Abstract
The article presents the results of recreation demand analysis for the Šumava National Park, and focuses on a sensitivity analysis to the specification of the shadow price of recreation (travel cost). In the sensitivity analysis, I examine whether it matters to use calculated rather than theoretically correct perceived travel costs, and also discuss whether the travel costs stated by respondents are reliable.
Based on the results, the perceived costs directly stated by respondents may be considered unbiased and may be validly employed in the analysis of recreation demand and other travel-based models. The relative absolute error of respondents stating the travel cost is stable for close and far trips, and does not alter with trip frequency or socio-economic characteristics of the respondents.
To examine the effect of the specification on the recreation demand results, several recreation demand models are estimated, using truncated stratified negative binomial model. Regardless of the travel cost specification employed, the performance of the models is very similar.
On the other hand, the effects on estimated consumer surplus are apparent for calculated travel cost specifications based on less precise estimation of travel distance and time. When relying on the results of existing simplified studies in decision-making, it is therefore necessary to reduce the bias caused by simplification of the study; in the Czech case, it is sufficient to consider an interval estimate of consumer surplus.