Popping the Question: The In uence of Survey Design on Estimated Visitor Spending in a Region

Regional scientists often use surveys to collect spending data as a rst step in conducting state and local economic impact studies. This paper examines the in uence of survey question type and wording on the information obtained about the expenditures of over 2,100 cruise passengers that visited Bar Harbor, Maine. Results show that average total spending amounts reported per passenger are not statistically dierent between surveys using open-ended ($104 per passenger) and categorical ($107 per passenger) expenditure questions. On the other hand, sample averages from surveys asking two dierent passenger spending questions-about group and personal purchases-are statistically dierent in ve of eight product categories (e.g., meals and drinks, souvenirs) considered. A comparison of these survey results to per- passenger spending amounts from a regression analysis of taxable restaurant sales in the Bar Harbor region is used to estimate possible data reporting errors made by survey respondents. The study results suggest that Bar Harbor cruise passengers made expenditures of between $16.9 million and $23.2 million in 2017, and these visitors accounted for an estimated 3.6 to 6.0 percent of annual restaurant sales. The methods and results of this study may be of interest to regional scientists and organizations that conduct expenditure surveys used in economic impact analysis.