The Macroeconomics of Structural Transformation

Abstract

This dissertation consists of three essays about the broad structural and technological changes to the US economy over the past fifty years, and their macroeconomic implications.

Chapter 1 looks at the relationship between the rise of IT and the rise in industrial concentration. Over the past four decades, most sectors of the US economy have become significantly more concentrated: a small number of large firms now have a much larger market share. Using administrative microdata on firms in every sector of the US economy, including new administrative data on firm-level IT investments, I show that the IT Revolution \\disproportionately benefited the largest firms - an instance of what we might call scale-biased technological change (ScBTC). I document that the largest firms have invested more intensively in IT than smaller firms have, particularly in the most advanced and proprietary kinds of IT. I also document that larger firms are generally more capital intensive than smaller firms, and operate at higher returns to scale, but that large firms invest an even greater share of their capital expenditures on IT than small firms do. I discuss how production function estimates and a model of firm heterogeneity and fixed costs might be used to show how the decline in the price of IT ultimately led to the rise of US industrial concentration, and how this can be distinguished from competing causal theories about a fall in competition or a rise in competition. Simple calibration exercises can quantitatively generate large increases in concentration from declines in the price of IT.

Chapter 2, joint work with Jan De Loecker and Jan Eeckhout, looks at the relationship between rising industrial concentration, and rising aggregate markups. We document the evolution of markups based on firm-level data for the US economy since 1955. Initially, markups are stable, even slightly decreasing. In 1980, average markups start to rise from 21\% above marginal cost to 61\% now. The increase is driven mainly by the upper tail of the markup distribution: the upper percentiles have increased sharply, while the median is unchanged. In addition to fattening upper tail of the unweighted markup distribution, there is reallocation of market share from low to high markup firms. This rise occurs mostly within industry for {\it all} industries. We also find an increase in profits. While there is some increase in overhead costs, the markup increase is in excess of overhead. We then discuss the macroeconomic implications of an increase in average market power, which can account for a number of secular trends in the last four decades, most notably the declining labor and capital shares as well as the decrease in labor market dynamism.

Chapter 3 presents a broad research agenda going forward for understanding the macroeconomics of structural transformation, and specifically outlines some speculative theories about our basic theories about economic growth, and the business cycle.