How long before the jobs market pulls the inflation trigger?

One of the thorniest questions facing the Bank of England's Monetary Policy Committee as it decides what to do about interest rates this morning is just how tight the labour market has become. To put it in plain English, are there so few people available for jobs which employers want to fill that wages are going to start climbing faster? If so, it will be a pretty good sign that the monetary brakes need to be applied to the economy again.

This always seems insanely hawkish to anybody who looks around them and sees that there are still many unemployed people and so much poverty, but the real problem in economic policy is not how to expand demand for employees - that's easy. The trouble is that expanding the economy when the people without work cannot do the jobs on offer - because they do not have the skills or they live in the wrong part of the country - will trigger inflation. As we know from bitter experience in Britain, this harms long-term growth and productivity and penalises low earners and small savers. Matching the people to the jobs is a precondition for running the economy at lower interest rates.

Even though unemployment is still higher than in 1979, there are real fears that the five-plus years of growth since the depths of the recession have soaked up the supply of labour that is both available and suitable.

Start with the anecdotal evidence. One JobCentre in Milton Keynes has acquired fame by announcing that it has three jobs for every person on its register seeking work. Throughout the South-east - which is, unfortunately, where the jobs are being created for the most part - it is plain that employers are having to fill low-paid jobs for cleaners and staff in fast- food joints with foreign students and immigrants.

Earlier this week a new survey of the construction industry by the Chartered Institute of Purchasing and Supply reported shortages of almost every type of building worker imaginable, from brickies to the most skilled electricians. Other surveys have reported skill shortages - that is, worker shortages - in other industries from information technology to retailing.

Anecdotal evidence is not enough, however. What about the statistics? They have shown a modest pick-up in earnings growth as unemployment has declined. But most economists have been surprised and cheered by the fact that the fall in unemployment has been so great for so little acceleration in pay. In the 12 months to December the number of unemployment benefit claimants declined by about 470,000 - a jobless rate of just 5 per cent, or below what many experts would have considered to be "full" employment. Meanwhile, underlying average earnings growth climbed from 4.5 per cent in November 1996 to 4.75 per cent a year later.

Unfortunately both indicators are very flawed as measures of the tightness of the labour market, as the Office for National Statistics (ONS) has indicated by its announcement earlier this week of improvements to the published statistics starting in April. The claimant count measure of unemployment has been distorted - or fiddled, if you prefer - by at least 10 changes to the benefit rules since 1979, all except one tending to reduce the headline jobless total. An alternative measure of unemployment is available, based on the Labour Force Survey rather than figures collected from benefit offices, which conforms to the international definition of unemployment as the number of people seeking work and available to work within the next two weeks.

The two measures - the claimant count and the LFS measure - generally move in the same direction. When unemployment is high they are close together. When it falls they move apart, reflecting the fact that the drop in the LFS total is limited by the number of discouraged or marginal workers who start looking for work again during good times. The gap is now 500,000, about as high as it has ever been since the survey began.

Although a less distorted measure has been available, commentators have always focused on the claimant count. It is monthly, while the survey is published quarterly, and it is more up to date. In future the ONS will publish a rolling three-month average of unemployment as measured by the survey, so although it will still lag behind it will provide new and more reliable information about how many people are finding work - or joining or withdrawing from the workforce - each month.

The statisticians are also improving their presentation of the earnings figures. In the past they have based each month's "underlying average earnings" on a three-month average of the past two actual figures and an estimate for next month. This bizarre method is to be replaced by a rolling three-month average, which will simply smooth out volatile fluctuations in actual pay growth. This will mean fewer revisions, so each new figure gives a better guide to inflationary pressure in the jobs market.

The latest figures showed underlying earnings growth picking up to 4.75 per cent, driven by both the booming service sector and the subdued manufacturing sector. This rate is at the border of what the Bank of England has said it thinks to be consistent with meeting the inflation target.

The Monetary Policy Committee has, therefore, probably spent some time discussing whether the increase in earnings growth reported last month will be sustained or even accelerate further. A worrying light is shed on this question if private and public sector pay are distinguished. The pay bill freeze in the public sector means growth in earnings is holding down the overall average. Private sector pay growth has climbed well past 5 per cent, according to City economists who calculate them separately. From April, the ONS will itself publish rates for private and public sectors separately.

What this means is that the mini-miracle that some spy in the recent performance of the UK jobs market - falling unemployment for almost no cost in terms of higher wage inflation - is at least partly a sleight of the statistical hand. The figures that have been put under the monthly spotlight in the past have exaggerated the favourable split between jobs and pay. More important, they have hidden the fact that the good news is the result of a public sector incomes policy. That affects some 6 million out of 22 million employees in Britain.

Little wonder that Gordon Brown, the Chancellor of the Exchequer, and other ministers have been exhorting us all to restrain ourselves in demanding pay rises for the sake of the long-term health of the economy. The silent incomes policy will not be sustainable if private sector pay pulls ever further ahead of public sector wages. The improved ONS figures will make the chasm plain to see from April.

As for the Bank's decision today, burrowing into the labour market figures suggests that out of all the available inflation indicators they are flashing the most insistently red. Upgrading the skills and aptitudes of the 1.9 million unemployed in order to move the metaphorical traffic lights much further back is a bigger task which does not impinge on today's interest rate decision.