Testing the effects of ordering and salience on sustainable food product choices in the UK

The challenge

Producing, transporting, storing and disposing of food products has a significant impact on the environment, and accounts for up to a fifth of the UK’s greenhouse gas emissions.[1] The Climate Change Commission among others has highlighted a shift towards more sustainable (and healthier) diets in which meat and dairy are reduced in favour of plant-based products as a key stage on the path to Net Zero. But while large proportions of people say they have started or are willing to do this, habit, culture, marketing, availability and many other drivers continue to push them in the opposite direction.

The environment in which purchase decisions are made is one of those drivers of choice, and there are numerous examples of consumers being ‘nudged’ into picking certain products when these are given greater prominence and salience. An increasing proportion of grocery shopping is done through online supermarkets, so there is interest in understanding how changes to that environment can affect consumer decisions and potentially support more sustainable choices.

The Food Standards Agency (FSA) wanted to know whether increasing the salience of products with a lower carbon footprint within an online supermarket setting can shift consumers towards making more sustainable choices. Specifically, the FSA was interested in the effects of product ordering, and of making that ordering explicit. When consumers see a set of products from which they can make a purchase choice, does showing the lowest-carbon options first affect the choices they make? Moreover, does telling them that this is how products have been ordered have any (additional) effect?

What we did

We conducted a three-armed online randomised controlled trial (RCT) in which we asked participants to shop for a meal for two in a simulated online supermarket. We showed products from the range available in the online grocery store of a major retailer; these were randomly ordered in the control arm, but ordered by carbon footprint in the other two arms.

In one of these arms, we placed a statement at the top of each product page to inform participants that the products had been ordered from the most to the least environmentally sustainable (see right); the other arm did not have this message. Participants were given a budget that was high enough to cover the cost of the most expensive products on the list, and the option to enter a prize draw to receive the items they chose plus any change from the budget.

What we found

We took steps to ensure choices within the experiment were as realistic as possible, and found that neither of the ordering interventions affected the probability of choosing more sustainable products. Contrary to some other findings in the literature, participants’ choices were not affected by a product's position.

There was however a distinct pattern of preferences between products in each category – participants favoured certain products in each category, regardless of price – which was the same between arms, despite the fact that the products appeared in different places on the page. This suggested that consumer choices for grocery food products were mainly driven by prior preferences, and that they might be too in-grained to be changed by subtle rearrangements of choice architecture. It is a useful example of the importance of examining specific questions and hypotheses, rather than relying on wider, less targeted evidence.

[1] Garnett, T., Smith, P., Nicholson, W., & Finch, J. (2016). Food systems and greenhouse gas emissions. University of Oxford: Food Climate Research Network.