The Electoral Information Hypothesis Revisited

The connection between political and bond market equilibration is studied. The conventional wisdom on the subject, the Electoral Information Hypothesis (EIH; Cohen 1993, Alesina, Roubini and Cohen 1997), is criticized on theoretical and empirical grounds. Using a framework proposed by Garrett and Lange (1995) and data from three of the most developed bond markets in the world a revised hypothesis is constructed and tested. The revised hypothesis recognizes the way formal political and bureaucratic institutions mitigate the effects of democratic politics on bond markets as well as the “fat tailed” distributions and volatility clustering commonly found in financial time series. The results show that empirical support for the revised EIH extends beyond the American case, but this support is not universal. More specifically, where democracy is of the majoritarian type and central banks are weak (the U.K. until recently) the revised EIH holds. Once placed on sounder statistical footings, the U.S. case—one of mixed majoritarian and consensual democracy and a moderately strong central bank—also supports the revised EIH. However, the revised EIH does not receive support in Germany, which has a consensual form of democracy and a strong central bank. The empirical power of the revised EIH thus is shown to vary by institutional context.