Returns to language skills in transition economies

Returns to English proficiency

English language proficiency is an important asset in the globalized world. The importance of English language skills has grown in transition countries as they have increased their openness to trade and foreign investments. Evidence from the Visegrad Four countries of CEE—the Czech Republic, Hungary, Poland, and Slovakia—shows that English language knowledge is in high demand [3]. Across the four countries, over half of the 74,000 job advertisements accessed via job portals as of July 2015 had English proficiency as a requirement. Looking at individual countries, in the Czech Republic 28% of the vacancies required English language skills, while in Poland 64% of vacancies did (Figure 1). However, English language skills among these countries’ populations are not yet well-developed. As Figure 2 demonstrates, the share of those able to have a conversation in English ranges from 20% in Hungary to 33% in Poland. English language proficiency appears to be less common in the FSU countries (see the Illustration). Data from a collection of nationally representative surveys conducted in the South Caucasus in 2013 show that only around 19% of people in Georgia, 16% in Armenia, and 8% in Azerbaijan speak English at an intermediate or advanced level (author’s calculations based on data in [1]). It is therefore not surprising that proficiency in English is highly rewarded in these countries’ labor markets: it leads to a significant increase in the probability of having a job [4].

What about Russian language skills?

During the Soviet era, Russian proficiency was essential to a successful career in the member republics of the Soviet Union, and Russian proficiency remains highly prevalent in many successor countries over two decades after transition. The Illustration depicts the percentage of people who spoke Russian (at intermediate or advanced level) in the three FSU countries of the South Caucasus in 2013. Accordingly, Russian is by far the most common of the foreign languages spoken in Armenia, Azerbaijan, and Georgia. Over 80% of individuals in Armenia, 70% of individuals in Georgia, and close to 40% of those in Azerbaijan are proficient in Russian (calculations based on data in [1]). This is not surprising since the incentives to learn Russian remain strong [4]. The language commonly maintains its status as a second language and medium of communication between various ethnic minorities within the countries. It also maintains its status as the lingua franca in interactions with Russia and other FSU countries. Russian proficiency therefore pays off in the labor market: existing estimates suggest that Russian language skills significantly increase one’s probability of employment in at least some parts of the FSU [4]. Moreover, evidence from Kazakhstan suggests that individuals who are less fluent in Russian (these happen to be those who classify themselves as bilingual in Russian and Kazakh) may in fact be penalized in the labor market in the form of lower earnings [6].

Language skills and occupations

Language proficiency as a form of human capital is particularly valuable in jobs that are reliant on communication skills and have higher-level language requirements. Russian proficiency has been found to contribute to selection into public sector jobs and into jobs in foreign organizations, businesses, or non-governmental organizations in the three FSU countries of the South Caucasus [4]. This is consistent with higher language proficiency requirements in these jobs. Moreover, Russian proficiency increases selection into jobs in trade, services, and communications—sectors intensive in communication skills [4]. Language skills have also been linked to the complexity of jobs. Evidence from the Visegrad Four countries suggests that high-skill occupations are more demanding in terms of English language requirements relative to medium- and low-skilled occupations [3]. However, the same study shows that this observation does not hold for all foreign languages: it does not find a relationship between the demand for German language skills and the complexity of an occupation.

Language–skill complementarity

Studies in various country contexts have shown that individuals with higher levels of education might gain more from knowing a language than those with lower levels of education [7]. This is referred to as language–skill complementarity, and appears to exist in transition countries as well. An analysis of returns to Russian proficiency by educational attainment demonstrates that the returns are higher for those with post secondary education, suggesting complementarity between language skills and education in the generation of labor market rewards in FSU countries [4]. Work experience is another indicator of the quality of skills valued in a given labor market. Older workers will typically have acquired more work experience compared with younger workers. Consistent with language–skill complementarity, the returns to Russian language skills appear to be higher for older cohorts [4]. However, these relationships are apparent among men only. For women, there is in fact an indication of some substitutability between Russian proficiency and general skills: Russian proficiency is significantly related to the employment probability of young and medium-educated women only.

Gender bias in rewards for language skills?

Gender gaps are a persistent feature in labor markets internationally as well as in transition countries. The evidence on whether transition has resulted in an improvement or deterioration of women’s relative position in the labor market is mixed. The different returns received by men and women for the same characteristics are often interpreted as evidence of discrimination. Recent evidence from a group of western European countries suggests that men receive higher returns for various foreign language skills than women, a finding that is indicative of some form of discrimination against women linked to this specific type of human capital [8]. Another piece of cross-country evidence, on the other hand, suggests that foreign language proficiency reduces the risk of unemployment more for women than for men [2]. The limited evidence on this from transition countries is inconclusive. In Armenia and Azerbaijan, proficiency in Russian language skills apparently increases women’s employment probability more than it increases that of men [4]. The fact that Russian language proficiency on average is more prevalent among men compared to women is coherent with this pattern. On the other hand, the labor market value of speaking Russian is very high for men in Georgia, while for women it appears to be non-existent—a pattern that suggests the possibility of gender bias in returns to language skills [4].

Minorities in multilingual societies

Since the collapse of the Soviet Union in 1991, various programs to promote national languages have been initiated in the newly independent states. These include single-language policies that made the national language the official state language as well as educational reforms that promoted the use of national languages in instruction. Evidence shows that such nationalist policies may have hurt the outcomes of minorities. One example is a reform implemented in Ukraine’s school system in 2009/2010. The reform obliged all linguistic minority students to take a standardized school exit test in Ukrainian, the state language, thus denying them access to translated versions of the test. There is evidence to suggest that the reform resulted in a decline in the number of subjects taken by minority students in schools with Hungarian and Romanian/Moldovan languages of instruction, especially those demanding linguistic skills, such as History and Biology [9]. Nevertheless, even full language proficiency in the national language does not guarantee equitable outcomes for minorities in places where ethnic tensions are present. Evidence from Estonia and Latvia shows that knowledge of Estonian and Latvian, the sole official languages in the two Baltic countries respectively, is not closely related to an income premium for male minority workers (only the public administration sector shows a positive association) [10]. Another study of Estonia finds that proficiency in Estonian may in fact increase the risk of downward mobility for minorities in the labor market [11]. One possible explanation is that in the course of transition, an ethnic majority background was increasingly treated as an advantage for obtaining high-status jobs. It is likely that mainly people with relatively good Estonian language skills applied for such jobs, and then had to compete with ethnic Estonians. In contrast, those with poor or no proficiency in Estonian might have settled for lower-skilled jobs in which ethnic discrimination was less prevalent.

Language and education

Language policies promoting or discouraging the use of various languages in transition countries may have important implications for school quality and economic returns to schooling. In many FSU transition countries, schools with Russian language instruction have traditionally provided higher-quality education than schools with national language instruction. Policies that encourage school instruction in national languages are likely to negatively affect returns to human capital if they are not accompanied by appropriate investment in the quality of those schools. Evidence from Kazakhstan illustrates this point: the Kazakh language continues to yield a negative wage premium 13 years after it was made the official state language—a fact that is largely explained by the comparatively poor quality of schools with Kazakh as the language of instruction [12].

Alongside promotion of national languages, various efforts to increase the prevalence of English have been underway in transition countries, many of these through education systems. One example is university instruction in English, consistent with a worldwide trend to offer English language instruction. However, such changes may have negative consequences for the formation of human capital without considerable investments to address the linguistic and methodological challenges that arise in the process of making them. Indeed, a case study from Central Asia demonstrates that the switch to English-only instruction may negatively affect students’ academic outcomes when instructors and students are less than proficient in English [13].