Economic Policy Panel
Trade policy and unemployment

Work-sharing, unemployment, trade policy and fiscal policy were among the topical European policy issues discussed by the Economic Policy Panel at its third meeting in Paris on April 24- 25. The Panel, comprising leading European and American economists, meets twice a year to discuss specially commissioned analyses of important economic policy issues. The papers discussed at the Paris meeting will be published in the third issue of the journal, available in October 1986. The first two issues of Economic Policy have already established it as the leading European forum for the discussion of topical policy issues.

In the first paper presented to the Panel, entitled 'Work Sharing: Why? How? How Not ...', Jacques Dreze (University of Louvain) argued that work-sharing measures, although desirable in principle, have been of rather limited effectiveness in combating European unemployment. Dreze argued that it is useful to contrast 'regular' jobs and 'casual' jobs. Because 'regular' jobs typically involve substantial hiring and training costs, the number of such jobs offered by firms is unlikely to respond to short-run fluctuations in wages, and pay in such jobs will not vary with economic conditions. The position of those with 'casual' jobs is quite different, however. During prolonged periods of recession, firms require fewer hours to be worked, and this reduction is borne largely by people without 'regular' jobs. These individuals cannot price themselves into work by offering to work at lower wages unless firms' sales prospects improve.

There is therefore, according to Dreze, a need for government measures to encourage work-sharing. This could take the form of encouraging early retirement, increased use of part-time workers, or a reduction in working hours. European governments have experimented with each of these approaches, but the only schemes which seem to have had any success are those such as the UK Job Release Scheme, which encourage early retirement in conjunction with mandatory replacement. In schemes where replacement is not mandatory, as in the Netherlands, only around 10% of the workers who retire are replaced. Measures to encourage part-time work have not reduced unemployment, but have encouraged more women to join the labour force. Job-splitting by male workers has been virtually non-existent, partly because of the difficulty of providing part-time jobs on more than a half-time basis. Measures to reduce the number of hours in the working week have usually raised hourly labour costs. Dreze argued that these policies will not induce firms to increase hiring in a recession unless firms are already planning to hire new workers.

Despite this limited success, Dreze speculated that increased part-time work is the most effective form of work-sharing. It is only feasible, however, if individuals are prepared to extend their working schedules outside the conventional period of business activities. Firms must be able to operate, say, six days a week while employees work a four-day week (with Saturday working two weeks out of three). A move in this direction, however, requires fundamental changes in social attitudes.

In his discussion of Dreze's paper, Torsten Persson (Institute for International Economic Studies, Stockholm) suggested that the analysis of unemployment and assessment of work-sharing could benefit from theoretical approaches other than those based on contract theory; 'insider-outsider' models and models of union behaviour might well produce different results. In particular, these approaches might suggest that many of the measures advocated by Dreze would increase labour costs or encourage increased wage demands. Persson also thought that labour market institutions and contractual arrangements might change with the introduction of work-sharing measures. It was important therefore to evaluate whether the conclusions of the analysis were sensitive to assumptions concerning these institutional details. In the ensuing discussion, some Panel members were critical of Dreze's 'disequilibrium' approach, advocating instead an equilibrium approach which focussed on the determinants of the natural rate of unemployment.

Unemployment and labour market behaviour was also the concern of Christopher Pissarides (LSE), in his paper 'Unemployment Flows in Britain: Facts, Theory and Policy'. Studies of aggregate unemployment often focus on the 'stock' of unemployed. Pissarides took a different approach, and examined the 'flows' into and out of unemployment. He found that changes in unemployment in the United Kingdom have been governed mainly by changes in the rate at which unemployed workers find jobs. Movements of workers in the other direction, from employment to unemployment, have generally been unimportant, except since 1979 when redundancies have played a role.

Pissarides based his conclusions on a theoretical model of the labour market in which firms and workers simultaneously search: firms to recruit suitable workers and workers to find desirable jobs. The flow of workers from jobs into unemployment shows little variation (except after 1979); Pissarides therefore took this as exogenous. His model did, however, explain the flow out of unemployment into jobs, in terms of the level of unemployment, vacancies, unemployment benefits and other variables. Vacancies were determined primarily by firms' demand for labour. These relationships together determine the evolution of unemployment and vacancies over time. He then provided estimates of the two key relationships in his model and used them to analyse the determinants of the 11.5 percentage point rise in UK unemployment between 1974 and 1983.

According to his estimates, nearly two-thirds of the rise in unemployment is attributable to a fall in the demand for new workers or to increased redundancies. No more than a third can be attributed to increased 'choosiness' of employers or workers. Pissarides's analysis suggests that increased unemployment benefits raised the unemployment rate by less than one percentage point between 1974 and 1983. Employment protection legislation greatly affected turnover in the labour market but had little overall effect on unemployment. Trade union resistance to cuts in real wages also raised unemployment by around one percentage point, but Pissarides's analysis indicated that most of the reduction in demand for new workers and increased redundancies is the result of lower demand for output.

The labour market, like that for any other good, serves to bring together buyers and sellers of the good concerned. Pissarides went on to examine the efficiency of the labour market in matching workers to jobs, noting that an extra vacancy not only increases the probability that an unemployed worker will find a job but also reduces the probability that another firm will be able to fill its vacancies. In order to operate efficiently, the labour market requires a level of vacancies between 30% and 50% higher than the level of unemployment, according to Pissarides's calculations. This was common in the United Kingdom during the 1960s, but the number of vacancies is currently only 10% of the number unemployed. Pissarides concluded that there are microeconomic as well as macroeconomic arguments in favour of measures to increase the number of jobs on offer. This could be accomplished either through an increase in aggregate demand or, if that is ruled out on counter-inflationary grounds, through the provision of employment subsidies.

Richard Layard (LSE) discussed the relationship of Pissarides's work to more conventional analyses of unemployment such as his own. He argued that they were complementary rather than conflicting. The Panel discussion also highlighted the fact that Pissarides's results focussed on the microeconomic issue of the efficiency of the labour market for a given state of aggregate demand. His analysis did not address the macroeconomic question of whether there was scope for policies to stimulate aggregate demand without leading to an increase in inflation.

Protection is bad for the economy as a whole, according to conventional economic wisdom. While import tariffs yield revenue to the government and benefit domestic producers, this is outweighed by the loss suffered by consumers because of higher prices. The case for import protection may be much stronger than conventional wisdom suggests, according to CEPR Programme Director Alasdair Smith and Research Fellow Tony Venables (University of Sussex and CEPR). In 'Trade and Industrial Policy under Imperfect Competition', they pointed out that the conventional case against protection assumes that markets are highly competitive. In many markets, however, there are only a few firms and there appear to be economies in large-scale production. Do the usual arguments in favour of free trade still hold? Smith and Venables observed that in such an environment international trade is potentially even more beneficial than when markets are perfectly competitive. Not only does trade allow consumers access to the cheapest source of commodities, but it also allows domestic firms to reap the benefits of large-scale production, increases competition and increases the range of products available to consumers.

Smith and Venables warned, however, that this did not imply that free trade is necessarily the best policy for a country to follow. Protection permits domestic firms to expand further the scale of their operations, allowing these firms to produce at lower costs because of economies of scale. The protected home market also allows an even more diverse range of products to be developed. Finally, the increase in profits may encourage new domestic producers to enter the protected industry.

In order to evaluate the significance of these arguments, Smith and Venables constructed a small illustrative model of an imperfectly competitive industry, incorporating recent developments in the theory of imperfect competition and international trade. The parameters of the model were assigned numerical values using data from the UK footwear and refrigerator industries. They then analysed the impact on each industry of a 10% export subsidy or a 10% import tariff. Provided other countries do not retaliate, Smith and Venables analysis suggests that the economy as a whole is slightly better-off under either of the protectionist policies than under free trade. Smith and Venables stressed that this did not imply that governments should adopt protectionist policies: the gains from such policies are quite likely to turn into losses if other countries retaliate, and there was a strong case for resisting protectionist measures through institutions such as the GATT. Interventionist trade policy, they argued, was 'a genie waiting to be let out of its bottle' and an open invitation to special pleading by special interest groups.

Paul Krugman (MIT), amidst a profusion of puns concerning cold feet, welcomed the attempt to quantify the implications of recent theoretical advances in industrial economics. The rather static nature of the model used by Smith and Venables was a limitation, however. The treatment of strategic interactions among firms and among countries was also a crucial consideration, according to Krugman. It was important to know how policy conclusions were affected by alternative assumptions about these interactions.

It is commonly believed that expansionary fiscal policy will increase employment, but that it will also lead to an increase in inflation, and a deterioration in the trade balance. The effects of a fiscal expansion are considerably more complicated, argued Jose Vinals (Stanford University and Banco de Espana) in his paper, 'Fiscal Policies, Employment and the Current Account'.

Governments are rightly concerned about the long-term effects of a trade deficit, because a country cannot live beyond its means indefinitely. The short-run effects of a deficit are also important, Vinals argued, because a trade deficit may affect a country's ability to borrow in the future and hence its ability to cope flexibily with unforeseen events. Consequently the impact of fiscal policy on the trade balance is an important question.

This impact is complex, however. Economic theory suggests that the effect of a fiscal expansion on employment and the trade balance depends on (1) whether the unemployment is caused by a low level of demand for output or excessive real wages; (2) how the fiscal expansion is financed; (3) whether or not the exchange rate is flexible; and (4) whether the fiscal expansion is temporary or permanent.

Vinals argued that where unemployment is due to insufficient aggregate demand, a temporary increase in government spending, if suitably financed, can increase employment without worsening the trade balance. Expansionary fiscal policies will not reduce unemployment, however, if this unemployment is due to excessively high real wages. A deterioration in the trade balance is inevitable, Vinals argued, only when the fiscal expansion is temporary and financed by sales of government debt.

Vinals used this theoretical framework to explain the varied experiences of the United States, the United Kingdom, France, Germany and Spain over the last decade. These countries differ widely in their economic circumstances and in their chosen fiscal policies. Yet in each case, Vinals's model correctly predicted the response of employment and the trade balance to changes in fiscal policy.

In the United States there was a switch from surplus to deficit in 1980, which intensified following the introduction of the Economic Recovery Act of 1981. There have also been important fiscal policy changes in Europe in recent years. The two most striking cases of fiscal restraint are those of West Germany and the United Kingdom, the only two countries to reduce budget deficits (as a percentage of GDP) between 1978 and 1984. But not all European countries followed the British and German example. In fact, Italy, Spain, Belgium and Denmark followed expansionary fiscal policies at various times. Italy substantially enlarged its actual and adjusted budget deficit in 1981 and 1982; Spain embarked on a path of large deficits after 1981, one year before the Socialists were elected to office; Belgium went through a period of high adjusted deficits from 1979 to 1981, which were later reduced; Denmark experienced large adjusted deficits from 1982 to 1984. The French case is, however, not as clear-cut. The initial Mitterrand years (1981-1983) were characterized by a worsening of the actual budget deficits, but the adjusted deficits were quite small and moderate by comparison to the other countries considered.

Vinals focussed on the periods which preceded and followed major fiscal policy changes in these countries. He found a significant positive association between changes in the budget deficit and employment for West Germany, the United Kingdom and Spain. The fiscal tightening in Germany and the United Kingdom was accompanied by a worsening of the employment situation. In the case of Spain, the fiscal expansion which took place after 1982 did not reduce unemployment, but employment at least fell at a reduced rate. In the case of the United States and France, Vinals found no evidence of a positive connection between fiscal expansion and employment.

Conventional wisdom suggests that budget deficits are accompanied by current account deficits. Vinals found that this is not always confirmed by the evidence. There was a positive association between fiscal and trade deficits for the United States, Germany and the United Kingdom, but no relationship for France and Spain. This evidence suggests that the link between budget deficits and current account deficits is far from straightforward. Yet Vinals's analysis, which incorporated the state of the labour market, financial policy and the exchange rate regime, is able to explain the response of the current account to the diverse fiscal policies pursued by European countries. Assessments of fiscal policy should take these other factors into account, Vinals argued, and not rely on the conventional wisdom that a trade deficit is the inevitable consequence of a fiscal expansion.

The papers discussed at this Panel meeting will be published with the Panel's comments in the third issue of Economic Policy, available in October. The April 1986 issue of Economic Policy is available now.

The Panel will meet again in London on October 16-17. Topics for discussion at the fourth Panel meeting will include: profit- sharing, banking deregulation and integration, relative wage dispersion, poverty programmes in the United Kingdom, and export credit guarantees.

The first volume of Economic Policy will consist of three issues, November 1985, April and October 1986. Each volume thereafter will consist of two issues. Subscriptions are available at the institutional rate of #16.00/$30.00/Ffr 240, and at an individual rate of #9.00/$15.00/Ffr 120 per volume. The special three-issue Volume 1 will be available at these rates for a limited period. Individual subscribers must order direct from Cambridge University Press. Further information can be obtained from The Subscriptions Manager, Cambridge University Press, the Edinburgh Building, Shaftesbury Road, Cambridge CB2 2RU, England or 32 East 57th Street, New York, NY 10022, USA.