An encouraging sign for US tech innovation
AEIdeas
It would be really bad for the US economy and workers if the pace of American innovation sort of petered out. Just search “great stagnation” to find out why (if it’s not intuitive). Less tech innovation would be a particularly worrisome. From “How fast are semiconductor prices falling?” by researchers David Byrne, Stephen Oliner (of AEI), and Daniel Sichel:
A stalling out of innovation in this sector likely would have broader implications for the economy, as semiconductors are an important general-purpose technology lying behind machine learning, robotics, big data, massive connectivity, and many other ongoing advances. Indeed, adverse developments in the semiconductor sector ultimately would damp the growth potential of the overall economy. … [But] if technological progress and attendant price declines were to continue at a rapid pace, powerful incentives would be in place for continued development and diffusion of new applications of this general-purpose technology.
Anyway, whether or not this is happening is difficult to determine. One way: Look at whether the price of tech stuff, like computer chips, is declining, adjusted for quality. According to government data, chip prices haven’t fallen much in recent years — just 9% from 2008 through 2012 — after rapid declines from the mid-1980s onward, especially during the late 1990s and early 2000s.
That smells like innovation stagnation. Yet there was also continuing improvement in chip performance. How to explain the disconnect? Well, maybe the government is calculating the numbers wrong. Bryne, Oliner, and Sichel point out that right when price declines slowed, Intel changed how it priced its chip and introduced new models:
Prior to the mid-2000s, Intel generally introduced new chips at the technological cutting edge and lowered the list prices of existing chips to remain competitive on a price-performance basis. However, by 2006, Intel had shifted to a new paradigm in which it largely kept the list prices of existing chips unchanged and began introducing new chips both at the frontier and at lower performance levels.
So Intel changed how it did business, but government price trackers didn’t change their methodology in response. If they had, by better gauging quality changes, the official stats would look quite different:
Our preferred hedonic index of [microprocessor unit] prices tracks the PPI closely through 2008. However, from 2008 to 2012, our preferred index fell at an average annual rate of 39 percent, while the PPI declined at only a 9 percent rate. Given that MPUs represent about half of U.S. shipments of semiconductors, this difference has important implications for gauging the rate of innovation in the semiconductor sector.
None of this means Washington should forget about innovation and its key role in economic policy. Innovation is a critical lens through which proposed tax, regulation, and spending should be viewed. But this new paper does provided an important data point suggesting the era of US economic growth doesn’t need to end anytime soon.
Follow James Pethokoukis on Twitter at @JimPethokoukis, and AEIdeas at @AEIdeas.
