Carl Mosk
University of Victoria


Did unlimited supplies of "cheap" labor fuel Japan's first seven decades of industrialization? Or is the period 1870 to 1950 more noteworthy for innovation on the demand side? specifically for the creation and refining of the nenko system, with its emphasis on skilling and retention of dedicated workers, and on wage payment according to age and seniority? In this paper I develop a labor segmentation model of the prewar Japanese economy in which supply matters more in the case of unskilled and semi-skilled labor, and demand matters more in the case of highly skilled labor. The model welds together a labor surplus framework for analyzing the female labor force in textiles with a particular type of efficiency wage model for the male labor force in heavy industry (a gift exchange model). Using this model I explain movements in wage differentials over the period 1880 - 1940 (male versus female; male white collar versus male blue collar), and the rationale behind the introduction and diffusion of the nenko system.


The prewar Japanese labor market segmented along both economic and social lines. In the social dimension sharp cleavages were delineated by sex, educational attainment; in the economic sphere the degree to which premium wages were required to elicit effort were paramount. Social and economic segmentation were interdependent and the family system bridged the gap between the two spheres.

In my version of the gift exchange model I differentiate between two types of gift exchange: general market specific gift exchange defined in terms of demographic characteristics (e.g.: age and sex as proxies for the family responsibilities an individual is expected to shoulder) and occupational and educational attainment; and firm specific gift exchange which attaches to individuals according to their within enterprise seniority, job assignment and productivity assessment. Market specific gift exchange rested on the elite samurai concept of the family, according to which the household head (invariably male) bore full responsiblity for the welfare of his dependents. Since the cost engendered by family responsibilities tended to increase steeply with age up until an individual's early 50's (given the age specific marriage and fertility characteristic of the era) gift exchange employers designing "fair" wage systems naturally used age as a primary determinant of remuneration. Firm specific gift exchange also reflected Tokugawa period samurai concepts: internal competition for ranking had been widely utilized in the fiefdoms to classify and pay warrior-bureaucrats, and the idea was carried over as a pay system for the elite white collar workers.

In the prewar economy the scope of the gift exchange market was narrow. For example self employed and family workers, ubiquitous because most of the enterprises in prewar Japan were either family managed farms or shops, had a direct vested interest in the fruits of production and hence required no effort inducing wage incentives. In routine government jobs and in most military positions, discipline could be maintained by non-pecuniary methods. Again in private sector jobs with short training periods and/or where a piece work system with "checkers" could be implemented on a low cost basis to monitor a worker's output and/or in work done on a put-out basis, effort could be readily monitored or was inconsequential or could be perfected remunerated through a piece work or examination system. Thus rank and file workers in light industry, for instance in textiles and food processing, were excluded from the gift exchange labor market. Indeed gift exchange was largely limited to the small (before 1920) heavy industrial sector, becoming entrenched in metals, machinery and chemicals during or before the interwar period. In these firms effort, flexibility, skill and the willingness to pass on skills to fellow workers was paramount.


The vast majority of textile workers in late Meiji and interwar Japan were young women, recruited from agricultural villages on contracts of two to five years duration signed under the authority of their household heads who dealt with representatives of the company. Thus calculations concerning the attractiveness or non-attractiveness of signing the contract were first and foremost calculations made by the patriarchal head of the household, and the opportunity cost of a girl's daily labor as far as the head was concerned was what she was likely to contribute to the household, that is was a day's worth of labor input on the fields farmed by the household. The proper way to conceive of opportunity cost is in time units. The wage floor for the textile industry was set at the marginal product of an hour's worth of labor time in agriculture which was positive, even though the marginal product of a worker was zero (because workers could be removed from the household without output being diminished, the remaining household members working harder). Surplus labor supply conditions characterized the interwar agricultural and light manufacturing labor markets in the sense that workers were in excess, but not in the sense of hours worked. For this reason as daily labor productivity in agriculture gradually crept upward so did daily wages in textiles. ry was able to readily import and adapt a foreign technology which could be combined with labor drawn from a inexpensive elastically supplied supply pool, management was able to disregard the bargaining power of labor, directly controlling workers, and wages were determined exogenously by the supply price of agricultural productivity. But in heavy industry the technology was far more sophisticated; the processes far more capital intensive; the skill formation period long; and monitoring costs far higher. And the price of sabotage was substantial. Hence "fair" wages which kept workers satisfied were a sine qua non for workers from whom effort and/or loyalty was sought. Initially the supply of effort supplying workers was restricted to elite white collar employees, who were responsible for running and managing the daily operations of the corporation.

Securing a pool of dedicated managers was essential to firms struggling over the control of the shopfloor. Initially male blue collar labor was supplied to firms through labor bosses. Corporations feared that if they continued to rely on an indirect style of management, working through the labor bosses, they would be saddled with a strike prone system of industrial relations, being held captive by their labor subcontractors. For this reasons heavy industry, spurred on by the example of direct control which light industry had managed to implement, gradually developed through trial and error experimentation a system of direct control in which nenko gift exchange principles served to motivate and bind workers. At first this gift exchange was limited to the white collar worker elite who had the responsibility for training and directing the blue collar rank and file.

But with time, the relative economic and social status of the white collar elite declined vis-a-vis rank and file blue collar workers. Gift exchange gradually diffused down to the shopfloor. In part this came about because government intervened. But the main reason lies in the changing supply/demand balance in the markets for white collar and blue collar workers. The supply of technical graduates from vocational schools, colleges and universities expanded rapidly and this eroded the relative status and earnings of white collar workers; blue collar supply expanded more slowly relative to demand and this tended to enhance the relative position of blue collar workers.


If heavy industry wages are set in a non-surplus labor supply/ gift exchange manner, we would expect nominal wage increases to be less sensitive to price index increases, and more sensitive to productivity gains, than they are in the case of light industry.

Analysis of time-series data for the interwar period on the capital intensive transportation and communications, and utilities, sectors, bears this out. In particular nominal wages in these sectors are far less sensitive to price index changes than they are in textiles.


In closing I would like stress the interrelationship between gift exchange and labor segmentation in late nineteenth and twentieth century Japan. During Japan's first seven decades of industrialization the labor market subdivided into three main submarkets: a "cheap" female labor market in light industry; an elite male white collar gift exchange labor market in heavy industry; and a male blue collar market initiated dominated by labor bosses. Seeking control over blue collar workers in this market, industrialists attempted to internalize it by bringing the training of workers under the aegis of the elite managerial class. In doing so the industrialists created an unstable social equilibrium in the firm: blue collar workers desired either greater control over their skills or the gift exchange accorded the white collar elite. Changing supply and demand in the labor market settled the issue. Gift exchange diffused down to the ranks of the shopfloor and Japan's modern system of industrial relations blossomed forth.