Climatic burden of eating at home against away-from-home: A novel Bayesian Belief Network model for the mechanism of eating-out in urban China

Abstract

Dietary patterns of eating away-from-home (AFH) considerably differ from those of eating at home in urban China, thus generating varied carbon footprints. However, few studies have investigated the effect of eating places on diet-related climatic burden, and few have modelled the mechanism under the condition of eating-out because the decision of consumers on whether to eat AFH or at home is determined by multiple non-linear socioeconomic factors. Here, we compared the carbon footprints of eating at home and AFH using household survey data from 12 Chinese provinces, and developed a Bayesian Belief Network (BBN) model to identify key factors of eating AFH. Our findings show that eating AFH leads to higher climatic burdens though respondents consume less food on average than when eating at home. However, in urban areas, the carbon footprint generated increases more rapidly from eating at-home than when eating AFH. The BBN model was found to have strong capability to predict the possibility of eating out with an accuracy of 89%. Although diet patterns and embedded carbon footprint vary considerably across provinces from northeastern to southwestern China, sufficient evidence could not be found to support the influence of geographic factors on the decision of respondents to eat AFH at large scale. Instead, individual occupation and income were found to be the two key contributors. Thus, merely estimating the carbon footprint of food consumption is currently not sufficient, but social and economic elements need to be quantitatively considered to differentiate the eating-place effect on diet-related climatic burden.

Keywords: Bayesian Belief Network; Carbon footprint; Climate change; Eat away-from-home; Socioeconomic transitioning; Urbanization.