Valuing the state of water in New Zealand using the experienced preference method

ABSTRACT

This article employs the experienced preference method to value the state of water in New Zealand. In doing so, this article represents a clear contribution to both the literature on non-market valuation and public policy regarding the preservation of water quality. The results show that a one-unit increase in satisfaction with the state of water bodies an individual has visited, is associated with a 1.6 per cent greater likelihood to report being very satisfied with their life. This positive link is robust to the estimation technique employed and a range of control variables. We find the value depends greatly on the functional form imposed on income. Specifically, the use of a cube root functional form yields an implicit willingness-to-pay (WTP) estimate of NZD 11,003, approximately half the size of the base model estimate. Uniquely, we couch this point estimate in the context of a broader distribution of implicit WTP estimates. It is hoped that these results will inform future applications and development of the experienced preference method and support public policy.

Acknowledgements

We are grateful to Statistics New Zealand for providing the data from the New Zealand General Social Survey. Statistics New Zealand publishes information for the benefit of the government, businesses, communities, and individuals. Although they take great care to ensure that the information is as correct and as accurate as possible, and updated regularly, they do not guarantee this and accept no legal liability whatsoever arising from, or connected to, the use of any material provided by Statistics New Zealand. We thank Nicholas Rohde, Jen-Je Su and other staff within the Department of Accounting, Finance and Economics at Griffith University for their helpful comments. We would also like to thank the anonymous reviewers who provided several helpful comments on an earlier version of the manuscript.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. All monetary estimates are reported in NZD. As at 6 April 2017: 1 NZD = 0.92 AUD; 0.70 USD; 0.65 EUR.

2. Individuals who report ‘hasn’t been to any of them’ or ‘don’t know or refused’ are removed from the regression.

3. All monetary estimates are reported in NZD. As at 1 August 2017: 1 NZD = 0.94 AUD; 0.75 USD; 0.64 EUR.

4. This transformation may potentially also capture collinear omitted variables. However, it is not clear what these omitted variables may be or how they might be controlled for. Nevertheless, we have included numerous controls in the model which could be expected to ameliorate possible omitted variable bias.

5. We use the OLS coefficient estimates here rather than the average marginal effects obtained from the ordered probit model for illustration, as the average marginal effects obtained from the ordered probit model depend on the specific values of income. The OLS coefficient estimates, in contrast, do not depend on the specific value of income. This facilitates the straightforward illustration of the infra-marginal changes in income levels. A complete set of OLS model estimates are available on request.