Most published work on the Japanese economy in English to date has been written from an American perspective and tended to highlight the differences between its workings and those of other industrialized countries. European studies, though so far few in number, allow comparisons with different economic models and institutions. The first workshop of CEPR's European Network on the Japanese Economy, which is supported by a grant from the Daiwa Anglo-Japanese Foundation, took place in Oxford on 12/13 November. The workshop was organized by Jennifer Corbett, University Lecturer on the Social and Economic Development of Contemporary Japan at the University of Oxford and Research Fellow in CEPR's Applied Microeconomics programme, and hosted by the Nissan Institute of Japanese Studies at St Antony's College, Oxford.

David Campbell (University of Essex) opened the workshop with `A Critical Analysis of Selected Life Cycle Studies on Japan', outlining the results of various recent studies of Japanese savings behaviour which he suggested are flawed in several respects. The definitions of transfer wealth in many of them are inconsistent with the commonly accepted definitions of Modigliani and KotlikoffSummers; there are serious deficiencies in the data they use; and they incorporate many unrealistic assumptions about demographic trends and savings behaviour which may substantially affect their estimates of accumulated wealth and life-cycle saving. Campbell concluded that a number of the claims made in some of these studies lack foundation and that life-cycle explanations of wealth accumulation cannot be rejected. He cited recent work in the US which shows very unequal distributions of wealth and saving and suggested placing greater emphasis on examining diversity of savings behaviour and, in particular, estimating saving by non-worker Japanese households.

John Muellbauer (Nuffield College, Oxford, and CEPR) noted that these criticisms of micro-data studies also apply to aggregate studies and time-series analyses in which demographic factors are particularly important. Jennifer Corbett suggested that the bequest motive may be an important reason for saving even if wealth accumulation results from life-cycle behaviour. Campbell agreed and noted that intergenerational wealth transfers which occur while the older generation is still alive are classified as life-cycle wealth. Asked about available data for non-salaried worker households, Campbell noted that consumption data and annual data on incomes are collected but may be unreliable. Tax survey data may remain a viable alternative.

Hideki Yamawaki (Université Catholique de Louvain) then presented `International Competitiveness and the Choice of Entry Mode: Japanese Multinationals in U.S. and European Manufacturing Industries'. The results of a logit regression analysis indicated that Japanese firms are more likely to enter via acquisition rather than green-field investment in industries in which they enjoy less of a competitive edge. Japanese firms' greater use of acquisition when diversifying into the US suggests that they are seeking US technologies, but it may also partly reflect the greater openness of the US market for corporate control.

Mari Sako (LSE) suggested extending the paper to consider the distinction between acquisition and capital participation; many studies have suggested that the distinction between joint venture and wholly-owned subsidiary is important, so comparative information on Japanese firms' strategy in the US and Europe would be of interest. She suggested that the automobile industry should be treated as a special case and noted that the use of US patents to measure technological advantage may introduce a bias: alternative European patent data is also available.

In `Careers, Ranks and Earnings Profiles in Internal Labor Markets: Evidence from U.K. and Japan', Kenn Ariga (Kyoto University) and Giorgio Brunello (Università degli Studi di Venezia) argued that existing studies that stress the steepness of earnings profiles in Japan cannot distinguish between human capital models and alternatives and thus provide no support for the common view that in-firm training is more important in Japan. Their new estimates, based on including a `rank variable' which proxies a worker's position in the firm hierarchy in the earnings function, suggest that earnings profiles are not robustly steeper in Japan than in the UK. Net earnings profiles earnings differences within ranks are in fact steeper in the UK; adding `between ranks' differentials yields the steeper Japanese profiles as expected. Apparently, wages attached to job descriptions are much more important in Japan than the conventional view suggests.

Stephen Nickell (Institute of Economics and Statistics, Oxford, and CEPR) questioned the appropriateness of comparing ranks internationally or across companies with different structures and suggested that the rank variable as defined here could be picking up tenure effects as well as unobserved individual ability. Merely including a rank variable did not provide an adequate test of human capital models against alternatives.

Seiichi Kawasaki (University of Stirling) opened the second day by presenting `Japanese Bonuses: Rent Shares, Profit Shares or Disguised Wages?', written jointly with Robert Hart. Their regression analysis of cross-sectional and pseudo-cohort panel data on wage structure suggested that bonuses in Japan should be viewed as workers' rent shares for firm-specific human capital rather than profit shares or part of normal wages: bonuses exhibit a steeper tenure profile than wages and a `mountain shape' as bonus payments initially rise with age but later fall.

Mary Gregory (St Hilda's College, Oxford) suggested as an alternative explanation that the expansion of higher education may have reduced the returns to extra education in terms of wages and bonuses, whose differences appear more powerfully associated with formal education than tenure. Giorgio Brunello suggested that the humped shape of bonus payments may support the profit-sharing hypothesis if profits fall as human capital declines. Brunello and Kenn Ariga also doubted the reliability of the pseudo-cohort panel data since Ministry of Labour statistics cover a varying sample of the workers of the establishments surveyed. Gianni Toniolo (Università degli Studi di Venezia and CEPR) suggested that trade unions may represent the interests of longer-tenure workers if bonuses are determined by bargaining, which may provide an institutional explanation of the results. Jennifer Corbett noted that Brunello's interpretation required bonuses to be distributed on the basis of individuals' contributions to profits; Kenn Ariga explained that there is some discretion in their distribution, but Kawasaki cited time-series evidence which suggested they are no more flexible than wages.

Konosuke Odaka (School of Oriental and African Studies, London, and Hitotsubashi University) concluded the workshop with `The Changing Nature of Capital Formation in Japan: 1780-1980'. His evidence from a number of regional and micro data sources suggested that output and investment displayed positive growth during the latter half of the eighteenth century, levelled off in the first half of the nineteenth century, and began to grow again after the Meiji Restoration. As the earlier rise in output was accompanied by neither population growth nor imports of technology, investment was financed out of the rising economic surplus. This reinforces other recent research which has revised upward the starting conditions for growth in Japan.

Gianni Toniolo noted that recent research on Scandinavian economic development had led to a similar upward revision of estimates of output levels for the pre-industrial period. The downward revision of the gap between potential and actual output suggested that the role of `catch-up' in Japan's growth may have been exaggerated. Jennifer Corbett asked about the extent of discontinuity in growth at the end of the nineteenth century. Odaka replied that the real industrial revolution probably began at the end of the Russo-Japanese war: while the Meiji Restoration brought great political discontinuity, the continuity in economic development may have been greater than previously thought.