Abstract
The proportion of New Zealand households living in owner-occupied dwellings has declined steadily since the early 1990s. The unemployment rate declined steadily as well, except for upward shifts due to the late 1990s Asian Financial Crisis and the Global Financial Crisis a decade later. Research initiated by Andrew Oswald in the 1990s posits that declining homeownership and declining unemployment are linked and that the causality runs from high homeownership leading to high unemployment. The international empirical evidence for this hypothesis is rather mixed. In this paper, we revisit the issue with New Zealand census data for commuting-defined labour market areas from 1986 until 2013. Allowing for spatial spillovers in our data, we apply a general nesting spatial-econometric model. We also consider the potentially different impacts of freehold and mortgaged homeownership. Generally, the evidence that a declining homeownership rate contributes to a lower unemployment rate is statistically fragile. When statistically significant, an increase of 1% in freehold ownership appears to increase a labour market area’s unemployment rate by between 0.07 and 0.2% points. The effect is relatively stronger in those labour market areas where freehold ownership is still relatively low.
