چگونه اولین کار برای زندگی حرفه ای فرد در ژاپن مهم است؟

Exploiting annual information on the work status of female workers from the Japanese Panel Survey of Consumers (JPSC), this paper examines how an individual’s job status immediately after graduation, referred to as “first job,” matters for his/her future job career. Using the ratio of regular employees in the labor force in the year preceding an individual’s graduation as an instrument for the first-job status (i.e., regular job or not), we confirm that even for women, whose retention rates are lower than those of men because of marriage and childbirth, individuals’ first-job status has a significant effect on their job status in the future. We further find that the effect gradually declines over the years and effectively disappears around 10 years after graduation. Finally, we find that the first-job effect is reversible: no negative effect of failing to obtain a regular job at graduation is observed if an individual can secure regular employment within a reasonable time period after graduation.

New school graduates just having completed their education and in search of a job inevitably will have to contend with the macroeconomic situation in their year of graduation. While those who graduate in a good year are likely to find a desirable job, those who graduate in a bad year may be forced to start with a less attractive job or find themselves unemployed due to a lack of employment opportunities. In Japan, the unemployment rate rose substantially during the 1990s and has remained high since then due to the prolonged period of slow growth and repeated recessions following the burst of the bubble economy. In particular the unemployment rate of the young, which increased from 4% to 10%, is causing concern, as young workers who are unable to find a good job upon graduation tend to remain jobless or work as part-time (or temporary) employees in subsequent years. This negative cohort effect impinging on those who graduated during the protracted recession, and especially during what came to be called the employment “ice age” (around 1998–2002), has become an important policy issue in contemporary Japan. Recent academic studies have found that macroeconomic conditions at labor market entry have a significant impact on individuals’ working conditions not only in their entry year but also in subsequent years. Oreopoulos et al. (2012), for example, focusing on Canadian college graduates, found substantial initial earning losses for those who graduated in a bad year, which linger on for up to around 10 years before dissipating. Similarly, Kahn (2010), using U.S. data for white male college-educated workers who graduated in a bad year, found long-run negative effects on wages as well as on occupational attainment. Several studies on other countries also find persistent cohort effects from macroeconomic conditions at graduation, including Brunner and Kuhn (2010) for Austria and Ohtake and Inoki, 1997 and Kondo, 2008, and Genda et al. (2010) for Japan.1 While the persistence of cohort effects in the labor market is well established, the mechanisms underlying these effects are not necessarily well understood. Although there are various possible theoretical explanations of the persistency of such effects, such as search costs, the accumulation of human capital through work experience, or signaling effects, that is, the stigma caused by the failure to find a job on graduation, there are few empirical studies examining the pertinence of the different explanations, since few datasets containing the necessary information on individuals’ employment history are available. Due to the limited availability of data on individuals’ entire job career history, many studies exploring the underlying mechanisms have focused on the effect of individuals’ first job obtained immediately after graduation. The reason is that if an individual’s first job obtained at graduation matters for his or her later career, the persistence of cohort effects can be attributed – at least in part – to individuals’ initial success (or lack thereof) in the job market. Moreover, it would mean that graduating in a bad year affects individuals differently rather than affecting all individuals in a particular cohort evenly, which has important implications also for policy makers. Against this background, a key study is the one by Oyer (2006), who examined the work status of economics Ph.D.s in relation to their first job obtained at graduation. Instrumenting macroeconomic conditions with the demand for economists in the year an individual graduated, he found that there is a causal link between the quality of an economist’s first job and that of his/her position 3–15 years later. Focusing on individuals’ research productivity, he further found that, for academics, getting a good first job increases publication productivity in the following 10 years. Oyer (2006) interprets the result as indicating that the first job matters in terms of the development of task-specific human capital, which affects an individual’s future career. Similar studies on individuals in Japan are relatively scarce. An exception is that by Kondo (2007), which shows that individuals’ current employment status is closely linked to their first job. Specifically, she finds that even several years after graduation the probability of being a regular employee is substantially lower if an individual failed to obtain a job as a regular employee at graduation. She suggests that a possible reason for this finding is the signaling effect: in Japan, companies rarely upgrade temporary or part-time workers to regular worker status, and potential employers cannot distinguish between lucky individuals who obtained a full-time regular job at graduation and possibly more able, but unlucky, individuals who failed to do so. There are reasons to believe that the first-job effect is likely to be more important in Japan than in other countries. Since lifetime employment is an important element of the employment system in Japan, workers are implicitly assumed to begin their job career immediately after graduation and continue to work in the same firm until they reach retirement. This means the standard route of recruitment is the recruitment of new graduates, and if an individual fails to find a desirable job at graduation, it becomes very difficult to find alternatives in subsequent years, since the failure to land a job as a regular employee at graduation is sometimes regarded as a stigma by potential employers. This conjecture is consistent with the results of a study by Genda et al. (2010) focusing on less-educated male workers in Japan and the United States, which finds persistent negative effects of the unemployment rate at graduation in Japan, while in the United States such effects are only temporary. Against this background, this paper seeks to examine in more detail how the job obtained at graduation, which is what we mean by “first job” here,2 matters for an individual’s subsequent job career in Japan, using micro-data for female individuals taken from the Japanese Panel Survey of Consumers (JPSC). While the presence of first-job effects in Japan has already been established by Kondo (2007), the structure of her dataset, i.e., pooled cross-sectional survey data from 1999 to 2002, prevented her from investigating the processes and mechanisms underlying these effects. For this study, however, we have long-run panel survey data for the period 1993–2007 covering individual female workers in Japan and including their employment history (from graduation), thus allowing us to examine the underlying mechanisms in detail. Using this dataset, which is not only considerably larger than that employed by Kondo (2007) but also covers a much longer period, including the so-called employment “ice age” (ca. 1998–2002) and more recent years, we first test whether her central findings on workers in Japan are supported. We then investigate how the first-job effect evolves with the passage of time after graduation by examining the career records of individuals collected by the JPSC. Finally, taking advantage of the long-run panel, which allows us to track the career progression of individuals following school/university graduation on an annual basis, we examine whether the first-job effect differs depending on individuals’ career path in the first few years after graduation. The findings can be summarized as follows. First, our results indicate that even for female workers, whose retention rates are lower than those of men because of marriage and childbirth, the employment status immediately after graduation matters for the employment status in subsequent years. Second, the effect gradually declines over the years and effectively disappears around 10 years after graduation. Third, we find that the first-job effect is reversible: no negative effect of failing to obtain a regular job at graduation is observed if an individual can secure regular employment within a reasonable time period after graduation; however, this by no means implies that the consequences of failing to find a job at graduation are trivial. Due to recruitment practices in Japan, which focus only on new graduates, obtaining a desirable job becomes more difficult in the years after graduation. If bad years carry on for a prolonged period – as was the case during the “lost decade” more generally and the employment “ice age” around the turn of the millennium in particular – a large number of unfortunate new graduates will lack the opportunity to dispel the unfounded stigma attached to not finding regular employment upon graduation and are therefore likely to experience negative effects throughout their career. The rest of the paper proceeds as follows. The following section describes the data used and the empirical strategy of our analysis. Next, Section 3 reports the results of our empirical analysis with regard to the three key questions we address: (1) whether the existence of a first-job effect can be confirmed; (2) how long the first-job effect persists; and (3) whether the initial effect is contingent on the career path taken by individuals in the first few years after graduation. Section 4 concludes.

Using the employment histories of female workers constructed from micro data of the Japanese Panel Survey of Consumers, this paper examined how finding a job as a regular employee at graduation matters for an individual’s future employment career. We confirm that even for female workers (whose retention rates are lower than those of men), an individual’s job status at graduation matters significantly for the future job status (i.e., regular job or not). The effect gradually declines and effectively disappears around 10 years after graduation. However, the observed first-job effect depends on the post-graduation career path taken by the individual: even if individuals failed to find a regular job at graduation, if they managed to find regular employment at least once within the first few years after graduation, their probability of being in regular employment thereafter was not notably lower than that of those who found regular employment immediately at graduation. That being said, though, because of hiring practices in Japan, those who failed to land a regular job at graduation will find it much more difficult to do so in later years unless labor market conditions improve dramatically. These empirical findings provide some clues to understanding the mechanisms underlying the negative cohort effect associated with entering the labor market during a recession and provide some indications for a policy response in prolonged recessions such as during the employment “ice age.” As for the background mechanisms, it can be argued that the burden of the negative cohort effect from a recession falls disproportionately on individuals who are unsuccessful in their first-job hunt (rather than falling evenly on all individuals in that cohort). Moreover, our finding that the first-job effect becomes irrelevant if an individual is fortunate enough to find a job as a regular employee within a reasonable time period after graduation suggests that, in the context of the Japanese employment system, the first-job effect results at least partially from the negative signal or stigma attached to having been unsuccessful in the first round of job hunting at graduation. Our findings also highlight the need for policies to assist those who lose out in the first-job hunt, especially when the economy is experiencing prolonged stagnation such as during the “lost decade” or the employment “ice age.” As it appears to be possible for individuals to make up for the negative effect caused by the bad luck of graduating when macroeconomic conditions are unfavorable if they can find a job as a regular employee within a reasonable time period, systematic policies to provide support for the employment of young workers who failed to find a (good) job in their first round of job hunting could provide substantial benefits. In this context, the recent proposal by the Japan Association of Corporate Executives (Keizai Doyukai) that recruiting firms should treat all workers that have graduated within less than 3 years as if they were new graduates is an encouraging step in the right direction.