Abstract:
Choosing where to live is one of the most important questions that individuals will consider during their lifetimes. Understanding the answer to this deceptively simple question can offer insights into how urban cities can improve their transportation infrastructure, housing construction and spatial planning. In urban economics, the conventional theory of residential location choices is a utility optimisation decision for single worker households choosing a residential location. Provided that all commuting costs and prices for land and rents depend on location, residential location decisions are thus subject to these constraints. However, from the von Thünen (1826) model on urban land use to Alonso (1964), Mills (1967) and Muth’s (1969) bid-rent theory, most literature on residential location choices largely neglects the differences between residential tenure type and household composition and does not capture the uncertainty that households face during the decision process of choosing residential location.
This thesis uses two-level data, including aggregate and individual census data from 2001 to 2018 for Auckland, New Zealand, to address three questions related to residential location choices: 1) What role does home purchase affordability play in commuting patterns and spatial distribution across workers? 2) Do occupational differences and multiple workers in the household matter in residential location choices? and 3) To what extent do suboptimal residential location choices lead to a spatial mismatch amongst key worker homeowners? Key workers are chosen as the focus of this thesis because this group of low- to moderate- class professionals, who provide essential services for the community, usually is not qualified for subsidised homeownership but are not able to afford to access homeownership near their job locations at the market price in cities where housing is severely unaffordable. According to the 2020 housing affordability survey (Blain & Holle, 2021), Auckland was ranked as the fourth least affordable market globally. The COVID-19 pandemic has further highlighted the dependence of cities and their populations on these workers.
The contribution of this thesis is twofold. First, this thesis extends Kain’s (1968) spatial mismatch hypothesis to consider the occupation of households. The aggregate census data analysis in this thesis confirms that workers’ sub-optimal residential location choices will cause a jobs-housing mismatch and lead to negative externalities that create additional commuting costs. The results indicate that for key workers, a one per cent increase in housing unaffordability (i.e., mortgage repayment relative to the annual income) results in an extra two-kilometre commuting distance – the equivalent to a $90 million deadweight loss a year. Second, this thesis also fills the knowledge gap by examining households’ residential rental location choices with multiple potential breadwinners under uncertainty. By extending Crane’s (1996) residential locational choice model, this thesis develops a two-worker, two-period, two-centre (2W×2P×2C) model to demonstrate the optimal residential rental location choices in multiple worker households. Using the Integrated Data Infrastructure (IDI) from Statistics New Zealand, empirical tests of the model suggest that multiple-worker households are less inclined to pay a rental premium to live close to the city centre than are single-worker households. To further understand how the characteristics of individual wage earners affect their residential choices under uncertainty, the rigidity of their job locations and work hours are analysed. The individual-level analysis confirmed the theoretical prediction that key workers with higher certainty job locations are less willing to pay higher rent to live close to the city centre. In contrast, workers taking public transport or working long hours prefer to pay a premium to reside in city centres, thus enabling shorter commutes and more job opportunities.
Residential location choice is a crucial driving force in urban dynamics. The evidence from key workers in one of the least affordable cities in the world articulates that to better analyse the choice of residential locations, homeownership affordability (versus renter affordability), job locations uncertainty, and commuting uncertainty are pivotal in the theory of residential location choices. In addition, the traditional assumption of single-worker households also prevents us from understanding the big picture of residential mobility in an urban city. This thesis gives prominence to the imminent need to refine the existing residential location choice models in the literature. The estimation of social costs associated with the housing-induced spatial mismatch in this thesis also pinpoints that the cost of homeownership could be exorbitant for an urban city, reminding us that when solving “The Problem of Social Costs” (Coase, 1960), one must analyse the costs of the action involved.