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

Abstract

We use unobserved components methodology to establish a New Zealand common cycle from economic activity data for 14 regions, and to assess the extent to which the region‐specific cycles are additionally important. We then aggregate the 14 regions to 5 regions, and estimate a similar common cycle. At this level of aggregation, we can assess the statistical significance and relative strengths of influence on the common cycle of monetary and fiscal policy variables and several external shock variables. Our results show that structural breaks associated with New Zealand’s major economic policy reforms of the mid‐1980s through to the early 1990s play an important role, and that New Zealand’s region‐specific growth cycles have exhibited considerable diversity. The variance contributions of region‐specific cycles dominate common‐cycle contributions, a result consistent with multivariate findings for Australasia, but contrary to evidence for the US and for Australian states. We also establish that during key periods, terms of trade and net immigration variables have had distinctive procyclical influences on the common cycle, that real government expenditure has had a modest crowding‐out role, and that monetary policy has had no additional significant influence.

A Multi-Regional Computable General Equilibrium Model for New Zealand

“…50 Higher levels of regional detail would be possible if regional employment data were used to further split national industry output, but either some industry detail would have to be sacrificed or the model streamlined to reduce the rate that its size rises with the number of regions since, at the current level of disaggregation, the model with the extensions of chapters 3 and 4 approaches the limits of what a standard desktop computer can handle without special regard given to memory management. Hall & McDermott (2007) observed a broad dichotomy in the behaviour of urban vs. rural regions over the business cycle and in later work (Hall & McDermott 2011) found it useful to aggregate their original 14 regions of analysis to the equivalent of our 5 regions for the purpose of considering the influence of fiscal and monetary policy, and external shocks, on the regional cycles. The current level of regional detail therefore appears sufficient in that context.…”