Imaoka Hideki
University of Shimane
For a long time already the Japanese' economy has been a source of wonder and puzzlement. Its performance has been remarkable while at the same time its practices and institutions seem to differ markedly from those existing in other developed economies. These feelings have certainly something to do with the fascination that the Japanese culture and society generally inspire. Perhaps in no other area of economic activity are these societal and cultural aspects as important as in. the working of labor markets and in the makeup of labor practices at the firm level. Observers of the Japanese labor markets have been in particular impressed by its low rate of unemployment which, in spite of a prolonged economic crisis, has remained consistently below the 3 percent mark. In addition, labor relations seem to be, at least when measured by the number of days lost as a result of strikes, much more harmonious than in other advanced countries. Yet a third feature that impresses outside observers is the level of cooperation that seems to exist at the level of the firm and that finds its expression in sophisticated practices such as the quality control circles (QCCs). Some of these practices have been applied, with relative success, to other countries, both developed and developing.
These observations have impressed among many observers the perception that Japan is unique. In this, they are supported by many Japanese who also believe that their culture, history, and environment, have produced a society that is not only unique but also unreadable by foreigners. The "culturalist" school of Japanese labor relations was thus born1, and from the beginning it recognized two main characteristics that were to be associated with the Japanese way: the lifetime employment system (shuushin koyou seido) and the wage seniority system (chingin nenkou joretsu). The former refers to the understanding that employment will normally last for the whole useful life of the employee. Associated with this is the total commitment of employees to the fortunes of their enterprise, in such a way that the relation between employers and employees goes beyond the contractual type of relations that would be usual in other countries. The latter refers to the relation between time of service (which often is related to age) and levels of wages as well as of rank within the enterprise. Besides these two, there are some other particular features of the Japanese labor markets that were highlighted. In the first place, trade unions at the shop level (rather than industry-wide federations or national confederations) are the most significant labor organizations from the point of view of labor negotiations. Secondly, and also in relation with this, labor negotiations are synchronized and take place every year during the spring year. This is the famous spring offensive, or shuntou2. In the third place, a large portion of the annual income of workers takes the form of bonuses, usually paid twice during the year, in June and again in December. Finally, the rate of female participation in the labor markets is lower than in other developed countries.
The popularity (particularly in the West) of the culturalist school has lasted until now, in spite of the mounting criticism of many economists. In particular, most Japanese labor economists disagree with the explanations offered by the culturalist school for the specific performance of the Japanese labor markets. We will discuss these arguments in the next section. However, in spite of a very basic disagreement, most Japanese economists would accept that, at least to some extent, the preceding features differentiate the Japanese system of employment from that of other developed countries.
The controversy surrounding the proper characterization of the Japanese system of employment must be of course placed within the context of the ongoing debate about the relative merits of the Japanese and American economic systems as well as about the future of the relations among both countries. Some of the questions being asked about the Japanese system of employment are also asked about other particularly Japanese economic practices and sound familiar: is Japan different? If so, is it changing? Is change necessary to the survival of Japan in the new international environment? The issue of adaptation to new internal and external realities is of utmost importance.
Let us finally place this characterization within the context of current economic conditions. As is well known Japan has endured for the last four years an acute recession, which followed the longest expansionary period of the post-war, the so-called Heisei boom (1987-1990). The productivity and GDP growth show the relatively small variation in the unemployment rate, albeit the latest figures for 1995 place this indicator at around 3.1 percent. It can be finally noted that, in spite of large drops in productivity during the current recession, the real wage rate has suffered a very small erosion. These numbers confirm the casual observation that the current recession, however long and deep, has not unduly strained labor conditions.
Let us start by presenting the basic facts on the Japanese labor force.
In 1993 the labor force was composed of some 66 million people. As is well known, unemployment is comparatively low in Japan, and in 1993 only 1.7 million people were out of a job, corresponding to a rate of open unemployment of 2.57 percent. Some 80 percent of all engaged persons were employees, the vast majority in non-agricultural sectors. If we exclude public servants, we can see that more than 72 percent of employees work in firms with less than 500 workers, and that more than 54 percent of employees work in firms with less than 100 workers. These figures express well how important small and medium firms are in the Japanese economy and labor market. The well-known and large firms, which we might identify with those firms employing at least 1000 workers, account for about 21 percent of all employees in the private sector. If anything, the importance of the small and medium firms has increased in the last 20 years. While the total number of employees increased by about 42.8 percent in this period, the number of those employed in firms with more than 500 workers increased by 39.1 percent, and that of those employed in firms with more than 1000 workers increased by a mere 30.3 percent. Another observation that can be drawn from these data is that the proportion of workers employed in the manufacturing sector (and also in the public sector) has decreased steadily, while those employed in the retail and wholesale trade and in the services sector has increased accordingly.
Let us look now into the various modalities of employment. In addition to managing positions and regular employees, the two most important other modalities are part-timers and "arubaito". Even though part time employees are not regular employees, do not belong to the firm union and have a tenuous contractual relation, in some cases they can work many hours. In principle, part time employees may work as many as 35 hours a week. The proportion of part time employees is higher in Japan than in most other developed countries, and it has shown an increasing trend at least in the last 20 years. It is also to be noted that few male workers practice this modality, while on the other hand female workers adhere to it widely, a pattern that has only changed slightly in the last years. This fact is related, of course, to the division of social activities within the family. Women usually take up jobs, either regular or part time, early in their adult life. They often quit after marrying or giving birth and, when the children grow up, may return to work, most often as part-timers. "Arubaito" (from the German arbeit) refers to work contracted by the hour or the day. This modality is widely practiced in restaurants, gas stations, and similar non-career building occupations. As it can be figured out from the large proportion of workers in the 15-24 years old group that are employed in this modality, students make up the largest contingent of "arubaito" workers. We will see later that the fact that a relatively large proportion of all employees belong to either of these two modalities is one of the factors behind the flexibility exhibited by the Japanese labor market during the recent recession.
With these facts and figures in mind, let us proceed to an examination of the two features most often associated with the Japanese employment system, namely the lifetime employment system and the wage seniority system. As said above these are considered by the culturalist school to be the defining, and in particular unique, characteristics of the Japanese employment system. As such, they have attracted the attention of critics of this school, most of them well-known Japanese labor economists3.
Ideally, the lifetime employment system implies that workers enter employment at a given firm immediately after graduation and remain in that firm until the age of retirement. Empirical studies of the degree to which this ideal approximates actual conditions have looked at two sorts of surveys. One is to ask from employees whether they have or not been working in the same firm since graduation. Surveys of this type, rather predictably, yield very low proportions of lifetime employees. According to Tachibanaki (1984), and using data for the years 1979-80, as little as 9.8 percent of male employees in the 50-54 years old age group belong to this category. Even for male employees in the 30-34 years old group the proportion is only 33.4 percent. Rates for female employees are much lower. The survey also shows that lifetime employment becomes more common the larger the size of the firm and the higher the educational level of the employees. This type of survey, however, can be criticized as it interprets the concept of lifetime employment too literally. The point of the culturalist school, which this literal interpretation would miss, is that corporate employment in Japan is highly cohesive (see, for more details on the concept of cohesiveness, Imaoka (1989)). A second type of survey asks employees how long they have stayed with the same firm. As a matter of fact, these data appear, with a great degree of disaggregation, in the yearly Census on Wages (Chingin Sensasu) compiled by the Ministry of Labor of Japan. Comparable data are not compiled so systematically in other developed countries, but still some comparisons are possible on the basis of existing statistics. Table 1 presents the result of a comparison between Japan and the United States conducted by Tachibanaki (1984) with data for the years 1977-78. Substantial differences in length of service are apparent, especially for female workers and for male workers less than 54 years old. The fact that length of service Is relatively low for employees older than 54 years is due to the mandatory retirement age (60 to 63 years usually) that is prevalent in large firms, which induces many employees to move rather early to other positions. Table 2, which displays more recent data for Japan, shows that length of service patterns have not changed much for workers in the 30-64 years of age group, while for other age groups the trend is for a decrease in length of service.
Table 1. Length of service in the same firm, Japan, 1977 and the United States, 1978
| Japan (1977) | USA (1978) | |||
| Age | Males | Females | Males | Females |
| 16-19 | 2.12 | 2.12 | 0.96 | 0.83 |
| 20-24 | 4.17 | 3.62 | 1.83 | 1.65 |
| 25-29 | 7.12 | 5.88 | 3.13 | 2.82 |
| 30-39 | 12.37 | 9.10 | 5.95 | 4.11 |
| 40-54 | 19.52 | 15.32 | 11.93 | 7.34 |
| 55-64 | 19.38 | 19.27 | 16.74 | 11.18 |
| Above 64 | 21.28 | 22.14 | 18.33 | 12.36 |
Source: Tachibanaki (1984)
Table 2. Length of service in the same firm, Japan, 1977-93
| 1977 | 1985 | 1993 | ||||
| Age | Males | Females | Males | Females | Males | Females |
| 16-19 | 2.1 | 2.1 | 1.1 | 1.0 | 1.0 | 1.0 |
| 20-24 | 4.2 | 3.6 | 2.7 | 2.8 | 2.5 | 2.5 |
| 25-29 | 7.1 | 5.9 | 5.4 | 5.4 | 5.0 | 5.0 |
| 30-39 | 12.4 | 9.1 | 11.1 | 7.9 | 10.2 | 8.1 |
| 40-54 | 19.5 | 15.3 | 17.7 | 10.0 | 18.7 | 11.0 |
| 55-64 | 19.4 | 19.3 | 15.1 | 14.5 | 18.3 | 13.6 |
| Above 64 | 21.3 | 22.1 | 12.1 | 14.5 | 11.6 | 15.3 |
Source: Tachibanaki (1984), Imaoka (1989) and Ministry of Labor of Japan (1994b)
Koike (1988) did another comparative study, in his case between Japan and the countries of the European Community (at that time, Belgium, Britain, France, Italy, Luxemburg, the Netherlands, and West Germany). The survey, however, dealt with a rather different question, namely the composition by length of service of the male labor force in manufacturing. He found that, when looking at the aggregate figures, the proportion of Japanese workers with more than 20 years of service (data from 1970-71) was consistently lower than that for European workers (data from 1972). The comparison became even stronger when the figures for Japanese employees of large firms were used, albeit in the case of the European Community only aggregate data were available. From this analysis Koike concluded that claims to the effect that lifetime employment is a Japanese characteristic are misplaced. Koike's analysis, however, seems to be incomplete. In order to reach conclusions about the cohesiveness of the employment structure, from data on the composition by length of service of the labor force, additional comparative data on the age structure of the labor force would be needed. This kind of data is not presented nor discussed in Koike (1988) and it seems not to be available.
Let us move on now to the second feature, the wage seniority system. Again ideally, this means that wage levels as well as rank within the firm are in direct relation to the length of service in the firm. Empirical verifications of this feature, in practice, consist of measuring wage differentials by age. This is due, of course, to the difficulty of collecting both consistent global statistics on rank within the enterprise and associative data of wage rates according to length of service. The hypothesis to be tested, then, is that about the existence of a strong positive relation between age and wage level. Data for the year 1993 indeed confirm this hypothesis for workers up to 54 years old, and also show that the relation is stronger for male workers. The fact that wage levels start to decrease after this age is consistent with the fact, noticed above, that employees at this age start to move to other jobs with a higher frequency. The hypothesis is maintained consistently when we disaggregate by educational level. Sustaining this hypothesis proves only one part of the second basic tenet of the culturalist approach, which in addition maintains that this trait is unique to Japan. Koike (1988) did comparative analyses of the wages-age relation for Japan and the USA on the one hand, and for Japan and the countries of the EC on the other hand. He concluded that this sort of positive correlation between wages and age is not unique to Japan and that it can also be found in the USA and the countries of the
EC. However, the relation holds in those countries only for white-collar workers. Its validity also for blue-collars could indeed be considered a feature of the labor market which is unique to Japan, which Koike termed "blue-collarization".
From this brief survey we may conclude that the two principles of culturalist school appear to hold rather consistently. However, it is difficult to conclude that they apply uniquely to Japan or that they derive from cultural aspects. To some extent, the principles also hold in the labor forces of other developed countries. The difference between Japan and other countries, barring more conclusive evidence, seems to be one of degree.
Unions
One of the distinguishing features of the Japanese system of labor relations is that most trade unions are organized at "the level of the enterprise, and that it is there that lies the center of trade-union activities and decision-making power. Exceptions to this rule are the Seamen Union and, among public workers, the Teachers Union and the Municipal Workers Union4. At the firm level, more than one union may be established, and indeed some 13 percent of all firms have two or more unions. Enterprise-based unions are organized into sector-wide labor federations and these, in turn, belong to labor confederations. Since 1987, and after a long and difficult process of unification, there has been one major confederation of labor grouping workers in the private sector, best known by its Japanese acronym Rengo (Japanese Private Sector Trade Union Confederation). In 1993, its membership comprised about 62 percent of all unionized workers.
The role of trade unions is to represent its members at the time of labor negotiations and also, in a more routine way, to manage day to day relations between labor and management. Only regular employees can be members of a union. In addition to part-time employees or employees with fixed-term contracts, employees holding management or supervisory duties are also barred from union membership. Trade unions flourished after the end of World War II, and in 1950 about 46 percent of all workers belonged to a union. This coverage has decreased steadily since then and in 1993 only 24.2 percent of all workers were members of a union. In spite of that, the coverage of trade unions is much higher in Japan than in the United States, although well below that existing in most European countries. In general, the larger the size of the firms, the higher the proportion of unionized workers. There are also important differences depending on the sector of the economy. Public workers have the highest rates of unionization, wholesale and retail trade has one of the lowest rates, and manufacturing is slightly above the average.
It is generally recognized that labor relations are more harmonious in Japan than in most other countries. The fact that most unions are enterprise-based may help in explaining this, as most potential conflicts can be identified and handled at an early stage. Another possible factor could be that labor negotiations are coordinated, happening around March of every year (this is the famous shuntou, or spring offensive), and that their results hold for a whole year. An examination of the pattern of spring wage increases, together with the observation above that unionization rates are higher in larger enterprises, supports the hypothesis that unions have been successful in obtaining higher wages for their members. This hypothesis is also confirmed by an earlier albeit much more detailed study by Taira (1970). He also noticed that this successful "shuntou unionism" has as a counterpoint the increase in the use by firms of nonregular workers, as part-timers and "arubaito".
Legislation
Most of the legislative framework underlying labor relations derives from laws enacted in the years immediately after the end of World War II, and in particular bears the mark of the American occupation authorities. Although labor legislation is quite complex, some of the most important labor laws are: the Labor Relations Law (September 1946); the Trade Unions Law (June 1946); the Labor Standards Law (April 1947); the Employment Security Law (November 1947), all of them subsequently revised and updated. Both the laws on Labor Relations and on Trade Unions establish the basic rules and institutions for relations between labor and management. In addition, the Labor Relations Law establishes a Central Commission for Labor Relations and Commisions for Labor Relations at local levels, to provide support for negotiations between labor and management. The Labor Standards Law defines working conditions, such as working hours, procedures for payment of wages, general safety standards, and other standards relating to the workplace environment. In addition to setting these standards, inspection bodies are established and penalties for infractions are set.
Legislation aimed at employment stability and at alleviating "the consequences of unemployment is comprehensive. The Constitution of Japan establishes that everybody who is willing and able to work has the right to do so, and the Employment Security Law is explicit in its assignment to the government of the main responsibility in maintaining equilibrium in the labor markets. Bureaus of Employment Security are set at central and local levels to take the necessary steps to this effect. Some of the activities of these Bureaus are to channel information so as to match supply and demand and to offer vocational training to displaced worker. On various occasions, and according to the circumstances, special laws have been enacted to strengthen this sort of activities by the government authorities. The Employment Insurance Law (December 1974) set up a national system of unemployment compensations. All regular workers of firms are eligible to participate in this scheme, which is administered by the national government. Operational and overall financial backing is provided by the state, while workers and firms divide equally a premium of 1.5 percent of wages. As of March 1993 some 32.8 million workers belonging to 1.8 million establishments were covered by this scheme which, through the fiscal year of 1992 (April 1992 to March 1993), paid almost 900 billion yen (10 billion dollars) to 571 thousand beneficiaries (monthly average). Part-time and "arubaito" workers, due to their type of working arrangements, are rarely eligible for any unemployment compensation. During the fiscal year of 1992 slightly less than 50 thousand day workers (monthly average) received unemployment compensation.
This brief discussion of labor legislation illustrates that Japanese labor policies are active in promoting the level of employment and regulating labor conditions. The government establishes annual and pluriannual labor plans, which contemplate measures to retrain workers, provide relief for sunset industrial sectors, and so on. Emphasis has been placed recently in the reduction of hours of work, and several measures have been taken to this end. Through the working of labor legislation and institutions, and the execution of labor policies, it can be judged that the government provides an important contribution to the flexible operation of labor markets.
It has long been noticed that the rate of unemployment in Japan is low if compared with those of other industrial countries. In addition, it appears that this variable is not very sensitive to business conditions. Even during the current recession, probably the deepest in the post-war period, the rate of unemployment has barely exceeded 3 percent. Again, several factors have been cited to explain that Japan is not so different, after all, from other countries. Chief among them are the low rate of participation of female and young people in the labor force, differences in definitions of unemployment, and the relatively high proportion of people employed in family and rural businesses (where employment is inherently more stable). Even taking account of these factors it still remains, as Ito (1992) convincingly shows, that the unemployment rate in Japan is lower and less responsive to business conditions than in other developing economies.
Comparative analysis of employment adjustment
Following Ono (1989) we will proceed to analyze the main supply and demand factors behind employment in several periods and across several industrial countries. Emphasis will be on two generally recessionary periods, 1973-76 and 1979-82, and the countries considered, besides Japan, are Canada, the USA, France, West Germany, and the United Kingdom.
Let Q stand for real GDP, E for employment, H for hours per worker. We can define productivity a by the formula:? = Q/EN
If, in addition, we consider active labor population F=?N, where ? is the rate of labor participation and N is the population 15-64 years old, and define the employment rate e by e = E/F, we may combine all of these relations into E = Q/a?
If we instead consider rates of change, we are left with the final expression
E=Q-a-?-H-N
which provides us with a decomposition of employment into its demand (a,Q,H) and supply (?, N) factors.
Let us refer to Table 3 for the purposes of the international comparison, and consider first the period 1973-76, which roughly corresponds to the first oil shock. For all countries subject to this analysis there were drops in the employment rate and also in the number or hours worked (per person). In addition the rates of growth of GDP and of productivity were lower than in the previous period (1970-73). Thus, the behavior of the demand factors was not qualitatively different across these countries, although the rate of growth of total labor hours (Q-a), while positive in Canada and the USA, was negative for all other countries, especially so for W.Germany. Turning now to the supply factors, let us first consider N. We see that its rate of growth had the largest drop in W.Germany and Japan, and that it was low anyway in Japan and in the European countries if compared with the values for Canada and the But perhaps the most interesting variable is the participation rate, which increased in all countries except Japan and W.Germany. It can thus be seen that there were many national differences regarding the behavior of the supply factors, with some similarities between W.Germany and Japan. In the end, the superior performance of Japan, reflected in the very small drop of employment observed during this period, seems to be due more to the low growth of labor supply rather than to a particularly good behavior of its productive sectors.
Таblе 3. Rаtеs оf chаngе оf the еmр1оуmеnt rаte аnd of its explaining factors
(реrсеntаgе сhаngе in three-year period)
| e | Q | a | Q-a | b | H | N | |
| Canada 1970-73 1973-76 1976-79 1979-82 1982-85 | 0.13 -1.66 -0.34 -3.84 0.54 | 20.40 13.67 12.54 1.71 13.22 | 10.26 7.77 2.79 1.99 4.15 | 10.14 5.90 9.75 -0.28 9.07 | 3.59 2.21 4.11 1.53 2.34 | -0.25 -2.27 0.26 -2.84 2.92 | 6.67 7.62 5.72 4.87 3.27 |
| USA 1970-73 1973-76 1976-79 1979-82 1982-85 | 0.06 -2.96 2.00 -4.08 2.76 | 13.08 2.93 12.12 -0.46 13.40 | 2.84 -0.04 0.90 1.91 1.80 | 10.24 2.97 11.22 -2.37 11.60 | 2.24 2.07 3.96 1.19 1.75 | 2.26 -1.47 0.25 -3.23 4.11 | 5.68 5.33 5.01 3.75 2.98 |
| Japan 1970-73 1973-76 1976-79 1979-82 1982-85 | -0.12 -0.74 -0.10 -0.28 -0.27 | 21.86 6.20 16.40 11.55 13.21 | 21.48 10.26 10.35 9.05 8.98 | 0.38 -4.06 6.05 2.50 4.23 | -0.38 -1.67 1.76 0.74 0.19 | -2.87 -4.34 2.13 -0.38 1.23 | 3.75 2.69 2.26 2.42 3.08 |
| France 1970-73 1973-76 1976-79 1979-82 1982-85 | -0.25 -1.86 -1.62 -2.37 -2.38 | 17.62 8.77 10.49 3.35 3.39 | 17.87 12.54 10.99 6.72 7.02 | -0.25 -0.37 -0.50 -3.37 -3.63 | 0.14 0.49 0.78 -1.68 -1.92 | -2.90 -4.59 -1.92 -3.19 -2.01 | 2.76 2.19 2.26 3.87 2.70 |
| W.Germany 1970-73 1973-76 1976-79 1979-82 1982-85 | -0.45 -3.07 0.77 -3.59 -1.77 | 12.26 4.02 10.30 0.88 7.00 | 13.60 12.41 7.80 4.67 7.56 | -1.34 -8.39 2.50 -3.29 -0.56 | -0.94 -2.32 -0.70 -1.77 -0.26 | -2.30 -3.21 0.68 -2.61 0.10 | 2.35 0.21 1.75 4.18 1.37 |
| UК 1970-73 1973-76 1976-79 1979-82 1982-85 | 0.02 -2.77 0.24 -6.33 -1.04 | 13.08 2.18 6.97 -2.49 9.97 | 12.75 6.79 3.65 6.39 4.69 | 0.33 -4.61 3.32 -8.88 5.28 | 0.88 1.38 0.58 -1.36 1.37 | -0.95 -3.85 1.00 -2.97 3.06 | 0.38 0.63 1.50 1.78 1.89 |
Let us move on to the period 1979-82, which is associated with the second oil shock. Again, for all countries employment and the number of hours per worker diminished while the rate of growth of GDP was also lower than in the previous period, thus characterizing a recessionary period. This time the rate of growth of productivity was lower for all countries except for the USA and the U.K. The latter two were also the only ones experiencing negative growth and, consequently, also those with the largest fall in the employment rate. If we examine the factors coming from the demand side, Japan was, during this period, the only country where the rate of growth of total labor hours worked (Q-a) was positive. Total labor hours fell the most in the European countries. In all countries hours per worker fell, but by the least in Japan. The demand factors for the countries surveyed behaved thus quite differently during the period 1979-82 than in the first oil shock. As for the supply factors, again we notice differences. The Japanese rate of growth of working age population (N ) was again one of the lowest (together with that of the U.K.). The rate of labor participation decreased in all European countries, rose slightly in Japan and more markedly in Canada and the once again superior Japanese performance on employment is to be attributed this time to both demand and supply factors.
What emerges from this interesting analysis by Ono is that, overall, supply factors such as the rate of labor participation and the low growth rate of the population are as important, if not more, than factors related with the productive sector. It is these factors that are more clearly related with the traditional characteristics of the Japanese system of employment, such as the lifetime employment system. Although this can be seen as a blow to the "culturalist" theories of labor, we will see later that these theories may find a new life in the explanations of the apparent adjustment of the labor participation rate to business conditions.
Fluctuations in employment and the business cycle
Similar findings to Ono's come from a study by Ministry of Labor of Japan (1994a). They examine the empirical evidence on absolute and relative fluctuations in employment. Regarding absolute fluctuations in the level of employment, we will distinguish two periods, 1963-75 and 1976-92, and use as a measure of fluctuation the standard deviation of employment.
Table 4 presents the results of such observations for Japan and major developed countries. It can be seen that employment has generally been stable, and that it became more stable during the second period in France, Germany and Japan and less for the other three countries of the sample. During the second period, Japan had the most stable level of employment among major developed countries, except for France1. Table 4 also shows the correlation coefficient between the rate of economic growth and rates of change in employment for the same periods and countries, a measure of relative stability of employment. Again Japan exhibits the most stable situation, especially so during the second period. The general observation in relation to these data is that, if anything, stability of employment in Japan has become more marked after the first oil shock.
Table 4. Relation between economic growth and changes in Employment (percent)
| Change in employment (standard deviation) | Correlation coefficient between economic growth and change in employment | |||
| 1963-75 | 1976-92 | 1963-75 | 1976-92 | |
| Japan USA Canada France Germany UK | 1.25 1.44 1.41 1.10 1.80 1.19 | 0.80 1.78 2.21 0.71 1.17 1.97 | 0.449 0.673 0.798 0.860 0.864 0.428 | 0.019 0.887 0.792 0.758 0.604 0.649 |
Source: Ministry of Labor of Japan (1994a)
As was pointed out earlier, an important factor to explain stability of employment in Japan arises from the way in which employment adjusts in response to changes in output. At a first level, a change in total labor hours worked will happen. At a second stage, it is useful to decompose this figure as the product of total number of workers and of hours per worker. A recent study from the Ministry of Labor of Japan6 found that (1) adjustment in terms of total labor hours is important in Japan, albeit smaller than in most other countries; (2) the contribution of adjustment of per capita labor hours is larger in the Japanese case than in most other countries; and (3) the contribution of adjustment of employment in the Japanese case is by far smaller than in other countries.
Tools of employment adjustment and income effects
Let us explore now in more detail the instruments and methods used by Japanese enterprises to deal with the question of employment adjustment. A comparative survey conducted in 1988 by the Ministry of International Trade and Industry (MITI (1988)) noticed that few Japanese enterprises (1.8 percent of those surveyed) resort to the firing of employees for the sake of employment adjustment. The instruments favored, when preparing long-term employment plans, are instead adjustment in the amount of overtime (44.9 percent) and in new hiring (37.7 percent). On the other hand, 21.1 percent of enterprises surveyed in the United States resorted to the firing of employees for the purpose of employment adjustment.
A more recent survey, referring to the period October - December 1993, found that the major instruments used by Japanese enterprises for the sake of short and medium term employment adjustment were control of overtime (26 percent of enterprises surveyed), reduction or suspension of mid-career hirings (18 percent), transfer within the enterprise (11 percent), and transfer within the group of related enterprises (8 percent).
Overall, it can be said that the Japanese enterprises do not appear to be willing to fire their employees for the purpose of employment adjustment. They prefer to adjust their total labor costs by applying flexible labor hours and flexible wage policies. Another dimension of the flexibility of Japanese labor markets must also be highlighted, one that is internal to the enterprise or to the group of related enterprises (keiretsu). The big enterprises use often the instrument of transferring employees within the enterprise or group of enterprises. In a study conducted over the year 19927 about the motives behind flows of personnel in workplaces, it was found that 59.3 percent of these movements were due to transfers to other positions within the same firm, 5.2 percent to transfers within the same group of enterprises, 25.4 percent were leaving the firm for contractual reasons (retirement or end of fixed-term contract), and 10.1 percent were fired due to business conditions Undoubtedly all these tools of adjustment reflect on per capita labor income. It is interesting to examine the increase in labor share that corresponds to a one percent increase in GDP and decompose it into its two components, namely the increase in the level of employment and the increase in per capita labor income. Such an exercise was carried by the Ministry of Labor of Japan for the six major OECD countries8. In the case of Japan it is found that almost 100 percent of the labor share increase is explained in terms of per capita labor income. Exactly the opposite situation is found for Canada, while for the USA more than 90 percent of the labor share increase is accounted for by increases in the level of employment. For the other countries of the sample it is also true than more than half of the change in the labor share must be attributed to changes in the level of employment.
Female participation in the labor market
Let us look now more closely to the labor participation rate, in particular the female participation rate. If we refer to Table 7.12 for the recent figures, we can see that the total rate of participation in the labor market is quite stable, with a maximum variance of 2.4 percentage points. Both the female and the male participation rates have separately a higher variance (3.1 and 2.7 percentage points respectively), the male rate being much higher than the female rate. Female participation has been increasing while the male rate has declined slightly. These figures confirm the general stability of the conditions in the labor market but they also indicate that some changes are slowly developing, which tend to increase the importance of women in these activities.
Table 5. Labor participation rate by gender, Japan, 1980-93(percentage)
| Year | Total rate | Female rate | Male rate |
| 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 | 63.3 63.3 63.3 63.8 63.4 63.0 62.8 62.6 62.6 62.9 63.3 63.8 64.0 63.8 | 47.6 47.7 48.0 49.0 48.9 48.7 48.6 48.6 48.9 49.5 50.1 50.7 50.7 50.3 | 79.8 79.8 79.5 79.4 78.8 78.1 77.8 77.3 77.1 77.0 77.2 77.6 77.9 78.0 |
Source: Ministry of Labor of Japan (1994a).
It was already noted before that, during the period 1973-76, only Japan and W.Germany exhibited a drop in participation rates. If this rate is further examined to differentiate according to gender we find an interesting difference between Japan and the other surveyed countries. Namely, while in all countries (including Japan) the participation rate for males drops, the corresponding rate for females decreases for Japan (and also for W.Germany, albeit the figure is in this case insignificant) and increases for all other countries. In addition, the drop in the Japanese male participation rate is, except for the U.K., the lowest. This is quite striking and demands further examination.
It is generally considered that there are two potential effects that adverse business conditions may have on the labor participation rate. On the one hand, the need, actual or expected, for supplementary income may drive new entrants into the labor market. This effect will produce an increase in the participation rate. On the other hand, the difficulties in getting a job, as well as the usually depressed wages, may induce people to wait or simply stay away of the job market. As the empirical evidence shows, in many countries the disincentive effect is dominant for males while the opposite effect is dominant for females. The case of Japan is different. The participation rate of males is little changed, while that of females goes down sharply. This phenomenon has been analyzed by several Japanese authors, among them Ono (1981), Higuchi, Hayami and Seike (1988), and Higuchi (1992). In what follows we will make mainly reference to the latter work (by Higuchi).
The study by Higuchi compares female labor participation for Japan and the U.S.A. It attempts to explain the differences in behavior by looking at the differentiated effect of several possible factors that would determine the female labor participation rate. The most important among these factors are: level of education, place of residence (size of city), age, family structure, number of children, and income of head of household. The study finds that having higher education has a larger effect in the female labor participation rate for the U.S.A than for Japan. The second factor, the size of the city of residence, has a similar effect in both countries, namely, the participation rate is lower for big cities. The third factor, age, is examined for the female population aged 30-44 years. Here we see that, while in Japan the labor participation rate increases with age, the contrary happens in the The fourth factor, family structure, refers to whether people live or not in nuclear families (including parents or other relatives of the wage earner or spouse). As would be expected, females living in nuclear families have a higher labor participation rate in both countries. As for the potentially negative impact of children in female labor participation, Higuchi found that this effect is much smaller in the USA than in Japan: female rates of participation were respectively 36.8 and 17.1 percent in the USA (white people) and Japan when the age of the children was less than 5 years, and again 51.7 and 28.7 percent in the case of children of more than 5 years of age. The final factor, income of the head of the household, has as expected a negative impact on female labor participation. However, this effect is much larger in Japan (with an elasticity of -0.2383) than in the USA (the elasticity is -0.0924).
Bonus payments as an adjustment tool
As said before, one of the peculiarities of the Japanese employment system is the bonus payments, which are received by blue collar workers as well as white collars. Paid twice a year (summer and end of year), they rarely represent less than one fifth of total annual remuneration and may in some cases be equivalent to more than six month's wages. This peculiar institution has been much studied and there are many interpretations of it. One of the main issues is whether the bonuses payments are used as an adjustment tool or not. Casual examination of Table 6 seems to support the existence of a procyclical relation between the business cycle and bonus payments, especially when the latter are measured in relation to the monthly This sort of result is maintained if one looks at data for the whole economy, and also if firms are broken down by size. In all cases we observe that the ratio of bonus payments to the monthly wage goes down during a recession and that it increases during an expansionary period.
Going beyond these observations, some authors (see e.g. Weitzman (1984)) have asserted that bonus payments are the expression of a deep link between employers and employees, what has been termed the "share economy". In this view, the variation in bonus payments is positively associated with the change in profits. Ono (1989) put these ideas to an empirical test. He estimated a regression with the rate of change in per capita bonus payment as a function of the rate of change in per capita profits, the rate of change of the consumer price index, and other independent variables. As expected, positive relations were found for both variables. The profit effect was not as strong for regular workers as for managerial workers, and there was in all cases a strong relation between the rate of change in bonus payments and the rate of change in the consumer price index. A similar regression was run using instead the change in per capita wages for regular workers. In this case, the relation with the rate of change in profits was statistically insignificant, while that with the rate of change in the consumer price index was high. This also confirms the common view that wages are less flexible than bonus payments.
Table 6. Change in bonus payments in manufacturing industry and their relation to monthly wagesi, Japan, 1980-93(percentage, number of months)
| Year | ||||
| Summer | End of year | |||
| Change | n.of month | Change | n.of month | |
| 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 | 10.0 5.1 4.4 1.9 3.1 3.7 1.5 0.3 4.0 8.5 7.8 4.5 2.7 -2.0 | 1.48 1.45 1.40 1.37 1.36 1.39 1.38 1.36 1.35 1.42 1.44 1.54 1.54 1.43 | 7.3 5.7 3.1 1.5 3.8 4.3 -2.3 2.6 6.5 8.6 7.3 2.2 -0.9 -1.1 | 1.67 1.64 1.58 1.53 1.55 1.56 1.50 1.52 1.54 1.61 1.66 1.71 1.65 1.53 |
Note. Data for enterprises with at least 30 employees.
Source: Ministry of Labor of Japan (1994a).
In this paper we have reviewed several mechanisms of adjustment in the labor market and their relation to the traditional practices associated with the Japanese system of employment. The insensitivity of the unemployment rate to recessions, well documented during the first oil shock and basically confirmed during the current recession, seems to be explained to a large extent by adjustments in the female labor participation rate and in labor hours. Average weekly hours of work went from 174.0 in 1990 to 159.4 in 1993. Both of these mechanisms of adjustments can be explained in reference to the lifetime employment system. The low female participation rate may be due to the fact that the (male) head of household has a secure job in most cases. The reduction in labor hours is a rational response to recession by firms (with government prodding), given the fact that Japanese firms cannot easily alter the number of their employees. Thus, the rigidity implied by the lifetime employment system is moderated by the flexibility of adjustment in other variables.
An interesting illustration of the working of the lifetime employment system comes from the evolution of productivity in the manufacturing sector. Typically in Japan productivity goes below its trend line during a recession and above it during an expansion. The opposite is typical instead of the This fact is explained by the different handling of their labor forces. As it was well documented in the previous section, while it is common in the to lay off workers during a recession, Japanese firms practice "labor hoarding". This practice, though it gives some rigidity in the short term, has served the Japanese economy well in the long run. As the current recession seems to be nearing its end, there is no evidence that the behavior of Japanese firms has changed in any substantial way from this pattern.