Modeling the influence of social factors on the dynamics of gray wages

Abstract

The problem of whitewashing wages of a significant share of Russians escalated during the pandemic, post-covid recovery and, especially, against the background of geopolitical shocks. The unprecedented expansion of state support measures provided targeted and officially has
prompted entrepreneurs and citizens to get out of the shadows. Despite the obvious positive changes, the problem of salaries “in envelopes” has not eradicated itself: in modern publications, scientists return to it repeatedly. When assessing this socio-economic phenomenon,
science relies mainly on methods of regression, panel analysis, instrumental variables, etc. In this work to assess the influence of social factors on the dynamics of gray wages we used the gradient boosting method, linear modeling, random forest and a naive Bayesian classifier. As
the initial information, we used the results of the surveys of the Russian Monitoring of the Economic Situation and Health of the Population by Higher School of Economics, which are freely available. The result of the simulation is the dependence of the receiving wages method (official, “in an envelope”) on a number of factors. The organization’s industry affiliation, size, ownership, employee’s education, duration of vacation, satisfaction with professional growth
and working conditions have the strongest influence. The applied value of the results obtained is the possibility of generating control effects, both at the state level and at the level of business entities in the direction of leveling the identified reasons for receiving gray salaries and reducing the scale of the hidden wage fund.

Keywords:

salary, gray salary, gradient booster, linear models, random forest