Introduction
In 2022, the global number of forcibly displaced people reached a record number of 100 million (United Nations, 2022). While most displaced individuals reside in developing countries (World Bank, 2024), immigration remains a significant topic in contemporary European public discourse and politics. The year 2015 marked a peak in the arrival of refugees and migrants (hereafter RMs) to the European Union (EU), the highest since World War II (International Organization for Migration, 2015). This event is often considered as a turning point in the so-called “refugee crisis” (Spindler, 2015). It has certainly exacerbated the EU’s increasingly restrictive approach to immigration, characterised by policies of externalisation, deterrence and detention1 (Wagner, 2015; Guiraudon, 2018; Majcher et al., 2020). Organisations such as Human Rights Watch (HRW) and Médecins Sans Frontières (MSF) have highlighted how these policies directly and detrimentally affect RMs, both in their attempts to reach EU soil and after their arrival, endangering their survival, health and well-being2 (HRW, 2018; Benvenuti et al., 2024).
Greece is one of the EU’s key entry points for RMs and thus carries substantial responsibility for their initial and increasingly prolonged reception3 (International Rescue Committee, n.d.). Unfortunately, the living conditions for RMs in the country have been precarious for years, manifested in various and complex forms of destitution (Greek Council of Refugees [GCR], 2022). This paper focuses on one of the most fundamental forms of destitution: the deprivation of food.
Background
The Greek context
In October 2021, 33 humanitarian aid organisations wrote a joint letter to the Greek government, warning of the increasing disruption of food access in the country and its especially severe consequences for RMs (Psaropoulos, 2021). The letter was prompted by a policy that took effect in the previous month, which limited government catering in Greek refugee camps to individuals with legal asylum status. The organisations predicted a hunger crisis threatening those who fall out of this category, such as individuals whose asylum applications had been recently granted or rejected.
As of January 2022, the International Rescue Committee (IRC) reported that about 40% of residents in Greek refugee camps did not have access to food. Minors were highly affected; teachers reported that primary school children often went without food throughout the school day (Smith, 2022). However, food access disruption does not only affect residents of refugee camps. According to the humanitarian organisation INTERSOS Hellas (2022), over 5,000 members of RM populations in Athens alone do not have constant access to food and are at daily risk of survival.
The existing data on Greece’s overall food security underscores the urgency of the issue. According to a 2020 survey, 96.9% of households were reported as food secure (Hellenic Statistical Authority, 2021); however, food insecurity has been rising consistently over the last ten years (Ekathimerini, 2021). Moreover, food inflation is on the rise (World Bank, 2022) and living costs have surged due to recent global events such as the Covid-19 pandemic and the war in Ukraine, which are predicted to exacerbate poverty and food insecurity among Greek residents, and especially marginalised groups (Sipsas, 2022). In this context, a lack of initiative at the policy level is concerning. As Dimitra Kalogeropolou, Director of the IRC Greece, states, “It ought to be untenable that people are being left without food in Greece, a country with the resources and the means to provide food and safety to everyone” (IRC, 2022, para. 4).
While humanitarian initiatives highlight the extent to which RMs are affected, there is a lack of conclusive data on this issue. For example, the aforementioned statistics on food security include households from all socio-economic backgrounds but exclude those who are homeless, undocumented, or live in refugee camps (Hellenic Statistical Authority, 2022).
Furthermore, RMs’ legal status plays a significant role in their experiences. An individual’s classification, whether as an asylum seeker, recognised refugee, migrant with a residence permit or undocumented migrant, determines the applicability of different policies and ministries responsible for them. In this study, the term “refugees and migrants” (RMs) has been selected to capture this variety of classifications and their associated implications. However, it is important to note that this term can be problematic, as it suggests a simple distinction between refugees and migrants which in reality, may be arbitrary and rooted in systemic injustices. Moreover, in humanitarian contexts, the term ‘migrants’ often excludes those individuals commonly called ‘expats’, who typically have legal entry and residency usually based on EU passports or high-skilled job contracts. While a detailed discussion of the terminology is beyond the scope of this work, it is essential to note that RMs are understood to include those who have been forcibly displaced and are in need of protection, regardless of their varying legal statuses.
Food Access in Greece
In its most common definition, food access, one of the four components of food security, is defined as “access by individuals to adequate resources (entitlements) for acquiring appropriate foods for a nutritious diet.” (Food and Agricultural Organisation of the United Nations, 2006, p.1). Consequently, food access disruption is understood as the opposite condition, namely, persistent lack of sufficient access to food.
Recent studies have highlighted the issue of food access disruption in Greece. The most comprehensive study is provided by Konstantinidis (2022) who analyses food security among Greek households in relation to the economic crisis between 2009 and 2015. He identifies three key findings that recur throughout much of the recent literature. First, he finds that residents of urban Athens are more vulnerable to insufficient access to food, a finding that aligns with other sources (Skordili, 2013; Bernaschi & Leonardi, 2022).
Secondly, Konstantinidis’ emphasis on low income is corroborated by Penne and Goedemé (2021), who examine the relationship between low income and food insecurity in 21 European countries., Their findings indicate that residents of (sub)urban Greece are strongly affected by inadequate income for accessing a healthy diet. Bernaschi and Leonardi (2022) extend this notion by highlighting the interplay between economic and humanitarian crises as one important determinant of food access disruption in Greece. This supports Awais and Sarwar Shah’s (2019) conclusion that the vulnerability to food insecurity among undocumented migrants in Athens is linked to their low social and economic status.
This increased vulnerability of particular groups represents the third recurring finding. Konstantinidis (2022) identifies populations particularly affected by food access disruption, including lower-income households, single mothers, and immigrant families from non-EU countries. Skordili (2013) expands this list to include the urban poor, people with disabilities, the elderly, and single households with no family ties.
In addition to these three recurring findings, the relevant literature also addresses choices in sourcing food (Skordili, 2013; Bernaschi and Leonardi, 2022). Finally, there is a body of research that explores the relationship between food security, malnutrition, and chronic disease among older adults (Gkiouras et al., 2020; Grammatikopoulou et al., 2019). Although these studies primarily focus on the healthcare implications of food security, they contribute to understanding the high complexity and multiple levels of food access and security.
Altogether, the recent literature reveals two key limitations. First, the focus tends to be on the broader concept of food security rather than on food access specifically. While many sources, particularly Konstantinidis (2022) and Penne & Goedemé (2021) do substantially discuss food access based on its official definition, the concepts of food security, destitution, and others are not always clearly defined and demarcated. This conceptual clarity matters as food access alone can be measured using different indicators. The importance of distinguishing and evaluating different components of food security separately has been raised in previous research (Leroy et al., 2015).
Secondly, it is striking that the recent literature predominantly discusses the economic dimension of food access in Greece. While this research interest is justified considering the recent economic crisis, there appears to be a gap in analysing the significance of food access in Greece beyond its economic component. While Skordili (2013) highlights the issue’s complexity by analysing limited food sourcing, no study comprehensively investigates the extent to which other factors may play a role. Additionally, Loopstra et al. (2015) have pointed out the significance of this knowledge gap at the European level. They argue that economic factors influencing food affordability cannot be the sole determinant of food access, given that European countries exhibit varying prevalence rates despite comparable rates of regression.
Food Access Among Refugee and Migrant Populations
While Awais and Sarwar Shah (2019), Konstantinidis (2022), and Skordili (2013) have noted that members of RM communities may be especially affected by food access disruption in Greece, it is essential to situate the general knowledge on this topic.
In one of the most recent contributions, Gingell et al. (2022) examine the determinants of food security among individuals from refugee backgrounds residing in high-income countries, identifying the lack of access to culturally appropriate foods as a key factor contributing to food insecurity. This aligns with studies such as those by Moffat, Mohammed and Newbold (2017), which emphasise the cultural dimensions of food insecurity. Hardison-Moody et al. (2016) further explore this finding by highlighting the importance of community-based approaches in their study of a diverse group of RMs in North Carolina.
Finally, a growing body of research focuses on systemic factors, underlying food insecurity. Nabulsi et al. (2020), who study food insecurity among Syrian refugees in Lebanon, emphasise the impact of dependency on formal assistance as well as economic and health disparities, corroborating Awais and Sarwar Shah’s (2019) characterisation of the vulnerability created by lacking legal papers. Additionally, Deschak et al. (2022) identify extreme food insecurity among migrants in transit. These studies highlight the need for interventions that strengthen the legal and economic position of RMs.
Improving Food Access
The literature on improving food access substantially focuses on evaluating or informing interventions at the policy-making level. One of the most comprehensive contributions is made by Wegener et al. (2012) who studied regional policy and planning approaches in Canada. Their conceptual model emphasises the engagement of multisectoral stakeholders, aligning with a consensus on the need for multidisciplinary solutions (Bublitz et al., 2021; Poole et al., 2015).
Another key theme is the significance of inequities in food access (Bublitz et al., 2021), predominantly identified in urban contexts. Several case studies focus on the impact of and possible interventions for so-called food deserts, namely the areas lacking access to full-service supermarkets (Costley, 2018). While some studies identify food deserts as crucial targets for food access improvement interventions (Wang et al., 2018), others do not find the concept to be relevant (Helbich et al., 2017).
Moreover, the importance of a theoretical foundation for addressing food access is emphasised. Leroy et al (2015) criticise the lack of theoretical models guiding the conceptualisation and evaluation of food access beyond policy-making. Highlighting the importance of a well-structured use of indicators, the authors develop a model for measuring food access at the individual and household levels. Based on similar concerns, the most comprehensive model for the analysis of food access is provided by Freedman et al. (2013). Establishing the domains and dimensions of nutritious food access, their model has served as a basis for multiple case studies (Freedman et al., 2018; Ibrahim et al, 2019; Khandelwal & Kotval-K., 2020).
Altogether, it emerges that most publications are based on case studies in North America. While this does not necessarily make their conclusions inapplicable to other geographical and/or legal contexts, the lack of food security research in the European context has, indeed, been criticised (Konstantinidis, 2022; Penne & Goedemé, 2021).
Research Gap
Overall, there is a gap in the knowledge specifically addressing interventions for food access disruption in Greece. The problem is indeed prominent, especially in Athens. While the existing literature mainly focuses on the economic dimension of the issue, other dimensions and related drivers remain understudied. Moreover, although the general literature on food access provides incentives for further research and practical interventions, it also emphasizes the context-based nature of the issue. Similarly, the literature on food security among RMs is specific to its respective study areas. Food access is not only country-specific but also varies between urban and rural areas, implying the need for a specific scope of analysis.
Theoretical Framework
As identified through the literature review, a framework that captures the context-specific complexity of food access disruption is essential for elaborating possible interventions. The conceptual model by Freedman et al. (2013) stands out as the most comprehensive and applicable to the given research problem, providing clear guidance in identifying the domains and dimensions relevant to addressing food access. This classification allows for an in-depth understanding and serves as a comprehensive basis for conceptualising practical implications.
The model consists of five domains with a total of seventeen related dimensions, which collectively constitute nutritious food access (see Figure 1).
Figure 1
Multicomponent Model of Nutritious Food Access (Freedman, Blake & Liese, 2013).
It must be noted that the given framework uses the concept of “nutritious food access”. As the literature review revealed, definitions in food security research tend to be inconsistent. This study focuses on food access and its disruption which, in its first instance, refers to the precarious situation of ensuring survival. This definition aligns with the framework’s implication of a certain nutritional food quality. As emphasised in the employed definition, food access cannot be assessed without considering its nutritional quality. Therefore, while this study’s definition is based on the same understanding employed by Freedman et al. (2013), the term ‘food access’ is employed for the sake of coherence.
Methodology
Informed by Freedman et al.’s framework (2013), this study aimed to identify the key components that need to be considered in efforts aiming to address food access disruption among RM populations in Athens. This was achieved through a comprehensive analysis of the relevant dimensions and related unmet needs.
The research scope was narrowed to the municipality of Athens, Greece, given its identified relevance as a specific geographical and urban context as well as considering the study’s feasibility.
Data Collection
The research design was of a descriptive, qualitative nature. According to Aspers and Corte (2019), qualitative research is a process that enables making “significant new distinctions”, suitable to capture an issue’s multitude of different layers. Such a qualitative approach is the foundation of Freedman et al.’s (2013) framework that informed this study, highlighting its ability to allow participants to highlight key nuances and identify prior unknown concepts in their own words.
Moreover, this study’s approach was hybrid because, while strongly based on Freedman et al.’s (2013) theoretical framework, it was adapted iteratively. Both secondary and primary data sources were used, applying three different data collection methods to answer both sub-questions.
Desk Research
Desk research was conducted in all research stages. Next to the preliminary review that informed the research design, secondary data was analysed by applying the theoretical framework. This way, themes that further informed the primary data collection were conceptualised. Additional desk research was further conducted at any stage when seen fit.
Qualitative Survey
In the period between 11 October and 2 November 2022, a qualitative survey with open-ended questions was conducted among 20 adult members of RM communities in Athens. All participants, beneficiaries of a humanitarian, non-governmental organisation (NGO) providing food assistance, were recruited through the NGO. Participants were informed about the research and their participation, and consent was obtained orally. This was done to ensure sensitivity to the participants’ safety situation and because some of themare illiterates. The surveys were then conducted in a separate physical space from the organisation’s setting.
While most of the qualitative surveys were done in English, two interpreters assisted the researcher where needed and feasible, which was the case for seven surveys. Of these, two were conducted in French and five in Farsi.
In contrast to a quantitative survey, qualitative survey questions are open-ended. The method was chosen to gather participants’ insights in their own words, especially appropriate given that the research topic is understudied. In this sense, the qualitative approach did not limit the response options to the researcher’s preconceived notions but allowed participants to introduce undiscovered concepts and describe them in depth.
The use of qualitative surveys was preferred over conducting qualitative interviews since it allowed a bigger sample size. Given the research context, the length of interviews would have restricted the number of participants decisively, and required a much higher time investment, decreasing the likelihood of participation. Although the qualitative approach did not seek a high quantity of participants for statistical representativeness, the diversity of the target group (in terms of nationality, legal status, and other factors) made a larger sample preferable. Capturing a broader range of experiences enhanced the quality of the findings.
Upon deciding on the survey format, a key factor was the target population’s characteristics. Méndez & Font (2013) dedicated a book to the challenges of conducting survey research among ethnic minorities and immigrant populations, concluding that there is ”no clear-cut best set of practices that can be followed by researchers and practitioners.”
The survey was conducted face-to-face, with the researcher reading out the questions and typing the participants’ answers manually, anonymising any personal identifiers that the participants agreed to share. The data was stored in a protected folder accessible only to the researcher.
While a face-to-face mode requires substantially more time than a written, self-completed survey, it was preferred based on the following considerations. Firstly, the researcher could observe the participants’ reactions, and note those worth considering, such as hesitation. Secondly, the face-to-face mode did not exclude illiterate participants. Lastly, face-to-face surveys have been found to increase participation rates, especially among marginalised groups (Neuman, 2012).
Finally, the researcher decided to follow the conversational format of face-to-face surveys. As West et al. (2018) highlight, conversational interviewing differs from standardised interviewing because it allows the researcher to rephrase or explain questions if the participant does not understand them, after initially reading the question as it is in the survey script,. The authors highlight that, while this format demands more time than standardised interviewing, it may decisively reduce bias and increase response accuracy. This is because different interpretations of the standardised questions may decrease the consistency of responses, especially in heterogeneous groups (Neuman, 2012). Given the target group’s diverse backgrounds and varying level of English, this consideration was critical.
Garbarski, Schaeffer, and Dykema (2016) further informed the application of conversational interviewing by finding that beyond addressing misunderstandings, additional “responsive behaviours” by the interviewer, such as basic utterances that acknowledge what is heard, may enhance the quality of data. Given the sensitive nature of the subject, this approach required a guideline for non-verbal interviewing behaviour. It was imperative to seek an appropriate balance between maintaining a professional demeanour and acknowledging the participants’ contributions. To deal with this challenge, the interviewer drew on the findings by Foucault Welles et al. (2022), which stated that an appropriate use of smiling, nodding, and direct eye contact significantly increases matched responses. The interviewer found the use of these three behaviours to be adequate and manageable.
To test the chosen method, a pilot qualitative survey was conducted, which led to the refinement of survey items and endorsement of the design.
Expert Interviews
To corroborate the qualitative survey data, three in-depth interviews with relevant experts in the field were conducted between 11 November and 5 December 2022. All participants were staff of three different humanitarian organisations working closely with the target population. The interview guide was semi-structured, designed in relation to the survey elements while encouraging participants’ own input. The interviews were conducted via video calls, with the participants receiving the interview script and signed consent forms in advance. All interviews were audio recorded and transcribed manually.
Ethical Implications
The more fluid format of conversational interviewing necessitated an increased awareness of interviewer bias. Although scholars endorse its quality, this approach demands a considerate and skilful implementation (Garbarski et al., 2016). One particular challenge was the practice of rephrasing a question, given its risk of asking a leading question or one that deviates from its initial meaning. Thus, when a question had to be rephrased, it was most often only simplified. If a respondent still did not understand, the question was skipped. Attention to nonverbal reactions was crucial in this process. Some participants did not specify whether they did not understand the question or did not want to answer it, which, given ethical implications, required careful assessment based on nonverbal clues. In most cases, the researcher rephrased the question once and then continued. It was important to be aware that for various reasons, including cultural norms, participants may be uncomfortable stating their reluctance to answer. Instead, they may react with nonverbal behaviours that may or may not be recognised as discomfort. An empathetic approach was prioritised to ensure that the respondents did not feel forced or negatively affected by their participation.
Such a sensitive approach was critical throughout the whole process, and the face-to-face format required the setting to be as safe as possible. Another essential issue, which could not be fully mitigated, was the researcher’s own background and privileged position (Neuman, 2012). It was important to be conscious of the existing power imbalance and its implications.
Finally, the effect of working with an interpreter had to be considered. Conducting qualitative surveys in participants’ primary or secondary languages was a huge advantage as it allowed access to non-English speakers. This not only ensured understanding of the questions but also allowed participants to express themselves in their native language, which is always preferable. Moreover, the familiarity created by a shared language and/or culture may increase trust and comfortableness to speak, as became especially evident with the Farsi interpreter. At the same time, the researcher needed to be aware that working with an interpreter meant a gradual loss of control over the process. Thus, it was important that the interpreter had a clear understanding of the research aim and approach.
Sampling
For both the qualitative survey and the expert interviews, purposive sampling was applied, which involved fulfilling a set of criteria.
Qualitative survey respondents had to reside in the metropolitan area of Athens and hold the status of recognised refugee, recognised migrant, asylum seeker, undocumented migrant or refugee, or rejected migrant or refugee. They had to be affected by food access disruption, defined as not having daily access to food. Since all participants were sought from among the NGO project’s beneficiaries, this approach not only decisively increased accessibility to the target group, it also allowed for an eligibility assessment to verify that respondents met the sampling criteria.
Everyone who was available and willing to participate was invited, considering the logistics and resources of the researcher. The initial sample size of over 30 participants was not met, as only 20 qualitative surveys were conducted, which is elaborated on in the limitations section. Noteworthy characteristics of the survey sample include that 16 out of 20 participants are female. At the time of the surveys, 12 participants held the legal status of recognised refugees, and seven were single parents. Their ages ranged from 27 to 44, with an average age of 37. Finally, participants represented eight countries of origin: Afghanistan, Cameroon, Congo, Eritrea, Ethiopia, Iran, Nigeria, and Senegal.
Regarding the expert interviews, participants had to demonstrate knowledge and experience in working with RM communities affected by food access disruption in Athens to be eligible. The recruitment of the three experts was facilitated by the same NGO, and one of the experts, who was interviewed, was a staff member of the organisation.
Given the highly sensitive context of forced migration and associated humanitarian work, all research participants have been anonymised and will be referred to using numbered abbreviations: P for qualitative survey participants and E for experts.
Data Analysis
In line with the research methodology, all gathered data were evaluated qualitatively applying the method of thematic analysis as outlined by Braun & Clarke (2006). While strongly based on the theoretical framework, this method allowed for flexibility to let new topics emerge. As the authors suggest, certain methodological decisions are imperative before delving into the analysis itself. The key considerations included adopting a hybrid approach, namely, starting deductively while allowing the data to reveal more themes. Additionally, in line with the research objective, themes were identified at the semantic level, rather than the latent level. Lastly, the analysis was conducted within a realist/essentialist paradigm.
The analysis process involved six phases (Braun & Clarke, 2006). First, the researcher familiarised herself with the data by reading through all transcripts analytically, followed by a first round of generating initial codes. In the third step, she started searching for themes, transferring all codes into an Excel sheet, and arranging them to generate categories. Freedman et al.’s (2013) 16 dimensions had been incorporated beforehand and listed as themes if there were codes related to them. Codes that did not fit the pre-existing themes were rearranged until a first logical order emerged, supported by sketching a first thematic map. In the fourth step, themes were reviewed by systematically revising all the data and finalising the thematic map. In the fifth step, the final themes were refined until complete. The themes were then grouped into eight categories, five of which were related to the framework’s domains, while three were newly generated. Finally, illustrative data excerpts were added for the written representation of the findings.
Quality Assurance
To enhance the study’s quality, the principle of trustworthiness was upheld, ensuring that the research findings are credible, transferable, confirmable, dependable, and authentic (Cope, 2014).
To ensure credibility, method and data source triangulation were applied (Carter et al., 2014), with the expert interviews being particularly valuable in substantiating the qualitative survey findings and providing a broader perspective. Their relevance was enhanced given the sensitivity of the issue, which may have prevented survey participants from disclosing certain aspects. Data source triangulation was applied by having a diverse sample.
While the transferability of this research is limited given its aim not to generalise a phenomenon but to study a highly specific context, findings may still be transferable, particularly when supported by a detailed description of the research process. This description also contributed to establishing confirmability, enabling a discussion of limitations. A strength of the study is the systematic provision of direct quotes from participants, which supports understanding how the researcher derived her conclusions. However, individual qualitative interpretation supposes the possible impact of bias.
The study’s documentation allowed for deriving recommendations for its reproduction, enhancing dependability, along with the extended survey conduction period of several weeks without observed changes in data patterns. Finally, authenticity was prioritised by including direct quotes, aiming to represent participants’ experiences as accurately as a secondary presentation of knowledge can (see Limitations).
Results
The findings are presented in two sub-topics: the dimensions of food access disruption and related unmet needs.
The Twenty Dimensions of Food Access Disruption
The identified twenty dimensions of food access disruption among RM populations in Athens include eleven dimensions provided by Freedman et al. (2013) and nine that were inductively generated. Some dimensions were slightly amended in their meaning; for example, “quality and variety of food sold” was changed to “quality and variety of food available.” This adaptation was essential to accurately reflect the target group’s reality, where food is not only purchased in stores but can also be acquired in different ways, such as through government assistance or civil society support programs.
The 20 dimensions were grouped into eight domains, five of which correspond to the domains outlined in the theoretical framework. Figure 2 presents a final overview of all the dimensions and domains, which are elaborated upon in the following sections.
Figure 2
20 Dimensions of Food Access Disruption among Refugees and Migrants in Athens.
Note. Own elaboration based on the framework by Freedman et al. (2013).
1 The Economic Domain
“If I have money I eat. I don’t have money, so I don’t eat. When you have money, you eat, and everything is okay.” (P8)
The economic domain includes three dimensions: household finances, food costs, and store incentives. Almost every participant identified the lack of funds for purchasing food as a strong driver of food access disruption, as P8 illustrated. This aspect tended to be associated with difficulties in finding sufficient and/or legal work, emphasising the lack of employment as a key challenge. As summarised by P17, “If I get a job, I have the money to get anything. Financial assistance and the projects are okay, but they cannot give me what I want. Then I can get things for my baby.”
A small number of participants mentioned the high cost of food as a challenge. E1 expressed her concern about the effect of the current inflation rate on people’s vulnerability to food access. High food costs can be perceived both directly and indirectly, particularly when they reduce the capacity of food assistance services. Finally, store incentives may increase someone’s food access, as noted by P4, who emphasised the value of discount vouchers offered by supermarkets.
2 The Supply Domain
“It is difficult. I have to eat what I see.” (P17)
The supply domain includes the dimension of variety and quality of available food, referring to the extent to which perceived nutrition quality and variation in supply influence food access. Most participants, such as P17, highlighted that while they wish for more varied and nutritious options, they often accept whatever food is available.
However, some participants indicated that unless products meet a certain quality standard, they rather skip those items. As P14 stated, “Of course, if it’s not healthy I won’t eat it.” In this sense, food access is diminished. E3 supported this finding by adding that especially readymade foods distributed by food kitchens may be perceived as unhealthy and, therefore avoided.
These contradictory data highlight the variability of individual situations within a diverse sample, allowing for different degrees of selectiveness and implying differing relevance of the supply variable. At the same time, this finding underscores how “nutritious food” is defined from an individual perspective, determined by cultural and personal factors. This aspect re-emerges within the social and household domain.
3 The Spatial-Temporal Domain
“I need to move quite a distance to find my taste.” (P14)
Within the spatial-temporal domain, two relevant dimensions are travel distance and time, as well as housing and infrastructure. Firstly, the travel distance from an individual’s home to the relevant food sources, along with the associated time and transportation costs, may impact food access. Only a few participants, such as P14, articulated this challenge, which likely varies depending on their location within the urban context of Athens.
Secondly, the type of accommodation (or lack thereof) emerged to be relevant. Two participants described how their housing conditions directly decreased their food access. For example, P11 lacked sufficient infrastructure which inhibited her ability to prepare food. “I have a house, but I cannot use electricity.” Furthermore, unstable accommodation is a common problem among the target population. In the experience of E3, access to both housing and food are the most essential and intersected needs for her clients.
4 The Household Domain
“Now I pick up the vouchers, I don’t eat, it is just for my children.” (P9)
Household composition has a two-fold effect on food access. The first dimension is the number of caretakers present. For single parents, accessing food can be more difficult as they bear the sole responsibility for their family, fulfilling both caretaking and income needs. P10 explained, “I’m looking for a job but because I take care of my son it’s difficult for me. I can only work when he is at school.”
Moreover, the amount, age, and nutrition needs of children matter. A recurring finding is participants’ prioritisation of their children’s food access over their own, which may lead them to skip meals, as illustrated by P9. Special nutrition needs also pose a key challenge. As explained by P4, “It is okay for me but for my child, he is 3 years, I need better food. If I find vouchers, I can buy him some milk or fish.” E1 substantiated that the nutritional needs of adults and children exacerbate the degree of households’ food access disruption.
5 The Social Domain
“Sometimes we can go to an Iranian restaurant with traditional food. But just four times a month we get a voucher for this.” (P6)
The social domain comprises multiple dimensions. The first dimension, cultural pathways, traditions, and norms, was highlighted by both survey participants and experts who emphasised the importance of culturally appropriate food in various ways. Different cultural backgrounds imply variances in eating culture, which fulfils important psychological and social functions. It was striking that some participants, when asked about their access to healthy foods, identified it as sufficient as long as they could access foods from their own culture, indicating a lack when these options were unavailable. For example, P6 replied to a question about food quality by referring to the frequency of receiving meal vouchers for an Iranian restaurant. This aspect’s relevance is evident given the earlier finding that some individuals may not consume food they regard as unhealthy. Furthermore, cultural norms regarding gender may play a role. E1 observed that some women deem it culturally unfavourable to express their opinion or “complain,” which may inhibit them from communicating their needs and consequently reduce their food access.
Discriminatory practices constitute the second dimension. P16 described the racism she faces when procuring food in her neighbourhood as a key barrier. P3 indicated how discrimination indirectly diminishes food access by affecting employment opportunities. “Everywhere you go they do not give you a chance. They see your face and… There is no approach.”
On the other hand, support through social structures, such as a community or friends, may enhance food access (P18). Varying perceptions regarding community responsibilities are noteworthy in this regard. P16, relating her neighbourhood experiences, became emotional while identifying the absence of solidarity. “It’s not affordable for everyone. It’s not like everyone can eat. And they don’t give to others.”
Lastly, civil society support, such as providing staple food, vouchers, or other services, is decisive in enhancing food access. Most participants identified this support as essential. P17 said, “There is no other way than the organisation.” However, while expressing their gratitude, participants stated the support’s insufficiency. E1 further identified accessibility as problematic, excluding people unaware of the services or those unable to attend due to disabilities. E3 added that several NGOs deem undocumented people ineligible for their services. While civil society support may enhance food access significantly, it often comes with conditions.
6 The Personal Domain
“It is okay only for hungry, but not good.” (P8)
The personal domain comprises three dimensions. First, an individual’s health status may necessitate specific dietary requirements that, if unmet, limit food access. E1 noted that the high prevalence of disabilities among RMs creates logistical barriers to accessing food. Additionally, she suggested that limited knowledge regarding nutrition may exacerbate health conditions. In this sense, access to nutritious food access is indirectly decreased due to the resulting health problems. This underscores the importance of considering individuals’ differing perceptions of nutritious products, especially when available food varies across different geographical contexts and cultures. E1 added that the ability of support services to provide fresh products may also be limited.
Lastly, individuals’ food-related identities and preferences may play a role. When asked about their access to food that meets their preferences, most participants described themselves as not being selective because they cannot afford to be. As P2 explained, “Sometimes you would like some other things like cheese or so, but you cannot be picky.” While some participants mentioned the challenge of finding food that matches their taste, they all accepted this reality, recognising that taste cannot be their priority.
A striking observation involves recurring reactions of hesitation, laughter, or embarrassment when participants were asked about food quality and preferences. For example, one participant flinched and laughed but did not verbalise her response. Such reactions may be interpreted as an inhibition to “complain” or even consider taste as relevant. P2, for example, stated, “I’m happy because I get food. When you get food, you’re happy.” While the low prioritisation of taste is logical in a situation of food access disruption, this has important implications considering the functions of food beyond merely addressing hunger.
7 The Psychological Domain
“I am mentally very bad. Before, when I ate cheese and bread, it was very good. Now it isn’t good, mentally I am bad, so anything is not tasty.” (P9)
An unprecedentedly emerging domain was the psychological one, comprising three dimensions. Firstly, negative associations with seeking support, such as shame, were identified as significant challenges. P2 emphasised this when asked about her biggest barrier to food access. She shared, “Shame. Of course, you feel shame to ask for food. If you have to go and to ask someone for food, it is very difficult.” Another participant (P3) highlighted her distress about being “seen like a beggar.” Considering the dependence of some participants on support services, this dimension is significant.
The second dimension is mental health issues and overwhelmedness. Introduced inductively and distinct from the dimension of health, this dimension describes a systemic condition directly related to the experience of food access disruption itself. P9’s testimony illustrated this aspect. Other participants expressed their frustration about not having enough food, especially given their caretaker responsibilities. E3 emphasised how the detrimental psychological effects of experiencing insufficient food access may in turn affect one’s resources and abilities to procure food. P9 indicated the effect of her feeling overwhelmed. “I have a lot of challenges (…). I need a doctor, I don’t have money, how can I buy food? Also, the language is a problem, everything. I don’t know how I can find the food.”
Lastly, agency in making choices regarding the type and preparation of food was shown to be a strong determinant. E2 recalled how in her experience, increased agency leads to improved nutrition as individuals can organise food access according to their needs, knowledge, and preferences. This aspect became evident during the survey, as multiple participants, when given the space to express additional thoughts, highlighted their wish for more agency. P13 emphasised how his food access would improve. “If I could get small amounts of money, I could utilise them better. I could get something I definitely like, and I would be better able to organise my spendings.”
8 The Legal Domain
“Once I have my legal document everything is okay, I think I won’t come for any assistance anymore.” (P14)
Finally, the legal domain encompasses two new dimensions. The first, and among the most strongly and frequently highlighted, is legal status, referring to the type of documentation that determines one’s ability to reside, work and access basic services. As P20 described, “The legal work would help the most. I am doing illegal work as a hairdresser.” E1 further emphasised how the fear of being arrested due to lack of documentation restricts free movement, thereby limiting options for accessing food.
Eligibility for certain social allowances is also based on legal status and is thus crucial in securing food access. E1 and E3 highlighted that being without documents often leads to exclusion from support services, exacerbating food access disruption. Similarly, P5 pointed out that the government’s withdrawal of cash support presents a major challenge for her family.
Key Unmet Needs
The multitude of possible dimensions indicates related unmet needs. Given the heterogeneity of the target population, various aspects emerged. As E3 noted, “The needs are wide and vast.” This study identified eight most urgent needs, which are visualised in Figure 3.
Figure 3
Key Unmet Needs of Refugees and Migrants Lacking Food Access in Athens.
Note. Own elaboration.
1 The Need for a Sufficient Amount of Nutritious Food
“My wish it to have enough food for my son.” (P10)
Almost every participant described the amount of food they access as insufficient, as illustrated by P10. The urgency of this issue is evident; adequate food is a basic requirement for survival and long-term health. E1 emphasised her concern about the consequences of chronic malnutrition, which poses serious threats to both individual and public health among RMs. In addition to the need for food that meets general nutritional standards, the lack of specific products, such as milk for babies and children, is particularly critical. Young children’s nutritional needs are a key concern. As P14 expressed, “It is okay for me but for my child, he is 3 years, I need better food.” In this context, E1 noted that project beneficiaries frequently request vegetables for their children. She emphasised that while food assistance services are aware of this need, they usually cannot provide fresh products due to a lack of refrigeration. Finally, there is a lack of food that meets specific dietary requirements, which exacerbates existing health issues and can lead to hospitalisations, further impacting public health (E1).
2 The Need for Legal Recognition/Protection
“For me, my only issue is really getting the legal document.” (P14)
Obtaining legal residence in Greece is one of the most urgent unmet needs. As mentioned earlier, legal status is a crucial determinant of food access, as it underpins the ability to gain employment, move freely, and access support services. Experts also emphasised its link to mental health issues. E2 explained how lengthy legal procedures can deprive individuals of their productivity, contributing to increased emotional distress alongside a prolonged state of uncertainty. In this sense, legal recognition is strongly connected to agency and dignity. As P3 expressed, “I don’t want to get food from others, I want to work. People see you as a beggar and I don’t want that.”
3 The Need for Employment (Assistance)
“Food is not my problem. Of course, it is very important but what I need is work.” (P3)
Following the emphasis on legal recognition, employment emerges as a key unmet need that obstructs food access, as exemplified by P3’s quote. The ability to work is further connected with other barriers which may require assistance. Key factors include lack of fluency in the Greek language, health issues, and the challenges of being a (single) parent. As P18 highlighted, the intersection of these factors is especially challenging. “I don’t know how to speak so I cannot work. I am alone with two kids.” E3 supported this finding by recalling that the RMs she works with primarily seek assistance with their legal documents and employment, aiming to become independent from food assistance. “And they say that if they manage to do that, then they don’t need the food.”
4 The Need for Agency to Make (Food) Choices
“It’s automatically critical. We can’t go where we want, we don’t eat what we want, and the children don’t eat healthy.” (P16)
As mentioned earlier, the ability to make food choices is a key determinant in improving food access, as it leads to better resource utilisation. Conversely, the lack of agency has a significant impact, as highlighted by P16. This need also relates to the mental health dimension. While the primary function of food is to address hunger and ensure survival and health, E2 expanded on its psychological and social benefits:
“Allowing the people to cook their own food, it doesn’t have to do only with the quality and the fairness of having food, proper food. It’s also something that relaxes you, makes you feel better. So, it makes you feel that you are not a victim that all the other people decide about. You can at least decide and reprise the role of having the family meet, and you decide what you eat, and you all gather and you’re together.”
In this sense, having agency enhances well-being in multiple ways, which in turn may improve food access. As E2 summarised, “It’s connected to essential freedom.”
5 The Need for Consideration of Cultural Factors
“In my homeland, we eat healthy food. Here we try to eat healthy food.” (P11)
Culture affects several dimensions of food access. Beyond its social dimension, it plays a crucial role in agency and psychological well-being. One issue is the lack of culturally appropriate food, for example in civil society services (E3). But there is a wider need to consider relevant cultural factors. For example, as implied above, cultural background influences of what constitutes nutritious food (P11). Lastly, cultural factors may affect how individuals participate, express their needs, and evaluate the assistance they receive.
6 The Need for Specific Support for (Single) Parents and Other Groups with Certain Risks and Needs
“I have a big problem, I am HIV positive, that makes everything difficult. I have a small baby, and I stay alone.” (P4)
P4’s testimony underscores another need that connects multiple dimensions of food access disruption: the necessity of tailored support for specific groups. Parents, in particular, require assistance, given the findings that their food access is impacted by unique challenges. These challenges include the need to prioritise their children’s well-being, the difficulty of balancing childcare with income generation, and the mental stress of not being able to provide adequately. Such deprivation may force RMs to resort to desperate measures. For instance, survival sex is a significant concern in Athens, particularly affecting women and unaccompanied minors (Digidiki & Bhabha, 2020).
Lastly, as E1 highlighted, health issues and disabilities require specialised assistance to ensure access to essential services. The lack of stable housing is another critical factor. When several of these factors intersect, food access disruption is exacerbated. Given the complexity of these issues, targeted assistance becomes essential.
7 The Need for Improved Access and Sharing of Information
“We do know about the state services provided for refugees but there are a lot of organisations in the field that may lack communication and information shared on this field.” (E1)
Given the relevance of civil society and governmental support, the need for improved access to information emerges. E1 highlighted that certain groups may be eligible for specific services but are unaware of this fact. Her testimony points to a lack of information sharing among stakeholders. While various practical barriers to accessibility may be at play, as outlined in a report by the Greek Council of Refugees (Karagiannopoulou et al., 2020), addressing information gaps among RMs is crucial and can be supported through language assistance (Hannides et al., 2016).
8 The Need for Participation
“So, I would opt for having advocacy that comes from below, that comes from the affected population, and that it is more respected in the state. Because we know that the state discusses with communities from time to time, but I don’t think the state is really listening.” (E2)
Lastly, RMs must be able to actively participate in their communities. E2, in addition to identifying the importance of inclusion in policy-making, emphasised the need for civil society organisations to engage RMs and enable them to advocate for themselves. In this sense, participation in public discourse is closely tied to agency, dignity, and psychological well-being. E1 and E3 emphasised this aspect by discussing the importance of dignified approaches that build trust, enabling participants to identify and articulate their main needs, which would ultimately enhance food access.
Conclusions
This research addressed the question: What are the key components that efforts aiming to address food access disruption among refugee and migrant populations in Athens need to consider?
The answer is two-fold. The first part of the results provides an overview of all the identified dimensions that underlie food access disruption among RMs, which are grouped into domains, resulting in the comprehensive model visualised in Figure 2. At this stage, the model intentionally refrains from prioritising any specific dimension, allowing it to encompass the diverse realities of participants, where key drivers may differ considerably. For example, if one participant’s primary barrier to food access is racism, this aspect may be irrelevant to others; nonetheless, the model acknowledges it as an equally important point. Given the target group‘s diversity, capturing all these experiences is crucial. Thus, the model proves to be a valuable contribution in two ways. First, by expanding the theoretical framework’s domains and dimensions, it addresses the existing knowledge gap in the literature. Second, it maps all identified elements that multi-component approaches need to consider.
The second part of the analysis complements the model by synthesising the key unmet needs experienced by RMs, which involves prioritisation. The eight identified unmet needs connect multiple previously identified dimensions and domains, thereby complementing and distinguishing the two parts of the findings. It also recognises intersectionality, as addressing each dimension individually would provide an incomplete picture of the needs. For example, one of the most highlighted dimensions concerns household finances, which is strongly linked to unemployment, which is also driven by factors such as legal status, single parenthood, and health issues. In this sense, the need for employment (assistance) encompasses the given situation most accurately.
The guiding question can finally be answered through three key conclusions. First, the importance of acknowledging each dimension becomes evident. The results indicate that RMs are far from being one homogenous group; therefore a one-size-fits-all solution will not improve everyone’s reality equally.
Secondly, considering this diversity, any endeavour aiming to address food access disruption among RMs must acknowledge the complexity of the issue. This demands multi-component approaches that consider the interconnectedness of the different dimensions, such as the psychological dimension that spans several identified needs. In this sense, it is acknowledged that mental health issues arising from the experience of food access disruption can be addressed by targeting these identified needs effectively. While different stakeholders will focus on their respective areas of expertise, recognising synergies among them is imperative.
Finally, the synthesis of results reveals the many layers that constitute food access disruption among RMs, both in terms of its causes and implications. Despite the participants’ diverse situations, certain structural factors underlie their shared experiences of food access disruption. Where such systemic inequalities emerge as a consequence of certain dynamics, particularly legal barriers, addressing these drivers at their root is indispensable.
To conclude, the findings offer a comprehensive overview of food access disruption among RMs, pointing out focal points of urgency. This information may facilitate different stakeholders’ in-depth understanding and its translation into practical strategies for designing and adapting interventions that effectively address food access challenges faced by RMs in Athens.
Limitations
This study is not without limitations, which are significant for both practice and future research.
First, given the emphasis on qualitative surveys, it is crucial to consider the positions of the participants involved. Despite all ethical considerations and measures, the research topic, and thus participation, remained sensitive. Given the face-to-face survey context, participants may not have felt entirely comfortable answering questions fully, which could affect the results. While the researcher had the advantage of accessing a setting where some level of trust was already established, the importance of fostering trust cannot be overstated. As identified by Méndez & Font (2013), individuals belonging to ethnic minority and/or immigrant populations often have valid reasons to hesitate to participate in research. Researchers must always keep participants’ situation and safety in mind, especially for those who are undocumented, as revealing any sensitive information may jeopardise their situations. Lastly, the potential impact of participating in research must be constantly scrutinised from an ethical standpoint, and researchers must ensure that participants do not feel pressured to participate or to provide answers they believe are desired.
Another main limitation was the language barrier, which had two implications. Firstly, it prevented individuals who did not possess a sufficient level of English or who did not speak one of the interpreters’ languages from participating. Secondly, for those participants who agreed to be interviewed in English, the choice of language may still have been a shortcoming. Depending on their English fluency, their responses may not have fully reflected their true thoughts as they would have in their native language. While this research lacked the resources for a multilingual team of interviewers, it is strongly recommended for any future study with adequate means. Not only would native language interviewers increase the study’s quality by enhancing the accuracy of responses but they would also help build greater trust. Overall, enabling participants to express themselves in their native or fluent second language of choice should be prioritised for both quality and ethical reasons.
Another limitation is the relatively small qualitative survey sample. In addition to logistical challenges and the study’s limited resources, a primary reason for this was the language barrier, which impeded participation when those approached did not understand English sufficiently and no suitable interpreter was available. In some instances, after beginning the survey, the researcher realised that a participant did not fully understand the questions; in these cases, she slowly aborted the survey and discarded their answers. Although additional expert interviews and desk research were conducted to supplement the data, a bigger sample that includes a broader range of perspectives would enhance the research findings.
Finally, researcher bias is always a consideration. In this study, efforts were made to mitigate its impact through quality assurance measures, including extended ethical reflections and consultations with peers and mentors. Future research should continue to implement these steps and, if resources allow, go beyond them, most importantly through researcher triangulation and the inclusion of native language interviewers.
Given the dynamic conditions that RMs face, the findings of this study are inconclusive. While their specificity to the context of Athens is its main strength, any research aiming to transfer this approach must engage in a critical review of the model. The diversity of the sample in terms of legal, ethnic, and demographic aspects is another strength, revealing both patterns and differences. Research focusing on a specific subgroup might paint a rather different picture.
Ultimately, future research should combine quantitative and qualitative methods. While quantitative designs may add significant value in identifying (representative) patterns, qualitative approaches are key to exploring specific aspects in greater depth. This study raises numerous follow-up questions, such as, What shapes RMs’ perceptions of nutritious food, and to what extent does this influence their food access? How is food access in its bare ‘survival function’ demarcated from its long-term health, social and psychological functions? And above all, what role does the classification of individuals as ‘refugees or migrants’ play in constructing systemic inequalities, affecting rights as fundamental as the right to food access? While these questions were touched upon, a more in-depth exploration was beyond the scope of this research. It is hoped that this study will, at the very least, highlight the critical importance of food access disruption and incentivise further exploration of this issue.
Notes
[1] Detailed examples of policy responses are outlined by Wagner (2015) who chronicles the legislative changes made at the national level, such as those regarding the duration of refugee status, financial benefits and family reunification. Guiraudon (2018), on the other hand, details the EU council’s “emergency measures” such as military patrolling, an emergency relocation plan, the creation of EU “hot spots” and an agreement with Turkey. Majcher et al. (2020) systematically analyse and criticise the expansion of immigration detention regimes, focusing on their duration and conditions around the EU.
[2] See HRW (2018) and Benvenuti et al. (2024) for detailed documentation of the implications affecting RMs’ access to basic resources such as housing and healthcare, up to the point of putting them at risk of violence and trafficking.
[3] While the frequency of the different migration routes fluctuates, Italy, Greece and Spain remain primary countries for first arrivals due to their positions at the EU’s external land and sea borders. As recorded in the Operational Data Portal of UNHCR, out of 144.925 arrivals to the EU in 2024, 29% (42.096) were registered in Greece (data updated 13 October 2024).
Acknowledgements
I would like to express my gratitude to all my professors and peers at Windesheim Honours College for their support throughout my academic development. I am especially thankful to my tutor Agnes Zenaida V. Camacho, whose support was crucial to this research. I am eternally grateful to everyone who supported me during my time in Athens. Above all, this article is dedicated to my research participants who trusted me with sharing their experiences.
Competing Interests
The article is based on the internship research report submitted by the author as a requirement for completing her bachelor’s degree at Windesheim Honours College. During her internship, the author received guidance and counselling from Agnes Zenaida V. Camacho, the Editor-in-Chief of this journal. This counselling did not include the assessment of the author’s internship reports. In compliance with Glocality journal policy, this submission underwent the same rigorous peer-review process as other submissions. All editorial decisions were made by the Editor assigned to this article.