Abstract
New Zealand experienced two natural experiments with respect to state-provided social housing after 1990. First, while continuing to acquire new state houses, the National Government substantially reduced the overall state house stock by selling a greater number of houses either to existing tenants (through the Home Buy scheme) or, if the house was vacant, to other purchasers (vacant sales). From 1999, the Labour-led government ended home buys, greatly reduced vacant sales and increased acquisitions, resulting in a major increase in the state house stock. We examine determinants of the spatial distribution of home buys, vacant sales and acquisitions over the period 1991-2006, focusing on levels of, and changes in, local deprivation status and house prices as determinants. Having modelled the determinants of each category, we test whether home buys, vacant sales, and acquisitions in an area over one five-year period had an effect on changes in local deprivation and house prices in the succeeding five-year period, after controlling for initial levels of, and prior changes in, deprivation and house prices. We find that state house acquisitions in an area led to a subsequent rise in local deprivation, consistent with the policy aim of providing housing to those most in need. While vacant sales had no material effects, a greater number of home buys in an area led to increased local real house price appreciation over the subsequent five year period. This finding, based on the results of a politically-driven natural experiment, is consistent with the hypothesis that a scheme that transforms existing tenants into homeowners (at the same location) improves community outcomes for the surrounding neighbourhood.