Female Ownership of Firms and Regulation Experience

Abstract

The presence of gender disparity in de jure rules across the world is relatively well-known. Many studies show that this disparity is detrimental to female labor participation. Our focus is different in that we examine whether firms with females amongst ownership experience elevated time costs and burdens associated with government regulations relative to their male-owned counterparts. In this sense, we are interested in both de facto and de jure differences in governmental regulation. Using firm-level data and two alternative matching methods, our results suggest that firms with at least one female owner report that senior management spends a larger percentage of their time dealing with regulations. We also find that construction permits take approximately 4–7 days longer to obtain for these same firms. Lastly, we find that female-owned firms perceive labor regulations to be a larger obstacle to business operations. In all cases, these effects tend to be largest in countries with the most disparity in de jure rules. However, the gap remains even in the most de jure equal countries.

Keywords:

• Regulation
• discrimination
• gender
• inequality

JEL codes:

• J16
• L51

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The data and code that support the findings of this study are available from the corresponding author upon reasonable request.

Notes

1 See, for example (Bertay et al., 2020; Cuberes & Teignier, 2014; Duflo, 2012; Kleven & Landais, 2017) for reviews. Goldin (2023) provides an excellent review of the history of gender inequality within the US.

2 There is also a complementary literature on the effects of specific policies aiming to promote equality, such as the Equal Pay Act in the United States (Neumark & Stock, 2006), childcare provision (Powell, 1998), and parental leave laws (Baum, 2003).

3 Islam et al. (2019) focuses on female participation defined as management, ownership, and employment. Hallward-Driemeier and Gajigo (2015) focuses on employment in specific occupations.

4 There are some exceptions to this. For example, in a study of Viet Nam, Pham and Talavera (2018) find no evidence of discrimination against female entrepreneurs despite being a country heavily dependent on social networks.

5 This is calculated assuming management works a typical 40-hour work week.

6 We use the version of this index that is gender adjusted – that is, the version that accounts for the fact that the laws may be unequal (Fike, 2017).

7 Economic freedom is not a measure of market competition per se, but economically free countries tend to experience more entrepreneurship (Angulo-Guerrero, Pérez-Moreno, & Abad-Guerrero, 2017; Bjørnskov & Foss, 2013; Nyström, 2008), income growth (Grier & Grier, 2021), and innovation (Zhu & Zhu, 2017). Djankov, La Porta, Lopez-de Silanes, and Shleifer (2002) find that limited governments have a lighter regulation of entry for start-up firms. As such, it has been used as a proxy for market competition (Emerson, 2006) and some use the term synonymously with ‘pro-market’ institutions (e.g. Bennett & Nikolaev, 2021).

8 This theory is more applicable to discrimination in the labor market, but it is possible that reducing discrimination in one realm (i.e. employment) could lead to a reduction in discrimination overall.

9 For example, labor participation, education, and health (Fike, 2018; Grier, 2023; Stroup, 2008, Stroup, 2011).

10 This test is similar in spirit to the analysis of female entrepreneurs in Muravyev et al. (2009).

11 See Frank, Lambsdorff, and Boehm (2011), Serra and Wantchekon (2012), Swamy, Knack, Lee, and Azfar (2001), and Dollar et al. (2001) for discussions of the gender-corruption literature.

12 Most studies discussed here also examine female-managed firms but for brevity and given our focus on ownership, we summarise only the female ownership results.

13 These were absent in the Breen et al. (2017) study though they did include several country level controls for robustness.

14 The surveys sometimes have a question concerning the share of female ownership, but it is not consistently included and significantly limits our sample.

15 In South Asia, over 80% of women in non-agricultural jobs are in informal employment; in sub-Saharan Africa this number is 74%; and in Latin America and the Caribbean it is 54% – see UN Women (2015).

16 Table A1 in Appendix A provides a comprehensive list of all sectors and our reclassification.

17 Note that our estimate of employees considers only full time or ‘permanent’ employees; this ignores temporary workers. We use full time employment as this is what the World Bank uses to categorize firms according to firm size (small, medium, or large).

18 The categories are Freedom of Movement, Contracting Freedom, Property Rights, Business Freedom, Financial Rights, Legal Status, and Freedom to Work, and each category contains a series of questions such as “Can a woman open a bank account in the same way as a man?”.

19 Another concern is that the gender of the respondent could bias the responses. Clarke (2022) studies this idea with regards to corruption. However, the World Bank no longer publishes data on the gender of the respondent due to a change in the WB’s personal data policies. This was confirmed in email correspondence between one of the authors of this paper and a senior economist at the World Bank on January 23rd, 2023. Instead, the World Bank does include an indicator of perceived truthfulness of responses. We reproduce our main results using only observations coded as truthful in the Supplementary Materials file finding the results to be unchanged. Some statistical significance is lost due to a smaller sample size as this variable is not always available.

20 The categories are Freedom of Movement, Contracting Freedom, Property Rights, Business Freedom, Financial Rights, Legal Status, and Freedom to Work, and each category contains a series of questions, such as ‘Can a woman open a bank account in the same way as a man?’

21 All of such variables are obtained from third parties such as International Country Risk Guide, the Global Competitiveness Report, and the World Bank’s Doing Business project, which mitigates any potential biases of the researchers which countries are more economic free.

22 The adjustment factor is given by AdjArea2 = $\frac{\mathit{\text{Area}}2+\mathit{\text{Area}}2*\mathit{\text{GDI}}}{2}.$

23 These countries are Qatar, Saudi Arabia, United Arab Emirates, Bahrain, and Jordan.

24 We also estimate our main results using Ordinary Least Squares with analogous controls (covariates listed in Section 3.2.1 and country, time, and sector fixed effects) finding similar results. These estimates are available in our Supplemental Materials file.

25 Split sample results for all outcomes are available upon request.

26 Similar to Table 2, these results are presented for only the third nearest neighbor. Full results available in an online Supplementary Materials file.