Home | Business News | Browse by Publication | I | International Advances in Economic Research

Common trends and common cycles in per capita GDP: the case of the G7 countries, 1870-2001.

Article Excerpt
Abstract In this paper we analyze per capita incomes of the G7 countries using the common cycles test developed by Vahid and Engle (Journal of Applied Econometrics, 8:341-360, 1993) and extended by Hecq et al. (Oxford Bulletin of Economics and Statistics, 62:511-532, 2000; Econometric Reviews, 21:273-307, 2002) and the common trend test developed by Johansen (Journal of Economic Dynamics and Control, 12:231-254, 1988). Our main contribution is that we impose the common cycle and common trend restrictions in decomposing the innovations into permanent and transitory components. Our main finding is permanent shocks explain the bulk of the variations in incomes for the G7 countries over short time horizons, and is in sharp contrast to the bulk of the recent literature. We attribute this to the greater forecasting accuracy achieved, which we later confirm through performing a post sample forecasting exercise, from the variance decomposition analysis.

Keywords Common trends Common cycles * Permanent and transitory components

JEL C10-C22-E30

Introduction

Identifying the forces that induce fluctuations in macroeconomic aggregates has become an important topic in macroeconomics because its motivation is to investigate the relative importance of supply (permanent) and demand (transitory) shocks in the generation and propagation of business cycles. Keating and Nye (1999) employ post-war and pre-war period data to examine the impact of demand and supply side shocks on output for the G7 countries. Their main finding is demandshocks explain most of the short-run output variance for Italy (77%), the United Kingdom (UK) (97%), Germany (67%) and France (74%), and large proportions in the case of the United States of America (USA) and Canada (47% and 46%, respectively). However, in the case of Japan, they found most of the variations in output for short time horizons were due to supply shocks.

Centoni and Cubadda (2003) examine the relative importance of permanent and transitory shocks on the US business cycles by modelling per capita gross domestic product (GDP), per capita investment and per capita consumption in a multivariate framework over the period January 1974 through April 2001. They find that demand shocks explain the bulk of the variations in GDP (82%) and investment (86%), but supply shocks explain the bulk of the variations (57%) in consumption. Hartley and Walsh (2003) find that demand shocks explain the bulk of the variations in output for Germany, France, The Netherlands, and the USA--around 70-80%. Further, they find an even greater role for demand shocks in explaining variations in output for Italy and the UK--over 90%. In another study, for Italy from 1974-1994, Gavosto and Pellegrini (1999) find supply shocks are important in explaining output variability at all frequencies over the 20 years. For instance, after one quarter, supply shocks explain over 70% of the variation in output and, even after 20 years, the incidence of supply shocks still explains around 75% of the variation in output.

The above-mentioned studies differ in their approach to modelling the relative importance of permanent and transitory shocks in explaining variations in macroeconomic aggregates. For instance, many of the studies already mentioned use vector autoregressive (VAR) models, imposing long run restrictions in many cases. In a departure from the extant literature, Hartley and Walsh (2003) and Hartley and Walsh (1992) use a method of moments procedure to estimate the parameters of a structural model of output variation.

In this paper, our main innovation is our extension of the work on the role of permanent and transitory shocks in explaining variations in income by considering the per capita income levels of the G7 countries with an application of the common trend and common cycle methodologies. The joint use of common trend and common cycle restrictions is important for two reasons. First, a correct imposition of the common cycle restrictions provides more accurate estimates from a dynamic model (in our case, the VAR model), leading to a more accurate measurement of the relative importance of permanent and transitory shocks. (1) The second reason relates to the issue of different time horizons when measuring the relative importance of shocks. While the relative importance of permanent and transitory shocks should not differ greatly for long time horizons, they do differ in short time horizons because the short-run dynamics are imposed only by one of them (Issler and Vahid 2001). (2) In light of this, the aim of this paper is threefold:

1. To examine whether per capita GDP for the G7 countries shares long-term (common trends) and short-term (common cycles) features,

2. To examine the importance of permanent and transitory shocks on per capita GDP for the G7 countries, and

3. To examine whether the transitory components or the permanent components are important in explaining the cyclical behaviour in per capita GDP for the G7 countries.

The...