraduate assistants to do most of the work, but the quality of assistantship turned out to be patchy. The PI was forced to take some of the work back to avoid jeopardizing the quality of the report. After accounting for the time PI spent in writing the report, the budget is about enough to cover the total cost.

Summary of Proposal Objectives vs. Research Accomplishments Proposal objectives

Research accomplishments

1. Conduct three case studies covering the participants in supply chains of a typical BIM-enabled project in three national contexts (U.S., UK, China).

China: 1. Pilot case studies (1)  Maotai Distillery extension project (2)  Guian Foxconn (3)  Xixian bonded area service center (4)  Shanghai Tongji University Sports Center (5)  Guangzhou 21 Line Rail Transit Project 2. Case studies (1)  Shanghai World Center Project (2)  Shanghai Leisure Park Project UK: 1. Pilot case studies Network Rail’s East Coast Main Line Project 2. Case studies (1)  Anglian water (@one alliance) (2)  Environment agency (Water and Environment Management; WEM) US: 1. Pilot case studies (1)  Sutter Medical Center, Castro Valley (2)  Autodesk Waltham project (3)  Sutter Health Fairfield Medical Office Building

(continued) 171

172

Incentivizing Collaborative BIM-Enabled Projects

Summary of Proposal Objectives vs. Research Accomplishments (continued) Proposal objectives

Research accomplishments

(4)  Cardinal Glennon Children’s Hospital Expansion (5)  Encircle Health Ambulatory Care Center 2. Case studies San Francisco International Airport (IPD) 2. Build a theory for IPD parties’ responses In Chapter 4, this research sets out a theoretical framework, and terms the to a given incentive system by drawing life cycle theory of BIM diffusion (Figure 2), drawing upon transaction cost on the methods of agency approach and economics and the principal-agent theory. The merit of this framework is to adbehavioral approach. dress BIM incentivization issues for countries at different BIM maturity stages. Figure 3 sets out a framework by which BIM incentivization issues can be addressed in their own right. As this research takes a broader view in evaluating the impact of BIM-incentivization measures, this framework is always embedded in national business environments and delivery environments in the analysis of this research. 3. Develop hypotheses and test them via econo- The table below provides the full list of the hypotheses tested and metric analysis to form an evidence base to corroborated. guide the design of incentivization systems for BIM-enabled collaborative governance.

Level 1 National environment

Level 2 Delivery environment

Level 3 Incentive structure

BIM embedded in conventional delivery systems.

The choice of delivery system is not explicitly modeled, but its impact is captured.

The impact of incentivization is not explicitly modeled, but the awareness of its importance is measured.

BIM embedded in conventional delivery systems.

The degree of governance alignment is measured by its impact on the strength of motivations to BIM participation.

The impacts of incentivization are analyzed with consideration of behavioral biases.

Hypotheses tested H1: The degree of BIM application can raise the perceived importance of BIM incentivization. H2: Perceived importance of BIM incentivization will have a positive effect on IPD acceptability. H3: The degree of BIM application can improve the quality of communication. H4: Better communication quality will lead to greater IPD acceptability. H5: The degree of BIM application can improve the quality of collaboration. H1: The presence of behavioral biases in the project team has a positive impact on the effectiveness of incentivization in BIM-enabled projects. H2: The strength of incentivization created by the delivery environment can positively impact the utilization of BIM in the project. H3: The greater utilization of BIM in the project can improve project performance. H4: The performance of a BIM-enabled project can positively affect the awareness of advanced incentivization measures needed for propelling effective BIM participation. H5: The effect of incentivization in the current project has a direct and positive impact on the awareness of advanced incentivization measures needed for propelling effective BIM participation.

(continued)

Appendix

Level 1 National environment

Level 2 Delivery environment

Level 3 Incentive structure

BIM possibly embedded in conventional or integrated delivery systems.

The impacts of project attributes on project governance design are addressed.

The joint effects of project attributes and project governance features on the use of incentivization measures are addressed.

BIM possibly embedded in conventional or integrated delivery systems.

Governance alignment is assumed.

173

Hypotheses tested H1: Greater project complexity will lead to early involvement of project contractors. H2: Greater project complexity will result in the greater use of incentive pools as motivational mechanisms. H3: The employment of incentive pools increases the use of group-based incentives. H4: The employment of group-based incentives has a direct impact on the joint development of project goals. H5: Subjective evaluation of project performance has a negative effect on the extent of decisions being made jointly in the project. H6: Early involvement of project participants positively impacts joint decision making. H7: Joint decision-making process has a positive impact on the effectiveness of incentives mechanisms. H8: Experienced contractors do not exhibit endowment effect, and thus experience positively influences the use of group-based incentive mechanisms in construction projects. H9: Loss aversion due to the endowment effect has a negative impact on the effective use of financial incentives. H11: Effective incentives have a positive impact on the effective use of BIM. H1: Resistance to use of new technology will have negative impacts on the external support. H2: BIM incentivization can reinforce the external support. H3: BIM incentivization will increase the perceived ease of use. H4: External (organizational) support will have positive impacts on perceived usefulness. H5: External (organizational) support will have positive impacts on perceived ease of use. H6: Perceived ease of use would have positive impacts on perceived usefulness. H7: Perceived usefulness will have positive impacts on behavioral intention to use. H8: Perceived ease of use would have positive impacts on behavioral intention to use. H9: Behavioral intention to use will have positive impacts on actual system use. H10: Actual system use will have positive impacts on project performance.

References Abdirad, H., & Pishdad-Bozorgi, P. (2014). Developing a framework of metrics to assess collaboration in integrated project delivery. In Proceedings of the 50th Annual International Conference of the Associated Schools of Construction, Virginia Polytechnic Institute and State University, VA. Alchian, A. A., & Demsetz, H. (1972). Production, information costs, and economic organization. The American Economic Review, 62(5), 777–795. Alderfer, C. P., & Bierman, H. (1970). Choices with risk: Beyond the mean and variance. The Journal of Business, 43(3), 341–353. American Institute of Architects. (2007). Integrated project delivery: A guide. New York, NY: American Institute of Architects. American Institute of Architects. (2010a). Integrated project delivery: For public and private owners. New York, NY: American Institute of Architects. American Institute of Architects. (2010b). On compensation: Considerations for teams in a changing industry. New York, NY: American Institute of Architects. American Institute of Architects. (2011). Primer on project delivery (2nd ed.). New York, NY: American Institute of Architects. Anderson, J. C., & Gerbing, D. W. (1988). Structural equation modeling in practice: A review and recommended two-step approach. Psychological Bulletin, 103(3), 411. Anderson, L. R., & Holt, C. A. (1997). Information cascades in the laboratory. The American Economic Review, 87(5), 847–862. Ansar, A., Flyvbjerg, B., Budzier, A., & Lunn, D. (2014). Should we build more large dams? The actual costs of hydropower megaproject development. Energy Policy, 69, 43–56. Antle, R., & Demski, J. S. (1988). The controllability principle in responsibility accounting. Accounting Review, 700–718. Anvuur, A. M., & Kumaraswamy, M. M. (2007). Conceptual model of partnering and alliancing. Journal of Construction Engineering and Management, 133(3), 225–234. 175

176

Incentivizing Collaborative BIM-Enabled Projects

Arayici, Y., Coates, P., Koskela, L., Kagioglou, M., Usher, C., & O’Reilly, K. (2011). Technology adoption in the BIM implementation for lean architectural practice. Automation in Construction, 20, 189–195. Arrow, K. J. (1965). Aspects of the theory of risk-bearing. Helsinki, Finland: Yrjö Jahnssonin Säätiö. Ashraf, H. W. (2008). Implementing BIM: A report from the field on the issues and strategies. Proceedings of the 47th annual meeting of invited attorneys. June 2008. Seattle, WA. Australasia, B. (2012). National building information modelling initiative: Volume 1 Strategy. Retrieved from https://buildingsmart.org .au/wp-content/uploads/2014/03/NationalBIMIniativeReport _6June2012.pdf Australian Government. (2011). National alliance contracting guidelines: Guide to alliance contracting. Retrieved from Canberra on September 5, 2018 C:/Users/bf298/Desktop/National_Guide_to _Alliance_Contracting.pdf. Autodesk White Paper. (2008). Improving building industry results through integrated project delivery and building information modeling report on integrated practice. Retrieved from http://images .autodesk.com/adsk/files/bim_and_ipd_whitepaper.pdf Baddeley, M. (2013). Behavioral economics and finance. London, UK: Routledge. Baddeley, M., & Chang, C. (2015). Collaborative building information modelling (BIM): Insights from behavioral economics and incentive theory. Retrieved from http://www.rics.org/us/knowledge/research /research-reports/collaborative-bim/ Baddeley, M., Curtis, A., & Wood, R. (2004). An introduction to prior information derived from probabilistic judgements: Elicitation of knowledge, cognitive bias, and herding. Geological Society, 239, 15–27. Badenfelt, U. (2008). The selection of sharing ratios in target cost contracts. Engineering, Construction and Architectural Management, 15(1), 54–65. Badenfelt, U. (2010). I trust you, I trust you not: A longitudinal study of control mechanisms in incentive contracts. Construction Management and Economics, 28, 301–310. doi:10.1080/01446191003587745 Bagozzi, R. P., & Yi, Y. (2012). Specification, evaluation, and interpretation of structural equation models. Journal of the Academy of Marketing Science, 40(1), 8–34. Baiman, S., & Rajan, M. V. (1995). The informational advantages of discretionary bonus schemes. Accounting Review, 557–579.

References

177

Baker, G. (1992). Incentive contracts and performance measurement. Journal of Political Economy, 100(3), 598–614. Baker, G., Gibbons, R., & Murphy, K. J. (1994). Subjective performance measures in optimal incentive contracts. The Quarterly Journal of Economics, 109(4), 1125–1156. Baker, G., & Gil, R. (2013). Clinical papers in organizational economics. In R. Gibbons & J. Roberts (Eds.), The handbook of organizational economics (pp. 193–212). Princeton, NJ: Princeton University Press. Baker, G., Jensen, M. C., & Murphy, K. J. (1987). Compensation and incentives: Practice vs. theory. The Journal of Finance, 43(3), 593–616. Banker, R. D., & Datar, S. M. (1989). Sensitivity, precision, and linear aggregation of signals for performance evaluation. Journal of Accounting Research, 27(1), 21–39. Bansal, V. K. (2011). Application of geographic information systems in construction safety planning. International Journal of Project Management, 29, 66–77. doi:10.1016/j.ijproman.2010.01.007 Barberis, N., & Thaler, R. (2003). A survey of behavioral finance. Handbook of the Economics of Finance, 1, 1053–1128. Baron, J. N., & Kreps, D. M. (2013). Employment as an economic and a social relationship. In R. Gibbons & J. Roberts (Eds.), The handbook of organizational economics (pp. 315–341). Princeton, NJ: Princeton University Press. Baron, R. M., & Kenny, D. A. (1986). The moderator–mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51(6), 1173. Basu, A. K., & Kalyanaram, G. (1990). On the relative performance of linear versus nonlinear compensation plans. International Journal of Research in Marketing, 7(2), 171–178. Baxter, P., & Jack, S. (2008). Qualitative case study methodology: Study design and implementation for novice researchers. The Qualitative Report, 13(4), 544–559. Becerik-Gerber, B., Jazizadeh, F., Li, N., & Calis, G. (2012). Application areas and data requirements for BIM-enabled facilities management. Journal of Construction Engineering and Management, 138, 431–442. Bernstein, P. G., & Pittman, J. H. (2004). Barriers to the adoption of building information modeling in the building industry. Autodesk Building Solutions, White Paper, 1–14.

178

Incentivizing Collaborative BIM-Enabled Projects

Bew, M., & Richards, M. (2008). BIM maturity model. In Construct IT Autumn 2008 Members’ Meeting. Brighton, UK. Bikhchandani, S., Hirshleifer, D., & Welch, I. (1998). Learning from the behavior of others: Conformity, fads, and informational cascades. The Journal of Economic Perspectives, 12(3), 151–170. BIS (Department for Business, Innovation and Skills). (2011). A report for the government construction client group building information modelling (BIM) working party strategy paper. Retrieved from http://www.bimtaskgroup.org/wp-content/uploads/2012/03 /BIS-BIM-strategy-Report.pdf Black, C., Akintoye, A., & Fitzgerald, E. (2000). An analysis of success factors and benefits of partnering in construction. International Journal of Project Management, 18(6), 423–434. Bolton, P., & Dewatripont, M. (2005). Contract theory: Cambridge, MA: MIT Press. Boukendour, S., & Hughes, W. (2014). Collaborative incentive contracts: Stimulating competitive behaviour without competition. Construction Management & Economics, 32, 279–289. doi:10.1080/01446193 .2013.875215 Brennan, M. D. (2011). Integrated project delivery: A normative model for value creation in complex military medical projects. Doctoral dissertation, University of Illinois at Urbana–Champaign. Bresnen, M., & Marshall, N. (2000). Motivation, commitment and the use of incentives in partnerships and alliances. Construction Management & Economics, 18(5), 587–598. Brewer, W., & Mendelson, M. I. (2003). Methodology and metrics for assessing team effectiveness. International Journal of Engineering Education, 19(6), 777–787. Broome, J. (2002). Procurement routes for partnering: A practical guide. London, UK: Thomas Telford Ltd. Broome, J., & Perry, J. (2002). How practitioners set share fractions in target cost contracts. International Journal of Project Management, 20(1), 59–66. Browne, M. W., & Cudeck, R. (1993). Alternative ways of assessing model fit. Sage focus editions, 154, 136–136. BSI (2013). PAS 1192-2:2013 Specification for information management for the capital/delivery phase of construction projects using BIM, London: BSI Limited. BSI (2014). PAS 1192-3:2014 Specification for information management for the operational phase of assets using BIM, London: BSI Limited.

References

179

BuildLACCD. (2016). LACCD building information modeling standards. Los Angeles, CA: BuildLACCD. Bull, C. (1987). The existence of self-enforcing implicit contracts. The Quarterly Journal of Economics, 102(1), 147–159. Bushman, R. M., & Indjejikian, R. J. (1993). Accounting income, stock price, and managerial compensation. Journal of Accounting and Economics, 16(1), 3–23. Bushman, R. M., Indjejikian, R. J., & Smith, A. (1995). Aggregate performance measures in business unit manager compensation: The role of intrafirm interdependencies. Journal of Accounting Research, 33, 101–128. Bygballe, L. E., Jahre, M., & Swärd, A. (2010). Partnering relationships in construction: A literature review. Journal of Purchasing and Supply Management, 16(4), 239–253. Cabinet Office. (2011). Government construction strategy. London, UK: HMSO. Cabinet Office. (2014). New models of construction procurement. London, UK: HMSO. Camerer, C. F., & Hogarth, R. M. (1999). The effects of financial incentives in experiments: A review and capital-labor-production framework. Journal of Risk and Uncertainty, 19(1–3), 7–42. Cao, D., Li, H., Wang, G., & Huang, T. (2017). Identifying and contextualizing the motivations for BIM implementation in construction projects: An empirical study in China. International Journal of Project Management. 35(4), 658–669. Cao, D., Wang, G., Li, H., Skitmore, M., Huang, T., & Zhang, W. (2015). Practices and effectiveness of building information modeling in construction projects in China. Automation in Construction, 49, 113–122. doi:10.1016/j.autcon.2014.10.014 Chang, C. (2013). Understanding the hold-up problem in the management of megaprojects: The case of the channel tunnel rail link project. International Journal of Project Management 31(4), 628–637. Chang, C. (2013a). Understanding the hold-up problem in the management of megaprojects: The case of the channel tunnel rail link project. International Journal of Project Management, 31(4), 628–637. Chang, C. (2013b). A critical analysis of recent advances in the techniques for the evaluation of renewable energy projects. International Journal of Project Management, 31, 1057–1067. Chang, C. (2013c). When might a project company break up? The perspective of risk-bearing capacity. Construction Management and Economics, 31, 1186–1198.

180

Incentivizing Collaborative BIM-Enabled Projects

Chang, C. Y. (2014a). Assessment of empirical techniques for the study of BIM incentivization: Roadmap for knowledge accumulation. Working paper. London, UK: Bartlett School of Construction and Project Management, University College London. Chang, C. Y. (2014b). Book review, Handbook of organizational economics. Construction Management & Economics, 32(3), 309–313. Chang, C. Y. (2014c). Incentivizing building information modelling projects: Key questions, assessment, and research opportunities. Working paper. London, England: Bartlett School of Construction and Project Management. Chang, C. Y. (2014d). Principal-agent model of risk allocation in construction contracts and its critique. ASCE Journal of Construction Engineering and Management, 140(1), 04013032. Chang, C. Y. (2015). Risk-bearing capacity as a new dimension to the analysis of project governance. International Journal of Project Management, 33(6), 1195–1205. doi:10.1016/j.ijproman.2015.02.003 Chang, C. Y., & Chen, S. (2016). Transitional public-private partnerships model in China: Contracting with little recourse to contracts. Journal of Construction Engineering and Management, 142(10), 05016011. Chang, C. Y., & Chou, H.Y. (2014). Transaction-cost approach to the comparative analysis of user-pay and government-pay public-private partnership systems. Journal of Construction Engineering and Management. 140:04014039-1-04014039-11. Chang, C. Y., & Howard, R. (2016). How to incentivize BIM participation? Conceptual framework and empirical evidence. London, UK: Bartlett School of Construction and Project Management, University College London. Chang, C. Y., & Ive, G. (2000). A comparison of two ways of applying transaction costs approach (II): The case of construction procurement routes, mimeo. London, UK: Bartlett School of Graduate Studies, University College London. Chang, C. Y., & Ive, G. (2007). Reversal of bargaining power in construction projects: Meaning, existence and implications. Construction Management and Economics, 25(8), 845–855. Chang, C. Y., & Ko, J. W. (2017) New approach to estimating the standard deviations of lognormal cost variables in the Monte Carlo analysis of construction risks. Journal of Construction Engineering and Management, 143(1), Article 06016006. 10.1061/(ASCE)CO .1943-7862.0001207.

References

181

Chang, C. Y., & Mills, G. (2016). Implementation lessons of introducing dual governance: The case of London Underground’s STAKE system. London, UK: Bartlett School of Construction and Project Management, University College London. Chang, C. Y., Pan, W., & Howard, R. (2017). Impact of building information modelling implementation on the acceptance of integrated delivery systems: Structural equation modelling analysis. ASCE Journal of Construction Engineering and Management 143(8), 04017044. Chang, C. Y., & Qian, Y. M. (2015). An econometric analysis of holdup problems in construction projects. Journal of Construction Engineering and Management, 141(6), 04015004. Chen, G., Zhang, G., Xie, Y. M., & Jin, X. H. (2012). Overview of alliancing research and practice in the construction industry. Architectural Engineering and Design Management, 8, 103–119. doi:10.1080/17452 007.2012.659505 Chen, J. Y., & Miller, B. L. (2009). On the relative performance of linear vs. piecewise-linear-threshold intertemporal incentives. Management Science, 55(10), 1743–1752. Chen, Y. Q., & Jiao, J. S. (2011). Influence of project delivery system and payment methods on project cost performance, Journal of Tongji University (Natural Science), 2011(09), 1407–1412. Cohen, J. (2010). Integrated project delivery: Case studies. AIA National, AIA California Council, AGC California and McGraw-Hill. Crotty, R. (2011). The impact of building information modelling: transforming construction. Oxon, UK: SPON Press. Daniel, W. M., Chan, P., Lam, T. I., Chan, A. P. C., & Wong, J. M. W. (2010). Achieving better performance through target cost contracts: The tale of an underground railway station modification project. Emerald Insight, 28(5/6), 261–277. Darrington, J. W. (2010). Addressing human motivation in construction contracts. The Voice of the Construction Industry, 40, 18–19. Darrington, J. W., & Howell, G. A. (2011). Motivation and incentives in relational contracts. Journal of Financial Management of Property and Construction, 16, 42–51. doi:10.1108/13664381111116070 Davis, F. D. (1989). Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Quarterly, 13(3), 319–340. Davis, F. D., Bagozzi, R. P., & Warshaw, P. R. (1989). User acceptance of computer technology: A comparison of two theoretical models. Management Science, 35(8), 982–1003.

182

Incentivizing Collaborative BIM-Enabled Projects

Deci, E. L. (1971). Effects of externally mediated rewards on intrinsic motivation. Journal of Personality and Social Psychology, 18(1), 105. Demian, P., & Walters, D. (2014). The advantages of information management through building information modelling. Construction Management and Economics, 32, 1153–1165. doi:10.1080/01446193 .2013.777754 Dossick, C. S., Neff, G., & Homayouni, H. (2009). The realities of building information modeling for collaboration in the AEC industry. Construction Research Congress: Building a Sustainable Future. Seattle, Washington, April 5–7, 2009. Eadie, R., Browne, M., Odeyinka, H., McKeown, C., & McNiff, S. (2013). BIM implementation throughout the UK construction project life­ cycle: An analysis. Automation in Construction, 36, 145–151. Eadie, R., Odeyinka, H., Browne, M., McKeown, C., & Yohanis, M. (2014). Building information modelling adoption: An analysis of the barriers to implementation. Journal of Engineering and Architecture, 2, 77–101. East, B., & Carrasquillo-Mangual, M. (2012). The COBie guide: A commentary to the NBIMS-US COBie Standard. Building SMART Alliance, 1–125. Eastman, C., Teicholz, P., Sacks, R., & Liston, K. (2011). BIM handbook: A guide to building information modeling for owners, managers, designers, engineers, and contractors. Hoboken, NJ: Wiley. Eisenhardt, K. M. (1989). Building theories from case study research. Academy of Management Review, 14(4), 532–550. El-adaway, I. H. (2010). Integrated project delivery case study: Guidelines for drafting partnering contract. Journal of Legal Affairs and Dispute Resolution in Engineering and Construction, 2(4), 248–254. El Asmar, M. (2012). Modeling and benchmarking performance for the integrated project delivery (IPD) system. PhD dissertation, Madison, WI: University of Wisconsin–Madison. Eriksson P., & Westerberg M. (2010). Effects of procurement on construction project performance. Working paper. Division of Entrepreneurship and Industrial Organisation, Lulea University of Technology, Sweden. Ertel, D., Jeff. W., & Laura, V. (2001). Managing strategic alliances: A cross-industry study of how to build and manage successful alliances. Brighton, MA: Vantage Partners. Eswaran, M., & Kotwal, A. (1984). The moral hazard of budget-breaking. The Rand Journal of Economics, 578–581.

References

183

Feltham, G. A., & Xie, J. (1994). Performance measure congruity and diversity in multi-task principal/agent relations. Accounting Review, 429–453. Flyvbjerg, B. (2014). What you should know about megaprojects and why: An overview. Project Management Journal, 45(2), 6–19. Flyvbjerg, B., Garbuio, M., & Lovallo, D. (2009). Delusion and deception in large infrastructure projects: Two models for explaining and preventing executive disaster. California Management Review, 51(2), 170–193. Freeman, J., Weil, S. A., & Hess, K. P. (2006). Measuring, monitoring and managing knowledge in command and control organizations. In Virtual Media for Military Applications (pp. 2–10), Neuilly-sur-Seine, France: RTO. Frey, B. S., & Jegen, R. (2001). Motivation crowding theory. Journal of Economic Surveys, 15(5), 589–611. Frey, B. S., & Oberholzer-Gee, F. (1997). The cost of price incentives: An empirical analysis of motivation crowding-out. The American Economic Review, 87(4), 746–755. Gibbons, R. (1997). Incentives and careers in organizations. In D. Kreps & K. Wallis (Eds.), Advances in Economics and Econometrics: Theory and Applications (Vol. II, pp. 1–37). Cambridge, England: Cambridge University Press. Gibbons, R. (1998). Incentives in organizations. The Journal of Economic Perspectives, 12(4), 115–132. Gibbons, R., & Roberts, J. (2013). The handbook of organizational economics. Princeton, NJ: Princeton University Press. Gibbs, M., Merchant, K. A., Van der Stede, W. A., & Vargus, M. E. (2004). Determinants and effects of subjectivity in incentives. The Accounting Review, 79(2), 409–436. Gjesdal, F. (1988). Piecewise linear incentive schemes. The Scandinavian Journal of Economics, 90(3), 305–328. Glaser, B., & Strauss, A. (1967). The discovery of grounded theory: Strategies for qualitative research. London, UK: Wiedenfeld and Nicholson. Gneezy, U., & Rustichini, A. (2000). Pay enough or don’t pay at all. The Quarterly Journal of Economics, 115(3), 791–810. Grossman, S. J., & Hart, O. D. (1983). An analysis of the principal-agent problem. Econometrica: Journal of the Econometric Society, 7–45. Grubb, M., Brophy Haney, A., & Wilde, J. (2009). Plugging the gap in energy efficiency policies: The emergence of the UK carbon reduction commitment. European Review of Energy Markets, 3(2), 33–62.

184

Incentivizing Collaborative BIM-Enabled Projects

Hair, J. F., Ringle, C. M., & Sarstedt, M. (2011). PLS-SEM: Indeed a silver bullet. Journal of Marketing Theory and Practice, 19(2), 139–152. Harris, J. D., Johnson, S. G., & Souder, D. (2013). Model-theoretic knowledge accumulation: The case of agency theory and incentive alignment. Academy of Management Review, 38(3), 442–454. Hart, O., & Holmstrom, B. (1987). The theory of contracts. In T. F. Bewley (Ed.), Advances in Economic Theory, Fifth World Congress of the Econometric Society (pp. 71–155). Hartmann, T., Van Meerveld, H., Vossebeld, N., & Adriaanse, A. (2012). Aligning building information model tools and construction management methods. Automation in Construction, 22, 605–613. doi:10 .1016/j.autcon.2011.12.011 Hayes, R. M., & Schaefer, S. (2000). Implicit contracts and the explanatory power of top executive compensation for future performance. The Rand Journal of Economics, 273–293. HM Government. (2015). Digital Built Britain: Level 3 building information modelling—Strategic plan. Retrieved from https://assets.publishing .service.gov.uk/government/uploads/system/uploads/attachment _data/file/410096/bis-15-155-digital-built-britain-level-3-strategy.pdf HM Treasury. (2010). Infrastructure cost review: Main report. London, UK: HMSO. HM Treasury. (2013). Infrastructure procurement routemap: A guide to improving delivery capability. London, UK: HMSO. Holmstrom, B. (1979). Moral hazard and observability. The Bell Journal of Economics, 10(1), 74–91. Holmstrom, B. (1982). Moral hazard in teams. The Bell Journal of Economics, 13(2), 324–340. Holmstrom, B., & Milgrom, P. (1987). Aggregation and linearity in the provision of intertemporal incentives. Econometrica, 55(2), 303–328. Holmstrom, B., & Milgrom, P. (1991). Multitask principal-agent analyses: Incentive contracts, asset ownership, and job design. Journal of Law, Economics and Organization, 7, 24–52. Hong, Y., Chan, D. W., Chan, A. P., & Yeung, J. F. (2011). Critical analysis of partnering research trend in construction journals. Journal of Management in Engineering, 28(2), 82–95. Howard, R., Restrepo, L., & Chang, C. Y. (2016). Addressing individual perceptions: An application of the unified theory of acceptance and use of technology to building information modelling. International Journal of Project Management. 35(2), 107–120. doi:10.1016/j.ijproman.2016.10.012

References

185

Hu, L. T., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling: A Multidisciplinary Journal, 6(1), 1–55. Ichniowski, C., Shaw, K., & Prennushi, G. (1997). The effects of human resource management practices on productivity: A study of steel finishing lines. American Economic Review, 87(3), 291–313. Ive, G., & Chang, C. Y. (2007). The principle of inconsistent trinity in the selection of procurement systems. Construction Management and Economics, 25(7), 677–690. Jensen, M. C., & Murphy, K. J. (1990). Performance pay and top-­ management incentives. Journal of Political Economy, 98(2), 225–264. Jolls, C., Sunstein, C. R., & Thaler, R. (1998). A behavioral approach to law and economics. Standard Law Review, 50(5), 1471–1550. Jöreskog, K. G. (1963). Statistical estimation in factor analysis: A new technique and its foundation. Stockholm, Sweden: Almqvist & Wiksell. Just, D. R. (2014). Introduction to behavioral economics. Hoboken, NJ: Wiley. Kadefors, A. (2004). Trust in project relationships—inside the black box. International Journal of Project Management, 22, 175–182. doi:10.1016/S0263-7863(03)00031-0 Kahneman, D. (2011). Thinking, fast and slow. New York, NY: Farrar, Straus and Giroux. Kahneman, D., Knetsch, J. L., & Thaler, R. H. (1991). Anomalies: The endowment effect, loss aversion, and status quo bias. The Journal of Economic Perspectives, 193–206. Kahneman, D., & Lovallo, D. (1993). Timid choices and bold forecasts: A cognitive perspective on risk taking. Management Science, 39(1), 17–31. Kahneman, D., & Tversky, A. (1979). Prospect theory: An analysis of decision under risk. Econometrica, 47(2), 263–291. Kamenica, E. (2012). Behavioral economics and psychology of incentives. Annual Review of Economics, 4(1), 427–452. Kandel, E., & Lazear, E. P. (1992). Peer pressure and partnerships. Journal of political Economy, 801–817. Kerr, S. (1975). On the folly of rewarding A, while hoping for B. Academy of Management Journal, 18(4), 769–783. Klein, B., Crawford, R. G., & Alchian, A. A. (1978). Vertical integration, appropriable rents, and the competitive contracting process. Journal of Law and Economics, 21(2), 297–326.

186

Incentivizing Collaborative BIM-Enabled Projects

Kline, R. B. (2011). Principles and practice of structural equation modelling (3rd ed.). New York, NY: Guilford Press. Kline, R. B. (2016). Principles and practice of structural equation modeling (4th ed.). New York, NY: Guilford Press. Knetsch, J. L. (1989). The endowment effect and evidence of nonreversible indifference curves. The American Economic Review, 79(5), 1277–1284. Kohn, A. (1993). Why incentive plans cannot work. Harvard Business Review, 1–7. Kreps, D. M. (1997). Intrinsic motivation and extrinsic incentives. The American Economic Review, 87(2), 359–364. Ku, K., & Taiebat, M. (2011). BIM experiences and expectations: The constructors’ perspective. International Journal of Construction Education and Research, 7(3), 175–197. Kuran, T., & Sunstein, C. R. (1999). Availability cascades and risk regulation. Stanford Law Review, 51(4), 683–768. Lahdenperä, P. (2012). Making sense of the multi-party contractual arrangements of project partnering, project alliancing and integrated project delivery. Construction Management and Economics, 30(1), 57–79. Lambert, R. (2001). Contracting theory and accounting. Journal of Accounting and Economics, 32(1), 3–87. Lambert, R., & Larcker, D. F. (1991). Executive compensation, corporate decision making, and shareholder wealth: A review of the evidence. Boston, MA: Harvard Business School Press. Latham, M. (1994). Constructing the team. London, UK: HMSO. Lawler, E. (1990). Strategic pay: Aligning organizational strategies and pay systems. San Francisco, CA: Jossey-Bass. Lepper, M. R., Greene, D., & Nisbett, R. E. (1973). Undermining children’s intrinsic interest with extrinsic reward: A test of the “overjustification” hypothesis. Journal of Personality and Social Psychology, 28 (1), 129. Levinthal, D. A. (2011). A behavioral approach to strategy—What’s the alternative? Strategic Management Journal, 32(13), 1517–1523. Lindblad, H. (2013). Study of the implementation process of BIM in construction projects: Analysis of the barriers limiting BIM adoption in the AEC industry. Unpublished MSc Thesis. List, J. A. (2003). Does market experience eliminate market anomalies? Quarterly Journal of Economics, 118(1), 41–71.

References

187

Liu, Y., Van Nederveen, S., & Hertogh, M. (2016). Understanding effects of BIM on collaborative design and construction: An empirical study in China. International Journal of Project Management, 35(4), 686–698. Löfstedt, R. E. (2004). The swing of the regulatory pendulum in Europe: From precautionary principle to (regulatory) impact analysis. Journal of Risk and Uncertainty, 28(3), 237–260. Love, P. E. D., Davis, P. R., Chevis, R., & Edwards, D. J. (2011). Risk/ reward compensation model for civil engineering infrastructure alliance projects. Journal of Construction Engineering and Management, 137(2), 127–136. Love, P. E. D., Matthews, J., Simpson, I., Hill, A., & Olatunji, O. A. (2014). A benefits realization management building information modeling framework for asset owners. Automation in Construction, 37, 1–10. MacCallum, R. C., Browne, M. W., & Sugawara, H. M. (1996). Power analysis and determination of sample size for covariance structure modeling. Psychological Methods, 1(2), 130–149. Macher, J. T., & Richman, B. D. (2008). Transaction cost economics: An assessment of empirical research in the social sciences. Business and Politics, 10(1), 1–63. MacLeod, W. B., & Parent, D. (1999). Job characteristics and the form of compensation. Research in Labor Economics, 18, 231–259. March, J. G., & Shapira, Z. (1987). Managerial perspectives on risk and risk taking. Management Science, 33(11), 1404–1418. McCloskey, D. N. (1990). Storytelling in economics. Narrative in culture: The Uses of Storytelling in the Sciences, Philosophy, and Literature, 5–22. McGraw-Hill Construction. (2014). The business value of BIM for construction in major global markets. Retrieved from https://www .icn-solutions.nl/pdf/bim_construction.pdf McIntosh, C. N. (2007). Rethinking fit assessment in structural equation modelling: A commentary and elaboration on Barrett (2007). Personality and Individual Differences, Department of Civil and Environmental Engineering, National University of Ireland, Cork, Ireland. 42(5), 859–867. Meng, X. (2012). The effect of relationship management on project performance in construction. International Journal of Project Management, 30, 188–198. doi:10.1016/j.ijproman.2011.04.002 Milgrom, P. R. (1981). Good news and bad news: Representation theorems and applications. The Bell Journal of Economics, 12(2), 380–391. Milgrom, P. R., & Roberts, J. D. (1992). Economics, organization and management, Upper Saddle River, NJ: Prentice-Hall.

188

Incentivizing Collaborative BIM-Enabled Projects

Mohr, J., & Spekman, R. (1994). Characteristics of partnership success: partnership attributes, communication behavior, and conflict resolution techniques. Strategic Management Journal, 15(2), 135–152. Monteiro, A., Meda, P., & Pocas Martins, J. (2014). Framework for the coordinated application of two different integrated project delivery platforms. Automation in Construction, 38, 87–99. doi:10.1016 /j.autcon.2013.11.003 Moore, L. J., Woodward, M. J., & Grogono-Thomas, R. (2005). The occurrence of treponemes in contagious ovine digital dermatitis and the characterisation of associated Dichelobacter nodosus. Veterinary Microbiology, 111(3), 199–209. Mosey, D. (2009). Early contractor involvement in building procurement: Contracts, partnering and project management. Hoboken, NJ: Wiley-Blackwell. Mourshed, M. (2006). Interoperability-based optimisation of architectural design. Doctoral dissertation. Mulaik, S. A., James, L. R., Van Alstine, J., Bennett, N., Lind, S., & Stilwell, C. D. (1989). Evaluation of goodness-of-fit indices for structural equation models. Psychological Bulletin, 105(3), 430. Murphy, K. J. (1999). Executive compensation. Handbook of Labor Economics, 3, 2485–2563. Murphy, K. R., & Cleveland, J. (1995). Understanding performance appraisal: Social, organizational, and goal-based perspectives. Thousand Oaks, CA: Sage. National Association of State Facilities Administrators. (2010). Integrated project delivery: For public and private owners. Las Vegas, NV: NASFA. Ni, H. B., & Wang, Y. (2015). Information technology in Chinese construction industry in 2015: Depth of using BIM and its development. Beijing, China: China City Press. O’Donoghue, T., & Rabin, M. (1999). Doing it now or later. American Economic Review, 89(1), 103–124. Organization for Economic Cooperation and Development. (1997). Regulatory Impact Analysis: Best Practices in OECD Countries. Retrieved from Paris. Pearl, J. (2009). Causality: Models, reasoning and inference (2nd ed.). New York, NY: Cambridge University Press. Perry, J. G., & Barnes, M. (2000). Target cost contracts: An analysis of the interplay between fee, target, share and price. Engineering Construction and Architectural Management, 7(2), 202–208.

References

189

Pfeffer, J. (1998). Six dangerous myths about pay. Harvard Business Review, 76(3), 109–119. Pocock, J. B., Hyun, C. T., Liu, L. Y., & Kim, M. K. (1996). Relationship between project interaction and performance indicators. Journal of Construction Engineering and Management, 122(2), 165–176. Potts, K., & Ankrah, N. (2014). Construction cost management: Learning from case studies. London, UK: Routledge. Prendergast, C. (1999). The provision of incentives in firms. Journal of Economic Literature, 37(1), 7–63. Prendergast, C., & Topel, R. (1993). Discretion and bias in performance evaluation. European Economic Review, 37(2), 355–365. Prendergast, C., & Topel, R. H. (1996). Favoritism in organizations. Journal of Political Economy, 958–978. Radner, R. (1986). Repeated partnership games with imperfect monitoring and no discounting. The Review of Economic Studies, 53(1), 43–57. Raju, J. S., & Srinivasan, V. (1996). Quota-based compensation plans for multiterritory heterogeneous salesforces. Management Science, 42(10), 1454–1462. Rebitzer, J. B., & Taylor, L. J. (2011). Extrinsic rewards and intrinsic motives: Standard and behavioral approaches to agency and labor markets. Handbook of Labor Economics, 4, 701–772. Roberts, J. (2007). The modern firm: Organizational design for performance and growth. Oxford, UK: Oxford University Press. Rogers, E. M. (1983). Diffusion of innovations: A cross-cultural approach (3rd ed.). New York, NY: Free Press. Ross, M., & Sicoly, F. (1979). Egocentric biases in availability and attribution. Journal of Personality and Social Psychology, 37(3), 322–336. Ryan, R. M., & Deci, E. L. (2000). Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being. American Psychologist, 55(1), 68. Saad, D. A., & Hegazy, T. (2014). Behavioral economic concepts for funding infrastructure rehabilitation. Journal of Management in Engineering, 31(5), 04014089. Samuelson, W., & Zeckhauser, R. (1988). Status quo bias in decision making. Journal of Risk and Uncertainty, 1(1), 7–59. Saxon, R. G. (2013) Growth through BIM. London, UK: Construction Industry Council. Retrieved from http://www.cic.org.uk/publications/

190

Incentivizing Collaborative BIM-Enabled Projects

Scherer, F. M. (1964). The weapons acquisition process: economic incentives. Boston, MA: Division of Research, Graduate School of Business Administration, Harvard University. Schoemaker, P. J. (1982). The expected utility model: Its variants, purposes, evidence and limitations. Journal of Economic Literature, 20(2), 529–563. Schweitzer, M. (1995). Multiple reference points, framing, and the status quo bias in health care financing decisions. Organizational Behavior and Human Decision Processes, 63(1), 69–72. Sebastian, M. T. (2010). Dielectric materials for wireless communication. Amsterdam, The Netherlands: Elsevier. Shelanski, H. A., & Klein, P. G. (1995). Empirical research in transaction cost economics: A review and assessment. Journal of Law, Economics, & Organization, 11(2), 335–361. Shevlin, M., & Miles, J. N. (1998). Effects of sample size, model specification and factor loadings on the GFI in confirmatory factor analysis. Personality and Individual Differences, 25(1), 85–90. Shore, B. (2008). Systematic biases and culture in project failures. Project Management Journal, 39(4), 5–16. Simon, H. A. (1955). A behavioral model of rational choice. The Quarterly Journal of Economics, 69(1), 99–118. Smith, N. J., Merna, T., & Jobling, P. (2006). Managing risk in construction projects. Hoboken, NJ: Wiley. Stake, R. E. (1995). The art of case study research. Thousand Oaks, CA: Sage. Sterman, J. D. (2000). Business dynamics: systems thinking and modeling for a complex world. Boston, MA: McGraw-Hill. Succar, B. (2009). Building information modelling framework: A research and delivery foundation for industry stakeholders. Automation in Construction, 18(3), 357–375. Succar, B., & Kassem, M. (2015). Macro-BIM adoption: Conceptual structures. Automation in Construction, 57, 64–79. Sunil, K., Pathirage, C., & Underwood, J. (2015). The importance of integrating cost management with building information modeling (BIM). In 12th International Postgraduate Research Conference (IPGRC 2015), 10–12 June 2015, Salford Quays, UK. Tang, W., Duffield, C. F., & Young, D. M. (2006). Partnering mechanism in construction: an empirical study on the Chinese construction industry. Journal of Construction Engineering and Management, 132(3), 217–229.

References

191

Thaler, R. (1980). Toward a positive theory of consumer choice. Journal of Economic Behavior & Organization, 1(1), 39–60. Thaler, R., & Sunstein, C. R. (2008). Nudge: Improving decisions about health, wealth, and happiness. New Haven, CT: Yale University Press. Thompson, R. D., & Ozbek, M. E. (2012). Utilization of a co-location office in conjunction with integrated project delivery. Paper presented at the 48th ASC Annual International Conference, Birmingham City University, Birmingham, UK. April 11–14, 2012. Thomsen, C., Darrington, J., Dunne, D., & Lichtig, W. (2009). Managing integrated project delivery. McLean, VA: Construction Management Association of America (CMAA). Turnbull, P. W., & Meenaghan, A. (2001). Diffusion of innovation and opinion leadership. European Journal of Marketing, 14(1), 3–33. Tversky, A., & Kahneman, D. (1974). Judgment under uncertainty: Heuristics and biases. Science, 185(4157), 1124–1131. Tversky, A., & Kahneman, D. (1991). Loss aversion in riskless choice: A reference-dependent model. The Quarterly Journal of Economics, 106(4), 1039–1061. Van Dijk, F., Sonnemans, J., & Van Winden, F. (2001). Incentive systems in a real effort experiment. European Economic Review, 45(2), 187–214. Venkatesh, V., Morris, M. G., Davis, G. B., & Davis, F. D. (2003). User acceptance of information technology: Toward a unified view. MIS Quarterly, 27(3), 425–478. Victoria Department of Treasury and Finance. (2010). The practitioners’ guide to alliance contracting. Victoria, Australia: State of Victoria, Australia. Walker, D. H., & Hampson, K. (2008). Procurement Strategies: A Relationship-based approach. Maldon, MA: Blackwell. Whyte, J. (2012). Building information modelling in 2012: Research challenges, contributions, opportunities. Design Innovation Research Center, Working Paper, Number 5. University of Reading. Williamson, O. E. (1985). The economic institutions of capitalism. New York, NY: The Free Press. Williamson, O. E. (1991). Economic institutions: Spontaneous and intentional governance. Journal of Law, Economics, & Organization, 7, 159–187. Williamson, O. E. (1996). The mechanisms of governance. Oxford, UK: Oxford University Press. Winch, G. (2001). Governing the project process: a conceptual framework. Construction Management & Economics, 19(8), 799–808.

192

Incentivizing Collaborative BIM-Enabled Projects

Winch, G. (2012). Managing construction projects, Hoboken, NJ: Wiley-Blackwell. Wiseman, R. M., & Gomez-Mejia, L. R. (1998). A behavioral agency model of managerial risk taking. Academy of Management Review, 23(1), 133–153. Wong, P. S. P., & Cheung, S. O. (2005). Structural equation model of trust and partnering success. Journal of Management in Engineering, 21(2), 70–80. Wright, S. (1934). The method of path coefficients. The Annals of Mathematical Statistics, 5(3), 161–215. Wu, M. L. (2009). Structural equation model-operation and application of AMOS. Chongqing, China: Chongqing University Press. Wu, Q. (2012). A research on the influencing path of trust on management performance of Integrated Project Delivery team. Unpublished master’s thesis. Tianjin, China: Tianjin University. Yeung, J. F. Y., Chan, A. P.C., & Chan, D. W. M. (2007). The definition of alliancing in construction as a Wittgenstein family-resemblance concept. International Journal of Project Management, 5, 219–231. Yin, R. K. (2013). Case study research: Design and methods. Thousand Oaks, CA: Sage. Zaheer, A., McEvily, B., & Perrone, V. (1998). Does trust matter? Exploring the effects of interorganizational and interpersonal trust on performance. Organization Science, 9(2), 141–159. Zhou, X., & Swan, P. L. (2003). Performance thresholds in managerial incentive contracts. Journal of Business, 76(4), 665–696. Zuppa, D., Issa, R. R. A., & Suermann, P. C. (2009). BIM’s impact on the success measures of construction projects. Computing in Civil Engineering. Reston, VA: American Society of Civil Engineers.

About the Author Dr. Chen-Yu Chang is an infrastructure economist with a tenured position at the Bartlett School of Construction and Project Management, University College London. He is the director of the Bartlett Infrastructure Center as well as the founding director of the UK-China Infrastructure Academy. The latter is cosponsored by the chancellor of the exchequer of the United Kingdom and the chairman of the National Development and Reform Commission of the People’s Republic of China. He holds a bachelor’s degree in civil engineering, a master’s degree in construction management from National Taiwan University, and a doctoral degree in construction economics from University College London. He has acted as primary supervisor to six doctoral candidate students, and second supervisor to three doctoral students. Over the past 15 years, his research has covered a broad range of issues associated with project organizations in general, and PPP/PFI in particular. Twenty-­six academic articles have been published under his name in leading international project/construction/engineering management journals. He is generally credited with the building of several PPP the­ ories in relation to holdup problems (Chang, 2013a; Chang & Ive, 2007; Chang & Qian, 2015), risk evaluation (Chang & Ko, 2017), risk allocation (Chang, 2013b, 2014d), project governance (Chang, 2013c, 2015), choice between government-pay and user-pay PPP systems (Chang & Chou, 2014), infrastructure financing (Park & Chang, 2013), and Chinese PPP (Chang & Chen, 2016). He has advised Taipei city government and New Taipei city government on how to implement government-pay PPP systems in the cities. In partnerships with the China International Engineering Consulting Corporation, he successfully held a high-profile PPP Forum in Beijing in March 2016 under the sponsorship of Foreign and Commonwealth Office’s China Prosperity Fund.

193

Index A Actual system use, 115, 116. B Behavioral biases: 9, 22, 23, 29; optimism bias, 23; present biases; Prospect Theory, 23; reference point, 23. Behavioral intention to use, 115, 116. Building information modeling (BIM): 18, 22, 23, 24, 31, 32, 33; clash detection, 33; diffusion, 18, 21, 38; external support; handbook, 8; implementation, 10, 20, 46, 49, 53, 82; Level, 56, 79; life cycle theory of, 23; mandates, 21; maturity, 22; utilization, 67, 88. C Case studies, 30, 32. Collaborative, 6, 8, 13, 4, 83, 87, 89. Common goal, 14, 100. Communication, 11, 19, 20, 21. Contract driven, 39, 41. Contract form, 28. Coordination, 11, 24, 25. Criteria of good fit, 65, 71, 117. Critical success actors (CSF), 13, 19. D Data collection, 60, 83. Delivery environment: 4, 44, 114; integrated, 6, 8, 23, 67, 70. Delivery system, 5, 9, 23, 42, 73, 119.

E Early involvement, 90, 100, 108. Economizing: first-order, 4; second-order, 4. Efficiency driven, 39. External support, 6, 115, 116. F Financial incentivization: 46, 48; Financial incentives, 47, 93. G Gain/pain share, 18, 26, 28. Group-based (team-based) reward, 16, 17. H Heuristics, 9. I Incentives: 12, 16, 23, 25, 27; alignment, 25; pool, 8, 26; indicator, 17, incentivization: 14, 19, 28; definition of, 8; effectiveness, 23; monetary (explicit), 15, 59; measures of, 12, 19; system, 11. Integrated project delivery (IPD): 8, 9, 13, 14, 19, 21, 26, 27, 36, 37; acceptability, 5, 19; parties, 8, 30; Interview, 33, 83, 111. J Joint decision making, 96, 100, 108. 195

196

Incentivizing Collaborative BIM-Enabled Projects

L Linear contact, 25, 28. M Motives: 25; motivation: intentional, spontaneous, intrinsic (extrinsic), 15. N National environment, 10. P Perceived usefulness, 74, 114, 116. Principal-agent theory (agency theory), 9, 14, 15, 16, 17. Project complexity, 45, 96, 100, 108. Project governance, 11. Project performance, 14, 27; objective (subjective), 17, 18, 25; metrics, 18. R Regulation, 7, 64. Reliability, 60, 72, 115.

Risk-averse, 14. Risk sharing: 9, 27; gain/pain share, 12, 26.

S S-curve: 4, 5, 67, 69. Structural equation modeling (SEM), 12, 21, 30, 31; confirmatory factor analysis, 31; path analysis, 31. T Target cost, 12, 25; setting, 12, 26, 28. Technology acceptance model (TAM), 9, 29, 30. Transaction cost economics (TCE), 11, 23, 30. Transaction cost: 23, 30, 43, 44. Trust, 12, 19. U User resistance, 9. V Validity, 60, 72, 117.

INCENTIVIZING COLLABORATIVE BIM-ENABLED PROJECTS

T

he use of digital representations to aid in projects—Building Information Modeling (BIM)—is gaining traction worldwide as an effective and beneficial approach to executing projects that can reduce errors and improve project results. These crucial benefits are prompting many businesses and governments to mandate its use, while others search for ways to grow and incentivize the practice. In Incentivizing Collaborative BIM-Enabled Projects: A Synthesis of Agency and Behavioral Approaches, Chen-Yu Chang, PhD (LON), explains the current state of BIM use in three countries: China, the United Kingdom, and the United States. Following multiple case studies in each country, Professor Chang explores the explicit and implicit motivators that may drive BIM participation and the factors that can influence its effectiveness. The case studies offer multiple perspectives on why and how BIM-enabled projects are adopted and provide a lens for understanding BIM at varying levels of maturity and in different environments. Understanding the motivators behind the current state of BIM-enabled projects facilitates the design of incentive structures for future BIM projects. As BIM continues to evolve and transform the industry, this research provides valuable insight into the best practices for incentivizing its use. This theoretical approach gives researchers and organizations new tools and ideas to help build their own strategies to encourage BIM use and better understand its place in managing projects.

PMI.org

ISBN: 978-1-62825-623-9

U.S. $34.95