Co‐movement of Australian State Business Cycles*

Abstract

We use a variety of techniques to examine the nature and degree of co-movement among Australian state business cycles. Consistent with the results of Dixon and Shepherd (2001), we find that these cycles move quite closely together, with particularly strong links between the cycles of the larger states. We then seek to understand the causes of this co-movement by using an unobserved components model to distinguish between various shocks and their transmission. Our model implies that the major source of this co-movement in state activity is the commonality of shocks affecting state cycles. In contrast, spillovers of region-specific shocks appear to play only a minor role in creating co-movement, though region-specific shocks themselves have a moderate influence on cyclical fluctuations. These findings are consistent with the results of recent studies for the United States, Canada and Europe, where common shocks have also been found to dominate regional cyclical activity. JEL Classification Numbers: C22, E32, R11

Unobserved Components Business Cycles for New Zealand. What Are They, and What Might Drive Them?*

“…Kouparitsas (2002, fig. 1) has established a common cycle for the USA which has turning points that closely match those of the NBER Dating Committee, and Norman and Walker (2007, fn 19 and figure 4) present a weighted average common cycle for Australia that has a correlation of 0.79 with a HP‐filtered cycle for domestic final demand. Here, we provide evidence in a New Zealand context that UC estimation can be used to derive credible common growth cycles.…”

“…For the USA, Kouparitsas (2002, p. 30) finds that its BEA regions are largely driven by common sources of disturbance and that they have similar responses to a common shock. In a relatively similar vein, Norman and Walker (2007, pp. 360, 373) conclude for six Australian states that the major source of fluctuations in states’ economic activity is shocks which are common to all states.…”

“…The particular model we use is the dynamic multiple indicator multiple causes model. This model was used by Kouparitsas (2001, 2002) to study regional business cycles in the USA, by Norman and Walker (2007) to study state business cycles in Australia, and by Hall and McDermott (2008) to assess implications for an Australasian common currency by determining whether there are asymmetric shocks across regions of Australasia.…”

“…Multivariate methods are more advanced for examining growth cycles than are classical cycle methods, despite the well‐known deficiencies associated with trend removal. For example, using unobserved components (UC) methodology, Kouparitsas (2002) finds that the eight US Bureau of Economic Analysis (BEA) regions are largely subject to common sources of disturbance, and Norman and Walker (2007) conclude for six Australian states that the major source of the state fluctuations is shocks which are common to all states. But Norman and Walker’s variance analysis also shows that each overall state cycle is driven partly by fluctuations specific to that state, in particular for Western Australia, but also for New South Wales and Tasmania.…”

Comovement of Chinese provincial business cycles

“…Regional business cycle analysis started in the mid 1990s for the US (Carlino and Sill, 1997) and was later applied to European countries (Spain, Barrios and Lucio, 2003; and the UK, , as well as Australia (Norman and Walker (2007)) and Japan (Wall, 2006). More recently, Owyang et al (2005), Kouparitsas (2002) and Crone (2005) provide deeper analysis of business cycle characteristics and co-movements for US states.…”