The links between product and labour markets lie at the heart of many important policy issues, including the effects of demand management policy on employment and the relative importance of reforms in the product and labour markets in stimulating economic activity. The traditional Keynesian models identified two main links between product and labour markets: first, increased labour demand raises purchasing power and thereby raises product demand; and second, increased product demand stimulates employment, either directly or indirectly via reductions in the real wage.
There is now broad agreement that while these Keynesian `spillovers' may be operative in the short run, when wages and prices are sticky, they are unlikely to be reliable in the longer run, when wages and prices respond flexibly to market conditions. Some recent research has focused on the imperfectly competitive interactions between the product and labour markets. It examines such issues as how monopoly power in product markets affects wage and employment decisions and how the bargaining process between firms and workers influences output, investment and innovation. In this context, the effectiveness of policies designed to change the degree of competition in the product and labour markets have received particular attention.
This body of research was the subject of a CEPR workshop on `The Links Between Product and Labour Markets', held in London on 23 February. The workshop was organized by Research Fellows David Begg and Dennis Snower, and the financial support of the Employment Market Research Unit (EMRU) at the UK Department of Employment is gratefully acknowledged.
Alistair Ulph (University of Southampton and CEPR) and David Ulph (University of Bristol and CEPR) presented a paper on `Bargaining Structures and Delay in Innovation', in which they examined the effects of unionized labour markets on the willingness of firms to engage in R&D expenditure and introduce new technologies. They modelled an innovation race between two firms and studied the influence of different bargaining structures on the value to each firm of winning a patent. The model suggests that allowing a union to bargain over the implementation of technology may lead it to take actions that will discourage the firm from winning an innovation race and thereby make the union worse off in the long run.
Zmira Hornstein (EMRU) remarked that the newspaper industry in the UK provided an example of a union's ability to delay the introduction of a new technology. Christopher Kelly (HM Treasury) questioned whether the model was specified correctly, since most firms invest directly in R&D rather than competing over existing innovations. The authors replied that the results of their model were robust with respect to direct R&D investment. Alan Manning (LSE and CEPR) pointed out that whether innovation increases social welfare remains an open question, particularly in the presence of strategic over-investment by firms. David Begg (Birkbeck College, London, and CEPR) noted that the analysis could be extended to cover the influence of unions on any type of investment, not merely the adoption of new technologies.
Giuseppe Bertola (Princeton University and CEPR) presented a paper on `Flexibility, Investment and Growth', in which he considered the effects of labour market imperfections in the context of an endogenous growth model with irreversible investment decisions and imperfect labour mobility. It is a standard result that labour turnover costs will unambiguously reduce the responsiveness of employment to both labour and product market shocks in partial equilibrium. Bertola showed, however, that this need not hold in general equilibrium. The paper thus provided a framework for analysing the impacts of industrial policies and of various institutional features on growth. Bertola noted an interesting implication of the trade-off between static and dynamic efficiency when obstacles to labour mobility cannot be eliminated: governments that provide some form of insurance to `bad' firms may be able to offset the adverse growth effects of high labour turnover costs and inefficient wage patterns.
Dennis Snower (Birkbeck College, London, and CEPR) remarked that the case for insuring `bad' firms hinges importantly on Bertola's questionable assumption that labour turnover costs are exogenously given. In practice, government subsidies will generate increased economic rents, which may give incumbent workers an incentive to engage in more rent-seeking activity and thereby raise firms' labour turnover costs.
Dennis Snower then presented a paper on `Transmission Mechanisms from the Product to the Labour Market', written jointly with Assar Lindbeck. He first noted that conventional Keynesian and neo-classical transmission mechanisms tend either to rely on wage-price rigidities or to imply countercyclical real wage movements. While some of these transmission mechanisms are undoubtedly important in practice, it is doubtful whether they tell the full story in the longer run, particularly in countries where real wages are strongly procyclical (e.g. the US).
Snower suggested four main channels whereby product demand variations may be transmitted to the labour market under flexible wages and prices, while allowing for procyclical and acyclical real wage movements: endogenous adjustment of the price elasticity of product demand; endogenous change in the imperfectly competitive interactions among firms; entry and exit of firms in response to product demand variations; and endogenous change in the marginal product of labour. Snower argued that the first two channels are `weak reeds' on which to hang a theory of the macroeconomic effectiveness of demand-side shocks, while the latter two are more promising. In particular, if nominal wages are temporarily sluggish, a rise in product demand may induce new firms to enter and remain in the market even after the wage sluggishness has disappeared. Furthermore, a rise in product demand may stimulate the marginal product of labour via induced changes in capital utilization and investment, and changes in industrial infrastructure.
Stephen Nickell (Institute of Economics and Statistics, Oxford, and CEPR) suggested that Snower's argument was fundamentally Keynesian in its assumption that sluggish wages and prices are required to make the suggested transmission mechanisms work. Snower replied that in the traditional Keynesian models the real effects of product demand variations tend to disappear when wage-price sluggishness disappears, whereas in his analysis the real effects remain. George Alogoskoufis (Birkbeck College, London, and CEPR) argued that product demand management policies could affect the labour market by changing the wedge between producer and consumer wages.
Stephen Nickell presented a paper on `Unions and Investment in British Industry', written with Kevin Denny, in which they argued that unions can influence investment either directly, by impeding the installation of new technology and raising the cost of investment, or indirectly via real wages. The results of their empirical model indicated that a firm with a recognized union and average union density had an investment rate around 23% lower than an equivalent firm with no recognized union, if product wages are held constant. If the effects of unions on wages are taken into account, a firm facing competitive conditions in the product market with a recognized union and average density had an investment rate some 13% lower than a similar firm with no recognized union, while the corresponding relative reduction for the `non-competitive sector' was a mere 4%. The authors also found that the negative impact of unions on investment had weakened slightly in recent years.
Alan Manning argued that if unions have bargaining power over investment, one might expect a positive (rather than a negative) effect on investment, particularly when labour and capital are complements. Nickell replied that his analysis did not imply that unions are necessarily opposed to investment, but that they nevertheless add to the cost of installing new technology. Patrick Minford (University of Liverpool and CEPR) observed that it appeared paradoxical for the competitive sector to have a larger drop in investment than the non-competitive sector as a result of union activity. John Muellbauer (Nuffield College, Oxford, and CEPR) suggested that the analysis should make allowance for the number of small firms, since the interactions between firm size, interest rates and access to capital may exert significant influences on investment.
Mark Stewart (University of Warwick) presented a paper on `Union Wage Differentials, Product Market Influences and the Division of Rents', in which he sought to specify the circumstances under which unions are able to establish wage differentials. His results showed that while unions are able to attain average wage differentials of some 8% in negotiations with firms in the `non-competitive' sector, i.e. those that have a degree of product market power, they appear unable to extract any significant wage differentials from firms facing competitive market conditions. Stewart concluded that it is incorrect to view the effectiveness of unions independently of the product market conditions in which they operate.
Dennis Snower argued in the ensuing discussion that market share would have been a better measure of product competition than the number of firms, but Stewart replied that the relevant data are unreliable. John Muellbauer noted that Stewart's results in the competitive sector were dominated by large coefficients on the variables for establishment size and foreign ownership. Stewart agreed that establishment size may be acting as a proxy for some other aspect of unionization.
Sushil Wadhwani (LSE and CEPR) presented a paper on `The Determinants of Wage Flexibility in Japan: Some Lessons from a Comparison with the UK using Micro Data', written with Giorgio Brunello, in which they sought to explain why wages are more flexible and output tends to be more variable in Japan than in the UK. Their empirical results revealed that the responsiveness of wages to unemployment in large firms was similar, and that the responsiveness of wages to unemployment significantly greater in small firms than in large firms, in both countries. They also found that insider considerations are at least as important in wage determination in large firms in Japan as in the UK.
The authors concluded that the main reason for Japan's greater wage flexibility is to be found in its comparatively large proportion of small firms, and argued further the low level of variation in Japanese unemployment relative to output may be attributed to the relative ease with which displaced workers move from the primary to the secondary sectors during recessions.
Mathias Dewatripont (Université Libre de Bruxelles) observed that large Japanese firms tend to perform better than their British counterparts and wondered whether this might be explained by the fact that they are typically less vertically integrated, and are thus better able to pass their adjustment costs over to sub-contractors.
Zmira Hornstein concluded the workshop with a paper on `The Concerns of Policy Makers', in which she drew attention to a number of areas of research to which the workshop had made significant contributions but which remained in urgent need of further work. These included an assessment of the extent to which competition in the product market is necessary for the achievement of efficient labour market outcomes, and the identification of the channels whereby product markets affect labour markets. Other major areas for further research included the degree to which the `insider power' of the professions accounts for restrictive entry conditions and the main determinants of insider power