for an annual GEP, NPP, or NRP. Households who had not visited any federal recreation lands in the
past two years were also screened out of the analysis, on the assumption they would not be part of the
relevant market either. After screening, the RDD sample provided 529 cases for analysis and the NPF
sample provided 1491 cases. Further details regarding survey design, sampling, and weighting can be
found in Taylor et al. (2006).
In general, the survey questionnaire can be broken into five stages:
1) Awareness of NPP or GEP: The interview began with an introduction, some screening questions,
and a short series of items about the household’s recent recreational experiences on federal lands.
The interviewer then asked whether the person was aware of the NPP or GEP. This question
grounded the person in an outside option that is similar to the NRP. If YES, the person moved to
Stage 2 (described below). If NO, the person moved to Stage 3.
2) Status Quo: The interviewer then asked whether the household has purchased a NPP or GEP.
The person then moved to Stage 3.
3) Purchase of the NRP: After the interviewer provided a short description of the NRP (which has
approximately the same attributes as the GEP; see footnote 2), we asked if the household would
be willing to buy - binary choice, YES or NO - the NRP at a randomly selected bid value. For
some respondents, the random bid value was the current GEP price. For these cases, the question
allows us to compare hypothetical purchasing decisions to the real purchasing decisions in Stage
2. The experiment then continues to stage 4.
4) Follow-up Valuation Question: To more precisely pinpoint the valuation distribution, the
interviewer then asks a follow-up valuation question. If a YES in Stage 3, we increase the price
and ask YES or NO again; if a NO in Stage 3, we lower the price and ask again.