NSW Department of Environment and Climate Change Envalue Study Database

Rationale for Benefit Transfer
The valuation of environmental qualities is fundamental to the efficient allocation of resources in accordance with the preferences of individuals and society. The importance of establishing values for such goods as clean air and water is widely acknowledged however environmental valuation studies are expensive to undertake. Benefit Transfer can reduce both the cost of research and the time required assigning values to environmental qualities of a specific site. This means that economic analysis can be carried out in a more timely and efficient manner. The term 'benefit transfer' was first used by Desvousges, Naughton & Parsons (1992) to describe the transfer of monetary valuations determined by research as applicable to the site studied, to a different site. The site where an existing study was conducted is referred to as the 'study site', while the site that proposes to make use of the transferred values is referred to as the 'policy site'. Economic analysis is routinely carried out in government agencies to determine the impacts of proposed policies and government legislation. In addition to the application of benefit transfer to public decision making and academic research, legislative requirements for the preparation of Environmental Impact Statements have increased the need for improved valuation and pricing of environmental resources. The advantages of benefit transfer are apparent but come with the caution that results can be misleading unless the values are carefully applied. The quality of the original study site and its similarity or difference to the policy site is of great importance. Areas of difference can include population size, socioeconomic characteristics and the magnitude of the change under consideration. The method used to transfer benefits can also have an effect on the quality of the results obtained for the policy site. This section describes some of the important concepts underlying benefit transfer, and recommends certain procedures to be followed to ensure acceptable accuracy.

Approaches to Benefit Transfer
The three broad areas that must be assessed in applying primary studies to benefit transfer are the quality of the study, the extent to which conditions at the study site are similar to those at the policy site and the method used to transfer the values.

Quality of the Primary Study
The process of benefit transfer can do nothing to correct for fundamentally flawed estimates. Allowance can be made for known biases or for differences between the sites but it remains critical that the primary studies selected are of adequate quality and are assessed using a sound methodology. It is important that an indication of the level of confidence that can be placed on the estimates is determined. The approach taken in the development of 'Envalue' has been to define a set of criteria for evaluation of primary studies. An assessment of the quality and applications possible for a particular study can then be made on the basis of the information provided in the database. Where underlying bias or variability in the final estimates is suspected, the criteria provide some guide to the direction (and in some cases the magnitude) of the possible bias. As well as providing an indication of the quality of individual studies, 'Envalue' offers the opportunity to compare studies on related environmental goods to establish trends and the nature of any relationships. A study that lies outside the general range of estimates or is inconsistent with an established trend may be indicative of flaws in the study design.

Criteria for Benefit Transfer
Desvousges, Naugton & Parsons (1992) proposed five criteria against which the accuracy of benefit transfer could be assessed. Based on their proposal, a smaller and simpler set of criteria has been adopted. A. The study site and policy sites are similar B. The environmental change under consideration at the policy site is similar to the proposed change at the study site. C. The socioeconomic characteristics of populations and other site details are similar. Depending on the extent to which these criteria are satisfied and the degree of accuracy needed, there is some choice as to the level of sophistication to be adopted for benefit transfer. The amount of data required and its level of aggregation is important to the cost and accuracy of the final transfer method.

Direct Transfer
Transferring Aggregate Benefit Estimates
Some studies report estimates for an entire region, state or the nation. Such direct estimates are of little use in terms of rigorous benefit transfer. Consequently they have been omitted from 'Envalue' unless accompanied by sufficient definition of the methodology and other important factors in the estimation so that a workable understanding of the estimation process can be gained

Transferring Mean Benefit Estimates
Direct transfer of benefits on a per person or per household basis may be attempted in cases where the conditions between the study and policy sites are not expected to influence the valuation of benefits significantly. The total value of the benefit at the policy site is then obtained by multiplying the mean estimate by the size of the affected population. However there are some difficulties with the application of this method. The population in the policy area may not value the environmental good in question to the same extent as that of the study area. The population could be different in terms of demographic structure, income, education, religion, ethnic group or other socioeconomic characteristics. Even if the population characteristics are the same, the amount and type of substitutes for the environmental good may be different at the study site from those available at the policy site (OECD, 1994). Direct Transfer is the least rigorous method of benefit transfer and recent studies have cast doubts on its use even in cases where sites are seemingly similar, if a degree of precision is required (Kirchhoff, Colby and Lafrance, 1997). It is suggested that Direct Transfer methods may be of most use where an indication of direction is required for policy makers conducting a preliminary assessment.

Adjusted Transfer of Mean Benefit Estimates
Adjustment of benefits may become necessary for two broad reasons. Firstly biases may have been identified in a source study in which case adjustments can be made in the hope that the result will be a more realistic set of data. Alternatively an attempt may be made to adjust for different conditions present at the study site to those encountered at the policy site. Such differences can be found in the socioeconomic characteristics of the sites, in the policy or project proposed or in the availability of substitute goods and services available at the sites (OECD, 1992). The process requires adequate information on both the mean values of the important parameters at the study and policy sites, and the functional relationship between (mean values of) the important parameters (and the mean value) of the estimated benefit. This approach generally provides a higher degree of accuracy than direct transfer and is adequate in some cases provided the variation about the mean value of each of the parameters is relatively small.

Transferring Demand Functions
The most accurate level of benefit transfer is to transfer the demand function as a whole rather than as discrete points such as the mean value. The increased accuracy usually requires considerably more data and greater complexity in the benefit transfer computations. The demand curve is transferred in functional form with coefficient values estimated in the primary study. An estimate of the total benefit at the policy site is then obtained by first substituting the values of the independent variables in the demand function that remain constant at the policy site (such as the socioeconomic characteristics of the population), then integrating the demand function over the range of values of the remaining independent variables (such as the environmental quality variables). Similarly, for benefits valued through the dose-response approach, transfer of dose-response relationships is the preferred approach. This approach allows more refined manipulation of the results at the policy site, for example to take account of substitution effects, such as alternative recreation sites in a travel-cost model. It is important that the assumptions underlying unit value and usage are consistent because total benefits depend upon both the benefit per unit and the level of use. Transferring the demand function ensures that the estimate of total benefits is based on estimates of use and value derived from the same data set. If direct transfer of benefits is carried out, it is easy for the underlying assumptions and data for value and use to be different (OECD, 1992). The distribution of population around the policy site is another important consideration in the transfer of values. Even when two sites share the same average unit values, the implementation of the benefit transfer approach based on unit values may yield a different estimate of total benefits than an approach based on transferring the entire demand function. The transfer of the demand function can take account of the socioeconomic characteristics of the population and its distribution. (OECD, 1992)