January 26, 2003
The Medium-Term Outlook for the World Economy: Five Questions That Need Answering

The Medium-Term Outlook for the World Economy: Five Questions That Need Answering

What one thinks about the medium-term future of the world economy depends on the answers one gives to five questions about productivity growth:

1. Rapid American productivity growth has continued through the recession. What conclusions should we draw from this? This question has two possible answers. The first answer is that changes in America's labor market have eliminated the old pattern by which firms tried to hold onto productive and experienced workers through the trough of the business cycle, because they knew they would be wanted soon and would be very expensive to replace. Such "labor hoarding" meant that measured productivity dropped in American recessions. Perhaps this institutional factor has now been erased from the American economy. However, if it has not been erased--if "labor hoarding" still exists--that means that the underlying productivity growth trend of the American economy has continued to accelerate, and that the future of Americna growth for the next ten years is very bright indeed.

2. Why are there such huge differences in productivity growth between the U.S. and mainland Europe--especially in information technology-heavy services? One possible answer is that European firms have, like American ones, invested heavily in information and communications technologies but that they have, in contrast to American firms, not yet figured out how to reorganize their work flow in a way to make their computers more than decorative paperweights. This suggests that Europe's problem is a deficiency of management consultants. The second possible answer--the one that Alan Greenspan strongly favors--is that firms will not undergo the unpleasant bureaucratic disruptions necessary to increase productivity until they can smell the profits from reorganization, profits that come either from reducing costs (by firing workers) or from meeting greatly expanded boom-time demand. Since the European Central Bank's policies make a boom in Europe nearly inconceivable, and since it is nearly impossible to fire workers on a large scale in Europe, no productivity boom.

3. How will the current intellectual property wars be resolved? Will they be resolved in a way that greatly increases the profits of CD and movie companies and that slows the adoption of broadband and other advanced information technologies? Or will they be resolved in a way that implicitly or explicitly confiscates a bunch of the intellectual property of CD and movie companies, but that gives consumers and other users enormous incentives to adopt broadband and other advanced information technologies? It is clear to me that the second would be better for economic growth, but that the first is more likely.

4. What is the size and salience of our current industrial revolution compared to past industrial revolutions? This is the one question to which the answer is clear. The classic British industrial revolution saw technological progress that reduced the price of making maybe 3% of GDP by 5% a year--a direct leading-sector technology-driven acceleration of 3% x 5% = 0.15% per year. Today's ongoing industrial revolution in information technology, communications technology, and biotechnology is reducing the price of making maybe 10% of GDP by at least 15% per year--a direct leading-sector technology-driven acceleration of 10% x 15% = 1.5% per year. Our modern leading sectors have ten times the technological-revolutionary relative salience of the leading sectors of the classic Britis industrial revolution.

5. The developing world: what does the forthcoming transformation of the service sector into a tradeables sector mean for the world economy? In the second half of the nineteenth century, the invention of the iron-hulled steam-driven ocean-going steamship meant that trans-oceanic trade was no longer restricted to the precious luxuries (silks and spices) and the biological products with interesting properties (coffee, tea, tobacco, sugar, opium) of the pre-industrial era. Instead, staple agricultural and industrial products--flour, meat, leather, steam engines, steel rails, locomotives, furniture, basic apparel, and so forth--became tradeable across oceans as well. Moreover, the trans-oceanic telegraph cable allowed for the rapid transmission of news around the world and for the first time made run-of-the-mill not-risk-loving investors comfortable making large scale investments in enterprises on other continents. The consequences were enormous.

Now we are looking forward to a world in which all kinds of white-collar paper-shuffling and information-processing--jobs that have been the monopoly of European-derived language-speakers located in the wrold's industrial core, and the monopolization of which has played a key role in boosting the standards of living of the industrial core's white-collar workers--can now or soon will be able to be performed anywhere in the world by anybody literate in one of the languages of the industrial core. What effect will this have on the distribution of income within the industrial core? What effect will this have on the rate of growth of the developing world. Even if first-world barriers to agricultural and textile and apparel imports remain, EC customs will be unable to restrict trade in bits over the internet: if Algerians are not allowed to grow citrus and export it to Europe, they will be able to process French-language documents.

Save for (4), I do not have nicely-potted answers to any of these questions. But on the answers to these questions hinges a great deal of our vision of what the world economy will be like in two decades...

Posted by DeLong at January 26, 2003 11:56 AM | Trackback

Email this entry
Email a link to this entry to:

Your email address:

Message (optional):


Why is net productivity rather than productivity per worker-hour used in determining the relative influence of IT in Europe?

Imagine that one company manages to create a perfect, fully roboticized factory. The workers get 10 hour per week maintainence jobs rather than 40 hour per week production jobs, and spend the rest of the time on the beach. Productivity does not increase, but productivity per worker does.

You can come up with a host of similar examples (I think), but it seems to me that if Europe has (1) more productive workers, and (2) a different preferred balance of money-time, the comparison of productivity is off.

I would like to know what I have missed.


Posted by: Brennan on January 26, 2003 08:12 PM

The US-Europe productivity gap is much smaller than believed - once you control for statistical differences in how GDP-per-hour-worked is derived. US price deflators are much more flattering to productivity than Europe's. EU GDP per hour worked has been growing only about 0.3%/year slower than in the US, a gap which may well prove transitory. Ditto for Japan.

Posted by: Peter vM on January 27, 2003 12:24 PM

"Moreover, the trans-oceanic telegraph cable allowed for the rapid transmission of news around the world and for the first time made run-of-the-mill not-risk-loving investors comfortable making large scale investments in enterprises on other continents. The consequences were enormous."
Actually this role was played by the house of Morgan- J. P. Morgan and his father operated the equivelent of investment banks on both sides of the Atlantic. I believe it can be said that much of the huge flow of investment into the US in the 19th century rested on British investors confidence in Morgan's character. It also allowed him to play the role of central banker in a time when the US had no cnetral bank. It was who made use of that telegraph (and quicker mail dleiveries) that mattered.

Posted by: Lawrence on January 27, 2003 06:19 PM

Five brief anwers for five interesting questions:

1/ Evolutionary theory may help us here. In times of slow change learned behaviour has a fitness premium, hence the importance of grandparents, older workers, or collective "wisdom". But in times of rapid and accelerating change - like the present - the premium for learned behaviour drops, and the advantage goes to rapid reactions on the fly. Hence the devaluation of the existing workforce. The future of growth is bright, for those who have work.

2/. Try looking back at Gregg Clark's piece about cotton machinery in UK, US and India: the machines were the same, but the productivity was way different. What cultural factors may be at work here?

3/. Now try an analogy with bacteria and genetic mutation. My guess is that the CD and movie industries are going to have a hard time of it given that the rules of the game keep changing. Right now Verizon is trying to shield Kazaa. And if they fail, presumabely someone will be bright enough to invent a software which better protects 'privacy', or maybe the Verizons of this world will disappear under the weight of wireless P2P and those coming enormous hard drive storage capacities. Who knows, but the smart money says the copyright protection people will be one step behind and not one step ahead (in part for reasons already explained in answer to 1 and 2).

4/. The current INDUSTRIAL revolution is going to dwarf the previous ones. Only one detail, it's happening in China (and principally for demographic reasons).

5/. Tradeable and automated services are going to be important, but don't forget with an aging population one hell of a lot of the growth in services will be in-situ: medical, domestic and care related etc. This also speaks for the global equalisation of incomes, but in this case the work doesn't go to the people, the people come to the work.

Posted by: Edward Hugh on January 28, 2003 03:21 AM


I agree with the above posts in that it seems weird to obsess so much with faster US productivity growth when a)it's only a phenomena since 1995, b) affected massively by the method of calculation and c) not so hot if you use TFP rather than just labour productivity


Posted by: James on January 28, 2003 09:07 AM
Post a comment

Email Address:



Remember info?