A two day conference on Growth, Trade and the Labour Market, organised by the Centre for Economic Performance (CEP) and CEPR, took place at CEP on 11/12 July 1997. Financial support was provided by the Centre for Economic Performance. Twelve papers were presented, covering a broad range of issues, from mechanisms to encourage innovation to empirical evidence on the effects of international trade on employment and wages. The conference organizers were Christopher Pissarides (CEP, LSE and CEPR) and Danny Quah (LSE and CEPR).

In ‘Agglomeration and Economic Development: Import Substitution versus Trade Liberalisation’, Diego Puga (LSE) and Anthony Venables (LSE and CEPR) analysed a model of development in which international inequality in the location of industry and income levels is supported by agglomeration of industry in a subset of countries. A model of economic geography was used to investigate the equilibrium conditions for agglomeration. Focusing on this equilibrium, Puga and Venables then argued that development may not be a continuous process of convergence by each country, but instead may involve countries moving in turn from a poor group to a rich group of countries On the role of trade policy in promoting industrialization, they showed that, while both import substitution and trade liberalization may successfully attract industry, welfare levels are higher under trade liberalization.

In ‘A Mechanism for Encouraging Innovation’, Michael Kremer (MIT) argued that patents create monopoly-price distortions and generate insufficient incentives for original invention. This is because inventors cannot fully capture the consumer surplus or the spillovers of their ideas to other researchers. Also, socially wasteful expenditures are incurred on ‘inventing around’ patents. Theoretically, these distortions could be eliminated if governments purchased patents from inventors at their social value and transferred them to the public domain. In practice, though, it is difficult to determine the social value of inventions. Kremer proposed a mechanism under which the private value of patents would be determined in an auction. Governments would then offer to buy out patents at this value, plus a fixed mark-up that would roughly cover the difference between the private and social costs of inventions. Inventors could decide whether to sell or retain their patents, with those purchased by the government typically being placed in the public domain. However, to provide auction participants with an incentive to truthfully reveal their valuations, government

occasionally sell patents to the highest bidder. The pharmaceuticals industry was seen as an example well suited to the proposed mechanism.

The relationship between the import-export behaviour of French firms and their labour market outcomes was examined by Francois Kramarz (ENSAE-CREST) in ‘International Competition, Employment and Wages: The Microeconometrics of International Trade’. Basing his analysis on matched multiple firm-level sources as well as matched worker-firm data, Kramarz looked at the effects of trade on employment, skill structure, wages, wage structure, within-firm inequality and the propensity to separate from workers. The basic components of the dataset were the customs forms submitted by firms during the period 1986–90. Kramarz found first, that employment decreased as the imported share of firms’ purchases rose, but the effect was small. Second, the firm-level skill structure changed as trade theory would predict, i.e. an increase in imports from, or exports to, high-skill countries induced an increase in the proportion of low-skilled workers (and conversely). Again, however, the effects were small. Third, no relationship was found between wage evolution and trade behaviour. Finally, the individual-level probability of changing firms or people losing their jobs, controlling for all individual characteristics, was found to be related, not to increases in the firm’s own trade activity, but to the behaviour of the firm’s competitors. Steve Nickell (Institute of Economics and Statistics, Oxford, and CEPR) questioned some of the variables in the analysis, more particularly the combinations of variables used in some of the regressions.

An alternative explanation of inequality was offered in ‘Obsolescence of Capital’ by Boyan Jovanovic (New York University). Previous explanations have used differences either in initial conditions or in government policies. In this paper, inequality arises because of finite capacity in the capital goods sector. Jovanovic combined a vintage capital model with an assignment model to demonstrate the mechanism. He based his analysis on four assumptions: capital quality and labour skill are complements; new technology is embodied in new machines; capital and labour are used in fixed proportions; and assignment is frictionless. The balanced-growth path was found to be accompanied by a non-degenerate distribution of skill and machine quality around the trend. The reason for this is that society cannot give everyone a frontier machine all the time because of the fixed-proportions assumption and because it is too costly. The complementarity assumption means that those who do get the best machines will want to get more skill. Consequently, inequality in machines promotes inequalities in skills, incomes and wealth. Jovanovic suggested that, applied to the whole world, the model can help explain the non-convergence puzzle. In particular, the flatter the marginal cost of improving machines, the greater is the inequality.

Daron Acemoglu (MIT and CEPR) provided an alternative view of the link between skill-biased technical change and wage inequality. In ‘Why do New Technologies Complement Skills? Directed Technical Change and Wage Inequality’, Acemoglu built on the models of Aghion and Howitt and Grossman and Helpman by allowing technical change to be directed to different groups, skilled and unskilled. He argued that when the proportion of skilled workers in the labour force is larger, the market for skill-complementary technologies is also larger and more effort will be spent in upgrading the productivity of skilled workers. The reason is that, once invented, most technologies are non-rival goods: they can be used by many firms and workers at some low marginal cost. The larger market means inventors will be able to get higher returns. The argument could also be used to explain trends in the college premium in recent decades: an increase in the relative supply of skilled workers reduces the skill premium in the short run, but also leads to creation of new skill-complementary technologies. This ‘skill-biased’ technical change may lead to a higher skill premium in the long-run. Thus the rapid increase in the proportion of college graduates in the US labour force may have been the cause of both the decline in the college premium in the 1970s and the very large increase in inequality during the 1980s.

Whether skill-biased technology shocks can also explain both the rise in unemployment in Europe and the rise in wage dispersion in the US in the 1980s was the question posed in ‘Unemployment Responses to "Skill-Biased" Technology Shocks: The Role of Labour Market Policy’ by Dale T Mortensen (Northwestern University) and Christopher Pissarides (CEP, LSE and CEPR). Specifically, the paper asked whether more income support and less labour-market flexibility in Europe was preventing the downward wage adjustment needed to maintain employment levels among lower-skilled workers in response to technology changes favouring those with more skills. Using their own (1994) search-theoretic model of the labour market, the authors concluded that these arguments were formally valid. Skill-biased shocks were embedded in the model, which was then calibrated for the United States and Europe. The results, taking account of the known policy differences, were found to be quantitatively consistent with the US and European experiences.

Similar conclusions were reached in ‘Unemployment Versus Mismatch of Talents: Reconsidering Unemployment Benefits’, a paper by Ramon Marimon (EUI, Universitat Pompeu Fabra, NBER and CEPR) and Fabrizio Zilibotti (Universitat Pompeu Fabra and CEPR). They developed an equilibrium search-matching model with risk-neutral agents and ex-ante firm and worker heterogeneity. Unemployment insurance has the standard effect of reducing employment, but also helps workers to a suitable job in this model. Its predictions were consistent with the contrasting performances of the European and the US labour markets in terms of unemployment, productivity growth and wage inequality. To show this, Marimon and Zilibotti constructed two fictitious economies with calibrated parameters which differed only in the degree of unemployment insurance. The economies were then assumed to have been hit by a common technological shock which enhanced the importance of mismatch. The shock reduced the proportion of jobs which workers regarded as acceptable in the economy with unemployment insurance (Europe), thus doubling unemployment. In the laissez-faire (US) economy, unemployment remained constant, but wage inequality increased more and productivity grew less owing to the larger mismatch.

The links between fertility choice, the dynamics of income distribution and economic growth were investigated in ‘Demographic Transition, Income Distribution and Economic Growth’ by Momi Dahan (Bank of Israel) and Daniel Tsiddon (Tel Aviv University and CEPR). The authors used a simple model of fertility choice, in conjunction with the well documented differences in returns to human capital across rich and poor, to model the co-determination of family size and investment in human capital. They concluded that fertility choice, income distribution and economic growth affected each other in a way that produced inverted U-shaped dynamics. The first stage was characterized by an increase in the average rate of fertility and by widening income inequality, while income growth was uncertain. Only in the second stage did average fertility fall and income become more equally distributed. At this later stage, too, the economy became more human capital-abundant and growth of income per capita took off. The model thus explained some of the documented facts about epochs of demographic transition without relying on ‘near rationality’ arguments or non-economic objectives.

In ‘Just Can’t Get Enough: More On Skill-Biased Change and Unemployment’, Marco Manacorda (LSE and University College, London) and Alan Manning (LSE and CEPR) re-examined the importance of a fall in the (relative) demand for the less skilled in explaining the problems of labour-market performance in the OECD countries. They proposed a new measure of skill mismatch and, using data from France, Germany, Italy, the Netherlands, the United Kingdom and the United States, they demonstrated how the measure can be used to assess the importance of skill-biased change in understanding labour-market changes in recent years. Their findings suggested that increased skill mismatch was by no means a universal phenomenon, though it did seem to have occurred in the United States, the United Kingdom and Germany. There was much discussion about the definition of neutrality used by Manacorda and Manning, with Daron Acemoglu and Steve Nickell, in particular, unconvinced by their approach to the problem.

The macroeconomic role of financial intermediation was the focus of ‘Agency Costs in Models of Economic Growth’ by Costas Azariadis and Shankha Chakraborty (both University of California, Los Angeles). Their paper incorporated a credit market with asymmetrically informed borrowers and lenders into a two-period overlapping-generations model. The agency cost upon which they focused was the cost of state verification. There were two types of good – a capital good and a consumption good. Final goods were produced from labour and capital using a standard neoclassical production function, but capital formation was credit-financed. Capital was produced from bank loans using a constant-returns technology subject to idiosyncratic shocks, the realized values of the shocks only being verified at a cost by the financial intermediaries. The latter dealt with the moral hazard problem associated with asymmetric information and diversified idiosyncratic project risks. As a result, depositors were paid a sure deposit rate while borrowers received debt contracts which verified project outcome in some states. The authors found that, with non-convex verification costs depending on future economic activity, i.e. with expectations playing a non-trivial role, up to three steady states might exist, with the intermediate one not necessarily unstable. Furthermore, the equilibrium capital stock correspondence could propagate and amplify technological shocks. Numerical calculations showed that the scale of such costs for dynamics of this nature was consistent with existing evidence on the cyclical behaviour of the interest-rate spread between loan and deposit rates.