Research Unit for Statistical and Empirical Analysis in Social Scienses

>>Japanese

Hi-Stat Vox No. 17 (December 9, 2010)

Introduction

As stated by North (1991), institutions provide the incentive structure of an economy. And as such, institutions are an important determinant of economic performance. To establish “good” institutions, however, is remarkably difficult, and to reform “bad” institutions is even harder. How do institutions evolve, and when do inefficient institutions persist? Although some progress has been made, we are still far from understanding the mechanisms of institutional change, theoretically or empirically.

In studying the determinants of institutional developments and the reasons for institutional diversity, economic historians have distinct advantages over economists of the modern period. History is a fertile ground for studying institutions and institutional change, because we observe a wide variety of economic institutions in diverse political settings over a long period of time, marked occasionally with historical accidents and natural experiments. In this essay, in order to demonstrate the richness of history and the power of comparative analysis, I would like to contemplate on the historical origins of the employment practices in the U.S. and Japan.

Postwar HRM Practices in the U.S. and Japan

When Japanese manufacturers, be it steel, automobiles or consumer electronics, rapidly gained market shares in the U.S. in the 1970s and 1980s, there was intense scrutiny for understanding the sources of their high labor productivity. What people discovered was an entirely different set of human resource management (HRM) practices pertaining to blue-collar workers employed by Japanese firms.

To put it simply at the risk of oversimplification, the employment relations in leading U.S. companies (such as U.S. Steel, General Motors, and General Electric) were characterized by (1) explicit and detailed employment contracts enforced by a legal third party, (2) adversarial relations between management and industry-wide unions, (3) low investment in employees' education and training, (4) narrow job definitions and finely-graded wages based on job titles, and (5) limited but contractual job security based on elaborate seniority rules.

By contrast, the employment relations in Japanese firms (such as Nippon Steel, Toyota, and Hitachi) were characterized by (1) implicit and discretionary employment contracts enforced internally through reputation, (2) cooperative labor-management relations based on an enterprise union and joint consultation, (3) high investment in employees' firm-specific human capital, (4) flexible job assignments and wages based on seniority plus performance evaluations, and (5) implicit (and thus noncontractual) guarantee of “life-time” employment.

Why did we observe such a striking contrast in the HRM policies pertaining to blue-collar workers between the two leading economies? Were these American-style and Japanese-style HRM practices both economically rational? Since when did they differ? Motivated by these questions, I have explored the historical origins of the employment relations in the two countries (Moriguchi 2000, 2003, 2005). In doing so, I first proposed a game-theoretic framework as a basis for conducting comparative historical analysis.

Theoretical Framework for Understanding Institutional Change and Persistence

Recent applications of game theory have greatly advanced our understanding of economic organizations and institutions. One of the major insights arising from comparative institutional analysis, proposed by Aoki (2001), Greif (2005), and others, is that institutional diversity can be modeled as multiple equilibria where distinctive institutions correspond to different equilibria of an appropriately defined game. Following this literature, I conceptualize an “employment system” as an equilibrium outcome of a game played by three players, management, labor, and government, who pursue their own objectives as they strategically interact with one another. In other words, an employment system is defined as a set of mutual best responses and the beliefs that support those best responses.

At the level of a firm, repeated interactions between management and workers shape employment relations, taking technology, labor markets, and state regulations as given. Using a model of repeated employment game with noncontractable human capital, I show that the Japanese-style HRM policies described above can be seen as a set of complementary practices that constitute an implicit contract equilibrium (ICE), in which management offer employment security to encourage human capital investment. I also show that, in the same employment game, the American-style HRM policies can be modeled as a set of complementary practices constituting an explicit contract equilibrium (ECE), in which management introduces contractable proxy (e.g., job titles, seniority) for noncontractable human capital (Moriguchi 2003).

At the level of an economy, the government chooses a legal framework and social welfare programs to maximize its objectives. Government policies thus affect strategy sets and payoffs of private agents (i.e., management and labor). At the same time, firm-level employment relations impact government payoffs by changing the level of industrial production, income distributions among constituencies, and the demand for social welfare. The strategic interactions between private agents and the government give rise to equilibrium employment systems that consist of a set of complementary institutions, where the presence of strategic interdependencies typically leads to a multiplicity of such equilibrium systems.

If employment systems in the U.S. and Japan that emerged by the 1970s can be viewed as two distinct equilibria, understanding the evolution of employment systems is equivalent to understanding the dynamic process of equilibrium selection. In conducting a historical analysis, I introduce two empirically important factors relevant to the process of equilibrium selection (Moriguchi 2000). The first factor is the existence of “unanticipated shocks,” such as wars, recessions, and foreign invasions, that exogenously change the parameters of the game or the nature of strategic interactions. The second factor is what I name “institutional capital” (e.g., physical capital, human capital, reputations, organizational knowledge, case laws, administrative expertise) that the players accumulate over time as they play particular equilibrium strategies.

Consider initial conditions in an economy, characterized by parameters of the game and institutional capital determined by prior history. When an unforeseen event occurs, it changes the parameters and induces endogenous strategic responses by management, labor, and government. Importantly, the players derive their best responses contingent on existing institutional capital that was formed without anticipating the event. Institutional capital accumulated in the past thus shapes and constrains today's decisions, creating path dependence in institutional evolution. As a particular equilibrium is realized in the economy, there is subsequent accumulation of institutional capital associated with this equilibrium, which, in turn, forms new initial conditions for the next unforeseen event, upon which further institutional development will be built.

What is crucial in determining the direction of institutional development is the relative size of the two factors: an unanticipated shock and institutional capital. There is institutional continuity or persistence due to the cumulative nature of institutional capital. As the players sink investments and accumulate institutional capital that reinforces and stabilizes the existing institutional structure, an employment system tends to converge to a given equilibrium. This is what I call the self-reinforcing process of institutional development. A possibility of institutional change, however, arises from an unanticipated shock that triggers a shift in the strategic responses of the players. When the magnitude of the shock is sufficiently large relative to the depth of institutional capital, the institutional trajectory shifts its course, diverging towards a new equilibrium. This is what I call the process of bifurcation. A temporary shock may have a lasting impact on the subsequent institutional development by initiating an endogenous process that amplifies the initial effect of the shock through the accumulation of institutional capital.

Comparative Historical Analysis of the U.S. and Japan

Based on the above theoretical framework, I trace institutional trajectories in the U.S. and Japan from the beginning of the 20th century to the post-WWII period. As I summarize below, the historical analysis reveals surprising turns of institutional developments in the two countries: namely, similar initial conditions around 1900, parallel developments towards the ICE in the 1910s and 1920s, the process of bifurcation after the early 1930s in which the U.S. trajectory diverged towards the ECE, and the process of institutionalization towards the respective equilibria during and after WWII.

At the turn of the 20th century, the employment relations in large American and Japanese factories were commonly characterized by (1) simple and short-term employment contracts, (2) the lack of systematic HRM policies, (3) the absence of organized labor, and (4) wages determined by spot labor markets. In both countries, production workers regularly changed their employers seeking higher wages or better working conditions. Employers, in response, invested little in them and resorted to dismissal whenever business conditions deteriorated. Personnel management was delegated to foremen who relied on direct monitoring and coercion to induce work effort. Consequently, high job turnover, low work commitment, and little trust between management and workers were the workplace norms both in the U.S. and Japan.

From the similar starting points, employment practices in leading firms in the two countries developed in parallel during the 1910s and 1920s. In particular, WWI created similar economic, political, and social conditions in the U.S. and Japan. In response to rapid industrial expansion, labor shortage, surging labor movements, and the increasing demand for social legislation, business leaders in both countries began to seek more stable workforce, industrial peace, and higher labor productivity. Progressive employers voluntarily introduced a variety of corporate welfare programs for blue-collar workers, including nonwage benefits (e.g., paid vacation, health insurance, corporate pensions, employee housing), corporate training and education, internal promotion and greater employment security, and employee representation plans (also called “company unions” in the U.S. and “factory committees” in Japan). This was the movement known as corporate welfarism, also called “welfare capitalism” in the U.S. and “employer paternalism (keiei onjo shugi)” in Japan. It was essentially an implicit long-term contract that intended to reward workers' noncontractable human capital such as loyalty, commitment, and firm-specific skills and experience. In both countries, leading proponents of corporate welfarism were large firms in capital-intensive industries, including International Harvester, General Electric, Standard Oil of New Jersey, General Motors, and Bethlehem Steel, in the U.S., and Mitsubishi Shipyards, Hitatchi Engineering Work, Shibaura Engineering Work, Sumitomo Electric Wire, and Nippon Steel, in Japan.

During the 1920s, the proponents of corporate welfarism in the U.S. and Japan gradually won trust from their employees and the public. At the end of the 1920s, these firms employed about 20% of production workers in the respective economies: although small, it was a sizable and growing minority. When the Great Depression unexpectedly hit both economies in 1929, major employers in both countries instituted work sharing and made considerable effort not to layoff their core workers. As the Depression deepened in the U.S, however, a majority of the proponents of corporate welfarism reneged on their promises and resorted to mass dismissals, resulting in a collapse of managerial reputations and a loss of firm-specific human capital. By contrast, the much shorter and less severe Depression in Japan allowed most employers to keep their promises without resorting to unilateral dismissals.

This initial bifurcation triggered the endogenous formation of labor laws in the two countries during the 1930s. In response to the economic and social conditions set off by the Great Depression, the governments in the U.S. and Japan attempted to facilitate economic recovery through legalization of business cartels and trade unions. In the U.S. newly formed industrial unions sharply increased their membership and political voice under the National Industry Recovery Act. By contrast, union legislation was repeatedly blocked in Japan by business leaders who maintained better reputations compared to their American counterparts. While the Japanese government gradually developed labor laws that were complementary to corporate welfare provisions, the U.S. government eventually established a new legal framework during the New Deal that seriously undermined the viability of corporate welfarism. Most importantly, company unions were outlawed in the U.S. and replaced by craft or industrial unions. Subsequently, American employers and unions negotiated collective agreements in the atmosphere of mutual distrust that called for third-party enforcement.

To further evaluate the impact of the Great Depression, I also examine the variation within the U.S. using firm-level data. The data indicate that the firms that were hit harder by the Depression abandoned corporate welfarism earlier and to a greater extent, and that the greater repudiations, in turn, were associated with the higher employee support for outside unions and more explicit employment contracts concluded under the New Deal (Moriguchi 2005).

The Second World War induced the U.S. and Japanese governments to enact stringent labor regulations to maximize wartime production. The respective governments designed wartime measures building largely on the employment practices and legal frameworks that had developed prior to the war. In addition, the strategic reactions and selective compliance of private agents made regulations that were compatible with prevailing employment relations more effective. Consequently, wartime government regulations contributed greatly to the institutionalization and diffusion of existing HRM practices in the respective economies (Moriguchi 2000). The emerging difference between the employment systems in the U.S. and Japan in the late 1930s was reinforced during WWII, leading to further divergence of the two institutional trajectories.

Consequently, by the mid-1940s, two distinctive sets of HRM practices were taking shape in major manufacturing firms in the U.S. and Japan. Employment relations in U.S. establishments were governed by explicit and detailed contracts between employers and industrial unions, whereas employment relations in Japanese establishments were based on implicit and long-term contracts between management and factory committees.

While the U.S. more or less continued down the same institutional path after WWII, the postwar U.S. occupation of Japan (1945-53) was a potential bifurcation point for Japan's institutional development. While Japan went through a sharp recession and a drastic labor law reform a la New Deal during the occupation, Japanese workers and employers essentially reestablished employment relations based on implicit contracts that relied on internal enforcement mechanisms. Most notably, industrial unionism failed to gain majority support from workers in Japan, and in many large manufacturing firms, employees eventually organized themselves into enterprise unions. The continuity of the Japanese trajectory reflected the depth of institutional capital, such as mutual trust, firm-specific skills, and workplace norms, which had been accumulated throughout the prewar and wartime periods.

Concluding Remarks

The comparative empirical analysis above highlights the historical contingency and path-dependence of institutional developments that eventually gave rise to distinctive employment systems in the U.S. and Japan after WWII. The timing and magnitude of unanticipated shocks played an important role in the process of equilibrium selection in the repeated employment game. Once selected, a set of complementary laws and institutions (such as collective bargain laws and state welfare provisions) began to form around the equilibrium, reinforcing the path of institutional development.

Since the 1980s, the employment practices in the U.S. and Japan have continued to evolve. On the one hand, inspired by the Japanese-style practices, many manufacturing establishments in the U.S. adopted “innovative” HRM practices with a varying degree of success, while searching for yet another way to improve labor productivity. On the other hand, under the strain of the Great Recession and the lost decade of the 1990s, Japanese firms have not only adjusted and redefined their HRM practices but also been experimenting with new policies, often looking into American practices as a possible role model. Given the path-dependence in institutional developments, understanding the history will shed better light on the present and future courses of employment practices in the U.S. and Japan

References