*Names in bold indicate Presenter
A hedonic price analysis developed in this paper shows that the rate of decline in prices for microprocessors and computer memory increased by 50% or more in the second half of that decade, compared with the early 1990s. The fast pace of technical innovation in semiconductors overall has been identified in some studies as a primary factor underlying increases in the pace of IT innovation over this period, and even more importantly, in productivity growth in the overall U.S. and global economies.
Somewhat alarmingly, however, the analysis also shows a sharp decline in the pace of price‐performance improvement in microprocessors beginning around 2003‐2004. A similar deceleration in declines in quality‐adjusted computer prices seems to have mirrored these trends in semiconductor prices. The slowdown in semiconductors appears be having a broad and continuing ripple effect in reducing the pace of progress in computers, the deployment of IT in the global economy, and may very well play an important role in what now appears to have been a post‐millennium slowdown in productivity growth in the American economy.
This paper also explores to what extent the more rapid pace of technical progress in microelectronics in the late 1990s was attributable to policies affecting technological innovation in semiconductor manufacturing and to what extent it was due to non-policy factors. I conclude that absent a new wave of innovation in microprocessors, the apparent slowdown in the pace of technical change will have a continuing ripple effect in reducing the pace of progress in computers, the deployment of IT in the global economy, and ultimately reduce gains in productivity across IT‐using sectors of the global economy in coming years. I also conclude that there is evidence that policy interventions of the late 1990s and early 1990s—e.g., a significant increase in public funding of semiconductor R&D, decisions to encourage creation of multi‐firm semiconductor R&D consortia, and in parallel, the creation of an International Technology Roadmap for Semiconductors, supported by virtually all semiconductor equipment, materials, design, and production firms around the world, and its coordination of private and academic R&D projects —played some role in accelerating the pace of technological innovation in semiconductor manufacturing. Finally, I consider whether it is likely that PCs are no longer at the cutting edge of technological change in information technology—whether today it is the Internet and communications to which the economic center of gravity for IT innovation has shifted— and the extent to which this may provide a new set of opportunities for policies intended to improve future rates of innovation in these sectors.