1 Introduction
This paper focuses on the fundamental determinants of degree of financial integration -- or more precisely
of the intensity of capital controls -- in the European Union (EU) over the period 1973-1993.1 Macroeconomic
evidence seems to support the view that the integration between European financial markets has increased
in recent years (see Lemmen and Eijffinger, 1993, Frankel, Phillips and Chinn, 1993 and Lemmen and
Eijffinger, 1995a). It fosters the impression that remaining differences in national economic and financial
structures are unimportant. Moreover, it may tend to overstate the pressure towards, and hence the speed
of integration. An obvious question, then, is what are the main determinants of financial integration. A
thorough understanding of the determinants of the intensity of capital controls may provide important insight
into the process of financial and monetary integration in Europe and may help in policy formulation.
This paper improves analysis started by Epstein and Schor (1992) and Alesina, Grilli and Milesi-Ferretti
(1994) in a number of aspects. First, previous investigations of the determinants of capital controls (Epstein
and Schor, 1992, Alesina, Grilli and Milesi-Ferretti, 1994, Gruijters, 1994, Milesi-Ferretti, 1995 and Grilli
and Milesi-Ferretti, 1995) typically constructed dummy variables or capital control indices to measure the
degree of financial integration. As we will argue, these measures are problematic because they do not account
for different intensities of capital controls. In our opinion, this aspects is of crucial importance for policy
analysis. We compute deviations from closed interest parity to measure the intensity of capital controls.
Resulting negative (positive) deviations from closed interest parity are associated with capital export (import)
restrictions. Second, Alesina, Grilli and Milesi-Ferretti (1994) apply their analysis to the financial markets
of 20 OECD countries. Grilli and Milesi-Ferretti (1995) apply their analysis to 61 developed and less-
developed countries. In practice, integration attempts across the world are more of a geographical nature.
The EU financial markets can be seen as an excellent sample to test for the fundamental determinants of
financial integration because they are in the process of institutional integration. Most legal barriers have
been either removed or are scheduled to be removed. Furthermore, the EU has adopted a stronger form
of harmonization for its financial services -- a policy of mutual recognition whereby member states within
the EU have agreed to allow financial intermediaries from other states to operate under home country rules
and supervision. Third, Alesina, Grilli and Milesi-Ferretti (1994) apply Maximum Likelihood Estimation
to estimate various logit/probit models. We innovatively apply a panel data approach to estimate various
fixed-effects models. The panel data approach allows us to include those author’s political variables (the
country-specific effects) along with other explanatory variables. Furthermore, several new explanatory
variables were included in our analysis: the realized depreciation of the exchange rate, the unemployment
rate, the productivity in the business sector, the ratios of government deficit, domestic credit and broad
money over gross domestic product and the ratio of broad money over narrow money. Contrary to Alesina,
Grilli and Milesi-Ferretti (1994) and related papers who emphasize the importance of various political
variables, we argue that various measures of national economic and financial structure have considerable
explanatory power in the determination of capital controls. We find high realized inflation rates, high
government deficits, high current account deficits, high credit expansion to the economy, low productivity
in the business sector and low availability of sophisticated deposit instruments to be the main determinants
of the intensity capital export restrictions. According to the evidence, remaining differences in national
economic and financial structures, should be of greater interest to policymakers.2
The paper is organized as follows. Section 2 provides a meaningful measure of the degree of financial
integration, which is the dependent variable in our empirical analysis. This paper focuses on the price aspect
of financial integration. That is, financial integration is expected to lead to price convergence. As we will
argue, the price measure -- i.e. the closed interest rate differential -- is particularly suited to measure the
1 The concepts of "financial integration" and "intensity of capital controls" are used interchangeable.
2 Our argument is also demonstrated by the events during the exchange crises in the European Monetary System.