Ongoing Projects
Supplier Search and Market Concentration [JMP]
Abstract: I analyze how falling search costs - driven by improvements in communications and transportation technology since the late 1990s - have affected firm size distribution and allocative efficiency. Lower costs have enabled smaller firms to source inputs internationally, reflected in the declining size of median and lower-percentile importers. Meanwhile, top importers have also grown, driving up market concentration. I suspect these trends originate from the structure of the intermediate goods market. In Swedish import data, I find that input prices exhibit rising variance and are negatively correlated with productivity. Based on my empirical findings, I develop a quantitative search-and-bargaining model in which search costs serve as both expansion cost and entry barrier. I find that decreasing search cost sometimes favor the most productive firms and sometimes smaller ones. Consequently, market concentration and efficiency respond non-monotonically to changes in search costs.
When Unified Market meet Local Markets: How Big Firms Drive Local Price Dynamics?
Inflation Persistence and a new Phillips Curve
with Marcus Hagedorn, Juan Llavador & Kurt Mitman
Abstract: Auclert et al. (2024) recently argued that, to first order, menu-costs models deliver the same New Keynesian Phillips Curves as time-dependent models in response to AR(1) shocks. We show here that when considering a broader class of shocks, menu-costs models can generate qualitatively and quantitatively different Phillips curves than implied by time-dependent models. Shocks to the growth rate of nominal demand generate inflation persistence in the model, in line with the data, but at odd with the standard time- dependent NKPC. Changes in the extensive margin of price adjustment in the menu-cost model generate history dependence that is captured by the lagged inflation rate. Once we control for lagged nominal demand growth, the explanatory power of lagged inflation drops significantly. The reason is that nominal demand growth is a second determinant of inflation in the Phillips curve in menu-cost models and inflation therefore inherits the persistence of the process for nominal demand.