Country of origin or EU/non-EU labelling of beef? Comparing structural reliability and validity of discrete choice experiments for measurement of consumer preferences for origin and extrinsic quality cues

• Published in: Food quality and preference 2014-06, Vol.34, p.50-61
• Main Authors: Lagerkvist, Carl Johan, Berthelsen, Thomas, Sundström, Kristian, Johansson, Helena
• Format: Article
• Language: English
• Subjects: Beef; Choice experiments; Economics; Economics and Business; Ekonomi och näringsliv; Food labelling; Food quality; Food Science; Hierarchical Bayesian estimation; Livsmedelsvetenskap; Nationalekonomi; Probability Theory and Statistics; R-index; Samhällsvetenskap; Sannolikhetsteori och statistik; Social Sciences
• ISSN: 0950-3293; 1873-6343; 1873-6343
• DOI: 10.1016/j.foodqual.2013.12.009

•Preferences for beef labelling origin and quality cues was studied using choice experiments (CE).•CEs, are widely used to elicit attribute importance.•CEs with and without presence of price was compared.•Gaps for attribute importance was supported, but more so on levels than in orderings.•CEs reveal violations of transitivity and dominance. The use of food labels and packaging information on place of origin and credence quality has increased, but it is still unclear how such information relates to consumer decisions and choices concerning the relative valuation of attributes. Valid measurement methods are needed to assess the importance attributed by the public to such information, but also to inform the industry about market demand. Discrete choice experiments (CEs) are frequently used to assess consumer preferences for food attributes. However, suitable methods are needed to reveal the probabilistic nature of preference data, so that heterogeneity can be explicitly accounted for. An R-index measure was modelled to account for choice probabilities in this study. Two surveys of Swedish residents, without (n=506) and with (n=278) a price vector, compared attribute importance for origin and extrinsic quality labelling of beef in CEs. Problems of structural reliability owing to unstable associations in attribute importance between formats were supported, but on levels rather than orderings. The R-index analysis of CE data revealed violations of attribute transitivity and dominance. Extending CE analysis to include choice probabilities may be useful for consumer profiling and for optimisation of product information. Labels with specific country-of-origin information instead of a wider EU/non-EU designation were the most determining attribute in this study.