We suggest a new way to identify salient features of the Russian labor market. Parameters of basic macroeconomic models pertinent to the Russian labor market are compared to a sample of other countries. We find that estimated values of Okun's coefficient and the elasticity of real wages to labor productivity in Russia are typical for emerging markets. What really distinguishes the labor market is that the elasticity of real wages relative to unemployment in Russia is very high by international standards. The overall conclusion is that the Russian labor market can be characterized by a combination of serious structural problems (such as low employee mobility, the significant size of the shadow sector, etc.) and solid macroeconomic performance, verified by the persistently low rate of unemployment in recent years.

Russian labor market, cross-country comparative analysis, macroeconomic models, migration

At first glance, the Russian economy has rather standard general characteristics. In a sample of 20 large economies (including 10 developed countries and emerging markets), Russia ranked 14th in average unemployment (7.0%) and 11th in the coefficient of variation for this indicator (21%) for the period 2000 to 2016 (Fig. 1).

Fig. 1

Average rate and coefficient of variation for unemployment, 2000 through 2016 (%).

Source: Authors’ calculations based on IMF data.

However, we can investigate Russia's specific features further if we switch from examining “static” characteristics to examining “dynamic” ones that demonstrate its labor market reaction to shocks. For example, Fig. 2 represents the scale of production shocks for the same sample of countries during the “great recession” and the labor market reaction to them. These indicators were calculated based on Okun's law: the horizontal axis represents changes in the GDP growth rate; the vertical axis represents the acceleration (or deceleration) of unemployment trends (2009 compared to 2008 in both cases). Russia's economy experienced the greatest deceleration in growth among all of the sampled countries, with its growth rate declining by 13.1 percentage points (p.p.), while unemployment demonstrated moderate acceleration (rising by 2.0 p.p. in 2009 compared to 0.2 p.p. in 2008). A simplified calculation (based on two points) of Okun's coefficient (see Section 3.1), correlating unemployment rates with economic growth rates shows that the absolute values for only two countries in the sample (Italy and Malaysia) were lower than in Russia.

Fig. 2

Difference between GDP growth rates and changes in unemployment in 2009 compared to 2008 (p.p.).

Source: Authors’ calculations based on IMF data.

Of course, these calculations are more illustrative in nature; for example, they do not take into account changes in the average hours worked by employees. Some countries actively used this mechanism to adapt to the crisis, but in Russia, this was used only to a small extent. According to the OECD, the average Russian employee worked 1.3% fewer hours in 2009 than in 2007, whereas this workload dropped by 3.3% in Germany, and 4.1% in the U.S. Moreover, the countries differed slightly on indicators such as when production began to decline and the scale of anti-crisis programs related to the labor market, among others. Nevertheless, Fig. 2 demonstrates that the Russian labor market is notable regarding the nature of its reaction to negative shocks. As a a result, the question remains: was Russia's reaction too weak or was the reaction in other countries too strong? The Okun coefficient calculated for Russia (–0.14) does not seem to be abnormally low in terms of long-term relationships, whereas for some countries this indicator is close to or even above one (in absolute value), which appears to be an excessively strong reaction. Still, it is unclear how these estimates should be interpreted, as the observed labor market trends include both long- and short-term relationships.

Another important aspect of the Russian labor market is that it is one of the few in the world where the share of labor income in GDP has grown in recent years (Brazil is another example; however, its trend is considerably less pronounced). If we look at GDP and employee compensation data in Russia (calculated according to a modified method after 2011) in a comparable form to that of earlier years, we can see that the proportion of wages to GDP grew from 40% in 2000 to 54% in 2016 (Fig. 3). For other countries, this indicator has remained comparably stable (as in the eurozone) or declined (as in the U.S. and Poland) during the same period. The literature (for instance, Grossman et al., 2017; Guerriero and Sen, 2012) cites several possible reasons for a reduction in labor income as a percent of GDP. However, the opposite trend in Russia requires an explanation.

Fig. 3

Compensation of employees in GDP (%).

Source: Authors’ calculations based on OECD and Rosstat data.

On the whole, the Russian economy (including the labor market) also is notable due to the significant regional differentiation for most of its economic indicators.

The Russian labor market has been examined in a large number of studies that have considered its structural and institutional characteristics. Many of the most important results have been gathered into a series of collective monographs (Gimpelson and Kapeliushnikov, 2011, 2014, 2017). At the same time, there are only sporadic studies on the macroeconomic characteristics of the Russian labor market before 2015. Rare exceptions include, in particular, the estimate of natural unemployment (NAIRU) by Akhundova and Korovkin (2006), the “wage curve” (Shilov and Meller, 2008), and the Phillips curve (Gafarov, 2011). In addition, several studies describe failed attempts to build basic macro models. For example, Akhundova et al. (2005) concluded that the connection asserted by Okun's law was unstable and its evolution uncertain in the Russian economy.

Notwithstanding the scant research into mechanisms for macroeconomic adjustment in the Russian labor market, the notion of the “Russian model of the labor market” formulated in macroeconomic terms is generally accepted. This notion was first introduced by Layard and Richter (1995) who noted that during the initial years of market reforms in Russia, unemployment grew slower than in other transitional economies despite the considerably greater production decline. Since then, the combination of a rather strong reaction by wages and a relatively weak reaction by the number of workers to production shocks has been considered the main attribute of the Russian labor market. Gimpelson and Kapeliushnikov and other authors have returned to this subject on several occasions, citing arguments in favor of similar reactions to subsequent shocks (Gimpelson and Kapeliushnikov, 2011, 2015).

However, we believe this common notion has a number of weaknesses:

1. It is based on separate, sporadic observations that are unrelated to each other through any macroeconomic model. One of the implications is that it is impossible to tell whether the observed reactions reflect a short- or a long-term adjustment, namely, whether we are dealing with a shift in, or a temporary divergence from equilibrium.
2. The observations used pertain to the crisis periods (1998, 2009, etc.). However, critical situations are always unique, both in terms of the shocks experienced and the reactions to them. In particular, special measures taken by the government (often including temporary changes in the “rules of the game”) play a substantial role during such periods. As a result, labor market responses to crisis shocks may provide a distorted view of “normal” mechanisms. For example, the analysis conducted by Darius et al. (2010) demonstrated that in 2008 and 2009, the reaction of the labor market in most countries turned out to be much stronger than would have been expected based on past experience.
3. The question regarding which economic mechanisms determine the specific attributes of the Russian labor market reaction to shocks is still open (or, rather, has not been even raised).

The above demonstrates that the generally accepted idea of the Russian labor market model represents more of an expert judgment than a well-grounded proposition.

In this paper, the results from building basic macroeconomic models for the Russian labor market, carried out over the past several years, are compared to the estimates of similar models for other countries. A systematic cross-country analysis of the parameters from these models allows us to finally formulate the distinguishing properties of the Russian labor market model. This approach also reveals macroeconomic mechanisms underlying patterns of adjustment to shocks typical for the Russian labor market.

Russia's population has relatively low spatial mobility (Bell et al., 2015; OECD, 2011). From 2002 to 2010, registered internal migrants made up only 1.4% of the population in Russia. By comparison, from 2000 to 2006, this indicator averaged 13.7% in the U.S., 14.6% in Canada, and 4.6% in Japan.2 One indirect proof of the low mobility in Russia is the comparatively large inter-regional differentiation in a number of economic indicators (Guriev and Vakulenko, 2012). We use cross-country comparisons to confirm it this (Fig. 4a). Evidently, the average value of the gross regional product (GRP) logarithm per capita in Russia was one of the lowest between 2000 and 2005, whereas the differentiation by GRP was higher among the countries in the sample that also have quite large territories.

Fig. 4

GDP and unemployment interregional variation by country.

Source: Authors’ calculation for Russia; Che and Spilimbergo (2012) for remaining countries^a.

^aChe and Spilimbergo (2012) calculate the population-weighted sum of absolute gaps between regional GDP per capita and national average, thus adjusting for the fact that different countries have different numbers of subnational units. They also take into account the fact that data are not available for some regions; they calculate national averages as population-weighted averages using the available sub-national data.

Interregional differences in wages and unemployment rates in Russia are also significant. Fig. 4b shows a cross-country comparison in unemployment rates. In Russia, the differences in unemployment rates are among the highest. Out of the countries in the sample, only in Mexico, Germany, and Italy are they higher than in Russia.

Looking at the trends for interregional variation within Russia in recent years (Fig. 5), we can see that variation by GRP and wages declined during the 2000s. In terms of the unemployment rate, interregional variations fluctuated considerably between years, having no pronounced trend.

Fig. 5

Variation between Russian regions in real wages, unemployment, and GRP per capita.

Note: The graph shows $\sqrt{\frac{1}{2}\sum _{i=1}^N{p}_i{(X_{it}-{\overline{X}}_t)}^2}$ where Xit is the log of real income (or real wage, or unemployment or real GDP per capita) in region i in year t, and ${\overline{X}}_t$ is the population average log of real income (or real wage, or unemployment, or real GDP per capita, respectively) in year t. P and Pi are the population of Russia and of region i, respectively.

Source: Rosstat's official data; authors’ calculations.

Thus, interregional variation within Russia is higher than in other countries, however, it has been reduced in recent years. At the same time, the intensity of migration in Russia is lower than in other countries of comparable size, and remains rather stable. In recent years, the barriers and incentives for migration have been declining concurrently. No economic barriers were visible for migrants during the 2000s that might have held them within their own regions (Guriev and Vakulenko, 2015). However, despite the fact there are no more “poverty traps,” which serve as the main barriers to migration, migration flows have not increased in Russia, possibly due to the significant reduction in interregional differentiation of per capita income, wages, and unemployment (Guriev and Vakulenko, 2015), namely, due to poorer incentives for migration.

The main reason behind the reduction in interregional differentiation of average per capita income in Russia is, in the opinion of Guriev and Vakulenko (2012), higher mobility of capital, which has increased due to the development of the financial sector and real estate market. The role of migration in reducing interregional differences within Russia is reviewed in Vakulenko (2016). It is noted that migration may lead to both a decrease and increase in interregional differentiation, as it influences both the demand for and supply of labor. The empirical results of Vakulenko demonstrated that although internal migration within Russia affects per capita income and wages in the short run (an outflow of migrants from a given region leads to growth in wages and per capita income), the effect of migration is low and, as a result, it makes no relevant contribution to reducing interregional differentiation. The effect of external migration is also insignificant (Vakulenko, 2016). The effects of migration on interregional convergence of income and GRP per capita in other countries are presented in Table 7. Migration leads to a convergence of GRP per capita and income in 40% of the studies reviewed, and to convergence in unemployment rates in 50% of the studies but the number of such studies is small.3 Thus, according to those studies, it is hard to confirm that migration helps in convergence. Therefore, Russia does not differ from the majority of countries in this respect.

Table 7Download as CSV 

Aggregated results regarding the effect of migration on interregional convergence.

Indicator	Migration leads to convergence (% of studies)	Number of studies
GRP per capita	40	15
Income per capita	42	12
Unemployment rate	50	4

Source: Authors’ analysis based on Table 1 from Vakulenko (2016).

Our analysis allows us to formulate the following general observations about the macroeconomic characteristics of the Russian labor market.

1. The Russian labor market is already formed and mature. The key links between the Russian labor market indicators are sustainable and long-term in nature (the indicators are co-integrated), are clearly pronounced (the coefficients differ from zero), and have the expected signs. In many cases, we can determine the direction of the causal relationships; which is also expected. Another sign of maturity is the applicability of Okun's law.
2. The strength of most correlations for the Russian labor market is typical for “emerging markets,” with only one exception, namely, a high elasticity of real wages relative to unemployment.
3. “Non-traditional” mechanisms function in the Russian labor market. We found neither signs of a mechanism of “effective wages” (the effect of wages on productivity) — the impact is going only from productivity to labor remuneration — nor signs of utilizing employment in the public sector as a social policy tool (Gurvich and Khazanov, 2016).
4. The Russian labor market is effective from a macroeconomic point of view. The high elasticity of real wages relative to unemployment ensures that the unemployment rate quickly returns to equilibrium. The asymmetry of reactions is also less visible than in other countries.
5. The efficiency of the Russian labor market signifies its capability to self-adapt. The quick return of the economy to a state of full employment after shocks provides an argument against using fiscal and monetary stimulation measures.4
6. On the whole, the main feature of the Russian labor market from a macroeconomic point of view (the real “Russian model of the labor market”) is its normality, namely, the absence of rigidities found in other countries.
7. The Russian population is less spatially mobile than in other countries of comparable size, and the degree of this mobility is rather stable over time. At the same time, the interregional differentiation by labor market indicators is comparatively large in Russia; however, it has been decreasing over the past several years.

Our findings have solid institutional grounds: many common sources of rigidity in the labor market are either absent or quite weak in Russia. The following factors can be noted:

1. the modest role of trade unions in forming labor remuneration (OECD, 2011);
2. a lack of rigid constraints on the dismissal of employees (also not always observed in practice);
3. comparatively low mandatory minimal wages (20% of the average wage as of mid-2017, and around 10% on average in the 2000s);
4. significant (over one-third) share of bonuses, and other flexible components in the structure of wages (Gimpelson and Kapeliushnikov, 2011).

On the whole, according to the World Economic Forum (2016) estimates, Russia occupies a rather high position in terms of flexibility in determining wages (23 out of 138 countries). At the same time, the Russian labor market rates comparatively poorly on other indicators: in terms of collaboration between employees and employers, Russia is ranked at just 103 (World Economic Forum, 2016).

In sum, we can conclude that the Russian labor market can be characterized by a combination of serious structural problems (such as low employee mobility, the significant size of the shadow sector, etc.) and good macroeconomic performance, which is confirmed by the persistently low rate of unemployment in recent years.

This situation is evidently driven by the specific features of the institutional environment noted above. However, institutions do not appear exogenously but are shaped endogenously. We believe that the Russian labor market, unlike goods markets, has been fortunate to avoid excessive regulation due to not being a potential source of “administrative rent.” For this reason, the market mechanisms are not significantly distorted, and therefore, the labor market is highly efficient from a macroeconomic point of view.