Underground low-voltage electricity networks have several advantages over overhead networks including reliability of supply, safety and improved visual amenity. The economic viability of replacing existing overhead networks with new underground networks depends on the value of these benefits to households, but no complete value estimates are available. In this thesis, the focus is on benefit valuation, with two methods used to gather information regarding the benefits of undergrounding in a case study area - the hedonic price method, in which household preferences are revealed through purchases in the real estate market; and, the stated choice method, in which preferences are elicited using a survey. The empirical data were collected in Canberra, Australia.
Evidence from the stated choice survey suggests the value of benefits to households would be at least A$6,838 per household on average, with a 95 per cent confidence interval of A$5,444 to A$8,253. The demand function underlying this estimate is broadly consistent with evidence from the hedonic price study, which measures the benefits to a marginal purchaser in the real estate market. Some 31 per cent of households in the study areas chose to pay an estimated price premium of 2.9 per cent (between 0.4 and 5.3 per cent with 95 per cent confidence) to purchase a property serviced by underground networks (holding constant other property characteristics).
In addition to estimating values for policy use, the thesis focuses on one challenge associated with applying each method. The hedonic price method can be difficult to apply in cities where retrofit undergrounding has yet to take place, since the type of network infrastructure tends to be highly correlated with other property characteristics, such as building age and distance to the central business district. In this thesis, it is shown that this problem can be overcome by analysing data from areas with a mix of underground and overhead infrastructure where properties are otherwise relatively homogeneous.
Turning to the stated choice method, this thesis supports a growing body of literature finding violations of the assumptions typically made when analysing responses to a sequence of choice tasks; in particular, the assumptions of truthful, independent response and stable preferences. Preferences stated in the first of a sequence of binary choice tasks were not significantly different from those stated in an incentive compatible single binary choice task, but, in subsequent choice tasks, willingness to pay for undergrounding was lower and responses were influenced by cost levels observed in past questions.
An understanding of the decision processes (or heuristics) underlying this influence is required to ensure value estimates are not biased. The response patterns identified by novel econometric analysis in this thesis are consistent with certain types of heuristics involving strategic misrepresentation and/or value learning. The results suggest practitioners should test whether the standard assumptions are violated, and, if they are, account for the violations either by econometric techniques, such as the models presented in this thesis, or by simple sample selection approaches, such as focusing on responses to the first question presented to each respondent. Principal Supervisor: Jeff Bennett