October 23, 2003

Dale Jorgenson Reworks the European and Japanese Data

He finds that the information technology revolution is taking hold throughout the OECD--and has been since the mid 1990s:

Economist.com | Economics focus: ...Mr Jorgenson uses data for Europe and Japan which are adjusted to incorporate price deflators and measures of software expenditure similar to those used in America. Unfortunately, the detailed information needed to make these adjustments is available for most economies only up to 2000. Even so, the results are striking. For instance, they suggest that Japan's GDP grew by an annual average of 2.1% in the second half of the 1990s, compared with only 1.4% according to official statistics.

Employing these revised data, Mr Jorgenson finds that in all G7 economies, not just America, a boom in IT investment helped to boost growth in the second half of the 1990s. Indeed, the contribution to GDP growth from IT capital spending was almost as big in Japan as in America%u2014although it was offset by a fall in investment of other sorts. All of the European economies also saw a marked increase in their IT capital stock, albeit smaller than in America. As in Japan, in many European countries this was partly countered by weaker non-IT investment.

The second factor that other countries have in common with America is that during the second half of the 1990s all of them, except Italy, enjoyed faster productivity growth--if it is measured properly. The most popularly used measure is labour productivity (output per worker-hour), which slowed in Europe and Japan after 1995. But a better measure is "total factor productivity" (TFP), which captures the efficiency with which both capital and labour are used. Output per worker-hour can be boosted simply by giving each employee more capital to work with, as well as by genuinely improving efficiency. Since 1995, TFP growth has accelerated almost everywhere.

Despite all the talk about Japan's inefficient economy, in the second half of the 1990s its TFP growth quickened to 1.1% a year, much faster than America's 0.6%. Official statistics imply that Japan was slow to adopt IT. But the internationally comparable data suggest that in the late 1990s the total direct contribution of IT to GDP growth (investment in computers and software, and TFP growth in the IT sector) was roughly the same in America and Japan (see chart). Japan's GDP growth was slower because of a fall in non-IT investment and weak employment due to deficient demand. Britain, Germany and Canada also enjoyed rates of overall TFP growth close to America's. Curiously, productivity in the IT sector actually grew faster in Europe than in America, but this was offset by slower TFP growth in other industries.

In other words, America's growth resurgence due to IT investment is not unique. The snag is that elsewhere it has been partly disguised by the poor performance of investment in other things. What about the period since 2000? Although in the five years before that date virtually all of America's increase in labour-productivity growth came from either increased IT capital or productivity gains in the IT sector, TFP growth has since quickened in other areas as well. Will this be repeated in Europe and Japan? Recent studies suggest it takes time for firms that invest in computers and software to work out how to reorganise their business practices and to retrain staff in order to reap the full efficiency gains. If so, these benefits may yet appear in Europe and Japan. America was the first big country to embrace IT, so it is hardly surprising that it has been the first to benefit...

There are subtleties in the measurement of total factor productivity that I do not think the Economist's writer has gotten correct. Suppose that total factor productivity in making computers booms: the economic growth produced by higher computer-manufacture productivity is ascribed to higher TFP. But this boom in computer-manufacture productivity will reduce the price of computers. Businesses will find that their computer-buying budgets go further: they will buy more computers, and thus make their workers more productive. This is an increase in labor productivity. Is it an increase in TFP? The answer is either "No" (because the key is having more computers, not using your current computers more effectively) or "Yes" (because the only reason you have more computers is because of higher TFP in computer-manufacturing). How, exactly, you do your growth accounting is important. And I haven't cracked Dale's paper yet to figure out exactly how he does his. Posted by DeLong at October 23, 2003 11:00 AM | TrackBack


Here is an article by a fellow Berkeleyan:


The Mixed Bag of Productivity

RECENTLY productivity has been growing at a rate of about 4 percent a year. For the country as a whole, this means that each year we can work as much as we did last year and consume 4 percent more; or we can consume as much as we did last year and work 4 percent less.

That's got to be a good thing, right?

Well, it depends on whom you ask. In truth, those productivity gains have resulted in some people's working 100 percent less, with the rest of us consuming 4.01 percent more. If you are one of the unemployed, chances are you are less enthusiastic about the productivity gains than are those who have enjoyed the increased consumption.

Strangely, productivity growth is not getting much blame for the "jobless recovery." Criticizing technological progress is downright un-American. On the other hand, criticizing foreign trade is a traditional pastime here, as in every other country.

To economists, trade and productivity growth have a lot in common: each allows you to produce more with less.

James Ingram's economics text tells the story of how an entrepreneur built a factory that was significantly more productive than his competitors' plants, and was hailed far and wide for his brilliance.

But then his dirty little secret was revealed: all he was doing was importing goods from abroad through the back door....

Posted by: anne on October 23, 2003 11:23 AM
Post a comment