It is a great pleasure to be here today and to say a few words about the meaning and importance of the GDP data and the National Income and Product Accounts. I would like to make some personal observations, based on the ways in which these data have been important to my work.

I started my career as an academic researcher with a strong interest in macroeconomics and it is hard to imagine how I or anyone else could have talked about the U.S. economy and the business cycle without timely and accurate information about GDP—or GNP as we used back then. One issue that I explored twenty years ago was the extent to which the economy has become more stable over time.

I carried out a series of econometric tests examining the response of consumption, investment and inventories to cyclical shocks. But it turned out the most compelling evidence came from simply plotting the growth rate of GNP over time. The resulting chart, subsequently reproduced in the New York Times, showed a dramatic decrease in the volatility of GNP in the postwar period.

I argued, as I still believe, that sound and cautiously active monetary and fiscal policy, together with automatic stabilizers, have been important to the increased stability of the macroeconomy. Others have disagreed as to whether policy is really stabilizing and even whether the economy has become more stable. But of course this debate would not have been possible without good historical GNP data.

More recently I, along with a team of researchers, have worked on a number of studies trying to understand the differences in productivity across countries. This work has been based on the scrutiny of very detailed, micro-level firm and industry data. But each study began with an aggregate analysis that featured GDP per capita as the best overall measure of economic performance across countries. We selected industries to study largely on the basis of whether they would add to our understanding of cross-country GDP per capita differences. By starting with a micro data base and building up to tell a macro GDP level story, I believe this work has added to understanding of the reasons for economic performance differences across countries. It would not have had the same impact or validity without good underlying GDP data.

A major task in my current position is to work with others in preparing the Administration's forecast for budget purposes. GDP and its growth over time are the centerpiece of this exercise. We recognize the tremendous uncertainty in trying to predict GDP ten years into the future and the need to make policy decisions which recognize that uncertainty. But we rely on a solid starting point for our work - the GDP data prepared here at the Department of Commerce.

Working with our forecast, however, makes me realize the need for continued progress. Why was there a slowdown in productivity growth in 1973? Has the trend of productivity growth increased in the 1990s? How is the digital revolution affecting businesses and consumers? Does the rise in the stock market reflect an increase in intangible capital accumulation by companies? These are tough questions, central to any forecast, and finding the answers would be helped by better GDP data.

Major improvements have been made—the shift to chain indexes, the use of better price indexes both by BEA and the Bureau of Labor Statistics and the incorporation of software investment. But more needs to be done to capture a rapidly changing economy where services account for much of GDP, where the digital revolution is in full flight, and where quality changes may be as important as quantity changes. If our statistical agencies are to keep up with the demands of the economy they need to be adequately funded to support the wealth of commitment and expertise of the people that work there.

My experience as an academic, as an economist in the private profit-making sector and as a member of the Administration tells me that good GDP data are vital to high-quality research, a greater understanding of the U.S. economy, and, ultimately, to sound decision making.