The Determinants of UK Business Cycles

Abstract

This paper considers the causes of post-war UK business cycles. Using an extended stochastic growth model we construct estimates of productivity and preference shocks both of which are highly persistent, volatile and potentially capable of explaining UK business cycles. We find the productivity term is the dominant explanation of UK output fluctuations but our estimated preference shift is crucial in understanding employment movements. A variety of Granger causality tests establish whether these productivity and preference terms are unpredictable and so can be potentially considered as the cause of UK business cycles or whether they are themselves Granger caused by other variables. The aim of business cycle theory is to identify the causes of business cycles and outline the various propagation mechanisms through which these produce cyclical fluctuations. The purpose of this paper is to study the former-using UK data we investigate the main sources of UK post-war business cycles. Our aim in doing so is to discriminate between competing theories of the business cycle as well as to increase our understanding of particular episodes of UK post war economic history. To achieve this aim we focus on two fundamental sources of business cycle fluctuations: productivity shocks and a preference shift, or more generally a supply and demand shock. We choose these because (i) they can be easily calculated using only published data, avoiding the need for any econometric identification assumptions (ii) they represent the simplest extension to the neoclassical stochastic growth model which can explain observed UK business cycle patterns. They therefore form a natural benchmark with which to assess Prescott's (1986) claim that this model can account for both growth and business cycles. However, while our main focus is on these two shocks our methodology is broad enough to allow for a wide range of potential causes of UK business cycles. In Section 1 of the paper we outline some stylised facts of the UK business cycle and motivate more rigorously our focus on productivity and preference shocks. In Section 2 we provide estimates of these shocks, discuss their univariate properties and examine their correlation with key business cycle variables. We find both shocks to be very volatile and persistent with the productivity variable accounting for a large proportion of output fluctuations and preference shifts more important in explaining employment movements. Simulations show that in combination with a neoclassical growth model these shocks are capable of matching UK business cycles. However, if