Have UK and Eurozone Business Cycles Become More Correlated?
Massmann, Michael, Mitchell, James, National Institute Economic Review
Recent estimates suggest that the UK business cycle is closer to the Eurozone business cycle than it was in the early 1990s. This paper investigates whether this phenomenon has been accompanied by increased correlation between UK and Eurozone business cycles. Considering a range of alternative measures of the business cycle we find, using 40 years of monthly industrial production data, no clear evidence for a sustained increase in correlation between UK and Eurozone business cycles. Instead, in the 1990s, the correlation between UK and Eurozone business cycles has been volatile relative to historical levels. It is only recently, i.e. since 1997, that the UK has become more correlated with the Eurozone, although the level of correlation is lower than against non-Eurozone countries. Importantly, the strength of these relationships is sensitive to how the business cycle is measured. Care should therefore be exercised when using business cycles estimates to test the relationship between UK and Eurozone business c ycles.
UK entry into European Monetary Union (EMU) is contingent on both parliamentary and popular votes, and the Government's Five Economic Tests being judged as satisfied. The first of these much discussed, and interpreted, tests asks "whether there can be sustainable convergence between Britain and the countries of the single currency". (1) An alternative version of the first test, written by HM Treasury (1997), asks, "Are business cycles and economic structures compatible so that we and others could live comfortably with euro interest rates on a permanent basis?" (2)
A possible measure of compatibility between business cycles is the absolute size of the cyclical disparity between the UK and the Eurozone. Indeed, available evidence indicates that this has declined to historically low levels. The most recent estimates, for the beginning of 2000, suggest that the UK business cycle is closer to the Eurozone business cycle than it was five or ten years ago. Importantly, the UK is 'closer' in the sense that the root mean squared difference between the UK business cycle and that of the Eurozone, as well as its largest economy Germany, has declined relative to peaks at the time of German unification and the UK recession in the early 1990s, see chart 1. The period of stability accompanied by the UK's move away from exchange-rate to inflation targeting, and Bank of England independence, has decreased the amplitude of cyclical variations and brought UK and Eurozone business cycles closer together. (4)
In this paper we investigate whether accompanying this decline in the root mean squared difference between the UK and Eurozone business cycles there has been, or indeed is growing evidence for, increased correlation between them. As a check on our results, we also examine whether the correlation between the UK cycle and those of important non-Eurozone countries, such as the US, has decreased. In addition we ask whether the phase between the business cycles has shifted over time. This lets us consider the possibility that the UK might have moved increasingly in line with the Eurozone but at a lead or lag.
There is related work that looks at the correlation between the UK's business cycle and those of other industrial nations. In particular, Artis & Zhang (1997, 1999), using monthly industrial production data from 1961 to 1995, find that the UK business cycle is more in line with that of the US than the Eurozone countries. Nevertheless they confirm that membership of EMU, or the ERM before it, has promoted correlation between participating countries' business cycles, in particular through the discipline imposed by a fixed exchange-rate that helps transmit shocks from one country to another. (5) Certainly it is widely believed that a monetary union can only operate successfully if the business cycles of the member countries are correlated; see, for instance, Christodoulakis et al. …