The Why of More or Less: Evidence from Spain on Gender Segregation at the Establishment Level of the Firm

IZA Logo

IZA Seminar

Place: Schaumburg-Lippe-Str. 9, 53113 Bonn

Date: 19.09.2006, 12:15 - 13:30


Presentation by 

Kevin T. Reilly (Leeds University Business School)


This paper examines the determinants of gender composition at the establishment level of the firm. Motivation for this examination is two-fold. First, an increasing concern of policymakers about the gender biased labour market outcomes. Second, the well-known stylized fact that in many dimensions {occupation, industry and establishment} the proportion of women has a tendency to widen the gender wage gap. While much is known world-wide about the gender wage gap effects, little evidence exists about why we observe the gender biased employment outcomes. This paper will contribute to this understanding using a new matched worker-establishment data set (Wage Discrimination and Job Characteristics Survey, 2002) from Spain to identify regularities in the gender composition across 460 private sector establishments. To accomplish our goal we use as our measure of gender composition in the establishment the proportion of females among full-time employees, which is the standard measure used in the wage gap literature. We show that this distribution is positively skewed with an average proportion of females of 0.26 that is six points larger than the median and argued this distribution is characterised by a high degree of gender segregation. To examine the validity of factors that can explain this distribution we use both unconditional and conditional (Two- Limit Tobit) forms of analysis. Not surprisingly we find that the industrial structure is an important determinant in the way most researchers would expect but also that regional-cultural differences’ within Spain play no role in this distribution. There is strong evidence presented that human capital investment by the establishment has the effect of reducing the proportion of females in the establishment. This is consistent with the theory of human capital in the context of lower labour force attachment by women relative to males. The evidence also suggests that establishments that are more dynamic in their decision-making in other dimensions have significantly more females in their establishment. Taken together we argue that this evidence is strongly suggesting that both supply and demand for labour aspects must be considered in formulating policy in this area.

Download complete paper   
For more information, please contact