How do firms change their employment after adopting AI? We answer this using two unique Danish data sources: the official firm-level IT usage survey, which includes AI-related questions since 2022, and a matched employer--employee register that records every wage payment at monthly frequency. Together, they identify the firms that first used AI in 2023—the earliest measurable cohort—and follow them month by month. Adopters are larger than non-adopters and employ more workers in AI-exposed occupations, so we compare each firm not against other firms but against the employment path it was already on. We find that adopters fall behind that path significantly more than non-adopters do: $6\%$ two years on, 11% by late 2025. The gap comes entirely from adopters with fewer than 100 FTEs, and runs through reduced hiring rather than layoffs, concentrated in AI-exposed occupations—25% below trend, against 7% elsewhere. Wages do not respond: the entire adjustment runs through employment. These effects are so far too small to show up in economy-wide employment: shrinking AI adopters employ too few workers to move the aggregate, and while AI-exposed industries do shed employment, an exact decomposition attributes only about a quarter of that to adoption.

We study the transmission of firm-specific export shocks through labour markets, examining how these shocks influence workers' wages at competing firms via labour market competition. Using rich Danish administrative data, a flexible, empirically driven definition of workers' outside options, and an identification strategy based on world import demand shocks, we show that both own-firm and competitor export shocks raise incumbent workers' wages. These shocks also trigger competitive worker reallocation, as expanding exporters poach workers from other firms. The results indicate spillover effects of export shocks extending to non-exporting firms. 

This paper studies the long-lasting effect of a credit supply shock on firms' economic activity and labour force composition. Using a novel matched bank–employer–employee dataset from Denmark covering 2000 to 2016, I exploit differences in bank exposure to the freeze of the international interbank market in 2009. Firms connected to exposed banks reduced employment by 15 per cent more than the control group, driven by a sharp contraction in hiring and a higher separation rate alongside a 20 per cent drop in exports. The contraction co-moved with an increase in average wages, reflecting a shift towards a higher-skilled workforce. Firms that maintained high exports and employment did so by eroding liquid assets, while downsizing firms retained liquidity as precautionary savings. The effect was long-lasting and altered firms' composition throughout the post-2009 period. Contrary to much of the literature, I find no differential effect on the investment rate. 

This study explores the dynamics of cash holdings across the entire population of Danish firms from 2003 to 2018, revealing a significant post-Global Financial Crisis surge in cash reserves, predominantly among small to medium-sized enterprises (SMEs) - a group often overlooked in existing research. On the contrary, cash holdings of larger Danish firms are stable and lower than those of public firms in the U.S. We show that small firms' entry and exit dynamics account for the post-crisis rise in cash holdings. We document correlations of cash holdings with firm characteristics consistent with those documented for large public firms, particularly a strong negative correlation between cash holdings and leverage. Using Oaxaca-Blinder decompositions, we show that the lower leverage of firms entering after the financial crisis fully explains the entry contribution to the increase in cash holdings. The contribution of firm exit cannot be explained by changes in the characteristics of exiting firms, in line with a direct effect of liquidity on survival outcomes during the crisis and its aftermath. Our results stress the critical role of firm dynamics on cash holdings' development.

This paper proposes a model of the joint effect of changes in the availability of secured and unsecured household credit on unemployment, where the two types of credit can act as substitutes. Extending the Mortensen–Pissarides framework, I develop a two-sector search and matching model with inter-sectoral worker mobility, a housing market and two frictional goods markets. Households face endogenous borrowing constraints when financing random consumption opportunities. Financial innovations that expand credit availability generally raise firm revenues and lower unemployment, but when shocks are strong enough to trigger reallocation across sectors, complex dynamics emerge and shocks to one credit market can adversely affect the price and availability of credit in the other. Calibrating with the Survey of Consumer Finances and time-series data on extracted home equity, I find that the rise in home-equity-secured credit explains only a small share of employment fluctuations between 2002 and 2010, while the contraction in unsecured credit is an important determinant of the post-recession drop in employment. 

Foreign labour plays an increasingly important role in the Danish economy. Around one in seven employed persons in Denmark were of foreign origin in 2024, half of whom had lived in Denmark for less than ten years. The shift towards more knowledge-intensive industries means foreign workers are increasingly employed in highly paid positions, to a greater extent than in the EU overall. Model-based calculations indicate that foreign labour responds to cyclical developments in Denmark, which may reduce the risk of wage pressure amplifying economic fluctuations. 

An important monetary transmission channel works through the impact of interest rate changes on non-financial firms’ investment decisions. Higher monetary policy rates affect firms’ investment decisions by raising the costs of external financing and lowering aggregate demand for goods and services. Using 20 years of firm-level data from Denmark, we find that an increase in monetary policy rates is associated with a reduction in corporate investment, particularly by smaller firms. 

Employment increased significantly in Denmark, the euro area and the US during the post-pandemic expansion, with stronger growth relative to GDP than in previous expansions. The high job intensity of the recent expansion in Denmark has several causes, including sectoral shifts, hiring of less productive labour and increased use of labour in production. This was supported by sharply rising demand, increased supply, and wages rising more slowly than product and input prices. The pressure on the Danish labour market has nonetheless eased over the past two years. 

We estimate a new measure of pension wealth gains covering the working-age population in Denmark during 2015–2022, in collaboration with Statistics Denmark. By linking this data to existing income and tax registers at the individual level, we provide a full picture of net financial wealth gains across the age distribution, laying the basis for further analysis such as the impact of monetary policy on net capital gains in Danish households. The data can be accessed through Statistics Denmark's research data portal.

Technical notes: PENSFLOW data · imputation of Aldersopsparing payouts

We show that the cash holdings of a typical Danish company increased after the 2007/08 financial crisis, especially among small companies. This coincides with a decrease in the use of loan financing, suggesting a decline in loan demand or supply. The development is also driven by a strong selection effect: companies with low cash holdings were especially likely to exit during and after the crisis.