Khat Production and Rural Household Food Security

1. Introduction

1.1. Background

Food security remains a central measure of rural household welfaregiven that it reflects the extent to which households can securesufficient and reliable access to food through own production, income,or market purchases. In Eastern Africa, and particularly in Kenya andEthiopia, this issue has become increasingly important as smallholderlivelihoods face growing pressure from land fragmentation, climatevariability, and unstable agricultural markets. Within this context,khat (Catha edulis) production has emerged as a majorlivelihood strategy because it is a high-value perennial cash crop withrelatively rapid and frequent returns compared to many staple crops.Existing evidence shows that income derived from khat production canstrengthen household purchasing power and improve market-based access tofood, especially where subsistence farming is constrained by limitedland and rainfall uncertainty (Abafita & Bulga, 2020; Cafer, 2018; Hussein et al., 2023).

This articles is concerned with the relationship between khatproduction and rural household food security. It acknowledges khatproduction is understood not only as cultivation, but as participationin the wider khat value chain, including harvesting, selling,transportation, and related household livelihood activities. Foodsecurity is conceptualised as the household’s ability to secure foodthrough both direct food purchases and income-mediated productionsupport, such as the purchase of farm inputs that sustain future foodaccess. This conceptual linkage is important because khat is notconsumed as a staple food; rather, its contribution to food security isindirect, operating through cash earnings that households use to buyfood and support agricultural production. In the current study area,this is reflected in the finding that most households reported usingkhat proceeds to purchase both food and farm inputs, indicating thatkhat contributes simultaneously to immediate consumption and longer-termlivelihood resilience.

However, the relationship between khat production and food securityremains contested in the literature. On the one hand, studies have shownthat khat income can act as a short-term buffer against food shortagesby increasing households’ ability to purchase food during difficultperiods (Cafer, 2018; Hussein et al., 2023). On the other hand, otherstudies warn that the expansion of khat may displace food crops, reducecrop diversity, and increase dependence on market-mediated food access,thereby potentially weakening long-term food availability andnutritional resilience (Gebrehiwot et al., 2016; Ademe et al., 2017; Dessie, 2015). This tension makes the khat–food security relationshipanalytically important: the crop may simultaneously function as a sourceof food access and as a source of structural vulnerability, depending onhow households allocate land, labour, and income.

The relationship is further mediated by household-level social andeconomic factors. Research on rural livelihoods indicates that cashincome does not automatically translate into better food outcomesbecause expenditure behaviour, market access, and intra-householddecision-making influence how income is converted into food quantity,food quality, and agricultural reinvestment (Ruel & Alderman, 2013; Barrett et al., 2001). Gendered divisions of labour are also relevant,since income from cash crops may be controlled differently from foodprovisioning responsibilities, affecting the extent to which cashearnings improve household food welfare (Bikketi et al., 2016). In thisarticle these issues are important because respondents differ byeducation level, marital status, gender, and type of engagement in thekhat value chain; yet the descriptive and empirical evidence indicatesthat the positive valuation of khat for food security is broadly sharedacross these groups rather than concentrated among one specificsubgroup.

In this article, structural functionalism and symbolic interactionism are used as theories to interpret economic activities such as khat production as parts of a wider social system that helps households maintain stability under changing external conditions. Fromthis perspective, khat can be viewed as an adaptive livelihood subsystemthat supports food access by generating income. Symbolic interactionism complements this view by emphasising that household decisions are also shaped by shared meanings, social norms, and locally embedded understandings of what khat represents within the rural economy.Together, these theories help explain why khat’s contribution to foodsecurity cannot be understood only in economic terms; it is also asocially embedded livelihood practice whose role is shaped by communitymeanings, household priorities, and welfare strategies.

Against this background, thIs article examine the contribution ofkhat production to rural household food security in Igembe SouthSub-County by specifically: (i) describing the main forms of householdparticipation in the khat value chain and the principal ways in whichkhat proceeds support food security; (ii) assessing how households ratethe importance of khat to their food security; and (iii) testing whetherthat perceived importance varies across household characteristics suchas gender, education, marital status, and primary khat activity. Thearticle advances the claim that khat functions as a broad-based,income-mediated food-security mechanism in the study area, supportingboth immediate food access and productive reinvestment, while itspositive value is widely shared across household groups rather thanbeing confined to a narrow social or occupational category.

1.2. Problem Statement

Despite the growing importance of khat in rural household economies,there is limited context-specific evidence on how khat productionaffects access to formal education in rural Kenya. Existing work hasoften focused either on khat’s contribution to income and welfare or onits negative social consequences, but rarely on the contradictionbetween education support through income and education disruptionthrough labour participation and school-related behavioural effects.While some studies suggest that cash-crop income can enhance schoolingby easing financial constraints, others show that rural households maytrade schooling for labour when economic pressures intensify (Islam& Hoque, 2022; Hussein et al., 2023). In Igembe South, where khat isstructurally embedded in household livelihoods, the absence of empiricalevidence on this education paradox creates a significant gap inunderstanding whether khat enables, constrains, or simultaneously doesboth in relation to formal education access.

2. Literature Review

Cash crop adoption is commonly linked to income growth and livelihooddiversification in rural economies, but its implications for foodsecurity remain contested because the mechanisms are multi‑directional.One strand of scholarship argues that cash income improves households’ability to purchase food, smooth consumption, and manage risk,particularly when subsistence production is uncertain. Yetdiversification can also increase exposure to market volatility,dependence on purchased food, and susceptibility to price shocks. Intheir foundational synthesis on rural Africa, Barrett, Reardon, and Webb(2001) show that diversification outcomes depend on assets, marketaccess, and the “push–pull” factors that shape households’ entry intocash and non‑farm activities. This implies that the cash crop–foodsecurity nexus cannot be assumed beneficial; it is conditional onstructural and institutional contexts that shape households’ capacity toconvert income into stable food access.

Within food security frameworks, improvements in income andagricultural productivity are often expected to enhance food accessandpotentiallydietary quality, yet evidence shows that income gains donot automatically translate into nutrition improvements.Nutrition‑sensitive agriculture literature emphasises that food securityinvolves not only availability and access but also utilisation andstability, which require attention to care practices, health services,women’s time, and intra‑household allocation. Ruel and Alderman (2013)demonstrate that agricultural and social programmes frequently show weakor inconsistent nutritional impacts because they are not designed withexplicit nutrition goals, gender responsiveness, and behaviouralpathways. This is relevant for cash crops like khat because incomeeffects may raise food purchases without improving diet diversity orchild nutrition outcomes, especially when household allocation normsweaken the income‑to‑nutrition pathway. Consequently, rigoroushousehold‑level evidence is needed to distinguish “more food” from“better diets,” and short‑term consumption smoothing from long‑termresilience.

Khat (Catha edulis) occupies a distinctive position amongcash crops because it is high‑value, frequently traded, and sociallyembedded in producing regions, which can shape both productionincentives and consumption patterns. While khat is widely discussed inhealth and policy arenas, fewer studies focus directly on itsassociation with food security using validated or comparable measures.Cafer (2018) specifically addresses this gap by demonstrating a positiveassociation between khat production and food security among surveyedfarmers in Ethiopia, challenging the dominant narrative that centresonly on displacement of food crops. However, Cafer (2018) also cautionsthat positive food security outcomes may mask vulnerability pathwayscreated through natural‑resource pressures (notably irrigation),unregulated extraction, and the longer‑term social costs of khatconsumption. This dual framing suggests khat can function as an adaptivelivelihood strategy while also reproducing structural risks thatthreaten sustainability.

Complementary empirical work indicates that khat can contributesubstantially to rural household income, strengthening food purchasingpower during periods of production shortfalls. Hussein, Wondimagegnhu,and Misganaw (2023) report that khat constitutes a major share of farmincome in northwest Ethiopia and that household characteristics such asland allocation, working capital, and farming experience shape incomedependence on khat. These findings reinforce the income pathwayargument: households that successfully commercialise khat can improveconsumption capacity and potentially stabilise food access. Yet the samestudy notes broader concerns associated with expanded khatproductionmarket and resource implicationswhich can destabilise benefitswhen shocks occur. From a food security perspective, this points to atension between immediate consumption smoothing and longer‑termvulnerability created by specialisation and exposure to pricefluctuations.

A major counter‑argument centres on land‑use substitution, where khatexpansion may reduce staple production and increase dependence onpurchased food. Gebrehiwot et al. (2016) document a shift in Ethiopianagroforestry homegardens from diverse self‑subsistence systems towardgreater khat orientation, driven by favourable markets, shrinking farmsizes, irrigation access, and reduced supports for food crop inputs.Importantly, they argue that such changes may conflict with nationalpoverty and hunger reduction objectives because they alter localproduction systems and potentially weaken food availability at thehousehold level. This suggests that even when households experienceincome gains, the broader food system may become more fragile if foodcrop diversity declines and if households become structurally dependenton cash income. The land‑use substitution literature thereforeemphasises long‑term resilience and diet quality risks that may notappear in short‑term income metrics.

Household‑level survey evidence complicates a simple displacementnarrative by showing that khat engagement may interact with landallocation and crop diversity in nuanced ways. Ademe et al. (2017)examine khat production/consumption implications for land used in cropproduction and on‑farm crop variety and report statistically significantdifferences across household categories, indicating that khat can shapeproduction portfolios and cropping decisions. While such results can beinterpreted as supporting displacement concerns in contexts where khatmono‑cropping grows, they also imply that some households may maintainor even adjust crop mixes strategically depending on land tenure andlocal constraints. The key implication is that substitution effects arenot uniform; they vary across ecological zones, household assets, andgovernance conditions. Therefore, evidence from different settings(including Kenya) is essential to avoid over‑generalising from onecontext

Environmental and resource pathways represent an additional mechanismthrough which khat production can influence long‑term food security.Dessie and Kinlund (2008) show that khat expansion interacts with forestfrontiers and can alter settlement patterns and local crop productiondynamics, intensifying human–environment pressures. Although khatexpansion does not fully explain forest decline on its own, it canaccelerate land conversion and increase human activity in sensitiveareas, with implications for ecosystem services and agriculturalsustainability. Food security outcomes are thus shaped not only byhousehold income and land allocation but also by changes inenvironmental conditions that affect productivity and resilience. Thisis particularly relevant where irrigation and water withdrawals enableintensive khat cultivation, potentially generating resource competitionthat undermines other agricultural activities. Such pathways highlight why short‑term welfare gains may coexist with long‑term ecological vulnerability.

Beyond khat‑specific studies, broader evidence on industrial/cashcrops in Sub‑Saharan Africa indicates that food security impacts arestrongly dependent on the dominant impact mechanism being studied.Jarzebski et al. (2020) systematically review industrial crop production and show that much of the literature focuses on food access and availability, while utilisation and stability are less examined,limiting conclusions about comprehensive food security outcomes. This isdirectly relevant to khat because many studies infer “food securityimprovements” from income or consumption quantity without measuringdietary diversity, nutritional status, or stability under shocks. Thesystematic review also underscores that governance and institutionalarrangements shape who benefits and how, suggesting that marketstructures and policy oversight are central moderators. Therefore, khatresearch should explicitly map mechanisms (income, land substitution,resource pressure, gender allocation) to the four food securitydimensions to avoid partial interpretations.

Khat’s role as a livelihood resilience strategy is best understood insettings where food crops are unreliable due to climate variability orconstrained resources. Cafer (2018) frames khat as potentially adaptive in such contexts, enabling households to secure cash in environmentswhere subsistence crops face production risk. Yet resilience isconditional: if khat production increases dependence on irrigation andintensifies resource depletion, then resilience can erode over timethrough environmental degradation. Moreover, unregulated extraction ofwater resources and the health/social vulnerabilities linked to khatconsumption may create compounding risks that undermine welfare. Thus,resilience is not simply the presence of income, but the sustainabilityof the livelihood system and the capacity to withstand shocks. This reinforces the need for studies that capture both short‑term consumption improvements and long‑term vulnerability pathways.

Food security outcomes are also mediated by expenditure patterns andfinancial behaviour, shaping whether cash incomes are used to strengthendiets or diverted toward non‑food consumption. While khat income canincrease household purchasing power, evidence across cash‑crop systemsshows that income is not always allocated to nutrition‑enhancing goodssuch as diverse foods, health services, or productive inputs thatimprove long‑term food availability. Nutrition‑sensitive programmeevidence highlights that weak nutrition outcomes often stem from missingbehavioural pathways and insufficient attention to women’s empowermentand time constraints. Ruel and Alderman (2013) argue that enhancingnutrition outcomes requires deliberate design featurestargeting,actions, and service qualityrather than relying on income growth alone.Applied to khat, this means the same income stream can generatedivergent food security outcomes depending on household decision rules,knowledge, and bargaining dynamics.

Gender dynamics significantly complicate the conversion of khatincome into household food security because responsibilities for foodprovisioning and control over cash income can be unevenly distributed.Evidence from Kenya on gendered divisions of labour indicates that womenoften bear disproportionate household and care responsibilities, whichcan influence food preparation, dietary decisions, and the labouravailable for subsistence production. Where men control cash croprevenue, households may experience allocation inefficiencies, weakeningthe income‑to‑food pathway even when overall earnings rise. Thisgendered structure can also shape labour allocation decisions thatreduce time for food crop farming or household food preparation.Therefore, any analysis of khat and food security must considerintra‑household power relations, not only aggregate income.

Recent evidence also suggests that crop switching toward khat can beassociated with heightened food insecurity under certain conditions,indicating the importance of context and governance. A farm householdanalysis in eastern Ethiopia shows that households shifting from coffeeto khat experienced higher food insecurity rates than non‑shiftinghouseholds, implying that khat adoption may not universally secure foodoutcomes. This finding strengthens the argument that cash croptransitions can have mixed welfare effects depending on agronomicconditions, market access, labour costs, and institutional support.Importantly, it underscores that khat’s income advantages do notnecessarily translate into improved food security when broaderconstraints or vulnerabilities dominate. Such evidence makes a strongcase for country‑ and region‑specific analysis rather than generalisingacross the Horn of Africa.

Overall, the literature reveals three critical gaps that justify thepresent study in Kenya. First, although khat is shown to increase incomeand sometimes correlate with improved food security, few studiesdisaggregate the income‑to‑food pathwaysdistinguishing foodquantity, diet quality, and stability across shocksparticularly inKenyan rural settings (Cafer, 2018; Jarzebski et al., 2020). Second,land substitution and resilience perspectives are often treated ascompeting narratives rather than interacting mechanisms within the samehousehold livelihood system (Gebrehiwot et al., 2016; Ademe et al., 2017). Third, socio‑relational factorsgender allocation, norms, andbehavioural driversare under‑integrated into empirical measurement offood security outcomes (Ruel & Alderman, 2013; Bikketi et al., 2016).

3. Methods

3.1. Research Design

An exploratory cross‑sectional survey design using quantitativehousehold data was used. The exploratory approach was appropriatebecause empirical evidence on khat production–welfare linkages(including food security) remains limited and context‑specific in ruralKenyan settings, and exploratory designs are well suited to generatingempirically grounded hypotheses and clarifying “what, why, and how”relationships in under‑studied social phenomena. The cross‑sectionalstructure enabled the study to capture household characteristics andwelfare indicators at a single point in time, providing a snapshot ofhow engagement in the khat value chain relates to food security outcomesamong rural households.

3.2. Participants and Sampling

The study was conducted in Igembe South Sub‑County, Meru County,Kenya, a predominantly agrarian setting where khat is widely cultivatedand traded; the site was purposively selected because khat production isintensive across the sub‑county and is widely perceived to shape ruralhousehold welfare, including food access and livelihood stability. Theunit of analysis was the rural household engaged in khat‑linkedactivities (production, business/trading, and brokerage/agency), whilethe unit of observation comprised household‑level welfare and livelihoodindicators, including food security status, household income, assetownership, and core socio‑demographic characteristics. The targetpopulation included adult household heads (≥18 years) who participate inthe khat value chain as farmers, traders/business actors, orbrokers/agents; household heads were selected because they typicallymake key decisions regarding production choices and householdexpenditure allocation (including food purchases).

A total of 147 households was targeted using stratified samplingacross three heterogeneous strata of khat engagement: khat farminghouseholds (n=90), khat business/trading households (n=44), and khatbrokers/agents (n=13). Sampling followed a multi‑stage approach: first,clusters (villages/locations) were selected using probabilityproportional to size (PPS) to reduce operational cost while retainingrepresentativeness; subsequently, the sample was distributed acrossselected communities as follows Akachiu (11), Antubeiga (13), Athi (12),Athiru Gaiti (10), Kibuline (9), Kangeta (9), Kanthiari (8), Kiengu(10), Kirimampio (12), Mukululu (12), Maua (10), Mioli (9), Muringene(8), Nguyuyu (7), and Njia (8) yielding the planned total n=147. Withineach selected village, households were drawn from village lists usingrandom selection procedures, with purposive identification applied onlywhere necessary to locate households that matched specific strata (e.g.,broker/agent households), while maintaining proportional representationacross the sub‑county.

Although 147 households were targeted, the field dataset indicateDthat 125 respondents participated, while 120 cases were used in severalanalyses. thIS ImplIES that some questionnaires had incompleteresponses; therefore, results are interpreted using the valid n for eachanalysis. The participant profile shows a respondent pool that waspredominantly male (63.20% male; 36.80% female) and largely composed ofolder household heads, with the largest proportion falling in the 51–60years bracket. In terms of marital status, 47.30% were married, 14.50%widowed, 14.30% divorced, 12.70% separated, and 11.20% single. Educationattainment was distributed as 27% secondary, 23% primary, 16%undergraduate, 14% college, 10% no education, 7% vocational training,and 3% postgraduate, indicating that most respondents had at least basicschooling.

3.3. Data Collection instruments

The study collected primary data using a structured household surveyquestionnaire administered to adult household heads in Igembe SouthSub‑County. The questionnaire was appropriate for this study because itenables consistent measurement of household‑level characteristics andwelfare indicators across a geographically dispersed rural population,while also allowing standardised comparison across households engaged indifferent khat value‑chain activities (farming, business/trading,brokerage/agency). Consistent with the broader exploratory design, theinstrument combined closed‑ended items (to support quantification andmodelling) with a limited number of open‑ended prompts (to capture briefexplanatory responses on household welfare pathways).

The questionnaire was organised into thematic modules aligned withthe study objectives on khat production and household welfare. A firstmodule captured household socio‑demographic characteristics, includingrespondent gender, age category, marital status, and highest educationattainment. A second module captured khat value‑chain engagement,measuring the household’s primary role(s) in khat activities (forexample, picking/harvesting, farming scale, selling scale, transport,brokerage/agency, and multiple‑activity engagement). A third modulemeasured economic and livelihood indicators relevant to food securitysuch as income category ranges, banking access, saving versus spendingbehaviour, borrowing trends, and the perceived importance of khat tohousehold income and welfare.

Crucially, the food security component of the instrument was designedto capture both the presence and the pathway of khat’s contribution tofood security, rather than relying on a single indicator. Key itemsincluded: whether households reported that khat contributes to foodsecurity (Yes/No); the nature of contribution, measured by household useof khat proceeds (buying food directly; buying farm inputs; buying bothfood and inputs); and the importance rating of the khat value chain forhousehold food security using an ordered response scale (unimportant;lowly important; moderately important; important; highly important). Inaddition to food security, the instrument captured welfare‑relevantoutcomes that are conceptually linked to food security in rurallivelihoods such as household participation in schooling (school feesfinancing), access to healthcare expenditures, and selected welfareindicators (e.g., savings/investments).

3.4. Data Collection Procedures

Data were collected using a structured household questionnaireadministered to household heads at their residences. The instrumentincluded both closed‑ended and limited open‑ended items to capturesocio‑demographics, household livelihood engagement in the khat valuechain, and welfare outcomes particularly food security indicators. Priorto participation, enumerators explained the study purpose and obtainedinformed consent. Only respondents who voluntarily consented wereinterviewed.

3.5. Data Analysis

3.5.1. Descriptive analysis

The analysis starts with descriptive statistics to present a clearpicture of how rural households report khat’s contribution to foodsecurity and how they evaluate its importance. Two survey‑basedindicators are summarised using counts and percentages: (i) the natureof contribution to food security (buying food directly; buying farminputs; buying both food and farm inputs), and (ii) the importancerating of the khat value chain to household food security, recorded onan ordered five‑category scale from unimportant to highlyimportant. Reporting percentages makes the descriptive results easyto interpret for non‑statisticians because it answers the question “outof all households interviewed, what share reported each response?”

To compute percentages for any category k we divide thenumber of households in that category by the total number of validresponses for that question and multiply by 100. The same rule appliesto all descriptive distributions in this article (including thefood‑security contribution categories and the importance ratingcategories). This is experessed in Formual [1].

The nk​ is the frequency (number of households) observed incategory k, and N is the total number of householdswith valid responses for that particular question (for example,N=120).

3.5.2. Empirical model

After the descriptives, the study estimated one regression model toexplain household differences in the importance rating of khat to foodsecurity. The dependent variable is ordinal (ranked), so the appropriatesingle model is Ordered Logistic Regression. This model is suitablebecause it respects the natural ordering of the outcome without assumingthat the “distance” between categories is equal. First, define thedependent variable. Let Yi denote household iii’s importancerating:

Meaning of the coding (what the numbers represent): 1= Unimportant 2=Lowly important 3= Moderately important 4=Important 5=Highly important.Next, we define the predictors. Let Xi represent the set ofhousehold characteristics used to explain variation in Yi​. In thisstudy, Xi​ contains variables already collected (gender, age group,marital status, education attainment, and optionally khat value‑chainrole):

Meaning of the predictors in equation (3): Male _i​= 1 if respondent is male, 0 if female. AgeGroup _i ​=respondent’s age category (as recorded in the survey). MaritalStatus_I = marital category (single, married, separated,divorced, widowed. EducationLevel _i​ = highesteducation category (no education, primary, secondary, vocational,college, undergraduate, postgraduate) . KhatRole_i =khat value‑chain engagement category (if included) Now theordered logit model is written in a simple, readable way as “odds ofbeing at or above a threshold.” For each threshold j (movingupward in the ordered scale), the model estimates:

The understanding of equation (4) is that: Pr (Y_i≥j∣X_i)is the probability that household i rates khat as at least categoryj (e.g., "important or highlyimportant") given its characteristics X_i⋅1. α_j are modelthresholds (one for each cut-point in the ordered outcome). β are the effects of the predictors(gender, age, education, etc.). If a predictor increases the left-sidequantity, it increases the likelihood of being in higher importancecategories. Finally, because non-statisticians interpret odds ratiosmore easily than raw coefficients, results will be reported as oddsratios obtained by exponentiating each coefficient:

Interpretation of equation (5): If OR_r > 1,predictor r increases the oddsof reporting higher importance of khat for food security. 1 If OR_r < 1,predictor r reduces thoseodds. For instance, (interpretation only): OR = 1.50 means " 50% higher odds," holding other variablesconstant.

3.6. Ethical Considerations

The study adhered to core ethical principles in social research.Participation was voluntary and based on informed consent obtained priorto data collection. Respondents were assured of anonymity andconfidentiality, and no personally identifying information was reportedin outputs. Data were used strictly for academic purposes and handled tominimise any risk of harm or disclosure to participants.

4. Findings

This section presents the findings in two complementary stages.First, the descriptive results summarise the sample profile, the mainforms of engagement in the khat value chain, and the two corefood-security indicators used in the study: the nature of khat’scontribution to food security and the perceived importance of the khatvalue chain to household food security. Second, the empirical resultsexamine whether these perceptions vary systematically across respondentgroups using bivariate ordered logistic regression models fitted fromthe cross-tabulated counts supplied for the ordered outcome variable.This two-step approach is appropriate because the descriptive resultsestablish the overall response pattern, while the regression analysisassesses whether the tendency to rate khat as more important for foodsecurity differs by gender, education, marital status, or primary khatactivity.

4.1. Descriptive Results

Table 1 presents a descriptive overview of the key variables relevantto this manuscript. The table consolidates four blocks of information:respondent characteristics, primary activity in the khat value chain,the nature of khat’s contribution to food security, and the importancerating assigned to khat in relation to household food security. Thisoverview is important because it shows both the structure of the sampleand the distribution of the food-security indicators before inferentialmodelling is undertaken.

Note. Of the 147 households targeted, 125 respondents participated;however, valid responses for several food-security analyses were basedon N = 120 due to incomplete questionnaires. Frequencies forthe importance-rating categories were derived from the reportedpercentages using the valid base N = 120 and rounded to thenearest whole number.

The descriptive results show that the respondent pool waspredominantly male (63.20%, n = 79) and that marriedrespondents formed the largest marital category (47.30%, n =59). Education levels were relatively broad, but the largest proportionswere those with secondary education (27.00%, n = 34) andprimary education (23.00%, n = 29), indicating that mostrespondents had at least basic schooling. These characteristics suggestthat the sample reflects household decision-makers with varied butgenerally adequate educational exposure to understand livelihood choicesand welfare outcomes.

With respect to participation in the khat value chain,picking/harvesting was the most commonly reported primary activity(26.10%, n = 31), followed by small-scale selling (22.40%,n = 27). A further 17.10% (n = 20) reportedinvolvement in multiple khat-related activities, indicating that manyhouseholds are not restricted to a single role within the value chain.The prominence of labour and trading roles suggests that khat functionsnot only as an agricultural commodity but also as a broader livelihoodplatform capable of supporting income generation across multiple pointsin the local economy.

The food-security indicators strongly underscore khat’s perceivedvalue in the study area. Most households reported that khat contributesto food security by enabling the purchase of both food and farm inputs(61.67%, n = 74), whereas fewer households used khat proceedsfor food only (25.83%, n = 31) or inputs only (12.50%,n = 15). This pattern suggests that khat income supports foodsecurity through both consumption and production pathways. At theattitudinal level, the majority of respondents rated khat as eitherimportant (41.21%, n = 49) or highly important (28.86%,n = 35) for household food security, with only a very smallminority rating it as unimportant (4.23%, n = 5). Collectively,these findings indicate that khat is widely regarded as a central pillarof household welfare and food access in the study context.

4.2. Empirical Results

To assess whether the perceived importance of khat to household foodsecurity varies across respondent groups, the study estimated bivariateordered logistic regression models using the supplied cross-tabulatedcounts. The dependent variable was the five-level ordered responsemeasuring the importance of khat to household food security, coded from1 = unimportant to 5 = highly important. Because the availableinformation consisted of cross-tabs rather than a row-level dataset,each predictor was entered separately in its own ordered logit model.The resulting estimates therefore represent unadjusted bivariateassociations, not a full multivariable model with simultaneouscontrols.

Note. β = ordered logit coefficient; SE = standard error; OR = oddsratio; CI = confidence interval; Ref. = reference category. Higheroutcome values indicate greater perceived importance of khat tohousehold food security. ** denotes statistical significance atp < .05. No coefficient met the p < .05threshold in the current bivariate models, so no estimate is marked with**. These models were estimated from the supplied cross-tabulated countsand therefore represent unadjusted bivariate associations, not a singlemultivariable model.

The empirical results indicate that none of the tested predictors hada statistically significant association with higher perceived importanceof khat to household food security at the conventional 5% threshold. Forexample, the coefficient for male respondents was positive (β = 0.143),corresponding to an odds ratio of 1.154, but the association was notstatistically significant, p = .678. This suggests that malerespondents had slightly higher estimated odds of placing khat in ahigher food-security importance category than female respondents, butthe evidence is insufficient to conclude that this difference issystematic in the wider study population.

A similar pattern was observed for education. Relative to respondentswith no education, all education categories produced odds ratios greaterthan 1.00, implying a general tendency toward higher importance ratingsamong more educated groups; however, none of these effects reachedstatistical significance. The largest odds ratio was recorded forvocational education (OR = 1.713), but its confidence interval was wideand crossed 1.00, indicating substantial uncertainty around theestimate. In practical terms, this means that although education mayappear descriptively aligned with stronger perceptions of khat’simportance, the available data do not support a statistically reliableeducation gradient.

The same conclusion applies to primary khat activity and maritalstatus. None of the activity categories differed significantly from thepicking/harvesting reference group, and none of the marital categoriesdiffered significantly from the single reference group. Some categories,such as transportation (OR = 0.612) and divorced status (OR = 0.515),showed lower estimated odds of assigning higher importance to khat, butthese effects were statistically non-significant and imprecise. Overall,the empirical evidence suggests that the positive valuation of khat as acontributor to food security is widely shared across household groups,rather than being concentrated among specific demographic oroccupational categories.

Short implication. The lack of statistically significant groupdifferences suggests that khat’s perceived role in food security may bestructurally embedded across the local livelihood system. As a result,food-security interventions in khat-producing areas should not assumethat reliance on khat is concentrated only among a narrow subset ofhouseholds; instead, it appears to function as a broadly normalisedlivelihood strategy within the sampled community.

Figure 1 displays the empirical results as a black-and-white scatterplot of odds ratios with 95% confidence intervals. The x-axis shows theestimated odds ratio, while the y-axis lists the predictor categories(excluding reference categories). The dashed vertical line at OR = 1.0marks the point of no difference relative to the reference group. Pointsto the right of the line represent higher estimated odds of rating khatas more important for food security, whereas points to the left indicatelower estimated odds. Confidence intervals that cross the null lineindicate non-significant effects.

Figure 1 echoes the interpretation of Table 2. Most odds-ratioestimates cluster relatively close to the null line at OR = 1.0, and allconfidence intervals cross that line, visually confirming the absence ofstatistically significant differences across the available predictors.The figure is useful because it allows the reader to compare thedirection, magnitude, and precision of each estimate at a glance. Forinstance, vocational education lies furthest to the right among theeducation categories, while divorced status lies furthest to the leftamong the marital-status categories, yet both intervals remain wide andoverlap the null line. The visual pattern therefore supports theconclusion that perceptions of khat’s food-security importance aregenerally positive and broadly similar across respondent subgroups.

5. Discussion

The findings suggest that the contribution of khat to rural householdfood security in Igembe South is best understood as an income-mediatedlivelihood mechanism rather than as a direct agricultural food source.The pattern shown in Table 1where most households reported using khatproceeds to purchase both food and farm inputsindicates that khatsupports food security through a dual pathway: it improves immediateaccess to food while also strengthening the productive base throughreinvestment in agriculture. This matters analytically because it shiftsthe debate away from a narrow question of whether khat “feeds”households directly and toward a broader understanding of how cash-cropincome sustains food access and production capacity simultaneously. Inpractical terms, the evidence implies that households are not using khatincome merely for short-term consumption; they are also treating it as aresource for maintaining future livelihood continuity, which is acentral feature of rural food-security resilience.

This interpretation is consistent with the strand of literature thatpresents khat as an adaptive livelihood strategy in contexts wherefood-crop production is uncertain. Cafer (2018) argues that khat canfunction as a resilience mechanism by improving household food securitythrough cash earnings, while Hussein et al. (2023) likewise show thatkhat contributes substantially to rural household income in Ethiopia,thereby strengthening household welfare and consumption capacity. Thepresent findings speak directly to that argument because the high shareof households purchasing both food and farm inputs suggests that khatincome is being translated into both present and future food-securityneeds. At the same time, the findings also point to the importance ofrecognising how khat income is used, not simply whether it is earned.This is important because agriculture-to-nutrition research shows thatincome gains alone do not guarantee improved food outcomes unlesshouseholds allocate resources in ways that support food access, care,and productive reinvestment (Ruel & Alderman, 2013).

The strong positive skew in the food-security importance ratingsshown in Table 1 and Figure 1 further suggests that khat’s role inhousehold welfare is not marginal but deeply embedded in everydaylivelihood practice. Most respondents rated khat as eitherimportant or highly important for food security, andonly a very small minority considered it unimportant. Error! Filenamenot specified.The meaning of this pattern is not simply that households“like” khat economically; rather, it indicates that khat has acquiredthe status of a normalised welfare resource within the local ruraleconomy. In other words, households appear to perceive khat not as asupplementary crop, but as a central mechanism through which they managefood access, livelihood stability, and production maintenance. Thisinterpretation fits well with symbolic interactionism, because thecrop’s perceived value is shaped not only by economic returns but alsoby locally shared meanings and practical understandings of what khatrepresents within the community.

The empirical results in Table 2 and Figure 1 deepen thisinterpretation by showing that the positive valuation of khat for foodsecurity does not vary significantly across the social and livelihoodcategories tested in the available models. None of the bivariate orderedlogistic regressions for gender, education, marital status, or primarykhat activity produced statistically significant differences in the oddsof rating khat as more important for food security. Substantively, thismeans that the perceived value of khat is not concentrated only amongone privileged groupsuch as better-educated households, male-headedhouseholds, or tradersbut appears to be broadly shared across thesampled community. This is an important finding because it suggests thatkhat’s contribution to food security in Igembe South is systemic ratherthan subgroup-specific. Viewed through a structural-functionalist lens,this indicates that khat has become part of the wider social andeconomic arrangement through which households maintain welfare underlocal rural constraints.

At the same time, the results should not be interpreted as evidencethat khat is unambiguously beneficial in all respects. The literaturereviewed earlier cautions that khat expansion can displace food crops,narrow agricultural diversity, and weaken longer-term food-systemresilience even when short-term purchasing power improves (Gebrehiwot et al., 2016; Ademe et al., 2017). The present findings do not refute thoseconcerns; rather, they refine them. In this study area, householdsclearly experience khat as a positive food-security resource, but thatpositive valuation may coexist with longer-term structural trade-offsthat are not fully captured by perception-based or short-term welfareindicators. This means the study aligns with the literature showingshort-run food-access benefits from khat income, while also leaving roomfor the argument that heavy dependence on khat could produce futurevulnerabilities through land-use substitution or reduced cropdiversification.

A further implication concerns the social distribution offood-security benefits within the household. Prior literature suggeststhat cash-crop income can fail to improve household food outcomes whenintra-household allocation patterns favour non-food uses or when thecontrol of income and the responsibility for food provisioning areseparated (Bikketi et al., 2016). In the current study, however, thewidespread use of khat proceeds for both food and farm inputs suggeststhat, at least at the reporting level, households are allocating khatincome in welfare-supporting directions. This does not eliminate thepossibility of hidden inequalities in control over income, but it doesindicate that khat earnings are visibly tied to food-security functionsat the household level. From a practical standpoint, this implies thatdevelopment interventions should not focus only on discouraging khatdependence; they should also improve how khat income is managedforexample, through household budgeting, agricultural input support, andfood-security planningso that the welfare gains from the crop are betterprotected and possibly expanded.

Overall, the discussion points to a clear conclusion: in IgembeSouth, khat appears to operate as a widely normalised, income-basedfood-security strategy that is recognised across household groups andlinked to both consumption and production needs. The study thereforecontributes to the literature by showing that the significance of khatfor food security is not only economic but also social and systemic.This helps explain why khat remains deeply embedded in rural livelihoodsdespite the wider policy and scholarly debate over its long-termeffects. For policy and practice, the implication is that any attempt toimprove food security in khat-producing regions must engage with khat’sactual role in household survival strategies rather than treating it asan external or marginal livelihood activity.

6. Conclusion

6.1. Summary

The article concludes that khat production contributes positively toshort-term rural household food security in Igembe South Sub-Countythrough income generation, particularly by enabling households topurchase both food and farm inputs. The descriptive findings showed thatmost households used khat proceeds to support both immediate food accessand productive reinvestment, while the majority also rated khat asimportant or highly important for food security. Theempirical results further indicated that this positive perception didnot differ significantly across the household categories examined,suggesting that khat’s food-security value is broadly shared across thesampled community rather than concentrated in one specific demographicor occupational group.

These findings suggest that khat functions as a widely embeddedlivelihood mechanism that supports food security through market-basedfood access and farm-level reinvestment. However, this contributionshould be interpreted cautiously, since the wider literature and thestudy’s analytical framing indicate that dependence on khat income mayalso carry longer-term risks related to land-use substitution, reducedfood-crop diversity, and household reliance on cash-mediated foodaccess. The study therefore concludes that khat is neither simplybeneficial nor wholly detrimental; rather, it plays a complex andstructurally embedded role in rural household food security.

6.2. Limitations and Future Research

This study is limited by its cross-sectional design and by the use ofcross-tabulated counts for the regression analysis, which means theempirical models are bivariate and unadjusted rather than fullymultivariable. In addition, the outcome used in the regression reflectsperceived importance of khat to food security, which does not captureall objective dimensions of food security such as dietary diversity orseasonal food stability. Future research should therefore usehousehold-level longitudinal data and more direct food-security measuresto assess whether khat’s short-term benefits are sustained over time oroffset by longer-term livelihood and land-use risks.