Abstract
Housing prices in Auckland have persistently increased in the last decade. To fast-track development of housing, Special Housing Areas (SHAs) were created in September 2013 as a measure to improve affordability. It is not clear the extent of the success (or failure) of SHAs as they were disestablished by May 2017. This paper investigates the causal effects of the SHAs programme on housing prices and the implications on affordability. We used a dataset comprising more than 170 thousand sales transactions between 2011 and 2016 in Auckland. Our approach consists of a Difference-in-Difference approach where the treatment consists of tranches of land designated as SHAs. Our results indicate that the SHAs caused an average price increase of approximately 5% and did not contribute to increases in the likelihood of affordable transactions. These findings are robust across several specifications and question the effectiveness of the SHAs on improving affordability.
Disclosure statement
No potential conflict of interest was reported by the authors.
Notes
1 The authors also conducted (informal) interviews with planners and developers, which revealed that a lack of monitoring and enforcement, and the incentive for developers to first build market-rate houses and postpone (maybe indefinitely) the construction of affordable houses, likely contributed to the failure of the program.
2 In most cases the after-treatment period is in October 2013. For SHAs located in the New Lynn Area Units it is November 2013; for the Albany Area Units it is May 2014; and for Otahuhu Area Units it is June 2014. We code our after-treatment indicator accordingly.
3 The AUs are non-administrative geographic areas, roughly the same size as suburbs, within urban Auckland normally containing a population of 3,000–5,000 people (Statistics New Zealand 2016). These are defined for the purpose of taking a census and are also used to define public school zones and even to proxy for the local labour market. The introduction of AUs fixed effects also controls for time-invariant unobserved neighbourhood characteristics.