The sectoral analysis of business cycles: the role of aggregate and disaggregate shocks
This dissertation is a contribution to the analysis of disaggregate and aggregate business cycles. The motivation for the study comes from the need to assess whether shocks which initially affect specific sectors can induce aggregate output fluctuations. This can occur not only because aggregate output is the sum of sectoral output, but also because of trade linkage across sectors. Two main purposes of this dissertation are to investigate empirically the causal direction between aggregate and disaggregate output and to examine the relative importance of aggregate and disaggregate shocks. The study utilizes monthly industrial production indices for twelve disaggregated industries in Korea;The theoretical basis of this study is a multi-sector business cycle model developed by Long and Plosser (1983). This model is used to motivate a trivariate Vector Autoregressive (VAR) model, called Sector-by-Sector model, to establish the causal link between sectoral and aggregate output fluctuations. Then the theory is used to motivate a restricted Multi-sector VAR model. The latter is estimated by Seeming Unrelated Regression (SUR) method. The innovations from the SUR estimation are used to establish the source of shocks to the economy. The errors are assumed to be due to aggregate, industry and sectoral shocks. The variances of these shocks and parameters which determine the impact of the shocks on sectoral growth rates are estimated by method of moments. In addition, factor analysis of the estimated innovations is employed to determine the number of common shocks in the economy;In the Sector-by-Sector analysis, each sectoral output in durable manufacturing industry has a strong causal link to other sectors in the economy. Sectoral output in mining industry has the strongest causal link to aggregate output. The Multi-sector analysis demonstrated that one common shock was sufficient to describe aggregate innovations in the Korean economy. Shocks to individual sectors generate almost two-thirds of the innovations in the growth rates for the Korean economy. Aggregate shocks are responsible for just over one-quarter of the innovations in long run growth rates. These findings are supportive of the "weak" version of real business cycle theory.