Language-Related Differences in a Contingent Valuation Study: English Versus Spanish

To test the similarity of English- and Spanish-speaking households responses to a contingent valuation survey, phone interviews were conducted in both languages regarding two forest fire prevention programs. While there were similar response rates, there were significant differences in the most frequent reasons given for refusing to pay. In the pooled logit model, the language intercept and bid interaction variables were insignificant in both programs. The likelihood ratio test of separate logit equations showed no statistical difference between English- and Spanish-speaking households responses to either program. Mean benefits reported by Spanish-speaking households were about one-third lower than English-speaking households, although the difference is not statistically significant.