The jobless total has now been falling for nearly two years, having peaked at just under 3 million. But in only two quarterly periods during that time have companies told the Department of Employment that they have taken on more new staff than they axed.So while the jobless total dropped by 350,000 between the end of 1992 and last summer, employers claim to have shed 70,000 more workers than they took on.
This always looked suspicious. It implied that unemployment was falling because huge numbers of people were withdrawing from the labour market altogether - choosing neither to take a job nor to claim benefit. But evidence from the rest of the economy suggested this was most unlikely.
For one thing, the economy's production of goods and services is now growing by more than 4 per cent a year. This is well above the 2-2.5 per cent rate that has been enough to stimulate job creation in the past. The Chancellor of the Exchequer and his chief economic adviser, Alan Budd, have both told the Treasury Select Committee in the last few days that this trend growth rate may have risen because of the labour market reforms of the 1980s. But, even if they are right, the economy has surely been growing quickly enough for long enough to have generated hundreds of thousands of net new jobs by now.
The picture of continued job shedding painted by the survey of employers has also been at odds with the Employment Department's Labour Force Survey (LFS). This measures employment by asking a sample of 60,000 people every quarter whether they have work or not. It showed a 250,000 rise in employment between the end of 1992 and last summer.
The surge in job creation reported by companies in the third quarter means that both measures of employment are now moving in the same direction, enabling us to be much more confident that the recovery in the labour market is sound.
Michael Portillo, the Secretary of State for Employment, argued last week that the employer survey might be slower than the LFS at picking up jobs created by new businesses, because the Inland Revenue took some months to process their Paye registrations.
Companies may also be reluctant to admit to creating jobs that involve short-term contracts, or offer only a few hours' work a week to avoid the need to register for PAYE or to pay employees' national insurance contributions.
But we cannot be certain that the employer survey will remain upbeat. The biggest component of the third- quarter employment rise is the supposed creation of 52,000 new jobs in banking, insurance and finance, which looks rather implausible at a time whenthe likes of Barclays and TSB are "downsizing". The Employment Department will also recall that there was a surge of 101,000 in employment in the same quarter last year, which was more than reversed just six months later.
Further doubts are raised by evidence from the LFS that the total hours worked in the economy on average each week dropped from 816 million in the spring to 813 million in the summer. This suggests demand for labour may even be ebbing, in which case the third-quarter surge in employment is probably either illusory or dominated by part-time jobs.
None the less, it does look on balance as though the labour market is tightening. The number of people who said in the LFS that they were not looking for work because they believed there were no jobs available was 11,000 lower in the summer than a year earlier at 170,000. And the filling and notification of new vacancies to Job Centres are running at four-year and five-year highs respectively.
So what are the implications for policy? Job creation and falling unemployment threaten upward pressure on inflation, if they mean that employees push more aggressively for pay increases. Pay demands will rise if the jobless total is pushed below the so-called "natural" rate of unemployment, putting upward pressure on costs, forcing suppliers to charge more for their goods, and thus generating a spiral of relentlessly rising inflation.
The US Federal Reserve takes the natural rate of unemployment very seriously as an indicator of inflationary pressure. But the "natural" rate of unemployment appears to have been much less stable in Britain than in the US. This helps explain why the Treasury and the Bank of England pay more attention to the dreaded "output gap" - the difference between the economy's output and what it could produce if available labour and capital equipment were being used efficiently. In theory, however, the concepts are equivalent: unemployment should be stable at its natural rate when there is no output gap and the economy is growing at its long-term trend rate.
To the extent that they are meaningful concepts at all, the output gap and the natural rate of unemployment are almost impossible to measure with any accuracy. Most estimates put the natural rate at about 8 per cent of the workforce in Britain, compared with an actual unemployment rate of 8.8 per cent. If so, the pace of the fall in unemployment would seem to justify the recent gentle upward nudge to interest rates.
But some economists, such as Patrick Minford at Liverpool University, believe the natural rate is far lower than 8 per cent and that the Chancellor is therefore taking unnecessary risks with the recovery by raising rates.
Kenneth Clarke's decision to include a package of work-incentive measures in the Budget reflects a concern that the natural rate is still quite high and has to be reduced. As he said in his Budget speech: "We have to do more to reduce unemployment in ways which are consistent with sustained growth and low inflation."
Several of the Chancellor's measures were designed to boost employment by cutting the cost to firms of creating low-paid jobs. These include more subsidies for companies taking on the long-term unemployed and cuts in employers' national insurance contributions. Other measures sought to ameliorate those features of the social security system that might deter individuals from taking low-paid jobs or from working longer hours.
The Treasury has been very cagey about the impact it expects these measures to have on the jobless total. Officials are estimating tentatively that they could take around 100,000 people off the count. Mr Portillo conceded last week that the net effect might be small, as the long-term unemployed might get jobs at the expense of people already in work or the short-term unemployed.
But this would be no bad thing. Cutting long-term joblessness cuts the natural rate of unemployment, even when overall unemployment remains the same. This is because the long-term unemployed become demotivated and increasingly unattractive to employers, who assume their rivals have chosen not to employ them for good reason. So "churning" the stock of unemployed people should mean they exert more downward pressure on wages - and therefore inflation - by enabling them to compete more effectively for available jobs.
Mr Clarke's package is a step in the right direction. The Labour Party rightly argues that the package is not ambitious enough and that it is being accompanied by undesirable cuts in spending on training. But it would be well advised to greet the principle more enthusiastically than it has so far done. Only with a bipartisan commitment to further cuts in the natural rate of unemployment can the British economy sustain a recovery that will generate the fabled "feel good factor" for which we are all waiting.Reuse content