When does transition increase the gender wage gap?

In the context of underlying stability in female participation rates, the gender wagegap, measured by the log of monthly wages, more than doubled in Belarus from 1996 to 2006. In this respect, the country has experienced a variant of the transition which occurred in the former Soviet Union where relative female wages fell by more than female participation. We have used the Machado and Mata (2005) analysis of the gender gap distribution. This reveals that the effect of coefficients on observed characteristics in widening the gap was increasing over time, especially in the lower and middle deciles of the wage distribution. At the same time, the effect of the characteristics themselves in reducing the gap was shrinking. The decomposition of changes in the gap over time, based on Juhn et al. (1991), confirms that the contraction of women's relative wages has been caused both by a deterioration in the observed characteristics of female workers and by the associated remuneration. Changes in the residual wage distribution tend to slightly reduce the gap rather than, as is the case elsewhere, to increase it. The analysis carried out in line with Neuman and Oaxaca (2004) suggests that the increased gap was not caused by sample selection. Instead, two observed factors are found to be mainly responsible for the results: hours of work have increased for men more than for women and women have experienced segregation in low-wage industries.