«ABSTRACT Title of Dissertation: MULTINATIONAL CORPORATIONS AND DEVELOPING COUNTRIES: ENTRY MODE, TECHNOLOGY TRANSFER AND PERFORMANCE REQUIREMENTS ...»
Title of Dissertation: MULTINATIONAL CORPORATIONS AND DEVELOPING
COUNTRIES: ENTRY MODE, TECHNOLOGY
TRANSFER AND PERFORMANCE REQUIREMENTS
Gary Wayne Anderson, Jr., Doctor of Philosophy, 2002
Dissertation directed by: Professor Roger Betancourt
Department of Economics
This dissertation develops and tests a theoretical model of multinational corporation technology transfer to affiliates in developing countries. A bi-lateral moral hazard model is used to analyze a multinational corporation’s decision to enter a new market via a wholly owned subsidiary, a joint equity venture or an arm’s length contract. The model demonstrates that governments can enact Pareto improving policies that increase technology transfer from foreign investors. However, frequently employed and recommended interventions such as export promotion policies and local content regulation lower joint venture profits and decrease the incentive to transfer technology.
Empirical testing is conducted using a simultaneous equation limited dependent variable model. The results indicate that the determination of foreign ownership shares at the plant-level is consistent with the bi-lateral moral hazard model. Further, there is little if any indication that ownership is shared in international joint ventures as a means of sharing risk.
MULTINATIONAL CORPORATIONS AND DEVELOPING COUNTRIES: ENTRY
MODE, TECHNOLOGY TRANSFER AND PERFORMANCE REQUIREMENTSby Gary Wayne Anderson, Jr.
Dissertation submitted to the Faculty of the Graduate School of the University of Maryland, College Park in partial fulfillment of the requirements for the degree of Doctor of Philosophy 2002
Professor Roger Betancourt, Chair/Advisor Professor Ingmar Prucha Professor Arvind Panagariya Associate Professor Daniel Vincent Associate Professor Bartlomiej Kaminski ©Copyright by Gary Wayne Anderson, Jr.
ACKNOWLEDGEMENTSOne of the benefits of working on a dissertation near the Library of Congress is that the library has copies of all dissertations on microform. Whenever I needed to look up some derivation which was contained only in a dissertation and not subsequent published work, I began by reading the authors’ Acknowledgements sections. I found this personal glimpse into another student’s struggle to complete his/her dissertation to be both reassuring and inspirational. I feel that I owe it to not only all of those people who have helped me complete this dissertation, but also to any future graduate students like myself who enjoy reading more than just scholarly prose to write this particular section.
While working on my dissertation, I have been fortunate to be surrounded by an extremely patient group of people. Advisors, family members and friends have all contributed to this dissertation in their own ways. Roger Betancourt has not only given superb advice and encouragement but he has also given me the freedom to make this dissertation a much more personal accomplishment. All of his efforts have made me a much better economist. Ingmar Prucha’s insistence on rigor has both made this a better dissertation and taught me valuable lessons. Anand Swamy and Brian Fikkert were instrumental in the early stages of my research and continued to provide valuable advice. As I go through life, perhaps the most important advice they gave was to go to Indonesia to conduct research. I believe the advice was: “If you want to be a development economist, then you really ought to have some actual experience in a developing country”. While living in Indonesia, Anggito Abimanyu provided me financial support, gave me access to a wealth of data, and offered innumerable insights.
Daniel Vincent and Bart Kaminski.
In addition to the technical guidance of my committee and other economic advisors, the support of my family and friends played a crucial role throughout the dissertation process. First and foremost, my wife Jessica has been most tolerant and understanding of all of my moods and frustrations. None of this could have been done without her support and understanding. I am fairly certain that my parents understand relatively little of the technical aspects of the work I have done over the past several years. They may even wonder why I have chosen to pursue a career which has often taken me far away from the family that I value so much. Of this they should be certain, everything that I have accomplished – and everything my sister Sara has accomplished
- is due to the love support and encouragement that they have always given us. My
father-in-law sent each of his children off to college with a rather simple directive:
“Remember what you’re there for.” Neither he nor my mother-in-law has ever been shy about offering me encouragement. They have been generous not only with their advice and support but with their refrigerator as well. Everything was greatly appreciated. To the rest of my family and friends who I have neglected to mention explicitly, please know that I realize that finishing a dissertation is not something accomplished alone.
List of Tables
List of Figures
Chapter 5: Policy Implications
Appendix A: Non-Restrictive Nature of Linear Contracts
Appendix B: Existence and Uniqueness of the Nash Equilibrium Solution........... 68 with a Linear Contract Appendix C: Detailed Solution
Appendix E: Data Description
Appendix F: Estimation Methodology
Appendix G: Two Limit Tobit Model
Appendix H: Empirical Model
Table 3.1 Policy Comparison
Table 4.1 Output shares by Ownership Group
Table 4.2 Frequency Distribution of Foreign Start-Ups and Total Plants.
....... 48 by Equity Class Policy Comparison Table 4.3 Variable Definitions
Table 4.4 Industry Level Analysis
Table 4.5 Variable Definitions and Summary Statistics
Table 4.6 Standard 2-Limit Tobit
Table 4.7 Two-Stage 2-Limit Tobit
Table 4.8 Endogeneity Tests
Figure 2.1 Contract Outcomes and Skill Parameters
Figure 2.2 Dependence of Critical Skill Levels on Production Function.
.......... 21 Parameters Figure 2.3 Dependence of Critical Skill Levels on Technology Production...... 22 Function Parameters
1.1 INTRODUCTION Foreign direct investment (FDI) flows to developing countries have risen to an unprecedented level in the 1990s. Aitken and Harrison (1999) note that in 1997 FDI represented about 40% of all public and private capital flows to developing countries.
Alongside this increased level of multinational corporation (MNC) engagement in the developing world, another trend must be noted. The nature of such involvement has also been transformed. Most significantly, MNCs have increased their usage of alternatives to wholly owned subsidiaries such as joint equity ventures (JVs) and non-equity interfirm contractual arrangements.1 These arm’s length (non-equity) arrangements will be referred to as technology purchases (TPs).2 Additionally, MNCs have sought important contributions from local counterparts that have had a significant impact on the ultimate success of an investment project.3 Finally, it has become evident that policy makers in less developed countries (LDCs) seek out MNCs in order to gain access to their intangible assets such as advanced technology or marketing capability rather than the MNCs’ stock of physical capital. The model of multinationals developed in this dissertation brings each of these observed tendencies to the forefront. In so doing, this 1 Oman (1989) argues these alternative forms of investment are the single largest form of MNC involvement in the developing world since the 1970s. Caves (1996), Rugman (1986) and Beamish and Banks (1987) all note that few of the alternative forms of contracting and equity arrangements between MNCs and developing country firms have been viewed through the metric of formal economic modeling.
2 This dissertation follows existing literature and employs a very broad concept of technology purchase.
A TP may include royalty payments, management fees, patent and trademark infringement rights, copyright fees or any number of other licensing arrangements. Generally, these terms refer an arm’s length (i.e., no equity involvement) exchange. However, the contracting mechanism employed in this paper also explains why a MNC may use these mechanisms in conjunction with equity.
3 Beamish and Banks (1987), Miller et al. (1997) and Gomes-Casseres (1989) present evidence of both the type and importance of local contributions.
previously escaped examination by formal economic theory.
1.2 LITERATURE REVIEW AND CRITIQUEThere are a number of weaknesses with current theoretical explanations of FDI.
Chiefly, as noted in the introduction, MNCs have many available entry-mode options and current theory offers little insight into this choice. Models of FDI which rely heavily upon “New Trade Theory” do not address the causes or impact of shared ownership and assume away the possibility of contractual arrangements. Krugman (1983) who rules out licensing by assumption believing it to be “relatively unimportant in practice” is fairly typical. Even those papers which model MNCs’ choice between direct ownership and contracting (e.g., Horstmann and Markusen (1987, 1996)) offer no insights into a MNC’s choice between shared and sole ownership.
A second weakness is that prior models have mis-characterized the knowledge transfer process. Most models of FDI assume that MNCs possess knowledge-based assets that can be costlessly or inexpensively transferred to an additional location. Prior research calls into question this non-rivalry assumption. Teece (1976) finds that it “is quite inappropriate to regard existing technology as something that can be made available at zero social cost.” Indeed, the author finds that transfers within the same company are quite costly. Contractor (1985) echoes this point.
Several explanations have been offered for the costly nature of such transfers.
For example, knowledge may be embodied in the individuals who employ the expertise.
Contractor (1985, p. 19) argues that “technology transfer is not the transfer of only
of a production or distribution capability that may also require interpersonal contact of a long duration.” Thus, technology transferred by MNCs may closely resemble North’s (1990) concept of tacit knowledge. However, even though some research has analyzed the impact of costly transfer (e.g., Wang and Blomstrom (1992)), there has been little analysis as to the reason for the costliness. In particular, there has been little or no formal analysis of the market failures that make such transfers costly.4 This observation that knowledge transfer may be costly is particular damaging to models in which MNCs choose equity over arm’s length transfer because they fear disclosure of proprietary information to third parties. Finally, given Mansfield and Romeo’s (1980) finding that as a rule technological secrets are not revealed more rapidly under licensing than they would be with trade or other entry modes, only one reliable conclusion remains. The fear that proprietary knowledge may become public information may indeed affect the decision to engage in a particular economy or the type of activity conducted (see Lee (1996)), but it will not have a strong impact on the decision among trade, licensing and FDI.
There are also questions regarding the strategic and dynamic implications of current explanations of entry-mode. Contrary to the prediction of Horstmann and Markusen (1987), licensing arrangements may act as a stepping stone to future expansion via FDI. Contractor cites several examples and Giannitsis (1991) concludes that “licensing was often the preferred instrument for entering the Greek market before deciding on other forms of investment (subsidiary, share participation, acquisition of 4 There has been a great deal of discussion of these market failures (see Caves (1996) for a summary) but
Markusen’s model. The authors note that “were the home-country only to license for a finite number of periods, the licensee would always have the incentive to dissipate [the value of the license] before [it] was revoked.” In a later paper, Horstmann and Markusen (1996) acknowledge this weakness and simply assume that MNCs can write short-term contracts. Additionally, Fikkert (1994) and Giannitsis (1991) find that equity investment and other forms of technology purchase are complementary activities. In all current models, equity investment is a substitute for licensing which arises when transaction costs to third parties are prohibitively high.
Finally, many explanations of FDI only explain import substituting investment and are frequently tied to retailing and distribution networks. While some foreign investors seek to capture market share in the host country, others target export markets or involve both production and distribution activities. While no claim is made that the model developed in this dissertation is a general model of all multinational activities, the problems of moral hazard and hidden action are relevant to both import substituting and export-oriented activities as well as production and distribution.
1.3 SUMMARY OF MODEL AND RESULTS The model developed here relies upon bi-lateral moral hazard. The essence of this approach is that collaboration between MNCs and local partners is best modeled as an interaction between two input providers, each subject to an incentive constraint. The market failures arise because the actual amount of these inputs provided by a particular relatively little formal modeling. Horstmann and Markusen (1996, 1987) are two exceptions.