A Quantitative Study of the Replacement Problem in Frictional Economies

The question of how technological change affects labor markets is a classical one in macroeconomics. A standard framework for addressing this question is the matching model with vintage capital and exogenous technical progress. Within this framework, it has been argued that the impact of technological change on labor market outcomes differs according to the mechanism through which the new technology enters the economy. In particular, it matters whether: (1) new capital replaces old capital by destroying the job and displacing the worker (Schumpeterian creative-destruction) or old capital can be "upgraded" to the frontier technology (Solowian upgrading); (2) firms make the technology adoption decision unilaterally (hold-up), or the investment decision is surplus-maximizing (efficient investment). Our main finding is that for quantitatively reasonable parameter values the specific details of the model for how technology is introduced and who decides on investments do not matter for the equilibrium outcomes of our main variables of interest: unemployment, wage inequality, and labor share. The intuition for this "equivalence result" is that these models will yield significantly different implications only if the matching process is very costly and time-consuming, but our calibration shows that this meeting friction is minor