Association for Public Policy Analysis & Management

Governments, foundations and individual donors currently spend billions of dollars subsidizing entrepreneurship training and assistance programs, but there is little rigorous evidence about the effects of many of these program on business creation or employment outcomes. The U.S. Department of Labor (DOL) has conducted several experimental evaluations of targeted initiatives to support self-employment as a reemployment strategy. These initiatives generally increased entry into self-employment, and in some cases—but not universally—they produced impacts on self-employment that were still apparent 1.5 to 3 years after study enrollment.

This paper considers a new self-employment training model for dislocated workers that we designed to address challenges facing this group identified based on past research and feedback from frontline staff. The Self-Employment Training (SET) pilot program was targeted to individuals who were unemployed or underemployed, and it included three important features:

1. A screening mechanism to identify potential participants with expertise or expertise that was related to their proposed businesses. The majority of start-ups fail within two years, but past research indicated that field-specific knowledge is generally linked to a higher likelihood of success in self-employment.
2. A range of personalized services to help SET participants navigate the early stages of entrepreneurship. This included 12 months of case management, customized training, and technical assistance.
3. Financial support to offset business start-up costs. SET participants who met business-related milestones were eligible for seed capital microgrants of up to $1,000.

The pilot was conducted in four sites—Chicago, Cleveland, Los Angeles, California, and Portland (Oregon)—between 2013 and 2017. Overall, 1,981 study enrollees were randomly assigned to either the SET program group or a control group with equal probability. Primary data sources for the evaluation were a baseline survey and an 18-month follow-up survey.

The paper includes experimental findings that draw on a report we prepared for DOL that will be released in the coming months, including the following key results:

• SET led to greater receipt of personalized assistance and seed capital, produced a sustained increase in self-employment activity, and led to modest increases in the rate of employment in any job (through self-employment or a wage/salary job).
• SET increased self-employment rates for all demographic groups, and may have been particularly effective for those without recent small business experience.
• There was no discernable impact of SET on earnings over the period covering 7 to 18 months after study enrollment.

The paper will also present new results from a nonexperimental analysis of seed capital, which is a particularly novel component of SET. Financial capital constraints are often noted as a major barrier to business creation—especially among disadvantaged groups—and SET program providers noted that the microgrants filled unmet needs. Hence, we are assessing the quantitative evidence supporting these qualitative conjectures. Our results are based on whether seed capital use was correlated with more successful outcomes, and whether there were differences in impacts across experimental subgroups that differed in their likely access to credit (for example, based on home ownership and credit history).