Language-related differences in environmental benefits estimation: Evidence from a mail survey

In contingent valuation studies, failing to accommodate populations with limited language skills might yield biased estimates. In the United States, there are many residents primarily fluent in Spanish. This study uses conditional logit models applied to data from a bilingual (English and Spanish) conjoint choice mail survey to evaluate the effects of language proficiency on estimates of the economic benefits of contaminated site cleanup. Results indicate that language does have significant effects on welfare estimates. The results suggest that mail surveys addressing environmental issues that may affect a linguistically diverse population should be designed at the outset with multiple languages in mind.