The labor market implications of generative artificial intelligence (GenAI) have been discussed since the launch of ChatGPT. So far, the most reliable studies on the labour market effects of GenAI have been conducted by studying online labour markets. These studies compare jobs that are more exposed to GenAI with less exposed jobs before and after the launch of ChatGPT. Studies have found that demand for work has decreased in tasks more exposed to GenAI.
In this brief, we present the results of a study in which we compare earnings and employment development in occupations more exposed to GenAI with occupations less exposed to GenAI before and after the publication of ChatGPT using Statistics Finland’s Incomes Register. The results show that earnings have increased faster in exposed occupations than in non-exposed occupations. There have been no differences in employment development between these groups. In Finland, GenAI has so far not caused any negative labour market effects. The results suggest that GenAI has increased labour productivity and thus labour demand. Of course, in the longer term (beyond the 20-month ex post window in our study), the results might be different as GenAI evolves and its use expands.
