Institutional context and the employment effects of artificial intelligence in European SMEs

Abstract

Using novel data from the Survey on Access to Finance of Enterprises covering the last quarter of 2025 and an Inverse Probability Weighted Regression Adjustment approach, we assess the conditional treatment effects of artificial intelligence (AI) use on the likelihood of employment growth among Small and Medium-sized Enterprises across twelve EU countries, under selection on observable firm characteristics. Given the cross-sectional nature of the data and the contemporaneous measurement of AI use and employment change, these estimates should be interpreted as conditional associations rather than as definitive causal effects. We find that AI use is associated with a positive conditional effect on the probability of employment growth, with no significant association with the likelihood of employment decline, robust to alternative estimators and leave-one-country-out specifications. The estimated conditional effects are also strongly heterogeneous across firm characteristics. Positive responses are concentrated among small and medium firms and among young and mature firms, while micro and old firms show no significant effects. Services and trade exhibit positive employment-growth associations, whereas construction is the only sector associated with a significant increase in the likelihood of employment decline. Institutional context also appears to matter. The association between AI use and employment growth is stronger in countries with higher innovation performance and governance quality, and weaker where employment protection is stricter and political centralisation is greater. These findings suggest that AI integration and institutional conditions may be important for employment-enhancing digital transformation.