The use of systems models to identify food waste drivers

Abstract

In developed countries, the largest share of food waste is produced at household level. Most studies on consumers' food waste use models that identify covariates as significant when in fact they may not be, particularly where these models use many variables. Here, using EUlevel Eurobarometer data from 2013, we use alternative analytical methods that avoid these problems (Bayesian Networks) to identify the impact of household characteristics and other variables on self-assessed food waste. Our analysis confirmed that the country, the age of the respondent, the status (student/non-student), and a belief that the family wastes too much are related to the level of self-assessed food waste. But we found no evidence that waste behaviours differ between people living in urban and rural areas, and little support of a difference between genders. Households from lower-income EU countries (e.g. Portugal, Greece, Bulgaria, Cyprus and Latvia), as well as students and young adults tend to report higher levels of food waste. Hence, the adoption of an EU strategy based on the concept of subsidiarity, and of country-level policy measures targeting different age groups is suggested. Furthermore, our analysis shows that policy makers need to be wary of relying on analysis based on large datasets that do not control for false-positives, particularly when sample sizes are small.

Routledge Handbook of Food Waste

“…Understanding why age and country-level differences occur may be of paramount importance for designing better food waste policy interventions, and needs further research via multiple methods. Nevertheless, the probabilistic understanding of the drivers of food waste that have been showcased in Grainger, M. J. et al (2018aGrainger, M. J. et al ( , 2018bGrainger, M. J. et al ( , 2018d allows future action and research.…”

“…The use of systems models to identify food waste drivers - Grainger, M. J. et al (2018a) This paper investigated the drivers of household food waste using Bayesian Networks to identify the impact of household characteristics and other variables on self-assessed food waste. Using EU-level Eurobarometer data from 2013, the study confirmed that the country, the age of the respondent, the status (student/non-student), and a belief that the family wastes too much are related to the level of self-assessed food waste.…”

“…As with Grainger, M. J. et al (2018a), the application of the machine learning algorithms has enabled new insight into the drivers of household food waste. Again, it is interesting to note that some of the drivers identified as important by previous literature, such as awareness of the food waste problem and shopping habits, here are found as not important.…”

Model selection and averaging in the assessment of the drivers of household food waste to reduce the probability of false positives

“…Some authors have detected geographical differences in the individual behaviours towards food waste among EU countries, due to factors such as per-capita income, and citizens’ perception of sustainability issues [ 7 , 9 – 11 ]. Although developed countries present high per-capita incomes, hunger in these countries is a reality: e.g., approximately 4% of US households, and more than 5% of Australian households are experiencing hunger [ 12 ].…”

Toward the creation of novel food waste management systems: A network approach

“…Recent estimates of food loss and waste generation range between 194-389 kg per person per year at the global scale, and between 158-298 kg per person per year at European level (Corrado and Sala, 2018), which translates into 88 million tons of food annually wasted, or equivalent losses of around 143 billion Euro (FUSIONS, 2016). Since the European Union (EU) is globally the second most significant contributor in terms of per capita food losses and waste at consumption and pre-consumption stages (FAO, 2019), it is unsurprising that debates on food waste management are thus becoming increasingly vocal amongst European stakeholders at national and local level (Gustavsson et al, 2011;Priefer et al, 2016;Grainger et al, 2018a).…”