Vol.33-No.11 November 1,1994
I. Introduction
Many OECD countries have experienced declines in union density since the late 1970s and early 1980s. The country that experienced the most dramatic union decline following France and the U.S. is Japan (Visser, 1991).
In Japan, postwar labor reform stimulated a burst of union organizing activity so that 45.3% of employees were union members by 1947. Following this initial wave, the level of union density declined to a little more than a third of employees by 1955. Subsequently, union density remained remarkably stable for the twenty-year period of high economic growth between 1955 and 1974. Since 1975, however, union density has declined sharply from 34.4% in 1975 to 24.2% in 1993.
While declining union density might be irrelevant to economies in some countries, this is not the case with the Japanese economy. Enterprise unionism, an idiosyncratic structure of organization in Japan, has been considered a key institutional pillar of the Japanese labor market, along with long-term employment, intrafirm skill formation, and the seniority wage system. Given this structure, the decline in union density signifies the erosion of an important labor market institution in Japan.
To investigate the reasons for the decline in union density, I have organized a research project for the Japan Institute of Labour (JIL). The full results are reported in the Japan Institute of Labour (1993) and Rebitzer and Tsuru (1994). This essay summarizes the main results based upon a sample survey of individual workers and their attitude towards unions and potential unionization.(1) The paper proceeds as follows. The second section reviews major factors which have been identified in the foregoing literature. The third part focuses on the process of union organizing. Special attention is paid to the incentives for employers to resist unions and the incentives for employees to join unions. The paper closes with a brief summary of the implications of the analyses.
II. Employment Shifts or Failure of Union Organizing?
1. Determinants of the Decline in Union Density
To better understand the nature of "de-unionization" in Japan, it is important to describe the process of declining union density. Two factors have played a critical role in the process: (1) the changing composition of employment in heavily and lightly unionized groups (employment shifts), and (2) the slowdown of organizing activity by unions in the union organizing process.
In Japan, unionization rates differ widely by gender, firm size, and industry. From this perspective, it is possible that sectoral shifts of employment among these groups is the primary cause of the declining unionization rate. However, previous statistical studies have revealed that the employment shifts account for only a modest proportion of the change in union density. For example, the shifting composition of the labor force between the secondary and tertiary industries explains only 20 to 30% of the drop of union density since 1975 (Freeman and Rebick, 1989; Ito and Takeda, 1990).
2. The Decline in New Union Organizing
An alternative explanation for declining union density is changes in the rate of new organization within sectors. The rate of new organization, which is defined as the ratio of newly organized workers divided by the number of employees, declined sharply from 0.45% in 1975 to 0.14% in 1993. Freeman and Rebick (1989) demonstrate that nearly 70% of the drop in union density can be accounted for by the falling rate of new organization.
However, the difficulty of organizing new members might be attributed to the fact that new, non-union firms are found in inherently non-union sectors in the Japanese economy. This structural explanation suggests that the decline in the rate of new organization is the natural result of sectoral shifts rather than changes in the incentives influencing the actions of employers and non-union workers. To evaluate this possibility, I estimate the determinants of the probability of union membership. After controlling for individual characteristics (such as gender, educational attainment, firm size, etc.) of workers, the probability of being union workers in new firms is substantially lower than the probability in old firms. This indicates that much of the decline in the rate of new organization is not the result of employment shifts. In the next section, I look more closely at the determinants of organizing success.
III. Employer Resistance or Employee Disinterest?
1. Employer Incentives to Resist Unions
Employers have an incentive to resist unions to the extent that unions increase the cost of labor. In the United States, for example, high and increasing union wage differentials have been mentioned as the primary cause of increased employer resistance to organizing efforts in the 1970s and 1980s (Blanchflower and Freeman, 1992).
Investigations of the importance of employer resistance in the decline of unions in Japan has been hampered by the absence of precise estimates of the union-nonunion wage differential. Using the data from the JIL survey, it is now possible to estimate the union wage premium at the level of individual workers. Table 1 summarizes the results. For men, the effect of unions on monthly salary and annual earnings are virtually non-existent. Similar results are obtained for women. On this basis, it appears that Japanese employers have little incentive to resist new union formation.(2)
Additional data also suggest that employer resistance is not very strong. The number of unfair labor practices complaints as a percent of the number of employees dropped from 0.20% in 1975 to 0.06% in 1993 (Rebitzer and Tsuru, 1994). According to this indicator, the level of employer resistance appears to have declined, rather than increased, in the 1970s and 1980s. In sum, employer resistance is not the determining factor in the decline in union density in Japan.
2. Employee Incentives to Join Unions
The preceding observation leads us to the other aspect of the process of union organizing: the attitude of workers towards unionization. The JIL survey asked non-union respondents whether or not they thought it would "be better for workers at your workplace to form a union?" 43% of male, non-union workers and 38% of female non-union workers answered "yes" to this question.
Non-union workers are more likely to support unions if they are dissatisfied with wages, benefits, and other working conditions. For this reason, those who support unions are also more likely to believe that unions improve wages and other working conditions than those who do not support unions. Table 2 compares the perceived consequence of the formation of a union on various aspects of working conditions. Only 14.9% of non-union men who would not support unions believe that a union in their firm would improve wages. In contrast, 54.0% of non-union men who would support a union believe that a union would improve wages. A similar pattern is observed for every aspect of the employment relationship for women as well as men.
The expected effect of unions by non-union workers poses a dilemma for Japanese unions because, as seen in Table 1, Japanese unions do not have much effect on wages. Of course, unions might attract new members if they improve important, non-monetary aspects of employment relationships.In particular, unions might be able to improve communication between labor and management. This "voice" effect has, in other countries, been favored to reduce turnover by eliminating the sources of dissatisfaction (Freeman and Medoff, 1984).
I evaluate the importance of the union "voice" effect by comparing the proportion of union and non-union workers who do not tell anybody about problems that they have in their workplaces. The results are presented in Table 3. The proportion of union workers who do not communicate problems is substantially lower than non-union workers. Clearly, the presence of unions alters the lines of communication in the Japanese firm.
The union voice effect is important because it might enable unions to attract non-union workers even if they do not improve wages substantially. To examine this possibility, I have estimated the union effect on an individual's job dissatisfaction and propensity for job change. Table 4 summarizes the coefficients of major variables. As shown, union membership does not have a statistically significant impact on either job dissatisfaction or propensity to quit. There is no doubt that Japanese unions "voice" various issues to management. It is not clear, however, that the effect of this voice is highly valued by individual employees.
IV. Conclusion
This essay examines why union density has declined in Japan. While various factors have contributed to the decline, one of the most significant determinants is the disinterest of non-union workers in unionization.
From this perspective, it appears that reversing the decline of union density may be difficult. To expand their membership, Japanese unions have recently begun to organize part-time workers and quasi-managerial employees who have traditionally been unorganized (Nakamura, Sato, and Kamiya, 1988). However, if union decline is largely due to the disinterest of regular, non-union workers, then current efforts by unions may not be sufficient to reverse the trend.
Notes
I am grateful to the members of the JIL project, Hiroki Hayashi (Hitotsubashi University), Machiko Osawa (Asia University), and James B. Rebitzer (Massachusetts Institute of Technology) for their help.
(1) A survey was conducted for the Japan Institute of Labour of union and non-union employees aged 18-59 who lived within 30 kilometers of central Tokyo.We randomly selected 2,800 names from municipal records for sampling. From this list, we excluded people who were not currently employed or who were self- employed or who were managers. Of the resulting 1,736 individuals, 1,104 agreed to be interviewed, yielding a response rate of 64%. All interviews were conducted in person at the residence of the respondents between July 2 and July 24, 1992.
(2) The similar conclusion that unions have little effect on wages is obtained by Tachibanaki and the Japanese Trade Union Confederation Research Institute for Advancement of Living Standards (1993).
References
Blanchflower, David G. and Freeman, Richard B., "Unionism in the United States and Other Advanced OECD Countries," Industrial Relations, Vol.31, No.1, Winter 1992, 56-79.
Freeman, Richard B. and Medoff, James L. What Do Unions Do?New York: Basic Books, 1984.
Freeman, Richard B. and Rebick, Marcus E., "Crumbling Pillar? Declining Union Density in Japan," Journal of the Japanese and International Economies, Vol.3, No.4, December 1989, 578-605.
Ito, Masanori and Takeda, Yukihiko, "Rodo Kumiai Soshikiritsu no Suii to sono Henka Yoin (The Change of Organizing Rate and its Causes)," Rodo Tokei Chosa Geppo (Monthly Labor Statistics and Research Bulletin), Vol.42, No.6, June 1990, 6-14.
Japan Institute of Labor, Rodo Kumiai Soshikiritsu Teika no Kitei Yoin (Determinants of the Decline in Union Density), Tokyo: The Japan Institute of Labour, 1993.
Nakamura, Keisuke, Sato, Hiroki, and Kamiya, Takuhei. Rodo Kumiai wa Honto ni Yakuni Tatteiru no ka (Do Labor Unions Really Have a Useful Role?), Tokyo: Sogo Rodo Kenkyusyo, 1988.
Rebitzer, James B. and Tsuru, Tsuyoshi, "The Limits of Enterprise Unionism: Prospects for Continuing Union Decline in Japan," Paper Prepared for the 10th World Congress of the International Industrial Relations Association, September 1994.
Tachibanaki, Toshiaki and the Japanese Trade Union Confederation Research Institute for Advancement of Living Standards, Rodo Kumiai no Keizaigaku (Economics of Labor Unions), Tokyo: Toyo Keizai Shinposya, 1993.
Visser, Jelle, "Trends in Trade Union Membership," OECD Employment Outlook, July 1991, 97-134.
previous page next page MENU Special Topic Title Index-Industrial Relations