The World’s Largest Open Access Agricultural & Applied Economics Digital Library This document is discoverable and free to researchers across the globe due to the work of AgEcon Search. Help ensure our sustainability. Give to AgE con Search AgEcon Search http://ageconsearch.umn.edu [email protected] Papers downloaded from AgEcon Search may be used for non-commercial purposes and personal study only. No other use, including posting to another Internet site, is permitted without permission from the copyright owner (not AgEcon Search), or as allowed under the provisions of Fair Use, U.S. Copyright Act, Title 17 U.S.C. The Non-Market Value of Abel Tasman National Park, New Zealand: A Choice Modelling Application Peter Lee1, Sue Cassells2, John Holland1 1 Institute of Agriculture and Environment, Massey University 2 School of Economics and Finance, Massey University Contributed paper prepared for presentation at the 57th AARES Annual Conference, Sydney, New South Wales, 5th-8th February, 2013 © Copyright 2013 by Authors’ names. All rights reserved. Readers may make verbatim copies of this document for non- commercial purposes by any means, provided that this copyright notice appears on all such copies. ii The Non-Market Value of Abel Tasman National Park, New Zealand: A Choice Modelling Application Peter Lee1, Sue Cassells2, John Holland1 1 Institute of Agriculture and Environment, Massey University 2 School of Economics and Finance, Massey University Abstract National parks and protected areas form the basis of global conservation initiatives and provide a raft of benefits in the form of various consumptive and non-consumptive uses. However, it is extremely difficult to express these benefits in monetary terms. The lack of economic values for these protected areas often results in sub-optimal conservation outcomes. Non-market valuation techniques can be used to estimate monetary values for these key environmental assets. This research applied the choice modelling approach to assess the value of non-market goods and services associated with Abel Tasman National Park in New Zealand. A standard multinomial logit model was used to analyse visitor preferences and derive welfare measures. The results indicate park users were willing to pay an actual cash value for the ecological and recreational attributes of the park. These monetary values can be used to guide future development, inform resource allocation decisions and ensure adequate conservation financing. Keywords: Choice experiments, stated preference, willing to pay 1 1. Introduction National parks and protected area networks play an integral role in the conservation of biodiversity and the provision of other benefits associated with the maintenance of ecological integrity (Chape, et al., 2008; Stolton & Dudley, 2010). However, the economic benefits associated with national parks and protected areas have always been difficult to quantify in monetary terms. As an economic good, the provision of protected areas by the market is often subject to a number of market failures. These failures primarily stem from the fact that protected areas exhibit varying degrees of non-rivalry and non-excludability and generate positive consumptive externalities (Dixon & Sherman, 1990; Turner, 2002). As a consequence of these market failures, the benefits associated with protected areas are typically not subject to economic valuation by the price mechanism, often resulting in an implicit zero value. This leads to protected areas being undervalued and underfunded relative to other government fiscal and budgetary considerations (Dixon & Sherman, 1991). However, national parks and protected areas provide a range of benefits including education, recreation and tourism, ecosystem services and various non-use values (Putney, 2003; Stolton & Dudley, 2010). An expression of these benefits in dollar terms would help to ensure the efficient allocation of conservation resources. Specifically, monetisation of these benefits can be used to justify continued public investment in protected areas and provide an impetus for the continuation of conservation activities in the face of competing development interests (Dixon & Sherman, 1990). The increasing importance of non-market valuation for protected areas must also be seen in the global context of a declining funding base for conservation and increasing calls for the adoption of a consumer-oriented approach to protected area management. Within this context, non-market valuation can play an increasingly important role by providing protected area managers with information regarding visitor preferences, the goods and services provided by these areas and how these benefits are able to be captured (WCPA-IUCN, 1998, 2000). Currently, New Zealand’s conservation estate faces a number of challenges in terms of declining budgetary allocations, calls for increasing commercialisation and renewed pressure from development interests (Haque, 2006; Office of the Parliamentary Commissioner for the Environment, 2010). Consequently, the application of non-market valuation techniques to protected areas in New Zealand is of particular relevance given the urgent need to ensure the efficient allocation of conservation resources. 2 Despite this, only a handful of non-market valuation studies have been conducted in New Zealand’s national parks. These studies have typically utilised either the contingent valuation or travel cost methods to estimate recreational values for these critical elements of natural capital. One emerging stated preference technique which shows considerable merit in its application to non-market valuation is the choice modelling (CM) approach. This paper makes a contribution to the valuation literature by applying the CM technique to a New Zealand national park case study. Several studies have used the CM approach to derive economic values for the various attributes which characterise national parks and protected areas in other locations. Hearne and Salinas (2002) elicited tourist preferences for the provision of recreational infrastructure in Braulio Carrillo National Park in Costa Rica. In a similar context, Hearne and Santos (2005) analysed tourist preferences for the development of the Maya Biosphere in Guatemala. In another nature-based recreation study, Naidoo and Adamowicz (2005) assessed the benefits associated with biodiversity conservation in Uganda’s protected areas. Within the developed country context, Juutinen et al. (2011) elicited visitor preferences for the development of Oulanka National Park in Finland. The study combined both ecological and recreational attributes of the park and assessed the welfare impacts of alternative management options. This study most closely resembles the type of choice model which this research intends to apply within the New Zealand setting. The aim of this study is to determine the economic value of some of the non-market goods and services associated with Abel Tasman National Park (ATNP) in New Zealand. This paper is structured as follows. Section two will provide a brief description of the selected national park case study. Section three will provide an overview of the CM technique. Section four will detail the methodology employed in the development of the choice experiment. Section five will present the results of the choice model. The final section will discuss these findings and conclude the paper. 2. Case Study ATNP is the smallest national park in New Zealand and is located in the Tasman region of the South Island. The park covers approximately 230 km2 and falls under the definition of a category II protected area as developed by the International Union for Conservation of Nature (Department of Conservation, 2008). The park is administered by the New Zealand Department of Conservation (DOC) and was established in 1942 as a result of concerns regarding the loss of native flora and fauna in this area. Prior to this, the area had been 3 subject to a range of land use activities including agriculture, quarrying, mining and timber milling. Pre-European Maori also used the area for settlement and subsistence agriculture (Department of Conservation, 1997). Currently, the park is experiencing a period of ecological recovery and rehabilitation as a result of the cessation of previous land use activities. This is evident from the rapid changes in the structural composition of the park’s ecosystem, with regenerating bush giving way to more dominant forms of native vegetation. The geographical coverage of the park includes a diverse array of physical landforms which provide a habitat for a range of threatened and at risk native flora and fauna. However, these biological communities face ongoing threats from both the presence of invasive alien species and visitor induced pressures (Department of Conservation, 2008). ATNP is also highly regarded for its scenic values and recreational opportunities and has been used to promote New Zealand as a premier tourist destination. The key characteristics of the park include a rugged, forested interior, golden beaches and a pristine marine environment. The coastal track is the most popular walking track in New Zealand with 151,000 visitors annually. Visitor activity is heavily concentrated in the coastal region with 95% of visitor use being within 500m of the coastline. For several decades, there has been a perception among visitors that ATNP suffers from overcrowding, particularly along the coast during the summer months (Department of Conservation, 2008). A major constraint on DOC’s ability to develop ATNP is whether additional development would compromise the values which the national park is intended to protect. Further development has the potential to negatively impact on wilderness values and ecologically sensitive areas (Department of Conservation, 2008). Accordingly, management of the park is largely dictated by the need to reconcile development pressures with the intrinsic values enshrined in ATNP. This provides a unique context in which to undertake a non-market valuation study. 3. The Choice Modelling Approach The relatively recent development of CM has largely been in response to criticism directed towards the contingent valuation method and traditional conjoint analysis (Bennett & Blamey, 2001). The conceptual microeconomic foundations of the CM approach are based on Lancasterian consumer theory. The characteristics theory of value states that consumers derive utility from the characteristics or attributes of a good as opposed to the actual good 4 per se (Lancaster, 1966). Accordingly, the fundamental premise of the CM approach is that an environmental good can be decomposed into a number of attributes and associated levels. A payment vehicle attribute is usually included to facilitate the calculation of welfare estimates. Consumers are presented with a series of choice sets which consist of several hypothetical management alternatives characterised by different attribute levels. Respondents are required to select their most preferred management alternative in each choice set (Hanley & Barbier, 2009). In order to explain consumer choice within a utility maximising framework, CM has integrated the Lancasterian model of consumer behaviour with random utility theory (RUT) developed by McFadden (1974). RUT presupposes that an individual’s utility can be divided into an observable deterministic component ( ) and an unobservable random stochastic component ( ) (Holmes & Adamowicz, 2003). Assuming these two components are additive, a generalised utility expression for each alternative can be expressed by equation 1 (Boxall, Adamowicz, Swait, Williams, & Louviere, 1996). (1) The deterministic component of utility can be explained by the attributes included in the CM study. The random component is a result of the analyst having imperfect information regarding all the determinants of utility (Holmes & Adamowicz, 2003). This leads to the inclusion of an error term which is able to capture the effect of these unobserved influences (Louviere, 2001). The deterministic component of utility can be further decomposed and expressed as (Hensher, Rose, & Greene, 2005): (2) Where there are attributes and represents the parameter coefficient relating to attribute alternative and ASC is an alternative specific constant. The ASC captures the average influence of all unobserved factors on utility. This representative portion of utility is often assumed to be linear in attributes for computational ease but can also be represented in quadratic or logarithmic form. The coefficients show the relative importance of each attribute and their effect on utility (Hensher, et al., 2005). With regard to the error component, as the analyst has practically no information about the unobserved elements a number of maintained assumptions exist. Collectively, these 5 assumptions are referred to as the independently and identically distributed (IID) condition. The IID condition assumes that all error terms are derived from the same underlying distribution and are uncorrelated with other error terms (Hensher, et al., 2005). Applying RUT specifically to the choice model, each individual selects an alternative which maximises their utility. The inherent uncertainty caused by the random element ensures the analyst is restricted to modelling the probability of an individual choosing a particular alternative (Hensher, et al., 2005). The probability of an individual selecting alternative over alternative can be expressed as (Hensher, et al., 2005): Prob=Prob (3) i Where represents the entire choice set. The fact the error term cannot be measured, transforms a consumer’s standard utility maximisation rule to a random utility maximisation rule. Equation 3 can be rearranged to express this as (Hensher, et al., 2005): Prob=Prob (4) i Expression 4 states that the probability of an individual choosing alternative is equal to the probability that the difference in the unobserved sources of utility is less than or equal to the difference in the observed sources of utility (Hensher, et al., 2005). Assuming the error term exhibits an extreme value type 1 or Gumbel distribution, a standard multinomial logit (MNL) model can be used for CM purposes. The probability of a respondent selecting alternative is given in equation 5 (Hensher, et al., 2005). Prob= (5) i The probability of an individual selecting an alternative is modelled as a function of the key design attributes and respondent socioeconomic, demographic and attitudinal variables. Alternatives with higher levels of desirable attributes have a higher probability of being selected (Bennett & Adamowicz, 2001). One restriction or assumption embodied in the MNL model is the behavioural condition known as the independence of irrelevant alternatives (IIA). This condition states that the 6 probability of a respondent selecting an alternative is independent of the presence or absence of other alternatives in a choice set. A result of the IID assumption, the IIA condition implies that the unobserved attributes are identical for each alternative (Hensher, et al., 2005). The results of the choice model can be used to derive Hicksian consistent welfare estimates. The CM approach can provide two forms of welfare estimates. First, implicit prices for individual attributes can be obtained by estimating the marginal rate of substitution between the non-monetary and monetary attribute as shown in equation 6 (Hanley & Barbier, 2009): (6) Where IP is the implicit price, is the parameter coefficient for the non-market attribute and is the parameter coefficient for the monetary payment vehicle. Implicit prices should be interpreted as the marginal willingness to pay for an attribute ceteris paribus. Measures of willingness to pay can also be obtained for situations involving changes to multiple attributes. Compensating surplus can be calculated by multiplying the difference in utility between two states of the world with the negative of the monetary coefficient. (7) Where CS is the compensating surplus, is the utility associated with the alternative management option, is the utility associated with the status quo option and is the parameter coefficient for the monetary payment vehicle. 4. Methodology A choice model was developed to assess visitor preferences and derive non-market values for ATNP. The first major design stage involved defining the attributes and associated levels which characterise the environmental and recreational aspects of the park. A comprehensive literature review was carried out in order to compile an initial list of attributes. This list was presented to a focus group with previous recreational experience in New Zealand’s conservation estate. Participants were asked to indicate which attributes played an important role in determining consumer choice and were given an opportunity to add additional attributes. A refined list of attributes was then provided to DOC staff for comment and final approval. All the attributes identified in the focus group were considered to be relevant by park managers ensuring there was little divergence between public and policy perspectives. 7