Business creation is economically important, and unemployment precedes the creation of a substantial share of new firms. Yet, most research has focused on analyzing the effects of unemployment insurance policies on re-employment outcomes, ignoring self-employment. In this paper, we analyze how the potential duration of unemployment benefits, a fundamental design choice of unemployment insurance systems, affects whether new firms are founded out of opportunity or necessity and their growth potential. To this end, we construct a comprehensive dataset on German firm founders that links administrative social insurance information with business survey data. Exploiting reform and age-related exogenous variation in the potential duration of unemployment benefits, we find that longer potential benefit duration implies longer actual unemployment and, as a consequence, more necessity entrepreneurship and worse startup outcomes in terms of sales and employment growth. We explain this overall effect of potential benefit duration through a mix of compositional and individual-level duration effects. Our findings underline that new firms started out of unemployment are a highly heterogeneous group and suggest that the (optimal) design of unemployment insurance systems has important externalities on whether innovation- and growth-oriented firms are started out of unemployment.

企业创建在经济上很重要，而失业率先于新公司的创建。然而，大多数研究都集中在分析失业保险政策对再就业结果的影响上，而忽略了自营职业。在本文中，我们分析了失业救济金的潜在期限（失业保险制度的基本设计选择）如何影响新公司是出于机会还是必要性及其增长潜力而成立的。为此，我们构建了一个关于德国公司创始人的综合数据集，将行政社会保险信息与商业调查数据联系起来。利用改革和与年龄相关的失业救济金潜在期限的外生变化，我们发现，更长的潜在福利期限意味着更长的实际失业时间，因此，在销售和就业增长方面，企业家精神更加必要，创业结果更差。我们通过组合和个人层面的持续时间效应来解释潜在收益持续时间的总体影响。我们的研究结果强调，从失业开始的新公司是一个高度异质的群体，并表明失业保险制度的（最佳）设计对创新和增长型公司是否从失业开始具有重要的外部性。