Speaker
Description
Keywords: Food Labelling, Consumer Choice, Discrete Choice Experiment, Willingness to Pay, Traffic Light System, Fat Content, Carbon Footprint
1. Introduction
Individual food choices influence both personal health and the broader environmental ecosystem. To achieve emission reduction targets by 2050, it is crucial to decrease the demand for high-emission foods and transition towards sustainable diets. Simultaneously, promoting healthier dietary choices is essential for improving public health by reducing the prevalence of diet-related diseases. Recent studies have also highlighted that a more sustainable diet could lead to improved health as a co-benefit. Consequently, aligning the goals of healthier and more sustainable food consumption necessitates effective policy guidance.
This study explores the effectiveness of food labelling as a potential policy tool to promote sustainable and healthy food consumption. Specifically, this research employed a consequential Discrete Choice Experiment (DCE), using a ready-meal salad as the experimental product, to assess participants’ willingness-to-pay (WTP) for salads featuring different levels of healthiness and environmental attributes, under different labelling systems. The study aims to address the following research questions: (1). What is the willingness to pay (WTP) among individuals for environmental and health gains? (2). How does the dual-labelling system – simultaneously presenting carbon and fat content information on the packaging – affect the WTP for these environmental and health gains?
The paper contributes to the literature in two dimensions: First, the paper contributes to the literature on the effectiveness of food labelling on food consumption. A key limitation in current literature lies in its predominant focus on individual labels, neglecting the potential interplay between multiple labels. Additionally, much of the previous research relies on surveys or hypothetical experiments, potentially suffering from the hypothetical bias issue. This study addresses these gaps by comparing outcomes of a consequential experiment involving both single and dual labelling, aiming to offer a clearer understanding of how labelling influences consumer choices.
Second, it adds valuable insights to the field of motivation crowding theory within the multi-labelling framework. In the context of consumer consumption, the interaction of information yields mixed results. This study adds to the literature by presenting empirical evidence of label proliferation, particularly between environmental and health traits, in the domain of food products.
2. Research Design
This project employs a consequential DCE approach using a ready-meal salad as the experimental product to examine how different information influences consumers’ preferences and willingness to pay (WTP) for low-carbon and low-fat features of food. This experiment is designed to examine the between-group effects, and respondents were randomly assigned to one of the three treatments: a single label (including either fat content or carbon footprint) and a dual label (featuring both fat content and carbon footprint) with the product labels representing carbon footprint, fat content, or a combination of both.
To encourage respondents to reveal their true preferences, this experiment follows a consequential design. At the end of the experiment, one choice task was randomly selected and revealed to participants. If a participant chose to purchase a product from the selected choice set, the product was provided with its cost deducted from the participation fee; if a participant did not buy any salad in the selected choice set, they received the full participation fee.
The choice sets in each treatment remained consistent, with the only distinction being the labels presented to participants. This product was evaluated based on two non-monetary attributes—carbon footprint and fat content—each of which was categorized into three tiers: high, medium, and low. To communicate these attributes to consumers, the packaging incorporated a traffic-light label system. Additionally, the product was priced across a spectrum of nine distinct points, ranging from £2.0 to £6.0, distributed evenly.
3. Estimation Strategy
The choice data from the DCE will be modelled based on the Random Utility Model. It is assumed that respondent n in treatment t will choose option j which yields the highest utility given the choice task. The utility is separated into two parts, specified in equation (1), with V_ntj representing the part of the utility observed by the research, and ε_nj representing random error capturing the individual factors outside of the model.
"U_ntj "= " V_ntj "+ " ε_nj " = " ∑_(s=1)^S▒〖γ_sjt A_sjt 〗 "+ " β_t P_jt "+ " ε_nj " (1)
A_sjt includes all the dummy variables for carbon and fat attribute levels in Table 1. And γ, β are parameters representing the marginal utilities of the corresponding attribute and price. The WTP for a certain attribute level is calculated as:
wtp = -γ_sjt/β_t (2)
4. Results
The study employs a randomized controlled design with 236 participants assigned to three treatments in a consequential discrete choice experiment (DCE). In summary, the results indicate three key findings:
- Consumers exhibit a higher average WTP for salads with reduced fat compared to reduced carbon levels.
- The WTP patterns for both features demonstrate a non-linear characteristic. Respondents are willing to pay more to avoid products that are high in carbon or fat, but this willingness does not increase proportionally when the reduction is from a medium to a low level. This indicates a diminishing marginal WTP for additional reductions beyond the initial decrease from high levels.
- The addition of dual labelling does not significantly influence the WTP for either low-carbon or low-fat attributes.
4. Conclusion
This research uses a Consequential DCE to explore the WTP among individuals for environmental and health gains and the dual-labelling information crowding effect in one experiment setting. The preliminary results show that the WTP for private gains (low-fat feature) is significantly higher than public gains (low-carbon feature); The WTP for these features presents a non-linear feature; and no significant information crowding effect is found. The results from this study offer insight into both the motivation crowding theory and the potential of policy intervention in the food sector.
