Mapping food insecurity risk in neighbourhoods

By Dianna Smith, University of Southampton, Lauren Rixson, Public Health England, & Nisreen Alwan, University of Southampton

There is a rare week or even day when food insecurity (where someone is unable to access food which meets dietary and nutritional requirements), often called food poverty, isn’t in the news. Between positive news stories like Marcus Rashford being awarded an MBE for his campaigning about child food poverty, we also hear about the negative impact of changes to the benefit system – greater demand for aid like food banks and a reminder of the ‘heat or eat’ dilemma facing households. With the loss of a temporary increase in Universal Credit benefits, the most vulnerable will struggle further to meet basic needs and go further into debt, according to charities like Citizen’s Advice.

Food insecurity in England is not a new social challenge, however, there has been very little known about who experiences food insecurity and where they live across the UK as data are not regularly collected, unlike in other countries including the United States. This changed from 2019, when regular measures of household food insecurity were included in the annual Family Resources Survey administered by the Department for Work and Pensions. What is still not available is the detail of where more people experience food insecurity in the countries of the UK, despite calls for this local data collection from 2008 onwards. This is where geographers are making an impact.

How can mapping risk help?

Back in 2018 the first author of this blog published a method of predicting risk of household food insecurity in England. Using data made available from the Trussell Trust, we identified that there was no statistically significant relationship between where food banks are located and either more deprived areas, or our measure of area risk that was based on how many people lived in higher-risk types of households or who claimed benefits. What happened next was a lot of requests from local authorities who were working with charities to support their communities in accessing food, either from food banks or other options like food pantries or community fridges. Though resources like food banks are usually charities, they often receive funds from local government to support the services. It’s important to know where these and other resources, like activities to help children and families experiencing holiday hunger, may be most needed so funding and effort is prioritised effectively. Mapping out food banks and assuming that shows where more people need help isn’t the most efficient option, because many people who are food insecure do not go to food banks. We know that food banks need volunteers and space to hold food, so these resources also influence where they open. We made our data available online for free, and then looked to improve on what we had created.

How we worked to improve the food insecurity mapping

Over the course of about a year, we did a scoping review to find out about newer literature to describe households in England and the UK who were food insecure, to consider new sources of data that would make our risk mapping more accurate. We also interviewed people who work in food banks and local authorities to ask about the people they had been helping. We found that mental health and disabilities and poor educational attainment were recurring themes in the literature and our interviews, as well as the characteristics we already included with benefits claimants and people living in low-income households. We added a new aspect to the measure, to account for structural predictors of food insecurity based on the review and interviews: access to good download speeds, bus stops, medium or large supermarkets (where food is less expensive) and places that employ at least 100 people. Now we have a measure that focuses on population characteristics (composition) and data about areas (context). After comparing our results that rank neighbourhoods in England (Lower Super Output Areas, typical population of about 1,500) with outcomes we know are associated with food insecurity (percentage of children claiming free schools meals or obese at year 6 and the 2019 Index of Multiple Deprivation) we found that the new measures were related to these outcomes, so the patterns we mapped are likely to be accurate. In other words, the areas with higher risk of food insecurity also had a higher percentage of children who were claiming free school meals or were obese.

Exploring the geography of risk, for the top decile of risk in the Compositional domain, 32.5% of the LSOAs were located in the North West region with 96% in urban areas. When comparing the two domains, Red LSOAs are in the top 10% most risk for both domains, while blue LSOAs are in the top 10% of risk for one domain but the lowest 10% for the other (Table 1).

Index rank	Simple	Complex	Urban	Rural
			Compositional	Structural	Compositional	Structural
1	Middlesbrough 003F	Tendring 018A	Wirral 016E	East Lindsey 006B	Tendring 018A	Carlisle 001D
2	Blackburn with Darwen 006E	Wirral 016E	Wirral 011C	Forest Heath 003G	Wakefield 039D	Northumberland 019C
3	Wirral 016E	Wakefield 039D	Blackpool 006A	East Lindsey 006A	County Durham 025B	Northumberland 003B
4	Wirral 011C	East Lindsey 017D	Wirral 027C	East Cambridgeshire 004A	Allerdale 005B	Eden 002D
5	Birmingham 050B	Wirral 009A	Stockport 004B	Herefordshire 009B	County Durham 045F	Teignbridge 003C
6	Birmingham 121B	Scarborough 012B	Middlesbrough 007E	East Lindsey 012C	County Durham 059F	Northumberland 007D
7	Wigan 009C	Stockport 004D	Stockport 004D	East Lindsey 006C	Wigan 031A	Eden 001C
8	Kingston upon Hull 017E	Stockport 004B	Wirral 008C	Central Bedfordshire 007A	County Durham 051E	Herefordshire 020C
9	Wirral 008C	Knowsley 006B	Blackpool 010A	Tower Hamlets 025E	County Durham 051A	Allerdale 002D
10	Oldham 014B/Middlesbrough 011B	Cheshire West and Chester 040B	St. Helens 014D	East Hampshire 004A	County Durham 051D	Eden 006C

Table 1 Ten most at risk LSOAs based on Rank, where 1=highest risk

The importance of geography

For most areas, we recommend mapping the population-level data (composition), and you can see a map here that shows how risk varies across England. However, in more rural areas especially, we suggest mapping the contextual measure alongside and identifying areas where there is poorer access to resources like employment, more affordable food, bus stops and good download speeds. All of these factors can affect household incomes and influence the risk of food insecurity alongside the population risk factors. The data that go into our measures are freely accessible, and we have made our new risk measures available to whoever will find it useful to plan resources in local areas. This might be local government teams looking at possible interventions or groups developing a food poverty action plan to help their local communities. Our new measure uses smaller neighbourhoods and complements the food insecurity maps published earlier this year which are based on survey data from the Food Foundation.

Overall, geographers are playing an important role in bringing together data to support groups that are assisting people in our country through difficult times. We are then making these new data resources available to enable targeting of precious resources from charities to the places where they can make a real difference to people. Alongside this we share our research to raise the issue of food poverty. Hopefully, we will reach a point in the future where such risk maps will no longer be needed.

About the authors: Dr Dianna Smith is a lecturer in GIS and health geography at the University of Southampton. Her research interests include food access/deserts, food insecurity, diet-related health, environmental influences on health and collaborating with public sector and civil society partners to measure health inequalities. She and Dr Alwan are leading research projects funded by NIHR ARC Wessex to develop the indicators shown here, and a new project to understand the health and social impacts of food aid interventions. Dr Alwan is Associate Professor in Public Health at the University of Southampton and Honorary Consultant in Public Health at University Hospital Southampton NHS Foundation Trust. She leads epidemiological and public health research in maternal and child health towards optimising the wellbeing of families and preventing future chronic disease. During the pandemic, she also focused on the recognition and the quantification of morbidity from COVID19, having initiated the call to Count ‘Long Covid’.