Preferences for the future of the Southport Spit: evidence from a choice experiment

ABSTRACT

The Southport Spit, an undeveloped open space at the northern end of the Gold Coast beaches, has been subject to contentious development proposals. This study employs a choice experiment to elicit residents' preferences for the future of the Spit in terms of development focus, maximum permitted height of development and extent of developed space. From an online choice experiment survey, we estimate a latent class model that reveals two strongly opposing viewpoints and a third, more nuanced, viewpoint. A pro-conservation segment opposes development of either a cruise ship terminal or casino, prefers low-rise development and opposes any increase in the extent of developed space. In contrast, a pro-development segment favours building a cruise ship terminal, a casino, or both a cruise ship terminal and a casino, prefers medium-rise development and has no clear preferences for how much of the available footprint should be developed. This study contributes to the small body of literature employing choice experiments to elicit the land-use preferences of urban communities; findings may help to explain why, to date, so little progress has been made towards reaching consensus on the future of the Spit.

KEYWORDS:

• Southport Spit
• choice experiment
• land-use planning
• environmental management
• urban policy

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 unless stated otherwise, all figures are in AUD

2 The IIA property states that the ratio of the choice probabilities of any pair of alternatives is independent of the presence or absence of any other alternatives in a choice set. This property is a condition underpinning the multinomial logit model (the usual starting point model for analysis of choice experiment data). When the IIA property does not hold, an alternative model must be employed – common alternatives include latent class and mixed logit models, we employ the former here.

3 Note that comparing parameter estimates across segments is not informative because of scale differences (Louviere, Hensher, and Swait 2000).

4 The open space preservation and recreational space charge was presented to respondents as an amount they would be charged on their rate bills. Therefore, other factors being equal, we would expect respondents to prefer options with lower charges – yielding a negative coefficient for the charge attribute in model estimation. This was the case for Segments 1 and 2 (see Table 3). Segment 3 had a positive coefficient for the charge attribute, suggesting this segment, all other factors being equal, prefers options with higher charges - this is not an expected result and thus not discussed further.

5 Note the negative coefficient of the charge even though alternative scenarios were presented with lower charge levels.