Willingness-to-pay for a bus fare reform: A contingent valuation approach with multiple bound dichotomous choices

摘要

The aims of this study are twofold: to measure travelers' willingness to pay (WTP) for bus travel under various travel distance for performing a bus fare reform, and to evaluate the reliability and validity of contingent valuation approach of dichotomous choice with multiple follow-up questions (DCm). To achieve so, this study designs triple-bound dichotomous choice contingent valuation scenarios, and employs multilevel interval regression to capture possible endogeneity within individuals. The estimated models using data with all three bounds, with the first two bounds, and with any specific single bound mostly gave consistent parameter significances and effect directions. However, the WTP estimated using the single third bound model demonstrated a different pattern from the other models, suggesting possible weariness effects. The analysis results also revealed yea-saying and free-riding effects, implying that respondents tended to say yes if their first two responses were yeses, and say no if their first two responses were noes. The yea-sayers had high income and low frequency in using public transit. On the other hand, the free riders significantly less supported the bus fare reform. Under well control of bound and path effects, WTP estimated by the proposed models was consistent with the WTP estimated with actual mode choice data, implying an anchor effect of current stage-based bus fares on travelers' WTP for future distance-based bus fare scheme. This study suggests that researchers who use DCm to evaluate WTP should investigate internally inconsistent responses caused by psychological and technical factors; DCm provides robust WTP estimates if the survey has been carefully designed with potential bound and path effects well controlled. More importantly, those internally inconsistent responses also provide information that is useful to fare reform. (C) 2016 Elsevier Ltd. All rights reserved.