Technological innovations and automatization are essentially changing jobs and job market tendencies around the world – Lithuania is not an exception. The average likelihood of a job position to be automatized in Lithuania is as high as 54 % (the second largest rate in OECD after Slovenia). In addition, a fifth of jobs face a high risk of automatization (above 70 %). According to the polarization hypothesis, automatization replaces routine work that requires medium-level competencies to be carried out, while jobs that require low or high skills are not affected by automatization – therefore, the demand for both of these skill groups is increasing. Thus, polarization in demand for skills is created (Acemoglu, Autor, 2010).
Is this polarization taking place in Lithuania? So far, no clear tendencies of the polarization of skills have been recorded. One of the main explanations of this is that employment in the field of manufacturing, which is most vulnerable to automatization, has so far remained stable and has not suffered too much (Galdikienė, 2020, p. 168). However, while analyzing skills and their change in Lithuania, regional differences are rarely taken into account. Actually, it may be that the polarization of skills takes place, but only on a regional level. Vilnius and Kaunas – cities that have strong scientific bases, are known for their universities and economic activities of higher additional value – might be one step ahead in the cycle of innovations than the rest of Lithuania.
The project will analyze the importance of geography to changes of skills in Lithuania – that is, the divide between Vilnius/Kaunas and the rest of the country and whether or not it conceals the polarization of skills. The answer to this question is relevant in practical terms, because it may also reveal which regions have already been hurt by the automatization or may be hurt the most in the future. During the project, data on the polarization of skills in Lithuania on a regional level will be systematized, aggregated and analyzed statistically.