Nonparametric techniques are frequently applied in recreation demand studies when researchers are concerned that parametric utility specifications impart bias upon welfare estimates. A goal of this paper is to extend previous work on nonparametric bounds for welfare measures to allow for measurement errors in travel costs. Haab and McConnell (2002) state that issues in travel time valuation continue to be topical in the recreational demand literature. This paper introduces a bootstrap augmented nonparametric procedure to precisely bound welfare when price data contains measurement error. The technique can be extended and becomes more convenient relative to other approaches when more than two site visits are made by a single recreationist. These techniques are demonstrated in a Monte Carlo experiment.
