Measuring the occupational segregation of males and females in Pakistan in a multigroup context

Occupational segregation, the unequal distribution of men and women across occupations, is a major contributor to gender disparities in labor market outcomes, impacting pay gaps, job quality, and employment. Economic theories like compensating differentials (women prioritizing non-pecuniary benefits) and crowding (lower wages due to female concentration in limited occupations) attempt to explain this. The devaluation theory offers a sociological perspective, suggesting lower wages for women's work due to societal norms and undervaluation. In Pakistan, a patriarchal society with low female labor force participation (22% compared to 84% for men), gender differences across sectors and occupations are stark. This study aims to provide a detailed analysis of occupational segregation in Pakistan, going beyond previous studies that only analyzed overall gender segregation, by examining several subgroups based on age, education, work type, and organization type, using the local segregation measures of Alonso-Villar and Del Río (2010).

The literature defines equality as uniform distribution of groups across occupations. Several methods exist for measuring occupational segregation, most suited to binary groups (male/female). However, indexes for multigroup segregation have emerged to account for diverse societal structures. Reardon and Firebaugh (2002) and Del Río and Alonso-Villar (2010) offered multigroup indexes, but Del Río and Alonso-Villar's local segregation indexes allow more detailed analysis of specific subgroups. Previous studies using various indexes (e.g., index of dissimilarity, Karmel MacLachlan index) have highlighted high occupational segregation across countries, showing its impact on female workers' pay and employment prospects. Studies on India also revealed high occupational segregation by gender and social groups. However, research on occupational gender segregation in Pakistan remains limited.

This study utilizes three waves (2013-2018) of the Pakistan Labor Force Survey's microdata on the civilian labor force aged 15-65, focusing on paid employees. The local segregation measures proposed by Alonso-Villar and Del Río (2010) are employed. These measures (Gini index variant G^s, generalized entropy index Φ^α, and dissimilarity index variant D^s) quantify the segregation of a target group (e.g., females) by comparing its occupational distribution to the overall distribution. These indexes range from 0 (no segregation) to 1 (total segregation). Overall segregation indices (mutual information index M, multigroup index of dissimilarity I^ps, multigroup Gini index G) are also calculated. The study analyzes occupational segregation by gender and several subgroups: age (young, middle-aged, elderly), education level (low, high), organization type (public, private), and hours of work (full-time, part-time). Segregation curves visually represent the cumulative employment proportion against the cumulative proportion of the target group.

The study finds substantial occupational segregation in Pakistan. Female segregation constitutes a large portion of overall gender segregation (83% based on the mutual information index M), despite females comprising only 14% of the total workforce. Local segregation indexes consistently show significantly higher segregation for females than males across all subgroups. Elderly workers of both genders exhibit higher segregation. Higher education doesn't reduce segregation; in fact, highly educated females show the highest segregation levels. In terms of region, females in rural areas have higher segregation than urban areas. Segregation is also higher for females in the public sector than the private sector. Finally, females working part-time show higher segregation than those working full-time. The age group analysis showed higher segregation for those above 45 years of age. Notably, elderly female workers had the highest level of segregation.

The findings confirm the presence of substantial occupational gender segregation in Pakistan. The high contribution of female segregation to the overall gap, despite low female labor force participation, highlights the systemic nature of the problem. The lack of impact of education on reducing segregation challenges the assumption that human capital investment alone can solve the issue. The results partially support both compensating differentials theory (females favoring certain work conditions) and devaluation theory (societal undervaluation of women's work). The observed differences in segregation across age groups, regions, and organization types indicate the interplay of various factors shaping occupational choices. The high segregation of highly educated women is a particularly striking finding.

This study provides detailed insights into occupational gender segregation in Pakistan, using local segregation measures. It highlights the disproportionate impact on women, the limited effect of education, and the role of societal norms. Further research could investigate the formal vs. informal sector divide and incorporate vertical segregation (hierarchical position within occupations). Policy recommendations include investment in education, promoting gender-friendly workplaces, and addressing societal gender norms.

The study is limited to the formal sector workforce. It focuses on horizontal segregation and doesn't analyze vertical segregation due to data limitations. Future studies should consider these aspects and examine the impact of other factors not included in this study.