Redundancies always have to be taken seriously, of course. But the trouble with monitoring company announcements is that it is difficult to be systematic. People pick out what they want to see, and many see only confirmation of their conviction that British industry is sliding down a sad spiral of decline. But one day's worth of bad news does not paint a realistic portrait of the job market.
A month's worth of newspaper reports presents a slightly fuller picture. The table shows the job losses and gains that made the news in June - only a tiny fraction of the actual totals. Even so, it shows that job losses and gains in a month chosen at random were pretty evenly balanced. In fact, there were slightly more new jobs than redundancies, although, not surprisingly, many more column inches written about the latter.
As the Stanford economist Paul Krugman has made it his mission to point out, you cannot base a sensible view of the world on headline stories about big companies. The reason, as far as the outlook for jobs and unemployment goes, is that flows of jobs created and destroyed in any given period are far greater than the net increase or decrease in employment. Rates of job turnover in industrial economies are astonishingly high. OECD figures show that total turnover or reallocation of jobs for most countries - that is the destruction plus creation of jobs as a proportion of the workforce - was in the range of 12-20 per cent from the mid-1980s to 1992. There was no marked transatlantic difference in turnover. The one exception was the UK, with an unusually low 9 per cent. And even that corresponds to about one in 11 jobs (that is, around two million) being destroyed or created every year. The high turnover rates compared with rates of net job change of between minus 2 and plus 4 per cent during the same years. It adds up to a lot of churning.
A new book* explores job turnover patterns in American manufacturing, toppling several myths in the process. One myth is that small firms create more jobs. This is true in the gross sense but not net, for small firms destroy a disproportionate number of jobs too. The authors write: "Because high job creation rates typify employers of all sizes, and because the manufacturing jobs base is dominated by large employers, large employers account for the bulk of job creation (and destruction)." The average US manufacturing plant has 1,600 employees.
Standard theories of the business cycle, which argue that economy-wide shocks cause aggregate fluctuations, are debunked too. The cyclical behaviour of employment across industries is extraordinarily varied. They do not react the same way in recessions.
Recessions are marked by a sharp increase in average job destruction rates and little change in job creation rates, meaning that job turnover rises too. The variation in job destruction over the business cycle is more pronounced among bigger and older firms. Smaller and younger firms display a much weaker pattern of boom and bust.
These facts suggest that a recession is a period of faster industrial restructuring rather than simply an economy-wide reaction to a common shock such as higher oil prices or a surprise increase in interest rates. An event like this is a trigger, but individual industries and companies react differently. The pace of restructuring subsides during a long recovery. The long-term trend decline in employment in manufacturing is due to slower job creation rather than faster job destruction.
As changes in the underlying job opportunities account for about half of the moves workers make between jobs or into and out of unemployment, unemployment also rises during recessions. During a recovery, moves into unemployment are dominated by new entrants and re-entrants to the labour market. During a recession, the rise in job destruction is the main reason.
Europeans, however, are far less likely to move from job to job via a short spell of unemployment, so even during good times inflows into unemployment are mainly due to job destruction. In America about one in 12 of the workforce moves into unemployment in an average quarter, and about the same proportion moves out. Both flows into and out of unemployment rise during a recession.
The study, based on the first comprehensive exploration of detailed statistics on individual US manufacturing plants during 1972-88, finds that rates of job creation and destruction are remarkably large. About one in 10 manufacturing jobs disappears every year and about as many new ones are created. The minimum in any year was one in 12.
The changes are concentrated on particular plants: two-thirds at plants which are expanding or contracting by more than 25 per cent. Plants that close account for fully a quarter of job destruction. "The message is plain: job creation and destruction in the manufacturing sector often involve dramatic events such as the start-up of a new plant or the death of an old plant."
Large scale job flows characterise all sectors of industry. Even shrinking industries display massive job creation and destruction, although the rate of turnover varies widely between industries.
However, high wage and capital-intensive industries experience smaller gross job flows. High wage jobs - typically in big firms - are more durable, and there is much faster net growth in jobs in high wage and high productivity industries. The figures confirm that job opportunities for the low-paid shrank during the 1970s and 1980s.
The book concludes: "This large-scale, pervasive, and continuous reshuffling of job opportunities bespeaks a constant reallocation of production activity in the US ... and other advanced market economies." What does this picture of constantly shifting activity, mainly within but also between industries, imply for economic policy?
For one thing, it highlights the pitfalls in the very popular types of policies targeted on particular categories of company - say small businesses or companies investing in particular regions. For businesses are likely to have extremely diverse reactions to the incentives on offer. The outcome in terms of the number of jobs created is unpredictable, with individual companies likely to behave in entirely different ways, the new research suggests. It also indicates that big businesses will create more and longer- lasting jobs than small companies.
Secondly, because workers face a high risk of their job vanishing - a 10 per cent chance in any year - there is a great advantage in having a workforce that is flexible in the sense of having strong basic skills. There is no sense in training people in specific skills because they could well find the relevant job opportunities vanishing.
The third conclusion is that governments should think carefully about the detailed impact of policy measures. Cuts in different types of spending such as defence or roadbuilding will trigger restructuring in certain industries or regions. These could have a bigger impact on jobs than changes in interest rates or taxes.
The final moral is that the constant and massive turnover in jobs - which Joseph Schumpeter, the favourite economist of free-marketeers, would have claimed as part of the "creative destruction" of capitalism - means it is impossible to draw any conclusions from one day's headlines. Dull as it is, we will have to wait for official statistics on the net change in employment to take the economy's temperature.
* Job Creation and Destruction, Steven Davis, John Haltiwanger and Scott Schuh, MIT Press, June 1996, pounds 23.50.
JOBS IN JUNE
300, US compressor manufacturer Copeland to build plant in Ulster
200, Elf and BP polypropylene joint venture in Grangemouth, Scotland
125, Japanese machine tools maker Matsuura opening plant at Coalville near Leicester
250, Sony says it cannot fill these new posts in South Wales TV monitor plant
1,000, Interconnection Systems, UK maker of printed circuit boards, to build pounds 120m Tyneside plant
300, Quintiles, US pharmaceuticals company, in new Scottish factory
Total gains: 2,175
20 axed by IBM in Greenock, Scotland
Up to 1,000 to be cut by retailer WH Smith in restructuring
900 redundancies in Hyder post-merger shake-up
Total losses: 1,920