Inflation Convergence Across Countries

Ball and Sheridan (forthcoming) show that OECD countries with a history of high inflation before the 1990s have subsequently experienced a larger degree of disinflation than countries with a history of low inflation. They label this process ‘regression to the mean’, and argue that it explains why those OECD countries which adopted inflation targeting experienced larger falls in inflation compared to other OECD countries. This paper explores further this phenomenon of convergence of inflation rates across countries. An extension of the analysis used by Ball and Sheridan finds that convergence in inflation rates also occurs in a much larger sample of countries over the same time period that they study. However, tests using data from the 1960s to the 1980s show that inflation convergence across OECD countries does not occur consistently over time. In contrast, when this analysis is repeated with data for US metropolitan regions, convergence occurs consistently since the early 1960s. Interpreted in the context of historical developments in monetary policy, the evidence suggests that rather than just being a mechanical occurrence, the convergence in national inflation rates experienced in the 1990s was brought about by monetary policy becoming more similar across countries, with authorities becoming more focused on achieving low inflation, especially in those countries which had been less successful in reducing inflation in the 1980s. In other words, the adoption of inflation targeting at least partly contributed to inflation convergence in the 1990s.