Bribe or die: gender differences in entrepreneurship in emerging markets

ABSTRACT

This paper examines the relation between gender differences in entrepreneurship and firm-level bribery, one of the most impactful business obstacles to private sector growth. Using data from a comprehensive survey covering 16,560 enterprises in 32 emerging economies, we find that female-led firms account for 19.2% of all enterprises, which is approximately 4 times lower than the share of male-led firms, and that female entrepreneurs experience a higher level of bribery than their peers. The impact of gender on bribery is more pronounced among firms located in countries that are more corruption-prone and becomes weaker among countries with a higher female graduation ratio at the tertiary level. In addition, female-run firms have a lower likelihood of obtaining a construction permit, securing a government contract, or holding an operating licence. Overall, the results suggest that a bribe-to-survive motive is a possible explanation for the higher level of bribery among female-led firms in emerging markets.

KEYWORDS:

• Women entrepreneurs
• bribery
• time cost
• gender differences
• emerging markets

JEL CLASSIFICATION:

• D23
• D73
• J16

Acknowledgments

We are grateful to the Editor, Professor David Peel, and to the anonymous referees for many helpful comments and suggestions. We thank Dr. Mai Hong Phan for consistent support and excellent feedback that has significantly improved the paper. This research is funded by National Economics University, Hanoi, Vietnam [Grant number 154/QÐ‐ÐHKTQD]. The usual disclaimers apply. Email addresses: [email protected] (Lan Thanh Nguyen), [email protected] (Hung T. Nguyen), [email protected] (Mia Hang Pham).

Data availability statement

Data are available from the data sources identified in the paper.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ A recent survey conducted by the Harris Poll suggests that a majority of Americans prefer to work for a female-led company as these firms are more likely to offer equal pay and include access to childcare. The poll survey results are available at: https://ml.globenewswire.com/Resource/Download/fec25583-ecb6-45f1-8b89-f9c40610065e (retrieved on 20 May 2020).

² The nexus between corruption and economic growth has been of interest for both empirical and experimental studies’ analysis in recent decades, and these findings so far are mixed. On one hand, corruption is considered as a necessary investment to ‘grease the wheels’ and bring firms advantages in operating and doing business more successfully (Wang and You 2012). On the other hand, corruption might ‘sand the wheels’ through eroding incentives for other investment activities and innovation, thus hampering firm growth and development (Beekman, Bulte, and Nillesen 2013).

³ As Hao, Chang, and Sun (2018) note, the inconsistent results on the relation between gender and corruption could be driven by a number of factors, including a non-representative sample, critical flaws in methodology, or biased or incredible results due to publication biases.

⁴ Using survey data from the Characteristics of Business Owners, Boden and Nucci (2000) find that female-owned firms have lower survival rates than their male-led counterparts. Similarly, Watson (2003) conducts a study of Australian small and medium enterprises and shows a higher probability of failure for female-controlled enterprises when excluding industry effects. The author notes that this difference becomes insignificant when industry effects are considered.

⁵ Let $\mathrm{\Delta }COR{R}_i$ denote the difference between corruption level with the treatment (female-run) and that without the treatment (male-run), given by $\mathrm{\Delta }COR{R}_i=E\left(COR{R}_{i1}|FEMAL{E}_i=1\right)-E(COR{R}_{i0}|FEMAL{E}_i=0)$.

Average treatment effect in population (ATT) is given as $ATT=E\left(CORR\right)=E\left(COR{R}_1-COR{R}_0\right)$

Average treatment effect on the treated (ATE) is given as $ATE=E(COR{R}_1-COR{R}_0|FEMALE=1)$

⁶ In untabulated results, we find no statistical difference between the characteristics of the treatment and control firms after the matching, which confirms the quality of the match.

⁷ Women entrepreneurs tend to have less networking than their male peers. As Singh, Vinnicombe, and Kumra (2006, 459) note, ‘networking refers to activities by individuals attempting to develop and maintain relationships with those with, or perceived to have, the potential to assist them in their work or career’. Women are shown to be limited in access to important organizational contacts (Jensen 2014). Similarly, Pham and Talavera (2018) show that gender stereotypes and constraints to network access prevent women from running their businesses effectively. Apparently, connections are helpful in forming relationships with other entities, which may then bring advantages for entrepreneurs’ careers (Forret and Dougherty 2004).

⁸ These findings are consistent with Robb (2002) and Klapper and Parker (2011), which show that female-owned enterprises typically underperform compared to male-owned firms, evidenced by their lower levels of sales, profits, and employment. In addition, Sabarwal and Terrell (2008) find that the small scale is a key factor to explain the lower profits of female-run firms in Eastern Europe and Central Asia. Similarly, Morris et al. (2006) suggest that not only do women entrepreneurs run smaller firms than their male counterparts but they also heavily remain in sectors with slower growth rates, such as services or retail.

⁹ We thank the referee for suggesting this analysis.

¹⁰ We further perform the regressions analyses on the association between female entrepreneurs and firm-level bribery measured by the bribe probability and the bribe amount. In untabulated results, we find a positive relation between the FEMALE dummy and the probability of bribery and the bribes amount, suggesting that female entrepreneurs tend to perceive a higher probability and level of informal payments than their male counterparts. The coefficients are, however, not statistically significant at the 10% level, which could be due to considerable missing data from non-responses to that question (e.g. missing data occupies approximately 20% of the full sample). As Wang and You (2012) note, responses to questions regarding bribes or informal payments are likely to be subject to self-reported biases due to undocumented, or even hidden, information, leading to inconsistency and potential measurement errors in empirical analyses. Nevertheless, we still observe some evidence of the positive relation between female entrepreneurs and firm-level bribery when adopting alternative measures of bribery.

¹¹ We thank the referee for suggesting this test.

Additional information

Funding

This work was funded by the National Economics University Research Grant [154/QÐ‐ÐHKTQD].