ECONOMIC VIEW : How much unemployment comes naturally?

Unemployment remains the favoured subject of study of academic economists. The OECD in Paris, always a solid indicator of the economic mainstream, almost appears to be becoming a jobs think-tank. The much- discussed jobs report of a year ago has continued to grow, with spin-off reports on particular themes to emerge soon in public; and with special studies of each member country to be added to the private collection of the organisation as well.

And at home, two weighty studies of unemployment have been added to the collection of material already available. One is the spring "Unemployment" edition of the Oxford Review of Economic Policy. Interestingly, almost a quarter of the journal was written by the prolific economists of the OECD. Of the rest, the most helpful statistical work concerns employment rather than unemployment, in Paul Gregg's and Jonathan Wadsworth's fresh documentation of the decreasing job tenure of the UK labour force.

The other recent study is a pan- European collaborative work from the Centre for Economic Policy Research, Unemployment: Choices for Europe. The UK representative was Charles Bean, of the LSE.

It would be fair to say that much of what is contained in these reports - and indeed in almost everything published on the subject - has a familiar ring to it. But there is one topic that is missing from almost all that is written. It does get a mention in the CEPR report, but it is a cursory and unwitting mention.

The missing issue is what level of unemployment we should consider to be full employment, or satisfactory. Some people appear to think there must be a job for all - they aim for zero unemployment. Others tend to hark back to the post-war boom, when unemployment would hit 2 per cent in recession.

Some, more realistically, take 3 per cent as a working definition of reasonably full employment - in line with the Beveridge report. On this latter basis there are people representing 5.5 per cent of the labour force whom we might expect to have a job but who don't have one.

Then there is the alternative notion of economists, who talk about the natural rate of unemployment. That is the rate that is compatible with steady inflation; it's estimated at about 8 to 10 per cent of the workforce in Europe, and about 6 per cent in the US.

Anything above that, and you are in an economic dip: anything below that, and inflation will be rising as workers no longer fear for their job, and the economy overheats. Assuming Britain has a natural rate of 8 per cent, we are only about half a per cent adrift at the moment.

But while the natural rate is a useful concept - it tells us not to overheat the economy to get unemployment down below quite high levels - it is not a level to which we should be resigned. We can always affect the natural level itself by taking policy action on incentives and training.

So the missing concept is the realistic or desirable level of the natural rate itself. What is the natural, natural rate? Is it 3 per cent or something different? In a telling assumption, made in the CEPR unemployment report, the achievable natural rate is taken to be 6.4 per cent. That is the level to which it is assumed reasonable labour market reforms could get unemployment.

The assumption was made for the French economy , and based on the fact that America's deregulated labour market achieves unemployment close to that in a normal year, and on the fact that among well-to-do workers, French unemployment is 6.4 per cent. None of the problems of low pay applies to this group, so the usual co-factors associated with a high unemployment risk are not involved. The 6.4 per cent rate is somehow fundamental.

Is this assumption reasonable? Although on the high side - the French economy is operating well below capacity - the answer must be yes. Historically, unemployment has been at levels much higher than 3 per cent.

Between 1885 and 1929, excluding world wars and the depression, unemployment averaged 5.3 per cent in Britain. The US rather worse before the First World War, and a little better after. All in all, then, 5 or 6 per cent seems to have a more natural precedent than either 3 or 8 per cent.

Alternatively, look at other markets and the usual degree of slack in those. The commercial property sector, for example, operates in a rather similar way to the labour market. There are significant "hiring and firing" costs in both - so companies usually don't hire property by the day any more than they hire labour by the day.

Well, I am told by those who work in it that a typical period in that market might see about 6 per cent of buildings unemployed, or vacant.

When occupation levels get above that, then new building starts and a boom begins. Indeed, in many sectors it would be a dream to operate at levels of more than 94 per cent of available capacity.

All this suggests a fundamental rate of unemployment of about 5 or 6 per cent. That level can be explained partly by people searching for jobs, partly by people being largely unsuited to work for personality reasons, but also from the fact that market frictions ensure the economy naturally operates with a 1 or 2 per cent degree of inefficiency.

If you imagine a giant, benevolent economic planner in the sky, sorting out the economy, you can imagine that getting every last person into a job would be a tall order. The same is true for an invisible hand.

As more and more people have jobs, it becomes more and more expensive for employers to recruit. At some point they don't try hard. At that point, you still have a few people out of work.

Under this view, the historical aberration was the period of very low unemployment enjoyed after the war. As much work could be done to explain that as is done to explain the high levels of unemployment today.

Most of all, if the conjecture about the natural, natural rate of unemployment is correct, then the big task is simply explaining to the public what realistic hopes they should have when it comes to solving this problem.