Computers and growth with frictions: aggregate and disaggregate evidence

Abstract

Not long ago, discussions of the computer revolution by economists centered on the disappointing payoff computers seemed to be having on aggregate, industry, or firm-level movements in productivity, but this disappointed has passed and the business pages are filled with stories heralding the arrival of the “new economy”. This paper investigates the role of computers, software, and communications equipment in the recent surge in U.S. productivity growth in a neoclassical model of investment and production in an attempt to clarify the potential importance of frictions in the transition to a more computer-intensive mode of production on the productivity effects of high-tech equipment. The estimated response of investment in high-tech equipment to its relative price is substantial and sluggish. The high-price elasticity of high-tech capital implies, in conjunction with reasonable assumptions about future declines in high-tech equipment prices and multifactor productivity throughout the economy, that trend GDP growth over 2001–2005 is likely to range between 3 percent and 3-$\text{3}\text{4}$ percent. The estimated investment frictions are suggestive of complicated dynamics in the short-run impact of high-tech investment on productivity; firm-level evidence suggests such short-run effects may be important, as do the large costs of training workers and installing high-tech capital.