Response from Stephen E. Oliner and Daniel E. Sichel
回应Foley和Michl关于经济增长来源的评论，强调资本深化与技术进步紧密相连，并说明增长核算在识别增长直接来源方面的价值。
Duncan Foley and Thomas Michl highlight an important point about the sources of economic growth. As they note, capital deepening and technological progress-while often regarded as independent sources of growth-are, in fact, closely connected: Firms adopt more capitalintensive methods of production in response to technological advances that reduce the relative price of capital services. The rapid growth in the stock of information technology capitalespecially computing equipment-clearly illustrates this point. Firms have invested heavily in this equipment because new models provide greater power at ever lower constant-quality prices. These price declines have occurred largely because scientists and engineers have found ways to pack more and more circuitry on each square inch of silicon. At root, the growth contribution from information technology capital deepening has been driven by efficiency gains in the design and production of these technologies. Indeed, as Robert Solow (1970) noted long ago, the growth models that underlie growth accounting exercises such as ours have the property that steady-state growth in labor productivity equals the rate of technological progress. We agree, as well, with a second key point in Foley and Michl's letter-that movements in both multifactor productivity (MFP) and capital accumulation are relevant forjudging whether a New Economy has emerged. Even if the growth of MFP remained constant across a broad swath of industries, rapid growth in the stock of information technology capital could be taken as support for a New Economy view. This is so because the robust accumulation of information technology capital is being spurred by technological advances in the part of the economy that produces this capital. Our only disagreement with Foley and Michl concerns the rather negative tone of their comments about the usefulness of growth accounting. Their letter, in effect, restates the well-known critique that growth accounting decompositions can identify only the proximate sources of growth, not the ultimate sources. We also mentioned this limitation in our paper. However, even the limited objective of pinning down the proximate sources of growth can answer many interesting and important questions, particularly during periods when the economy is not in steady state. To take a couple of examples from our paper, the growth accounting framework allowed us to estimate the contribution from the boom in information technology capital spending during 1995-1999 to the observed acceleration in labor productivity. It also enabled us to estimate how much of the rise in MFP growth during this period was centered in the computerand semiconductor-producing part of the economy. Answers to these questions are important for understanding the mechanisms through which economic and technological developments have affected growth and for assessing the prospects for growth in the future.