As the most food-insecure region in Africa, East Africa faces persistent challenges in ensuring adequate food supply, as climatic fluctuations, political instability, and economic hardships continue to undermine its ability to meet the dietary needs of its growing population. While previous research has predominantly examined climate change and political instability as primary drivers of food insecurity, the influence of rural development on food security outcomes remains insufficiently explored. Hence, this study investigates the relationship between rural development and food security while incorporating climatic and socio-economic factors using panel data from 12 countries between 2001 and 2020. Employing heterogeneous panel cointegration techniques, the findings derived from PCSE and FGLS estimators reveal that rainfall substantially improves food availability and utilization while diminishing food accessibility and stability. In contrast, higher temperatures negatively affect all four dimensions of food security. Moreover, population growth exerts a significant negative influence on food availability and stability, while food imports enhance food availability but simultaneously reduce accessibility and utilization. Furthermore, political stability is crucial in strengthening food availability and stability, whereas rural development significantly boosts food availability, accessibility, and utilization. Nevertheless, the Dumitrescu–Hurlin panel tests indicate bidirectional predictive linkages between population growth and food security, and a unidirectional linkage from temperature to food security. These findings propose targeted recommendations for East African authorities to strengthen food security policies and resilience.
rural developmentfood securityclimate changepolitical stabilityfood inflationfood importsThe author(s) declare that no financial support was received for the research and/or publication of this article.section-at-acceptancePeace and DemocracyIntroduction
The battle against hunger has emerged as a resounding global call for immediate and decisive action. As the world’s population continues to grow, meeting the dietary needs of billions presents an ever-pressing challenge. In 2021, an alarming 828 million people worldwide faced hunger, and projections suggest these numbers will rise if immediate action is not taken (FAO, IFAD, UNICEF, WFP and WHO, 2022). Food security, defined as a condition where all individuals have consistent access to nutritious, affordable, and safe food aligned with their dietary preferences, is crucial for human health, social stability, and economic development. Ensuring food security guarantees fundamental human rights, such as access to clean water and proper nutrition, while fostering productivity and economic progress (Abdi et al., 2024a, 2024b; Conceição et al., 2016). However, the four pillars of food security—accessibility, availability, stability, and utilization—are increasingly threatened by climate change, political instability, food price inflation, and underdeveloped rural areas. Weak governance in food-insecure regions further limits the capacity to mitigate and adapt to food security challenges, while strong institutional frameworks are critical for policy implementation and resilience-building (Connolly-Boutin and Smit, 2016; Yiadom et al., 2023).
Climate change poses a dual threat to global food security. On the one hand, it directly affects agricultural productivity through erratic rainfall, rising temperatures, water scarcity, altered soil conditions, pest outbreaks, and shifting crop growth timings (Chandio et al., 2023; Mubenga-Tshitaka et al., 2023). The Intergovernmental Panel on Climate Change [IPCC] reports that Earth’s average temperature has increased by 1.1 °C since the industrial era, with projections exceeding 1.5 °C within the next two decades (IPCC, 2021). These changes result in more frequent droughts and floods, land degradation, and declining agricultural yields (Adhikari et al., 2015). Climate change has already caused a 10% reduction in global agricultural yields between 2000 and 2019 (FAO, 2023). Regions such as tropical and subtropical areas, including East Africa, are particularly vulnerable due to their dependence on rain-fed agriculture and limited adaptation capacities (Kumar et al., 2021; Ntiamoah et al., 2022). On the other hand, climate change exacerbates food insecurity indirectly by fueling conflicts over dwindling natural resources such as land and water (Abdi et al., 2023a, 2023b). Competition for these resources often intensifies communal disputes, particularly between farmers and pastoralists migrating in search of better opportunities (Delgado et al., 2023; Warsame et al., 2023; Yiadom et al., 2023).
Conflicts and political instability further aggravate food insecurity, particularly in regions like East Africa. Recurring conflicts disrupt food production, displacement rates increase, and access to food supplies becomes severely limited (World Food Programme, 2024). According to the Global Report on Food Crises (2023), conflicts pushed over 117 million people into acute food insecurity globally. Moreover, conflicts influence food prices (Hussein et al., 2021). The Russia-Ukraine geopolitical crisis, for instance, disrupted wheat exports—both countries account for 25% of global wheat shipments—leading to sharp increases in food prices and reduced purchasing power (FAO, 2022; Karume et al., 2024). Consequently, food inflation rates exceeded 5% in low-income countries, deepening inequalities and worsening poverty (Abdi et al., 2024a, 2024b; World Bank, 2022). On the other hand, rural development plays a pivotal role in alleviating food insecurity by enhancing agricultural productivity and building resilience to climate and political shocks. Globally, 9.1% of the population experienced hunger in 2023, up from 7.5% in 2019, with rural communities bearing the brunt of this burden (FAO, 2024). Rural areas, home to 80% of the poorest populations, are characterized by limited access to infrastructure, education, and healthcare, which exacerbates food insecurity (Abdi et al., 2024a, 2024b; IFAD, 2021). Improved rural infrastructure, such as road networks and irrigation systems, facilitates access to markets, reduces post-harvest losses, and enables the adoption of climate-smart agricultural practices (Ali Warsame and Hassan Abdi, 2023; Chandio et al., 2022). However, rural development remains inadequate in many regions, leaving rural communities particularly vulnerable to hunger and poverty.
East Africa, a region marked by high climate vulnerability and persistent conflicts, faces immense challenges in addressing its food security crisis. With a population growth rate of 5.8%, it accounts for 44% of Africa’s food-insecure population, making it the most vulnerable on the continent (FAO, 2021). Rain-fed agriculture, which contributes 43% to the region’s GDP and sustains 80% of its population, is highly susceptible to climate-induced shocks such as droughts, floods, and locust infestations (Adhikari et al., 2015; Mubenga-Tshitaka et al., 2023). Climatic disruptions, including erratic weather patterns, water shortages, and pest outbreaks, have severely impacted crop yields, with projections showing declines of up to 72% for wheat, 45% for maize and rice, and substantial losses in other staple crops (Adhikari et al., 2015). Locust infestations alone have devastated 5,000 hectares of pastureland in Djibouti (World Bank, 2020). The region’s dependence on food imports further exacerbates its vulnerability. Countries like Ethiopia (39%), Djibouti (55%), Somalia (91%), and Eritrea (100%) relied heavily on Russian and Ukrainian wheat in 2021, making them susceptible to supply chain disruptions and price surges caused by geopolitical tensions (FAO, 2022; Karume et al., 2024). Such challenges disproportionately affect low-income households, which exacerbates food insecurity and deepening socio-economic disparities. As illustrated in Figure 1, rural development funding in Eastern Africa presents steady growth from 2000 to 2013. However, subsequent fluctuations demonstrate ongoing underinvestment, which hinders the region’s ability to strengthen agricultural resilience and infrastructure. Additionally, East Africa’s high fertility rate, averaging four to six children per woman, places further pressure on limited domestic production, which perpetuates cycles of poverty and malnutrition (World Bank, 2022).
Annual rural development funding in Eastern Africa (2000–2021).
Bar chart showing rural development funding in millions of U.S. dollars from 2000 to 2020. Funding increases from 2000 to 2008, peaks in 2016, and fluctuates slightly thereafter.
The adverse effects of climate change on food security have been extensively studied, particularly in vulnerable regions like SSA, where socioeconomic, political, and adaptive capacity limitations exacerbate food insecurity (Connolly-Boutin and Smit, 2016). Climate change has been linked to reduced agricultural productivity, as evidenced by Mubenga-Tshitaka et al. (2023), who found that rising temperatures negatively impacted East Africa’s agricultural yields in the long-term, while precipitation influenced productivity primarily in the short-term. Fagbemi et al. (2023) conceptualized that the change in climatic patterns had exacerbated the ever-increasing hunger in SSA countries. Similarly, Abdi et al. (2023a, 2023b) observed that increasing temperatures were negatively correlated with cereal production in East Africa, whereas precipitation enhanced yields. However, the impact of climate change on food security varies depending on the crops grown. For instance, Abdi et al. (2024a, 2024b) found that rising temperatures reduced the production of sorghum, rice, banana, and beans but increased sugarcane yields, while Adhikari et al. (2015) identified wheat, maize, rice, and soybeans as highly vulnerable to climatic shifts, with sweet potato, potato, and cassava being more resilient. Furthermore, extreme weather events such as droughts and floods significantly exacerbate food insecurity across Africa (Yiadom et al., 2023). They also affirmed the role of institutional quality in lessening these effects. Delgado et al. (2023) highlighted how climate-induced food insecurity has deepened inequality and social divides across the continent. In addition to these challenges, Misra (2014) reported the depletion of natural resources in SSA due to climate change and advocated for sustainable measures such as Soil Aquifer Treatment (SAT) and artificial groundwater recharge to address the growing crisis.
Changes in food prices significantly influence both the quantity and quality of food people can access, particularly in developing countries where a substantial share of household income is spent on food (Chandio et al., 2023). Rising food prices often reduce food security by forcing families to stretch their budgets, which makes it harder to afford adequate and nutritious meals (Zhou and Wan, 2017; Chavas, 2017). Recent global crises, such as the COVID-19 pandemic and the Russian invasion of Ukraine, have further exacerbated food insecurity in low- and middle-income countries, with higher global food prices driving up the cost of staple foods, especially in nations reliant on imports (Zereyesus et al., 2023; Okou et al., 2022). However, countries with stronger local food production tend to experience less inflation in staple food prices, which marks the importance of agricultural self-sufficiency (Okou et al., 2022). Inflation also plays a critical role in shaping food availability, as it disrupts prices and limits farmers’ access to inputs, leading to lower crop yields (Mumuni and Joseph Aleer, 2023). Although inflation has less impact on other dimensions of food security, lower inflation levels have been associated with reduced undernourishment, as stable prices enable households to purchase sufficient food (Candelise et al., 2021). Furthermore, Abdi et al. (2024a, 2024b) found that the effects of higher food prices on food security in 32 SSA countries were mixed, with both positive and negative outcomes depending on local contexts.
Political stability is paramount in shaping food security outcomes, as effective governance is deeply intertwined with food systems. The literature asserts that stable political institutions enhance food availability, accessibility, and utilization through supportive policies and efficient implementation. Zhou and Wan (2017) found that good governance boosts agricultural productivity and food distribution, thereby improving food security. Similarly, Smith and Haddad (2015) observed that stable governments are better equipped to sustain nutrition programs, combat food deficiencies, and promote healthier diets. In Somalia, Maxwell (2013) documented how political instability between 2011 and 2012 disrupted food systems, which exacerbated vulnerabilities and worsened food insecurity. On a broader scale, Soffiantini (2020) revealed that political unrest in Arab countries hindered food production and distribution, which led to widespread shortages and nutritional deficits. Abdullah et al. (2020), using a dynamic panel model of 124 countries from 1984 to 2018, found that political stability, democratic accountability, and good governance are positively associated with improved food and nutrition security. Conversely, they observed that conflict, corruption, and military involvement in politics undermine food security. Deaton and Lipka (2015) further demonstrated that political stability attracts investment in agricultural infrastructure and social safety nets while fostering policies that address food security challenges.
Many studies have proposed that rural development delivers an integral solution to the challenges posed by climate change, political instability, and food inflation by enhancing food security through improved infrastructure and technological advancements (Abdi et al., 2025). Physical infrastructure, such as roads and storage facilities, boosts agricultural performance and ensures food availability. Upgraded rural road networks reduce food system costs, which makes goods more affordable and accessible (Edeme et al., 2020; Pivoto et al., 2018). Fedderke and Bogetić (2009) reiterated that physical infrastructure significantly enhances labor productivity, which fosters sustainable growth and agricultural resilience. In Nigeria, Adepoju and Salman (2013) found that road access, soil quality, and extension services positively affect farmers’ productivity, while Boakye (2019) noted that inadequate storage facilities in Ghana result in more than half of cultivated crops failing to reach consumers, which stresses the necessity for improved warehousing. Similarly, in India, Lokesha and Mahesha (2017) demonstrated that improved rural roads facilitate crop diversification and profitability. Beyond physical infrastructure, access to electricity and ICT amplifies the impact of rural development by enhancing market connectivity and providing critical agricultural information (Abdi et al., 2025). Ezeoha et al. (2020) found that mobile phone and internet access improve food security by increasing market linkages, while Candelise et al. (2021) showed that access to electricity enhances food availability and nutrition in 54 developing countries.
East Africa’s food security crisis is shaped by the interconnected challenges of climate change, political instability, and economic pressures, yet the potential of rural development as a transformative solution remains underexplored. The region’s reliance on non-irrigated agriculture makes it particularly vulnerable to extreme weather events and erratic rainfall patterns that significantly reduce crop yields (Abdi et al., 2023a, 2023b; Adhikari et al., 2015; FAO, 2021). Political instability further exacerbates these issues, with governance failures and conflicts disrupting food systems and preventing the implementation of sustainable agricultural policies (Maxwell, 2013; Soffiantini, 2020). Additionally, the sharp rise in global food prices, driven by crises such as the Russia-Ukraine conflict, has strained household affordability, particularly in import-dependent nations that source their food externally (FAO, 2022). While substantial research has examined the links between climate change, conflicts, and food security (Adhikari et al., 2015; Delgado et al., 2023), rural development—through improved infrastructure, access to technology, and market connectivity—suggests an exhaustive pathway to lessen these vulnerabilities. This study aspires to bridge the gap by investigating how rural development, alongside climatic and socio-economic factors, influences food security in East Africa. Utilizing empirical data from 2000 to 2021, the analysis strives to present actionable insights into creating a resilient and sustainable food system for the region. To find reliable and region-specific outcomes, this study employs robust econometric methodologies, including panel-corrected standard errors (PCSE), feasible generalized least squares (FGLS), and the Dumitrescu and Hurlin panel causality test, to address limitations in prior research by accounting for heterogeneity and cross-sectional dependencies across East African countries.
This study makes a significant contribution to the existing body of knowledge on food security in East Africa by addressing critical research gaps and employing innovative methodologies. Initially, while much of the existing literature focuses on isolated factors such as climate change, political instability, or food price inflation, this study adopts an integrated approach to examine their combined effects alongside rural development. By integrating diverse variables—including climate change, governance quality, and rural infrastructure—this research proposes pioneering insights into the varied determinants of the region’s food security landscape. Subsequently, while previous studies primarily illustrate the challenges faced by food security, including climate change, political instability, and economic pressures, they often overlook the role of rural development as a climate resilience strategy. This study addresses this gap by demonstrating how investments in rural development, such as improved infrastructure, electricity access, and ICT, can mitigate the adverse effects of climate change, enhance agricultural productivity, and stabilize food systems. By positioning rural development as a solution rather than just scrutinizing the obstacles, the research shifts the narrative toward actionable strategies for resilience and sustainability. Rather than treating food security as a singular dependent variable, this study investigates its core components—availability, access, utilization, and stability—to uncover how climate change, political stability, and rural development influence these factors. By disaggregating food security, the research provides a profound understanding of the specific mechanisms through which environmental and economic variables impact different dimensions of food security. Finally, the study equips policymakers with evidence-based recommendations to foster rural development, promote climate-smart agriculture, and build resilient food systems in East Africa and beyond.
Materials and methodologyData sources and food security measurement framework
This study investigates the relationship between rural development and food security in East Africa, incorporating climatic and socio-economic factors. The analysis utilizes a balanced panel dataset spanning 12 East African countries—Democratic Republic of Congo, Ethiopia, Kenya, Madagascar, Malawi, Mozambique, Rwanda, Somalia, Tanzania, Uganda, Zambia, and Zimbabwe—covering the period from 2001 to 2020. These countries were selected based on their exposure to climatic volatility, institutional fragility, and macroeconomic shocks, which heighten their vulnerability to food insecurity. The dependent variable is food security, operationalized through a composite index that captures four core dimensions: availability, accessibility, stability, and utilization. These dimensions are represented by heterogeneous indicators—such as average dietary energy supply adequacy, GDP per capita, prevalence of undernourishment, food production variability, child malnutrition rates, and access to water and sanitation. To consolidate these variables into a single index, the study employs Principal Component Analysis (PCA), a data-reduction technique widely used in development economics and public health research to handle multidimensional constructs (Greco et al., 2019; Jolliffe, 2002). PCA transforms correlated indicators into uncorrelated principal components by capturing the maximum variance within the data. While PCA assumes linear relationships and enforces orthogonality, these assumptions serve the analytical purpose of simplifying interpretation and mitigating multicollinearity (Abdi and Williams, 2010).
Prior to extraction, all variables were standardized to address differences in units and scales to ensure that no single dimension dominates the composite measure due to measurement units. Justification for component retention follows a robust criterion, combining the Kaiser rule (eigenvalues >1), scree plot analysis, and cumulative variance thresholds—as recommended by Hair et al. (2019)—to ensure meaningful representation and minimize information loss. Though PCA’s limitations—such as its linearity assumption and orthogonality constraint—are acknowledged, the method remains a practical and theoretically grounded tool for summarizing complex constructs like food security. This is further supported by applications in similar contexts, such as Headey and Ecker (2013), who developed a PCA-based food security index for SSA, and Zezza et al. (2017), who used PCA to analyze household food security. More recently, Hamadjoda Lefe et al. (2024) applied PCA to construct a food security index for 40 SSA countries. On the other hand, the independent variables include climatic factors—mean annual rainfall and average temperature (from the Climate Change Knowledge Portal)—and socio-economic indicators such as rural development, food imports, food price levels, population growth, and political stability. Data were sourced from authoritative databases, including the Food and Agriculture Organization (FAO), World Development Indicators (WDI), and the Worldwide Governance Indicators (WGI). Table 1 provides a detailed overview of the variables utilized in the analysis and their corresponding data sources. Equations 1–4 present the mathematical formulation of the PCA-based index across the four food security dimensions. This relationship is represented as:
FSit=θ1AVAILit+θ2ACCit+θ3STABit+θ4UTILit
where FSit represents the food security index for country i at time t. The availability dimension, AVAIL, captures the adequacy of dietary energy supply, while accessibility, ACC, accounts for economic access to food. Stability, STAB, represents the resilience of food supply chains, and utilization, UTIL, assesses the nutritional and health-based quality of food. The coefficients θ1,θ2,θ3, and θ4 define the relative importance of each dimension in contributing to overall food security.
Variables description and data sources.
Variable
Code
Indicator
Source
Food security indices
Food security
FS
Availability
ADESA
Average dietary energy supply adequacy (percent) (3-year average)
FAO
Accessibility
GDPPC
Gross domestic product per capita, PPP, (constant 2017 international $)
FAO
PU
Prevalence of undernourishment (percent) (3-year average)
NPU
Number of people undernourished (million) (3-year average)
Stability
FPV
Per capita food production variability (constant 2014–2016 thousand int.$ per capita)
FAO
PFSV
Per capita food supply variability (kcal/cap/day)
Utilization
CHW
Percentage of children under 5 years affected by wasting (percent)
CHS
Percentage of children under 5 years of age who are stunted (modelled estimates) (percent)
FAO
CHO
Percentage of children under 5 years of age who are overweight (modelled estimates) (percent)
WAT
Percentage of population using at least basic drinking water services (percent)
SAN
Percentage of population using safely managed sanitation services (Percent)
Climatic indicators
Mean rainfall
RF
Average annual rainfall (mm)
CCKP
Mean temperature
TEM
Average annual temperature in (°C)
CCKP
Macroeconomic variables
Food prices
FPI
Food price inflation (value %)
FAO
Population growth
PG
Population growth (annual %)
WDI
Food imports
FM
Imported food excluding fish (value, millions USD)
FAO
Rural development
RD
Development Flows to Agriculture (Disbursement millions USD)
FAO
Political stability
PS
Political Stability and Absence of Violence/Terrorism: Estimate
WGI
Specifically, this study delineates food availability via the average dietary energy supply adequacy (ADESA). Moreover, food accessibility is formulated as a function of economic and nutritional access indicators, given by:
Accessibilityit=ω1GDPPCit+ω2PUit+ω3NPUit
where GDPPC denotes gross domestic product per capita, which reflects the economic capability of individuals to afford food. PU represents the prevalence of undernourishment, which indicates the proportion of the population consuming inadequate calories. NPU denotes the number of people undernourished by offering a numerical measure of food deprivation. The coefficients ω1,ω2, and ω3 capture the respective influences of these factors on food accessibility. Moreover, food stability is modeled to account for fluctuations in food production and supply, expressed as:
Stabilityit=δ1FPVit+δ2FSVit
where FPV represents per capita food production variability, which measures inconsistencies in agricultural output and FSV denotes per capita food supply variability, which reflects fluctuations in food availability due to external shocks. The coefficients δ1 and δ2 determine the extent to which these variables influence food stability. Food utilization is captured through multiple nutritional and hygiene-related factors, expressed as:
where CHW represents the percentage of children under five affected by wasting, an indicator of acute malnutrition, while CHS reflects the percentage of children under five affected by stunting, measuring chronic malnutrition. CHO accounts for the percentage of children under five affected by overweight, which highlights imbalances in nutritional intake. Additionally, WAT represents access to basic drinking water services, ensuring safe hydration, and SAN refers to safely managed sanitation services. The coefficients γ1,γ2,γ3,γ4, and γ5 measure the relative impact of these variables on food utilization.
Econometric model framework
Building on the conceptual foundations outlined by Abdi et al. (2024a, 2024b), Fagbemi et al. (2023), and Hamadjoda Lefe et al. (2024), this study models the functional relationship between climate change and food security in East African countries. The initial framework can be expressed in Equation (5) as follows:
FS=f(RF,TEM,W)
In this formulation, RF signifies average annual rainfall, TEM represents mean annual temperature, and W encompasses a set of control variables that potentially influence food security outcomes in the region. To address the broader research objectives, the model is extended to include key macroeconomic and institutional determinants. The extended model takes the following form (see Equation (6)):
FS=f(RF,TEM,FPI,PG,FM,RD,PS)
where FPI denotes food price inflation, PG indicates population growth, FM represents food imports, RD captures rural development, and PS accounts for political stability. This broader formulation allows for a more comprehensive assessment of both climatic and non-climatic determinants of food security in East Africa. To operationalize the empirical analysis, three model specifications are developed. The first model examines the effects of climate change on food security, while also controlling for selected economic influences. This is specified in Equation (7):
FSit=α0+φ1RFit+φ2TEMit+φ3FPIit+φ4PGit+φ5FMit+εit
where α0 is the intercept, φ1 through φ5 are the coefficients of the explanatory variables, and εit is the error term capturing unobserved factors. The second model introduces rural development to evaluate its direct contribution to food security, modifying the baseline model as show in Equation (8):
Based on theoretical expectations, RF, FM, RD, and PS are expected to positively influence food security. In contrast, TEM and FPI are anticipated to exert negative effects due to their adverse impacts on food production and affordability. The expected effect of PG is ambiguous, as it can increase food demand and pressure agricultural systems, yet potentially improve accessibility through labor supply expansion and market dynamics. The panel dataset covers 12 East African countries (i=1,2,…,N) over the period 2001 to 2020 (t=2001,2002,…,T).
Empirical techniqueCross-sectional dependence
A critical challenge in panel data analysis, with profound implications for estimation precision and inference validity, is the presence of interdependencies among individual units (Sarafidis and Wansbeek, 2012). These interdependencies often emerge from shared economic networks, spatial proximity, and deep-rooted cultural ties, making cross-sectional dependence (CSD) a probable concern in the East African context (Abdi et al., 2023a, 2023b). Economic integration through trade fosters financial interlinkages, geographical contiguity amplifies spillover effects, and cultural affiliations enhance social and economic interconnectedness. Ignoring CSD risks producing biased and unreliable estimates, which ultimately weakens the robustness of empirical findings (Sarkodie and Owusu, 2020). To address this, a preliminary diagnostic step involves testing for CSD among the panels under investigation. A widely employed technique for this purpose is the Lagrange Multiplier (LM) test, initially formulated by Breusch and Pagan (1980), along with its scaled variant proposed by Pesaran (2004). These tests assess the null hypothesis of cross-sectional independence—implying no contemporaneous correlation in the error terms across cross-sectional units—against the alternative hypothesis of significant interdependence.
Slope heterogeneity test
The second critical issue in panel data analysis is the determination of the homogeneity of slope coefficients. If the assumption of homogeneity is made without empirical validation, it can result in the omission of country-specific idiosyncrasies (Bedir and Yilmaz, 2016). Swamy (1970) introduced a test for slope homogeneity that assesses the dispersion of individual slope estimates from an appropriate pooled estimator. Pesaran and Yamagata (2008) argue that both the F-test and Swamy’s test require panel data models with a relatively small cross-sectional dimension (N) compared to the time dimension (T). To circumvent this limitation, they proposed a standardized version of Swamy’s test, referred to as the Δ˜ test, for assessing slope homogeneity in large panels. The Δ˜ test is asymptotically valid as (N, T) → ∞, without any constraints on the relative rates of increase of N and T under the assumption of normally distributed error terms. Pesaran and Yamagata (2008) subsequently derived the following standardized dispersion statistic (see Equation 10):
Δ˜=N(N−1S˜−k2K)
where k signifies the number of regressors, and S˜ represents Swamy’s statistic. Under the null hypothesis, with the assumption that (N, T) → ∞ and the error terms are normally distributed, the Δ˜ test adheres to an asymptotic standard normal distribution. To refine the small sample properties of the Δ˜ test, assuming normally distributed errors, one can employ the following bias-adjusted mean and variance (see Equation 11):
Δ˜adj=N(N−1S˜−E(Z˜iT)Var(Z˜iT))
In this context, E(Z˜iT)=k, and Var(Z˜iT)=2k(T−k−1)/(T+1).
Second-generation unit root test
Once cross-sectional dependence has been identified, the next crucial step is to examine the stationarity of variables and determine their order of integration. Traditional first-generation unit root tests operate under the assumption that panel units are independent across sections; however, this assumption is often violated when cross-sectional dependence exists (Breitung and Das, 2008). Neglecting this dependence can lead to misleading inferences and spurious regression results. To overcome this limitation, Pesaran (2007) introduced second-generation panel unit root tests, such as the cross-sectional augmented Dickey-Fuller (CADF) test and its extended variant, the cross-sectional augmented IPS (CIPS) test. These advanced techniques explicitly account for interdependencies among panel units and mitigate the influence of unobserved common factors. While both tests follow a similar methodological framework, the CIPS test offers an enhanced approach by incorporating cross-sectional averages. Fundamentally, these tests evaluate whether the data support the alternative hypothesis of stationarity or whether the null hypothesis of a unit root persists. The CIPS procedure builds upon the CADF model, aggregating the CADF statistics across cross-sections to compute the final CIPS statistic. The CADF test statistic is formally expressed as shown in Equation (12):
ΔYi,t=αi+βiYi,t−1+γiY¯t−1+δiΔY¯i,t+εit
where Y¯t−1 and ΔY¯i,t denote the cross-sectional averages of the lagged values and the first differences of the individual series, respectively. The computation of the cross-sectional mean follows Equation (13).
Y¯t−1=1N∑i=1NΔY¯i,t
The CADF statistic is obtained by averaging the individual CADF statistics CADFi as illustrated in Equation (14):
CIPS=1N∑i=1NCADFi
where CADFi represents the t-statistics derived from the CADF regression outlined in Equation (12).
Panel cointegration test
Before proceeding with the main analysis, it is essential to determine whether a long-run equilibrium relationship exists among the variables. This is accomplished through panel cointegration tests, specifically those developed by Pedroni (1999, 2004) and Kao (1999). Pedroni test is notable for handling differences between panels by including specific fixed effects and time trends in the cointegration regression. This approach allows the autoregressive (AR) coefficient to vary from one panel to another. Here’s a look at how the Pedroni panel cointegration method is typically set up. The general structure of the Pedroni panel cointegration equation is presented in Equation (15):
Yit=ρi+ω1iXit+ω2iX2it+……ωpiXpit+εit
In this setup, ρi and ωi represent the intercepts and slope coefficients, which can differ between cross-sections. The assumption is that Y, X, and p all have the same integration order of I(1). According to the null hypothesis, which suggests that there is no cointegration, the residuals εi would be integrated at I(1). If the p-value is significant at the 1%, 5%, or 10% levels, we would reject the null hypothesis and conclude that a cointegration relationship is present.
Panel model approach
FGLS, introduced by Parks (1967), was the first method used to examine the relationships among the study variables. This static panel technique is especially suited for long-term analysis. When dealing with non-spherical errors and an unknown covariance matrix, FGLS helps estimate the coefficients of a multiple linear regression model and their covariance matrices. The FGLS process involves creating an inverse matrix of the estimated variance–covariance matrix (EVCM) of the errors, which transforms the regression equation. FGLS is appropriate for panels with T greater than N (Biswas et al., 2022). Below is the basic setup for the FGLS models, as specified in Equations (16) and (17), respectively:
β̂=(X′Ω̂−1X)−1X′Ω̂−1yVar(β̂)=(X′Ω̂−1X)−1
In this context, Ω̂ takes into account implicit assumptions about varying error variances (heteroscedasticity), serial correlation, and dependencies between different cross-sectional units. On the other hand, OLS is applicable when ΩNT is assumed to have constant error variances (homoscedasticity), no time-based correlations, and no cross-sectional dependencies. The second method used to analyze the relationship between the variables is the PCSE technique, proposed by Beck and Katz (1995). Like FGLS, PCSE is well-suited for long-term analysis in static panel data. Panel data often present challenges such as autocorrelation, group-wise heteroscedasticity, and CSD. The PCSE method effectively addresses these issues (Doku et al., 2019; Sandow et al., 2021). Additionally, Nickell (1981) demonstrated that PCSE can reduce bias in dynamic models with fixed effects, particularly when there is slope heterogeneity. This technique is especially useful when the dataset contains more N than T (Biswas et al., 2022; Hoechle, 2007). The PCSE estimator is computed using the following formulas, as shown in Equations (18) and (19), respectively:
In this setup, X˜ and y˜ are the Prais-transformed versions of the explanatory and dependent variables. We use Σ˜ to represent our estimate of Σ. The equation ΩNT=Σ⊗π means that ΩNT is formed by taking the Kronecker product of Σ and π.
Dumitrescu-Hurlin granger causality
To investigate causality among the variables, this study utilizes the Dumitrescu–Hurlin causality test proposed by Dumitrescu and Hurlin (2012). This test is designed to uncover how variables influence each other by showcasing their effectiveness in diverse panel structures. It is particularly versatile and applicable whether N exceeds or falls short of T (Abdi et al., 2024a, 2024b). The method is adept at identifying causation within specific segments of the panel (Lopez and Weber, 2017), which is crucial for developing well-informed policy recommendations. Given that it accommodates CSD, the Dumitrescu–Hurlin test is classified as a second-generation test (Çeştepe et al., 2024). In this causality analysis, we outline the null and alternative hypotheses as follows:
In this procedure, we compare the null hypothesis with the alternative hypothesis. To do this, the test uses two statistics, Wald statistics (W) and the standardized statistics (Z). The formulas for these test statistics are given in Equations (20) and (21)
W¯=1N∑i=1NWiZ¯=N2K(W¯−K)Empirical investigation and analytical discussionDescriptive and preliminary analysis
Figure 2 illustrates the development flows to agriculture across Eastern African countries by asserting significant disparities in funding allocation over time. Kenya and Ethiopia emerged as the largest recipients, with Kenya recording the highest peak (144 million) in 2012. This reflects strong external support and policy prioritization of agriculture. Ethiopia also experienced multiple years of high inflows, particularly between 2012 and 2014, which suggests sustained donor interest that has facilitated improvements in agricultural productivity and food availability. In contrast, countries like Somalia, Zambia, and Zimbabwe receive consistently lower agricultural development flows, potentially due to weak institutional mechanisms or socio-political challenges, which exacerbate food insecurity by limiting agricultural output and market access. Mozambique, Uganda, and Rwanda exhibit moderate but steady increases in agrarian funding. Though challenges remain in sustaining rural livelihoods, this indicates gradual improvements in food security. D.R. Congo demonstrates fluctuating inflows, which stresses the variability of donor commitments and its continued vulnerability to food insecurity. The regional differences in Figure 2 suggest that countries with stable agricultural policies and more muscular governance structures tend to attract higher funding. At the same time, fragile economies struggle with limited inflows that further deepen food insecurity.
Development flows to agriculture across Eastern African countries.
Heatmap showing rural development funding in millions of USD from 2000 to 2021 for various African countries. Ethiopia had the peak funding in 2015 with 144 million USD, while Mozambique and Uganda also received significant amounts. The color intensity represents the funding scale, with dark blue indicating higher amounts and light yellow indicating lower amounts, as shown on the scale bar on the right.Summary statistics and correlation analysis
Table 2 presents a detailed statistical summary of the variability of central tendencies and distributional characteristics of the study variables. Rural development exhibits the highest mean value (18.234), which indicates its prominence in the dataset. However, political stability records the lowest mean (−0.883), which reflects the persistent challenges associated with governance and institutional stability. The data further reveal substantial variability in food imports, which exhibit the highest maximum value (685.400) and the second-largest standard deviation (57.534). In contrast, food security and population growth demonstrate the least dispersion with standard deviations of 0.574 and 0.585, respectively, indicating relative stability in these variables over time. The distributional properties reveal key data patterns. Food security, rainfall, population growth, and political stability are left-skewed, while average temperature, food prices, imports, and rural development are right-skewed. Food inflation, imports, and rural development trends have critical food security implications. High import variability and right-skewed food prices indicate unstable external reliance, worsening food insecurity when domestic production falls short. Meanwhile, uneven rural development suggests that while some areas receive significant investment, many lag behind, restricting agricultural capacity and resilience to inflation. Furthermore, the Jarque-Bera test results confirm that all variables except food security deviate from normality, as evidenced by probability values below the 0.05 threshold.
Descriptive summary of variables.
FS
RF
TEM
FPI
PG
FM
PS
RD
Mean
0.000
1006.725
23.816
15.136
2.766
0.661
−0.883
18.234
Maximum
2.879
1601.820
30.440
685.400
3.927
2.955
0.661
144.402
Minimum
−2.847
249.270
19.010
−13.110
0.524
0.043
−3.313
−0.038
Std. Dev.
1.270
347.877
2.622
57.534
0.585
0.574
0.908
20.719
Skewness
−0.120
−0.321
0.822
10.366
−1.552
1.467
−0.532
2.842
Kurtosis
2.647
2.611
3.757
113.318
6.560
4.798
2.620
13.452
Jarque-Bera
1.823
5.635
32.760
125998.200
223.148
118.362
12.747
1415.583
Probability
0.402
0.060
0.000
0.000
0.000
0.000
0.002
0.000
Observation
240
240
240
240
240
240
240
240
To assess the strength of relationships among the variables, this study conducted a correlation coefficient analysis, as presented in Table 3. The results reveal both positive and negative correlations among the variables. Positive correlations indicate that increases in certain factors enhance food security. Specifically, rainfall, food imports, and political stability positively correlate with food security, which suggests their contributions to improved outcomes in East Africa. Conversely, negative correlations stress variables that hinder food security. Higher temperatures, rising food prices, and population growth are linked to lower food security, which suggests the challenges these factors impose on the region. Notably, rural development exhibits a weak negative correlation with food security, implying that, in this dataset, increases in rural development are marginally associated with declines in food security across the region.
Analysis of pair-wise correlations.
Variables
FS
RF
TEM
FPI
PG
FM
PS
RD
FS
1.000
RF
0.134**
1.000
TEM
−0.479***
−0.042
1.000
FIN
−0.038
0.020
0.078
1.000
PG
−0.341***
0.171***
0.392***
−0.023
1.000
FM
0.037
−0.249***
0.304***
−0.020
0.005
1.000
PS
0.501***
0.326***
−0.662***
−0.076
−0.054
−0.330***
1.000
RD
−0.003
0.102
−0.065
−0.034
0.058
0.343***
−0.029
1.000
***, **, * denote significance levels at 1, 5% and 10%, respectively.
Evaluation of cross-sectional dependence and heterogeneity
It is essential to assess the presence of CSD before proceeding with panel estimation techniques. Table 4 presents the results of two widely used CSD tests—the Breusch-Pagan LM test and the Pesaran scaled LM test. These tests were employed to examine whether the residuals across countries in the panel are correlated, which would indicate the influence of unobserved common factors or regional spillover effects. The findings from both the Breusch-Pagan LM and the Pesaran scaled LM tests consistently reject the null hypothesis of cross-sectional independence at the 1% significance level for all variables. This suggests a statistically significant presence of cross-sectional dependence in the panel dataset, implying that shocks or changes in one country may affect others within the region. Additionally, the heterogeneity of the slope coefficients was evaluated using the Pesaran and Yamagata (2008) test. The results, presented in Table 5, indicate that the null hypothesis is rejected at the 1% significance level, as indicated by the significance of both Δ˜ and Δ˜ Adjusted. This confirms the presence of heterogeneity in the slope coefficients across the cross-sectional units, which implies that the relationships between variables vary significantly among countries in the dataset.
Results of cross-sectional dependence analysis.
H0: No cross-section dependence
Variable
Breusch-Pagan LM
Pesaran scaled LM
FS
299.226
20.3
[0.000]
[0.000]
RF
203.807
11.995
[0.000]
[0.000]
TEM
316.631
21.815
[0.000]
[0.000]
FPI
120.882
4.777
[0.000]
[0.000]
PG
458.288
34.144
[0.000]
[0.000]
FM
723.117
57.195
[0.000]
[0.000]
PS
220.447
13.443
[0.000]
[0.000]
RD
168.819
8.949
[0.000]
[0.000]
Heterogeneity analysis results.
H0: slope coefficients
Statistic
p-Value
Δ˜
3.334
0.001
Δ˜ Adjusted
4.714
0.000
Panel unit root and cointegration tests
To address the limitations associated with cross-sectional dependence in first-generation unit root tests and to ensure accurate inference regarding the stationarity properties of the variables, this study employed second-generation panel unit root tests—namely, the CIPS and CADF tests. The results, presented in Table 6, indicate a mixed order of integration across the variables. According to the CIPS test, rainfall, temperature, food imports, food prices, and rural development are stationary at level [I(0)], thereby rejecting the null hypothesis of a unit root. Conversely, food security, population growth, and political stability become stationary only after first differencing, indicating integration at order one [I(1)]. The CADF test results corroborate these findings, showing that rainfall, food imports, and rural development are stationary at level, while food security, temperature, food prices, population growth, and political stability are stationary at first difference.
Findings of second-generation unit root tests.
Variables
CIPS
CADF
Level
1st difference
Level
1st difference
FS
−1.642
−3.128***
−1.846
−2.531***
RF
−4.245***
−5.448***
−3.071***
−4.26***
TEM
−2.704***
−5.681***
−1.918
−3.773***
FPI
−3.274***
−4.696***
−2.187*
−3.316***
PG
−0.989
−2.381**
−1.549
−2.771***
FM
−2.876***
−4.630***
−2.260**
−2.920***
RD
−2.600***
−4.526***
−2.415**
−3.023***
PS
−2.022
−4.623***
−1.549
−2.749***
***, **, * denote significance levels at 1, 5% and 10%, respectively.
The cointegration among the variables was assessed using the Pedroni and Kao cointegration tests. The results, presented in Table 7, indicate a strong cointegration relationship among the variables. The Pedroni test results, including Modified Phillips-Perron, Phillips-Perron, and Augmented Dickey-Fuller, reject the null hypothesis of no cointegration at the 1% significance level, which confirms that the variables share a long-run relationship and move together over time. Similarly, the Kao test supports the presence of cointegration through the Modified Dickey-Fuller and Augmented Dickey-Fuller statistics. However, the p-values for the Dickey-Fuller, Unadjusted Modified Dickey-Fuller, and Unadjusted Dickey-Fuller t-statistics exceed the 5% significance level, which indicates that these tests fail to reject the null hypothesis.
Results of Pedroni and Kao cointegration tests.
Statistic
p-Value
Pedroni cointegration test
Modified Phillips–Perron t
5.492
0.000
Phillips–Perron t
2.288
0.011
Augmented Dickey–Fuller t
2.650
0.004
Kao cointegration test
Modified Dickey–Fuller t
−1.754
0.040
Dickey–Fuller t
−1.404
0.080
Augmented Dickey–Fuller t
−2.170
0.015
Unadjusted modified Dickey–Fuller t
−1.308
0.095
Unadjusted Dickey–Fuller t
−1.183
0.118
Empirical analysis—PCSE and FGLS techniques
Given the presence of mixed integration orders and confirmed cointegration among the variables, conventional long-run estimators such as FMOLS, DOLS, or ARDL-PMG were considered. However, these approaches produced inconsistent and unstable estimates, likely due to sample size constraints or convergence difficulties inherent in the data. As an alternative, this study adopts the PCSE and FGLS estimators applied to first-differenced data, which enable robust short-run analysis while controlling for cross-sectional dependence and heteroskedasticity. The findings indicate that rainfall exhibits a significant positive impact on food security using the PCSE and FGLS estimators. Specifically, increased rainfall favourably correlates with food security, which reinforces its vital role in East Africa’s rain-fed agricultural systems. This finding aligns with Adhikari et al. (2015), who affirm the importance of adequate rainfall in boosting agricultural productivity, and Abdi et al. (2024a, 2024b), who report a similar positive effect in Somalia. Similarly, Randell et al. (2022) and Hamadjoda Lefe et al. (2024) also report a positive relationship between rainfall and food security in SSA. These assert the importance of sustainable water management in addressing precipitation variability. However, the region’s rainfall variability remains a challenge, as frequent dry spells lead to crop failures while excessive precipitation accelerates soil erosion, both of which undermine food security. Given East Africa’s dependence on agriculture, ensuring water availability through efficient irrigation systems and climate adaptation strategies is essential to mitigating these risks and sustaining food production.
As anticipated, higher temperatures negatively influence food security due to their detrimental effects on crops throughout the plant life cycle. Elevated temperatures impair photosynthesis, disrupt fertilization processes, and limit nutrient and water uptake, ultimately reducing both the quantity and quality of crop yields. These findings align with Parthasarathi et al. (2022), who highlight the adverse physiological effects of rising temperatures on crop growth. Similarly, Abdi et al. (2023a, 2023b), Mubenga-Tshitaka et al. (2023), and Yiadom et al. (2023) emphasize the negative impact of higher temperatures on agricultural performance in East Africa, which reinforces the paramount role of temperature regulation in sustaining food production. However, the implications of rising temperatures extend beyond reduced yields. Increased heat stress accelerates soil moisture loss, intensifies pest infestations, and shortens growing seasons, further exacerbating food insecurity. In East Africa, where many farming systems rely on rain-fed agriculture, these effects pose significant challenges for smallholder farmers who lack access to irrigation and climate-adaptive technologies. Given these risks, targeted climate adaptation strategies—such as heat-resistant crop varieties, improved irrigation infrastructure, and agroecological practices—are essential for mitigating the adverse effects of temperature increases and ensuring food security in East Africa.
Across both models, the coefficient of food prices ranges from −0.00123 to 0.000242, indicating an insignificant relationship with food security. This suggests that while food prices may have both positive and negative effects, their overall impact remains statistically weak in the East African context. Unlike Abdi et al. (2024a, 2024b), who found significant mixed effects of food prices on food security in SSA, our findings suggest that food price fluctuations alone may not be a primary driver of food security outcomes in the region. While prior studies, such as Headey and Hirvonen (2022), argue that higher food prices can incentivize agricultural production and increase demand for unskilled labour, these effects may not be strong enough to translate into measurable improvements in food security. Similarly, although Gustafson (2013) highlights the adverse impact of rising food prices on household purchasing power, our findings suggest that other factors, such as income levels, market access, and government interventions, may mitigate the direct influence of food prices on food security. The insignificant results indicate the complexity of food security dynamics in East Africa, where multiple structural factors likely dominate over food price fluctuations alone.
In addition to climatic factors, several non-climatic factors significantly influence food security in East Africa. The findings from both PCSE and FGLS estimators reveal a consistently negative effect of population growth across all estimated models. Rapid population growth intensifies food demand, placing immense pressure on agricultural systems and often leading to the overexploitation of arable land (Muyanga and Jayne, 2014). Furthermore, urbanization exacerbates CO2 emissions, contributing to climate change and the depletion of natural resources, which further threatens agricultural productivity and food availability (Abdi et al., 2024a, 2024b). These findings align with empirical studies such as Miladinov (2023), who demonstrated a positive relationship between population growth—both rural and urban—and undernourishment, which demonstrates the growing strain on food supply systems. Similarly, Hall et al. (2017) projected that rapid population growth would be a key driver of food insecurity and undernourishment in Africa. In East Africa, where much of the population relies on subsistence farming, the connection between rapid demographic expansion and limited agricultural resources displays the urgency of approaches to sustainable food production, land management, and infrastructure development to mitigate food insecurity risks.
Food imports play a vital role in enhancing food security by increasing the availability of diverse, high-quality food options. The findings from both estimators confirm that food imports significantly and positively contribute to alleviating food insecurity, aligning with Arias et al. (2024), who affirm their role in supplementing domestic food production. Similarly, Smith and Glauber (2020) exhibit how food imports mitigate domestic production shortfalls, ensuring a stable food supply, particularly in regions vulnerable to climate-induced agricultural disruptions. In East Africa, where erratic weather patterns often constrain local agricultural productivity, food imports serve as a crucial buffer against production deficits and seasonal shortages. However, the broader implications of food imports on food security remain a subject of debate. While imports improve food availability, their long-term effects on food system resilience raise concerns, especially in light of disruptions from the COVID-19 pandemic and the Russia-Ukraine war. COVID-19 exposed vulnerabilities in global supply chains, while the Russia-Ukraine conflict disrupted grain markets, driving up food prices and worsening food insecurity in import-dependent regions like East Africa. Subramaniam et al. (2024) argue that heavy dependence on food imports may undermine local agricultural development, discourage domestic investment in food production, and create vulnerabilities in times of global supply chain disruptions. Additionally, fluctuations in global food prices can impact affordability, making food access unpredictable for lower-income populations.
Strikingly, political stability favorably influences food security across all model specifications, which affirms the importance of stable governance in ensuring food availability and accessibility. These findings align with Yiadom et al. (2023), who emphasize that institutional quality fosters political stability, which in turn alleviates key drivers of food insecurity. Stable governance creates an enabling environment for investment in agriculture, improves policy implementation, and enhances food supply chain efficiency. Similarly, Subramaniam et al. (2022) and Cassimon et al. (2022) support the critical role of strong institutions in formulating policies that address food supply challenges, improve the availability and accessibility of nutritious food, and reduce undernourishment—particularly in developing economies. In East Africa, where food security is often undermined by political instability, governance reforms aimed at strengthening institutional frameworks, reducing conflict, and ensuring policy continuity are essential for building resilient food systems and promoting long-term food security outcomes. Moreover, our findings indicate that while rural development contributes to food security by improving agricultural productivity and market access, its effect remains statistically insignificant across both PCSE and FGLS estimators. This suggests that rural development alone may not be a direct or immediate driver of food security in East Africa. Unlike Ghanem (2015), who highlighted the positive role of rural development in supporting smallholders and enhancing food security, our results imply that additional factors, such as infrastructure investment, financial inclusion, and access to technology, may be necessary for rural development to have a measurable impact (see Table 8).
Results from the PCSE and FGLS.
Variables
PCSE
FGLS
[1]
[2]
[3]
[4]
[5]
[6]
RF
0.000748***
0.000346**
0.00013
0.000732***
0.000510***
0.000201
(4.557)
(2.001)
(0.751)
(5.519)
(3.945)
(1.287)
TEM
−0.224***
−0.0755**
−0.138***
−0.237***
−0.115***
−0.159***
(−10.55)
(−2.456)
(−3.551)
(−12.16)
(−4.379)
(−4.306)
FIN
−0.000139
0.0000817
0.000237
−0.000335
−0.000315
0.000242
(−0.213)
(0.127)
(0.521)
(−0.530)
(−0.516)
(0.539)
PG
−0.426***
−0.594***
−0.271**
−0.341***
−0.466***
−0.254**
(−4.774)
(−7.018)
(−2.477)
(−4.412)
(−7.141)
(−2.499)
FM
0.507***
0.561***
0.503***
0.619***
0.629***
0.507***
(4.165)
(5.520)
(5.715)
(8.893)
(10.010)
(6.401)
PS
0.610***
0.409***
0.557***
0.395***
(6.181)
(4.522)
(7.429)
(4.738)
RD
0.000365
0.000963
(0.180)
(0.498)
Constant
5.425***
3.261***
3.785***
5.507***
3.598***
4.143***
(12.080)
(6.376)
(4.331)
(13.580)
(7.604)
(4.953)
Observations
240
240
240
240
240
240
Countries
12
12
12
12
12
12
z-statistics in parentheses. *** p < 0.01, ** p < 0.05, * p < 0.1.
Robust findings from food security index components
The estimation results, reported in Tables 9, 10, indicate the effects of climatic variables on the four dimensions of food security across East African countries. In particular, rainfall exhibits a two-sided effect: it improves food availability and utilization but tends to undermine accessibility and stability. This pattern suggests that while higher rainfall supports agricultural yields and enhances nutritional outcomes, it can simultaneously destabilize food markets and infrastructure, especially in areas prone to flooding or seasonal variability. These findings resonate with those of Randell et al. (2022) and Hamadjoda Lefe et al. (2024), who reinforce the role of sustainable water management—particularly irrigation and water governance—in enhancing food availability across SSA. However, unlike these studies, which primarily focus on the productive benefits of rainfall, our results demonstrate additional perspective that rainfall variability can also compromise food access and supply consistency, thereby broadening the understanding of rainfall’s role in shaping food security outcomes.
Robust results from the PCSE estimator.
Variables
Food availability
Food accessibility
Food stability
Food utilization
[1]
[2]
[3]
[4]
RF
0.000846***
−0.000299**
−0.000302
0.000266*
(5.218)
(−2.006)
(−0.212)
(1.861)
TEM
−0.0205
0.0645***
−0.283
0.0357
(−0.674)
(2.722)
(−0.910)
(1.105)
FPI
0.000274
−0.000559
−0.00741
0.000
(0.637)
(−1.020)
(−0.996)
(0.159)
PG
−0.116
0.436***
−2.769***
0.200*
(−1.434)
(6.968)
(−3.550)
(1.878)
FM
0.612***
−0.296***
0.491
−0.566***
(5.214)
(−2.970)
(0.647)
(−6.441)
PS
0.238***
−0.517***
2.470***
0.0839
(2.576)
(−7.482)
(3.133)
(0.980)
RD
0.0217***
0.00810**
−0.0752***
0.000161
(5.357)
(2.284)
(−3.740)
(0.0939)
Constant
−0.638
−2.843***
29.89***
−1.241*
(−1.311)
(−6.902)
(5.078)
(−1.736)
Observations
240
240
240
240
Countries
12
12
12
12
***, **, * denote significance levels at 1%, 5% and 10%, respectively.
Robust results from the FGLS estimator.
Variables
Food availability
Food accessibility
Food stability
Food utilization
[1]
[2]
[3]
[4]
RF
0.000963***
−0.000364***
0.000505
0.000350**
(6.847)
(−3.290)
(0.395)
(2.431)
TEM
0.0289
0.0885***
−0.317
0.168***
(1.059)
(5.078)
(−1.160)
(5.276)
FPI
0.00002
0.000178
−0.00799
−0.0001
(0.0595)
(0.422)
(−1.079)
(−0.0602)
PG
−0.260***
0.395***
−2.151***
0.134
(−3.898)
(8.533)
(−3.275)
(1.462)
FM
0.361***
−0.408***
1.082*
−1.369***
(4.255)
(−6.226)
(1.742)
(−18.700)
PS
0.224***
−0.474***
1.693**
0.358***
(2.686)
(−9.978)
(2.550)
(4.240)
RD
0.0151***
0.00581**
−0.0912***
0.0171***
(4.676)
(2.313)
(−4.981)
(5.317)
Constant
−1.491***
−3.045***
26.41***
−3.803***
(−3.352)
(−8.851)
(4.966)
(−6.207)
Observations
240
240
240
240
Countries
12
12
12
12
***, **, * denote significance levels at 1, 5% and 10%, respectively.
Temperature fluctuations exhibit a differentiated relationship with the dimensions of food security in East Africa, producing negative effects on food availability and stability, while showing positive associations with accessibility and utilization. The declines in availability and stability likely stem from the physiological stress that elevated temperatures impose on crop growth, leading to reduced yields and increased volatility in production. In contrast, the positive correlations with accessibility and utilization may reflect localized agroecological dynamics, where higher temperatures—within certain thresholds—extend growing seasons or improve productivity in warmer zones. This divergence underscores the non-linear and spatially contingent nature of temperature impacts, whereby the same climatic driver can simultaneously constrain output while enabling adaptive shifts in cultivation or food consumption patterns. These results are consistent with Parthasarathi et al. (2022), who document the inhibitory effects of rising temperatures on plant development, particularly in thermally sensitive regions, thereby supporting our findings on the adverse production-related consequences of warming.
The PCSE results, corroborated by the FGLS estimator, suggest that food prices are negatively associated with food accessibility and stability, but positively linked to availability and utilization. However, none of these relationships reach statistical significance. This outcome implies that, within the East African context, the influence of food prices on food security remains limited. One likely explanation is the predominance of subsistence agriculture, where smallholder farmers are not tightly integrated into formal markets and thus respond weakly to price signals. Furthermore, the prevalence of informal trade networks and government interventions—such as price controls and subsidies—may dilute the direct transmission of price fluctuations to households, especially in rural areas. The combination of positive and negative directional effects points to a more layered interaction between prices and food security, where incentives for production may coexist with affordability challenges. These findings partially align with those of Agyei et al. (2021), who reported that COVID-19-related price shocks adversely affected food availability in selected African countries. Yet, unlike their study, our results do not indicate statistically significant effects, which suggests a more muted or context-specific relationship between food prices and food security outcomes in East Africa.
Population growth presents a differentiated relationship with the various dimensions of food security in East Africa. While it appears to support food accessibility and utilization, it exerts a significant negative effect on availability and stability, as evidenced by the results from both PCSE and FGLS estimators. These outcomes suggest that rising population density places considerable strain on agricultural systems and natural resources, leading to diminished per capita food availability and greater exposure to fluctuations in supply. At the same time, the positive associations with accessibility and utilization may reflect improvements in market integration, labor force participation, and dietary diversity that often accompany demographic expansion, particularly in more urbanized settings. This duality speaks to the multifaceted nature of demographic change, where increased population can both challenge production capacity and simultaneously foster conditions that improve food access and use. Our findings intersect with those of Miladinov (2023), who identified a linkage between population growth and undernourishment; however, whereas that study emphasizes nutritional deficits, our analysis takes a broader approach, which provide a disaggregated view of how population dynamics shape food systems along both supply and demand channels.
The results indicate a multifaceted relationship between food imports and the dimensions of food security in East Africa. Food imports are found to significantly enhance food availability, suggesting that external sourcing plays a key role in supplementing domestic production and mitigating supply shortfalls. The effect on food stability is positive but statistically insignificant, implying that while imports may contribute to a more consistent supply, their stabilizing impact is not uniform across the region. In contrast, food accessibility and utilization are negatively and significantly affected by food imports, which may reflect affordability challenges linked to the cost structures of imported goods. Tariffs, transportation expenses, and currency-related trade costs can elevate retail prices, thereby restricting access for low-income households and limiting dietary diversity and quality. These findings are broadly consistent with Arias et al. (2024) and Smith and Glauber (2020), who argue that food imports can serve as a short-term solution to supply gaps but may also introduce market distortions and reinforce socioeconomic disparities in food access.
The results reveal the considerable role of political stability in shaping food security outcomes across East Africa. Political stability is positively and significantly associated with food availability and stability. This implies that a secure governance environment contributes to stronger agricultural performance and more reliable food supply systems. In a region where political unrest has frequently disrupted farming activities, market operations, and transport infrastructure, a more stable context appears to support sustained agricultural investment and the efficient functioning of food distribution networks. While political stability also proposes a positive relationship with food utilization, the association is not statistically significant. This suggests that improvements in governance may not be sufficient on their own to influence food quality and dietary practices, which often depend on complementary factors such as education, income, and public health infrastructure. In contrast, a significant negative effect is observed between political stability and food accessibility. This result may reflect persistent structural inequalities—such as rural–urban divides, economic exclusion, or post-conflict recovery gaps—that limit equitable access to food despite improvements in the broader political landscape. These findings are consistent with the observations of Yiadom et al. (2023), Subramaniam et al. (2022), and Cassimon et al. (2022), who document similar patterns in the uneven effects of governance on different dimensions of food security.
Rural development shows a significant positive association with food availability, accessibility, and utilization. This asserts its central role in supporting agricultural productivity, improving market integration, and promoting more favorable food consumption patterns in East Africa. These findings imply that investments in rural infrastructure, transportation networks, and local market systems enhance both the supply and accessibility of food, particularly for communities dependent on smallholder and subsistence farming. The observed positive effect on utilization may also be attributed to improved access to diverse and nutritious food options as rural livelihoods and local economies expand. However, the analysis also reveals a significant negative correlation between rural development and food stability. This result suggests that while rural expansion supports short-term gains in production and access, it may simultaneously introduce new risks—such as market volatility, land pressure, or resource degradation—that undermine the long-term consistency of food systems. This contrasts with Ghanem (2015), who emphasized the uniformly beneficial effects of rural development on food security, without addressing potential trade-offs between growth and stability.
Analysis of panel causality tests
To probe short-run predictive linkages, we apply the Dumitrescu and Hurlin (2012) panel Granger non-causality test, which accommodates slope heterogeneity across countries. The test evaluates whether lagged values of one variable improve forecasts of another; it speaks to temporal precedence, not structural causality, and in its canonical form assumes cross-sectional independence. As reported in Table 11, we fail to reject the null of no Granger-causality for the pairs rainfall–food security, food prices–food security, and food security–rural development, consistent with complex, potentially non-linear pathways not captured by a linear lag structure. In the short-run, shocks to rainfall or prices do not systematically improve forecasts of food security at the panel level, which suggests that policy responses premised on immediate pass-throughs may be unreliable. By contrast, we find bidirectional Granger-causality between population growth and food security. This indicates mutual predictive content (demographic pressures forecast subsequent food outcomes, and past food conditions forecast population dynamics). Moreover, we detect unidirectional Granger-causality from temperature to food security and from food security to food imports and food security to political stability. This implying that past temperature shocks help forecast food security, while past food security helps forecast subsequent import volumes and political stability.
Outcomes from Dumitrescu–Hurlin causality analysis.
Null Hypothesis: ≠
W-Stat.
Zbar-Stat.
Decision
RF ≠ FS
1.14
−0.006
No causality
FS ≠ RF
1.783
1.226
TEM ≠ FS
2.461**
2.526
Unidirectional
FS ≠ TEM
1.322
0.343
FPI ≠ FS
1.338
0.374
No causality
FS ≠ FPI
1.709
1.085
PG ≠ FS
2.429**
2.465
Bidirectional
FS ≠ PG
5.476***
8.306
FM ≠ FS
1.613
0.901
Unidirectional
FS ≠ FM
4.341***
6.131
PS ≠ FS
1.16
0.032
Unidirectional
FS ≠ PS
2.803***
3.183
RD ≠ FS
1.632
0.938
No causality
FS ≠ RD
1.964
1.575
The symbol ≠denotes the null hypothesis of no homogeneous causality across panel units—that is, the independent variable does not Granger-cause the dependent variable uniformly across all countries. ***, **, * denote significance levels at 1%, 5% and 10%, respectively.
The study’s estimates are interpreted as conditional associations rather than causal effects. Several endogeneity channels may bias inference: reverse causality, where food security can reshape rural development priorities and political stability; omitted common shocks such as climate anomalies, commodity cycles, and donor timing that co-move with regressors and outcomes; simultaneity in policy responses; and measurement error in institutional and development indicators. To mitigate these risks, we include country and year fixed effects and estimate the main specifications with PCSE and FGLS, using first differences where appropriate, to accommodate heteroskedasticity, serial correlation, and contemporaneous correlation across panels. In addition, we adopt the Dumitrescu–Hurlin panel Granger non-causality test to assess temporal precedence under heterogeneity; it does not provide structural identification and, in its canonical form, assumes cross-sectional independence. Accordingly, the reported linkages should be read as precedence patterns rather than causal effects, and residual endogeneity from unobserved common shocks, simultaneity, or measurement error may remain despite these safeguards.
Conclusion and policy implications
East Africa’s struggle for food security encapsulates the region’s broader challenges, including climatic volatility, political instability, and economic fragility. Ensuring food security not only mitigates malnutrition and lowers mortality rates but also fosters social cohesion and underpins long-term sustainable development. However, persistent climatic shocks—such as droughts, erratic rainfall, and rising temperatures—continue to undermine agricultural productivity by reducing crop yields, depleting natural resources, and heightening susceptibility to extreme weather events. These challenges are exacerbated by weak institutional frameworks, inadequate infrastructure, exposure to global price fluctuations, and constrained financial resources, all of which limit the effectiveness of mitigation efforts. The humanitarian cost of food insecurity remains dire; as of May 2023, hunger-related deaths in Somalia, Kenya, Ethiopia, and South Sudan alone ranged between 1,126 and 3,095 per day, equating to an hourly rate of 47 to 129 deaths (Oxfam, 2023). Without urgent and coordinated intervention, the region faces an escalating crisis marked by further loss of life, deepening human suffering, and the continued erosion of fundamental human rights.
Using PCSE and FGLS estimators, this study analyzed the influence of rural development, climatic variability, political stability, and food inflation on the four dimensions of food security in East Africa, based on panel data spanning from 2001 to 2020. The results reveal that rainfall significantly improves food availability and utilization but adversely affects accessibility and stability, likely due to disruptions caused by irregular or excessive precipitation. Rising temperatures are negatively associated with food availability and stability, which reflects the impact of heat stress on agricultural production and supply volatility. Population growth places a substantial strain on food availability and stability while showing a positive association with accessibility and utilization—potentially linked to shifts in labor dynamics and urban market development. Food imports contribute positively to food availability but negatively to accessibility and utilization. Political stability supports food availability and stability yet shows a negative association with accessibility, possibly reflecting uneven governance benefits across regions or population groups. Rural development positively and significantly affects availability, accessibility, and utilization, consistent with improvements in infrastructure and market access, but is negatively associated with food stability, which may reflect exposure to market volatility or transitional effects of rural transformation. In addition, the Dumitrescu–Hurlin panel Granger analysis shows no panel-wide evidence of Granger-causality between food security and precipitation, food price inflation, or rural development. By contrast, we find bidirectional Granger-causality between population growth and food security, and unidirectional Granger-causality running from food security to political stability and to food imports.
Drawing from the preceding findings, targeted policy actions in East Africa should reflect the multidimensional and context-specific nature of food security, addressing its four key dimensions. Strengthening irrigation systems, water harvesting infrastructure, and flood control mechanisms is critical for improving availability and ensuring the stability of food supply amid rainfall variability. Likewise, investment in heat-resilient crop varieties, agroecological practices, and localized climate advisory services can reduce climate-induced production risks and contribute to both stable supply chains and consistent access. Responding to population pressures will require a dual approach that promotes sustainable agricultural intensification and rural employment generation, thereby enhancing availability while supporting accessibility and utilization through improved livelihoods and dietary options. The region’s reliance on food imports underscores the need to strengthen regional integration under the AfCFTA, streamline trade logistics, and expand domestic food processing capacity—actions that can support availability while improving the affordability and nutritional quality of food, thus influencing both accessibility and utilization. Governance reforms are equally important: improving market efficiency, transparency in distribution, and rural service delivery is essential to ensuring that political stability translates into equitable food access and more stable food systems. Rural development strategies should be coupled with risk mitigation investments—including storage infrastructure, market regulation, and climate-resilient land planning—to protect stability while enhancing the other three dimensions. Moreover, the identified causal links between food insecurity, political stability, and food import dynamics point to the importance of integrated food security planning and early warning systems. A coordinated, forward-looking policy agenda rooted in these dimensions will be essential for strengthening climate resilience across East Africa’s food systems.
A key limitation is measurement: the composite food security index constructed with PCA integrates availability, accessibility, stability, and utilization into a single score, which can conceal trade-offs and distinct movements among the pillars. Although the panel approach accounts for CSD and slope heterogeneity across countries, country-specific estimations are absent, and important national differences in governance quality, agricultural investment, and infrastructure may therefore be blurred. Substantively, estimates are interpreted as conditional associations rather than causal effects. The Dumitrescu–Hurlin panel test is used to assess temporal precedence under heterogeneity, not structural causality, and in its canonical form assumes cross-sectional independence. Despite country and year fixed effects and the use of PCSE and FGLS to address heteroskedasticity, serial correlation, and contemporaneous correlation, residual endogeneity may persist through reverse causality, omitted common shocks, simultaneity in policy responses, and measurement error. The limited time dimension of the dataset further restricts fully disaggregated country-level regressions.
To preserve multidimensional detail, future research should pursue complementary approaches that retain pillar specificity, including dashboard frameworks, multi-criteria decision analysis (MCDA), and fuzzy-logic aggregation. Longitudinal techniques such as time-varying PCA and dynamic factor models can better capture how components evolve and interact under climatic and institutional shocks. Besides, we recommend that future studies with longer time series and finer-resolution data conduct country-level analyses to deliver country-specific policy guidance. To strengthen identification, researchers should deploy event-study and modern staggered difference-in-differences around exogenous reforms or eligibility thresholds with transparent pre-trend checks; develop instrumental-variable strategies with clearly defined relevance and exclusion, for example using donor allocation rules or administrative discontinuities; and employ factor-absorbing panel estimators such as common correlated effects (CCE) or interactive fixed effects to partial out latent common shocks. Methodologically, testing for state dependence and non-linearity—through threshold or quantile Granger causality and interactions with baseline aridity, market access, and price levels—together with subnational or higher-frequency data and microdata linkages, would reveal heterogeneous pathways and help triangulate results across complementary designs.
Data availability statement
Publicly available datasets were analyzed in this study. This data can be found at: https://databank.worldbank.org/source/world-development-indicators and https://www.fao.org/faostat/en/#data/RA.
Author contributions
AA: Validation, Data curation, Writing – review & editing, Supervision, Conceptualization, Methodology, Project administration, Investigation, Writing – original draft, Software, Resources, Visualization, Formal analysis. AM: Writing – review & editing, Methodology, Writing – original draft. SS: Writing – original draft.
Conflict of interest
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The authors declare that no Gen AI was used in the creation of this manuscript.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
ReferencesAbdiA. H.MohamedF. H.HalaneD. R.OmarM. O. (2025). Building food security through rural development: estimating climate change and crop production linkages in sub-Saharan Africa. Cogent Food Agric.11:249070671. doi: 10.1080/23311932.2025.2559984AbdiA. H.MohamedA. A.MohamedF. H. (2024a). Enhancing food security in sub-Saharan Africa: investigating the role of environmental degradation, food prices, and institutional quality. J. Agric. Food Res.17:101241. doi: 10.1016/j.jafr.2024.101241AbdiA. H.MohamedA. A.SugowM. O. (2023a). Exploring the effects of climate change and government stability on internal conflicts: evidence from selected sub-Saharan African countries. Environ. Sci. Pollut. Res.30, 118468–118482. doi: 10.1007/s11356-023-30574-w, PMID: 37917256AbdiA. H.SugowM. O.HalaneD. R. (2024b). Exploring climate change resilience of major crops in Somalia: implications for ensuring food security. Int. J. Agric. Sustain.22:2338030. doi: 10.1080/14735903.2024.2338030AbdiA. H.WarsameA. A.Sheik-AliI. A. (2023b). Modelling the impacts of climate change on cereal crop production in East Africa: evidence from heterogeneous panel cointegration analysis. Environ. Sci. Pollut. Res.30, 35246–35257. doi: 10.1007/s11356-022-24773-0, PMID: 36527558AbdiH.WilliamsL. J. (2010). Principal component analysis. WIREs Comput. Stat.2, 433–459. doi: 10.1002/wics.101AbdullahA.QingshiW.AwanM. A.AshrafJ. (2020). The impact of political risk and institutions on food security. Curr. Res. Nutr. Food Sci. J.8, 924–941. doi: 10.12944/CRNFSJ.8.3.21AdepojuA. A.SalmanK. K. (2013). Increasing agricultural productivity through rural infrastructure: evidence from Oyo and Osun states, Nigeria. Int. J. Appl. Agric. Apicult. Res.9, 1–10.AdhikariU.NejadhashemiA. P.WoznickiS. A. (2015). Climate change and eastern Africa: a review of impact on major crops. Food Energy Secur.4, 110–132. doi: 10.1002/fes3.61AgyeiS. K.IsshaqZ.FrimpongS.AdamA. M.BossmanA.AsiamahO. (2021). COVID-19 and food prices in sub-Saharan Africa. Afr. Dev. Rev.33, S102–S113. doi: 10.1111/1467-8268.12525, PMID: 34149241Ali WarsameA.Hassan AbdiA. (2023). Towards sustainable crop production in Somalia: examining the role of environmental pollution and degradation. Cogent Food Agric.9:2161776. doi: 10.1080/23311932.2022.2161776AriasJ.JankM. S.CardosoV. M.GilioL. (2024) The role of international trade in promoting food security. Inter-American Institute for Cooperation on Agriculture (IICA). Available at:https://test-assets-opsaa.iica.int/storage/resource/2024/09/0f35ff812bbb23235e5e4beb22a3bad5.pdf?utm_source=chatgpt.comBeckN.KatzJ. N. (1995). What to do (and not to do) with time-series cross-section data. Am. Polit. Sci. Rev.89, 634–647. doi: 10.2307/2082979BedirS.YilmazV. M. (2016). CO2 emissions and human development in OECD countries: granger causality analysis with a panel data approach. Eurasian Econ. Rev.6, 97–110. doi: 10.1007/s40822-015-0037-2BiswasS.BhattacharyaM.SadaranganiP. H.JinJ. Y. (2022). Corporate governance and earnings management in banks: an empirical evidence from India. Cogent Econ. Finance10:2085266. doi: 10.1080/23322039.2022.2085266BoakyeH. (2019). Achieving food security through efficient warehousing: case study of infrastructure for poverty eradication programme (IPEP), Ghana. Purchasing Supply Chain Manag.7, 1–11, Available at: https://www.eajournals.org/wp-content/uploads/Achieving-Food-Security-through-Efficient-Warehousing.pdf?utm_source=chatgpt.comBreitungJ.DasS. (2008). Testing for unit roots in panels with a factor structure. Economic Theory24, 88–108. doi: 10.1017/S0266466608080067BreuschT. S.PaganA. R. (1980). The Lagrange multiplier test and its applications to model specification in econometrics. Rev. Econ. Stud.47, 239–253. doi: 10.2307/2297111CandeliseC.SacconeD.VallinoE. (2021). An empirical assessment of the effects of electricity access on food security. World Dev.141:105390. doi: 10.1016/j.worlddev.2021.105390CassimonD.FadareO.MavrotasG. (2022). The combined effect of institutional quality and capital flows on food and nutrition security and undernourishment in sub-Saharan Africa. PLoS One17:e0275345. doi: 10.1371/journal.pone.0275345, PMID: 36215308ÇeştepeH.ÇetinM.AvcıP.BahtiyarB. (2024). The link between technological innovation and financial development: evidence from selected OECD countries. Int. J. Finance Econ.29, 1219–1235. doi: 10.1002/ijfe.2734ChandioA. A.AkramW.BashirU.AhmadF.AdeelS.JiangY. (2023). Sustainable maize production and climatic change in Nepal: robust role of climatic and non-climatic factors in the long-run and short-run. Environ. Dev. Sustain.25, 1614–1644. doi: 10.1007/s10668-022-02111-1ChandioA. A.JiangY.AminA.AkramW.OzturkI.SinhaA.. (2022). Modeling the impact of climatic and non-climatic factors on cereal production: evidence from Indian agricultural sector. Environ. Sci. Pollut. Res.29, 14634–14653. doi: 10.1007/s11356-021-16751-9, PMID: 34617217ChavasJ.-P. (2017). On food security and the economic valuation of food. Food Policy69, 58–67. doi: 10.1016/j.foodpol.2017.03.008ConceiçãoP.LevineS.LiptonM.Warren-RodríguezA. (2016). Toward a food secure future: ensuring food security for sustainable human development in sub-Saharan Africa. Food Policy60, 1–9. doi: 10.1016/j.foodpol.2016.02.003Connolly-BoutinL.SmitB. (2016). Climate change, food security, and livelihoods in sub-Saharan Africa. Reg. Environ. Chang.16, 385–399. doi: 10.1007/s10113-015-0761-xDeatonB. J.LipkaB. (2015). Political instability and food security. J. Food Secur.3, 29–33. doi: 10.12691/jfs-3-1-5DelgadoC.TschukertK.SmithD. (2023). Food insecurity in Africa: drivers and solutions. SIPRI Research Policy PaperDokuJ. N.KpekpenaF. A.BoatengP. Y. (2019). Capital structure and bank performance: empirical evidence from Ghana. Afr. Dev. Rev.31, 15–27. doi: 10.1111/1467-8268.12360DumitrescuE.-I.HurlinC. (2012). Testing for granger non-causality in heterogeneous panels. Econ. Model.29, 1450–1460. doi: 10.1016/j.econmod.2012.02.014EdemeR. K.NkaluN. C.IdenyiJ. C.ArazuW. O. (2020). Infrastructural development, sustainable agricultural output and employment in ECOWAS countries. Sustainable Futures2:100010. doi: 10.1016/j.sftr.2020.100010EzeohaA. E.ObiA.IgweA.EzeruigboC. (2020). The mobile phone revolution, the internet and rural electricity: what are the implications for food security in Africa?Inf. Dev.36, 603–622. doi: 10.1177/0266666919884991FAO. (2021). The state of food security and nutrition in the world 2021. Food and Agriculture Organization, Available at: https://www.fao.org/publications/fao-flagship-publications/the-state-of-food-security-and-nutrition-in-the-world/en/FAO. (2022). The state of food security and nutrition in the world 2022: Repurposing food and agricultural policies to make healthy diets more affordable. Food and Agriculture Organization. Available at: https://www.fao.org/3/cc7900en/cc7900en.pdfFAO. (2023). The impact of disasters on agriculture and food security: Avoiding and reducing losses through investment in resilience. Food and Agriculture Organization. Available at: https://www.fao.org/3/cc7900en/cc7900en.pdfFAO. (2024). The state of food security and nutrition in the world 2024. Food and Agriculture Organization. Available at: https://data.unicef.org/wp-content/uploads/2024/07/SOFI2024_Report_EN_web.pdfFAO, IFAD, UNICEF, WFP and WHO. (2022). The State of Food Security and Nutrition in the World 2022. Repurposing food and agricultural policies to make healthy diets more affordable. Rome, FAO.FagbemiF.OkeD. F.FajingbesiA. (2023). Climate-resilient development: an approach to sustainable food production in sub-Saharan Africa. Future Foods7:100216. doi: 10.1016/j.fufo.2023.100216FedderkeJ. W.BogetićŽ. (2009). Infrastructure and growth in South Africa: direct and indirect productivity impacts of 19 infrastructure measures. World Dev.37, 1522–1539. doi: 10.1016/j.worlddev.2009.01.008Food Security Information Network. (2023). Global report on food crises 2023. Available at:https://www.fsinplatform.org/report/global-report-food-crises-2023Global Report on Food Crises. (2023). FSIN and GNAFC. (2023). Global Report on Food Crises 2023. Food Security Information Network. Retrieved from Available at:https://www.fsinplatform.org/report/global-report-food-crises-2023/?utm_source=chatgpt.comGhanemH. (2015). Agriculture and rural development for inclusive growth and food security in Morocco. Brookings Global Working Paper Series. Available online at: https://www.brookings.edu/wp-content/uploads/2016/07/Agriculture_WEB_Revised.pdf.GrecoS.IshizakaA.TasiouM.TorrisiG. (2019). On the methodological framework of composite indices: a review of the issues of weighting, aggregation, and robustness. Soc. Indic. Res.141, 61–94. doi: 10.1007/s11205-017-1832-9GustafsonD. J. (2013). Rising food costs & global food security: key issues & relevance for India. Indian J. Med. Res.138:398.HairJ. F.BlackW. C.BabinB. J.AndersonR. E. (2019). Multivariate Data Analysis. 8th Edn. Boston, MA: Cengage.HallC.DawsonT. P.MacdiarmidJ. I.MatthewsR. B.SmithP. (2017). The impact of population growth and climate change on food security in Africa: looking ahead to 2050. Int. J. Agric. Sustain.15, 124–135. doi: 10.1080/14735903.2017.1293929Hamadjoda LefeY. D.Asare-NuamahP.NjongA. M.KondoweJ.MusakarukaR. R. (2024). Does climate variability matter in achieving food security in sub-Saharan Africa?Environ. Challenges15:100870. doi: 10.1016/j.envc.2024.100870HeadeyD.EckerO. (2013). Rethinking the measurement of food security: from first principles to best practice. Food Secur.5, 327–343. doi: 10.1007/s12571-013-0253-0HeadeyD. D.HirvonenK. (2022). Food inflation, poverty, and urbanization (Vol. 2134). Intl Food Policy Res Inst. Available online at: https://books.google.com/books?hl=en&lr=&id=BPGGEAAAQBAJ&oi=fnd&pg=PA1&dq=food+inflation+poverty&ots=4VupNn3Kkv&sig=j3hSYqcppI9P9Y6kiisi-TKRfUYHoechleD. (2007). Robust standard errors for panel regressions with cross-sectional dependence. Stata J.7, 281–312. doi: 10.1177/1536867X0700700301HusseinM.LawC.FraserI. (2021). An analysis of food demand in a fragile and insecure country: Somalia as a case study. Food Policy101:102092. doi: 10.1016/j.foodpol.2021.102092IPCC. (2021). Climate change 2021: The physical science basis. Intergovernmental Panel on Climate Change. Available at: https://www.ipcc.ch/report/ar6/wg1/IFAD. (2021). Rural development report 2021: Transforming food systems for rural prosperity. International Fund for Agricultural Development. Available at: https://www.ifad.org/en/w/publications/rural-development-report-2021JolliffeI. T. (Ed.) (2002). “Principal component analysis for special types of data” in Principal component analysis. (New York, NY, USA: Springer), 338–372.KaoC. (1999). Spurious regression and residual-based tests for cointegration in panel data. J. Econom.90, 1–44. doi: 10.1016/S0304-4076(98)00023-2KarumeK.MondoJ. M.KiyalaJ. C. K. (2024). “Drought, the war in Europe and its impacts on food insecurity in sub-Saharan Africa, East Africa” in Climate change and socio-political violence in sub-Saharan Africa in the Anthropocene: Perspectives from peace ecology and sustainable development. eds. KiyalaJ. C. K.ChivasaN. (Cham, Switzerland: Springer Nature Switzerland), 91–111.KumarP.SahuN. C.KumarS.AnsariM. A. (2021). Impact of climate change on cereal production: evidence from lower-middle-income countries. Environ. Sci. Pollut. Res.28, 51597–51611. doi: 10.1007/s11356-021-14373-9, PMID: 33988844LokeshaM. N.MaheshaM. (2017). Economic benefits of road infrastructure on agricultural development and rural road infrastructure development programmes of India and Karnataka. J. Res. Bus. Manag.4, 42–48.LopezL.WeberS. (2017). Testing for granger causality in panel data. Stata J.17, 972–984. doi: 10.1177/1536867X1801700412MaxwellD. (2013). “Food security and political stability: A humanitarian perspective” in Food security and sociopolitical stability. Ed. Christopher B. Barrett. (Oxford, UK: Oxford University Press).MiladinovG. (2023). Impacts of population growth and economic development on food security in low-income and middle-income countries. Front. Hum. Dyn.5. doi: 10.3389/fhumd.2023.1121662MisraA. K. (2014). Climate change and challenges of water and food security. Int. J. Sustain. Built Environ.3, 153–165. doi: 10.1016/j.ijsbe.2014.04.006Mubenga-TshitakaJ.-L.DikgangJ.Muteba MwambaJ. W.GeloD. (2023). Climate variability impacts on agricultural output in East Africa. Cogent Econ. Finance11:2181281. doi: 10.1080/23322039.2023.2181281MumuniS.Joseph AleerM. (2023). Zero hunger by 2030 – are we on track? Climate variability and change impacts on food security in Africa. Cogent Food Agric.9:2171830. doi: 10.1080/23311932.2023.2171830MuyangaM.JayneT. S. (2014). Effects of rising rural population density on smallholder agriculture in Kenya. Food Policy48, 98–113. doi: 10.1016/j.foodpol.2014.03.001NickellS. (1981). Biases in dynamic models with fixed effects. Econometrica49, 1417–1426. doi: 10.2307/1911408NtiamoahE. B.LiD.Appiah-OtooI.TwumasiM. A.YeboahE. N. (2022). Towards a sustainable food production: modelling the impacts of climate change on maize and soybean production in Ghana. Environ. Sci. Pollut. Res.29, 72777–72796. doi: 10.1007/s11356-022-20962-z, PMID: 35610457OkouC.SprayJ. A.UnsalM. F. D. (2022). Staple food prices in sub-Saharan Africa: An empirical assessment. Washington, DC: International Monetary Fund.Oxfam (2023) Hunger in East Africa to hit new peak with one person likely to die every 28 seconds despite G7 pledge to end famine. Available online at: https://www.oxfam.org.uk/mc/3anjjv/ParksR. W. (1967). Efficient estimation of a system of regression equations when disturbances are both serially and contemporaneously correlated. J. Am. Stat. Assoc.62, 500–509. doi: 10.2307/2283977ParthasarathiT.FirdousS.DavidE. M.LesharadeviK.DjanaguiramanM.ParthasarathiT.. (2022). “Effects of high temperature on crops” in Advances in plant defense mechanisms (London: IntechOpen).PedroniP. (1999). Critical values for cointegration tests in heterogeneous panels with multiple regressors. Oxf. Bull. Econ. Stat.61, 653–670. doi: 10.1111/1468-0084.61.s1.14PedroniP. (2004). Panel cointegration: asymptotic and finite sample properties of pooled time series tests with an application to the PPP hypothesis. Econometr. Theory20, 597–625. doi: 10.1017/S0266466604203073PesaranM. H. (2004). General diagnostic tests for cross section dependence in panels. Cambridge Working Papers. Economics 1240(1), p. 1.PesaranM. H. (2007). A simple panel unit root test in the presence of cross-section dependence. J. Appl. Econ.22, 265–312. doi: 10.1002/jae.951PesaranM. H.YamagataT. (2008). Testing slope homogeneity in large panels. J. Econom.142, 50–93. doi: 10.1016/j.jeconom.2007.05.010PivotoD.WaquilP. D.TalaminiE.FinocchioC. P. S.Dalla CorteV. F.de Vargas MoresG. (2018). Scientific development of smart farming technologies and their application in Brazil. Inf. Process. Agric.5, 21–32. doi: 10.1016/j.inpa.2017.12.002RandellH.GrayC.ShayoE. H. (2022). Climatic conditions and household food security: evidence from Tanzania. Food Policy112:102362. doi: 10.1016/j.foodpol.2022.102362, PMID: 37064798SandowJ. N.DuoduE.Oteng-AbayieE. F. (2021). Regulatory capital requirements and bank performance in Ghana: Evidence from panel corrected standard error. Cogent Economics and Finance9, 2003503. doi: 10.1080/23322039.2021.2003503SarafidisV.WansbeekT. (2012). Cross-sectional dependence in panel data analysis. Econometr. Rev.31, 483–531. doi: 10.1080/07474938.2011.611458SarkodieS. A.OwusuP. A. (2020). How to apply dynamic panel bootstrap-corrected fixed-effects (xtbcfe) and heterogeneous dynamics (panelhetero). MethodsX7:101045. doi: 10.1016/j.mex.2020.101045, PMID: 32939352SmithV. H.GlauberJ. W. (2020). Trade, policy, and food security. Agric. Econ.51, 159–171. doi: 10.1111/agec.12547SmithL. C.HaddadL. (2015). Reducing child undernutrition: past drivers and priorities for the post-MDG era. World Dev.68, 180–204. doi: 10.1016/j.worlddev.2014.11.014SoffiantiniG. (2020). Food insecurity and political instability during the Arab spring. Glob. Food Secur.26:100400. doi: 10.1016/j.gfs.2020.100400SubramaniamY.MasronT. A.LoganathanN. (2024). Imports and food security. Glob. J. Emerg. Mark. Econ.16, 7–24. doi: 10.1177/09749101221146422SubramaniamY.MasronT. A.SubramaniamT. (2022). Institutional quality and food security. Singapore Econ. Rev.67, 2099–2127. doi: 10.1142/S0217590820500046SwamyP. A. V. B. (1970). Efficient inference in a random coefficient regression model. Econometrica38, 311–323. doi: 10.2307/1913012WarsameA. A.AbdiA. H.AmirA. Y.Azman-SainiW. N. W. (2023). Towards sustainable environment in Somalia: the role of conflicts, urbanization, and globalization on environmental degradation and emissions. J. Clean. Prod.406:136856. doi: 10.1016/j.jclepro.2023.136856World Bank. (2020). World Bank announces $500 million to fight locusts, preserve food security, and protect livelihoods. Available at: https://www.worldbank.org/en/news/press-release/2020/05/21/world-bank-announces-500-million-to-fight-locusts-preserve-food-security-and-protect-livelihoodsWorld Bank. (2022). Food security update – June 13, 2025. Available at: https://thedocs.worldbank.org/en/doc/40ebbf38f5a6b68bfc11e5273e1405d4-0090012022/related/Food-Security-Update-117-June-13-2025-2.pdfWorld Food Programme. (2024). Global report on food crises 2024. Available at:https://www.wfp.org/publications/global-report-food-crises-grfcYiadomE. B.DziwornuR. K.MawutorJ. K.AmankwahR. F. (2023). Exploring the relationship between extreme weather events, urbanization, and food insecurity: institutional quality perspective. Environ. Challenges13:100775. doi: 10.1016/j.envc.2023.100775ZereyesusY. A.CardellL.AjewoleK.FarrisJ.JohnsonM.KeeJ.. (2023) International food security assessment, 2023–2033 (GFA-34) US Department of Agriculture, Economic Research ServiceZezzaA.CarlettoC.FiedlerJ. L.GennariP.JolliffeD. (2017). Food security indicators: the value of disaggregated data. Food Policy68, 56–68. doi: 10.1016/j.foodpol.2017.08.007ZhouZ.-Y.WanG. (2017). Food insecurity in Asia: Why institutions matter. Asian Development Bank Institute. Available online at: https://researchonline.jcu.edu.au/53219/
Edited by: Ilia Murtazashvili, University of Pittsburgh, United States
Reviewed by: Omarsherif Mohammed Jemal, Arsi University, Ethiopia; Stefania Di Giuseppe, Food and Agriculture Organization of the United Nations, Italy