Пространственные аспекты оценки кривой заработной платы в России

Журнал Новой экономической ассоциации. 2021. Т. 51. № 3. С. 69–101.

Demidova O., Тимофеева Е. А.

The wage curve is traditionally defined as the negative relationship between wages and unemployment rates (taking into account various control variables). It was empirically shown that the wage curve exists in some countries, including Russia. However, usually in such studies with data for the Russian regions, the mutual influence of Russian regions is not taken into account. This could create the problem of omitted variable bias. In this paper, we took into account the corresponding impact using spatial-econometric models and gave a detailed interpretation of the results. The estimate of the parameter reflecting the effect of unemployment on wages in the model without spatial effects is almost twice the corresponding estimate in the models where these effects are taken into account. According to panel data for 2005–2018 for 81 regions we estimated the partial marginal effects of changes in unemployment in one region on wages in the other regions. Similar calculations were made for the other variables. Using partial marginal effects, we found for each region: 1) the regions most affected by this one, 2) regions, changes in which will most likely affect the selected one. This is important, for example, for assessing consequences of government programs etc.

Keywords: безработица, unemployment, пространственные модели, предельные эффекты, spatial models, кривая заработной платы, wage curve, marginal effects