The Glass Ceiling Puzzle, Legal Institutions, and the Shadow Economy

ABSTRACT

Scholarly studies in economics, sociology, psychology, and management emphasize the low number of women in management as one of the main indicators of gender discrimination in the labor market. This study investigates the differences in the percentage of women in senior management across forty-five countries. The results of the regression analysis show that women are more represented in senior management in developing countries than in so-called “liberal Western democracies.” Women also participate more in senior management in countries in which prejudice and discrimination against women are greater. The study presents empirical evidence for two economic explanations for these puzzling results: the weak functioning of the legal system and the large size of the shadow economy.

KEYWORDS:

• Gender inequality
• glass ceiling
• legal institutions
• shadow economy

JEL Codes:

• J71
• O17

SUPPLEMENTAL DATA

Supplemental data for this article can be accessed at https://doi.org/10.1080/13545701.2018.1456672.

Notes

1 Data are available for twenty-four out of the forty-five countries for 2004. The other countries are included in the survey in later years.

2 The group of developing countries includes: Argentina, Armenia, Botswana, Brazil, Chile, China, Georgia, India, Indonesia, Latvia, Lithuania, Malaysia, Mexico, Peru, Philippines, Russia, South Africa, Thailand, Turkey, and Vietnam.

3 Countries that I categorize as “liberal Western democracies” include: Australia, Belgium, Canada, Denmark, Finland, France, Germany, Greece, Ireland, Italy, the Netherlands, New Zealand, Norway, Spain, Sweden, Switzerland, the United Kingdom, and the US.

4 Roland Inglehart and Pippa Norris's (2005) factor analysis suggests that the scale created with the WVS provides reliable indicators of cultural attitudes toward gender equality.

5 The reported results of diagnostic tests for normality reveal that the null hypothesis of normal skewness or kurtosis cannot be rejected at standard levels of statistical significance. In all cases where the Breusch–Pagan test suggests that the null of homoscedasticly distributed errors can be rejected at the 5 percent level of significance, we report White's standard errors. Outlier-robust estimations (available upon request) show that the reported results are not driven by outliers.

6 Unfortunately, the above-described variables for the availability of flexible employment arrangements and the availability of mentoring are available for only the last year in the sample. Hence, it is plausible to expect that the explanatory power of these variables in my model would be limited. I acknowledge this limitation and report that coefficients on other variables remain almost the same when variables for flexible employment arrangements and mentoring are not included in the model (available upon request).

7 The reported diagnostic tests suggest that the estimated models are well-specified with respect to normality, heteroscedasticity, and multicollinearity. The outlier-robust estimation (available upon request) suggests that the reported results are not driven by outliers.

8 Data are in constant 2011 international USD. The results are similar when data for the enrollment rates in primary schools, enrollment rates in secondary schools, or life expectancies are compared.

Additional information

Notes on contributors

Bruno Ćorić

Bruno Ćorić is Assistant Professor at the University of Split, Faculty of Economics, Split, Croatia. He received his BA in Economics from the University of Split in 2001, and his MA and PhD from Staffordshire University Business School, UK, in 2002 and 2008, respectively. Currently, he teaches macroeconomics at the undergraduate level and macroeconomic management at the graduate level. His major research interests include business cycles, open economy macroeconomics, and economic growth. He has published in Oxford Bulletin of Economics and Statistics, International Review of Economics, and Finance, Applied Economics, and Applied Economics Letters.