|
Why has Europe suffered high and persistent unemployment in the
1980s, while the United States has not? This puzzle has helped spur one
of the most important developments in macroeconomics in the 1980s, the
elaboration of models in which wages are the outcome of negotiations
between firms and incumbent workers. Observers have noted that, for a
variety of reasons, the unemployed seem to become disenfranchised from
the labour market, so that high unemployment does not curb the growth of
real wages. The debate surrounding the validity of these approaches has been fierce because of their implications for the effectiveness of reflationary policies. The desire to re-examine these theories and the evidence behind them prompted the Centre to hold a March workshop on `Wage and Employment Bargaining and Labour Market Dynamics', organized by CEPR Research Fellows David Begg and Dennis Snower. Financial support was provided by the Employment Market Research Unit of the Department of Employment and by the German Marshall Fund of the United States. The workshop explored the importance of insider power in wage and employment determination from a variety of different perspectives, theoretical and empirical. It examined different sources of hysteresis, the relative importance of insider and outsider forces in wage and employment determination, and evidence on both symmetric and asymmetric persistence of macroeconomic shocks. The first paper presented to the workshop, `Seniority Rules and Wage-Employment Determination', by Allan Drazen (Hebrew University of Jerusalem) and Nils Gottfries (Institute for International Economic Studies, Stockholm), considered how one might model the optimizing behaviour of insiders under stochastic conditions. In particular, Drazen and Gottfries examined how wages and employment should react to transitory shocks when current insiders take account of possible future developments in the economy. They also considered how temporary shocks can become permanent on account of the insiders' wage-setting activities. Employment and trade union membership change as a result of a shock, and there is no tendency for wages to change so as to re-employ the unemployed in better times. Drazen and Gottfries also analysed how seniority rules can protect incumbent workers. Commenting on the paper, James Malcomson (University of Southampton) argued that wages should depend on seniority rules and we should expect employment rules such as `last-in-first-out' to depress outsider wages. Richard Jackman (LSE) wondered why insiders did not set the wage so that there was no entry by new workers, as the logic of the model itself might suggest. The next paper, `Dynamic Wage and Employment Determination with Employment Adjustment Costs', by Ben Lockwood and Alan Manning (Birkbeck College, London) analysed the time paths of wages and employment when wages are negotiated by unions and firms and there are employment adjustment costs. The speed of adjustment in this model depended on bargaining power and the power of insiders, according to Lockwood and Manning. They found that, contrary to common belief, the presence of unions can increase the optimal speed of adjustment in the labour market. Their analysis suggests important ways of amending the conventional labour demand function so as to include new exogenous variables relevant to the dynamic determination of wages and employment. Stephen Nickell (Institute of Economics and Statistics, Oxford, and CEPR) suggested that the results on the speed of adjustment depended in part on the authors' assumption that the firm's revenue was a linear function of its employment. The effects of unions on the costs of adjustment were, Nickell believed, rather more complicated in practice. Lockwood replied that they were looking at the strategic aspects of wage-setting, and that unions of course played a role in this. In their paper, `Wage-Setting and Labour Market Adjustment in the Main Industrial Countries', George Alogoskoufis (Birkbeck College, London, and CEPR) and Alan Manning (Birkbeck College, London) examined the sources of unemployment persistence in more detail. Blanchard and Summers had argued that the difference between European and US unemployment levels could be attributed to the effects of hysteresis. Alogoskoufis and Manning re-estimated the model, introducing a quadratic rather than the linear trend used by Blanchard and Summers. Calculations based on the re-estimated model suggested a much smaller hysteresis effect, with little difference among the United States, Japan and the European countries. Alogoskoufis and Manning modelled the process of wage and employment de termination in greater detail. They departed from the Blanchard-Summers analysis by explicitly allowing for adjustment costs in labour demand as well as insider-outsider effects. They found marked differences between countries. France and the United States in particular were difficult to model in terms of a conventional downward-sloping labour demand curve, although the results for the other major countries were more conventional. Alogoskoufis and Manning concluded that the differences in hysteresis effects across countries were not large enough to explain the marked differences in unemployment. Persistence in labour demand varied much more across countries and this, the authors argued, was the most important determinant of differences in unemployment. Charles Bean (LSE and CEPR) noted that explicit measures of search behaviour were omitted from the equations. Nils Gottfries asked whether there was any distinction in behaviour between upswings and downswings. Alogoskoufis replied that measures of search behaviour were difficult to obtain for many countries and that it was difficult to allow for differences in behaviour between upswings and downswings within a conventional regression framework. Larry Katz (Harvard University) and Alan Krueger (Princeton University) presented the final paper of the morning session, entitled `Job Queues and Wages'. Focusing on the relationship between job queues and wage offers, Katz and Krueger examined data on job applications for a number of occupations, drawn from a cross-country survey, concentrating on those jobs that offer a legal minimum wage. Some economists have argued that minimum wage legislation would be ineffective, since workers and firms will conspire to evade the provisions of the law. In particular, it is argued, those jobs carrying the (excessive) legal minimum wage will carry fewer fringe benefits, such as on-the-job training or pleasant working conditions. If this is so, Katz and Krueger argued, jobs at the minimum wage and those offering wages either above or below it should attract roughly the same rate of applications, since they are equally attractive. The authors found, however, that this did not seem to be the case. The second part of the paper examined the effects of high wages on job applications and turnover. There appear to be more applications for jobs in industries that pay more, and furthermore better-paying industries seem to have a lower turnover rate. Katz and Krueger interpreted this as queuing by workers who are willing to wait longer for a better-paid job. Contrary to the beliefs of some economists, it therefore appeared that better-paid jobs do confer some rents on the workers who hold them. Commenting on the paper, Tony Meyrick (HM Treasury) suggested that more could have been done to model search behaviour by employers. Some jobs are advertised only on a very informal basis and others not at all: this sort of information would have been very useful in the study. The first paper of the afternoon session, `InsiderOutsider Forces in Wage Determination', was presented by Stephen Nickell (Institute of Economics and Statistics and CEPR) and Sushil Wadhwani (LSE and CEPR). The authors used data from individual firms to examine the importance of insider effects on wage-setting. This is a fruitful approach since the insider-outsider theory suggests that economic conditions inside firms may play a significant role in determining wages, employment and unemployment at the macroeconomic level. Nickell and Wadhwani estimated a wage equation at the level of the firm, including variables which represented both insider and outsider factors. The wage equation can be written in such a way that the relative importance of insider effects is captured by a single parameter. Insider forces in wage determination are captured by variables representing the incumbent workforce, union power and the firm's liquidity position. Outsider forces are measured by alternative wages prevailing in other industries, the aggregate unemployment rate and the value of unemployment benefits. Nickell and Wadhwani's estimates suggested that insider forces do play a role in wage determination, but that outsider forces also play a very significant role. The authors also discussed more recent work in which the strength of insider forces is allowed to vary over time. Interestingly, such models indicated that insider power varied significantly between booms and recessions. John Driffill (University of Southampton and CEPR) interpreted the results as implying there was more persistence in downswings. Stephen Nickell agreed, but suggested that this was true more or less by definition. David Stanton (Employment Market Research Unit) disputed the view that higher bargaining power implied greater insider power; this implication conflicted with the idea that the way to make labour markets more responsive to economic conditions is to decentralize wage bargaining. The final paper of the workshop, `Shock Persistence in the Labour Market: Some Empirical Evidence' by David Begg (Birkbeck College, London, and CEPR), Christopher Martin (Queen Mary College, London) and Dennis Snower (Birkbeck College, London, and CEPR), used the insider-outsider theory to investigate asymmetric persistence of anticipated macroeconomic shocks. The authors suggested that the wage and employment objectives of employed workers may depend on whether they face an economic upswing or downswing. In an upswing, when the labour demand curve shifts outwards, incumbent workers may have an incentive to drive up their wages and, in the process, discourage new hiring. They do not, however, have an equal and opposite incentive to drive down their own wages in a downswing. Thus the amount by which employment rises in an upswing will be smaller than the fall in employment in a downswing of equal magnitude. In short, the effects of temporary, anticipated macroeconomic shocks on wages and employment may persist due to an asymmetry in insiders' responses to upswings and downswings. The authors adopted two approaches to testing for these asymmetries. The first examined whether the insider workforce has a stronger influence on wages in an upswing than in a downswing. The insideroutsider theory predicts that this will be the case, since upswings may be less effective than downswings in changing the size of the insider workforce. Similarly, the theory predicts that macroeconomic shocks have a weaker influence on wages in an upswing than in a downswing. The authors' preliminary empirical results showed that the relevant coefficients in the UK wage equation are significantly different in upswings and downswings. This suggests that the UK labour market may suffer from a limited degree of asymmetric persistence. In contrast, the US wage equation exhibits negligible symmetric or asymmetric persistence. The second approach is to focus directly on shifts in the labour demand curve in order to investigate the amount of economic rent that insiders extract in the process of wage-setting. The insiders are assumed to set targets for the growth of wages and employment relative to the level of the previous period. When the labour demand curve shifts outwards, the insiders set the wage so that both wages and employment will increase. When the labour demand curve shifts inwards, the wage is set so that both wages and employment fall. The theory suggests that the trade-off between wages and employment should be steeper in an upswing than in a downswing. Begg, Martin and Snower estimated an econometric model which reflected this theory and found some evidence that wages do behave asymmetrically in upswings and downswings. They concluded that the insider-outsider hypothesis showed sufficient promise to justify more thorough empirical investigation. Patrick Minford (University of Liverpool and CEPR) caused a lively discussion by asking whether insider-outsider effects represented irrational behaviour on the part of workers, since it doomed their jobs to eventual extinction. Nils Gottfries suggested that the authors' second approach called for some form of smoothing across different labour markets. Begg agreed, but argued that this was done in the first part of the paper. The workshop concluded with a roundtable discussion of the policy implications of the insider-outsider approach. David Stanton found the work very interesting. Until recently, the declines in union power and in hiring and firing costs seemed to be having relatively little effect on employment. Instead, he argued, firms were using the opportunity to reorganize production and working practices. He also suggested that a closer link with research in industrial relations would be of benefit to both sides. George Alogoskoufis argued that the evidence for both insider effects and persistence in labour demand was strong. Policy prescriptions would depend on identifying the sources of this persistence. Jon Stern (HM Treasury) found the insider-outsider approach stimulating but was keen for more institutional detail to be included in the analysis, so that the determinants of labour market flexibility could be seen more clearly. |