Synchronization of business cycles: A spatial analysis

In this paper we extend and complement the existing literature by considering the synchronization of business cycles across multiple countries adopting spatial error and spatial lag models. Spatial analysis considers that interaction of observations (or spatial dependence) occurs across several locations in space. Distance matters in the sense that locations geographically close to each other exhibit similar values of variables of interest. Strength of spatial interaction declines as locations are more distant from each other. We start with basic spatial analysis which primarily relies on geographical distance and extend our model to consider "economic connectivity" based on trade relationship. For robustness checks, we look at three different measures of business cycles: Hodrick-Prescott filtered output volatility, output growth volatility, and volatility of residuals from a growth regression.

With data from 187 countries over the period of 1960-2007, maximum likelihood results suggest a strong spatial dependence of business cycles across countries. With the long span of our data, we are able to divide the sample into three subsamples to explore the dynamics of business cycle comovements. Over the three subsample periods in our study, the importance of geographical distance in affecting the comovements of business cycles is increasing from the Bretton Woods era (1960-1972) to the period of international shocks (1973-1986), while declining afterward (1987-2007). Instead, during the globalization era (1987-2007), economic distance matters more than geographical distance. Over the period of 1987-2007, countries located close to each other may not necessarily share similar patterns of business cycles. Instead, countries that have a close economic tie (measured by trade relationship) are more likely to exhibit a distinct comovement of their economic fluctuations. Furthermore, the economic connectivity and comovement of business cycles among OECD countries and between OECD and the rest of the world are stronger than the economic connectivity among non-OECD countries and between non-OECD countries and the rest of the world.