Filip Mellgren

Private Know-How, Creative Destruction, and Growth

Job Market Paper

Paper (PDF)

Abstract. New-to-firm and exiting products tend to be small. I document this pattern in Swedish manufacturing data and study its implications for creative destruction’s contribution to growth. I develop a growth model in which innovation creates private know-how that competitors cannot inherit. This know-how can protect large products from replacement, concentrating replacement among smaller products where the same proportional quality gain contributes less to growth. In the calibrated model, creative destruction contributes 0.58 percentage points less to annual growth than in an accounting benchmark that removes the association between productive scale and protection while preserving replacement frequency. A uniform increase in R&D subsidies raises growth mainly through improvements to existing products. Targeted subsidies also change research at firms receiving no additional support.

Monopsony Power and Creative Destruction

R&R at American Economic Journal: Macroeconomics

Coauthors: Isabella Maassen, Jonas Overhage.

Paper (PDF)

Abstract. This study examines the effects of monopsony on macroeconomic productivity and growth using an endogenous growth model with heterogeneous firms facing upward-sloping labor supply curves. We find that this framework can rationalize the prevalence of unproductive yet innovating firms that would otherwise be crowded out by more productive competitors. In addition, we show that monopsony is an important determinant of product market power. Calibration to U.S. data confirms previous findings that monopsony decreases static productivity. However, it also leads to higher growth. We estimate that a 1% narrowing of the markdown increases the present value of output by about 1.08%. (JEL O31, O40, J42)

Skill Premium and Markups

Coauthors: Anders Åkerman, Horng Wong.

Preliminary slides — November 2026: HTML · PDF

Abstract. In many high-income countries, the rising market shares and profitability of large firms are accompanied by a pattern of increasing concentration of highly skilled workers towards those firms, as well as rising employment of skilled workers as a share of the economy’s total workforce. We build a structural framework to explore whether skill-to-firm sorting contributes to the higher (and rising) profitability and markups of large firms, and whether skills are compensated for the markups they generate. Our model highlights a new channel for sorting: more efficient firms have a stronger incentive to employ more skilled workers to increase their markups. Since skilled workers are in short supply relative to less skilled workers, this channel strengthens demand for skilled workers beyond the role of technology. This result implies that skilled workers may be compensated for the markups they generate, raising the skill premium. Thus, product market competition can shape the relative wages and employment of skills in the economy. Models that do not consider sorting imply that wages and employment of all skill types are always penalized by higher markups. We apply our model to Swedish data and run a “race between education, technology, and market structure” to decompose the sources of changes in the skill premium.