Conference Agenda

In stated preference (SP) surveys, theoretical restrictions can be used in two ways by imposing the restrictions as is often done in attribute non-attendance (ANA) or testing the restrictions as in whether the data meet the specific requirements of ‘adding up’. However, one could just as easily impose the ‘adding up’ restriction rather than test for its existence. We propose a situation where this type of adding-up restriction may be valuable. In our application, survey respondents are asked to gauge their support for a measure to help their struggling municipal wastewater treatment plant meet required discharge levels. First participants are offered a choice between expanding or not expanding (status quo) traditional treatment capacity. As a follow-up, respondents choose whether to support expanding treatment capacity using alternative ecologically engineered treatment technology (EETT) in the form of constructed wetlands. Finally, participants are asked to choose their preferred wastewater treatment expansion choice among the two alternatives. Naïve estimates suggest that households are willing to pay (WTP) roughly the same amount for traditional expansion and wetland treatment expansion, but also exhibit a strong statistically significant WTP premium for wetlands relative to traditional wastewater treatment. Clearly, the results violate the scope test, but after imposing the adding up restriction the WTP estimates decrease for traditional wastewater treatment and increase for wetland alternatives yielding reasonable estimates of the wetland premium. These findings generally hold for a variety of model specifications including mixed logit and ANA. Finally, we introduce a fractional response mixed logit model to accurately handle instances when survey respondents report having no rank-order preference between EETTs and traditional wastewater treatment.

This study revisits and extends the discussion by Bateman et al. (2008) regarding potential deviations from Rational Choice Theory (RCT), focusing on two hypotheses – Coherent Arbitrariness (CAH) and Discovered Preferences (DPH) – and their manifestations in Discrete Choice Experiments (DCEs). Using data from 5,917 Polish respondents, we implemented multiple treatments varying the cost vector, choice set size, priming techniques, and the placement of a contingent valuation Willingness To Pay (WTP) question either before or after the DCE. Our findings show that the design of the cost vector systematically influences stated preferences and that attempts to mitigate this anchoring effect (e.g., by presenting an open-ended WTP question prior to the DCE) were unsuccessful. The persistence of cost vector effects over time aligns with CAH, while the lack of any discernible preference refinement challenges DPH. Further analyses reveal systematic variations in cost sensitivity and show that excluding respondents who violated the Weak and Strong Axioms of Revealed Preferences did not reduce cost vector effects. In light of these results, we recommend using diverse cost vectors to verify the robustness of DCE-based welfare estimates and call for further research into the broader impacts of heuristic behaviors in stated preference methods.

This study examines how exogenous and endogenous information affects consumers’ willingness to pay for labelled food products across six European countries using a choice experiment. While sustainability attributes are positively valued, additional information has minimal impact on attribute valuation. Instead, more information increases the valuation of unlabelled products, suggesting spillover effects in the valuation of information provision. Importantly, respondents randomised into additional information treatments exhibit higher choice consistency, underscoring the benefits of comprehensive information for stated preference studies. We do not find any difference between exogenously provided and exogenously sought information, both in terms of stated preferences and choice consistency.

This study investigates convergent validity between elicitation formats within contingent valuation methodology, with a particular focus on theory-driven formats that are designed to align with incentive compatibility. While earlier literature endorsed the single binary choice elicitation format as the only incentive-compatible format, Vossler and Holladay (2018) identified theoretical conditions under which payment card and open-ended elicitation formats may also satisfy incentive-compatibility, thereby encouraging truthful preference revelation. Nevertheless, empirical findings have revealed a lack of convergent validity across these formats, with behavioral biases proposed as a possible explanation. This study seeks to deepen understanding of this issue by comparing these three theory-driven (and arguably incentive-compatible) formats. Specifically, we assess whether single binary choice and payment card demonstrate robustness against bid-vector effects, and to what extent biases – such as anchoring, yea-saying, and strategic misinterpretation – influence responses in those formats. To explore this, we conducted a large-scale, multi-country survey with 12,343 respondents across Poland, Sweden, Hungary, Spain, Germany, and the Netherlands, examining consumers’ willingness to pay (WTP) for an EU-wide biodiversity enhancing programme. Our analysis reveals significant discrepancies in WTP estimates across formats, with Wald tests consistently rejecting convergent validity. However, analysis of choice frequencies suggests that convergent validity holds for lower bid levels, whereas higher bids introduce more pronounced divergences. We argue that these discrepancies stem from both behavioral biases and methodological factors, including bid vector design and econometric assumptions. Our results underscore the critical influence of bid structure on WTP estimates and highlight the need for further research to refine elicitation formats in contingent valuation studies.