ISSN: 2407-814X (p); 2527-9238 (e) AGRARIS: Journal of Agribusiness and Rural Development Research Vol. 7 No. 2 July – December 2021, Pages: 207-224 Article history: Submitted : April 7th, 2021 Revised : August 2nd, 2021 July 4th, 2021 Accepted : August 14th, 2021 Simbarashe Ndhleve1, Clarah Dapira2, Hlekani Muchazotida Kabiti1,*, Zibongiwe Mpongwana3, Elphina Nomabandla Cishe4, Motebang Dominic Vincent Nakin1, Stephen Shisanya2, Keitirele Patricia Walker5 1 Risk and Vulnerability Science Centre, Walter Sisulu University, South Africa 2 Directorate of Research and Innovation, Walter Sisulu University, South Africa 3 Research Chair in Sustainable Rural Development, Walter Sisulu University, South Africa 4 Faculty of Educational Sciences, Walter Sisulu University, South Africa 5 Malotwana Silvopastorall, Botswana *) Correspondence email: kabitihlekani@yahoo.co.uk Household Food Insecurity Status and Determinants: The Case of Botswana and South Africa DOI: https://doi.org/10.18196/agraris.v7i2.11451 ABSTRACT This study used a survey questionnaire to investigate the state and determinants of household food security in South Africa and Botswana. In South Africa, 1557 households in two District Municipalities within Eastern Cape Province participated in the study. In Botswana, data was collected from 506 in Southern Kanye, southeast District, and Gaborone households. The paper employed Household Food Insecurity Access Scale and Prevalence to investigate the state food security across the sample population. The Household Food Insecurity Access Prevalence was used to determine the characteristics of and changes in household food insecurity of the sample households. The linear and ordinal regression analyses were carried out to outline determinants of food insecurity in the region. Findings emanating from the current study show a high prevalence of food insecurity in both countries. Meanwhile, geographical location, household size, and socio-economic infrastructure emerged as common determinants of food security for Botswana and South Africa. The paper recommends state-mediated intervention to improve access to basic socio-economic infrastructure and address unemployment rates in both countries as key areas to improve food security status. Keywords: Food access, food availability, food security, Household Food Insecurity Access Score INTRODUCTION Globally, food insecurity is a serious socio-economic problem that manifests itself in the form of hunger, malnutrition, and stunting due to limited access to nutritious food. It is a problem that has received much scholarly attention for many decades and has been identified by various governments as a development priority area. However, it remains difficult to find long-lasting solutions to the problem. According to FAO, IFAD, UNICEF, WFP, & WHO (2019), 1 in 9 people, http://issn.pdii.lipi.go.id/issn.cgi?daftar&1420518152&1&& mailto:kabitihlekani@yahoo.co.uk https://doi.org/10.18196/agraris.v7i2.11 ISSN: 2407-814X (p); 2527-9238 (e) 208 Household Food Insecurity Status and ….. (Ndhleve et al.) corresponding to 821 million people, are regarded as not having access to enough nutritious food to eat. Africa houses the most severely deprived people, currently estimated at three-quarters of the world’s extremely poor (FAO et al., 2019; Gonzalez, 2015; Grobler, 2016; World Bank, 2015). The World Bank (2015) estimated that 56% of the world’s extreme poor are in Africa. It is a finding that also resonates with a report by the Grebmer et al., (2019) in 2019 Global Hunger Index, which reveals that of the ten hungriest nations in the world, 6 of them are in Africa. This implies a high prevalence of hunger and undernourishment in Africa compared to the rest of the world. This food insecurity challenge is further exacerbated by the fact that the African continent has been experiencing fast growth in population size, which is projected to double by 2050 (World Bank,2015) However, this population growth does not match per capita food production, which has declined by 20% between 1970 to 2000 in Africa (Abdulai, Barrett, & Hazell, 2004). Godfray et al., (2010) caution that tremendous food production growth does not always translate to sufficient access to protein and energy as micronutrient malnourishment is shown to be prevalent in countries experiencing increased agricultural production. Accordingly, if the challenge of food insecurity is to be successfully addressed, increased agricultural production should also be accompanied by interventions that directly address the key underlying structural issues responsible for food insecurity. For example, structural inequalities such as the unequal distribution of the means of production, especially land, limited access to agricultural inputs to grow food, and income inequality that curtails the poor's purchasing power to access food on the market all need to be addressed. As succinctly articulated by Sen (1990), food insecurity is not a function of food production, considering that the world produces sufficient food to feed everyone. Instead, it is a result of inherent socio-economic structural factors that disenfranchise the poor. Other factors such as lack of meaningful employment opportunities, failure of small-scale agricultural production, uneven food landscape, ever-increasing food prices, failure to afford food, and lack of access to production resources are continuously negatively impacting food security (Altman, Hart, and Jacobs, 2009; Bahta, Wanyoike, Katjiuongua and Marumo, 2017). The risk of food insecurity has multiple determinants framed around the four aspects, namely, accessibility, availability, utilization, and stability. Also, these four aspects are affected directly and indirectly by various socio-economic factors. The impacts follow socio-economics lines and are significantly different both geographically and across different times. Thus, in many African countries, hunger and malnutrition take place when there is plenty of food. This is mainly because most of the poor are not engaged in any meaningful economic activities that generate income and enable them to access nutritious food (Ahmed, Eugene, & Abah, 2007; FAO, IFAD, UNICEF, WFP, & WHO, 2018; Statistic South Africa, 2019). It is a challenge that does not even spare the economically stable African countries such as South Africa and Botswana (Asefa, 1991; Human Sciences Research Council, 2015; Bahta, Wanyoike, Katjiuongua, & Marumo, 2017). These two countries present striking similarities and differences that are significant in the analysis of food security. In terms of similarities, firstly, both countries have enjoyed high and sustained economic growth rates that are considered the best in Africa over the past decades (Acquah http://issn.pdii.lipi.go.id/issn.cgi?daftar&1420518152&1&& ISSN: 2407-814X (p); 2527-9238 (e) 209 AGRARIS: Journal of Agribusiness and Rural Development Research et al., 2013; Bahta et al., 2017). Secondly, reported rapid economic growth in these two countries is widely blamed for the failure to change the face of food insecurity (Bahta et al., 2017; van der Berg, 2006). In Botswana, significant variation in food access is widely reported across areas in the Southern and Southeast Districts. In South Africa, households that experienced severely inadequate food access were observed in the Northern Cape (13%), Mpumalanga (13%), and the Northwest (12%) (Statistic South Africa, 2019). Meanwhile, in Northwest (24.5%), Northern Cape (21%), Eastern Cape (20%), KwaZulu-Natal (19%), and Mpumalanga (19%), households experienced the highest proportions of inadequate access to food. The values are above the national average of 15.8% for all provinces (Statistic South Africa, 2019). The unequal food security status in both countries is not moving with economic growth, as evidenced by the fact that the number of poor households threatened by food insecurity has been increasing. The Global Food Security Index that assessed food security across 113 countries in 2015 ranked South Africa and Botswana at 41 and 46, respectively. However, the food security status in the two countries has further retrogressed, as shown by their 2018 rankings; South Africa has slipped by 4 points and Botswana by 6 points. Both countries are currently ranked at 45 and 52, respectively (The Economist Intelligence Unit, 2018). Although such changes do not represent a dramatic regression of the state of food security, in both countries, they do signal that people who are experiencing hunger are on the rise, albeit at a lower rate. Botswana and South Africa rank the highest on inequality levels in Southern Africa (Bahta et al., 2017; Statistic South Africa, 2019; Coates, Swindale, & Bilinsky, 2007; Statistics Botswana, 2018). Regarding differences, South Africa is a renowned net food exporter, producing enough staple food (Statistic South Africa, 2019) while Botswana is a net importer of food products, with imports currently accounting for about 90% of the national food supply (Bahta et al., 2017). These notable differences and similarities provide a comprehensive and deep base to critically examine the dynamics of food security as well as draw inferences that are widely applicable and can be used to inform future policy interventions. This deepening food insecurity crisis, as evidenced by the increasing number of food insecure households in South Africa and Botswana, served as a strong motivation for this study. Central to this food insecurity challenge is the dearth of up-to-date scientific data that sheds a nuanced picture of the challenge and details the current state of food security within poor households. The existing information is limited to a few scattered quantitative and aggregate indicators at the national level that provide superficial reading, especially when it comes to the state of food security at the household /local municipality level, together with key determinants of food security. This is the gap that the author identified in this paper. Accordingly, this paper intends to address this gap by sharing the findings of a study that was undertaken in South Africa and Botswana. The case study approach was employed to profile the state of food security in Botswana and South Africa and identify context- specific indicators of food security in the two countries. http://issn.pdii.lipi.go.id/issn.cgi?daftar&1420518152&1&& ISSN: 2407-814X (p); 2527-9238 (e) 210 Household Food Insecurity Status and ….. (Ndhleve et al.) RESEARCH METHOD Study Site This study was undertaken in two countries known as the economic powerhouses of Southern Africa, which are South Africa and Botswana as shown in Figure 1. In South Africa, data collection took place in two district municipalities of the Eastern Cape Province, namely OR Tambo, Chris Hani, and Buffalo City, a metropolitan municipality. Meanwhile, in Botswana, data were collected in three geographical areas: Southern (Kanye), the South East District (Ramotswa and Tlokeng), and Gaborone as shown in Figure 1. The socio-economic profiles of the two study areas are presented below. FIGURE 1. MAP SHOWING STUDY AREA SITES South Africa is known to be food secure at the country level as it produces enough food and even exports the surplus. However, these national figures often hide widespread hunger, especially at the individual household level. Food availability varies from urban (formal) and informal areas, rural (commercial farms), and traditional homelands. The urban areas have well-pronounced food distribution networks that ensure the availability of food all the time when compared to rural areas. The number of food-insecure households has been gradually increasing over the years, especially in predominantly rural regions and host the former homelands, such as the Eastern Cape Province. The province is chiefly rural, with 70% of its population residing in rural areas. It is ranked as one of the provinces experiencing the worst poverty in South Africa. According to the General http://issn.pdii.lipi.go.id/issn.cgi?daftar&1420518152&1&& ISSN: 2407-814X (p); 2527-9238 (e) 211 AGRARIS: Journal of Agribusiness and Rural Development Research Household Survey released by Statistics South Africa in May 2019, nearly 60% of households are reported to be receiving at least one social grant (Statistic South Africa, 2019). The province's unemployment rate is estimated at 35.4%, which is a figure that excludes individuals who have given up looking for employment. Accordingly, the actual unemployment rate is even higher at 42.5% if the expanded unemployment rate is factored in (Statistic South Africa, 2019). In the district municipalities where data collection was undertaken, the unemployment rates are as follows: Chris Hani 32.3%, OR Tambo 26.5%, and in the metropolitan city - Buffalo City 22.4% (Eastern Cape Socio Economic Consultative Council (Eastern Cape Socio Economic Consultative Council, 2017). In terms of poverty intensity, Chris Hani District is ranked at 43.1%, OR Tambo at 43%, and Buffalo City at 43%. Access to basic social amenities shows a huge disparity between rural and urban households, with 91% of households in BCM having access to piped water; 80.7% have access to electricity, and over 70.0% have access to sanitation services (Eastern Cape Socio Economic Consultative Council, 2019). Meanwhile, in district municipalities that are mainly rural, the proportion of households with access to basic social amenities is as follows: In Chris Hani, 72.7% of households have access to piped water, and 76.2% have access to electricity. In OR Tambo, 70.0% have access to electricity, only 11.6% of households have access to a flush or chemical toilet, and 37.7% have access to piped water. Within these settings, only 36% and 43.3 of the households in Chris Hani and OR Tambo, respectively, are involved in some form of subsistence agricultural production targeting to improve household food security (Eastern Cape Socio Economic Consultative Council, 2017). Both the two district municipalities and the province at large have great potential for agricultural production, which can contribute substantially to food security and employment. However, this potential remains relatively untapped. In fact, the study area has been experiencing some form of de-agrarian nation, and a recent research report by Statistics South Africa (2019) shows that only 27.9% of the province’s households are engaged in agricultural activities. Meanwhile, in Botswana, despite having high per capita income levels in Africa, the country grapples with the food security challenge at the country and household level. Unlike South Africa, which is food secure at the national level, Botswana is highly dependent on food imports, mainly because 70% of its total land comprises the Kalahari Desert. Botswana imports approximately 90% of its gross food supply from South Africa (Bahta et al., 2017). The same publication reveals that approximately 3500 square kilometers on the eastern and northern margins of Botswana is reported to be suitable for crop and livestock farming which is pursued by smallholder livestock producers. Livestock farming is the dominant agricultural activity in Botswana and it contributes about 70% of the agricultural GDP (Bahta et al., 2017; Statistics Botswana, 2017). Rainfall patterns are also sporadic and the country is plagued by water scarcity problems. These prevailing climatic conditions make the country susceptible to drought, thus dashing the hope of attaining food self-sufficiency (Nephawe, Mwale, Zuwarimwe, & Tjale, 2021). Accordingly, yields are poor, thereby exposing the masses, especially those in rural spaces, to hunger and malnutrition. When compared to their urban counterparts, households located in rural areas are the most vulnerable to hunger as food prices are reported to be much higher in rural spaces (Acquah et al., 2013; Bahta et al., 2017). The same authors http://issn.pdii.lipi.go.id/issn.cgi?daftar&1420518152&1&& ISSN: 2407-814X (p); 2527-9238 (e) 212 Household Food Insecurity Status and ….. (Ndhleve et al.) concur that the situation is further aggravated by the low income levels coupled with high unemployment rates prevailing in the rural areas, thereby limiting access to basic foodstuffs by the poor. Inequality is reported to be rampant, with about 20% of the population living below the poverty datum line, while 30% are either unemployed or under-employed (Statistics Botswana, 2017). This is the socio-economic context in which this study was undertaken. According to the Multi-Topic Household Survey, while poverty has been reduced to approximately 16% of the population, some 30% remain slightly above poverty line and are susceptible to a wide range of shocks (Statistic Botswana, 2018). Data Collection A cross-sectional survey design was employed with the aid of a semi structured questionnaire tool. Data was collected from 506 households in Botswana and 1557 households in South Africa. A stratified random sampling procedure aimed at stratifying households according to their socio- economic circumstances, namely, poor, better off and developed, was followed. In South Africa, data was collected from households in two District Municipalities, namely OR Tambo (662), Chris Hani (640), and one Metropolitan Municipality (255) that is Buffalo City, all located in the Eastern Cape Province. In Buffalo Municipality majority of the participants resided in informal settlements. In Botswana, data was collected in three geographical areas namely, Southern-Kanye (242) and South East District-Ramotswa and Tlokeng (108), and Gaborone (156). Respondents were the individuals responsible for food acquisition and preparations who were interviewed by trained enumerators. A pretested questionnaire was administered to collect data on household demographics,, socio- economic and food accessibility data using the Household Food Insecurity Access Scale (HFIAS). Data was analysed using SPSS. Descriptive statistical outputs such as: tables of frequencies, percentages and bar graphs on food security indicators covering accessibility, availability, utilisation and stability were generated. Separate linear and ordered logistic regression analyses were run for the two countries. HFIAS score, which is a continuous measure was used as the dependent variable for the linear regression whilst the categorical variable HFIAS prevalence was used as the dependent variable for the ordered logistic regression Measurement of Food Insecurity In this paper, household food insecurity status was measured by the HFIAS. The HFIAS measures the occurrence and frequency of food insecurity within households during the month prior to the survey (Coates et al., 2007). Four HFIAS indicators, namely Household Food Insecurity Access-related conditions (HFIA- conditions), Household Food Insecurity Access- related Domains (HFIA- Domains), Household Food Insecurity Access Scale Score (HFIAS score) and Household Food Insecurity Access Prevalence (HFIAS prevalence)) were employed to unpack the characteristics of and changes in household food security (access) of the sampled household. The indicators were computed for each household through questionnaire responses in relation to the various domains of food insecurity experience. The HFIAS score is a continuous measure derived from the sum of http://issn.pdii.lipi.go.id/issn.cgi?daftar&1420518152&1&& ISSN: 2407-814X (p); 2527-9238 (e) 213 AGRARIS: Journal of Agribusiness and Rural Development Research the responses to measurement questions, ranging from 0 to 27. A high score, the more food insecure (access) the household experienced and vice versa. Because HFIAS is a continuous measure, Household Food Insecurity Access Prevalence (HFIAP) was computed in order to classify households into food insecurity categories namely; food secure, mildly food insecure, moderately food insecure, severely food insecure. Regression Analysis The multiple linear regression was used to check for linear relationship between food security and the household socio economic factors. In this case, the assumption is that HFIAS score (dependent variable) have linear relationship dependent variables and is represented by the equation below where: yi=b0+b1 xi1+⋯+bz xiz+εi (1) whereby yi is the value of the i th case of the HFIAS score variable; z is the number of independent variables; bj is the value of the coefficient, j=0,1,...,z; xij is the value of the i th case of the jth independent variable and ↋i is the error term in the observed value. To ensure that all the determinants of household’ food insecurity status are captured accurately and conclusively, an ordered logistic regression was also used. The ordered categorical measure of household food insecurity namely HFIAS prevalence was used as a dependent variable. Since the dependent variable was categorical with more than two outcomes (1= secure; 2= mildly insecure; 3= moderately insecure; 4= severely insecure), the use of ordered regression was justified (Liu, 2009; Jega, Man, Latiff, & Wong, 2018). The same independent variables were used for both the linear and ordinal regression. The equation was modelled as follows: Y*=αj+Xβ+ε (2) Where, Y = observed ordinal variable which is a function of Y* = that is unobserved or unmeasured variable. The measurement model assumes that category 4>3>2>1Y =f(Y*) X = is the vector of independent variables β = is the regression coefficients to be estimated ε = is the error term αj = is the threshold or cut points Y* = is divided into some cut points or thresholds α1, α2, α3, α4, and α1< α2 < α3 < α4 Y ranging from 1 to 4. The food insecurity status category in which each household falls is expressed as: Y=1, Y*≤α1 2, α1 60 years 10(2.8) 34(9.5) 48(13.4) 267(74.4) 27(16.4) 58(35.2) 27(16.4) 53(32.1) Education No education 4(2.2) 16(8.6) 25(13.5) 140(75.7) 12(8.8) 70(51.1) 12(8.8) 43(31.4) Informal 0(0) 1(2.2) 7(15.2) 38(82.6) 6(13) 19(41.3) 9(19.6) 12(26.1) Primary 8(1.7) 24(5.1) 58(12.3) 383(81.0) 16(17.6) 27(29.7) 28(30.8) 20(22.0) Secondary 37(5.4) 83(12.1) 129(18.8) 437(63.7) 39(27.5) 54(38.0) 17(12.0) 30(21.1) Tertiary 17(10.6) 59(36.6) 31(19.3) 54(33.5) 35(38.0) 36(39.1) 12(13.0) 9(9.8) higher education levels are better placed to acquire employment which provides financial resources to ensure food accessibility. This finding might also be interpreted as signaling limited economic opportunities in both countries, that is Botswana and South Africa resulting in graduates being unemployed. Further clarification of this finding will be provided for by the regression analysis. Key Food Security Indicators The results of both linear and ordered logistic regression analyses with HFIAS and level of severity as the response variable and a set of indicator variables as explanatory variables for both Botswana and South Africa are shown in Table 4. Linear regression allowing for perceived categorical indicator and quantitative variables with the dependent variable being the HFIAS total score, which is a continuous http://issn.pdii.lipi.go.id/issn.cgi?daftar&1420518152&1&& ISSN: 2407-814X (p); 2527-9238 (e) 217 Household Food Insecurity Status and ….. (Ndhleve, et al.) variable. The linear regression analysis was employed to identify variables that are related to food insecurity and in continuous form. Ordered logistic regression representing the severity of food insecurity, and ranges from category secure (1) to severe insecurity (4), therefore ordinal variable implying the use of ordered logistic regression. TABLE 4. REGRESSION ANALYSIS RESULTS FOR HFIAS, HFIAP AND SOCIO-DEMOGRAPHIC CHARACTERISTICS VARIABLES SOUTH AFRICA BOTSWANA HFIAS SE Prevalence categories SE HFIAS SE Prevalence Categories SE Chris Hani (SA) and Southeast (BO) -2.424*** (0.367) -0.239* (0.143) -0.672 0.719 -0.326 0.279 BCM (SA) and Southern District (BO) 0.427 (0.526) 0.956*** (0.219) -1.128** 0.134 -5.859*** 1.142 Gender Female 1.276*** (0.332) 0.480*** (0.126) 0.788 0.526 0.196 0.223 Race Colored -1.543** (0.627) -0.865*** (0.233) White -2.688** (1.297) -1.179*** (0.392) Other 1.044 (5.772) 13.87 (761.6) Age 30-50 years 0.203 (0.455) 0.275 (0.170) 1.685* 0.995 0.490 0.367 51-60 years 0.389 (0.540) 0.330 (0.210) 1.840* 1.090 0.535 0.415 > 60 years -0.168 (0.551) 0.315 (0.216) 1.327 1.053 0.345 0.400 Education Informal -1.043 (0.570) -0.981 (0.578) -0.690 0.991 0.119 0.457 Primary -0.908* (0.501) -0.458** (0.217) -1.129 0.793 0.0520 0.378 Secondary -1.914*** (0.507) -0.321 (0.204) 02.209** 0.979 -0.547 0.431 Tertiary -4.853*** (0.683) -1.137*** (0.249) -0.972 1.184 -0.217 0.502 Income Total Income -1.76e-05*** 4.60e-06 -598e-06*** 1.84e-06 Salaries/ Wages -1.302** (0.542) -0.686*** (0.214) Pension/ Grants 0.463 (0.496) -0.00181 (0.208) Other 0.118 (0.696) -0.0303 (0.288) Disability status 0.910 (0.786) 0.990** (0.429) Livestock Cattle number -0.0790 (0.0562) -0.00321 0.0234) -0.0590** 0.0299 -0.0384*** 0.0126 Number of goats -0.0573* (0.0304) -0.0362*** (0.0130) -0.0284 0.0233 -0.00181 0.00899 Number of sheep_ -0.0241* (0.0123) -0.00910** (0.0043) -0.0383 0.0354 0.0184 0.0143 Number of pigs -0.0356 (0.0392) -0.0154 (0.0129) -0.109 0.799 -0.323 0.254 Number of chickens -0.0589** (0.0272) -0.0235** (0.0103) -0.00415 0.0145 0.00461 0.00618 Number of donkeys 0.0275 (0.236) -0.126 (0.0792) -0.787 0.0606 0.00176 0.0233 Number of horses 1.511 1.128 0.327 0.657 Household size 0.129** (0.0560) 0.0246 (0.0230) 0.163* 0.0957 0.127*** 0.0469 SEIAccess -0.450*** (0.0643) -0.118*** (0.0273) -0.886*** 0.146 -0.319*** 0.0661 Constant 18.79*** (1.175) 19.97*** 2.587 R-squared 0.273 0.25 *** p<0.01, ** p<0.05, * p<0. http://issn.pdii.lipi.go.id/issn.cgi?daftar&1420518152&1&& ISSN: 2407-814X (p); 2527-9238 (e) 218 AGRARIS: Journal of Agribusiness and Rural Development Research Firstly, household geographical location categorized according to district municipality explains a substantial variation in household food security across the investigated district municipalities. In South Africa, a household residing in Chris Hani was less food insecure by 2.424 in contrast to a household in OR Tambo, based on linear regression analyses. Similarly, the ordered logit for Chris Hani households being in a worse off food insecurity status category is 0.239 (p<1) less than OR Tambo households, while holding at constant the other variables in the model. The likelihood that households from Buffalo City are in a worse off food insecurity status is 0.956 (p<0.01) more than that of OR Tambo district municipality. In summary, the results reveal that Chris Hani residents are more food secure (less insecure) in contrast to the comparator district, OR Tambo. BCM residents are on average more food insecure than OR Tambo residents. The study results reveal substantial food security variability across district municipalities than has been found in previous research surveys. In Botswana, a household in the Southern District was less food insecure by 1.128 in contrast to a household in the Gaborone District, based on linear regression analyses. Similarly, ordered logit regression results explain that Southern District households are less severely food insecure (5.589; p<1) than Gaborone District residents whilst holding other variables held constant. Regarding geographical location, residing in the Southern District of Botswana significantly affects household food insecurity status under the order logit model and therefore is the only coefficient interpreted. This finding is inconsistent with existing empirical evidence that has established that households that are located further away from all- weather roads that ensure access to markets are more food insecure when compared to those located close to better infrastructure (Bahta et al., 2017). This finding highlights the rise of urban food insecurity, especially for those living within the low income neighborhoods (Jonah & May, 2019). Study participants from the urban settlement were mainly from the low income households as it was difficult to engage high income households due to security fences around the households. Indeed, urban areas do have adequate availability of foodstuffs which are not always accessible to everyone within the settings. Moreover, the cost of living in urban settlement is higher as compared to that of rural settlement. This affects the household food security status as some of the money is diverted to other household needs (Chakona & Shackleton, 2019). Accordingly, this might explain the findings of this study that show households residing in Gaborone and Buffalo Metropolitan were more food insecure when compared to other regions that are largely rural. Meanwhile, the explanation that can be used to elucidate the finding that households in Chris Hani are more food secure when compared to those in OR Tambo region, despite the fact that both are largely rural districts, is that data collection in this area was done in locations that are close to rural towns. Here, the food retail environment is more pronounced when compared to rural areas that are located further away from such small rural towns. Secondly, gender emerged as another significant correlate of food insecurity in this study but this was only the case in South Africa. Female-headed households are 1.276 scores more food insecure compared to the male-headed households and the ordered logit for females being more food insecure is 0.480 more than males while holding mother model variables http://issn.pdii.lipi.go.id/issn.cgi?daftar&1420518152&1&& ISSN: 2407-814X (p); 2527-9238 (e) 219 Household Food Insecurity Status and ….. (Ndhleve, et al.) constant. Such an observation is agreement with previous studies that also observed the feminisation of poverty in South Africa (Statistic South Africa, 2018). The feminisation of poverty in South Africa implies that female headed households have limited and unequal access to resources, some of which affect food production and access. Poverty and food insecurity are interlinked. This phenomenon is expressed more within the rural areas, wherein household head gender is a determinant of level of food security (Tibesigwa and Visser, 2016; Kassie et al., 2012; Altman, Hart, & Jacobs, 2009). Size of Household size was also a significant determinant of household food security in both Botswana and South Africa. A bigger household was found to be more food insecure than smaller household. This may be due to more members that are economically inactive possibly children and old people signifying that such households have a higher dependence ratio. Larger households imply more people to feed and more resources needed in order to ensure food availability. In South Africa, regarding race, Coloureds and Whites are more food secure than Black Africans. Relative to black Africans, Coloureds are -1.543 less food insecure or -0.865 less likely to be food insecure, while Whites are -2.688 less insecure or -1.179 less likely (odds) to be food insecure. The coefficients for White households are bigger (more negative) compared to Coloureds thus, implying White households are way better off (more food secure) than other population groups. These results are in agreement with the report by Statistic South Africa, which revealed that white-headed households have a higher average expenditure (four times higher) as compared to that of-black headed households in South Africa ( Statistic South Africa A, 2018). There were no observed significant differences in household food insecurity between different races in Botswana. Fourthly, a higher educational qualification increases the chances of being food secure, compared to lack of formal education in South Africa but there were no observed significant effects of education status on household food insecurity in Botswana. Relative to the absence of formal education, households with primary education, secondary education and tertiary education are more food secure in that order. The results for informal education were insignificant. These findings were aligned with the observations by Nwokolo, (2015) and (Ruhyana, Essa, & Mardianis, 2020) that higher levels of education are associated with increased household income, livelihood opportunities and food security. When analysed as a quantitative variable, income level significantly reduces household food insecurity in Botswana. High income levels are associated with increased household food security and vice versa (Fraval, Oosting, de Boer, Lannerstad, & van Wijk, 2019). Unlike Botswana, household income data in South Africa data was collected as a categorical variable. Compared to households relying on agriculture as their main source of income, having salaries/ wages and trade income (i.e. non-agricultural income) reduces food insecurity by -1.3 points and this was significant at p<0.01 in South Africa. This could be due to the structure of agricultural production among rural areas of South Africa which is mainly seasonal and rainfed such that seasonal production must provide for the family needs for the rest of the year (du Toit, 2019). In comparison to agricultural production, salaries and wages provide http://issn.pdii.lipi.go.id/issn.cgi?daftar&1420518152&1&& ISSN: 2407-814X (p); 2527-9238 (e) 220 AGRARIS: Journal of Agribusiness and Rural Development Research income and the means of accessing food throughout the year. Similarly, livestock ownership, such as cattle, goats, pigs and donkeys, reduces food insecurity in South Africa. In Botswana, only cattle ownership significantly reduces food insecurity. Livestock ownership is known in Africa for providing a hedge for small scale farmers against climatic risk and acting as a store of wealth which can be converted to required resources which include food and cash (Gwiriri, Bennett, Mapiye, Marandure, & Burbi, 2019). Households which own livestock are likely to dispose off in the event of food shortage in exchange for food. The results strongly show that food security dynamics have context specific manifestations mainly linked to household size and household food environment. The common determinants of food security for Botswana and South Africa that emerged from the analyses were household size, geographical location and socio-economic infrastructure (SEI). The variables in the ordered logistic regression model accounted for about 27% (South Africa) and 25% (Botswana) of the variance in food insecurity. Marsden and Sonnino (2012), argued that food security has spatial attributes, since its production, distribution and consumption depend on household geographical location. Access to socio-economic infrastructure reduces food insecurity; this resonates with Sustainable Livelihoods Framework (Scoones, 1998), which argue that access to many streams of assets is a pre-requisite for survival. CONCLUSION This study provides a valuable baseline for critical insights and comparative analyses of household food security status for Botswana and South Africa and across different socio- economic groups. The applicability of both the HFIAS and HFIP scale cannot be overemphasized. There is a high prevalence of food insecurity in both countries; however, the determinants of food security are different between the two countries. And in cases when they are similar, the magnitude of the effect is not similar. Geographical location, household size; gender, race, education; source of income; livestock ownership and socio-economic infrastructure (SEI) emerged as significant factors in South Africa and geographical location; household size; age; source of income; livestock ownership and SEI significantly predict household food security in Botswana. The list of significant indicators reveals similarities with previous studies on food security although differences in terms of magnitude can be observed. Replicating this study at micro scale could allow a complete outline of determinants of food security and their magnitudes. Attaining food security in both countries requires the adoption of strategies or programmes that are cognisant of these variables in order of their magnitude as they emerge from the analyses. The micro scale analyses should put emphasis on factors that emerged as significant predictors of household food security as per each country in order of their magnitude. Policymakers and program administrators should desist from simulating outcome of food security indicators across Southern Africa or any geographical jurisdiction without due analyses. Declaration of interest statement: There is no potential conflict to declare. http://issn.pdii.lipi.go.id/issn.cgi?daftar&1420518152&1&& ISSN: 2407-814X (p); 2527-9238 (e) 221 Household Food Insecurity Status and ….. (Ndhleve, et al.) Acknowledgement: This paper was drawn out of a bigger research project titled, “The African Food Security barometer: Towards Policy Driven”, funded by the Association of African Universities-AAU. We would like to acknowledge the study participants from Botswana and South Africa for making this study a success. REFERENCES Abdulai, A., Barrett, C. B., & Hazell, P. (2004). Food Aid for Market Development in Sub-Saharan Africa (No. DSGD Discussion Papper No.5). Washington, D.C. Retrieved from http://ebrary.ifpri.org/utils/getfile/collection/p15738coll2/id/71151/filename/71152.pdf Acquah, B., Kapunda, S., Legwegoh, A., Gwebu, T., Modie-Moroka, T., Gobotswang, K., & Mosha, A. (2013). The State of Food Insecurity in Gaborone, Botswana (J. Crush, ed.). Cape Town: African Food Security Urban Network (AFSUN). Retrieved from http://www.afsun.org/wp-content/uploads/2016/06/AFSUN17.pdf Ahmed, F. F., Eugene, C. E., & Abah, P. O. (2007). Analysis of Food Security among Farming Households in Borno State, Nigeria. Agricultural Economics, Environment and Social Sciences, 1(1), 130–141. Altman, M., Hart, T. G. B., & Jacobs, P. T. (2009). Household food security status in South Africa. Agrekon, 48(4), 345–361. https://doi.org/10.1080/03031853.2009.9523831 Asefa, S. (1991). Enhancing food access in Africa: The Botswana experience. Studies In Comparative International Development, 26, 59–83. https://doi.org/10.1007/BF02687175 Bahta, S., Wanyoike, F., Katjiuongua, H., & Marumo, D. (2017). Characterisation of food security and consumption patterns among smallholder livestock farmers in Botswana. Agriculture & Food Security Food Security, 6. https://doi.org/10.1186/s40066-017-0145-1 Chakona, G., & Shackleton, C. M. (2019). Food insecurity in South Africa: To what extent can social grants and consumption of wild foods eradicate hunger? World Development Perspectives, 13, 87–94. https://doi.org/10.1016/j.wdp.2019.02.001 Coates, J., Swindale, A., & Bilinsky, P. (2007). Household Food Insecurity Access Scale (HFIAS) for Measurement of Household Food Access: Indicator Guide (v. 3). Washington, D.C.: FHI 360/FANTA. Retrieved from https://www.fantaproject.org/sites/default/files/resources/HFIAS_ENG_v3_Aug07.pdf du Toit A. (2019) Agriculture, Value Chains and the Rural Non-Farm Economy in Malawi, South Africa and Zimbabwe. In: Scholvin S., Black A., Revilla Diez J., Turok I. (eds) Value Chains in Sub-Saharan Africa. Advances in African Economic, Social and Political Development. Springer, Cham. https://doi.org/10.1007/978-3-030-06206-4_12 Eastern Cape Socio Economic Consultative Council. (2017). The Eastern Cape socio economic review and outlook, 2017. Retrieved March 21, 2019, from Eastern Cape Development Corporation website: http://www.ecdc.co.za/media/3749/dedeat-sero- report-2017_digital-version.pdf Eastern Cape Socio Economic Consultative Council. (2019). The economic review of the Eastern Cape. Retrieved October 2, 2019, from Eastern Cape Socio Economic Consultative Council website: http://www.ecsecc.org/documentrepository/informationcentre/ecsecc- gdpreport2019q1_45652.pdf http://issn.pdii.lipi.go.id/issn.cgi?daftar&1420518152&1&& ISSN: 2407-814X (p); 2527-9238 (e) 222 AGRARIS: Journal of Agribusiness and Rural Development Research FAO, IFAD, UNICEF, WFP, & WHO. (2018). The State of Food Security and Nutrition in the World 2018. Building climate resilience for food security and nutrition. Rome: FAO. Retrieved from https://www.fao.org/3/I9553EN/i9553en.pdf FAO, IFAD, UNICEF, WFP, & WHO. (2019). The State of Food Security and Nutrition in the World 2019. Safeguarding against economic slowdowns and downturns. Rome: FAO. Retrieved from https://www.fao.org/3/ca5162en/ca5162en.pdf Fraval, S., Oosting, S. J., de Boer, I. J. M., Lannerstad, M., & van Wijk, M. T. (2019). Food security in rural sub-Saharan Africa: a household level assessment. Rade-Offs in Science - Keeping the Balance: Abstracts of the WIAS Science Day 2019, 14–14. Lunteren: Wageningen University & Research. Godfray, H. C. J., Beddington, J. R., Crute, I. R., Haddad, L., Lawrence, D., Muir, J. F., … Toulmin, C. (2010). Food Security: The Challenge of Feeding 9 Billion People. Science, 327(5967), 812–818. https://doi.org/10.1126/science.1185383 Gonzalez, C. G. (2015). World Poverty and Food Insecurity. Penn State Journal of Law & International Affairs, 3(2), 56–83. Grebmer, K. von, Bernstein, J., Mukerji, R., Patterson, F., Wiemers, M., Chéilleachair, R. N., … Fritsche, H. (2019). 2019 Global Hunger Index: The Challenge of Hunger and Climate Change. Bonn, Dublin: Welthungerhilfe, Concern Worldwide. Grobler, W. C. J. (2016). Perceptions of Poverty: A Study of Food Secure and Food Insecure Households in an Urban Area in South Africa. Procedia Economics and Finance, 35, 224– 231. https://doi.org/10.1016/S2212-5671(16)00028-9 Gwiriri, L. C., Bennett, J., Mapiye, C., Marandure, T., & Burbi, S. (2019). Constraints to the sustainability of a ‘systematised’ approach to livestock marketing amongst smallholder cattle producers in South Africa. International Journal of Agricultural Sustainability, 17(2), 189–204. https://doi.org/10.1080/14735903.2019.1591658 Human Sciences Research Council (HSRC). (2015). Food Security in South Africa: Key Policy Issues for the Medium Term–Position Paper. Pretoria, Human Sciences Research Council. Jega, A. A., Man, N., Latiff, I. A., & Wong, K. K. S. (2018). Flood Disaster Effect on SmallholderFarmers’ Food Security inKelantan: An Ordered Logistic Regression Analysis. IOSR Journal of Agriculture and Veterinary Science (IOSR-JAVS), 11(3), 42–50. Jonah, C. M. P., & May, J. D. (2019). The nexus between urbanization and food insecurity in South Africa: does the type of dwelling matter? International Journal of Urban Sustainable Development, 12(1), 1–13. https://doi.org/10.1080/19463138.2019.1666852 Kassie, M., Ndiritu, S., Sheferaw, B., and Bekele, A. (2012). Determinants of food security in Kenya. A gender perspective. 86th Annual Conference, April 16-18, 2012, Warwick University, Coventry, UK 135124, Agricultural Economics Society. DOI: 10.22004/ag.econ.135124 Koch, J. (2012). The Food Security Policy Context in South Africa. International Policy for inclusive Growth, Brazil. Country study 21. Accessed on March 13, 2019. https://ipcig.org/pub/IPCCountryStudy21.pd http://issn.pdii.lipi.go.id/issn.cgi?daftar&1420518152&1&& https://ipcig.org/pub/IPCCountryStudy21.pd ISSN: 2407-814X (p); 2527-9238 (e) 223 Household Food Insecurity Status and ….. (Ndhleve, et al.) Liu, X. (2009). Ordinal regression analysis: fitting the proportional odds model using Stata, SAS and SPSS. Modern Applied Statistical Methods, 8(2), 632–645 DOI: 10.22237/jmasm/1257035340 Marsden, T., and Sonnino, R. (2012). Human health and wellbeing and the sustainability of urban–regional food systems. Current Opinion in Environmental Sustainability 2012, (4): 427–430 Nephawe, N., Mwale, M., Zuwarimwe, J., & Tjale, M. M. (2021). The impact of water-related challenges on rural communities food security initiatives. AGRARIS: Journal of Agribusiness and Rural Development Research, 7(1), 11–23. https://doi.org/10.18196/agraris.v7i1.9935 Nwokolo, E. E. (2015). The Influence of Educational Level on Sources of Income and Household Food Security in Alice, Eastern Cape, South Africa. Journal of Human Ecology, 52(3), 208–217. https://doi.org/10.1080/09709274.2015.11906944 Ruhyana, N. F., Essa, W. Y., & Mardianis, M. (2020). Sociodemographic Factors Affecting Household Food Security In Sumedang Regency West Java Province. AGRARIS: Journal of Agribusiness and Rural Development Research, 6(1). https://doi.org/10.18196/agr.6189 Scoones, I. (1998). Sustainable rural livelihoods: A framework for analysis. Brighton: Institute of Development Studies, University of Sussex Sen, A. (1990). Gender and Cooperative Conflict. In: Tinker, I., Ed., Persistent Inequality, Oxford University Press Oxford, 123-148. Statistics Botswana. (2017). Botswana demographic survey report 2017. Statistics Botswana, Gaborone. Accessed on March 21, 2019. http://www.statsbots.org.bw/sites/default/files/ Botswana%20Demographic%20Survey%20Report%202017.pdf Statistics Botswana. (2018). Report. Botswana multi-topic household survey 2015/2016. Statistics Botswana, Gaborone. Accessed on March 21, 2019 http://www.statsbots.org.bw/sites/default/files/publications/BMTHS%20POVERTY %20STATS%20BRIEF%202018.pdf Statistic South Africa. (2018). Men, Women and Children: Findings of the Living Conditions Survey 2014/15. Statistics South Africa Library Cataloguing-in-Publication (CIP) Data Report no.: 03-10-02 (2014/15). Statistics South Africa, Pretoria. Accessed on March 21, 2019 http://www.statssa.gov.za/publications/Report-03-10-02%20/Report-03-10-02%202015.pdf Statistic South Africa. (2019). Towards measuring the extent of food security in South Africa: An examination of hunger and food adequacy/Statistics South Africa. Pretoria: Statistics South Africa, Accessed on November 20, 2019. http://www.statssa.gov.za/publications/03-00-14/03-00-142017.pdf The Economist Intelligence Unit. (2015). Global food security index 2015: An annual measure of the state of global food security. New York. Retrieved from https://foodsecurityindex.eiu.com/Home/DownloadResource?fileName=EIU Global Food Security Index - 2015 Findings %26 Methodology.pdf The Economist Intelligence Unit. (2018). Global Food Security Index 2018: Building resilience in the face of rising food-security risks. New York. Retrieved from https://foodsecurityindex.eiu.com/ Tibesigwa, B., and Visser, M. (2016). Assessing gender inequality in food security among small- holder farm households in urban and rural South Africa. World Development, 88(C), 33- 49. DOI: 10.1016/j.worlddev.2016.07.008 http://issn.pdii.lipi.go.id/issn.cgi?daftar&1420518152&1&& http://www.statsbots.org.bw/sites/ http://www.statsbots.org.bw/sites/default/files/publications/BMTHS%20POVERTY%20STATS%20BRIEF%202018.pdf http://www.statsbots.org.bw/sites/default/files/publications/BMTHS%20POVERTY%20STATS%20BRIEF%202018.pdf http://www.statssa.gov.za/publications/Report-03-10-02%20/Report-03-10-02%202015.pdf http://www.statssa.gov.za/publications/03-00-14/03-00-142017.pdf ISSN: 2407-814X (p); 2527-9238 (e) 224 AGRARIS: Journal of Agribusiness and Rural Development Research van der Berg, S. (2006). Public spending and the poor since the transition to democracy. In H. Bhorat & R. Kanbur (Eds.), Poverty and Policy in Post-Apartheid South Africa. Pretoria: Human Sciences Research Council Press. World Bank. (2015). Ending Poverty and Hunger by 2030. An Agenda for the global food systems. Washington DC, World Bank. Accessed on March 21, 2019. http://documents.worldbank.org/curated/en/700061468334490682/pdf/95768- REVISED-WP-PUBLIC-Box391467B-Ending-Poverty-and-Hunger-by-2030-FINAL.pdf http://issn.pdii.lipi.go.id/issn.cgi?daftar&1420518152&1&&