تجزیه و تحلیل سطح خود اشتغالی در طول چرخه زندگی

In this paper, we estimate the incidence of self-employment over a person’s life-cycle across different socio-economic groups and show to what extent self-employment rates differ across groups. The analysis utilizes data from the March supplements of the Current Population Survey. This paper shows that the probability of self-employment is increasing with age and education and is higher for men, whites, and married women compared to other groups. Females are less likely to be self-employed and the difference appears to widen in absolute terms over the life-cycle, but is largest in relative terms early in the life-cycle. We show that the gender gap is not due to marriage and the presence of children. The difference between an African–American male and the (white) benchmark is dramatic, particularly in a person’s middle age. In contrast, the difference between a high-school graduate and the (college educated) benchmark is relatively small and changes sign over the life-cycle. Young (age 34 or lower) high school graduates are more likely to be self-employed than otherwise identical college graduates, while the reverse is true for older individuals. The paper discusses explanations for these life-cycle profiles of self-employment.

Small firms and entrepreneurial activities are playing an increasingly important role in the economy. Self-employed workers comprise approximately 10% of the labor force, operate a large fraction of businesses, and provide jobs for one-tenth of all wage workers.1 However, the propensity for entrepreneurial activity varies greatly over the population, with women, blacks, and lesser educated individuals being less likely to be self-employed than others. Identifying the life-cycle pattern of these differences is important for increasing our understanding of this sector and for guiding public policy. Numerous studies have investigated how the incidence of self-employment varies across demographic groups and how the likelihood of becoming an entrepreneur depends on a number of socio-economic variables. The majority of these studies tend to look at the proportion of people who are entrepreneurs at any given point in time and documents that some groups, e.g., white males or older educated individuals, are more likely to be self-employed than other groups.2 A few of these studies also measure the extent of turn-over in the self-employment group.3 Finally, Georgellis and Wall, 2005 and Taniguchi, 2002, Lin, Picot, and Compton (2000), Fairlie (1999) and Kuhn and Schuetze (2001) investigate the relative contribution of entry and exit rates in the self-employment propensities of different groups.4 However, the above literature does not document to what extent self-employment rates differ across socio-economic groups over the life-cycle. This paper fills this gap by investigating the life-cycle variation of differences in self-employment rates using data from the Current Population Survey. We adopt a logit regression in which the self-employment levels are a function of employee characteristics and the state of the economy. We distinguish three employment states: wage/salary employment, self-employment, and non-employment, and use them to construct the self-employment rate conditional on employment, and estimate the probability that an individual of any given set of characteristics is self-employed at any given age. We are able to determine how the self-employment probabilities vary over a person’s life-cycle, for example, whether differences between groups are more pronounced in the early, rather than the late, years of their life. In other words, we investigate whether groups, such as white males, have persistently higher probabilities of self-employment over the life-cycle than other groups, or whether the relative self-employment rates change over the life-cycle. This paper concentrates on differences in self-employment rates due to gender, race, and educational attainment. We find that the probability of self-employment is increasing with age and education and is higher for men, whites, and married women compared to other groups. Moreover, these inter-group differences in self-employment rates vary dramatically over the life-cycle. With regards to gender differences, not all of the differences in self-employment rates between men and women can be attributed to the marriage and the presence of children. The effect of marital status and the presence of children in the household impacts the two genders differentially and in ways that are consistent with the traditional division of labor in a household. However, these two factors tend to increase, rather than decrease, the propensity of women to be self-employed, conditional on employment. We confirm the low propensity of self-employment of African–Americans compared to that of Whites.5 The relative gap declines somewhat for older individuals: by age 60, the proportion of non-blacks who are self-employed is only twice that of blacks. The low self-employment rates of African–Americans might be driven by two factors: (i) the disparity in financial wealth of Blacks and Whites and (ii) the lower incidence of self-employment by the parents of African–Americans. Young African–Americans are severely liquidity constrained relative to young whites and therefore are not likely to pursue self-employment (see Fairlie & Meyer, 1996). As those black Americans who desire self-employment accumulate savings over their lifetime, their proclivity to be self-employed approaches that of other Americans. In addition, by the time people reach middle-age, the impact of parental background on the propensity to enter self-employment is likely to wane. Finally, we confirm that better educated individuals are more likely to be self-employed. This finding is consistent with the Lucas (1978) model in which human capital enhances an individual’s managerial ability and, hence, increases the propensity to be self-employed. It is also consistent with the empirical research of Rees and Shah, 1986 and Fujii and Hawley, 1991, and Bates (1990). We find, however, that any education effects become apparent only for middle-aged and older individuals: for young individuals under the age of 34, the incidence of self-employment is the same regardless of educational attainment. We believe that this pattern arises from differential accumulation of human capital by the two groups. In the early stages of their careers, college graduates enter jobs that allow them to accumulate human capital. In turn, this capital allows them to enter into types of self-employment that they would not have been able to pursue in the absence of prior work experience. In contrast, high school graduates enter jobs which provide a lesser amount of human capital. Thus, their possibilities for self-employment do not expand at the same rate that of college graduates. This paper is organized as follows. Section 2 briefly describes the Current Population Survey and the construction of our data set. Section 3 presents the main results. The paper concludes with some remarks on the significance of the findings.

This paper shows a detailed analysis of the self-employment probabilities over the life-cycle and the impact of various socio-economic factors on these probabilities. The above analysis reveals that there are substantial differences across socio-economic groups in the likelihood that a person is an entrepreneur and that these differences are not, in general, constant over the life-cycle. The finding that self-employment propensities not only differ among demographic groups, but also vary differentially across groups over the life-cycle suggests a number of possibilities about the causes of these inter-group differences in self-employment rates. One possibility is that the causes of intergroup differences have differential impact over the life-cycle. Another possibility is that these causes are only present at some age groups. Of course, both possibilities may be present to some extent. A more detailed investigation of this issue is beyond this study and would necessitate a analysis of flows into and out of self-employment and how these flows vary over the life-cycle.