The second half of the 1990s saw an astonishingly large productivity boom in America. Moreover, the productivity boom did not come to an end when the recession began. Only a strong underlying tide of increased efficiency and productivity from the use of high-tech computer and communications equipment can reconcile the steep American "unemployment recession" coupled with the non-existent "output recession": The U.S. unemployment rate has risen by two full percentage points over the past six quarters, yet production has not fallen at all, but has risen by 2.4%.
But why was the boom so tightly confined to the United States? Why wasn't it a world-wide boom? Why wasn't it at least an industrial core-wide boom? There is, after all, nothing in the air or water that makes high-tech equipment work better in San Francisco or New York than in Frankfurt, Lyons, or Edinburgh.
The past four years have seen analysts puzzle over slow relative growth--especially slow relative growth in Europe--outside the United States. Some analysts from the Bundesbank suggested that it was all due to the undercounting of production in Europe, but this had a hard time explaining the component of the U.S. productivity acceleration found not in computer-making but in computer-using industries. The U.S. Federal Reserve moved toward a consensus that the real problem was too much European red-tape--that businesses invest heavily in high-tech equipment when they can smell immediate productivity gains from reorganization and restructuring, and that western Europe's steps toward economic reform and liberalization have so far been too small for businesses to be confident that they will be allowed to reorganize and restructure. The lost tribes of Keynesians offered an opposed diagnosis: that businesses invest heavily in high-tech equipment only when they need to satisfy increased demand, and European central bankers unwilling (as Alan Greenspan was in the 1990s) to take risks on the inflation side in order to expand employment have greatly slowed the transformation. Still others said that the U.S. boom was oversold--that because of the NASDAQ bubble the U.S. had pushed employment too high and invested too much in telecom infrastructure and computer equipment that would never be useful for much of anything, and that the U.S. boom was only a boom in production, not an increase in economic welfare.
It would be unlikely if any of these stories were completely false: some socially-unproductive investments were made during the NASDAQ bubble, some European firms have held back on investment plans because of slack demand, others have held back on investment plans because of various obstacles to economic flexibility, and there are important differences between American and many European statistical systems with respect to their ability to track economic growth.
Yet I cannot help but be reminded of America just a decade ago, when commentators and analysts speculated that all the investment in high-tech equipment the U.S. was undertaking was wasteful and dissipative, and people quoted Robert Solow's question, "Why do we see the computer revolution everywhere but the productivity statistics?" to each other. People in Washington D.C. observed the ferment of innovation in Silicon Valley, but dismissed the possibility that it would have a large macroeconomic impact. After all, Moore's Law of steadily and rapidly increasing computer power had been at work for three decades already, without having any effect on aggregate productivity.
What the conventional wisdom of a decade ago in America missed was that you do not expect growing sectors--even rapidly growing sectors--to have any impact at all on aggregate statistics until they achieve critical mass. This is an old lesson, first taught us by the observation that historians of technology think the British Industrial Revolution took place between 1780 and 1830 (when its key inventions were made) while social historians think it took place between 1830 and 1870 (when the new industries of steam, factory, and iron achieved critical mass, and turned Britain from a nation of shopkeepers to a nation of factory workers and industrialists).
And, indeed, a decade ago the Federal Reserve's Daniel Sichel pointed out that it should have been no surprise that back then the aggregate impact of the computer revolution was small: it was, after all, proportional to how fast computer prices were declining multiplied by the share of national income spent on high-tech multiplied by the share of total production attributable to the use of high-tech; all three of these numbers were not very large (although growing rapidly); and the product of three not very large numbers will be a small number. However, as computers and communications continued to diffuse throughout the American economy, expenditure on high-tech and the share of income attributable to high-tech continued to grow and did achieve critical mass. The result was the American productivity boom that is still ongoing.
Isn't the way to bet that much of western Europe now is in the same situation as the U.S. in the early 1990s, simply waiting for the next turn of the business cycle and the continued diffusion of technology to push high-tech's salience above the critical mass at which its macroeconomic effects become obvious? Already no one would deny that the computer and communications revolution are having their transformative effects on the economies of Finland, Sweden, Ireland, Australia, and perhaps one or two others in addition to the United States.
Isn't the way to bet that the rest of western Europe will be relatively close behind?