Developing Country Business Cycles: Characterizing the Cycle and Investigating the Output Persistence Problem

Identifying business cycle stylised facts is essential as these often form the basis for the construction and validation of theoretical business cycle models. Furthermore, understanding the cyclical patterns in economic activity, and their causes, is important to the decisions of both policymakers and market participants. This is of particular concern in developing countries where, in the absence of full risk sharing mechanisms, the economic and social costs of swings in the business cycle are very high. Previous analyses of developing country stylised facts have tended to feature only small samples, for example the seminal paper by Agénor et al. (2000) considers just twelve middle-income economies. Consequently, the results are subjective and dependent on the chosen countries. Motivated by the importance of these business cycle statistics and the lack of consistency amongst existing research, this thesis makes an important contribution to the literature by extending and generalising the developing country stylised facts; examining both classical and growth cycles for a sample of thirty-two developing countries.

One significant finding that emerges is the persistence of output fluctuations in developing countries and the strong positive relationship between the magnitude of this persistence and the level of economic development. The observation of procyclical real wages and significant price persistence indicates the suitability of a New Keynesian dynamic general equilibrium model with sticky prices, to explore this relationship; thus, the vertical production chain model of Huang and Liu (2001) was implemented. This model lends itself to such an analysis, as by altering the number of production stages (N) it is possible to represent economies at different levels of development. There was found to be a strong significant positive relationship between the magnitude of output persistence generated by the model and economic development. However, a very significant finding of this analysis is that the model overestimates output persistence in high inflation countries and underestimates output persistence in low inflation countries. This has important implications not only for this model, but also for any economist attempting to construct a business cycle model capable of replicating the observed patterns of output persistence.