آزاد سازی مالی، کمک های خارجی، و تحرک سرمایه: شواهد از 90 کشور در حال توسعه

The methodology of Feldstein and Horioka (Econ. J. 90 (1980) 314) is used to gauge the degree of capital mobility and accessibility to international financial markets following financial liberalisation. The sample consists of 90 developing countries divided into four regions: Africa, Asia, Latin America, and the Middle East. The sample period is 1975–1995, split into two periods. Our results indicate that for developing countries capital is relatively immobile. There is also evidence that access to international financial markets increases following financial liberalisation. Finally, including foreign aid in the saving–investment regression has an important positive effect on the saving coefficient.

In the past three decades, many developing countries have liberalised their financial markets and, in addition, opened up their capital accounts. An expected consequence thereof is that these countries' access to international financial markets should improve. Potentially, there are many gains from increased financial integration. For instance, international capital mobility is crucial to global resource allocation, since it helps to smooth consumption and reduce risk. Furthermore, it allows for investment, and hence growth, beyond the premises of domestic saving. Theoretically at least, unrestricted capital flows facilitate specialisation in the production of financial services, and so benefit the international economy. Competition from abroad is introduced in the financial industry and innovation is stimulated. All this creates dynamic efficiency. Under the circumstances that the global financial market is able to properly price the risks and returns inherent in financial claims, global saving can be allocated to the most productive investments. Thus, potentially there are important welfare gains to be made from external financial liberalisation. In the literature, two main methods are used to gauge the degree of international capital mobility: The first is by studying the rates of return on capital across countries, a common approach when interest is in analysing financial capital flows. Second, by studying the correlation between domestic saving and investment rates. Since the focus is on long-run real capital flows, this paper dwells on the second approach. However, it should be noted that even with external financial reform the possibilities for capital flow might be limited by obstacles, such as transaction costs, taxes, and official restrictions. The main objective of this paper is to gauge the degree of international capital mobility in the developing world. Another important objective of this paper is to determine whether financial market liberalisation has led to increased international capital mobility in the sense of Feldstein and Horioka (1980). To do this we use a dataset that runs from 1975 to 1995 and that consists of 90 developing countries from Africa, Asia, Latin America, and the Middle East. First, using cross-section as well as panel-data techniques the entire sample is employed to investigate the overall degree of capital mobility and also whether capital mobility has increased over time. Thereafter, to allow for inter-regional comparisons, the paper makes use of the four aforementioned geographical regions and repeats the exercise for each region. The first step follows previous research in its objective, but the dataset used for it, to our knowledge, is the largest used so far. Previously used datasets of similar size have been combinations of industrial and developing countries, whereas this dataset exclusively consists of developing countries. The second step, with its regional comparisons, is a clear extension of the literature. It reveals the different timing and evolution of capital mobility across regions following financial liberalisation, something that is otherwise hidden in full sample estimations. It also remedies any pooling of the data that is unjustified, at least the one that pertains to structural regional differences. Another extension of this paper is direct inclusion of foreign aid in the saving–investment regression. We do this because in many developing countries a significant part of investment financing comes from non-market flows, such as foreign aid. To the extent that foreign aid is important the saving–investment correlation will weaken and the Feldstein–Horioka regression will be mis-specified. As will be shown in the paper, the regression coefficient of saving on investment indeed increases when including foreign aid in the regression. The paper is organised as follows: Section 2 provides a theoretical background to the Feldstein–Horioka hypothesis and reviews the methodologies used to test it as well as alternative interpretations and critiques. Section 3 describes the data, while Section 4 discusses results from cross-section and panel-data analyses. Finally, Section 5 concludes the paper.

The purpose of this paper has been to gauge the degree of capital mobility and to determine whether financial liberalisation in developing countries has enhanced access to the international financial markets. High capital mobility and access to offshore funds would be reflected in low correlation between the domestic investment rate and the domestic gross saving rate. An economy with low capital mobility and, hence, high correlation between domestic saving and investment, would have a slope coefficient near unity. Enhanced access to international financial markets following financial liberalisation would be indicated by a falling correlation coefficient over time. The sample used to test this postulation covers 90 developing countries over the period 1975–1995, which is the largest sample used so far for this purpose. While the first step of the analysis followed previous literature, the second step introduced an innovation by estimating savings–investment correlations for Asia, Latin America, Middle East, and Sub-Saharan Africa separately. We included an interaction term do obtain evidence on whether the slope parameter is constant over time and enhanced the Feldstein–Horioka regression by the inclusion of foreign aid. Finally, we used poolability tests to obtain a ranking of regions with respect to relative openness. The results indicated that, except for the Middle East, overall capital mobility is low. However, it was shown that financial liberalisation improves access to international financial markets. This was true for the whole sample as well as for all regions individually. While we could easily conclude that the Middle East was most open, it was hard to obtain a ranking of Asia, Latin America, and Sub-Saharan Africa. Nevertheless, the conclusion for these regions' countries it that they still have quite a low capital mobility. The results obtained in this paper are thus consistent with common expectations with respect to capital mobility in developing countries. To a lesser extent, however, are the results consistent with the results of many other studies investigating the degree of capital mobility in developing countries. Finally, the paper made the important point that foreign aid should be included in the saving–investment regression. Omission of foreign aid, when aid flows are important in explaining investment in developing countries, would downward bias the saving coefficients and lead to the impression of more openness than is actually the case. Indeed, inclusion of foreign aid had the expected effect of increasing the slope parameters.