Gender Wage Inequality

1118 Words3 Pages

Shilpa Bista
Conceptual framework: Gender earnings gap in Tajikistan

To estimate the gender wage differential, this paper uses a standard approach, the Oaxaca (1974) decomposition, which entails decomposing the average male and female earnings differentials into two components. One component is the “explained” effect on the wage differential that occurs due to an average difference in human capital characteristics between men and women receiving equal treatment, and the second is the “effect of discrimination,” which indicates the “coefficient effect” on the wage differential that occurs either due to different treatment of men and women given that they possess the same individual characteristics, or due to differences that cannot be explained by the regression model (Oaxaca 1973).
The earnings equation for men and women is based on the Mincer (1974) human capital earnings function: ln⁡〖〖(W〗_i)=Z_i^' β+ μ_i 〗, i = 1,…,n where W_i is hourly wage rate of worker i, β is a vector of coefficients, μ_i is a disturbance term and Z_i^' is vector of individual characteristics such as education, experience, gender, as well as occupational, industrial, regional and economic indicators (Oaxaca 1974, 695; Bishop et al. 2005). Average natural logarithm earnings are calculated separately for male and female as functions of their individual characteristics.
The first step is defining a measure of discrimination, which Oaxaca (1974) expresses as follows:
D=(W_m/W_f -〖(W_m/W_f )〗^°)/〖(W_m/W_f )〗^° where D is the discrimination coefficient, W_m/W_f is the “observed male-female wage ratio” and 〖(W_m/W_f )〗^° is the “male-female wage ratio in the absence of discrimination” (694). Oaxaca’s (1974) measure of discrimination assumes that in the...

... middle of paper ...

...h, service activities (including hotel and restaurants and recreational services), government, transportation, and other activities (such as auto sales, computer-related activities, and research and development).
Both education and experience are assumed to increase earnings. For education, dummy variables are created to show three levels of educational attainment: (1) a person who has completed a higher education degree is considered well educated; (2) a person who has received any type of secondary degree is labeled as “fairly educated;” and those who have neither are “not educated.” Work experience is a function of years of education and age. Due to missing data on “actual number of years of work experience,” Oaxaca (1974) uses potential experience which is defined as: potential experience = (age) – (number of years of schooling completed) – 6 (697).

More about Gender Wage Inequality

Open Document