What are the distinctive features of corporate governance in Japan and Continental Europe? What is the relevance of these experiences for the economies in transition from socialism to capitalism? These questions were discussed at a joint CEPR workshop with ECARE, Université de Bruxelles on 28/29 October on `Competing Models of Capitalism'. The workshop was organized by Erik Berglöf (ECARE, Université Libre de Bruxelles and CEPR) and Jenny Corbett (St Anthony's College, Oxford, and CEPR), and financial support was provided by the Daiwa Anglo-Japanese Foundation's funding for CEPR's Japanese Economy Network.

In the first paper, `The Financing of Industry 1970-89: An International Comparison', Jenny Corbett documented the substantial problems in comparing the sources of industry finance in Germany, Japan, the UK and the US in the 1970s and 1980s. She argued that net sources and national income accounts provide the most appropriate and consistent information. Analysis of such data leads to the insight that there is no `market-based' Anglo-American pattern of industry financing. Rather, industry in Germany, the UK and the US is principally internally financed with small or negative contributions from market sources. Japanese industry &nbspis more externally financed with both banks and markets contributing larger shares. In the 1980s, a period of financial liberalization, all countries, except Japan, have seen more internally and less market financed industry. Oren Sussmann (Hebrew University of Jerusalem and CEPR) interpreted the fact that the ratio of retention to investment is higher in the US than in Japan as a result of different investment behaviour rather than different retention policies. Indeed, the ratio is generally low for high investment countries like Japan. He also demonstrated that the aggregation of financial flows over long periods underestimates the role of bank finance in providing liquidity.

In `The Choice Between Public and Private Debt: An Analysis of Post-Regulation Corporate Financing in Japan', Takeo Hoshi (University of California, San Diego) discussed the shifts from traditional bank to convertible bond financing of Japanese firms in the 1980s and the increasing heterogeneity of financing patterns among firms. Hoshi's work provides a theory of choice between bank debt and public debt. A simple model of moral hazard suggests that the use of bank financing will be negatively correlated with a firm's net worth. The model also suggests that the profitability of a firm and its use of bank financing will be negatively correlated if the manager's interest is sufficiently aligned with that of shareholders, and positively correlated if the manager puts sufficiently large weight on private benefits. Empirical investigation reveals that keiretsu firms' use of bank financing is negatively correlated with their performance, but owner-managed firms show positive correlation between profitability and the use of bank financing. Patrick Bolton (ECARE, Université Libre de Bruxelles, and CEPR) gave an alternative interpretation of the positive correlation between net worth and public debt financing. As reported in other studies, Japanese equities were highly overvalued in the 1980s. Firms with a high Tobin's Q (the measure of net worth used by Hoshi) had the highest incentive to issue convertible bonds since they took most advantage of this overvaluation. This explanation is corroborated by developments after the 1990s crash: no new convertible bonds have been issued since then.

In `Large Shareholders, Banks, and Managerial Moral Hazard: An Empirical Investigation', Yishay Yafeh (Hebrew University of Jerusalem) presented his findings on the nature of monitoring by large shareholders and debtholders in Japanese manufacturing firms. Rather than examining the indirect effects of monitoring on profitability, he studied the direct effects on firm behaviour, finding that both shareholders and banks holding large amounts of debt monitor firms across a decreasing range of activities, allowing scope for managerial moral hazard even if the monitored firm is not in financial distress. Denis Gromb (ECARE, Université Libre de Bruxelles) pointed out that the study could be extended to provide insights on the monitors' incentives to engage in monitoring activities. While Yafeh considers the fraction of shares held by the largest shareholder, a more precise account of the ownership concentration might provide evidence of free-riding among large shareholders. Moreover, different claims, like debt and equity, and the ownership structure of the monitors, could lead to different monitoring activity.

The next presentation was to a joint ECARE/CEPR lunchtime meeting, open to a broader audience of local academics and policy-makers. In `Corporate Governance in Germany', Jeremy Edwards (St John's College, Cambridge, and &nbspCEPR) &nbspquestioned &nbspthe popular view of the German model of capitalism as one in which banks, as shareholders and lenders, oversee companies. First, German joint stock companies account only for 10% of output and make little use of bank loan finance. Second, German banks are not major shareholders in joint stock companies: in aggregate, only about 10% of the shares of German companies are held by banks. Although it is possible that some large banks' exercise of proxy votes at shareholders' meetings results in good corporate governance, it is doubtful that there are sufficient incentives for banks to use their voting power for this purpose. Third, the role of banks is typically overemphasized relative to concentrated ownership: in total, there are only about 30 German companies which do not have a single shareholder with an ownership stake of at least 25%.

In the next paper , `Why Aren't Universal Banks Universal?', Sandeep Baliga (King's College, Cambridge) presented a moral hazard model where entrepreneurs can be financed by monitored (German) or nonmonitored (Anglo-Saxon) bank loans. The model assumes internal scale economies in monitoring, external ones on capital markets, and a trade-off between monitoring costs and agency costs from moral hazard. These assumptions imply that capital markets of the Anglo-Saxon style are efficient when there are a large number of small firms, and German-style loans are efficient for a small number of large firms. Persistence is possible in both types of economies, however. A German system can arise even when an Anglo-Saxon system would be efficient, and when firms can choose the scale of projects and banks can choose the method of finance, the Anglo-Saxon system can arise although consumers would prefer the German system. Ernst-Ludwig von Thadden (Universität Basel and CEPR) criticized the model for ignoring the crucial issue of bank refinancing, which makes the trade-off incomplete, and because the problem of market liquidity is oversimplified by assuming decreasing average costs of bond finance.

In `Firm Ownership Structure and Investment: Theory and Evidence from German Panel Data', Julie Elston (Wissenschaftszentrum Berlin) provided empirical evidence that independent firms in Germany were more liquidity-constrained than bank-held firms during the late 1970s and early 1980s. These results are consistent with Jeremy Edward's conclusion that the German system is not so different from the UK's in that liquidity constrains &nbspinvestment &nbspbehaviour for &nbspsome firms. Elston's results do refute the hypothesis that firm investment is independent of country specific institutional factors like the banking structure, however.

In `The Privatization of Ex-Zaibatsu Holding Stocks and the Emergence of Bank-Centred Corporate Groups in Japan', Hideaki Miyajima (Waseda University) investigated the Japanese post-war experience with the dissolution of &nbspthe zaibatsu and the resulting changes in corporate governance structures. Although an individual-centred ownership structure with employee ownership, equity finance and a market for corporate control was viewed as the most desirable system by the responsible institutions, a system of institution-centred ownership with cross shareholding, debt financing, and a main bank delegated monitoring system emerged in Japan. One of the most important factors for this development was the stock market collapse in 1949, which resulted in a transfer from individually held to institutionally controlled ownership.

In `Centralized Decentralization: Corporate Governance in the East German Economic Transition', Ernst-Ludwig von Thadden analysed the corporate governance structure of the Treuhandanstalt agency in charge of East Germany's transition to a market economy in a two-phase process. In the first phase, the Treuhandanstalt was set up as a strongly centralized privatization agency, while in the second phase, enterprises were tightly controlled until their privatization. Von Thadden argued that in order for such a centralized structure to be effective, two complementary features are important: a high degree of operational freedom and a time limit on the agency's existence. Gérard Roland (ECARE, Université Libre de Bruxelles, and CEPR) pointed out that the limited time horizon of the Treuhandanstalt may indeed be interpreted as a device against rent-seeking from insiders in state-owned enterprises (SOEs), since privatization reduces their lobbying ability. The commitment to speed may have had negative effects, however, allowing substantial rents to informed private investors who were able to purchase assets at very low prices. Treuhand officials were only too glad to accept quick deals to show `good' privatization performance, as measured by numbers of privatizations. Guido Friebel (ECARE, Université de Libre Bruxelles) pointed out that the Treuhandanstalt aimed to harden budget constraints of unprofitable enterprises through privatization. The buyers of such enterprises received high one-shot subsidies in exchange for employment and investment guarantees. Current experience in East Germany indicates that these firms' budget constraints will not be as hard as hoped, however.

In `Finance in Transitional Economies: The Case of Poland', Ron Anderson (Université Catholique de Louvain and CEPR) examined transition banking in Poland between 1989. He found that the state still dominates banking: 80% of bank assets are held by the banks which were spun off from the National Bank in 1989, and the state retains significant direct and indirect ownership shares in most of the banks newly created since 1990. Moreover, the state still provides indirect subsidies to favoured sectors, leading to soft budget constraints. Furthermore, the slow pace of privatization has restrained the development of equity markets, while defects in bankruptcy laws have resulted in creditors' passivity towards enterprises in financial distress. The Financial Restructuring Act removed some of these defects in 1993 and also recapitalized some of the largest banks. Nevertheless, the rapid increase of bad debts in 1994 makes it doubtful that the 1993–4 bank bailout will be the last of its kind in Poland.

In `Ownership and Control in Poland: Why State Firms Defied the Odds', Sweder van Wijnbergen (Universiteit van Amsterdam and CEPR) presented survey data of 64 large SOEs in Poland indicating that these firms have done better than expected. Painful adjustment measures were taken, including considerable labour shedding, but the data support the view that materials and energy use has become more efficient, while trends in unit labour costs are also encouraging. Two elements of corporate control have led to these developments. First, the state has stopped open-ended subsidies and contributed to wage restraint by an excess wage tax. Second, banks have played a powerful role in disciplining enterprises, but only after their own governance was reformed in the banking reforms at the end of 1991.

Wendy Carlin (Wissenschaftszentrum Berlin and CEPR) discussed Anderson and van Wijnbergen's divergent interpretations of banks' contribution to enterprise restructuring in Poland. Anderson's depiction of bad debts growing rather than contracting in transition stands in sharp contrast to van Wijnbergen's in which the banking system's commercialization in late 1991 was associated with a clear change of lending behaviour from simply funding losses towards using profitability as an indicator of creditworthiness. Some concern was expressed about the unqualified use by van Wijnbergen of profitability as an appropriate indicator of where resources should be allocated. Carlin pointed out that, in the sample used, successful firms were predominantly in sectors with lower import penetration and more rapid price increases, whereas unsuccessful firms could be found in more competitive sectors.

In a panel discussion on `Emerging Models of Capitalism: Eastern Europe's Choice' moderated by Richard Portes (CEPR and London Business School), Janusz Lewandowski (Institute for Market Economics, Gdansk, and former Minister of Privatization for Poland) focused on Polish privatization. Political constraints proved to be crucial for the reform of the enterprise sector. Due to insider resistance, the government was not able to recentralize ownership to allow subsequent privatization to outsiders. In fact, only about 10% of privatized enterprises were sold to outsiders. Still, given that these property rights are transferable, once capital markets are functioning, effective outsider control can emerge. This can be observed currently, though it is not banks which acquire ownership (for example, by debt/equity swaps), but large privatized companies. For instance, former &nbspforeign &nbsptrade organizations have developed into private conglomerates, which are more important for corporate control than banks.

Jan Mládek (Czech Institute of Applied Economics and former Deputy Minister, Federal Ministry of Economy, Prague) argued that the Czech government made use of a window of opportunity when designing and implementing the voucher privatization before insiders had organized their resistance. Voucher privatization resulted in a governance structure which is dominated by bank-controlled investment companies. These companies exercise control of the investment funds which govern enterprises. The funds exert control not only through supervisory boards, but also directly through the firms' management committees. In the long run, the issue of how to control these controllers will be decisive.

Peter Dittus (Bank of International Settlements) commented that competition in the banking sector is crucial for controlling the controllers and asked how the recent freeze of new banking licences in the Czech Republic and Poland should be evaluated against this background. Mládek answered that various bankruptcies have revealed the trade-off between competition and stability in the banking sector. Competition from foreign banks is thus very important. Van Wijnbergen asked to what extent banks in these countries can be considered fully private. Mládek stated that banks still redistribute their profits, which are often due to non-competitive margins, to loss-making enterprises. This shows that market behaviour has not yet totally replaced old connections. Portes raised the question of whether more of Poland's SOEs would be in private hands today if the window of opportunity had been used better in 1990. Lewandowski replied that stabilization had been considered critical in 1990, and that mass privatization through vouchers might have lead to threats of inflation at that time.

In `Suppliers' Associations in the Japanese Automobile Industry: Collective Action for Technology Diffusion', Mari Sako (London School of Economics) discussed the structure and functions of supplier's associations in Japan. She revealed that all Japanese car assemblers, except Honda, have such formal associations of part suppliers. Suppliers benefit from membership through better access to information about the assembler-customer, intra-suppliers exchange of know-how and technical guidance from the assembler. Hideaki Yamawaki (Université Catholique de Louvain) recommended analysing the disadvantages of such associations to complete the picture.

In `Rewards in the Afterlife: Late Career Job Placements as Incentives in the Japanese Firm', Mark Rebick (Oxford University) reported on his study of the Japanese labour practice of helping to find new jobs for men at the age of 60 after mandatory retirement. Rebick found that receiving assistance from the pre-retirement employer leads to 20% higher wages in the new job. Given that an average of five years is spent in postretirement employment, this wage differential has a present value to pre-retirement employees comparable to that of a major promotion. These results hold for both blue- and white-collar workers. Rebick concluded that the promise of post-retirement job placement may function as an important control device over older employees. Seiichi Kawasaki (Stirling University) pointed out that post-retirement placements are inseparable parts of the reward packages encouraging long-term competition among employees. Thus, they do not work as an incentive system for older employees, since the post-retirement jobs are already earmarked for winners of the long-term promotion race.