Journal of Food and Nutritional DisordersISSN: 2324-9323

All submissions of the EM system will be redirected to Online Manuscript Submission System. Authors are requested to submit articles directly to Online Manuscript Submission System of respective journal.

Research Article, J Food Nutr Disor Vol: 5 Issue: 5

Following the Crisis: Poverty and Food Security in Harare, Zimbabwe

Godfrey Tawodzera1, Liam Riley2* and Jonathan Crush3
1Senior Lecturer, Department of Geography, University of Limpopo, South Africa
2Banting Postdoctoral Fellow, Balsillie School of International Affairs, Wilfrid Laurier University, Canada
3CIGI Chair in Global Migration and Development, Balsillie School of International Affairs, Wilfrid Laurier University, Canada
Corresponding author : Dr. Liam Riley
Banting Postdoctoral Fellow, Balsillie School of International Affairs, Wilfrid Laurier University, 67 Erb Street West, N2L 6C2, Waterloo, Canada
Tel: 1-226-772-3080
E-mail: lriley@balsillieschool.ca
Received: July 25, 20164 Accepted: August 16, 2016 Published: August 22,2016
Citation: Tawodzera G, Riley L, Crush J (2016) Following the Crisis: Poverty and Food Security in Harare, Zimbabwe. J Food Nutr Disor 5:5. doi:10.4172/2324-9323.1000208

Abstract

Following the Crisis: Poverty and Food Security in Harare, Zimbabwe

Household food security in African cities has received increasing academic and policy attention in the past decade as the continent rapidly urbanizes. The African Food Security Urban Network has played a leading role in producing empirical research on the extent of household food insecurity and on its causal factors, but to date it has produced little longitudinal data. This paper addresses this gap by presenting the results of household food security surveys conducted in low-income neighbourhoods in Harare, Zimbabwe in 2008 and 2012. The analytical focus is on the changes that took place from the “crisis” situation in 2008, when the formal sector economy virtually ceased to function, to the situation in 2012 after new economic policies and a political detente had led to economic stabilization. The results show an overall improvement in food security but with important qualifications, such as the continued importance of non-monetized and informal food sources, continued problems with access to basic services and infrastructure, and the accrual of food security gains mostly among wealthier households.

Keywords: Food security; Crisis; Food insecurity; Food sources

Keywords

Food security; Crisis; Food insecurity; Food sources

Introduction

The African Food Security Urban Network (AFSUN) (www.afsun. org) was established in 2008 to broaden policy and academic discourses on food security in Africa, which has tended to focus on production challenges and rural development with an implicit assumption that because food is widely available in urban areas, hunger is not a significant problem [1]. Drawing on the broad definition of food security that guides the FAO-including the four pillars of availability, access, utilization, and stability -AFSUN’s research contributions to date have demonstrated that food insecurity is a significant problem in the low-income settlements that characterize many Southern African cities [2,3].
In 2008, AFSUN devised and implemented a household food security baseline survey, interviewing 6,453 urban households in nine countries in eleven cities across Southern Africa [3]. The survey gathered extensive information on household structure and composition, food sources, household poverty, income sources and household expenditures. The result was a large dataset amenable to analysis of various food-related thematic issues including urban agriculture [4,5], the supermarket revolution [6,7], HIV/AIDS and food insecurity [8], food governance [9], the impacts of infrastructural deficits on food security [10], migration and food security [11], the relationships between gender and food security [12,13] and the impact of economic shocks [14,15].
Few cases illustrate the underlying link between economic shocks and food insecurity better than Zimbabwe [15,16]. The causes of the country’s post-2000 economic and political crisis have been debated at length in the literature and need not detain us here [17-22]. However, its scope and scale need to be emphasized: there was a 40 percent contraction in GDP between 2000 and 2006; annual inflation increased from two-digit figures in 2000 to 231 million percent in July 2008; and external debt ballooned to USD 6 billion in 2008 [23]. Formal sector employment shrunk from 3.6 million in 2003 to only 480,000 in 2008. The economic crisis, which reached its nadir in 2008, was exacerbated by political oppression and conflict between the ruling ZANU-PF party and opposition groups seeking a change in government [24,25].
The effects of the economic and political crisis on ordinary Zimbabweans were profound, given the massive job losses resulting from economic decline, increases in other urban costs such as housing, water, electricity and transportation, and hyperinflation [22,26-29]. Out-migration from the country increased rapidly and many households became dependent on the remittances sent from South Africa, the UK and further afield [30]. Food remittances and the informal economy became critical to everyday survival [31,32]. The informal economy exploded in size, strength and diversity, though the Zimbabwean government’s assault on all forms of urban informality -called Operation Murambatsvina -interrupted its expansion for a brief period [29,33].
For most of 2008, the market was characterized by constant staple food shortages. Domestic food production continued to fall, shops virtually emptied of food, and hyper-inflation made it all but impossible to purchase what was left. The formal food system virtually collapsed and most foodstuffs could only be accessed on the parallel market. The situation was particularly grave in urban areas where households had to purchase most of their food. AFSUN implemented its baseline household food security survey in lowincome areas of Harare in Zimbabwe in late 2008. Among the 462 households surveyed, rates of food insecurity were extremely high, with 96% being food insecure [34]. Dietary diversity was lower than in any other city in Southern Africa. Households in low-income urban areas in Harare were far worse off in terms of all the food insecurity and indicators of poverty than households in the other ten Southern African cities surveyed by AFSUN.
After 2008, Zimbabwe’s political and economic situation began to improve. The complete collapse of the economy was averted in 2009 with the abandonment of the Zimbabwean dollar and a power-sharing agreement between the ruling and major opposition parties [25,35]. A Government of National Unity was inaugurated in February 2009 and was to last until 2013. The formation of this coalition helped stabilize the economy, arresting the precipitous decline in GDP, controlling inflation, introducing a multi-currency regime and improving the food supply [36]. Between 2009 and 2011, Zimbabwe’s GDP growth averaged 7.3%, making it one of the world’s fastest growing economies, albeit from a very low base. According to the World Bank, Zimbabwe experienced an economic rebound after 2009 “and with the support of record international price levels, exports of minerals- notably diamonds, platinum, gold, and other products- have injected new life into the economy” [36]. Zimbabwean trade flows rebounded with exports rising at 39% per year. Imports also rose quickly, averaging 34% per year from 2009 to 2011, in response to domestic demand. As the economy stabilized, domestic food production increased and shops restocked with food imported primarily from South Africa.
In order to assess whether political stability, renewed economic growth and the return of food to the stores were impacting positively on the food security of the urban poor, AFSUN repeated its household survey in 2012 in the same low-income neighbourhoods of Harare surveyed in 2008. This paper presents a comparative analysis of the 2008 and 2012 survey results. The central question is whether the desperate food security situation in Zimbabwe’s urban centres at the height of the crisis improved with political stability and partial economic recovery. In other words, did positive macro-economic trends translate into ground-level improvements in incomes, poverty levels and food security?

AFSUN Methodology

The 2008 and 2012 surveys in Harare were conducted using the same questionnaire and sampling procedure. In 2008, a total of 462 households were interviewed in three low-income neighbourhoods of Harare- Mabvuku, Tafara, and Dzivarasekwa. In 2012, a broader area of the city was sampled. For the purposes of comparison in this paper, we have extracted the data for households in the same three low-income neighbourhoods. As a result, the total number of households interviewed in 2012 was 351, slightly lower than in 2008 but still sufficient to make statistically meaningful comparisons. The AFSUN survey instrument used several standardized household food security assessment tools including the Household Food Insecurity Access Scale (HFIAS); Household Food Insecurity Access Prevalence (HFIAP); and the Household Dietary Diversity Score (HDDS) [37-40]. As well as collecting detailed demographic, socio-economic and health information about each member of the household, the AFSUN survey instrument asked about household food purchasing patterns including frequency of patronage of different types of retail outlet, the prevalence of urban agriculture, income and expenditure including on food, lived poverty (using the Lived Poverty Index), informal ruralurban and urban-urban food remittances and livelihood strategies.
Changing patterns of food sourcing
As in urban areas throughout Southern Africa, Harare households rely on a variety of food sources in the formal and informal sectors, and these food sources shape the stability, affordability, and accessibility of food by the urban poor [7]. Comparison of food sourcing patterns in 2008 and 2012 reflect important shifts that coincide with the restoration of the formal sector economy following dollarization. In particular, there was a significant increase in food sourcing from supermarkets and small retail outlets between 2008 and 2012. In 2008, only 30% of households accessed food at supermarkets, whereas in 2012 the vast majority (92%) were using supermarkets (Figure 1). Comparing Harare with other Southern African cities, it would appear that the low rates of patronage of supermarkets in 2008 was an anomaly due to the acute economic crisis and that Zimbabweans quickly resumed patronizing supermarkets once the option became available. Small shops and takeaways also increased dramatically in popularity, from being patronized by only 17% of households in 2008 to 73% of households in 2012.
Figure 1: Changes in Household Food Source.
Supermarkets are still used less frequently than other sources in part because they are often located in formally planned areas at some distance from the informal areas surveyed. The infrequent patronage of supermarkets also suggests that people access dry goods and bulk food supplies rather than fresh foods for day to day consumption. In both 2008 and 2012, the most common frequency of shopping at the supermarket (exclusive of “never”), was once a month (40% in 2008 and 48% in 2012) (Figure 2). This is in keeping with the frequency of wage and social grant payments and also suggests a pattern of buying staples, such as maize meal, in bulk on a monthly basis. The proportion of households shopping at supermarkets at least once a week decreased slightly from 29% to 27%. The sharpest increase was in the proportion of households shopping at supermarkets at least five days per week; this increased from 7% to 20% of households that patronized supermarkets.
Figure 2: Patrons’ Frequency of Using Food Sources.
Small retail outlets were used more frequently than supermarkets, although the most common use pattern shifted from monthly in 2008 to daily in 2012 (with an increase in daily use from 18% to 44%). The increased patronage of supermarkets and small shops by 2012 did not appear to displace other food sources, however, but rather expanded the range of sources used (Figure 1). Informal markets/street food remained the most popular source (97% of households in 2008 to 94% in 2012). In contrast to the use of supermarkets on a monthly basis, informal sources were most likely to be used on a daily basis. Informal vendors are definitely more accessible and offer foods that people consume on a daily basis, such as fresh produce and meat, and in small affordable quantities. Informal vendors are also likely to facilitate price negotiation, offer flexibility in the quantity of food purchased, and provide informal credit arrangements [31,41-45].
Urban agriculture (UA) has been a feature of the urban landscape in Harare since the early 1990s and continues to be an important supplemental food source [46-48]. In recent years, UA has been associated primarily with low-income households with inadequate financial resources and insecure livelihood opportunities [49,50]. The 2008 survey found that well over half (60%) of the households were engaged in UA as a food source (Figure 1). Furthermore, 70% of those producing some of their own food accessed it on at least a weekly basis. By 2012, however, the proportion of households reporting UA as a food source had declined to 46%; and only 59% of these households accessed their own production on at least a weekly basis.
Although still very high by regional standards, UA appears to have quickly declined in prevalence as purchased food became more freely available. Another indication of the UA decline was evident in responses to the question: “To what extent does the household rely on field crops and garden crops as additional livelihood strategies?” In the 2008 survey, 45% and 47% of households relied to some degree on garden crops and field crops respectively, whereas in 2012 these figures had fallen to 28% and 23% (Table 1). The reduction in the importance of UA suggests that for a subset of households it was a short-term response to an acute economic crisis. However, the continued importance of UA for many households after the economic crisis of 2008 suggests it is an enduring part of urban life for a sizeable number of households in Harare’s low-income neighbourhoods.
Table 1: Reliance on Urban Agricultural Livelihood Activities.
Food remittances are another important alternative to marketbased food sources in southern African cities [32,51,52]. The proportion of households receiving food remittances increased marginally from 42% in 2008 to 47% in 2012 (Table 2). Most of this came from an increase in households receiving food remittances from relatives in rural areas. Although there is considerable controversy about the impact of the Zimbabwean land reform, which saw over 4,000 white commercial farmers lose their farms, there is an emerging consensus that resettled black farmers are producing a great deal more than they used to produce [53]. This could explain the continued and even increased flow of food remittances from rural to urban households. Alternatively, the increase in 2012 may simply have reflected a better agricultural season in 2012 than in 2008 [54]. Although levels of food remitting increased slightly between 2008 and 2012, the proportion of recipient households who said they were very important or critical to survival dropped from 83% to 49% (Figure 3).
Table 2: Food Remittances from Rural and Urban Areas.
Figure 3: Importance of Food Remittances.
Changes in household food security status
Food security in Harare improved in 2012 relative to 2008 on all measurements. The comparison with the 2008 regional AFSUN survey suggests that Harare was in an extremely bad situation and by 2012 it had returned to a rate of food insecurity close to the regional norm for low-income neighbourhoods (Table 3). In 2008, Harare had a mean HFIAS of 14.7 and a median of 16.0. Only Manzini in Swaziland (a country ravaged by HIV and AIDS) had a higher mean (14.9) than Harare although its median score was lower (14.0). In 2012, both the mean and median scores in Harare were considerably lower, at 9.6 and 10.0 respectively, and seven other cities had higher mean and median HFIAS scores.
Table 3: HFIAS Scores in AFSUN City Surveys.
The HFIAP scores provide further insight into the absolute and relative improvement in food security status in Harare. In 2008, Harare had the lowest number of food secure households (2%) and the second highest number of severely food insecure households (72%). In 2012, the share of food secure households had increased to 10% and the proportion of severely food insecure households fell from 72% to 63% (Figure 4). The continuing high proportion of severely food insecure households in 2012 complicates the narrative of significantly improved food security suggested by the change in HFIAS scores. The discrepancy in the picture presented by changes in the HFIAS as opposed to the more incremental HFIAP redistribution is consistent with an argument that food security gains accrued primarily to a small group of households that benefited from economic stabilization, a point that is explored further in the discussion section below.
Figure 4: Distribution of Households in HFIAP Categories.
Aggregate household dietary diversity also appears to have improved between 2008 and 2012. The mean HDDS rose from 4.8 in 2008 to 6.5 in 2012 and the median HDDS from 5 in 2008 to 6 in 2012. A comparison of the distribution of HDDS scores at the two points in time shows that this was primarily a result of a significant fall in the number of households with extremely low HDDS scores and a rise in the proportion of households at the high end of the scale (Figure 5). While 29% of households were in the extremely low range of 1-3 in 2008, only 9% were still in this category in 2012. Only 12% of households had scores of 8 and above in 2008, compared to 30% in 2012.
Figure 5: Distribution of Household Dietary Diversity Scores.
The improvement in dietary diversity is reflected in more widespread consumption of foods from almost every food group (Table 4). The only food group consumed by a lower proportion of households in 2012 than in 2008 was vegetables (consumed by 92% of households in 2008 and by 83% in 2012). The most substantial increases were in the consumption of dairy products (12% in 2008 and 39% in 2012), meat (22% in 2008 and 50% in 2012), fruits (15% in 2008 and 41% in 2012), and sugar or honey (64% in 2008 and 83% in 2012). With the arguable exception of increased fruit consumption, the foods with the sharpest increases conform to a more typical urban diet associated with the dietary transition taking place throughout the Global South [55,56]. Even as food becomes more abundant and accessible, the increased consumption of fatty, calorie-dense, and processed foods could well be accelerating a nutritional transition with negative impacts on health in the longer term.
Table 4: Food Groups Consumed in Previous 24 Hours.
Livelihoods, incomes and expenditures
Can these changes in the food security status of households between 2008 and 2012 be explained with reference to changes in livelihoods, employment and incomes? Secure employment, reliable income sources, and an adequate food budget are certainly critical for ensuring household food security for most urban households who primarily rely on purchased foods to meet their nutritional needs [57]. A comparison of the 2008 and 2012 employment profile of household members suggests, however, that little changed in the labour market prospects for people aged 18-65.
Overall employment was only slightly different in 2012 (59% employed) than it had been in 2008 (58% employed). Unemployment figures were also very similar (at 42% in 2008 and 40% in 2012) (Table 5). There were clear shifts in patterns of employment. First, amongst the employed there was a move away from full-time towards greater part-time employment. The proportion of all working age adults employed full-time fell from 43% to 35% between 2008 and 2012 and the proportion employed parttime increased from 15% to 24%. Second, amongst the unemployed, a higher proportion was looking for work in 2012 (19%) than in 2008 (15%). There is a dearth of official employment statistics with which to compare these figures but they are strongly suggestive of an increasingly precarious employment market.
Table 5: Work Status of Household Members Aged 18-65.
Urban households worldwide reduce their vulnerability by drawing on multiple types of income sources rather than relying solely on paid employment [58-60]. Having a wider range of income sources means that if one source fails, for example through job loss or illness, the household has other sources to draw on to mitigate the consequences. In the context of Harare, multiple sources of income were necessary in order for a household to survive during the crisis of 2008 when the value of money was almost worthless [14,16]. In 2012, the pattern of multiple household income sources persisted. The proportion of households deriving income from informal sector businesses fell from 42% in 2008 to 34% in 2012, indicating a reduction in the importance of the informal sector, which had become critically important for survival during the crisis. At the same time, the fact that one-third of all households were still obtaining income from informal business activity in 2012 suggests the continued importance of informal income sources for many of the urban poor. The precarity of most informal sector incomes corresponds with the reduction of financial stability at the household scale suggested by the shift from full time to part time employment.
Despite very high rates of formal sector unemployment, the most significant income source in both 2008 and 2012 was wage work. However, the average monetary value of this income source appears to have increased nearly six fold from $77per month in 2008 to $440 per month in 2012. This change suggests that there might have been an improvement in household purchasing power, although the mean value of income from the second most common income source, informal business, declined from $155 per month in 2008 to $105 per month in 2012. The declining significance of incomes from informal business is linked to the reduction of demand in the informal sector as the formal sector improved, to police crack-downs on informal vending and to decreased profit margins and the elimination of the currency black market [61,62].
Turning to the other side of the balance sheet, changes in how much each household spent suggests that life in Harare was far more costly in 2012 than in 2008, even with higher wage incomes. Expenditure in every category was higher in 2012, including common expenditure types such as housing, education, debt service/repayment, and utilities. Expenditures on food and groceries increased but not as much as the cost of other basic necessities, rising from $57 per month to $91 per month. The sharp increase in expenditures on education suggests that young people in low-income households may increasingly be marginalized from more remunerative opportunities in the urban job market, thus reinforcing the intergenerational cycle of poverty.
Corresponding with the stabilization of the formal economy, there was a clear overall improvement between 2008 and 2012 in terms of access to a cash income: the proportion of households who always or many times went without a cash income in the previous year declined from 59% to 31% and the proportion that never went without improved from 11% to 20%. Going without cash incomes, even occasionally, means that households rely more on non-monetized livelihood activities, such as bartering, working for payment in kind, household production of basic needs (including urban agriculture), and drawing on social capital. Although the situation had clearly improved by 2012, almost half (49%) of all households had still gone without a cash income at points in the previous year, highlighting the persistent problem of precarious income sources for Harare’s urban poor.
Macro-economic growth, political stability and the return of food to the formal retail sector clearly all had a generally positive impact in these low-income areas of Harare. Food was not only more freely available; it was also more accessible. As a result, overall levels of food insecurity declined and dietary diversity increased. At the same time, it is clear that not all households benefitted equally from the changes and that many remained mired in unemployment, poverty and food insecurity. The following sections of the paper therefore disaggregate household food security status by a number of different variables to determine which types of household benefited from the changes and which did not.
Income, employment and food security
Income levels and income source are consistently strong predictors of food security status, particularly in the urban context where food is widely available but often inaccessible because of the cost. The cross-tabulation of household food insecurity scores with income tercile shows a remarkable consistency in the differences by group in 2008 and 2012 (Table 6). For both indicators, households in the lowest income tercile were consistently more food insecure than middle income households, and highest income households were the most food secure.
Table 6: Mean Food Insecurity Scores by Income Level and Income from Wages.
While this relationship is unsurprising, Table 6 also shows that the source of income influences food security scores: the gap in HFIAS between households with a wage income source and without a wage income source widened from 0.4 in 2008 to 2.8 in 2012. The gap in terms of HDDS (from 0.2 in 2008 to 0.7 in 2012) also widened, as households with a wage income benefited much more from improvements in household food security than households without a wage income. These findings confirm that improvements in household food security between 2008 and 2012 were much greater for a specific group of people, and that these people were living in households with a wage income earner. The economies of households receiving a wage income are more directly connected to the formal economy, such that policies directed at currency stabilization, food price stabilization through food imports, and the development of supermarkets, were more likely to benefit this group.
Changes to household dietary diversity also showed a greater improvement for the least poor households than the second least poor households (difference of 1.5 rather than 0.9). Tracking these changes in terms of the changing proportion of households in each category that were “severely food insecure” according to the HFIAP calculation shows even more starkly that the gains were accrued to households with higher incomes (Figure 6). Low income households had virtually the same prevalence of severe food insecurity in 2008 and 2012, whereas proportion of severely food insecure households in the high income tercile dropped precipitously from 63% in 2008 to 43% in 2012 (Figure 6).
Figure 6: Prevalence of Severe Food Insecurity by Income Tercile.
Household characteristics and food security
Different types of households tend to be more vulnerable to food insecurity in crises, for example if there is a higher dependency ratio or if the households relies on a woman as the primary income earner. These characteristics are captured in the categories of household size and household type Tables 7 and 8, where they are cross tabulated with the HFIAS, HDDS, and HFIAP scores to present the mean score for households in each category in 2008 and 2012. Household size is categorized into 1-5 members (58% in 2008 and 77% in 2012), 6-10 members (42% in 2008 and 22% in 2012), and greater than 10 members (approximately 1% in both surveys).
Table 7: Mean Food Insecurity Scores by Household Size.
Table 8: Mean Food Insecurity Scores by Household Type.
There was an overall improvement in food security scores amongst all household size groups. However, the magnitude of the improvement did differ: for example, the mean HFIAS for households with 1-5 members did not improve as much as for households with 6-10 members even though in 2012 as in 2008, households with fewer members had better HFIAS scores (Table 7). The improvement in HFIAS and HDDS scores was the same in both categories (1.7). In all food insecurity scores, larger households consistently scored worse than smaller households, most notably improving least in their HFIAS and HDDS scores (Table 7). The disadvantages for larger households lie in their higher dependency ratios in terms of children, the elderly and morbid members who are less able to contribute economically yet require a share of the household’s food.
Figure 7 presents a different perspective on the changing connection between household size and food security status by highlighting the change in the proportion of “severely” food insecure households by number of household members. There is not a clear linear relationship in either year between increasing household size and increasing likelihood of being severely food insecure, yet households with 8 or more members had the largest reduction in prevalence of severely food insecure households, from the highest rate (83%) in 2008 to the lowest rate (56%) in 2012. This shift supports the hypothesis that many larger households in 2008 were formed as a response to the crisis and included temporary, vulnerable members that exacerbated the effects of food insecurity.
Figure 7: Prevalence of Severe Food Insecurity by Household Size.
The link between household type and food security status helps to shed light on these trends. “Female-centred households” are defined as headed by a woman without a partner and include any combination of immediate relatives (including her children, siblings, parents and grandparents); male-centred households are headed by a man without a partner and including any combination of immediate relatives (including his children, siblings, parents and grandparents); nuclear households include a head and a spouse or partner, with or without children, but without other relatives in the household; and extended households have a head, a spouse or partner, immediate relatives and a combination of other members (relatives and nonrelatives).
In both 2008 and 2012 female-centred households were the most food insecure with the highest mean HFIAS score of any household type (16.1 in 2008 and 10.5 in 2012). However, the greatest HFIAS improvement was amongst male and female-centred households (5.3 and 5.1 respectively). Extended households showed the lowest overall HFIAS improvement, which does not entirely fit the narrative suggested in Figure 7. It does indicate that a subset of larger extended households experienced greater improvements in food security, perhaps because they were able to draw on a larger variety of income sources to reduce their vulnerability.
The representation of the results in terms of the percentage of severely food insecure households in each category provides nuance to the narrative that female-centred and larger households experienced a relatively rapid improvement after 2008. Extended households were the least likely type to be severely food insecure in 2008, suggesting that there were many extended households able to weather the crisis well, even as many others-perhaps those that formed as a response to the crisis itself -had extremely low HFIAS scores that lowered the group average (Figure 8). From the perspective presented in Figure 8, nuclear households had the greatest improvement and the gap between female-centred and other types of households actually widened in the interim. This suggests that while many female-centred households improved their HFIAS scores, they still fell within the “severely food insecure” range, and to a greater extent than among other household types. The 2008 survey therefore likely reflected that single women with dependents, many of whom were already the most vulnerable households, were more severely affected than men and married women by the economic crisis.
Figure 8: Prevalence of Severe Food Insecurity by Household Type.
In terms of dietary quality and diversity, female-centred, nuclear and extended households improved at about the same rate, while male-centred households improved the least, falling from the highest mean score in 2008 to the lowest mean score in 2012 among household types (Table 8). In fact, their overall HDDS showed very little improvement suggesting that the males responsible for procuring food for the household tended to stick with the same range of food items they had always eaten despite the availability of a wider range of food types. Extended households had the highest mean HDDS in 2012, which is consistent with the argument that they were able to draw on a wider range of income sources to diversify their diets.
Price increases and food security
While many factors contribute to food access, rapid price increases such as those experienced during the extraordinarily high inflation in 2008 forced consumers to cut back on purchases, reduce food consumption, sacrifice nutritional value for sustenance, and make trade-offs between food and other basic needs [14,15]. Respondents in both 2008 and 2012 were asked whether the household had gone without certain types of food because of the price of food over the previous six months. In 2008, a third of households (32%) experienced daily shortages due to food price increases (Figure 9). More than two thirds experienced going without food on at least a weekly basis and only 4% never went without food. In 2012, the proportion experiencing daily shortages due to food price increases had declined from 32% to only 4%. On the other hand, the proportion of households that had not experienced shortages due to price increases in the previous six months increased even more dramatically from 4% in 2008 to 51% in 2012. The economic stability in 2012 thus clearly enhanced the food security pillar of “stability” relative to 2008.
Figure 9: Frequency of Going without Food Due to Price Increases.
There was a shift in the types of foods that households had gone without. In 2008, the most common foods that people went without were dairy products (84%), eggs (83%), meats, poultry, or offal (79%), and roots or tubers (78%) (Figure 10). These food types tend to be rich in protein, fats, and micronutrients, and omitting them from the diet on a consistent basis would have long-term health consequences, especially for children. Going without dairy products is likely connected with the instability of the electricity system in 2008, which would be critical for storage of dairy products, as well as the reduced capacity for processors to operate. Vegetables, which do not require processing before distribution and can be stored at room temperature, were the least likely food category to be inaccessible because of price increases in 2008 and 2012. Vegetables are also widely grown in the city, such that their availability is not as widely affected by market forces.
Figure 10: Types of Foods Not Consumed Due to Price Increases.
Non-market food sources and food security
The role of price increases in preventing access to food highlights the importance of food sources outside the market, such as urban agriculture and food remittances. Cross-tabulating the food insecurity scores with the use of these food sources helps to substantiate this observation. In 2008, households receiving food remittances from rural areas had a lower mean HFIAS (13.6) than households that did not receive these remittances (14.8). In 2012, the opposite was true and households that received remittances had a higher mean HFIAS (10.1) than households that did not (9.7). Food remittances evidently played a significant role in reducing the food insecurity of remittancereceiving households in 2008. In 2012, by contrast, remittancereceiving households probably represented a group of generally vulnerable households that relied on remittances because they did not have other means to access food. There was little change in the relative difference in HDDS between these groups of households in 2008 and 2012 (Table 9).
The opposite trend appeared in cross-tabulations of household food insecurity indicators with households that accessed food from their own production and those that did not. In 2008, households growing some of their own food had a mean HFIAS of 14.8, which was higher than the mean HFIAS among households that never produced their own food (14.2) (Table 9). In 2012, this relationship was inverted and the households that did not produce any of their own food had a higher HFIAS (10.1) than those that produced some of their own food (9.3).
One possible explanation is that many vulnerable households, who did not normally produce their own food and, as a result, lacked the necessary tools, inputs, and knowledge of agriculture, were engaging in subsistence food production during the crisis. The swelling of the proportion of households engaging in UA in 2008 suggests that many vulnerable households were turning to this source as a coping mechanism, temporarily raising the food insecurity score for households engaged in UA relative to other households. The relatively positive score for households engaged in UA in 2012 might suggest that these are households that normally engage in UA, even beyond the acute crisis in 2008, and have the necessary skills and resources for viable urban agriculture enterprises. Moreover, for these households that normally produce food, UA appears to have a positive impact on their food security status. As with rural remittances, there were negligible differences in the HDDS score between these categories of households.

Conclusion

The status of household food security in low-income neighbourhoods in Harare was improved in 2012 relative to 2008, and yet persistently high rates of severe food insecurity demonstrate that the daily need to access adequate food continued to be a major challenge for most households. The stabilization of the formal economy by 2012 shaped household food access in some key ways: more households received income from wage work and supermarkets and small retail outlets were much more important food sources than they were in 2008 (although alternative food sources remained important). Most households continued to rely on a diverse set of livelihood strategies and food security strategies even under these improved economic conditions, drawing on non-monetary informal food sources such as rural remittances and urban agriculture in consistently high numbers. Food price increases were less of a problem in 2012 than in 2008, but they continued to impede many households from accessing food on a regular basis.
The findings reported here suggest that improvements in food security status have accrued mostly to the least poor households, and that vulnerability continues to exist for households experiencing a range of deprivations. This suggests a combination of two scenarios: households becoming less poor while also becoming less food insecure, and households that were not poor in 2008 that had higher food security scores in 2012. Because the survey did not track the same households, these trends represent a general widening of the food security gap in keeping with a widening poverty gap in lowincome urban communities. These trends show that food security status in Harare is inextricably linked to other dimensions of poverty, and that even within low-income neighbourhoods there is a wide differentiation in poverty rates and food security status among households.
The key lesson for policymakers is that even in the context of overall economic improvement, food insecurity remains endemic among the poorest segments of the urban population. Households are already accustomed to drawing on resources outside of the formal economy and improvements in employment income have not reversed that trend. These alternative livelihood strategies should therefore be considered as a normal part of urban life and supported with state resources that can improve equal access to food for the most marginalized groups and environmental sustainability of activities such as urban agriculture.
Even during the ostensibly stable period following the crisis, political tensions continued to reduce the state’s effectiveness in improving the everyday lives of most citizens, particularly urban residents who were seen as supporters of deeper political reform [24,29,63]. The 2013 election gave the ZANU-PF full control over government but subsequent policies have led to staple food supply problems, price increases, and vulnerability to changes in weather [25,54]. At the same time, formal and informal sector retailers have proliferated in Zimbabwe’s cities and continue to re-shape urban food networks and consumption patterns and sustain high levels of urban poverty.

References