Benefit transfer

Benefit transfer

Shortened version of: Navrud, Stale (2005): Section 3.4 "Benefit Transfer" in: European Commission: ExternE - Externalities of Energy - Methodology 2005 Update. EUR 21951 EN. Office for Official Publications of the European Communities, Luxembourg, 2005. ISBN 92-79-00423-9. pp. 23-29.

Definition

Decision-makers often need economic analyses to support decisions among different policy alternatives. When relevant economic values are unavailable and cannot be provided by dedicated valuation studies due to limited resources, then economic measures estimated in similar contexts and sites can provide a proxy for the estimates necessary for decision-making.

According to Rosenberger and Loomis (2001), benefit transfer is defined as the adaptation and use of existing economic information derived for specific sites under certain resource and policy conditions to new contexts or sites with similar resources and conditions. Brouwer (1998) defines it as a technique where the results of monetary (environmental or health) valuation studies, estimated through market based or non-market based economic valuation techniques, are applied to a new policy context. Some authors (e.g. Navrud, 2004) prefer the term 'value transfer', since in many cases damage estimates can also be transferred from previous studies (termed "study-sites") to new evaluation contexts ("policy-sites").

Unit value transfer - simple unit transfer

• To be used: If the well-being experienced by an average individual at the original study-site will be equivalent to the well-being experienced by the average individual in the policy-site. -> Direct transfer of the economic benefit or damage from the study-site to the policy-site.
• Not to be used: For transfer between countries with different income levels and costs of living.
• Method: Simplest method of transferring economic estimates from one site or context to another, based on using an estimate from a single relevant study-site or a range of point estimates if more than one study is considered relevant.

Unit value transfer - unit value transfer with income adjustment

• To be used: If the income in the policy-site differs from that of the study-site, e.g. for transferring values from studies in developed countries to developing countries.
• Method: Benefit value in the policy-site can be estimated by adjusting the benefit value in the study-site(s) by the ratio between income levels in both sites and the income elasticity of demand for the environmental good.

Where (Bp) is the adjusted policy-site benefit; (Bs) is the original benefit estimate in the study-site; (Yp) and (Ys) are the income levels; and (ß) is the income elasticity of demand for the analysed environmental good.

• To be considered: Benefit is sensitive to elasticity. Using an income elasticity equal to one would change the willingness-to-pay measure in the policy-site proportionally to the relative per capita income differential across the two areas of study, whilst income elasticity equal to zero would mean that no adjustment is considered for income differentials (Davis et al., 1999).

Function transfer - benefit function transfer

• Method: Benefit-function transfer involves the use of a willingness-to-pay function, derived in a study-site preferably using stated or revealed preference techniques, which relates willingness to pay to a set of characteristics of the study-site population and the environmental good. That is, benefit function transfers use a model that statistically relates benefit measures with study factors such as characteristics of the user population and the resource being evaluated. The transfer process involves adapting the benefit function to the characteristics and conditions of the policy-site, forecasting a benefit measure based on this adaptation of the function, and use of the forecast measure for policy analysis (Rosenberg and Loomis, 2001).
• Advantage to unit value transfer: In contrast to unit value transfer more information can be taken into account in the transfer process.
• Disadvantages: Information on data collection and model specification in the original study must be available. Same data must be available for the policy-site. Relevant variables for the policy-site might be excluded by the benefit function used for the study-site.

Function transfer - meta-analysis

• Usage: Using the results of many valuation studies, developed in different study-sites, for estimating a single benefit transfer function.
• Method: Meta-analysis is defined as the statistical summary of relationships between benefit estimates and quantifiable characteristics of studies. In meta-analyses, several studies are analysed as a group and each result of these studies is one observation in a regression analysis. The data for a meta-analysis are typically summary statistics from study-site reports and include quantified characteristics of the population, the study site's environmental resources, and the valuation methodology used.

Validity and reliability

Several factors were identified that can affect the reliability and validity of benefit transfers. Rosenberger and Loomis (2001) summarised these factors:

• One group of factors that affects the validity of benefit transfers includes number, quality and documentation of the available studies:
  • The quality of the original study greatly affects the quality of the benefit transfer process;
  • The limited number of studies investigating a specific environmental good restricts the pool of estimates and studies from which to draw information;
  • The documentation of data collected and reported can be a limitation.
• A second group of factors is related to methodological issues. For example:
  • Different research methods may have been used across study-sites, including what questions were asked, how they were asked, what was affected by the management or policy action, how the environmental impacts were measured, and how these impacts affect recreation use;
  • Different statistical methods for estimating models can lead to large differences in values estimated. This also includes issues such as the overall impact of model misspecification and choice of functional form;
  • There are different types of values that may have been measured in primary research, including use values and/or passive- or non-use values.
• A third group of factors concerns the correspondence between the study site and the policy site, which arises because
  • Some of the existing studies may be based on valuing activities at unique sites and under unique conditions;
  • Characteristics of the study-site and the policy-site may be substantially different, leading to quite distinct values. This can include differences in quality changes, site quality, and site location.
• A fourth factor is the issue of temporality or stability of data over time. If the existing studies occurred at different points in time, relevant differences between then and now may not be identifiable nor measurable based on the available data.
• A fifth factor is the spatial dimension between the study-site and the policy-site. This includes the extent of the implied market, both for the extent and comparability of the affected populations and the resources impacted between the study-site and the policy-site.

Conditions and limitations

Rosenberger and Loomis (2001) argue that some general conditions should be met to perform benefit transfers.

• The policy context should be carefully defined, identifying:
  • The extent, magnitude, and quantification of expected impacts from the proposed action;
  • The population that will be affected by the expected impacts;
  • The data needs, including the type of measure (unit, average, marginal value) and the degree of certainty surrounding the transferred data.
• The study-site data should also meet certain conditions:
  • Studies transferred must be based on adequate data, valid economic method, and correct empirical technique;
  • Contain information on the statistical relationship between benefits and socio-economic characteristics of the affected population;
  • Contain information on the statistical relationship between the benefits and physical/ environmental characteristics of the study site;
• The correspondence between the study-site and the policy-site should have the characteristics:
  • The environmental resource and the change in the quality or quantity of the resource at the study-site and the resource and expected change at the policy-site should be similar;
  • The markets for the study-site and the policy-site are similar, unless there is enough useable information provided by the study on own and substitute prices - other characteristics should be considered, including similarity of demographic profiles between the two populations and their cultural aspects

References

• Rosenberger, R.S. and Loomis, J.B. (2001). "Benefit transfer of outdoor recreation use values: A technical document supporting the Forest Service Strategic Plan (2000 revision)". U.S. Department of Agriculture, Forest Service. http://www.evri.ca/english/default.htm
• Brouwer, R. (1998), "Future Research Priorities for Valid and Reliable Environmental Value Transfer", CSERGE Working Paper GEC 98-28: London.
• Navrud S. (2004), "Value transfer and environmental policy" in T. Tietemberg and H. Folmer (eds), The International Yearbook of Environmental and Resource Economics 2004/2005: a survey of current issues. London: Edgar Elgar Publishers.
• Navrud, Stale (2005): Section 3.4 "Benefit Transfer" in: European Commission: ExternE - Externalities of Energy - Methodology 2005 Update. EUR 21951 EN. Office for Official Publications of the European Communities, Luxembourg, 2005. ISBN 92-79-00423-9. pp. 23-29.