The COE project is being organized into a vertical-and-horizontal matrix, grouped by country and by theme. The work of the Finance Group will become part of the vertical theme category, vertically spanning cross-country analysis. As Prof. Odaka has previously stated, we intend for the theme-centered analysis to make major theoretical contributions to historical economic research, so the Finance Group has adopted an approach which entails both compiling estimates and conducting analysis.

We will not deal with prewar data for the time being, but will first gather data from the wartime and early postwar periods. The main reasons for choosing this approach are: (1) The influence of colonialism means that the prewar and postwar data sources are different for some countries; (2) Again, because of colonialism, there are major differences in prewar and postwar economic systems for some countries; (3) Prewar information is not necessarily important for the major theoretical issues (discussed below) we are most interested in addressing at present. Nevertheless, the issue of verifying early conditions remains important, and we intend to gather as much information as possible from early periods on colonialism's institutional legacies and its influence on the intensity and diffusion of indigenous financial practices.

The ten subject countries in this area are South Korea, Taiwan, China, Thailand, the Philippines, Indonesia, Malaysia, Singapore, Japan, and India. The statistical work will be divided into compiling basic and applied time series. The basic time series for the financial systems of all ten countries will initially be compiled to establish a unified standard permitting the maximum degree of vertical comparison. The category names will be amount of currency, interest, price levels, exchange rates, reserve requirement, number of banks (numbers of head offices and branches), central bank balance sheets, deposit bank balance sheets, nonbank balance sheets, securities market data, matters related to government debt, and central government revenues. Included in applied time series will be sectoral data, data on term structures, data on direct credit systems, major enterprise data, and profit and loss statements of banks. This information is collected as necessary for individual country analysis.

So what can we expect to learn from an analysis of the financial data collected for these ten Asian countries? Following are results from our historical analyses of each country and from the analysis derived from our statistical cross-country methodology.

First, I'll discuss our individual country analysis. There are two basic topics. The first is the issue of the reform and evolution of financial institutions. The most important issue in this instance is to apply close study of historical institutional facts to discern patterns of institutional reform. This requires detailed fact-finding on the passage and implementation of basic laws as well as policymaking processes and the public's response. Viewed in a long-term perspective, two basic trends can be seen in the financial institutional reforms of developing countries. One is a shift from reforms promoting efficiency in providing loans to reforms favoring the mobilization of funds. More specifically, in the 1950s and 1960s, the dominant approach was to promote expansion of the directed credit system so that its information functions would yield a more efficient distribution of resources; but dissatisfaction with the results led from the 1980s, in the midst of burgeoning free market and liberalization ideologies, to a change toward emphasizing institutional reforms, especially reforms seeking to mobilize deposits. Second, there is a trend toward a progressive division of labor at first, and then toward universal banking. Until the 1960s, strong efforts were made to construct financial systems capable of performing differentiated tasks in areas such as directed credit systems, savings institutions, and housing finance. But beginning in the 1980s, the trend shifted toward encouraging the integration and universalization of functions. The first task in investigating financial institutional reform is to investigate basic patterns in each country and find out whether they accord with the patterns described above.

After verifying the nature of the basic patterns, the next task is to find out what doctrines and needs exerted the greatest impact, and who brought forth the financial institutional reforms. In other words, it is necessary to pursue investigation into the mechanism of institutional evolution and the subsequent political economic implications. For this purpose, we need to think about the influences of doctrines (for example, liberalization doctrine) on the efficiency of financial institutions. We must also consider responses to the external shock of the globalization of finance, based on advances in communications technology, and to changes in international financial standards.

One of the core topics in individual country studies at present is analysis of the relationship between capital procurement and growth. The first step in such investigation is verifying a country's long-term financing patterns. It is necessary to investigate a number of fundamental factors such as the weight of financing from domestic and foreign markets, the contrast between informal and formal markets, comparisons of regulated finance as opposed to finance generated through market mechanisms, internal versus external funds, and the relative merits of capital provision toward the government versus provision toward the private sector. After investigating the above factors, we will proceed to analysis of the relationship between real growth and financing patterns for individual countries. There are a large number of important topics. First, for example, is the relationship between financing patterns on the one hand, and the investment and savings rates, changes in industrial structure, and changes in growth rates on the other. A second issue requiring investigation is how financial systems responded to information costs and agency costs. And third, we need to think about the issue of relative magnitude of ex ante supply of and demand for capital on the various development phases. Regarding long-term aspects, there was excess demand for long-term capital at least, but we also need to undertake fresh thinking on other phases of development and other types of funds.

In addition to the above investigations, which deal with individual countries, three cross-county statistical comparisons are planned. First, we plan to conduct a comparative analysis of financial development patterns in Asian countries by conducting a basic statistical time series analysis of internationally comparable processes including the diffusion of banking institutions, the monetarization of economies, real interest levels, and financial system diversification. Second, we will apply statistical analysis to a study on the relationship between financial variables and growth. As a first step, we will substitute financial variables for variables related to infrastructure or education in a Barro-type endogenous growth model. Such an analysis is justified because it is possible that financial variables, by enhancing the efficiency of resource allocation, may be able to prevent diminishing returns. There is also the problem of initial values recently emphasized by D. Rodrik. This view holds that from early periods in Asia education levels were very high and resources very well allocated, and that these factors were related to high growth. We plan to conduct a quantitative economic check on whether early financial development levels (including prewar development levels and influences of colonial policies) became the origins of virtuous growth cycles. These analyses will be conducted by utilizing panel data.

The third and final theme will be an analysis of the effects of regulation and deregulation on financial development and economic growth. Some important ideas in this regard include R. McKinnon's emphasis on the mobilization of financial savings by raising real interest on deposits to equilibrium level, and achieving an efficient allocation of funds; and Stiglitz's belief that real interest rates on deposits should be raised but kept below equilibrium level, and that the contingent rents provided this way would work as incentives for investment in monitoring capability or branch networks. While McKinnon believes that liberalization contributes to growth, Stiglitz feels that some degree of regulation is necessary in financial markets. It is also possible to apply the quantitative investigation on the panel data of Asian countries before and after liberalization. One of our members is already conducting a preliminary analysis.

Hitotsubashi University, Institute of Economic Research