“Payment card or food parcel?” What people referred to food banks choose and effects on food insecurity, mental health, and wellbeing: a randomised, controlled trial.
THE PROBLEM
Since 2010, more and more people have been turning to food banks for assistance. Organisations like the Trussell Trust have been keeping track of this increase, noting the number of times individuals receive food parcels. For example, between 2022 and 2023, food parcels were given out nearly three million times, which was more than double the number from five years prior. A survey conducted by the Food Standards Agency revealed that around 4% of households used a food bank by January 2023.
With this growing demand, food banks are struggling to meet needs. They often rely on surplus food and donations, which can be inconsistent in quantity and quality. Many food banks have trouble getting sufficient supplies, especially during times of high demand. In 2022, two-thirds of the Independent Food Aid Network's members reported issues with their food supply. The Trussell Trust even had to make an emergency appeal to buy more food after using up its reserves.
There are concerns about whether the food provided by food banks is meeting the needs of those who use them. Food parcels are often pre-packed, and their contents may not match recipients' preferences or dietary requirements. Also, receiving food parcels is unlikely to address the multiple financial needs people using food banks face. Many individuals feel embarrassed or stigmatised about receiving food charity, and many people experiencing food insecurity do not use them. The sustainability and effectiveness of food banks has been questioned. Critics suggest that simply providing food parcels is not an effective response to food insecurity. Alternatives like offering money, vouchers, or access to social supermarkets are being considered, but there's a lack of solid evidence evaluating these alternatives.
RESEARCH QUESTION
Among people who are given a choice between receiving pre-packed food parcels or a payment method and means to purchase their own food, what proportion of people choose the payment method over food parcels?
Does providing people who are referred to a food bank a choice to receive either food parcels or a payment method and means to purchase their own food reduce household food insecurity and improve mental health and wellbeing?
LATEST UPDATE
Learn about the latest update on our trial here (PDF).
THE INTERVENTION
We conducted a research study at a food bank in London. The standard practice was to provide 6 pre-packed food parcels, available to collect weekly, to those referred to the food bank. In our trial, new referrals to the food bank were invited to participate. After receiving their first food parcel, they would then be randomly assigned to one of two groups:
Control Group: They would continue receiving pre-packed food parcels for five weeks, consistent with the food bank's standard practice.
Treatment Group: They would have the option to choose between receiving food parcels or a payment method to purchase their own food weekly for five weeks.
STAKEHOLDERS
Our research aimed to provide evidence is relevant to food bank users, food bank managers, food bank staff, food bank volunteers, and national policymakers.
SETTING
A food bank in London.
TARGET POPULATION
Food bank users.
KEY OUTCOMES
There were three main outcome measures for the trial:
The proportion of people choosing the payment card option over the food parcel option in the intervention group.
The difference in proportions of participants with moderate and severe food insecurity after at the study end point (6 weeks from baseline) in the control group compared to the treatment group.
The difference in mental health and well-being score at the study end point (6 weeks from baseline) in the control group compared to the treatment group.
EVALUATION DESIGN
The evaluation was developed in collaboration with stakeholders to minimise the burden on all participants, who were food bank users and may have been in vulnerable situations. We carefully selected study methods with this consideration in mind. Additionally, an adaptive design was implemented to assess the quality of data and participant burden early in the study. This ensured the collected information was of high quality while avoiding excessive burden, which could have led to dropout. Alongside the randomised controlled trial, the study included process and economic evaluations.
SAMPLE SIZE
A minimum of 260 participants (130 per control and per treatment group) were needed to participate in the study to provide 80% power at a 95% confidence level to detect a minimum effect size of a 7.5 percentage point difference in moderate and severe food insecurity between the treatment and control group. Given the potential for a high drop-out rate, we aimed to recruit a total of 364 participants.
RECRUITMENT
All new referrals to the food bank received invitations to take part in the study via the food bank’s usual communication channels with new patrons. Additionally, individuals could receive study information in person at the food bank. Every participant received user-friendly study materials, available in multiple languages and formats to cater to different needs. Members of the research team and research champions within the food bank were on hand to answer questions and provided support to all participants.
DATA COLLECTION
During recruitment, we gathered information on who joined, who participated, and who declined. We also noted participants' choices between food parcels and payment methods. Participants needed to complete questionnaires when they joined, halfway through, and at the end, focusing on food security, diet, mental well-being, and personal sense of dignity. The questionnaires could be filled out online, by phone, or in person. Participants also had to photograph items bought with the payment method or received in food parcels and send related receipts.
We interviewed some participants to understand their experiences with both food parcels and payment methods, and the interviews were to include important questions to inform the process evaluation. Interviews with food bank staff were conducted for the economic evaluation.
The process evaluation will incorporate qualitative methods to understand how the intervention was implemented, stakeholders’ views, and their impacts over time. The process evaluation aims to explore broader effects beyond the intervention itself. Methods used for the process evaluation include in-depth interviews and focus groups with stakeholders, and collection of process and implementation data from observations at intervention sites.
The economic evaluation aimed to capture the costs and benefits of adopting the new intervention relative to a comparator.
The wider impacts on nutritional quality of diets and environmental sustainability were to be modelled. The modelling focused on the potential impact of the intervention if it were taken up in a wider population.
The study sought approval from the ethics committee at the University of Liverpool. Further oversight of the trial was to be provided by the SALIENT consortium and its Independent Scientific Advisory Board. A protocol for the trial was published in advance of data collection.
This was to be the first randomised, controlled trial, which would test outcomes associated with offering people a choice of receiving a food parcel from a food bank or instead receiving a payment method to purchase their own food in comparison to providing a food parcel with no choice. It aimed to add robust quantitative evidence on outcomes associated with offering people a payment method, as well as qualitative insights into how and why different forms of support may or may not work for different people. It also quantified the costs of offering pre-packed food parcels in comparison to providing people a choice of providing food parcels and payment card.
We will produce publications reporting the main trial results and modelled impacts on health and environment, process and economic evaluations. The outputs of this research will be shared with our partner food bank and with national policymakers. We anticipate that these findings will be of great interest to other local authorities funding, or looking to fund, similar models and will seek to share our learning nationwide.
Rachel Loopstra - Co-investigator
Oyinlola Oyebode - Co-investigator - o.oyebode@qmul.ac.uk
Thijs Van Rens - Co-investigator - J.M.van-Rens@warwick.ac.uk
Claire Thompson - Co-investigator
Steven Cummins - Co-investigator
Juanita Bernal Alvarado - Researcher
Bea Savory - Researcher
Oliver Huse - Research fellow
Natasha Bayes - Postdoctoral researcher
Jess Brock - Researcher
Suruchi Ganbavale - Research Associate
