Conceptualising food banking in the UK from drivers of use to impacts on health and wellbeing: A systematic review and directed content analysis

Food banks have become commonplace in the UK as an emergency response to food insecurity. However, food banks are not a long-term solution to food insecurity and are often not accessed by those in need. In the context of the cost-of-living crisis, and increased food insecurity, this systematic review applied market/government failure theory, voluntary failure theory, and Radimer et al. ’ s (1990) domains of food insecurity to explore three important aspects relevant to the food banking experience: the drivers of food bank use; the limitations of the current food bank model; and the impacts of the food banking model for food bank clients. Empirical, peer-reviewed articles written in English with a UK food bank context and reporting relevant data to these aspects were eligible for inclusion. In total, 221 titles were identified using four databases (Web of Science, SCOPUS, PubMed, CINHAL Plus) in July 2022. The final sample of 41 articles (comprising qualitative, quantitative and mixed methods studies), were quality assessed using the Mixed Methods Appraisal Tool. Data were extracted and analysed through directed content analysis. Market and government failures were widely reported to drive food bank use. Insufficiency, paternalism and particularism represented key limitations of the food bank model. Negative health and psychological impacts of food bank use were prominent, yet social impacts were largely positive. Consequently, new solutions are needed to promote positive health and psychological impacts for food bank clients in the UK. The application of these findings to other high-income countries experiencing food insecurity should be determined.


Introduction
Food insecurity is defined as the lack of physical, economic, and social means to obtain sufficient quantities of safe and nutritious food to satisfy dietary needs and maintain a healthy and active lifestyle (Food and Agriculture Organisation, 2021).Radimer, Olson, and Campbell (1990) conceptualise the experience of household food insecurity as: "the inability to acquire or consume adequate quality or sufficient quantity of food in socially acceptable ways, or the uncertainty that one will be able to do so".
Events over recent years -including the COVID-19 pandemic and the cost-of-living crisis -have built upon a sustained period of harsh austerity policies, driving many individuals in the UK into food insecurity.
In their latest report, The Food Foundation (2023) reported that 17.7% of UK households experienced food insecurity in January 2023 -more than double the figure reported for July 2021, just before the cost-of-living crisis started to take effect (Institute for Government, 2022).Meanwhile, many of those already experiencing marginal food insecurity have been pushed into more severe forms of food insecurity (Pool & Dooris, 2022).Food insecurity has previously been associated with negative health and psychological outcomes within high-income countries (Afulani, Coleman-Jensen, & Herman, 2020;Bell et al., 2022;Gundersen & Ziliak, 2015;Maynard et al., 2018).Consequently, food insecurity has become a major public health concern in the UK which must be urgently addressed (Sosenko, Bramley, & Bhattacharjee, 2022;The Food Foundation, 2023).
In response to increasing levels of food insecurity in the UK, as well as other high-income countries such as America and Canada, emergency food aid has become an ever-normalised feature within modern Western society (Lambie-Mumford, 2016;Livingstone, 2015;Tarasuk, Fafard St-Germain, & Loopstra, 2020).The landscape of emergency food aid in the UK is heterogenous (Loopstra, 2018b;Power, Small, Doherty, & Pickett, 2020); emergency food aid largely originates from food banks (Sosenko et al., 2019) (also referred to as food pantries in the US), in which food is donated by the public or corporate food businesses such as supermarkets, and is then collected by individuals who have been referred to the service due to need.The majority of UK food banks operate formally, such as those within the Trussell Trust network, an independently-run national network of over 1300 food bank centres (Loopstra, 2018b;May, Williams, Cloke, & Cherry, 2020b;Trussell Trust, 2022b).Trussell Trust food banks utilise a voucher system, supplying food parcels designed to last for three days to individuals experiencing food crisis who have been referred from other agencies (Loopstra, 2018b;Trussell Trust, 2022e).During April 2021-March 2022, the Trussell Trust report distributing over two million food parcels to those in need (Trussell Trust, 2022d).Meanwhile, independent food banks (also known as community food banks) not affiliated with the Trussell Trust also exist and are usually more informal.Independent food banks may or may not utilise a voucher/referral system, and may be more flexible with what they provide, having their own policies and practices to address food insecurity at the community level (May et al., 2020b).Other forms of emergency food aid also exist within the UK on a smaller, more community-based scale, including food pantries (also known as community supermarkets), in which food can be purchased at a highly reduced price, and soup kitchens or community cafes, in which hot food is served to individuals experiencing food insecurity (Power et al., 2020).
While food bank use is becoming increasingly widespread across the UK (Loopstra & Lambie-Mumford, 2023), a number of limitations with the current food banking model are coming to light.For example, food banks are only designed to provide short-term relief in times of crisis (Trussell Trust, 2022e).Yet, the current demand has triggered increased long-term reliance by those persistently unable to afford sufficient food to meet household needs (Möller, 2021;Purdam, Garratt, & Esmail, 2016).Subsequently, concerns have been raised regarding the ability of food banks to provide an adequate, long-term source of nutritious food to promote health and wellbeing for people experiencing food insecurity (Dowler & O'Connor, 2012;Loopstra & Lalor, 2017;Pollard & Booth, 2019).Additionally, the scope of a food bank to support individuals is constrained by its reach, with evidence suggesting that high proportions of individuals experiencing food insecurity do not access food banks, but instead live with "hidden" food insecurity (Armstrong, King, Clifford, Jitlal, & Ibrahimi Jarchlo, 2022;Loopstra & Lambie-Mumford, 2023;MacLeod, Curl, & Kearns, 2019;Prayogo et al., 2017).Hidden food insecurity can be understood as individuals taking steps to conceal their food insecure status from others, not accessing public forms of food support and instead implementing coping strategies to cope with food shortages (Douglas, Sapko, Kiezebrink, & Kyle, 2015;Power, Small, Doherty, & Pickett, 2018;Puddephatt et al., 2020;Purdam et al., 2016).This may have particularly serious health impacts, as a US study by Zepeda (2018) found that participants who identified as food insecure but did not use food pantries (the US version of a food bank) reported nearly twice the level of actual hunger of food pantry accessors in the area.These results are likely to be similar in other high-income countries such as the UK, indicating the urgency of finding new and appropriate solutions to food insecurity.

Exploring the drivers of food bank use: market failure/government failure theory
Theoretical perspectives to understand the proliferation of food banks in high-income countries include the application of Seibel's (1989Seibel's ( , 1996) ) theories of 'Successful Failures' and 'Shunting Yards' by Ronson and Caraher (2016).Ronson and Caraher (2016) argue that food banks 'successfully fail' by continuously thriving and proliferating despite not making much tangible progress in their mission to reduce food insecurity, representing 'Shunting Yards' by acting as a distraction used by government to shirk responsibility of the intractable issue of food insecurity (Ronson & Caraher, 2016).However, an alternative perspective is offered by market failure/government failure theory (Weisbrod, 1977), a dominant economic theory (Salamon & Anheier, 1998), which conceptualises societal reliance on the voluntary sector as a result of the combined failures of the market and the government (Salamon, 1987;Weisbrod, 1977).Market failure results from a lack of efficient provision of public goods within the free market system (Dollery & Wallis, 2002;Rocha, 2007;Salamon, 1987;Weisbrod, 1977).Relevant types of market failures identified in the literature include unemployment, underemployment, and income inequality (Bartik, 1990;Weeden & Grusky, 2014).Subsequently, government failure is underpinned by the inability of the government to meet this unsatisfied demand (Salamon, 1987).Hence, individuals turn to the voluntary sector (e.g.food banks) to acquire the goods which are not sufficiently supplied through the market or the government (Dollery & Wallis, 2002;Salamon, 1987).
Much existing literature documents the rise in food bank use and describes sociodemographic, health and nutritional characteristics of food bank users, providing insight into who is most likely to be using food banks (Garratt, 2017(Garratt, , 2020;;Garthwaite, 2016a;Lambie-Mumford & Green, 2017;Lambie-Mumford & Green, 2017;Loopstra, 2017;Loopstra, Fledderjohann, Reeves, & Stuckler, 2018;Loopstra, Reeves, & Tarasuk, 2019;Power, 2022;Power, Yang, & Pybus, 2023;Prayogo et al., 2017).For example, Loopstra (2017) found that factors associated with accessing a food bank included unemployment, disability, being a single parent, being of working age and living in a single-person household, and having a very low fruit and vegetable consumption.Additionally, within the literature, financial drivers of food bank use, largely due to austerity and low income appear to be prominent (Garratt, 2017;Garthwaite, 2016aGarthwaite, , 2016b;;Jenkins et al., 2021;Lambie--Mumford, 2019;Lambie-Mumford & Dowler, 2014;Loopstra et al., 2018;MacLeod et al., 2019;Power, 2022;Prayogo et al., 2017).For example, Prayogo et al. (2017) found that of food bank users surveyed, the majority were in receipt of benefits, while Loopstra et al. (2018) found that benefit sanctioning in particular was closely linked to food bank use.Such financial drivers may be traced back to key downfalls within government policy and the market system.Therefore, examining the extent to which market and government failure theory can be applied to understand the drivers of food bank access in the UK may enable a new structural perspective of the issue and has potential to offer theoretical insights on the underlying forces driving people to use food banks.Poppendieck's (1999) seminal work 'Sweet charity?' was influential in critiquing the food bank model, highlighting seven key limitations: inaccessibility, inadequacy, inappropriateness, indignity, inefficiency, insufficiency, and instability.Yet, almost three decades later, food banks are still in existence, functioning by the same operational processes as ever.Applying an additional theoretical perspective may provide novel conceptual understanding of the limitations of the food banking model.Much like with Seibel's (1989Seibel's ( , 1996) ) theories of 'Successful Failure' and 'Shunting Yards', Salamon's (1987) voluntary failure theory problematises the role of the voluntary sector in addressing social problems caused by market and government failures.The theory states that the voluntary sector possesses inherent limitations in its ability to provide goods and services which are sufficient and appropriate to meet community needs (Salamon, 1987).These failures include: the limited ability of the voluntary sector to supply an adequate and reliable stream of resources (insufficiency); the tendency of the voluntary sector to focus on particular subgroups of the population and marginalise other groups (particularism); the inclination of the voluntary sector to place the needs of those it serves in the hands of those more fortunate (paternalism); and the manifestation of amateur approaches in place of professional services provided by trained advisors (amateurism) (Salamon, 1987).

Exploring the limitations of the current food banking model: voluntary failure theory
As previously described, a key limitation of the food bank model is that it cannot offer a long-term solution to food insecurity.As such, concerns have been raised regarding the ability of food banks to adequately support individuals experiencing food insecurity (Bazerghi, McKay & Dunn, 2016;Dowler & O'Connor, 2012;Oldroyd, Eskandari, Pratt, & Lake, 2022;Pollard & Booth, 2019).For example, a systematic review conducted by Bazerghi et al. (2016) found that while food banks may prevent immediate hunger at the time of access, they were constrained both in their ability to provide sufficient culturally appropriate and nutritional foods, as well as by their operational limitations, including restricted opening hours and reliance on funding.More recently, Oldroyd et al. (2022) conducted a mixed-methods systematic review exploring the nutritional adequacy of food parcels, the nutritional status of food bank clients, and the ability of food banks to address food insecurity within developed countries.Findings indicated that while food banks may be a lifeline for many in accessing food, the support provided is often still insufficient to meet nutritional recommendations, and individual cultural, health and social needs.Additionally, Power (2022) systematically examined how food banks may overrepresent Christian ideals, manifesting and reinforcing a broad range of not only religious, but also racial, class and gender inequalities.
The work summarised above may suggest applicability of Salamon's voluntary failure theory to the food bank context, exemplifying insufficiency and particularism of the current food bank model.Therefore, exploring the extent to which voluntary failure theory can be applied to understand the limitations of the food bank model could provide a novel theoretical insight.In the US, Jones and Deitrick (2020) have previously applied Salamon's theory to explore limitations of food pantries (the US equivalent of food banks) and hence, applying this theory within the UK context could provide a useful framework for understanding the specific limitations of the food banking model dominating the UK currently.

Exploring the impacts of the food banking model for food bank users: Radimer et al.'s four domains of food insecurity
The definition of food insecurity offered by Radimer et al. (1990) postulates four important domains in the development and persistence of food insecurity: quantity; quality, social; and psychological.These four domains may be considered to provide a conceptual framework for understanding the experience of food insecurity and thus its impacts upon individuals (MacLeod et al., 2019).
In a narrative review conducted in the US, food pantry (the US equivalent of food banks) clients showed lower diet quality and higher prevalence of depressive symptoms, alongside a number of diet-related diseases including diabetes, heart disease and obesity (Eicher-Miller, 2020).Meanwhile, previous research in the UK indicates that food bank clients experience significantly poorer health outcomes, including mental health outcomes, than the general population (Loh, Knight, & Loopstra, 2021).Yet, it is not clear whether it is these poorer health outcomes that lead to food bank use, food bank use which leads to these poorer health outcomes, or a combination of both.The negative mental health impacts of food insecurity have also been documented in a number of qualitative studies (Douglas et al., 2015;Garthwaite, 2016b;Power et al., 2018;Power & Small, 2022;Puddephatt et al., 2020), with Puddephatt et al. (2020) suggesting that the experience of negative health impacts as a result of food bank access may trigger a 'vicious cycle' of further negative impacts.
The applicability of Radimer et al.'s (1990) definition of food insecurity to a broad range of cultures and countries (excluding UK) has previously been shown by Coates et al. (2006).However, the experience of food insecurity within the UK may have nuanced differences to other countries, including the US, where Radimer et al.'s definition was developed.Given the proliferation of food banks across the UK, understanding the broad range of impacts on users, and their interconnectedness, is vital.Specifically, synthesising data under Radimer et al.'s (1990) four domains of household food insecurity (quality, quantity, social, and psychological) in the UK context may allow a holistic, up-to-date conceptualisation of the personal experience and impacts of UK food banking and, in turn, identify adjustments which could benefit those accessing food banking services.

Current research questions
This paper aims to offer novel conceptual insights on the drivers of food bank use in the UK, the limitations of the current model, and the impacts of the food banking model for clients.To do this, we use a theory-driven approach focused on the application of market/government failure theory, voluntary failure theory and Radimer et al.'s four domains of food insecurity to the UK food bank context.Applying these theoretical approach may provide a deeper understanding of the experience of food bank access, thus contributing towards reducing need and developing more appropriate and beneficial solutions.Despite a growing number of studies being published concerning different aspects of food banking in the UK context, these studies largely remain localised with regards to their geography and their focus.Synthesising research across drivers, limitations and impacts of food bank use within a systematic review will increase the utility of the findings by enabling a clearer conceptual understanding of the experience of food banking in the UK.
Hence, the current review will address the following three research questions.
1 To answer these research questions, we conducted a systematic literature review comprising qualitative, quantitative and mixed methods studies, and used directed content analysis to synthesize research on the topic of food banking in the UK.Directed content analysis aims to apply predetermined codes derived from a theory to the data, in order to validate or extend the theory within the context (Hsieh & Shannon, 2005).As such, the three questions detailed above were explored deductively through application of the aforementioned theories in order to gain further conceptual and theoretical understanding of the experience of food banking in the UK setting.We focused on the food banking model specifically, as this is the largest and most relied upon source of emergency food provision in the UK (Sosenko et al., 2019).

Methods
This systematic review was conducted and is reported following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) (2020) guidelines.

Search strategy
A search of the literature was conducted in July 2022 via the Web of Science Core Collection, SCOPUS, PubMed, and CINHAL Plus databases.The search terms were developed from preliminary reading and appeared as follows: "("food insecur*" OR "food poverty") AND (charit* OR "food bank*" OR foodbank* OR "food aid" OR philanthrop*) AND UK".The search terms aimed to gather articles which related to at least one of the research questions, and were used identically in all databases.

Eligibility criteria
Only empirical, peer-reviewed articles, written in English and based in the UK were included in the review.Given that 9.3 million adults in the UK reported moderate-to-severe food insecurity in January 2023 (The Food Foundation, 2023), the UK specific focus of the review is especially relevant and timely.The decision to include qualitative, quantitative and mixed methods studies was based around wanting to capture the full breadth of relevant literature on the landscape of food insecurity in the UK.Systematic reviews and other review articles were however, excluded.Grey literature was also excluded due to time constraints.Articles that referred to food insecurity without direct reference to food bank use were excluded, as were articles which discussed food banking but did not report data of direct relevance to exploring the drivers, limitations or impacts of food banking (e.g.those exploring prevalence of food bank use, demographic characteristics of food bank clients, and media coverage of food banks without exploration of the experiences of food bank use and the effects to those using them).Articles exploring other forms of food aid (such as hot food providers) were also excluded, and in cases in which more than one food support provider was mentioned, articles were only included if data regarding food banks could be extracted independently.Additionally, articles which were based within the context of high-income countries more widely were excluded, even if the UK was included in their scope, due to methodological difficulties of separating out UK data.

Study selection
Articles were identified from each database by a single author (NT), duplicates were removed, and the titles and abstracts of the remaining articles were screened against the eligibility criteria.Full texts of articles appearing to meet the eligibility criteria were then obtained and each was assessed for inclusion.An independent researcher verified inclusion of a random sample (25%) of these articles by screening the full texts.Any discrepancies were resolved by discussion.The reference lists of the sample were then examined to identify any additional relevant studies, and a final sample was composed (see Fig. 1).

Data extraction
Content analysis is validated within social science research, providing a useful method through which to conduct rigorous, systematic, theory-driven, and reproducible literature review (Seuring & Gold, 2012).Data extraction took place through directed content analysis methods, as described by Hsieh and Shannon (2005).To conduct directed content analysis, analytic themes are derived from the theory, and act as coding categories during the analysis (Hsieh & Shannon, 2005).As such, directed content analysis allows the relevance of data to be tested against a theory, in order to validate, extend, or undermine the theory (Hsieh & Shannon, 2005).
Within the present study, Weisbrod's (1977) theory of market failure/government failure, Salamon's (1987) voluntary failure theory and Radimer et al.'s (1990) four domains of food insecurity were applied to the literature to explore the extent to which they could contribute towards understanding three important aspects of the experience of food banking respectively: (1) the drivers of food bank use; (2) the limitations of the current food bank model; and (3) the impacts of the food banking model for food bank clients.Table 1 highlights the analytic themes derived from each theory, which were used to deductively explore these three aspects and address the research questions.
Directed content analysis has previously been used by Piaskoski, Reilly, and Gilliland (2020) in their systematic review to explore experiences of household food insecurity within rural populations in Australia, using the 5As of food insecurity (availability, accessibility, acceptability, adequacy, and agency) as analytic categories.By applying directed content analysis to the UK food banking context, it was hoped that new conceptual and theoretical insights could be obtained.

Quality assessment
Due to the inclusion of qualitative, quantitative and mixed methods studies, and the heterogeneity of study designs within the review, the Mixed Methods Appraisal Tool (MMAT) was used to assess the quality of included studies (Hong et al., 2018).Methodological quality was assessed by one author (NT) for each article and a score was given between one and five, five being the highest score for quality (see Appendix 1 and Supplementary Material).Inclusion of studies within the review was not dependent upon methodological quality, and low-quality studies were not excluded.

Descriptive statistics
The search yielded 221 titles (see Fig. 1).Of these titles, 104 were duplicates.After removal of duplicates, the titles and abstracts of the 117 remaining articles were screened for relevance, and 58 titles were excluded.The full texts of the remaining 59 articles were then obtained and screened against the inclusion criteria.A further 25 articles were excluded at this stage.In order of priority, exclusion was a result of non-UK basis (N = 4; of which N = 3 were based in the high-income country context), non-relevance to the drivers, limitations and impacts of food bank use (N = 6), non-food bank basis (N = 6) and non-empirical data/ not peer-reviewed (N = 9).The reference lists of the remaining sample were reviewed and an additional 8 articles meeting the inclusion criteria were identified, giving a final sample of 41 articles, summarised in Appendix 1.
Within the sample, there was a broad and varied representation of journal disciplines.Journals publishing articles in the field of sociology and social policy were the most represented (21.9% of included articles), followed by public health and health care (19.5%), geography (17.1%) and food and nutrition related journals (17.1%).Four articles were published in the Journal of Poverty and Social Justice (9.8%), meanwhile Social Science and Medicine provided two articles (4.9%), and two journals relating to third sector research contributed three articles (7.3%).Social Enterprise was the only journal representing business research, with one article included in the review (2.4%).Across the 41 articles included, a range of methodologies were utilised.The sample included mostly qualitative (61.0%) methodologies, with fewer authors choosing quantitative (26.8%) or mixed (12.2%)methods.The earliest article in the sample was published in 2013 with the number of articles published generally increasing per year.Subsequently, the most common year of publication was 2020 (24.4%).Of those specifying the type of food bank studied, Trussell Trust made up 41.5%, with independent food banks only representing 19.5% of the data, and articles studying both independent and Trussell Trust responses making up 14.6%.Generally, studies were of high methodological quality, with 95.1% scoring above 3 in the MMAT assessment.In this sense, evidence from poorer quality studies (<5%) affected the conclusions of this review minimally.The full results of the quality assessment are available in the Supplementary Material.

Content analysis
The literature was analysed according to the research questions and associated analytic themes (see Table 1).Any additional themes which arose as a result of data analysis are also discussed.

RQ1: to what extent can market failure/government failure theory be applied to understand the drivers of food bank use?
The first research question was explored utilising the analytic themes derived from Weisbrod's (1977) theory of market failure/government failure.Of the 41 studies included for data analysis, 29 addressed this research question (70.7%).Directed content analysis revealed that the reasons apparent in the literature for individuals experiencing food insecurity turning to charity could be clearly categorised into market and government failures, with 100% of these 29 studies pointing to either one or both failures in driving food bank use.

Market failure.
The role of market failures in driving the need for food banks largely affected those eligible to work, and was ascribed to low and stagnant income (44.8% of the 29 studies addressing this research question), unstable or precarious working conditions (31.0%), and sudden or unexpected job loss and reliance on zero-hour contracts (13.8% respectively).Unemployment was also reported to drive food bank use in those eligible to work (17.2%).These factors were also suggested to be compounded by rising living costs, including fuel, housing, and food (17.2%).

Government failure.
Meanwhile, it was clear that government failure lies in austerity welfare reforms, largely affecting the long-term unemployed or those unable to work due to illness or disability.Issues reported to directly initiate the need to access a food bank included benefit sanctions (48.3%), delays (31.0%) and changes (37.9%), insufficient benefits (20.7%), administration problems associated with receipt of benefits (N = 3), removal of the spare room subsidy (17.2%), and the introduction of Universal Credit (17.2%), a payment combining six previous benefits and tax credits which is paid to those on low income, unemployed, or ineligible to work (Gov.UK, 2023).

Additional themes.
In addition to market and government failures, the role of challenging personal circumstances (such as ill health (including mental health), domestic violence, homelessness, debt, and release from prison) and environmental factors including distance to and availability of budget shops in driving food bank use were discussed (17.2% and 3.4% respectively).The role of individual choices and behaviours was also mentioned (31.0%).For example, in studies by Power et al. (2020) and Power and Small (2022), both food insecurity service providers and participants from low income households who had not experienced hunger respectively reported beliefs that food bank use was partly driven by personal failings including poor financial management.Meanwhile Parr, Hawkins, and Dayson (2021) reported data from food bank and Trussell Trust managers, who believed people with food insecurity often possess the power to change their situation, yet choose not to use it.

RQ2: to what extent can voluntary failure theory be applied to understand the limitations of the current food bank model?
Within the sample, 29 studies (70.7%) highlighted limitations of the current food banking model.This research question was explored through the lens of Salamon's (1987) voluntary failure theory, with the four limitations of the charitable sector (insufficiency, particularism, paternalism, and amateurism) used as analytic themes.Paternalism, particularism and insufficiency were well represented in the included studies, with some debate surrounding the application of amateurism in this context.

Paternalism.
Paternalism describes the inclination of the voluntary sector to place the needs of those it serves in the hands those more fortunate.The paternalistic nature of food banks was the most widely reported of Salamon's limitations (65.5% of the 29 studies addressing this research question), manifesting through mechanisms of reliance: reliance on referral agencies; reliance on volunteers; and reliance on donations.
Within studies describing paternalism, reliance on referral agencies to issue vouchers represented a key paternalistic feature of food banks within the Trussell Trust model specifically, represented in 14 studies (48.3% of studies addressing this research question).Within such studies, the system of judging deservingness to receive a voucher is described as subjective, inconsistent, and responsible for creating a hierarchy of worthiness rated by those of a higher socioeconomic and morally distanced position (37.9%).One study also mentioned a reliance on the referral agents to complete the voucher correctly, and in a timely manner (3.4%) (May, Williams, Cloke, & Cherry, 2019), making access to food dependent upon the capacity of others.Meanwhile, rationing of food bank vouchers to three per year represented a paternalistic feature of UK food banks which hailed from reliance on referral agencies (17.2%), with studies reporting that crisis was only 'allowed' on a certain number of occasions, despite factors often beyond the clients' control persistently influencing their ability to afford food.
Reliance on food bank volunteers and their capacity to distribute food characterises another paternalistic feature of the food bank model (20.7%).Receipt of food was reported to depend upon the schedules of volunteers and the time that they could dedicate to volunteering, resulting in power relations between food bank volunteers and clients.
The final mechanism of reliance was characterised by reliance on donations (17.2%).The issues associated with this were reported to lie in the lack of agency of food bank clients -exemplified by the existence of a prescribed checklist provided by Trussell Trust food banks which dictates accepted foods, such as non-perishable staples including pasta and rice, tinned goods, and biscuits (May, Williams, Cloke, & Cherry, 2020a;Trussell Trust, 2023) -and transfer of choice into the hands of those with more power and resources (Beck & Gwilym, 2020;May et al., 2020a).There was also evidence of preference to donate food over money through fear that money will be spent on other commodities such as alcohol and drugs (Garthwaite, 2016b).Channelling of donations through supermarkets was also reported as an issue, being reliant upon the relationship between supermarkets and food banks (Iafrati, 2018;Surman, Kelemen, & Rumens, 2021).

Particularism.
The concept of particularism describes the voluntary sector's focus on certain subgroups of the population and marginalisation of others, and was clearly exemplified within 16 of the 29 studies addressing this research question (55.2%).Particularism was enacted through various routes, including the inability of food banks to meet the various socioeconomic, practical, cultural, religious and medical needs of a broad spectrum of individuals.Commonly, studies highlighted more than one type of particularism.
Particularism was most commonly enacted through socioeconomic routes (37.9%).Judgement of eligibility to receive food by referring agencies appeared to be based on socioeconomic position, excluding those who are struggling but not yet severely food insecure enough to warrant access.In addition (May et al., 2019), reported the referral system to be confusing and particularistic itself, with only those who are sufficiently well-informed being able to navigate the system and prove their need.
Practical particularism -relating to the practical difficulties that some subgroups of the population with food insecurity might face in accessing food from a food bankwas also a common theme (34.5%).Issues with transport to and from the food bank, and carrying received food home were quoted in six studies, encapsulating the majority of these practical difficulties, especially in cases of limited mobility and/or health issues (Lawson & Kearns, 2021;Puddephatt et al., 2020) or living in a rural community (May et al., 2020b).Additionally, operational characteristics such as restricted numbers of food banks or referral agencies, and impractical opening times were reported to impact food bank access (17.2%).One article also highlighted that food received from a food bank did not meet the practical needs of the clients due to discrepancies between cooking instructions and cooking facilities in their homes (3.4%) (Thompson, Smith, & Cummins, 2018).
Cultural aspects were revealed within 20.7% of the 29 studies addressing this research question.These aspects included a shortage of culturally acceptable foods within food banks, due in part to donations largely being shaped by white, middle-class ideals of necessary items (10.3%) (Power, Doherty, Small, Teasdale, & Pickett, 2017;Power et al., 2020;Williams, Cloke, May, & Goodwin, 2016).Similarly, overrepresentation of white individuals both accessing and volunteering at food banks was suggested to prevent those of other cultures knowing of, and feeling represented or welcome, in these spaces (6.9%) (Power, Doherty, et al., 2017;Williams et al., 2016).Meanwhile, Thompson et al. (2018) suggest that the inability to access food suited to cultural preferences also leads to food waste, due to food items being 'unfamiliar', 'unpalatable' or 'culturally inappropriate'.
Religious particularism was also represented by 13.8% of studies addressing this research question.While two studies (6.9%) by Power, Doherty, et al. (2017) and Power, Small, Stewart-Knox, Pickett, and Doherty (2017) found that food banks did not actually promote messages of faith, Trussell Trust food banks in particular were reflected as being underpinned by Christian morals and values, and providing those following Christian faith an opportunity to express charitable kindness, which may deter those of other faiths from accessing these services, (Denning, 2021;Williams et al., 2016).However, Power, Small et al. (2017) also highlighted that religious basing may be a benefit to food banks, as religious teachings might initiate more of a welcoming and caring atmosphere, making volunteers more committed to their role.
The medical needs of certain individuals may not be met through food accessed at food banks, representing another manifestation of particularism within UK food banks.For example, 13.8% of studies highlighted that those with specific dietary needs as a result of health conditions, including diabetes (Douglas et al., 2015;Thompson et al., 2018), and allergies or food intolerances (Garthwaite, Collins, & Bambra, 2015) may struggle to access appropriate foods at food banks.In such cases, symptoms caused by the food available may persist, exacerbating health problems.Additionally, one study (3.4%) reported food bank clients skipping medication that must be taken with food due to food shortage (Douglas, Maciver, & Yuill, 2020).
Solidifying these particularistic tendencies, only 6.9% of studies provide evidence of food banks attempting to accommodate the dietary needs of their clients (Douglas et al., 2020;Surman et al., 2021).Both of these studies provide this evidence in the context of independent or community-run food banks, with no evidence of such accommodation in Trussell Trust food banks within the sample.
3.2.2.3.Insufficiency.Insufficiency relates to the inability of charities to provide a steady and adequate supply of resources (Salamon, 1987), in this case, a sufficient quantity and quality of food at a stable rate.Insufficiency was represented in 15 articles addressing this research question (51.7%), through three routes: insufficient quantity, insufficient quality, and unevenness in supply.
Insufficient quality of foods appeared to be the biggest issue (44.8%), encompassing both nutritional and actual quality.Nutritional quality was referred to through high proportions of processed and nonperishable foods, items high in sugar and salt, low proportions of fresh fruit and vegetables and low micronutrient content (37.9%).Two studies (6.9%) discussed the actual quality of food received, relating to food being close to, or even past the best before date when distributed.
Insufficient quantity of food provided was reported in six studies (20.7%), with four of these articles being qualitative in nature.Yet, within the two quantitative studies, the results were contradicting, with one study indicating that the food supplied in a food bank parcel was sufficient in energy and macronutrients for three days (Preston & Burley, 2015), and the other stating that the food did not meet daily requirements for energy (Turnbull & Bhakta, 2016).
Meanwhile, only 13.8% of the studies describing insufficiency reported an unevenness in supply.This was reported to relate to instability in the food items received, a large dissociation between client needs and the content of food parcels, and temporally uneven donations.For example, Iafrati (2018) highlighted how links between food banks and faith-based organisations result in a peak in donations around September which relates to the Harvest Festival season, but does not coincide with peaks in demand during school holidays.
3.2.2.4.Amateurism.Amateurism is defined as the reliance on amateur approaches and volunteers, in place of trained professionals.Nine studies referred to amateuristic tendencies (31.0% of the 29 studies addressing this research question), and reported contradictory findings.Iafrati (2018) referred to local authority budget cuts which have required food bank volunteers to take on additional roles such as assessing need and signposting, which they may not be qualified to do.In addition, May et al. (2019) highlighted that outsourcing of decisions of worthiness to other untrained professionals may occur.The same study also highlighted a social disassociation between food bank clients and volunteers which could affect their understanding of food bank clients' situations, leading to amateur approaches (May et al., 2019).However, both these studies were reported as lower quality (MMAT scores of two and three respectively) within this review and so declarations of amateurism should be made with caution.Directly contradicting these findings, Power, Small et al. (2017) noted that volunteers are often not responsible for decisions of worthiness, with their role much more focused on welcoming clients and making them feel comfortable.Meanwhile, in three other studies, volunteers had often experienced food insecurity themselves, making them 'professionals' through lived experience in understanding the struggles of those accessing the food bank (Lawson & Kearns, 2021;Power & Small, 2022;Strong, 2020).Food bank volunteers were also reported as uneducated as to the nutritional quality of food parcels in one study (Garthwaite et al., 2015), despite Thompson et al. (2018) finding food bank volunteers to be working with nutritionists to promote health within food parcels.

Additional themes.
Additional limitations of the current food banking model were also highlighted in the literature which are not encapsulated by voluntary failure theory.Problematic public and media misconceptions of food bank clients were discussed (27.6%).For example, Garthwaite (2016b) reported that food bank clients are often misrepresented as lazy people who lack skills and make poor choices within the mainstream media (e.g. on TV programmes such as 'Benefits Street'), creating damaging stereotypes of individual blame which stick despite often not reflecting reality.This relates closely to the additional themes of personal choices and behaviours in RQ1.The deep ingraining of these stereotypes within public opinions may has cause them to become tightly interwoven within the food bank model and represent a significant limitation for food bank clients.
As a result of both expected or actually experienced limitations of food bank use, the use of coping strategies instead of (as well as alongside) accessing food banks was highlighted (34.5%).Tactics such as cooking in bulk, minimising food waste, stretching food out, meal planning and sacrificing food for children were all reported (Douglas et al., 2015;Power et al., 2018Power et al., , 2020) ) in order to mitigate from the insufficiencies and particularism which lead to inadequate access from food banks, or providing a buffering effect from the need to access food aid at all.Additionally, social circles appeared to be commonly relied upon to access food in place of food banks (Osei-Kwasi, Nicolaou, Powell, & Holdsworth, 2019; Power, Doherty, et al., 2017;Power et al., 2018;Puddephatt et al., 2020), yet this effect appeared to be culturally mediated, with a small minority of individuals (mostly white British) preferring to access a food bank than declare need to family (Purdam, Esmail, & Garratt, 2019;Strong, 2020).
Meanwhile, the inability of the current food banking model to address the root causes of food insecurity was indicated (6.9%), undermining structural change and placing reliance on charity, which may not be able to keep up with this demand (Parr et al., 2021;Puddephatt et al., 2020).Parr et al. (2021) described food banks as merely a sticky plaster, plugging the immediate crisis but with the complete inability to address the driving forces of food insecurity, nor act as an adequate solution to food insecurity.
Despite the limitations highlighted in the literature, some studies mentioned benefits to food bank use.Within 37.9% of studies addressing this research question, food bank clients and service providers alike reported the service as a lifeline, a key source of basic food which may otherwise be lawfully unobtainable, and allowing some breathing space for those suffering from the dramatic effects of austerity.Gratitude, hope, and reassurance were all mentioned to be implicit within the food bank model (10.3%) (Denning, 2021;Douglas et al., 2015;Purdam et al., 2016).Additionally, the provision of extra services, including cooking classes and signposting were described as beneficial to food bank clients, aiding their progression back into self-sufficiency (17.2%).

RQ3: to what extent can Radimer et al.'s four domains of food insecurity be applied to understand the impacts of food bank use?
The final research question was explored with consideration of Radimer et al.'s (1990) conceptual domains of food insecurity, with quantity, quality, psychological, and social aspects as analytic themes.The impacts for individuals accessing food banks appeared to fit well into these domains, even over three decades after their conceptualisation.Again, 29 of the 41 included studies (70.7%) addressed this research question.

Quantity and quality.
As the quantity and quality of the food available at food banks has a direct impact upon the recipient's physical health, these two components were combined into one componentphysical health -during analysis.Eight of the 29 studies (27.6%) provided evidence regarding the physical health implications of accessing food from food banks, which are closely linked to issues of insufficiency and particularism discussed above.Within these studies, the exacerbation of existing health conditions including irritable bowel syndrome, and poor self-rated health is mentioned (10.3%), having significant negative consequences to overall wellbeing (Garthwaite et al., 2015;Loh et al., 2021;Puddephatt et al., 2020).The nutritional composition of food parcels was discussed as negatively impacting health (10.3%) (Fallaize, Newlove, White, & Lovegrove, 2020; Forde & Solomon-Moore, 2019; Preston & Burley, 2015).There was also evidence of a proportion of the population with food insecurity going hungry rather than accessing emergency food from food banks (6.9%), having significant physical health impacts (MacLeod et al., 2019;May et al., 2020a).

Psychological.
Meanwhile, 24 studies addressing this research question (82.8%) provided consistent evidence for the negative psychological implications associated with food bank use.Negative emotions experienced as a result of food bank use were cited (58.6%), most often embarrassment (34.5%) and shame (41.2%), but also humiliation (Lambie-Mumford, 2013;Power et al., 2018), degradation (Lambie-Mumford, 2013; Lawson & Kearns, 2021), powerlessness and desperation (Douglas et al., 2015).Stigma associated with food bank use was explicitly discussed in thirteen studies (44.8% of studies addressing this research question).Meanwhile, low mood was reported as a result of the lack of choice available to food bank clients (Douglas et al., 2020), and continual uncertainty surrounding access to food manifested as anxiety and stress (Puddephatt et al., 2020;Williams et al., 2016).Anticipation of these negative emotions before first access, coupled with emotional and psychological damage as a result of access, was also mentioned in one study (Lawson & Kearns, 2021).

Social.
Social impacts of accessing food from food banks were represented in 51.7% of studies addressing this research question.However, only 20.7% concerned negative social impacts.These negative impacts included fear of being seen or judged for accessing a food bank (Lawson & Kearns, 2021;May et al., 2020b;Osei-Kwasi et al., 2019) and subsequent marginalisation and detachment from society (May et al., 2020b;Purdam et al., 2016;Surman et al., 2021).Only two of the studies which referenced negative social impacts did not allude to any countering positive impacts (6.9%) (Osei-Kwasi et al., 2019;Purdam et al., 2016).
In contrast with the negative physical health and psychological impacts, almost all studies reporting social impacts indicated one or more positive social outcomes (86.7%) potentially aiding in alleviating some of the psychological impacts of food insecurity and food bank access.Of these, 46.2% reported that accessing a food bank could ease expected shame and stigma feelings due to the welcoming, relaxing and nonjudgemental atmosphere created by volunteers.In particular, food banks were reported to provide a 'listening ear' to clients (Garthwaite et al., 2015;Lawson & Kearns, 2021).Additionally, 38.5% referred to the benefits of the sense of community and sharing of experiences, leading to solidarity and social connectedness which helped to relieve some of the psychological implications.

Discussion
This systematic review and directed content analysis aimed to explore important aspects of the food banking experience through a theoretical lens.Firstly, Weisbrod's (1977) market/government failure theory was applied to explore the drivers of food bank use.Then, Salamon's (1987) voluntary failure theory was applied to examine the limitations of the current food bank model.Finally, Radimer et al.'s (1990) four domains of food insecurity were applied to explore the impacts of the food banking model for food bank clients.Applying these three theories to the literature through directed content analysis methods allowed the relevance of the theories to the food banking context to be tested (Hsieh & Shannon, 2005), providing new conceptual and theoretical insights on the experience of food banking in the UK.
Resulting from the content analysis, a conceptual map of UK food bank use according to the three theoretical perspectives has been developed (Fig. 2).Fig. 2 encapsulates and visually summarises the key drivers, limitations and impacts of food bank use elucidated within this review through the theoretical lenses applied, enabling a holistic, up-todate conceptualisation of the experience of food banking from drivers of use, to limitations of access, to impacts on wellbeing.It is important to note that this diagram is by no means exhaustive, and there are likely many other (often elusive) predictors of, and barriers to, food bank use not yet captured by the extant literature.

The application of market/government failure theory to explore the drivers of food bank use
Both market and government failures were overwhelmingly presented as driving food bank use within included studies, thus validating the relevance of market/government failure theory in the context of food banking in the UK today.To our knowledge, the current review is the first to apply this theory to the UK food banking context, however, market failure theory has been individually applied in the context of food insecurity by Rocha (2007).Rocha revealed inherent flaws of the current food system within North America, allowing the defining of food insecurity as a form of market failure.This, in turn, enabled the identification of specific government responses to counter failings within the market.Within the context of food banking, describing market (and government) failures as drivers of food bank use allows a deeper understanding of the structural problems at play, and the subsequent solutions and government commitment required to tackle this issue, including more commitment to the provision of the living wage, more stable working conditions, and increased employment within the labour market.
In particular, results from this review highlight the significant role of inadequate welfare provision in driving food bank use.This has been mirrored in recent research conducted by the Trussell Trust (2022c), which directly attributes severe food insecurity and food bank access to failures of the government.For example, Trussell Trust data indicate that one in three of those living on Universal Credit cannot afford to buy sufficient food (Trussell Trust, 2022c), highlighting that the current social security system is inadequate to support those living in poverty, with significant changes required in order to ensure access to food as a basic necessity.Additionally, challenging personal and environmental factors were highlighted in this review as accelerating the journey to food bank use.The Trussell Trust (2022a) also provide evidence for this, reporting that low income due to accessing social security can be compounded by factors such as ill-health, lack of social support, and challenging life experiences.This highlights the presence of a complex web of drivers to food bank use which must be addressed in order to mitigate the effects of the current cost-of-living crisis (see Fig. 2).

The application of voluntary failure theory to explore the limitations of the current food bank model
With regards to the limitations of the current food bank model, studies included in this review highlighted strong evidence of insufficiency, paternalism, and particularism, in line with Salamon's (1987) theory.Meanwhile, the relevance of the amateurism domain of voluntary failure theory was less clear when applied to the context of food banking in the UK.Jones and Deitrick (2020) have previously examined the relevance of voluntary failure theory within food pantries in the US, which operate under the same model as food banks in the UK.The authors implemented a quantitative survey among service providers and found medium to strong evidence of failures in the four failure categories: insufficiency, paternalism, particularism, and amateurism.The present review corroborates their findings relating to particularism, paternalism and insufficiency, providing strong evidence of these failures in the UK food bank setting.These failures may have been more strongly evidenced within the present study due to the majority of participants within included studies being food bank clients, who feel such limitations more directly and personally than service providers, as surveyed by Jones and Deitrick (2020).Additionally, only a limited depth of understanding can be obtained from survey methodology.Over half of the studies included in the current review used qualitative interview and observation methodologies, being much more able to elucidate the depth and richness of the lived experience of food insecurity and the broad range of associated struggles (Sofaer, 1999).
However, amateurism was less well represented, and data regarding its existence and impact within the food banking model remain equivocal.Within other charitable contexts, for example the voluntary sport sector, amateurism is reported through the presence of undertrained staff, affecting working relationships with trained staff due to competition and judgement of capability (Misener & Misener, 2017).Yet within the food banking context, education and experience of volunteers was highly variable within included studies.It is likely that the heterogeneity of food banks within the UK has contributed towards this inconsistency, with different food banks having different training schemes for volunteers.Standardising training and education for food bank volunteers may reduce potential amateurism within the food bank model, yet this raises concerns regarding increased formalisation and bureaucracy within food banks, further institutionalising access, and may act as a backwards step for clients.
In addition to Salamon's failures, two additional themes arose from the literature review: (1) public misconceptions and institutionalised negative stereotypes of food bank users (and the subsequent implementation of coping strategies), and (2) the inability of the food bank model to address the root causes of food insecurity, thus acting as an adequate solution.Firstly, public misconceptions of food bank users as lazy and lacking in skills were highlighted, with media coverage of food banking reported to be a key driver of such discourses (Garthwaite, 2016b;Lawson & Kearns, 2021).Ross, Morgan, Jorm, and Reavley (2019) conducted a systematic review which concluded that negative media portrayals regarding individuals living with mental illness have Fig. 2. A conceptual map of food bank use in the UK using a theoretical lens *RQ1: Findings based on the application of market/government failure theory (Weisbrod, 1977) and additional themes highlighted within the review.Δ RQ2: Findings based on the application of voluntary failure theory (Salamon, 1987), with adaptations from Poppendieck (1999) and McIntyre et al. (2016), and additional themes highlighted within the review.ꓳ RQ3: Findings based on the application of Radimer et al.'s (1990) four domains of food insecurity.
the potential to influence public opinions, being linked to increases in stigmatising attitudes towards people living with mental illness.Similarly, applying this to the food banking context, negative media coverage of food bank users may play a strong role in influencing public opinions of food banking, and increasing society-wide stigmatisation of food bank users.Additionally, evidence synthesised in this review highlighted that even people working and volunteering within the food banking system may hold stigmatising attitudes towards food bank users (Parr et al., 2021;Power et al., 2020), the institutionalisation and ingraining of stigma, even within the food bank model itself.
The institutionalisation of stigma is problematic, as it is likely to be a major contributor to reluctance and resistance to access the service (Davis & Geiger, 2017;Douglas et al., 2015;Middleton, Mehta, McNaughton, & Booth, 2018), having further negative consequences, such as the manifestation of hidden food insecurity.For example, as a result of reluctance to access food banks, studies included in this review highlighted the implementation of coping strategies such as cooking in bulk, careful meal planning, minimising food waste and stretching out or sacrificing food to manage the symptoms of food insecurity and buffer the effects of inadequate food provision.However, coping strategies may not be relevant or sufficient to mitigate the effects of severe food insecurity.For example, consuming cheaper alternatives to popular food brands and shopping in budget supermarkets have been described as coping strategies within individuals experiencing food insecurity in high income countries (Puddephatt et al., 2020;Stone et al., 2024;Van Der Velde, Schuilenburg, Thrivikraman, Numans, & Kiefte-De Jong, 2019;Zigmont, Linsmeier, & Gallup, 2021), yet for those already consuming the cheapest options and still experiencing severe food insecurity, this is not a viable solution (Dowler & Lambie-Mumford, 2015).Therefore, the utilisation of coping strategies might be ineffective, resulting in increased hunger and malnutrition within individuals living with hidden food insecurity (Zepeda, 2018).Instead, alternatives to traditional food banks are urgently required which are independent from food banks and their negative public image, promoted positively within the public domain, and able to adequately support and empower individuals experiencing food insecurity to access sufficient nutritious food.
Additionally, the food bank model was reported to only provide a buffer to hunger, not able to address any of the structural causes of food insecurity or drivers of food banking.While positive impacts of the charitable sector may exist (such as the positive social impacts shown in this review), ultimately, charity cannot prevent the manifestation of food insecurity, driven by a lack of income from an ineffective welfare system and labour market (Smith-Carrier, 2020).Just as severe food insecurity and food bank access have structural drivers, to address the root causes requires structural solutions, going further than philanthropic giving and instead influencing policy and the food system as a whole, in order to reduce food insecurity.
Referring back to Poppendieck's (1999) conceptualisation of the limitations of the food bank model, indignity was highlighted as one of the seven key limitations, representing the experience of stigma and other negative affective reactions when accessing a food bank.Meanwhile, McIntyre, Tougas, Rondeau, and Mah (2016) aimed to update and extend Poppendieck's (1999) seven key limitations to account for more than just the operational characteristics of food banks, identifying an additional five emerging limitations in the contemporary literature.One such limitation was ineffectiveness, representing doubt over whether the food banks are capable of meeting their goal of reducing food insecurity (McIntyre et al., 2016).Through the application of Salamon's (1987) voluntary failure theory, evidence from this review has highlighted that indignity and ineffectiveness are still operative within food banks in the UK, representing more important limitations than amateurism.Therefore, adapting Salamon's (1987) voluntary failure theory to replace amateurism with both Poppendieck's (1999) indignity and McIntyre's (2016) ineffectiveness may improve the applicability of the theory to the UK food bank context (see Fig. 2).Future research might benefit from exploring this adaptation, particularly within other high-income countries.

The application of Radimer et al.'s four domains to explore the impacts of food bank use
Findings from the included studies regarding impacts of food bank use were well-aligned with Radimer et al.'s (1990) four domains of food insecurity; quality, quantity, psychological and social.Both the physical health impacts (made up of quality and quantity aspects) and psychological impacts were overwhelmingly negative with the latter appearing to be particularly pertinent, largely due to the experiences of shame and stigma.In another UK-based quantitative study not included in this review, Baumberg (2016) found that one third of welfare claimants experienced some level of stigma, either as a result of their own personal beliefs, or felt from others.Meanwhile, findings from RQ1 highlight the important role of inadequate welfare in driving food bank use.Taken together, this suggests that stigma experienced as a result of requiring access to welfare payments may be further compounded by stigma experiences when receiving food from food banks.Markowitz (1998) highlighted that stigma experiences have been linked to adverse effects on psychological wellbeing and life satisfaction, in particular, depression and anxiety.
More generally, experiencing food insecurity has been associated with poorer mental health outcomes in high-income countries (Afulani et al., 2020;Leung, Epel, Ritchie, Crawford, & Laraia, 2014;Maynard et al., 2018;Pryor et al., 2016).Additionally, in a study by Spinosa, Christiansen, Dickson, Lorenzetti, and Hardman (2019), lower socioeconomic status was linked to higher psychological distress, which in turn was associated with higher emotional eating and a subsequent higher BMI.From synthesising the literature within this review, it is clear that negative psychological impacts of accessing a food bank are highly prevalent and may compound existing negative psychological impacts of food insecurity, and the subsequent physical health impacts.This is likely to lead to further reluctance to access support from food banks, leading to hidden food insecurity, increased hunger and further impacting on physical health (Zepeda, 2018).This relationship between negative physical health and psychological impacts of food bank use emphasises the urgent need for more psychologically supportive solutions for individuals experiencing food insecurity, to promote not only psychological but also physical wellbeing.
Notwithstanding the above, the review findings suggest that accessing food bank services may provide positive social impacts by enabling individuals to become more socially connected to others in a similar position.In a recent study by Rotenberg, Surman, and McGrath (2021), food insecurity was related to loneliness, particularly for those in middle adulthood and the unemployed.Taken that individuals using food banks are likely to be experiencing severe food insecurity, it is likely that they will also be at risk of loneliness.Elsewhere loneliness has been shown to be associated with poorer health outcomes, including mental health, physical health, and general wellbeing (Park et al., 2020).Therefore, improved social connectedness through food bank access may work to reduce loneliness, having not only social benefits, but also improving physical and psychological health outcomes.As such, while the negative psychological impacts of food bank use may further exacerbate negative physical health impacts, positive social impacts may help to alleviate some of the negative psychological and physical health impacts.However, social benefits may be highly dependent upon the operational characteristics of the food bank.For example, Lawson and Kearns (2021) highlight the benefits of a social 'café style' model where there is opportunity for social interaction.Meanwhile, Rotenberg et al. (2021) found improvements to loneliness within an environment focused around communal food eating and sharing.In both of these examples, the operational format is different from the typical food bank model in the UK, in which food parcels are often received and taken away to consume at the home.There is currently limited UK-based research on the social impacts of food bank use, yet evidence from this review highlights that relationships between physical health, psychological and social impacts do exist (represented by the arrows in Fig. 2).Further exploration of these relationships would be valuable in developing more appropriate, supportive and holistic alternatives to the traditional food bank model.

Strengths and limitations
Strengths of this systematic review include the application of novel theories to improve understanding of three important aspects of the UK food bank experience: the drivers of food bank use, the limitations of the current food banking model, and the impacts of food bank use for food bank clients.However, there are also some methodological limitations.Firstly, grey literature, such as government reports, were not examined within this review due to time constraints but may have provided valuable additional insight.Similarly, including 'UK' in the search terms may have restricted the breadth of titles discovered by excluding those that specified a particular country within the UK (England, Scotland, Wales etc).Finally, the results of this review, the applicability of the theories to the food banking context, and any recommendations are specific to the UK context.It would be presumptuous to generalise the results of this study to other high-income countries, due to the heterogeneity of operational characteristics of food banks between countries.

Future recommendations
The current findings highlight the importance of upstream approaches to addressing food insecurity.Eradicating the need for food support is the most critical challenge which must be addressed, and this relies on government commitment to ensuring not only an adequate and affordable supply of food for all people, but the financial means to be able to purchase nutritious food, regardless of situation.The National Food Strategy (Dimbleby, 2021) outlined key government recommendations for aiding the most disadvantaged households in the UK to eat well and reduce inequalities.These recommendations include expanding the financial eligibility for free school meals, and of the Healthy Start scheme -a UK voucher scheme enabling low-income pregnant mothers and young families access to fruit, vegetables, vitamins and milk -using revenue obtained from implementing sugar and salt reformulation tax for large food businesses.Bringing these recommendations to fruition would likely increase access to healthy foods for more families experiencing food insecurity, counteracting some of the downstream effects of market and government failures, and potentially reducing the need to access food support.
Yet, while upstream change requires upstream commitment in the long term, individuals experiencing food insecurity require action now.Therefore, priorities for future research might also focus around developing novel food support models which help to minimise the insufficiencies, paternalistic and particularistic tendencies, and negative physical health and psychological impacts of current food bank services.Community-based initiatives such as community kitchens have been suggested as an alternative to food banks.Evidence from a systematic review conducted by Iacovou, Pattieson, Truby, and Palermo (2013) indicated that community kitchens may be well-placed to meet the nutritional, social and psychological needs of those with food insecurity, whilst maintaining dignity, and promoting the development of social and practical skills through capacity building.However there are also logistical, operational, and financial challenges associated with such operations, and shifting to a community kitchen model would require significant investment if it were to be on a national scale (Engler-Stringer & Berenbaum, 2007;Loopstra, 2018a;May et al., 2020a).Additionally, similar to food banks, community projects often rely on a dedicated volunteer workforce and stable supply of food.
There is also potential for social business models to address the limitations and negative impacts of the food bank model, presenting a solution to food insecurity which is itself financially stable.Social businesses exist fundamentally as community-based organisations (Pearce, 2003;Roy, Donaldson, Baker, & Kerr, 2014) and have the capacity to broadly influence social, economic, and environmental challenges within communities by blending the activities of the third and trading sectors (De Leeuw, 1999;Kay, Roy, & Donaldson, 2016;Roy, Baker, & Kerr, 2017;Roy et al., 2014).Through this, social business models may present opportunities for individuals experiencing food insecurity to reintegrate back into the market, thus restoring dignity and minimising shame relating to food bank access (Power et al., 2018;Power et al., 2020;Vlaholias, Thompson, Every, & Dawson).Social businesses can also create employment opportunities within communities thereby reducing reliance on volunteers which in turn, works to address root causes of food insecurity within communities.Lindberg et al. (2018) provide preliminary evidence of positive results from the implementation of social businesses to tackle food insecurity in Australia, and evidence from other high-income countries, such as the UK is now needed.

Conclusion
This systematic literature review and directed content analysis has taken a theory-driven approach to synthesize research into food banking in the UK.Included studies suggest that market and government failures drive food bank use in the UK, with the current food banking model having significant limitations (insufficiency, particularism, paternalism, indignity, and ineffectiveness) and damaging physical health and psychological impacts.Meanwhile, food bank access may provide some positive social impacts for clients.
The environment in which this research is situated is highly fluid and interwoven with wider political, social, economic and environmental factors, with the current cost-of-living crisis making it particularly timely.The findings from this review and the resultant conceptual map allow a holistic and current visualisation of experiences of food bank use in the UK through the lenses of market/government failure theory, voluntary failure theory and Radimer et al.'s four domains of food insecurity.This can provide a framework for targeting specific areas for improvement and hopefully make small steps towards addressing the intractable problem of food insecurity in the UK.
. To what extent can market failure/government failure theory be applied to understand the drivers of food bank use? 2. To what extent can voluntary failure theory be applied to understand the limitations of the current food bank model?3. To what extent can Radimer et al.'s four domains of food insecurity be applied to understand the impacts of food bank use?

Table 1
Analytic themes derived from the respective theories.