Recent progress towards economic and monetary union in Western Europe has highlighted the importance of regional policies, while the transformation of Eastern Europe has focused attention on the role of finance in development and may also entail a substantial diversion of international capital flows. Recent research on endogenous growth has addressed the relationships among growth, convergence and capital market imperfections, focusing on financial intermediaries' abilities to enhance the economy's production opportunities, capital markets' contribution to the real economy, and the effects of gaining access to international capital markets on developing countries. Many of these issues were discussed at a conference on `Finance and Development in Europe', held in Santiago de Compostela on 13/15 December 1991, which was made possible by financial support from the Regional Government of Galicia. The conference was organized by Alberto Giovannini, Associate Professor of Economics and Finance at Columbia University and a Research Fellow in CEPR's International Macroeconomics programme, and Guillermo de la Dehesa, Conseyo Delegado of the Banco Pastor and a member of CEPR's Executive Committee.

In his paper, `A Theory of Financial Development', Oren Sussman (Hebrew University of Jerusalem) developed a model in which the size of the market for financial intermediation and the number of banks both rise with increases in the capital stock; as each bank becomes more specialized and efficient, the industry becomes more competitive, the cost of intermediation falls and the `gross mark-up' the gap between banks' rates to borrowers and lenders reduces. With constant returns to scale in monitoring and perfect competition in banking, the mark-up and the share of banking in national income are unaffected by the growth of the capital stock. For a `spatial' model of a monopolistically competitive banking system, however, with geographically dispersed borrowers, monitoring costs increase with the distance between banks and firms, so banks enjoy market power. As they get closer together, however, monopoly rents fall, new banks enter, each serves a smaller market share, and the mark-up falls.

Sussman reported that calculating mark-ups for 81 countries from IMF data and testing these predictions econometrically indicated that a $1,000 rise in per capita GNP reduces the gross mark-up by about 0.2%, at least among middle- and high-income countries. He maintained that this weak negative correlation is in fact a strong result: intermediation is a non-tradable, whose relative price is strongly and positively correlated with GNP per capita, so rising labour costs should raise the cost of intermediation. These data suggest, however, that the effect of financial development dominates this rise in costs.

In his joint paper with Alec Levinson and Joseph Stiglitz, `Capital Market Imperfections and Regional Economic Development', Bruce Greenwald (Columbia University) noted that `traditional' theory assumes that common interest rates imply the neutrality of monetary policy across regions, while arbitrage in financial markets will offset the effects of any deliberate attempts to stimulate regional capital investment. With imperfect and asymmetrically distributed information, however, local institutions may have incentives to invest in their `home' regions to take advantage of superior information; the opportunities they pass up will then be adversely selected, with below-average rates of return that discourage outside investment. Such informational imperfections lead to sub-optimal capital allocations but also allow interventions to have positive effects that compensate, at least in part, for these misallocations.

Greenwald described their econometric tests of a model of capital market competition for data on manufacturing industries in the member states of the US during 1972-82. Their results indicated that aggregate local employment fluctuations were significantly influenced not only by the aggregate national employment in the sectors constituting the local economy, but also by aggregate local employment growth. This suggests that deteriorations of locally dominant industries in the US (such as the automobile industry in Michigan) impaired the financial health of local banks, thus reducing the availability of capital to other local industries. Capital market barriers of this type may be expected to impede efforts towards regional development in Eastern Europe, within the European Community and within existing national boundaries elsewhere.
Policies to transfer income and capital towards Italy's `Mezzogiorno' region have persistently failed to reduce its income gap with the North: despite massive external aid, total factor productivity has remained consistently lower, even in private manufacturing. In his paper, `Finance and Development: The Case of Southern Italy', written with Giampaolo Galli and Curzio Giannini, Riccardo Faini (Università di Brescia and CEPR) investigated whether the peculiarities of the region's financial sector may have impeded the channelling of capital to firms and projects with high social rates of return. Italy has more than 1,000 banks, of which the smaller ones including many with only one or two branches are concentrated in the South. Their operating costs as a share of total funds exceed those of banks elsewhere by about 20%, mainly due to differences in physical productivity, and their average loan quality is considerably worse. The average rates on banks' loans in Italy's 95 provinces are strongly correlated with their per capita GDP: this is explained in part by size and sectoral composition, since smaller and riskier borrowers are concentrated in the South, but more than half the observed differential must be attributed to weaker competition and higher bank costs. Greater information problems lead to more intense rationing, captive relations between banks and firms, and poorer screening; and even Southern firms borrow at less favourable rates from local than from external banks.

Faini reported that the results of their cross-section regression on individual loan contracts also provided indirect but strong evidence that screening is less efficient in the South. He concluded by calling for a reduction in government intervention: liberalization and increased competition are necessary, but not sufficient, for Southern Italy's financial system to allocate resources more efficiently.

In `The Role of Finance and Economic Development in South Korea and Taiwan', Yung Chul Park (Korea University, Seoul) analysed the role of the financial sector in these countries' rapid economic development, in particular its effectiveness in mobilizing savings and its efficiency in allocating credit. The intermediated credit market has played an important role in both economies because of informational asymmetries between lenders and borrowers. Even after a decade-long promotion of liberalization, however, their governments continue to play dominant roles in their financial systems: controlling interest rates, interfering in the allocation of funds, regulating entry, and determining the types of services offered.

Despite its severity and protraction, financial repression appears not to have been detrimental to growth and industrialization. Park attributed this curious result to the design of government intervention specifically to promote export- led development. In a credit-rationed economy, firms' success depends critically on access to credit, and export credit incentives in these countries have been contingent on export performance, which has made the cost of credit allocation to different industries and activities more transparent to policy- makers. Park found no evidence that the financial sector's growth has had positive effects on savings behaviour or raised the average productivity of investment: it appears rather to have adjusted passively to changes in the real sector.

Presenting his paper, `Regional Imbalances and Transfer and Compensatory Policies: The Case of Spain', written with Juan Ramón Cuadrado and Andrés Precedo, Guillermo de la Dehesa examined the behaviour of regional income differences across the European Community and within Spain in particular. Using data on GDP per inhabitant for 171 Community regions for 1968-88, he showed that regional income disparities were more than double those in the US. Such differences decreased until 1975-7, but they remained stable from the early 1980s onwards: regional convergence has reversed and now displays a `core periphery' pattern. In Spain, regional incomes converged strongly during 1964-79, although with significant concentration in the more industrial areas; convergence halted in 1979, worsened in 1983 and remained stable thereafter. Decomposing the behaviour of GDP per inhabitant into output per worker, participation and unemployment rates indicated that the early tendency towards equality of output per worker was reinforced by the other factors; from 1977 onwards output per worker continued to converge, while the other components operated in reverse.

De la Dehesa noted that taxes, public expenditures and the distribution of public investment have always had strong regional components in Spain, whose differentiated regional effects may have dominated those of more explicit regional development policies. An analysis of regional GDP and family disposable income per inhabitant during 1967-87 confirmed the clear rise in disposable income in certain backward regions due to transfers from the rest of the country. Regional data on public capital stock and investment are imperfect, but they also suggest a negative, significant correlation with GDP, indicating that public investment policy has been oriented towards inter-regional redistribution.

In `Convergence in the Closed and in the Open Economy', Daniel Cohen (Ecole Nationale Supérieure, Paris, CEPREMAP and CEPR) developed an overlapping generations model of growth for a small country whose output depends on human and physical capital and hours worked. For an open economy facing a constant world interest rate, investors instantaneously shift funds to close the gap between this rate and the marginal productivity of capital; for a closed economy, the capital stock adjusts more slowly, but the conditions for convergence towards a steady state are the same. Gaining access to world financial markets will therefore not enable an economy converging towards a steady state to `take off' on an endogenous growth path.

Cohen noted that previous studies of the effects of human and physical capital accumulation have assumed that these are fixed as proportions of income and found that growth converges to a steady-state value at a speed determined by the divergence between steady-state and initial income. If human capital accumulation as a fraction of GDP is an endogenous function of the economy's level of development, however, the closed and open economies may behave quite differently. Cohen's test on a sample of 66 large, non-oil producing countries for 1970-87 and on a sub-group classified by the World Bank as `severely and moderately indebted' indicated that physical capital growth in the high-debt countries was below average. Paradoxically, it was human capital accumulation that differed between the debtor countries and the sample as a whole.

In `International Financial Integration and Economic Development', Paul Krugman (MIT and CEPR) noted that the case for capital inflows as the key to development assumes that countries' production functions are the same, so that capital-scarce nations can export labour-intensive products in return for capital- intensive goods. In fact, poorer countries simply have worse production functions, so their marginal products of capital are lower than their capital/labour ratios suggest, and growth accounting also indicates that the pay-off to any increase in capital accumulation will be very small. Krugman also expressed scepticism about recent models in which external economies to capital accumulation cause the elasticity of output with respect to capital to exceed its share of GNP at market prices. Even if the social return to capital is much higher than its private return, if both are approximately constant then capital will tend to move to large economies rather than poor ones, and global financial integration may even promote inequality in per capita income.

Krugman then noted that history suggests that large-scale capital movements are unlikely to take place, even if they are beneficial. Even during 1972-81, the IMF's 15 `highly indebted countries' ran current deficits averaging only 3% of GNP, which financed less than 15% of their domestic investment. Before 1914, capital flows were much larger relative to income, but very little capital flowed from rich to poor countries: most flowed from Europe to the North America, Argentina and Australasia, which already had comparable or higher per capita incomes. Krugman concluded that international financial integration is potentially mutually beneficial, but both history and theory indicate that its importance has been grossly overstated.

In his paper, `The Effects of Competition for Funds Between Eastern and Southern Europe', John Flemming (European Bank for Reconstruction and Development) noted that Eastern Europe's entry into world financial markets might be expected to raise the cost of commercial funds to all borrowers and divert official aid away from Southern Europe, while the two regions might also compete in product markets. The East's higher levels of human capital suggest a greater potential for high-technology products if it has access to sufficient capital, however, so even if the South `loses' in the competition for capital it may nevertheless `gain' from reduced competition in product markets. Abolishing quotas on Eastern exports to the West would make the East a more attractive location for investment, which would make it more competitive in the capital market and also push up its real exchange rates and effective real wages, thus reducing its competitiveness in low- tech products.

Flemming maintained that improving the efficiency of investment could lead to a rapid growth of Eastern Europe's effective capital stock, raising output growth and voluntary savings and leading to a virtuous circle reminiscent of the West German recovery after World War II. The effects of capital and goods market competition might then be offset by market expansion and increasing returns, which would spill over into increased demand for capital goods and consumer durables in Europe as a whole.

In his paper, `Financing and Development in Eastern Europe and the Soviet Union: Issues and the Role of the EBRD', written with Robin Burgess, Philippe Aghion (European Bank for Reconstruction and Development and CEPR) explored the relation of finance to the problems encountered in recent efforts towards privatization and restructuring in the region. State-owned banks have no experience of operating on a commercial basis, so new private enterprises cannot secure access to credit, while the absence of legally enforceable contracts discourages sub-contracting and hence impedes specialization; most economic activity still remains in the hands of state-owned enterprises. The low incidence of bankruptcies probably reflects the expansion of inter-firm credits due to creditors' well-founded belief that the government will eventually bail out debtor firms, and no credible enforcement of the new bankruptcy laws can be expected until the social costs entailed by liquidation are substantially reduced. Until prices are market determined, it is in any case hard to identify which firms should survive and which need to be liquidated.

Aghion noted that privatization of state-owned enterprises remains quite limited, and the approaches designed at the beginning of the transition proved overly complex and ambitious. He suggested privatizing small firms through multi-part auctions, which are best organized at a local level. For medium to large enterprises, the most viable companies or sectors should be privatized quickly, before their skilled management and labour force move to other firms. Other viable state-owned enterprises should be restructured by writing off their inherited debts and breaking up those that not natural monopolies before privatization.

In his paper, `Taxation, Money, and Credit in Liberalizing Socialist Economies: Asian and European Experiences', Ronald McKinnon (Stanford University) suggested that former centrally planned economies in Asia may fare better in the transition than their East European counterparts, since the absence of major civil disorders and unmanageable external shocks may allow them a freer choice of liberalization strategy. In socialist economies, taxation is largely implicit and the system of enterprise credit entirely passive: as liberalization devolves effective property rights to enterprises, the government gives away its tax base, but with interest rates pegged below market-clearing levels it cannot issue bonds directly to cover its fiscal deficit. Governments therefore borrow from the banking system, which issues savings deposits and cash balances. Because of this monetary overhang, even reformist governments are tempted to maintain price controls to avoid inflation, which impede the reform process.

McKinnon therefore proposed establishing a taxation authority at the outset of reform: personal income tax should become the principal tax base as in the West, supplemented by a non- distortionary, value-added tax on enterprises. During the transition, the coexistence of both `traditional' and `liberalized' enterprises complicates the regulation of the provision of finance: free trade should prevail on the commodity account between the two sectors, but the government must monitor and limit the traditional sector's cash deficits to maintain control over the aggregate supply of internally convertible money. Liberalized enterprises should initially finance their investment only from the non-bank capital market and retained earnings; the premature development of ordinary commercial banking could lead to a disastrous loss of overall monetary control and encourage irresponsible bank lending.


The papers presented at this conference will be published by Cambridge University Press early next year, in a volume edited by Alberto Giovannini.