A Choice Experiment Model for Sustainable Consumer Goods: A Systematic Literature Review and Workflow Design

Abstract

Although various methodological biases have been shown, the choice experiment (CE) literature has confirmed the relevance of sustainability in consumers’ purchase choices. Analysing 186 case studies through the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) methodology, this study defines the state-of-the-art and future perspectives of the CE approach, allowing us—for the first time—to evaluate in a single study all variables to be considered for setting up the CE questionnaire, with a focus on the selection of attributes and levels, maximising the reliability of the result, and minimising potential method biases. This paper defines a standardised workflow to expand and refine a sustainability perspective that can potentially drive cross-cutting CEs in every consumer good by investigating the accuracy of characteristics driving the willingness to pay (WTP) a premium price for greener consumer goods. Most of the studies analysed in this article concern food products (92%), and around half (51%) focus on sustainability-related aspects, frequently described in generic terms. The results show how defining an adequate number and type of attributes and levels characterising the target product is crucial for a bias-reduced study. These need to be concrete and familiar, and using labels is essential to enhance informed choice, with sustainability being a far-reaching concept with multifaceted definitions. Moreover, choosing a neutral target product, defining the correct sample size, selecting a balanced and representative group of respondents, and using the right analysis model can also minimise potential bias.

Assessing Farmers’ Willingness to Pay for Adopting Sustainable Corn Traits: A Choice Experiment in Italy

“…However, to minimize potential bias and ensure a manageable number of choices for the participants, the levels were further reduced to 13. Indeed, previous studies have shown how the number of attributes and levels must be broad enough to describe the product adequately but not be excessively complex enough to overwhelm the customer, resulting in attention bias [17]. The 13 levels selected were distributed as follows: 3 levels for yield, 3 for yield under water stress conditions, 3 for disease resistance, and 4 for price (Table 1).…”