Implicit Contracts , the Great Depression , and Institutional Change :

This paper employs a game-theoretic framework and comparative historical approach to study an institutional change triggered by an extraordinary shock — the Great Depression. By characterizing “corporate welfarism” practiced by large manufacturing firms in the U.S. and Japan in the 1920s as implicit employment contracts based on internal enforcement, the paper reexamines the impact of the Depression on corporate welfarism and the subsequent evolution of employment relations in the two countries. Documenting parallel institutional developments in the U.S. and Japan during the two decades prior to the Depression, the paper identifies the early 1930s as a point of bifurcation at which two paths began to diverge. In the U.S., the repudiation of the implicit contracts by a majority of employers induced by the deep and prolonged depression caused a change in the expectations of workers and the public, leading to the advancement of explicit employment contracts based on third-party arbitration. By contrast, during a less severe depression, a majority of Japanese large companies kept their implicit contracts, while developing institutional arrangements to mitigate a negative impact of business fluctuation. The paper then attributes the irreversibility of the institutional change after the Depression to the endogenous formation of complementary institutions, such as union laws, enforcement agencies, and state welfare programs. ∗I am grateful to Lee Alston, Masahiko Aoki, Pierre Azoulay, Carliss Baldwin, Adam Brandenburger, Robert Gibbons, Claudia Goldin, Avner Greif, Robert Margo, Laura Owen, Warren Whatley, and Gavin Wright for their valuable comments and encouragement. I would also like to thank seminar participants at MIT Sloan School, University of Illinois, Northwestern University, University of Michigan, University of Pennsylvania, UCLA, and Stanford University for their helpful discussions. 6 Theoretical Appendix In this appendix, I develop a simple game-theoretic model that captures the essence of corporate welfarism. In the first section, I model spot contracting and corporate welfarism as two equilibria of the repeated game between the firm and workers with non-contractable human capital, and examine the existence and efficiency of these equilibria. In the second section, I introduce an exogenous shock that changes a parameter of the repeated game, and explore its impact on equilibrium selection. In the third section, I endogenize the contractability of human capital to introduce an option to write an explicit contract. It is shown that once an explicit contract is developed, it may eliminate an implicit contract equilibrium. The results indicate a role of history in the equilibrium selection and the irreversibility of the institutional change due to the endogenous investment in complementary assets. 6.1 Repeated Employment Game 6.1.1 Basic Stage Game Assume that there is an infinitely-lived firm who owns productive assets and employs workers. Using an overlapping generations framework, each worker is assumed to work for two periods, in which he is referred to as “young” and “old.” Workers contribute two types of labor inputs; contractable and non-contractable. In every period, a worker chooses the level of contractable effort e and the level of non-contractable effort x. Contractable labor inputs, such as work hours and the number of pieces assembled, are verifiable to a third party and legally enforceable. By contrast, non-contractable labor inputs, such as efforts to acquire better skills or other intangible human capital (e.g., loyalty, diligence, dependability), are unverifiable. Based on historical observations, I assume that non-contractable effort increases worker’s productivity one period later, whereas contractable effort improves contemporaneous productivity. I refer to non-contractable effort x also as “skill investment” and assume that higher x leads to the formation of better skill s in the subsequent period. Assume that the firm cannot directly observe skill investment x but observes the realized skill level s one period later. Assume also that both skill and output levels are unverifiable and non-contractable. In the following analysis, I say an employment contract is explicit if the contract is contingent solely on contractable variables, and is implicit if it is contingent on non-contractable variables. Consider the two-period employment game (i.e., stage game) played between the firm and an individual worker. The timing of the game is as follows. At the beginning of the first period, the firm offers an employment contract to a young worker. The young worker accepts or rejects the contract. If he rejects, he has an outside option that yields u per period in terms of utility. If he accepts, then he enters into employment relations with the firm and chooses the effort level (e1, x1). At the end of the first period, after observing the level of e1, the firm pays monetary compensation m1 to the worker. At the beginning of the second period, the firm observes the skill level s2 of the now old worker. The firm then again offers an employment contract. The worker either rejects or accepts the offer. If he rejects, he quits the firm The following model is based largely on Kanemoto & MacLeod (1989). See also Bull (1987); MacLeod & Malcomson (1989); Baker, Gibbons & Murphy (1994) for modelling implicit contracts. By this definition, even if a contract is “explicitly” spelled out, it is an implicit contract when it is contingent on unverifiable variables.