Skill Structure of Demand for Migrants in Russia: Evidence from Administrative Data
Authors: Simon Commander (IE Business School, EBRD and Altura Partners) and Irina Denisova (CEFIR, NES).
Using Russian Ministry of Labor administrative data for all legal migrant applications in 2010 and matching the migrant to the sponsoring firm, we find that there is some – albeit limited – evidence of firms using migrants to address high skill shortages. However, the overwhelming majority of migrants are skilled or unskilled workers rather than qualified professionals; a reflection of the low underlying rates of innovation and associated demand for high skill jobs.
Migration policy continues to be a priority in Russian economic policy. This is driven both by a demand for labor – given the unfavorable demographic trends of the last decades – and the easily available supply from the CIS countries. It is still not clear, however, what is the skills structure of the demand for migrants. Relatively new administrative data on demanded permissions to employ migrants sheds however some light on the issue.
In particular, we use the 2010 nationwide dataset ‘Job positions filled by migrants’ published by the Russia Federal Employment Service. The dataset gives detailed information on the applications for permits for migrants, including the 4-digit occupation, firm ID and the offered wage. The Federal Employment Service’s role is to approve or reject an application. In almost all cases documented in this dataset, approval was granted. Moreover, in 99% of the cases, the duration of the permitted contract was one year.
The data allow us to study the skill composition of demand for migrants from the legal sector, with the sizeable illegal labor migration staying beyond the scope of the study. The total number of applications for all of Russia in 2010 was just over 890,000, of which nearly 250,000 or 28% originated from firms in Moscow. The analysis below uses the permission data for the 21 most developed Russian regions (a full version of the paper is available as Commander and Denisova, 2012).
A breakdown of the number of requests in 2010 by skill type using the one-digit ISCO-88 classification (Managers, High-level professionals, Mid-level professionals, Service worker, Skilled agricultural workers, Craft and trades workers, Plant and machine operators, Unskilled workers) shows that over 70% of the requests were for skilled and unskilled workers. At the same time, about 17% of the total migration requests were for higher-level professionals (7%) and managers (10%). Among managers, nearly nine out of ten requests were for production or department managers with no more than 12% of managerial migration requests being for top-level executives. Among the category of high-level professionals, architects and engineers accounted for over two-fifths of requests.
Is the situation any different in the main urban labor markets? In Moscow a lower proportion – around two thirds of the migrant applications – were for skilled and unskilled workers. The starkest difference was that professionals working in IT accounted for a minute share of total high-level skill applications in Russia, but nearly 9% in Moscow. Thus, while there are some differences in the migration profile between Moscow and the rest of the country, the broad picture that emerges is one where migration policy and practice seem to be responding mainly to the apparent bottlenecks at the lower-skill end of the labor market.
Legal requests for migrants are massively dominated by requests concerning low-skill groups; and illegal migrants, as shown by anecdotal evidence, are mainly low skilled. At the same time, there is a sizeable demand for qualified migrants, managers and professionals. There are two potential motives to issuing a request for a qualified migrant: to economize on the costs of labor by substituting a local laborer with a migrant; or to fill in the gap of the scarce qualification/skills hardly available domestically. The two motives could be distinguished by looking at the wage offers associated with the posted positions and comparing them with wages paid in comparable occupations in the same region. The aim of the exercise is to see – particularly within the categories of higher-skilled applicants – whether they command any wage premium that might reflect their scarcity value.
Figures 1-2 plot the reported (relative) wage offers for two migrant skill categories: Department Managers (ISCO code=122) and Computing Professionals (ISCO code=213). The figures depict distributions of relative (to the region average) wage in logs, thus implying that the points around 0 are the wage offers at the level of regional average, above 0 means positive wage premium, and below 0 means negative wage premiums (economizing on the costs). Each figure also gives the mean search wage from the EBRD survey of recruiting agencies in 2010 (relative to the regional average).
Figure 1. Relative Wage Distribution, Production and Operation Department Managers (ISCO-88 Code: 122) Source: Authors’ calculations based on Rostrud 2010It is clear from Figure 1 that the wage offers for migrants do not identify any clear positive selection effect, in that migrants’ wages mostly fall below the survey search mean comparators. In the majority of cases, the offered wages also fall below the regional average wage thus implying that the motive is to substitute for cheaper labor.
The demand for migrants with skills of IT professionals is more complicated: there are those who offer wages below regional average, but there is also a large group of those ready to pay a wage premium to attract migrants (with log wage above zero). The search through recruiting agencies (the survey wage) would still require offering higher wages.
Figure 2. Relative Wage Distribution, IT Professionals (ISCO-88 Code: 213) Source: Authors’ calculations based on Rostrud 2010For further analysis, the migration dataset was mapped to the ORBIS (a dataset assembled by Bureau van Dijk) firm observations using the unique national tax identification code (so called INN). The ORBIS data includes information on firms’ balance sheets and simple performance data such as output per employee.
When looking only at demand from firms that lie in the top 10-20% of the productivity distribution (productivity is calculated as output per worker in the narrowly defined industry), the picture looks somewhat different: wage offers tend to lie above the average (Figure 3). It is likely that the most productive firms tend to offer wages higher than both regional average for the occupation and the survey-based search wages. This implies that the scarcity of skills on the domestic labor market is one of the more important motives behind the demand for migrants from high-productivity firms.
Figure 3. Relative Wage Distribution, Production and Operation Department Managers (ISCO-88 Code: 122), 10% Most Productive Firms Source: Authors’ calculations based on Rostrud 2010 and Orbis-RoslanaTo control for other firm characteristics, we run regressions relating the relative wage of a migrant to a set of firm and region characteristics, including measures of size and ownership, a measure of recent growth in the region, as well as the level and change in foreign direct investment in a given region since 2007. We also control for the tightness of the local labor market, using a measure of search wages raised in the EBRD survey compared to average wages in a region. The estimates are run with and without region, industry and occupation controls. The results show that relatively high wages tend to be associated with large and/or foreign-owned firms. Growth in a region or the level of FDI per capita are not systematically associated with the relative wage once controls enter the regression, suggesting that the relative wage is largely determined by firm-level features. The measure of labor market tightness enters positively but is insignificant whencontrolling for industry, region and occupation.
Overall, the data from the Russian Ministry of Labor that documents all applications for migrants to Russia in 2010 and allows matching the migrant to the sponsoring firm, show that there is very limited evidence of firms using migrants to fill high-skill jobs. In fact, the overwhelming majority of migrants, skilled or unskilled workers, were mostly originating from other states of the CIS. Furthermore, most were hired at relatively low wages in comparison to the occupation/region averages or the wages reported in the EBRD survey of recruiting firms. At the same time, there is a sizeable portion of demand for skilled migrants, which are offered wage premiums. The demand originates mostly from highly productive firms. Migration policy should acknowledge these different motives behind the demand for migrants.
▪
References
- Simon Commander and Irina Denisova “Are Skills a Constraint on Firms? New Evidence from Russia”, IZA Discussion Paper No. 7041, November 2012