Mortality and fluctuations in macroeconomic activity in New Zealand

To model the statistical relationships between mortality rates and macroeconomic conditions in New Zealand over the post-war era.

Time series methods are used to model statistical relationships between the macroeconomic unemployment rate and mortality rates across different age groups and morbidities.

Decreases in the age-adjusted mortality rate are two-and-a-half times as large during episodic increases in the national unemployment rate (p = 0.01). This short-run procyclicality is driven primarily by mortality among the elderly (66+) and middle-aged (36-65). There is also strong evidence that assault and self-inflicted harm mortality rates are countercyclical and have a long-run relationship with the national unemployment rate (p < 0.001 and p < 0.001).

Mortality exhibits both short-run procyclical and long-run countercyclical features. Accelerated reductions in elderly mortality during periods of rising unemployment are consistent with overseas evidence that tight labour markets have negative impacts on aged care. However, further research is necessary to uncover the causal channels, if any, that underpin this correlation in New Zealand data. Long-run variation in mortality due to violence, such as assaults and suicide, is closely linked to employment conditions, indicating that policy responses to violence and suicide should not preclude addressing a potential lack of employment opportunities.