摘要：One of the challenges in estimating economic impacts of recreational activity and natural resource use is collecting survey data when the underlying population is not perfectly known. We combine bootstrapping techniques with input output analysis to estimate impacts from recreational activity at a barrier island in Louisiana. Per capita expenditure point estimates are bootstrapped and the 1,000 bootstrapped samples are shocked into an input output model to develop confidence intervals of economic impacts between the two surveys. Results show that for many sectors of the economy, confidence intervals do not overlap between survey modes. These results suggest that researchers using multiple survey modes should think carefully about how both the mode and potential underlying differences in population may affect economic impact results.
