| Hamilton(2000) argues that the self-employed consistently earn less than comparable employed workers and concludes that most workers who choose to try self-employment must do so because they receive substantial non-pecuniary benefits, such as "being your own boss." However, Hamilton's analysis assumes that a worker's entry into self-employment is irreversible. In chapter 1, I treat this decision as reversible and compare the lifetime income of those who try self-employment to those who do not. I find that once I incorporate sectoral mobility and hours into the agent's decision problem, a majority of those who try self-employment fare better than those who do not. I explicitly control for selection by employing nearest neighbor matching and use PSID data to study the realized income paths of self-employed individuals and their "twins."A natural question is whether the experiences of women are similar: do women, on average, earn a premium for trying self-employment, or on the contrary, do women pay a premium for non-pecuniary benefits such as increased flexibility to choose work hours? Using the methods described in Chapter 1, I find that on average women who try self-employment sacrifice income for doing so, while working more hours, during the first 5 years. However, mothers of small children who try self-employment earn a premium for doing so, in part due to savings in child care costs. Interestingly, these mothers work at least as many hours, suggesting the availability of additional scheduling flexibility in the self-employment sector.The ability to smooth consumption through investment in a risky asset is important if the sectoral covariances of income with risky assets differ, which Rosen & Willen(2002) found to be the case with PSID data. In chapter 3, I corroborate their result using PSID data and the method from chapter 1, but find that this covariance depends upon the income measure used. When I instead use SIPP data from 1990 through 2004, I do not find strong evidence of a difference between the covariance of risky assets and incomes across sectors. However, there is evidence to suggest that the self-employed more actively invest in risky assets. |
