New paradigm theory gains ground

US productivity growth has exceeded 4% per annum, which is statistically unusual

THE POPULARITY of the "new paradigm" in American economics is one of the more extraordinary developments in the recent history of the subject.

Essentially a populist "theory" intended to explain the American "miracle" of the 1990s, it has until recently been difficult to find any serious macro-economist who has been willing to give it houseroom. As so often, Paul Krugman of MIT puts it best: "The reason I can't buy into the New Economy is that I can't bring myself to endorse a doctrine that I know to be just plain dumb."

The general view in academia is that the new paradigm will suffer the same fate as the "supply side economics" of the 1980s - just another passing fancy of the type which takes hold more easily in American society than in Europe.

Yet most business executives in the United States are convinced that a combination of globalisation and information technology has transformed the rules of the dismal science, especially as they apply to the western half of the US.

The battle between economic opinion and business opinion is neatly encapsulated on the newstands. The Economist says that the new paradigm "verges on claptrap", while Business Week reckons it heralds nothing less than "a new kind of math", which explains why the economy can sustain growth at 4 per cent per annum. And recent economic data in the US has been so encouraging that some very respectable economists, such as Alan Blinder, have started to ruminate about whether they may have been too hasty in being so rude about the whole idea.

The new paradigm comes in many shapes and sizes but, as Alan Blinder explains in a penetrative daily briefing published by the Group of Seven on 20 May, there are two basic ingredients which appear in all of its varieties.

The first is that the economy can sustain a much lower level of unemployment than in the past, without this leading to an acceleration in inflation. In other words, the unemployment rate which is likely to emerge on average over the whole economic cycle (variously known as the natural rate, the structural rate or the Nairu) has fallen from the 6 per cent estimated since the late 1970s to something much closer to today's 4.3 per cent rate.

This in itself would be important enough, but the second ingredient in the new paradigm is even more significant. It says that the underlying growth rate of productivity in the US economy has risen sharply in the 1990s.

The first ingredient is much less revolutionary, and much easier to accept, than the second. It is less revolutionary because large changes in the Nairu have occurred repeatedly in developed economies in the last 30 years, first in an upwards direction and then more recently (notably in the UK) in a strongly downwards direction.

Although the Nairu in the US has been much more stable than in other countries, a move of a point or two in either direction would be well within the confines of past experience. This is the sort of move that might be explained - stretching things a little - by the impact of globalisation on wage behaviour in the labour market or, more likely, by the rise in competition in product markets, holding down inflation for any degree of unemployment.

Even as recently as a year ago, there seemed little need to speculate in these directions, since wage increases were rising gently, just as they should have been if unemployment had fallen below the Nairu. But in the past few quarters there has been a remarkable deceleration in wage increases, despite the continuing drop in unemployment.

The jury on this issue will still be out for several more quarters, but even if it turns out that the Nairu has dropped by (say) 1.5 per cent of the labour force, we should be clear about the long-term effects.

These will be highly beneficial, but not transformatory. Of course, no- one should snub their noses at a permanent drop in the unemployment rate of 1.5 per cent, equivalent to a decline of one quarter in the jobless total over the cycle. But the GDP effects of this change will be limited, probably raising the level of output by around 1 per cent on a sustainable basis.

If the adjustment to this new Nairu is spread evenly over three years, it would boost GDP growth by 0.3 per cent per annum over this period. Unfortunately, though, the growth rate would inevitably return to normal once the adjustment phase was over - not much of an economic miracle.

The second ingredient in the new paradigm is, however, much more revolutionary. The contention here is that the long-term growth rate of productivity in the economy has risen sharply as a result of the cascading technical changes in information technology. Until now, mainstream economists have been exceptionally united in agreeing that the trend rate of GDP growth in the US is around 2.5 per cent, stemming from an increase in the labour force of about 1 per cent a year, and a rise in productivity of about 1.5 per cent a year.

Now consider what would happen if the trend in productivity growth rose by just 0.3 per cent per annum. After three years the level of GDP would be 1 per cent higher than otherwise, which is exactly the same effect as the drop in the Nairu discussed above. But the key point is that the growth rate would remain 0.3 per cent higher in perpetuity, so the impact on output would continue to accumulate.

After 10 years, GDP would be 3 per cent higher, over 30 years 10 per cent higher - a revolution indeed.

The only problem with these starry-eyed calculations has been that for most of the 1990s, there has been no sign of anything untoward happening in the US productivity statistics.

In January, Ed McKelvey of Goldman Sachs wrote The US Productivity Paradox - Miracle or Mirage, a masterly summary of the productivity debate, completely devoid of the usual hysteria and political overtones. Anyone with a serious interest in the new paradigm should read this paper. His conclusion? Simply that the strong rates of growth in US productivity in the second half of the 1990s were a normal catch-up after abnormally sluggish productivity growth in the early 1990s.

As the graph shows, the level of productivity in the US is still no higher than it "should" be, based on past experience. But the new factor is that the growth rate of productivity in the past few quarters has been exceptionally high, especially for such a late stage in the economic upswing.

In the past two quarters, productivity growth has exceeded 4 per cent per annum, which Goldman Sachs reckons is statistically unusual, given the state of the economic cycle.

And Mr Blinder now says that these data "raise for the first time the real possibility that trend productivity may have accelerated markedly - just as the New Economy crowd has claimed".

The next few quarters of productivity data will be crucial. For the time being, it is probably wise to remain a sceptic. But unless there is a marked slowdown in American productivity growth soon, some very famous economists may yet be forced to eat some very humble pie.