Author: Kevin Atkinson; Department of Economics, Appalachian State University, Boone, NC 28608
Economists prefer revealed preference data, yet some situations lack sufficient revealed preference information for economic analysis. Stated preference data, acquired from surveys asking respondents about their behavior under hypothetical scenarios, may be useful in such situations. Stated preference data is often biased, but revealed preference data also has limitations. Combining both types of data may be especially useful in many situations because it grounds the results from stated preference surveys in the reality of revealed preferences while using data that extends beyond what can be observed from the past. The purpose of this research is to investigate the predictive validity of stated preference data, a current topic of debate among economists (Hausman 2012, Haab et al. 2013). This project will inform that debate through a survey of mountain bike park recreation participants about proposed trail development scenarios and then collecting data to determine revealed preferences after the proposed scenarios become reality. The Rocky Knob Trails Survey was conducted during 2011 and 2012, garnering 302 nearly complete responses. During these years the average number of annual trips to Rocky Knob reported by survey respondents was 16. The trails were not yet completed, so we asked respondents how many trips they would take during a typical year after completion of the trails. The average number of annual trips to Rocky Knob during a typical year with 6 and 8 miles of trail reported by survey respondents is 24 and 60. One half of the survey respondents agreed to be interviewed after the trails were completed. To date, 99 have responded to a follow-up survey begun in November 2013. We asked respondents for the number of mountain bike trips they had taken to Rocky Knob during the past 12 months. No results are yet available since data collection is ongoing (data will be available in January). Regression models will be estimated to determine if stated preference trips accurately predict revealed preference trips. The dependent variable will be revealed preference trips taken as reported in the November 2013 survey. Independent variables will include stated preference trips reported in the previous survey, time between surveys, socioeconomic and other variables. Recreation demand models will then be estimated using both the revealed and stated preference data to determine if the stated preference data can be calibrated to predict accurately. Consumer surplus estimates from the demand models will be estimated.
If you happen to be going (and why wouldn't you not attend an undergraduate research conference on a Saturday morning?), be sure to tell him his presentation was awesome.