Most people know and want to eat healthier, but this isn’t translating into action.

We think that's because it's hard - the environments we live within are making it difficult for us to make the healthier choice.

Action is needed to transform our environments to make the healthier choice, the easier choice. Britain’s food retailers play a major role in our food environment - and are key to improving the nation’s diet.

That is why we unveiled a new policy for mandatory health targets for the largest grocery chains. This would incentivise action from grocery chains to adjust their shopping environments so the healthier choice is the easiest choice. This could be transformative with the potential to help four million people achieve a healthier weight.

This blog looks at the variation in data across geography and demographics to better understand the role health targets could play in making sure everyone has access to healthier food, no matter where they live.

The not-so-hidden income pattern in our food purchases

Re-analysing the food purchasing data from Kantar’s Worldpanel Division Take Home Data that we used to develop our Health Targets, we were able to break down the healthiness of food purchases by income group, deprivation, local authority and region.

Although most people - no matter their background - say they want to eat healthier, there is a measurable difference in the healthiness of what's being purchased depending on geography and income.

Unsurprisingly, we found that the wealthier you were, or the wealthier the area you lived, the healthier your food purchases tended to be - with the lowest income groups making the least healthy purchases.

The charts below show the average health score, as measured by a converted nutrient profiling model score*, by income group and areas of deprivation (using the index of multiple deprivation measures or IMD). It should be noted that these are not entirely linear trends, with there being greater variation for areas of deprivation.

However, these variations are relatively small - the difference between average health scores are only 1 or 3 points away from each other. This is less variation than we see across different retailers where there is a variation of 5 points from 63 for the worst performer and 68 for the best performer.

Geography seems to matter too

We are confident that your environment impacts your health. It is not surprising to us that our observational analysis also found geographical patterns - urban areas tended to make healthier purchases compared to rural ones, the south of England shopped more healthily than the rest of the UK, and the most deprived areas shopped the least healthily.

Regionally, Scotland had the least healthy average shop, with Wales coming in second, followed by the North of England.

London proved to be shopping most healthily. The top ten healthiest local authorities were all based in London and the South East. Some of the healthiest areas based on retailer food sales were not necessarily the most affluent. Some of the areas with the healthier scores were areas with the highest proportions of neighbourhoods among the most deprived (such as Tower Hamlets or Barking and Dagenham).

What’s driving variations in healthy purchasing?

There are four things that we think most influence the healthiness of purchasing:

1. the food available
2. the food itself and whether it's higher in calories (e.g. ‘junk food’)
3. promotions, advertising and placement
4. cost and convenience.

But from this analysis, we don't fully know what's accounting for the variations we’ve found.

Is it down mainly to personal preference or is it more to do with what's being sold in your community? Do people living in one area really have very different tastes to one another? Does everyone have access to the same choices and options? Are people with lower incomes more likely to be targeted by in-store promotions?

Almost everyone regardless of income or geography says they want to eat healthily - but for some reason that desire isn’t translating into what people are putting in their shopping baskets.

What’s getting in the way of people being able to make the choices they want to?

It might be easiest to think that this is driven by costs - after all our analysis found that higher income groups purchased healthier baskets. However, we also found that it’s not necessarily the case that the ‘high-end’ supermarkets sell healthier foods compared to the ‘budget’ ones - our initial research found that there wasn’t a strong link between the health score of a retailer and the average cost of its shopping basket.

Retailer’s responsibilities

A big part of this is understanding how much retailers are responding to demand and how much they are shaping it.

Anecdotally, we’ve heard that major retailers do cater to local demographics and markets, and sometimes decide what any given shop should stock based on what branches serving similar demographics are stocking. If this is true, it’s plausible that unhealthy patterns are being perpetuated and reinforced instead of being disrupted.

But the truth is that this is an aspect of our food environment that’s a little unknown. It’s hard to know what types of decisions businesses are making and why - and how health is factored into this decision making.

This is why we want to partner with businesses to better understand and identify ways in which they can prioritise health alongside commercial success. Our partnership with Asda demonstrates that retailers are committed to helping their consumers make healthier choices.

How could targets help to close the health inequality gap?

Our policy of mandatory health targets for major supermarkets has the potential to transform the way food is sold to us. It’s essential that these targets benefit everyone across the country, pushing all stores across the country to be healthier - regardless of their location or consumer base.

But if some areas and groups are already making healthier purchases than others, could retailers try and game a national target - by just selling even more healthy food to the people already shopping healthily, instead of improving the healthiness of everyone’s purchases?

Whilst in theory this is possible, we think it’s unlikely that businesses would take such an approach: it would be complicated, hard to implement and would be unlikely to help them achieve a mandatory target at a national level.

If they were to only target the most affluent and healthiest stores, this would require them to make shifts to be even healthier - which feels difficult practically.

And do we really think retailers would actively want to disregard the health of their consumers, given many consumers want healthier options and we know many retailers are already committed to healthier sales?

But to avoid any risk of that we recommend that if mandatory health targets are adopted, socio-economic and regional data should be monitored to prevent targets being 'gamed'.

This requirement to monitor and collect the data for regional breakdowns of health scores could also be supplemented by stronger measures. For example, penalising retailers if the difference between their highest and lowest scoring stores gets too wide.

We also know that approaches that don’t demand personal action are the most effective at improving health and diets.

By setting national targets about how healthy the average shopping basket should be, we believe we can shift the responsibility from the individual to the supermarket - making it easier and more achievable for all shoppers to make healthier choices regardless of their income or area. So that everyone - no matter where they live or what they earn - can benefit from the healthiest possible options and choices.

If retailers are able to achieve the targets we have proposed, this will impact all regardless of income or area. It would encourage all stores to become healthier. And if retailers implement targets fairly alongside measures to monitor impact, those in the least affluent areas stand to benefit the most and could see the greatest improvements to food on offer.

[*See technical appendix for details on the converted NPM score calculation. All NPM references pertain to a converted NPM score.]