Abstract

How does regulation affect entrepreneurship outcomes? We examine the effect of two regulatory policy mechanisms—costs and procedures—on entrepreneurship quality and quantity. Based on the national systems of entrepreneurship perspective, we apply public interest and public choice theories to hypothesize how regulatory costs and regulatory procedures can affect entrepreneurship quality and quantity differently. Using a multi-level approach, we test the direction, size, and shape of these effects with data on 51,330 innovation-oriented entrepreneurs (reflecting quality) and 871,241 entrepreneurs who started new ventures (reflecting quantity) across 76 countries during 2008–۲۰۱۷٫ We find that regulatory procedures in a country often have an inverted U-shaped relationship with entrepreneurship quality, suggesting that both too few and too many procedures might be detrimental when policymakers target innovation. We find that regulatory costs tend to have negative or inverted U-shaped effects on entrepreneurship quality and quantity. Our findings show that the way regulations are administered—by imposing financial costs or administrative requirements—is a boundary condition for entrepreneurship that affects the overall quantity of entrepreneurship and the innovation-centered quality of entrepreneurship.

Introduction

There is growing interest in how policies may better target specific goals (World Bank, 2023) and which levers can be adjusted to induce a particular entrepreneurship effect. Scholars and policymakers increasingly see regulation as a tool to influence entrepreneurship, and in some cases to target activities such as innovation, firm growth, and exporting. How the regulatory environment affects entrepreneurs is a complicated question. For example, should regulators simplify regulations or reduce filing fees for mandatory forms if their goal is to encourage more entrepreneurship in general? What if the goal is to achieve more innovation? Some policies favor prioritizing the “quality” of entrepreneurship, such as job creation, new products development, opportunity exploration, and innovation, versus generally a higher volume of entrepreneurship. Other perspectives consider whether the role of public policy is to focus on knowledge dissemination and creation via entrepreneurship. For example, will innovation be undersupplied without public support (see Mazzucato and Perez, 2015)?

The economic effects of differences in regulation on entrepreneurship are often not well understood (Audretsch et al., 2019; Fritsch et al., 2021). Entrepreneurship and institutions research examines which regulatory domains matter, such as tax policy, property registration, entry regulation, bankruptcy laws, and export permitting (e.g., Autio et al., 2014; Busenitz et al., 2000; Estrin et al., 2016, Estrin et al., 2019; Klapper et al., 2006), but how the effects are induced needs more attention (Audretsch et al., 2022). Findings on the direction and size of key relationships can be inconclusive (Stenholm et al., 2013; Estrin et al., 2013a), with some recent studies point to non-linear relationships (Braunerhjelm et al., 2021; Braunerhjelm, 2022; Audretsch et al., 2019). This could stem from the heterogeneity of both regulation and entrepreneurship; regulation varies widely in type and nature of change, and entrepreneurship can take many forms (see Klapper and Love, 2016; Braunerhjelm et al., 2015, Braunerhjelm et al., 2021; Chowdhury et al., 2019; Colombelli et al., 2016). Prior work often examined the rate of entrepreneurship activity, such as events, entry, and flows (e.g., Parker, 2009; Reynolds et al., 2002), and the research lacks consensus about determinants, including regulation, of entrepreneurship quality and quantity (see Cole et al., 2016; Colombelli et al., 2016; Sobel, 2008).

Results

We use a six-step approach to test our hypotheses. First, we test the significance of the country-year group variance for the dependent variable by excluding individual and country-year level independent variables and controls (i.e., the null random model) (specification 1, Table 4, Table 5) to justify the use of multilevel models (Rabe-Hesketh et al., 2005). Second, we add individual predictors and year and industry controls to test the effects of entrepreneurs’ internal resources and capabilities on change in the likelihood of engaging in innovation and new entry. Third, in addition to our individual variables, we include country-year explanatory variables in levels (specification 3, Table 4, Table 5). Fourth, we add country-year explanatory variables for regulatory costs and procedures in levels, together with the quadratic term, to establish our baseline specification and to test H1 and H2 (specification 4, Table 4, Table 5). Fifth, we add country-level control variables that predict how other socioeconomic conditions may affect entrepreneurship quality and quantity in a country (specification 5, Table 4, Table 5). Finally, we use individual-level controls, along with one-year lagged linear and quadratic terms of regulatory cost and procedures and other country-level variables (one year lagged), to test the role of regulation in shaping entrepreneurship.