Multiple Destination Trip Bias in Recreation Benefit Estimation

Abstract

Balanced and multiple use of the public lands has become the theme of much of federal land use policy.1 As additional and alternative uses of the public lands are proposed, the measurement of the value of publicly managed recreation facilities is necessary in order to make rational comparisons. For the most part these alternative uses have quantifiable benefits and costs determined through market operations. As a matter of social policy, however, access to many publicly managed recreation facilities is often provided at prices less than the market prices that would otherwise prevail. Since prices determined in a market economy usually provide the basis for calculating the benefits and costs of various commodities or activities, situations involving commodities or activities acquired without the use of market exchange pose certain difficulties for value estimation. It is necessary in these cases to derive visitor willingness-to-pay from observed nonmarket data or from direct consumer responses to questions about recreation values. One widely accepted technique for estimating the willingness-to-pay for a recreation site in cost-benefit studies is the Clawson-Knetsch travel cost method.2 Recently published regulations implementing revised Principles and Standards for Planning Water and Related Land Resources will result in increased application of this technique in federal planning. The advantages and disadvantages of using the travel cost method for estimating recreation benefits have been widely discussed (e.g., Pearce [1978]). The travel cost method assumes, among other things, that all travel costs are incurred exclusively to obtain access to the single specific recreation site being valued and thus differences in travel cost may be treated as a proxy for price. This study (1) examines the consequences of multiple destination trips on travel cost estimates of benefits, (2) suggests a means of correcting the bias that arises under such circumstances, and (3) obtains nearly identical value estimates from the travel cost method and from a direct question on willingness-to-pay. The magnitude of multiple destination bias is