Authors: Norman Gemmell , Arthur Grimes
This paper takes advantage of a quasi-natural experiment in local property tax reform that arose from the amalgamation of several local councils in 2010 in Auckland, New Zealand, to form a unitary local authority. The reform involved several tax changes including a shift in the base of the local tax (known as ‘Rates’ in New Zealand) from a land-value, to a capital-value, base; changes in the relative levels of Rates across the former councils; and changes in the level of a separate tax (Development Contributions) levied specifically on new and altered buildings.
These reforms provide opportunities to examine empirical support for a number of established hypotheses in the local property tax literature related to the level and structure of local taxation. Empirically, the exogenous nature of the New Zealand reforms enables more reliable estimates than hitherto of hypothesized effects of the tax changes on new property development arising from the tax switch (land to capital values), and changes in relative levels of both Rates and Development Contributions.
To test these hypotheses, we use difference-in-difference type regression analysis to examine how far observed changes in consents for new building development are consistent with predictions from our economic models, having controlled for a variety of other influences.
Our results suggest that there is little evidence of tax effects on new building development after the amalgamation, but there is stronger support for such effects on building alterations. Since our dataset covers only two post-amalgamation years, we speculate that this apparent difference may arise from the greater flexibility of building alterations to respond in the shortrun, compared with new development responses.