Urban Vulnerability to Food Insecurity Under Displacement Pressures: Evidence from Tigray, Ethiopia

Abstract

Food insecurity remains a pressing challenge in conflict zones, where disrupted livelihoods, mass displacement, and eroded institutional support increase household risk. The armed conflict in Ethiopia’s Tigray region erupted in November 2020, devastating infrastructure, displacing over a million people, and amplifying urban hunger. This study assessed the effects of war-induced internal displacement on the vulnerability of urban households to food insecurity (VFI) in Tigray’s host communities. Using cross-sectional data from 560 households surveyed in May–June 2024, we computed food insecurity using the Household Food Insecurity Experience Scale (FIES) and applied ordered logit regression to identify the drivers of VFI. The findings indicate that 14.46% of households were food-secure, with 21.43%, 35.54%, and 28.57% facing mild, moderate, and severe vulnerability, respectively. Significant predictors included household head age, education, widowed status (especially for women), and humanitarian aid receipt, allied with displacement scale and conflict damages, which elevated vulnerability. These results underscore the need for integrated interventions that blend emergency aid with livelihood restoration. Policies must target at-risk groups, rebuild assets, and enhance access to education and financial resources. Ultimately, facilitating the repatriation of internally displaced persons is vital for post-conflict recovery in the Tigray and analogous settings.

1. Introduction

Armed conflict remains one of the most destructive forces shaping household food security and welfare in Sub-Saharan Africa (Muriuki et al., 2023). Beyond its immediate toll on lives and livelihoods, war drives mass displacement that alters settlement patterns and places unprecedented strain on urban centres. As rural populations experience violence, towns and cities become hubs of refuge, creating sudden population surges that intensify competition for food, housing, employment, and basic services (Gebrihet & Gebresilassie, 2025a). In urban economies, where food security depends heavily on functioning markets and stable purchasing power, displacement-induced pressure often deepens existing vulnerabilities, producing new and complex forms of urban food insecurity (Gebrihet et al., 2025). The cumulative impact of prolonged conflicts entrenches vulnerability across generations, creating cycles of hunger, poverty, and social dislocation (Muriuki et al., 2023). Understanding how households navigate these pressures is crucial for developing evidence-based policies to address food insecurity and support sustainable post-conflict recovery.

Traditionally, food insecurity has been examined through structural factors such as poverty, land scarcity, environmental stress, and demographic pressure (Ellis & Mdoe, 2003; Gebre & Rahut, 2021). Recent scholarship underscores the significant effects of shocks, particularly armed conflicts. Empirical studies from South Sudan, the Democratic Republic of Congo, Yemen, and Syria demonstrate that conflict diminishes productive assets, reduces labour availability, and limits access to markets and services (Büscher, 2018; FAO, 2024; Hashim et al., 2021; Ibrahim et al., 2024; Jacobs & Kyamusugulwa, 2018; Muriuki et al., 2023; Sassi, 2021). Conflict transforms vulnerability through various pathways, including forced displacement, destruction of infrastructure, and erosion of social networks (Geremedhn & Gebrihet, 2024). The literature on resilience reveals that households respond heterogeneously, employing strategies such as migration, asset liquidation, income diversification, and reliance on informal community support (Demisse et al., 2024; Mossie et al., 2024). Despite this evidence, existing research largely focuses on national or regional aggregates, often overlooking household-level dynamics in localized conflict settings. For example, Gebrihet et al.’s (2025) study on food insecurity and coping strategies in the war-affected urban areas of Tigray, Ethiopia, did not identify the principal drivers of household vulnerability, particularly the role of displacement pressures and conflict-related losses of assets and livelihoods. Consequently, the mechanisms through which urban households in Tigray experience and adapt to food insecurity under displacement conditions remain underexplored.

The armed conflict in northern Ethiopia, particularly in Tigray, has produced one of the most acute displacement crises in Horn of Africa (Geremedhn & Gebrihet, 2024). Since 2020, successive waves of violence have forced hundreds of thousands of rural residents into urban centres such as Adwa, Axum, and Shire (Gebrihet et al., 2025). While these towns offered relative refuge from direct hostility, they were rapidly overburdened by the mass influx of displaced populations. Local food markets, already destabilized by warfare and blockade, confronted rising demand that outstripped supply, drove up prices, and further constrained availability. Host households, many of whom had suffered conflict-related livelihood losses themselves, were compelled to share scarce resources with displaced families, thereby deepening their own exposure to food insecurity.

This study has four interrelated objectives. First, it assessed the extent of household food insecurity in the urban setting of Tigray during the war. Second, it identifies the principal drivers of household vulnerability, paying particular attention to displacement pressures and conflict-related losses of assets and livelihoods. Third, it situates the Tigray experience within broader debates on conflict, displacement, and food insecurity, offering empirical insights into its relevance to other war-affected urban contexts. Finally, it advances policy recommendations designed to promote sustainable recovery, strengthen household resilience, and reduce vulnerability to food insecurity in both the immediate and long-term aftermath of the war.

2. Literature Review

Food insecurity remains a persistent challenge in conflict-affected regions, undermining livelihoods and socioeconomic development. The following section synthesis the conceptual foundation, theoretical framework, and empirical review of the study, identifies thematic patterns and gaps, and situates the study within scholarly debates on vulnerability and resilience in conflict settings.

2.1. Conceptual Framework

The conceptual framework for this study situates urban household food insecurity within the context of displacement caused by war in Tigray, Ethiopia. Armed conflict is the main driver of disruption because it destroys assets and livelihoods while forcing large numbers of people to move into urban areas (Araya & Lee, 2024; Weldegiargis et al., 2023). Displacement creates sharp increases in demand for food, housing, and basic services, placing additional pressure on host communities that were already weakened by the collapse of markets and the loss of income (Geremedhn & Gebrihet, 2024). Urban households, which rely more on markets and cash income than on self-production, become particularly vulnerable when supply chains fail and food prices rise under the weight of population pressure (Gebrihet et al., 2025; Teodosijević, 2003; Weldegiargis et al., 2023).

Food insecurity, the dependent variable in this framework, refers to the lack of consistent access to safe, adequate, and nutritious food. It is understood across several dimensions: availability, access, utilization, stability, agency, and sustainability (Banks et al., 2021; Su & Amrit, 2024). Urban host households are highly exposed because they depend on wages, remittances, or small-scale trade to secure food. When displaced populations arrive, competition for jobs increases, wages decline, and living costs rise, reducing households’ purchasing power. At the same time, disruptions to trade routes and supply chains restrict the availability of food, creating price increases and market instability. Together, these pressures reduce households’ ability to maintain adequate and stable consumption, producing both immediate and longer-term food insecurity.

Urban host households respond to these pressures by adopting a range of coping strategies. Some reduce meal frequency and dietary diversity, while others borrow food or money, rely on humanitarian aid, engage in informal work, or share resources within extended family and community networks. These responses provide short-term relief, but their effectiveness diminishes when displacement persists, and humanitarian needs remain widespread. As both displaced and host populations depend on the same limited urban resources, coping mechanisms quickly become overstretched, leaving many households in chronic or recurring food insecurity.

In this framework, displacement caused by war is the central independent variable. It undermines urban food security both directly and indirectly by increasing demand, straining markets, and weakening household entitlements. Food insecurity, the dependent variable, will be analyzed across the dimensions of availability, access, affordability, and stability. The framework therefore emphasizes how displacement reshapes urban vulnerabilities and influences the resilience of host households as they attempt to maintain access to adequate and nutritious food during prolonged conflict. Figure 1 presents the causal pathways linking armed conflict, displacement, and urban household food insecurity in Tigray, Ethiopia.

Armed conflict triggers mass displacement, which exerts economic, social, and institutional pressures on urban host households. These mediating factors, such as income loss, market disruption, gendered vulnerabilities, and asset depletion, increase the likelihood of food insecurity across multiple dimensions (availability, access, affordability, and stability). Humanitarian assistance and social capital can partially mitigate these effects, though their influence weakens under protracted siege conditions. The framework emphasizes the dynamic interactions between conflict, displacement, and household resilience in shaping urban vulnerability.

2.2. Theoretical Framework

Household vulnerability frameworks conceptualize food insecurity as a function of exposure to shocks, sensitivity of livelihoods, and adaptive capacity (Ellis, 2000; Moser, 1998). In the context of war-induced displacement, urban host households face shocks not from agricultural collapse but from sudden increases in population pressure, market disruption, unemployment, and rising living costs. Their sensitivity stems from their heavy dependence on market-based food access, with limited buffer mechanisms such as savings or social protection. Adaptive capacity is further constrained, as community support systems, humanitarian assistance, and informal safety nets are stretched thin by the simultaneous needs of both displaced and host populations.

Sen’s Entitlement Theory remains particularly relevant in explaining food insecurity under urban displacement (Sen, 1983). In Tigray’s urban centres, these entitlements have been severely disrupted: local labour markets collapsed, small businesses closed, and remittance flows were interrupted. Blockades and restrictions on humanitarian aid further limit entitlement (Clark, 2021). Displacement magnifies these pressures by intensifying competition for jobs and services, thereby reducing host households’ capacity to convert entitlements to food security (Gebrihet & Gebresilassie, 2025b; Geremedhn & Gebrihet, 2024). While Entitlement Theory explains access-based vulnerabilities, Devereux (2001) critiques them for underplaying political dynamics. In war-affected contexts, such as Tigray, state actions, aid restrictions, and broader governance failures can amplify household vulnerability, creating structural barriers to securing food.

The Sustainable Livelihoods Framework (DFID, 2008) provides a holistic lens by linking food security to households’ access to, and use of, diverse livelihood assets. In the urban host communities of Tigray, displacement has eroded multiple forms of capital: financial capital is weakened by unemployment and inflation; social capital is strained as hosts share resources with displaced families; physical capital, including housing, markets, and water systems, is overstretched; and political capital is undermined by governance failures and restrictions on aid (Natarajan et al., 2022). The SLF, when adapted to displacement contexts, illustrates how conflict reshapes the urban asset base and constrains households’ abilities to cope and adapt.

The human security theory posits that food security is integral to overall well-being, emphasizing how conflict simultaneously threatens physical, economic, and social survival (Fouinat, 2004; Jolly & Ray, 2006). Political economy perspectives complement this view by demonstrating how war and displacement reconfigure power relations in urban spaces, privileging households with access to humanitarian aid or remittances while marginalizing poorer host families. Structural barriers, such as disrupted markets, high food prices, and limited employment opportunities, further exacerbate inequality in food access (Geremedhn & Gebrihet, 2024; Israni, 2024).

Resilience theory adds insight by examining how households adapt to displacement pressure (Folke et al., 2010; Shilomboleni et al., 2024). In Tigray’s host communities, coping strategies include reliance on food aid, engagement in informal employment, borrowing, and dietary adjustments (Eshetu et al., 2025; Gebrihet & Gebresilassie, 2025a). However, resilience is uneven: households with diversified incomes, stronger networks, or external support adapt more effectively, while poorer households quickly exhaust their coping mechanisms. Prolonged displacement and conflict undermine recovery pathways, resulting in chronic food insecurity for large segments of urban populations.

These theoretical perspectives provide a comprehensive lens for understanding urban food insecurity in war-affected Tigray. They illustrate how armed conflict and displacement disrupt entitlements, strain urban assets, and limit adaptive capacity while also highlighting the mechanisms through which host households attempt to cope and build resilience. By integrating entitlement theory, livelihood approaches, human security perspectives, and resilience thinking, this study situates empirical investigation within a robust conceptual and theoretical foundation. This enables a nuanced understanding of how conflict-driven displacement interacts with structural, economic, and social factors to shape urban host communities’ vulnerability to food insecurity.

2.3. Empirical Review

Empirical studies highlight that food insecurity in conflict settings emerges from multiple determinants including socioeconomic status, education, gender, and social capital (d’Errico et al., 2023; Giannini et al., 2021; Sileshi et al., 2019; Gebrihet & Gebresilassie, 2025c). In urban contexts, displacement adds a critical layer to these vulnerabilities by increasing the demand for food, housing, and services, thereby driving up prices and intensifying competition for scarce jobs (Büscher, 2018; Jacobs & Kyamusugulwa, 2018). Host households with limited incomes or weak social networks are particularly disadvantaged as they must navigate market disruptions while accommodating displaced populations. This dynamic reflects findings from South Sudan, the Democratic Republic of Congo, and Yemen, where urban host communities face rising food insecurity following large inflows of displaced people (FAO, 2024; Hashim et al., 2021; Ibrahim et al., 2024).

Socioeconomic status, education, asset ownership, and livelihood diversification strengthen household resilience. d’Errico et al. (2023) found that households with diversified incomes and higher education levels exhibit a greater capacity to withstand shocks across 35 countries. Giannini et al. (2021) showed that livelihood diversification reduces climate-related vulnerability in Senegal. Gendered dynamics remain central to the understanding of household vulnerability (Gebrihet & Gebresilassie, 2025c). Similarly to rural settings, women-headed households in urban Tigray face disproportionately higher risks of food insecurity due to limited access to stable employment, credit, and social protection (Gebrihet & Gebresilassie, 2025b). Displacement compounds these challenges, as women in host communities often assume caregiving responsibilities for displaced relatives while managing fragile household food budgets (Geremedhn & Gebrihet, 2024). Studies from Syria and Yemen have also demonstrated how displacement reshapes gendered vulnerabilities, placing additional burdens on women to secure food and income under conditions of scarcity and disrupted services (Hashim et al., 2021; Ibrahim et al., 2024).

Urban households in host communities are highly exposed to food insecurity because of their dependence on markets rather than on self-production. In Tigray, disrupted trade routes and repeated blockades destabilized food supply chains, while displacement-driven population growth intensified demand, leading to sharp price increases and reduced affordability (Eshetu et al., 2025; Weldegiargis et al., 2023). Evidence from Shire and Mekelle illustrates that host households experienced declining purchasing power, even when food was available, as competition with displaced populations depressed wages and inflated the costs of living. Similar outcomes were observed in urban centres in South Sudan and the Democratic Republic of Congo, where displacement magnified market instability and eroded household-coping capacity (Büscher, 2018; Jacobs & Kyamusugulwa, 2018).

Coping strategies adopted by urban host households mirror broader patterns documented in conflict literature but are shaped by the unique pressures of displacement. In Tigray, these include reducing meal frequency and dietary diversity, borrowing money or food, relying on humanitarian assistance, and engaging in informal or precarious employment (Eshetu et al., 2025; Gebrihet & Gebresilassie, 2025a). However, as displaced and host households draw on the same overstretched resources, their coping mechanisms quickly weaken, leading to chronic food insecurity. Comparative studies from Yemen, Syria, and South Sudan confirm that prolonged displacement undermines host community resilience by exhausting both social support networks and humanitarian response capacity (FAO, 2024; Kafando & Sakurai, 2024; WFP, 2023).

Within Ethiopia, urban households are already vulnerable to food insecurity because of reliance on markets, inflation, and limited social protection (Gebresilassie & Nyatanga, 2023). Rural households depend on markets, face limited financial services, and confront environmental stressors such as locust infestations (Gebrihet & Gebresilassie, 2025a, 2025b). Structural vulnerabilities increase exposure when shocks disrupt supply chains, transport, or income-generating activities (Araya & Lee, 2024; Gebrihet et al., 2025). Social safety nets, income diversification, and livelihood interventions support resilience. Evidence from other regions of Ethiopia demonstrates that cash transfers, community support systems, and technological innovations reduce both chronic and shock-induced vulnerability (Sileshi et al., 2019).

Tigray illustrated the compounding effects of conflict on household food insecurity. Prior to 2020, most areas were food secure (IPC Phase 1) or stressed (Phase 2) owing to soil rehabilitation, irrigation, and safety net programs (Clark, 2021). The conflict between 2020 and 2023 disrupted agricultural production, destroyed infrastructure, depleted livestock, and restricted market access. Rural households face 77% food insecurity, while urban households experience 39% vulnerability, with female-headed households disproportionately affected (Araya & Lee, 2024; Gebrihet et al., 2025).

Dependence on rural agricultural production amplifies conflict-induced urban food insecurity. Eshetu et al. (2025) report that urban households rely on external markets, reducing immediate exposure, but maintaining vulnerability through disrupted supply chains. Conflict eroded traditional coping mechanisms such as social networks, savings, and community support. Evidence from Burkina Faso, South Sudan, Syria, Yemen, and Nigeria reinforces rural vulnerability to agricultural destruction, market disruption, and asset loss (FAO, 2024; George et al., 2021; Hashim et al., 2021; Kafando & Sakurai, 2024; WFP, 2023).

Despite extensive research on the relationship between conflict and food insecurity, important gaps remain. Most studies focus on national aggregates and overlook the specific vulnerabilities of urban host communities that accommodate displaced populations. Few investigations examine how displacement pressures, combined with asset destruction and livelihood disruption, shape food insecurity at the household level. Evidence from Tigray, which experienced one of the most acute sieges and humanitarian blockades in Africa, is particularly limited.

The Tigray case also presents a distinct pattern of displacement and urban food insecurity. Initially, urban residents moved to rural areas as conflict intensified in cities. Later, as fighting spread to rural areas, urban centres such as Shire, Adwa, and Axum became hosts to large numbers of displaced rural households. These urban populations were already struggling with livelihood losses and reliance on unstable markets. The regional siege and communication blackout worsened conditions by severing supply chains, halting remittances, and limiting humanitarian access. This combination of multi-directional displacement, market collapse, and isolation created a complex environment that intensified household food insecurity.

This study addresses these gaps by examining urban household vulnerability to food insecurity in post-war Tigray. Studying Tigray therefore contributes to a deeper understanding of the urban dimensions of vulnerability in protracted conflicts and highlights the importance of context-specific approaches in addressing food insecurity. The use of the Household Food Insecurity Experience Scale (FIES) and ordered logit regression provides a rigorous way to quantify household vulnerability and identify structural and displacement-related determinants.

3. Methodology

3.1. Description of the Study Area

Tigray is an Ethiopian regional national state. It was bordered by Sudan, Eritrea, Afar, and Amhara. It is divided into seven administrative zones: Northwestern, Western, Southern, Southeastern, Mekelle, Eastern, and Central. These zones are further subdivided into 93 Weredas (districts) and 814 ‘Tabia’ (the smallest administrative unit) (Clark, 2021; Araya & Lee, 2024). Seasonal labour and smallholder agriculture are the two main pillars of rural food security in this region. Most of the population resides in rural areas as smallholder farmers, rearing livestock, and growing cereals (Clark, 2021). The household food security of the Tigray has been significantly affected by the protracted armed war. A map of the study area (Tigray) is shown in Figure 2.

3.2. Source, Type, and Method and Data Collection

To examine the relationship between war-induced internal displacement and urban household vulnerability to food insecurity in war-affected Tigray, this study relied on primary cross-sectional, household-level survey data collected from urban households between May and June 2024. Data were collected using a structured questionnaire that was specifically designed for this study. The questionnaire was first developed in English, drawing on established measures of household vulnerability to food security, and was then translated into Tigrinya, the predominant local language of the region. This translation was prudently reviewed to confirm equivalence and to facilitate accurate understanding by both respondents and enumerators. The structured questionnaire was pretested on a small sample of urban households, and adjustments were made to improve its reliability, clarity, and appropriateness.

The data collectors received rigorous training prior to conducting the household survey, covering research (survey) ethics, interview techniques, and the technical aspects of the questionnaire. Supervision was conducted to minimize data collection and entry errors and ensure data quality. Written ethical approval was obtained, including informed verbal consent, from each participant, and confidentiality was strictly observed throughout the data collection process.

3.3. Sampling Techniques and Sample Size

To obtain the required number of sample households (

Table 1. Proportional sample size across the three towns in %(n).

Towns	Female	Male
Adwa	33.55 (102)	66.45 (202)
Axum	33.84 (112)	66.16 (219)
Shire	48.15 (195)	51.85 (210)
Total	39.33 (409)	60.67 (631)

Source: Authors’ computation. Note: Values in parentheses are number of household heads.

), this study employed a four-stage- sampling procedure. In the first stage, two zones (central and northwestern) are selected. Owing to resource constraints, this study focused on two urban zones (the central and northwestern zones). In the second stage, three towns–Adwa, Axum, and Shire–were randomly selected. In the third stage, three ‘Tabias’ (the smallest administrative units) were selected randomly using a lottery method. Finally, 1040 households were randomly selected using systematic sampling. The sample size of households was computed using the sample size determination formula developed by Yamane (1967), as follows:

$$n={\displaystyle \frac{N}{1+N\left(e^2\right)}}$$

(1)

where $N$ (=138,667) is the target finite household population, $n$ is the required sample household size, $e$ is the marginal error (0.05), and 1 is a constant. Data were collected using a structured questionnaire to examine gender disparities in the prevalence of food insecurity in war-torn Tigray, Ethiopia, using appropriate estimation techniques. The response rate was 100%; no missing response occurred.

3.4. Methods of Data Analysis

This study employed both descriptive and inferential statistical analyses to analyze household-level survey data collected from a sample of urban households.

3.4.1. Descriptive Statistical Analysis

A descriptive statistical analysis was employed to examine the characteristics of the sample of urban households. In addition, inferential statistical analyses, such as chi-square and t-tests, were employed to assess the existence of associations between continuous and categorical variables and outcome variables of interest.

3.4.2. Measurement and Modelling Approach

This study used two complementary statistical estimation techniques, namely, the Household Food Insecurity Experience Scale (FIES) and ordered logistic regression estimation techniques, which are specified below.

3.4.3. Household Food Insecurity Experience Scale (FIES) Measures

Household food insecurity can be computed using two primary estimation techniques: objective and subjective (Maxwell et al., 2003). Previous empirical studies have employed objective measures of food insecurity, relying on household food expenditure data to examine household food security status (Araya & Lee, 2024; Gebresilassie & Nyatanga, 2023; Weldegiargis et al., 2023). By contrast, subjective assessments capture individuals’ perceptions and lived experiences of food insecurity, offering nuanced insights into the challenges of accessing food.

This subjective measure of food security has gained prominence among sociologists and economists as a valuable complement to objective measures of food security (Gebresilassie, 2020; Gebresilassie & Nyatanga, 2023; Weldegiargis et al., 2023). Hence, this study used the Household Food Insecurity Experience Scale (FIES), developed by Ballard et al. (2013). FIES, a subjective food security measurement technique, is an experience-based metric that captures direct information from households regarding food access problems. It included eight standard questions covering different levels of food insecurity. Respondents reported whether, in the past one months, the households had experienced situations such as worrying about food, eating less preferred foods, skipping meals, or going without food for an entire day.

The FIES is a prominent measure of household food security and provides a valuable framework for classifying and understanding food security levels (Coates et al., 2007). It measures household food insecurity experiences using a series of eight dichotomous questions, with yes-or-no questions over the past 30 days about households’ behaviours toward, and experiences with, inadequate food (Coates et al., 2007; FAO et al., 2023). The estimated FIES scores ranged from 0 to 8, with higher scores indicating higher severity of household food insecurity (

Table A2. FIES measures of households’ food insecurity.

Measures of Food Security	Category Values	Household Categories	Cut-Off Points
FIES	0	Food-secure	[0]
	1	Mildly food-insecure	[1, 3]
	2	Moderately food-insecure	[4, 6]
	3	Severely food-insecure	[7, 8]

Source: Authors’ compilation from literature.

). The list of the eight series of questions of the FIES is presented in Appendix A 

Table A1. Questions in the FIES.

FIES Questions	During the Last 30 Days, Was There a Time When:
$Q_1$	Were you worried when you would not have enough food to eat?
$Q_2$	Were you unable to eat healthy and nutritious food?
$Q_3$	Did you eat only a limited variety of meals?
$Q_4$	Had you to skip a meal?
$Q_5$	Did you eat less than you thought you should?
$Q_6$	Did your household run out of food?
$Q_7$	Were you hungry but did not eat?
$Q_8$	Did you go without eating for a whole day?

Source: Adapted from Coates et al. (2007).

.

A household ending food security is measured based on survey data on the eight-standard experience-based Food Insecurity Experience Scale (FIES) questions (Coates et al., 2007; FAO et al., 2023) as described in Table A1. The metric has been widely used in the analysis of food insecurity (Gebrihet et al., 2025; Gebrihet & Gebresilassie, 2025a, 2025b). Based on the household’s yes/no responses to the eight FIES questions and adapting the FAO’s global reference scale of FIES (Coates et al., 2007; FAO et al., 2023), the household’s food security status has been classified into four categories (See Table A2). A household is food secure (=0): If the household responded ‘no’ to all the eight questions, i.e., if $Q_1$= /$Q_2$= /$Q_3$= /$Q_4$= /$Q_5$= /$Q_6$= /$Q_7$= /$Q_8$= /0; mildly food insecure (=1): If the household responds ‘yes’ to at least one of the first three FIES questions (i.e., if $Q_1$ = 1 or $Q_2$ = 1, or $Q_3$ = 1) and ‘no’ to the rest of the FIES questions (i.e., $Q_4$ = $Q_5$ = $Q_6$ = $Q_7$ = $Q_8$ = 0), Moderately food insecure (=2): If the household responded ‘yes’ to either $Q_4$ or $Q_5$ and ‘no’ to $Q_6$, $Q_7$, and $8$; severely food insecure (=3): If the household responded ‘yes’ to one or more of the last three FIES questions (i.e., $Q_6$ = 1 or $Q_7$ = 1 or $Q_8$ = 1), respectively.

The sum of the “frequency of occurrence” of the eight dichotomous food insecurity conditions over the past 30 days was computed as follows:

$$FIES={\sum }_{i=1}^8(Q_{1,a}+Q_{2,a}+Q_{3,a}\dots , Q_{8,a})$$

(2)

where $Q_{1,a}\_Q_{8,a}$ refer to the “frequency-of-occurrence” question (once or twice, sometimes, or often) and $x_i$ refers to the $i\mathrm{t}\mathrm{h}$ sample household.

3.4.4. Order Logistic Regression Analysis: Model Specification

To examine the drivers of urban household vulnerability to food insecurity, we employed an ordered logit regression estimation technique based on the estimates (categories) of the FIES approach. The dependent variable (outcome variable) was obtained from the computed FIES estimates, which classified households into four categories: food secure-(0), mildly food insecure (1), moderately food insecure (2), and severely food insecure-(3).

Accordingly, the ordered logit regression model (ordinal regression) is used as an appropriate estimation technique to predict an ordinal dependent variable given one or more explanatory variables. Explanatory variables can be continuous, categorical, or ordinal (Agresti, 2010; Long & Freese, 2001). When the outcome variable of interest (the dependent variable) is ordinal, the appropriate estimation technique is the ordered (ordinal) regression estimation model. Hence, to identify the factors influencing household vulnerability to food insecurity, this study employs an ordered logit regression estimation model as the dependent variable (household food insecurity status).

We assumed the existence of a latent or unobserved continuous variable (${FIES}_i^{*}$) that reflects the underlying propensity of the $i$th household to experience vulnerability to food insecurity. The latent variable model is specified as follows:

$${FIES}_i^{*}={\alpha }_0+{\alpha }_i{X}_i+{\epsilon }_i {\epsilon }_i~ \mathrm{Normal} (0,1)$$

(3)

where ${FIES}_i^{*}$ is the unobserved latent variable representing the underlying propensity for vulnerability to food insecurity for the $i$th household. $X_i$ is a vector of independent variables, ${\alpha }_i$ is the parameter vector to be estimated, and ${\epsilon }_i$ is the error term assumed to follow a standard logistic distribution.

Let ${\theta }_1,<{\theta }_{2,}<$ ${\theta }_3$, be the cut-off points (unknown threshold parameters), and define the probability of household vulnerability to food insecurity status in Equation (3) were computed as follows:

$${FIES}_i^{*}=\left\{\begin{array}{c}1 if {FIES}_i^{*}\le {\theta }_1, Food-Insecure\\ 1 {if \theta }_1<{FIES}_i^{*}\le {\theta }_2, Mildly Food-Insecure\\ 2 if {\theta }_2<{{FIES}_i^{*}\le \theta }_3, Moderately Food-Insecure\\ 3 if {FIES}_i^{*}>{\theta }_3, Severely Food-Insecure\end{array}\right.$$

(4)

where ${\theta }_1,$ ${\theta }_2$, and ${\theta }_3$ refer to the cut-off points or threshold level computed using FIES estimates, Equation (2).

In the ordered logistic regression model, we estimate the cumulative probability instead of the probability of an individual event. Hence, the cumulative probability that the $i$th sample household falls into the $j$ category (a response of less than or equal to $j$) is given as follows:

$$logit=P({FIES}_i^{*}\le j)={\displaystyle \frac{\exp \left({\alpha }_0+{{\alpha }_{j}X}_j\right)}{1+\exp \left({\alpha }_0+{{\alpha }_{j}X}_j\right)}}={\alpha }_0-{{\alpha }_{i}X}_i for j=1,\dots ,j-1$$

(5)

The logarithm of the probability that the $i$th sample household falls into the $j$th category using Equation (5) is computed as follows:

$$\log \left({\displaystyle \frac{p({FIES}_i^{*}\le j)}{p({FIES}_i^{*}>j)}}\right)={\alpha }_0-{\alpha }_j{X}_i, jϵ[1, j-1]$$

(6)

where ${\alpha }_j$ is the intercept and log is the log-odds of falling into category $j$ or below.

3.5. Description of Study Variables

Households were classified into male- and female-headed households to examine the drivers of food insecurity disparities between the two groups. The dependent variable in this study was a dummy variable indicating food security or insecurity based on FIES scores, with 1 representing food security and 0 representing food insecurity. Mild, moderate, and severe levels of insecurity were aggregated into household food insecurity using FIES scores. Independent variables (covariates), based on theoretical and literature evidence as well as data availability, were included as a set of independent variables, as presented in

Table A3. Description and definitions of variables.

Variables	Definition	Description	Expected Signs (Directions)
Dependent variable
Food insecurity	Urban household vulnerability to food insecurity	It is a categorical variable, 0 = food-insecure; 1 = mildly food-insecure; 2 = moderately food-insecure; and 3 = severely food-insecure.
Independent variables
Sex of the HH head	Sex of the household head	It is a dummy variable, 1 = Female; 0 otherwise.	Positive
Age of the HH head	Age of the household head	It is a continuous variable measured in years.	Negative
HH size (persons)	Number of household members	It is a continuous variable measured as the total persons in the household.	Negative
Marital status of the HH head	Marital status of the household head	It is a categorical variable, 1 = single; 2 = married; 3 = divorced; and 4 = widowed.	Positive/Negative
Education of the HH head	Educational attainment of the household head	It is a categorical variable, 0 = illiterate; 1 = primary school; 2 = high school; and 3 = above high school.	Negative
Religion of the HH head	Religious affiliation of the household head	It is a dummy variable, 1 = Christian; 0 otherwise.	Positive/Negative
Employment status of the HH head	Employment status of the household head	It is a dummy variable, 1 = employed; 0 otherwise.	Negative
Average income per capita	Average annual income per household member	The average annual income per capita, which is a continuous variables measured in Ethiopian Birr.	Negative
Number of IDP in a host HH	War-induced internally displaced persons (IDP) residing in the household	It is a continuous variable measured as number of IDPs in a host household.	Positive
Access to credit	Household’s access to credit facilities	It is a dummy variable, 1 = access to credit; 0 otherwise.	Negative
Humanitarian aid (Food aid recipient)	Household’s receipt of humanitarian aid	It is a dummy variable, 1 = received humanitarian aid; 0 otherwise.	Negative
Damage due to armed war	Households’ loss of life or productive assets during the armed war	It is a dummy variable, 1 = experienced loss or damage; 0 otherwise.	Positive

Note: HH refers to household.

in the Appendix A.

The dependent variable in this study was a categorical variable indicating insecurity based on the FIES scores, with 0, 1, 2, and 3 representing for food security, mildly, moderately, and 3 severely food insecurity, respectively.

Armed conflict: Exposure to conflict or armed war. This variable was selected based on theory and previous studies in the Tigray region of Ethiopia and other conflict-affected contexts (Gebrihet et al., 2025; Gebrihet & Gebresilassie, 2025a, 2025b).

The number of IDP per household: The number of war-induced internally displaced persons (IPD) who are dependent on their relatives, friends, and other guardians. It is the main outcome variable of interest in this study and was adopted based on George & Adelaja (2022). The definitions and descriptions of the variables used in this study are presented in Table A3 in the Appendix A.

3.6. Data Processing and Analysis

The data collected from the sample households were entered into Microsoft Excel and subsequently exported to Stata version 17 for analysis. The findings are presented in the Figures and Tables.

4. Results

4.1. Descriptive Statistics

Table 2. Descriptive statistics of categorical variables (%).

Variables	Adwa	Axum	Shire	Average	${\mathit{\chi }}^2$ Value
Gender of the household head
Female	24.21	21.4	24.3	71.94	4.13 ***
Male	77.81	78.6	75.7
Marital status of the HH head
Single	0.00	2.6	8.4	8.13
Married	66.79	73.4	78.6	74.67
Divorced	6.23	0.0	0.7	6.4
Widowed	22.13	19.7	9.3	9.4	2.64 ***
Education of the HH head
None	8.21	14.9	16.7	15.6	9.13 ***
Primary (Grade 1–6)	55.79	66.1	48.3	51.3
Secondary (Grade 9 = 12)	28.72	18.6	31.6	25.4
Tertiary & above	7.33	0.4	3.4	7.7
Religion of the HH head
Christian	96.32	99.1	89.9	97.4	4.32 ***
Muslim	3.68	0.9	10.2	2.6
Employment status of the HH head
Employed	46.53	41.1	47.6	42.2	3.67 ***
Unemployed	53.47	58.9	52.4	57.8
Access to credit
Yes	35.47	33.9	28.4	32.8	5.73 ***
No	64.53	66.1	71.6	67.2
Food aid recipient
Yes	64.53	56.1	64.2	61.3	3.15 ***
No	35.47	43.9	35.8	38.7
Damage due to armed conflict
Yes	75.72	81.13	74.14	76.51	8.63 ***
No	24.38	18.87	25.86	23.49
Total	28.4 (159)	32.5 (182)	39.1 (219)	100 (560)

Source: Authors’ computation. Note: HH refers to a household. *** p < 0.01.

presents the descriptions of the variables used in this study and the complete (100%) sample household response rate. Approximately 72% of the households were male-headed and 28% were female-headed. Most households were married (75%) and 9.4% were widowed. Most household heads (51%) had completed primary school, followed by 25% who had completed secondary school. Approximately 97% of the respondents were Christian and the remaining 3% were Muslim. Importantly, about 58% of the household heads were unemployed, while the remaining 42% were employed, with significant variations across towns; the highest unemployment rate was in Axum (59), followed by Adwa (54%) and Shire Towns (52%). Most households (61%) received humanitarian aid (food aid) to address their acute food insecurity, with slight variations across towns. Furthermore, most of the sample households (76.51%) had lost either the lives of members of their households or productive assets due to armed war (damage due to armed war), while the remaining 23.49% did not lose their productive assets or the lives of members of their household.

Table 3. Summary of statistics of continuous variables.

Variables	Mean/Share	Standard Deviation
Age of the HH head	53.25	11.42
HH size (persons)	4.31	1.85
Average annual income per capita	5309.23	32.028
Number of IDP in a host HH	2	3.32

Source: Authors’ computation. Note: HH refers to a household.

summarizes the key household characteristics (continuous variables). On average, the respondents’ age was 53 years, while the household size was four persons, which is almost equal to the national average of 4.4 according to the 2007 population census. On average, in Tigray, the average number of war-induced internally displaced persons per host household is two (number of IDP in a host HH) during the post-war period, implying that an urban household has hosted, on average, two internally displaced persons in their house; they could be their relatives or support them living within the households. Furthermore, on average, the annual income per capita of the sample household was 53,029 Ethiopian Birr.

4.2. Analysis of Urban Households’ FIES Indicators

Table 4. Percentage share of household Food Insecurity status using FIES, %(n = 560).

Towns	Food-Secure	Mildly Food-Insecure	Moderately Food-Insecure	Severely Food-Insecure	Total Food-Insecure
Adwa	2.71 (20)	4.73 (35)	7.97 (59)	6.08 (45)	19.59 (145)
Axum	3.65 (27)	5.27 (39)	8.78 (65)	6.89 (51)	20.14 (149)
Shire	4.46 (33)	6.22 (46)	10.14 (75)	8.78 (65)	25.13 (186)
Mean	14.29 (80)	21.43 (120)	35.54 (199)	28.75 (161)	85.71 (480)

Source: Authors’ computations. Note: Values in parentheses are the number of households.

presents the FIES estimates of urban household vulnerability to food insecurity in war-ravaged towns in Tigray, Ethiopia. The FIES analysis results show that approximately 14% of the sample households (one-sixth) were not vulnerable (food secure). Most households (86%) were vulnerable to food insecurity with varying levels of severity across towns. This figure shows that eight out of the ten sample households are vulnerable to food insecurity, indicating the severity of food insecurity in the war-ravaged urban settings of Tigray. Approximately 25% were found in Shire, followed by 20% each in Axum and Adwa.

On average, approximately 29% of households were found to be vulnerable to severe food insecurity, of which approximately 9% (the largest) were found in Shire Town. Similarly, approximately 36% of households are vulnerable to moderate food insecurity, of which the largest proportion (10%) is found in Shire Town. Furthermore, approximately 21.43% of the households were vulnerable to mild food insecurity; of this total, the highest (6.22%) was found in Shire Town, followed by Axum (5.27%) and Adwa (4.73%). These figures reveal that the largest proportion of the sample households are food insecure, with significant differences across towns.

4.3. Drivers of Household Vulnerability to Food Insecurity

The dependent variable was household vulnerability to food insecurity (food secure = 0, mildly food insecure = 1, moderately food insecure = 2, severely food insecure = 3). An ordered logit regression analysis was conducted to examine the drivers of war-induced internal displacement of households vulnerable to food insecurity in host communities in war-ravaged Tigray. The results of the regression analysis are presented in

Table 5. Drivers of household vulnerability to food insecurity.

Variables	Coef. (Std.Err)	Odds Ratio	p-Value	[95% Conf. Interval]
Gender of the household head (Ref. Female)	0.452 ** (0.194)	1.571	0.020	[1.073454	2.299511]
Age of the household head	−0.046 (0.062)	1.047	0.457	[0.9271673	1.183103]
Religion of the household head (Ref. Muslim)	−0.238 (1.019)	0.788	0.815	[0.1068529	5.814547]
Access to credit (Ref. = No)	−0.058 (0.066)	0.944	0.381	[0.8300968	1.073844]
Household size	0.011 (0.008)	1.656	0.177	[0.9948571	1.028441]
Household head level of education (Ref. = No)
Primary	0.001 (0.001)	0.997	0.359	[0.9993411	1.001821]
High school	−1.229 ** (0.619)	0.999	0.047	[−2.444667	−0.0151158]
Above high school	−0.0140 (0.008)	1.014	0.079	[0.9984001	1.030044]
Marital status of the head of household
Married	−0.602 (0.949)	0.999	0.526	[0.0852446	3.522719]
Divorced	0.021 (0.1887)	1.277	0.912	[0.7057512	1.477095]
Widowed	0.012 * (0.007)	1.153	0.090	[0.9980983	1.026703]
Average annual income per capita	−0.428 ** (0.201)	1.535	0.033	[1.034914	2.275162]
Number of IDP in a HH	0.419 ** (0.197)	1.521	0.034	[0.0326462	0.805243]
Food aid recipient (Ref. = No)	−0.031 * (0.169)	1.032	0.067	[0.9977811	1.066324]
Damage due to armed conflict (Ref. = No)	0.389 ** (0.195)	1.475	0.046	[1.00738	2.160692]
/cut1	−1.947 (5.669)			[−13.05697	9.163964]
/cut2	−0.700 (5.661)			[−11.79417	10.39361]
/cut3	0.371 (5.662)			[−10.72587	11.46682]
$\mathrm{Prob} > {\mathsf{\chi }}^2$	0.0431
$\mathrm{Pseudo} {\mathrm{R}}^2$	0.3375 ***
Number of Obs.	560

Source: Authors’ computations. Note: *** p < 0.01, ** p < 0.05, * p < 0.1. Values in parentheses are standard errors (Std. Err), and Ref. where denotes the reference category.

.

Accordingly, the results indicate that being a female-headed household has a positive and statistically significant effect on household vulnerability to food insecurity, with an odds ratio of 1.571, while maintaining the other factors constant. Female-headed households face higher risks, reflecting both gendered vulnerabilities and the loss of male labour during armed wars. The years of schooling of the household head was negatively associated with household vulnerability to food insecurity, highlighting the protective role of education in coping strategies, resource allocation, and restoration of lost livelihoods. The results indicate that the education status of the household head has a positive and significant effect on household vulnerability to food insecurity, implying that educated heads of households have relatively better food security than their non-educated counterparts. An educated household head can seek means to help the family address food insecurity. Hence, a household head who has completed high school has a negative and significant effect on household vulnerability to food insecurity, with an odds ratio of 0.999 when compared to a household head who has no formal education, while other factors remain constant.

Regarding the marital status of the head of the household, widowed women had a positive and statistically significant effect on household vulnerability to food insecurity, with an odds ratio of 1.153, while other factors remained constant, indicating that widowed women had the worst vulnerability to food insecurity compared to single and married women. Income is the most important factor that affects household vulnerability to food insecurity. The results indicated that the average annual per capita income of the household was negatively associated with household vulnerability to food insecurity, with an odds ratio of 1.535. Furthermore, the results show that the number of war-induced internally displaced persons (the number of IDP in a host HH) is positively associated with household vulnerability to food insecurity. This result indicates that, by controlling for other covariates, this variable is significantly associated with household food insecurity vulnerability. For each additional member of the internally displaced persons per household, the likelihood of the household moving into a higher food insecurity category increased by 1.521 factors. Similarly, a household that received humanitarian aid was negatively associated with vulnerability to food insecurity, with an odds ratio of 1.032. This implies that households that received humanitarian aid had a relatively better food security status than their non-humanitarian aid counterparts. Importantly, household vulnerability to food insecurity among those who had experienced damage to the household member or loss of productive assets due to armed conflict was more vulnerable to food insecurity than those who had not experienced damage due to armed war, which increased by a factor of 1.475.

These results underscore the structural and war-induced internally displaced drivers of household vulnerability to food insecurity in the war-ravaged urban settings of Tigray. Loss of assets or life (damage due to armed war), war-induced displacement, gender, widowed women, and income disruption increase household vulnerability to food insecurity, whereas education, humanitarian aid, and per capita income appear to be protective factors.

5. Discussion

This study highlights the severity of food insecurity among urban households in post-war Tigray, Ethiopia, where conflict-induced displacement has reshaped vulnerability dynamics in host communities. Drawing on cross-sectional data from 560 households, only 14.29% of households were food secure, whereas the remaining 85.71% experienced varying levels of insecurity: 21.43% mild, 35.54% moderate, and 28.75% severe. Using the Household Food Insecurity Experience Scale (FIES) and ordered logit regression, we identified key determinants of vulnerability, including female-headed households, widowed status, lower education, reduced per capita income, higher numbers of internally displaced persons (IDPs) per household, and war-related damage. Protective factors include higher education, particularly the completion of high school, and access to humanitarian aid. These findings directly respond to our research objectives by quantifying urban food insecurity in host communities and revealing how displacement interacts with structural vulnerability to exacerbate hunger.

These results underscore the multifaceted toll of the Tigray conflict on urban livelihoods. The predominance of moderate-to-severe insecurity (64.29% combined) reflects the systemic erosion of household entitlements, where war disrupts market access, income streams, and social safety nets. Female-headed and widowed households exhibited heightened vulnerability, with odds ratios of 1.571 and 1.153, respectively, highlighting the gendered burden of caregiving amid male fatalities or conscriptions (Gebrihet & Gebresilassie, 2025b). Interestingly, household size did not significantly affect vulnerability (p = 0.177), diverging from expectations in the resource dilution models. This reflects urban Tigray’s reliance on shared informal networks among extended families, which mitigates per capita pressures but can collectively increase the strain on host households.

The positive association between IDP numbers and vulnerability (odds ratio 1.521) illustrates a “contagion effect,” where each additional displaced person increases competition for food and income, stressing household coping capacity. The protective role of education (odds ratio 0.999) suggests that literacy facilitates income diversification and access to aid, whereas the modest effect of humanitarian support (odds ratio 1.032) indicates partial relief and highlights gaps in coverage. War-related damages (odds ratio 1.475) further entrench poverty as asset loss constrains recovery potential.

These findings resonate with and extend the existing literature on conflict and food insecurity. Our prevalence estimates align with humanitarian reports indicating 39% urban vulnerability at the conflict’s peak (Araya & Lee, 2024; Gebrihet et al., 2025), but our post-war focus reveals persistent systemic insecurity. The displacement-vulnerability link corroborates evidence from Yemen, Syria, and South Sudan, where sudden population inflows in host cities increase food prices and depress wages (FAO, 2024; Hashim et al., 2021; Ibrahim et al., 2024). However, this study adds value by quantifying household-level mechanisms using ordered logit regression, addressing gaps in research that often aggregate data regionally or emphasize rural shocks (Muriuki et al., 2023; Sileshi et al., 2019).

Gendered vulnerability patterns align with Devereux’s (2001) study, highlighting how political and institutional barriers exacerbate women’s marginalization in conflict zones (Geremedhn & Gebrihet, 2024). Unlike multi-country analyses suggesting universal benefits from livelihood diversification (d’Errico et al., 2023), our findings show that education has an outsized role in urban Tigray, where market-dependent livelihoods make literacy essential for informal trade recovery (Gebresilassie & Nyatanga, 2023). The limited efficacy of aid is consistent with Maxwell et al. (2014), who observed that rations stabilize households but rarely rebuild resilience, particularly in regions affected by protracted blockades (Clark, 2021).

Theoretically, this research advances vulnerability frameworks by integrating war-induced displacement into the sustainable livelihood framework (DFID, 2008), demonstrating how conflict depletes urban capital: financial (income loss), social (overstretched networks), and physical (infrastructure damage). By focusing on host communities, this study shifts the discourse from displaced persons as sole victims to shared urban precarity, reinforcing human security perspectives that frame food access as being inseparable from broader economic and social survival.

The findings reveal that displacement pressure substantially magnifies household food insecurity in urban Tigrays. Host families, already weakened by conflict-related livelihood disruptions, face heightened risks when compelled to share scarce resources with displaced populations. The observation that each additional internally displaced person (IDP) in a household increases insecurity risk by 52% underscores the compounding strain displacement imposed on fragile urban economies. This pattern resonates with broader evidence that urban displacement, unlike rural displacement, creates market-based rather than subsistence-driven food insecurity, as sudden demand shocks overwhelm disrupted supply systems (Glass et al., 2017).

The concentration of vulnerability among female- and widowed-headed households further illustrates how war reshapes social and economic roles to deepen precarity. Limited access to income-generating opportunities in conflict conditions, combined with caregiving burdens, leave these households particularly exposed. This aligns with Haysom’s (2021) argument that conflict not only reduces material entitlements, but also undermines social support networks that might otherwise mitigate shocks.

Taken together, these findings highlight the importance of examining displacement not only as a humanitarian challenge, but also as a structural driver of urban food insecurity. By tracing how war-induced population movements reshape household economies, this study advances debates that have traditionally focused on national- or regional-level aggregates while overlooking localized urban dynamics. This contributes to a growing recognition that urban centres are not merely sites of refuge but also frontlines of vulnerability, where displacement, market disruption, and social fragmentation converge.

6. Conclusions and Policy Implications

This study was guided by four interrelated research questions concerning the extent, determinants, and implications of urban household food insecurity in post-war Tigray. The analysis first assessed the level of food insecurity among urban households, then identified the socioeconomic and displacement-related drivers of vulnerability, compared the findings with evidence from other conflict-affected contexts, and finally considered the policy responses required for recovery. Drawing on primary cross-sectional data from 560 households and applying the Household Food Insecurity Experience Scale (FIES) alongside ordered logit regression, we found widespread insecurity: only 14% of households were food secure, while 86% faced mild to severe vulnerability. Key determinants include female- and widowed-headed households, lower education and income, higher numbers of hosted internally displaced persons (IDPs), and war-related asset losses. Higher education and humanitarian assistance have emerged as important buffers against food insecurity.

These findings highlight the compounded effects of conflict on urban livelihood. The displacement strains the market, reducing access to resources and intensifying competition within host households, exacerbating vulnerability. This study underscores the overlooked needs of host communities in conflict zones, revealing that interventions focused solely on rural areas or displaced populations may be insufficient. By emphasizing urban settings, this study offers guidance for more equitable recovery strategies in post-conflict regions.

These implications extend beyond Tigray, suggesting the need for integrated urban support systems that rebuild financial, social, and physical resources. Addressing food insecurity in host communities can prevent poverty cycles, strengthen resilience, and enhance overall human security. Future research should consider longitudinal approaches to track household recovery trajectories and incorporate institutional factors, such as governance and aid distribution, to provide deeper insights into resilience mechanisms.

This study highlights the urgent need for displacement-sensitive policies that prioritize gendered vulnerabilities, restore lost assets, and promote sustainable urban food security. Addressing these challenges is essential to fostering recovery, resilience, and long-term stability in post-conflict settings.

In line with these findings, the following important policy implications are suggested.

• The Regional Government of Tigray (social welfare office), in collaboration with the WFP and NGOs, should support urban households that host at least one IDP.
• The Regional Education Bureau, supported by UNICEF and local NGOs, should implement accelerated high school completion programs and vocational training in small-scale trade, tailoring, and food processing for the heads of households with low education.
• Microfinance institutions and NGOs specializing in post-conflict recovery should provide microloans for livestock restocking, small businesses, or household asset replacement to households that lost their assets during the conflict, coupled with financial literacy workshops to ensure the effective use of funds.
• The WFP, the Ethiopian Red Cross, and local NGOs should scale up food and nutritional support for severely food-insecure households.
• The Ministry of Urban Development, local municipal authorities, and the UN-Habitat should reconstruct damaged homes and basic infrastructure to enable safe IDP returns without overburdening host households.
• The Regional Trade and Industry Office, with NGO support, should provide grants or low-interest loans to urban traders and small business owners who have lost assets during the conflict to restore functional marketplaces.
• The local urban food security task force coordinated by the Regional Government should track aid delivery, monitor household vulnerability, and report monthly food insecurity trends.

7. Strengths and Limitations

This study provides valuable insights into gender-based disparities in household food insecurity in war-affected urban settings in Tigray, Ethiopia. To our knowledge, this is among the first surveys to explore gender disparities in the prevalence of food insecurity in this region using primary cross-sectional data collected between May and June 2024. The relatively large sample size enhances the reliability of our findings and allows for a detailed analysis of key vulnerability factors. However, this study had several limitations.

• This study relied on self-reported data, including the Household Food Insecurity Experience Scale (FIES), which may be subject to recall or reporting bias.
• The cross-sectional design limits the ability to track household vulnerability over time and constrains causal inferences regarding the relationship between food insecurity and its determinants.
• Institutional quality indicators, which could significantly influence household vulnerability, were not included because of resource constraints, thus limiting insights into how governance and institutional factors affect food insecurity in this context.
• While this study identifies key drivers of household vulnerability to food insecurity, it does not examine the effect of market price exposure, distance to distribution points, and local governance constraints due to data unavailability. Additionally, it does not examine potential interactions or combined effects among the variables included in the regression analysis.

8. Future Research Directions

Household vulnerability to food insecurity in the aftermath of the protracted armed conflict in Tigray is a complex multidimensional problem that requires multidisciplinary approaches. Future research should consider the following.

• The impact of armed conflict on health outcomes, including malnutrition, child mortality, and nutritional security, as well as on migration, unemployment, and labour market dynamics in war-affected urban areas.
• Employ longitudinal study designs to capture household recovery trajectories and resilience over time.
• Incorporate institutional quality indicators as independent variables to assess their effects on household vulnerability to food insecurity.
• Spatial analysis techniques, including satellite imagery, were used to map patterns of household vulnerability and recovery, providing a broader geographic perspective on food insecurity.