Abstract: How do digital technologies affect early-career employment and wage growth? Using administrative data from the Netherlands for 2006–2024, I document two concurrent trends across school-leaving cohorts: the probability of employment after leaving school fell, but early-career wage growth rose. I propose that the rise of digital technologies offers a unifying explanation for both trends. Following investment spikes in digital technologies, such as software and databases, firms reduce their early-career employment share and increase early-career wage growth. By enabling firms to codify their routines and knowledge, digital technologies substitute for early-career workers but augment the experience of those who do become employed. A calibrated heterogeneous-firm model that embeds both channels shows that the decline in the price of digital technologies is quantitatively important for both cohort trends. Substitution rather than augmentation is the dominant channel.

Selected for the IZA@LISER Summer School in Labour Economics (2026)

Presented: Sciences Po, Le Mans, IZA@LISER Summer School, ESPE Helsinki, AIEL Rome, Erasmus University Rotterdam, University of Amsterdam

Related: Using résumé and job posting data for the Netherlands, I find that generative AI has shifted hiring away from entry-level roles and towards experienced ones. However, I find no decline in entry-level employment in exposed occupations, consistent with scale effects offsetting substitution effects.

Abstract: We decompose the wage penalty of temporary work using matched employer-employee data from the Netherlands. Temporary workers earn 28 log points less per hour than permanent workers. Worker fixed effects explain half of this gap, while firm pay premia account for 28%. We find that most of the firm contribution reflects temporary workers sorting into low-paying firms rather than within-firm pay differences. As temporary employment differs substantially across demographic groups, we extend our framework to gender and native-migrant wage gaps, separating contract from firm sorting. Accounting for contract type explains part of the native-migrant firm pay-premia gap, but matters little for the gender wage gap.

Presented: Torino LABORatorio R. Revelli Workshop, Richmond Fed (CORE Week; co-author), 1st Aarhus Workshop on Labor Markets (co-author), Maastricht University (co-author), Scottish Economics Society, University of Amsterdam MInt Seminar, Spatial, International, and Macro Seminar Erasmus University Rotterdam, EALE Barcelona

Abstract: We study how revenue-neutral corporate tax reforms affect aggregate investment, productivity, and welfare in a heterogeneous firm model. Along the revenue-neutral schedule of rates and deductions, we show that a high-rate, high-deduction regime generates more aggregate investment and higher welfare than a low-rate, low-deduction one. More generous deductions act as an investment subsidy that benefits small to medium-large firms, which invest intensively relative to their size, and thereby foster capital accumulation despite the higher tax rate. However, the same mechanism reallocates capital away from the largest and most productive firms, reducing aggregate productivity. An event study exploiting firm-level variation in exposure to a 2005-2007 Dutch reform supports our model’s key mechanism.

Presented: Theories and Methods in Macroeconomics, Dutch National Bank XAmsterdam Macro Workshop, 5th Sailing the Macro Workshop Ortygia (co-author), University of Naples Federico II (co-author),  7th Annual CefES Conference (co-author)

Abstract: The U.S. aggregate labour share of income has declined significantly since the mid-1980s. The prevailing view in the literature attributes the decline to common, within-sector decreases in labour shares, thereby downplaying the contribution of sectoral reallocation. Using a new decomposition method, I show i) the long-term decline is overwhelmingly driven by sectors whose labour shares fell as they grew in size, ii) reallocation toward high-labour-share sectors would have offset nearly half of the total decrease, an order of magnitude more than previously found; iii) the within-sector effect is only half as large as other studies conclude, no larger than normal fluctuations over the business cycle.

 Presented: Erasmus University Rotterdam Brown Bag Seminar, Erasmus University Rotterdam PhD Seminar