A comparison of attribute non-attendance in online and mail choice experiment surveys for eliciting Colorado homeowners values for reducing wildfire risks
| Authors: | José J. Sánchez, Tom Holmes, John B. Loomis |
| Year: | 2026 |
| Type: | Scientific Journal |
| Station: | Pacific Southwest Research Station |
| DOI: | https://doi.org/10.1080/21606544.2026.2718961 |
| Source: | Journal of Environmental Economics and Policy |
Abstract
Although there have been several contingent valuation studies comparing online and mail surveys, only a few comparative studies have been conducted using choice experiments (CE). CEs portray each choice option as different levels of attributes. The assumption is that survey respondents make trade-offs between attribute levels when making their choices among alternative options. Recent research has demonstrated that respondents do not always pay attention to all attributes when making choices. Methods have been developed to estimate attribute non-attendance (AN-A) and its influence on willingness to pay (WTP). This study provides the first comparison of WTP values for environmental programmes obtained from online versus mail CE surveys in which the impact of AN-A is evaluated. A CE survey of homeowners was conducted to estimate homeowner WTP for alternative programmes that reduce wildfire risk and damage to their home relative to a status quo alternative. Using the equality-constrained latent class model to infer AN-A, we found that AN-A noticeably increased goodness of fit in both mail and online samples compared with multinomial logit and random parameter logit models. WTP results are substantially lower when accounting for AN-A.