HOUSEHOLD FOOD SECURITY IN UGANDA:AN EMPIRICAL ANALYSIS

Godfrey B. A. BahiigwaEconomic Policy Research Center

P.O. Box 7841Kampala, Uganda

Tel. 256-41-540141/541024Fax. 256-41-541022

Email: bahiigwa@eprc.or.ugor

eprc1@imul.com

December 1999

_________________________________

Acknowledgements

This research was made possible by funds from the Economic Policy Research Centre (EPRC)and the African Capacity Building Foundation (ACBF). I am grateful to my colleagues: RosettiNabbumba, Enoth Tumukwasibwe and Ben Paul Mugyereza who participated in the fieldwork. Ialso wish to thank Drs. John Okidi and Peter Mijumbi, for their technical comments. Finally,credit goes to the administrative team at EPRC for their logistical support that made thispublication possible.

HOUSEHOLD FOOD SECURITY IN UGANDA:AN EMPIRICAL ANALYSIS

Abstract

This study presents findings of a study on household food security in Uganda. Thestudy finds that during the first period (July-December 1997), 48 percent ofhouseholds in Uganda were food secure and 52% food insecure. During the secondperiod (January-June 1998), 59 percent of households in Uganda were food secure,while 41 percent were food insecure. The study established that food security variesfrom one season to the next, depending mainly on the weather pattern. Householdfood security also varies across regions, agro-ecological zones and districts. Thewestern region was found to be food secure in both periods. The central region wasfood insecure in the first season, but food secure in the second season. The situation inthe eastern region was about the same, roughly half the population were food insecurein both seasons. The northern region had 51 percent of the population food insecure inthe first season, but that figure rose to 74% during January-June 1998. Overall, thenorthern region was the most food insecure. In general, the three main causes ofhousehold food insecurity in both periods were inadequate rainfall, pests and diseases,and excessive rain, in that order. The incidence of pests and diseases is likely to beinfluenced by the weather fluctuations. The study findings suggest that in selectingpriority intervention areas, the Plan for Modernization of Agriculture should consideragricultural extension, rural credit and access to improved technologies as the threemost important areas.

1

Executive Summary

This study presents findings of a study on household food security in Uganda. The studyhad two main objectives. The first objective was to determine the food security status ofrural households during the period July 1997-June 1998. The second objective was toidentify the main determinants of household food security. The study was conducted in14 districts selected from the four geographical regions, covering six of the seven agro-ecological zones (farming systems) in the country. The survey was conducted duringMarch-June 1998, but households were asked about their food security status during twoagricultural seasons: July-December 1997 and January-June 1998. The reason forcovering two seasons was to assess if there are seasonal variations in household foodsecurity.

The study finds that during the first period (July-December 1997), 48 percent ofhouseholds in Uganda were food secure and 52% food insecure. During the secondperiod (January-June 1998), 59 percent of households in Uganda were food secure, while41 percent did not have enough food to feed themselves. The study established that foodsecurity varies from one season to the next, depending mainly on the weather pattern.However, despite the variation, the study finds that at any one point, at least 40 percentof households in Uganda do not have enough food to feed themselves. Indeed weatherfluctuations that result in either little or too much rain were identified as the major causeof food insecurity. In fact, during July-December 1997 the rains were not sufficient for agood harvest. However, due to El Nino, the rains were much better during January-June1998, although excessive in some instance, but that explains the improved food securitysituation during that period.

Household food security also varies across regions, agro-ecological zones and districts.The western region was found to be food secure in both periods. The central region wasfood insecure in the first season, but food secure in the second season. The situation inthe eastern region was about the same, roughly half the population were food insecure inboth seasons. The northern region had 51 percent of the population food insecure in thefirst season, but that figure rose to 74% during January-June 1998. Overall, the northernregion was the most food insecure. District level analysis revealed four categories:districts that were food secure during both seasons; districts that were food secure in thefirst period, but food insecure in the second period; districts that were food insecure inthe first period, but food secure in the second period; and, districts that were foodinsecure during both periods.

In general, the three main causes of household food insecurity in both periods wereinadequate rainfall, pests and diseases, and excessive rain, in that order. The incidence ofpests and diseases is likely top be influenced by the weather fluctuations. While 95percent of households depend on own production as the main source of food, in timesof food shortages, survey results reveal that the market is the main source of food.Dependency on the market is by far the main coping strategy for food insecurehouseholds.

Households indicated that to ensure household food security, they would expect threemain roles of the local government: providing information through extension andseminars, improving access to credit, and supply of improved seed. These results almosttally with those household suggestions for ensuring national food security. Farmer

2

education (extension/information), availability of improved seed, and access to creditwere given as the most important interventions, in that order.

These results are expected to help policy makers in selecting interventions that will leadto improved household food security. For example, the study findings suggest that inselecting priority intervention areas, the Plan for Modernization of Agriculture mightconsider agricultural extension, rural credit and access to improved technologies as thethree most important areas.

3

1. Introduction

Many policy makers in Uganda are concerned about food insecurity that affects thecountry from time to time. In fact, there are communities, such as those in Kotidodistrict, that are chronically food insecure. Several factors have been cited as possiblereasons for this vulnerability to food insecurity. They include, unreliable rainfall patterns;declining soil fertility; pests and diseases; lack of access to land by some potentialproducers; low commodity prices; reliance on traditional methods of production such asuse of unimproved seeds and animal breeds and use of the hand hoe; and, poorextension services. While many of these factors may affect the level of food production,they are not equally important, and no study has been done to determine which of theabove factors are more critical than the others. Yet, policy formulation and interventionmust be informed by empirical findings that help to prioritize among the manypossibilities.

While aggregate data are generally available at the national level, little work has been doneto understand the food security problem at the household level. Having national foodbalance data is not sufficient to understand the food security dynamics in the country,especially in the rural areas. To my knowledge, previous researchers in Uganda have notaddressed issues of food security definition and categorization. Several questions remainunanswered, for instance, what is a food secure or food insecure household and whatcriterion is used for this categorization? What factors determine a household’s foodsecurity status - are they the general ones given above or are there household specificcharacteristics that are equally or more important? What do households consider to bethe most critical factors that determine their food security status? These are only a few ofthe many questions that need answers and this study was designed to provide qualitativeand quantitative data that highlight the status of household food security in Uganda.Since food security is one of the main elements of the mission and strategic purpose forthe Plan for Modernization of Agriculture in Uganda, the findings in this study areexpected to provide useful information in selecting priority areas for intervention.

Most agricultural production comes from millions of rural households. This makes itinteresting for the study to focus on the food security status and its determinants forthese production units. Understanding the rural household, which is the source of foodsurpluses for the urban areas, has important implication for national food security as awhole. A threat to household food security ultimately threatens national food security.Food security at the household level is a prerequisite for national food security.Therefore this study was undertaken to gain insights into household food security and itsdeterminants, and obtain policy relevant information for possible intervention and policyformulation.

The main objective of this study was to assess the food security status of ruralhouseholds in Uganda, with a view to identifying the key determinants of food securityor food insecurity. Specific objectives of the study were to:• Determine the proportion of the population that was food secure at national,

regional, agro-ecological zone and district levels• Determine, from the household's point of view, the key factors that influence its

food security status• Asses the relevance of current government agricultural policies for household and

national food security

4

• Suggest policy strategies, based on research findings that would lead to sustainablefood security at both household and national levels.

Results of the study are expected to provide information that will guide economicplanners and policy makers in their effort to develop interventions programs andformulate policies that will ultimately lead to household and national food security. Morespecifically, the study provides rural household views on what needs to be focused on toensure both household and national food security, at least in the short to medium-term,given the low incomes of the rural population. The study also provides information onwhat rural households think should be the role of local government in ensuringhousehold food security.

Farming systems, cropping patterns, climatic and resource constraints within Uganda areso diverse that few food policies designed at the national level could be expected to havesimilar effects on rural household food security. This study provides regional and districtspecific information that can be used to develop interventions and formulate policies thatare tailored to the specific needs of those areas. Since planning is largely decentralized,the study findings are relevant beyond the central government.

2. Food Security

World food security was the main focus of the 1996 World Food Summit (WFS)organized by the Food and Agriculture Organization (FAO) of the United Nations. TheWFS reports indicate that after 30 years of rapid growth in agricultural production, theworld can produce enough food to provide every person with more than 2700 caloriesper day. This calorie intake is normally sufficient to ensure that all have access toadequate food, provided food distribution is not too unequal. Yet more than 800 millionpeople in the developing world, about 20 percent of their total population, suffer fromchronic under-nutrition. Millions more are vulnerable to food insecurity, they cannotguarantee access to sufficient food at all times.

Despite gains in food production and food security on a world scale, many countries andwhole regions failed to make progress in recent decades. Sub-Saharan Africa producesless food per person today than it did 30 years ago. The WFS reports further indicate thatworld agricultural growth to the year 2010 is expected to slow, but should still outpacepopulation growth. But not all regions and countries will share equally in these gains inproduction and nutrition. The situation in Africa south of the Sahara will deterioratefurther, while progress in South Asia will be painfully slow.

The United Nations, technically, defines a household as food secure when it has accessto the food needed for a healthy life for all its members (adequate in terms of quality,quantity, safety, and culturally acceptable), and when it is not at undue risk of losing suchaccess (Von Braun et al, 1996). This definition is generally acceptable, because it has thethree critical aspects (availability, access and risk). Access refers to the ability to obtainthe necessary food, either through own production or purchasing from the market. Riskarises from fluctuations in production or income. In Uganda’s context, especially in ruralareas where the majority of households depend on own production, the risk to foodsecurity arises more from fluctuations in production than from income because very littleof what they consume is purchased. However, for households that are net food buyers,the risk to food insecurity is due to fluctuations in their income.

5

In general terms, Uganda is regarded as self-sufficient in food production. In fact,achieving food self-sufficiency in food production has long been one of the majorobjectives of the agricultural sector. Yet agricultural and population statistics availableindicate that per capita food production in 1997 was 44 percent less than what it was in1970. As an indicator of food availability, Uganda had more food available per person in1970 than in 1997. That means the country's food crop production has not been keepingpace with increases in population.

In 1970 when Uganda's population was bout 9.78 million, total food crop output was14.1 million metric tons, while in 1997 with a population of about 20.4 million people,total food output was 16.5 million metric tons. In the 27-year period, population grew byabout 109 percent, while total food production grew by about 17 percent only (Figure 1and Appendix 1). From these statistics, it is clear that Uganda’s self-sufficiency in foodproduction is threatened. In the 1970s when the country’s food security status was better,soils were more productive. Over the years, soil productivity has declined, as evidencedby declining yield trends for various crops. Given the low or non-use of mineralfertilizers to augment natural soil fertility, yields will continue to decline.

If agricultural productivity is increased, Uganda has great potential to produce sufficientfood not only to meet domestic demand, but also have surplus to export to regional andinternational food markets. Statistics from the 1991 Population Census and Land Useindicate only 30 percent of cultivable land was under use. Harnessing this potentialrequires focussing on the country's production base: the rural agricultural sector, which isdominated by millions of subsistence farmers.

Figure 1: Indices of Food Crop Production and Population in Uganda (1970=100)

Food insecurity in Uganda has several dimensions. The first is low productivity of theagricultural sector, which means low production and hence reduced food availability. Thesecond dimension is poverty, with 44 percent (8.8 million) of the population living inabsolute poverty, according to 1997 household survey data (GoU, 1999). The majority of

0

50

100

150

200

250

Year

Inde

x

1970 1973 1976 1979 1982 1985 1988 1991 1994 1997

Production Per Capita Population

6

the poor households are rural based, with low incomes and, therefore, when own foodproduction fails, they are fully exposed to food insecurity. Since agriculture is the mainactivity of the rural poor, increasing their productivity will achieve two objectives:reducing rural poverty through increased household income, and achieving sustainablehousehold and national food security. The third dimension is the vulnerability to naturalhazards, such as poor rainfall that leads to low production and, therefore, increased risk.

Uganda's rural agriculture is predominantly subsistence and semi-subsistence farming. Itis predominantly rain-fed and, therefore, food supplies are susceptible to rainfallfluctuations. There is little, if any, stockholding at the household level, a situation thatmakes rural households find it difficult to go through off-season periods and times ofpoor harvest. Poor or lack of appropriate and affordable post-harvest technology athousehold level leads to food losses that have been estimated at about 30 percent. At thenational level, there is no specific buffer stock program that would release food onto themarkets during times of shortage to stabilize retail food prices during periods of lowsupplies. In the past, attempts by the government to purchase agricultural produce failedbecause of inefficiency inherent in the public sector. Such food reserves, whileadvantageous to consumers in maintaining low food prices, are very expensive tomaintain and tend to rely on government subsidies. In Uganda, where public sectormanagement is inefficient, buffer stock programs, if they are to evolve, should bemanaged by the private sector on a purely commercial basis. However, the private sectorhas not yet developed enough to maintain large food commodity stocks.

Estimates in the 1991 census indicate that about 89 percent of Uganda's population livein the rural areas and more than 80 percent of the rural labor force are employed inagriculture. It is estimated that there are about 2.5 - 3 million farm households in Uganda,80 percent of whom cultivate less than 2 hectares, but contribute about 70 percent ofagricultural production. Well over 50 percent of household production is consumeddomestically, although there is increasing commercialization, especially of food crops.These structural characteristics of Uganda's agricultural sector have importantimplications for food security, both at the household and national levels.

Modernization of the agricultural sector and improving the livelihood of the poor aremajor development objectives of the government of Uganda1. Achieving these objectiveswill require implementation of strategies in the PMA, which give special emphasis onsubsistence farmers. The strategies are aimed at increasing household incomes throughincreased shares of marketed production. This should be possible by promotingproductivity-enhancing technologies, along with improved extension service delivery.Agricultural transformation, an integral part of rural growth in Uganda, has to take placein order to increase food supply and incomes of the rural poor, and therefore improvetheir livelihoods.

1 The government has designed a Plan for Modernization of Agriculture (PMA) focusing on povertyeradication through agricultural transformation from being predominantly subsistence to beingcommercially-oriented.

7

3. Data Collection and Analytical Methodology

3.1 Data Collection

Primary data for the study were collected from 14 districts in the four geographicalregions of the country (Western, Central, Eastern and Northern) during March –June1998, but respondents were asked about their food security status during the period July1997 - June 1998. The districts for the survey were selected based on the followingcriteria:• food security status (chronic, transitory, stable) of the district, based on the

categorization in a 1995 study on food security• geographical location by region, to ensure the sample was nationally representative• agro-ecological zone within which the district is located, to capture the different

farming systems in the country

The selected districts by geographical region and agro-ecological zone are given inappendices 2 and 3, respectively. One hundred households were interviewed from eachdistrict, except for Tororo, Busia, Luwero and Nakasongola from which only fiftyhouseholds each were interviewed. The reason for this being that at the time of selection,Busia and Nakasongola were new districts, and district specific data were only availablefor the districts of Tororo and Luwero from which the former districts were created,respectively. Thus, the total number of households interviewed was 1200. Within eachdistrict, two counties were selected with a target of 50 households each. The countieswere selected based on spatial separation from each other, distance from theurban/administrative center and population density. Within each county, two sub-counties were selected, targeting 25 households each. At sub-county level, households welargely selected randomly, except that care was taken to include female headedhouseholds. Secondary data on production, population, etc and other global literature onfood security were obtained from several publications, both national and international.

3.2 Analytical Methodology

The definition of food security used in this study is based on the households'perspectives. The major food concern of a household is to have something to eat, at alltimes, basically having enough (in terms of quantity and availability) to feed thehousehold. The issue of quality (nutritional content) is secondary. For example, if ahousehold has a plot of cassava (a starchy food) that forms the main component of thediet, and can sustain the household from one season to the next, the household considersitself to be food secure. Yet, from a nutritional point of view, the household could befood insecure because such a diet would be deficient in proteins, oils and vitamins. It isthis very basic definition that this study adopts, and household data collection was basedon the concept of “enough” as perceived by the household. Therefore, a household isfood secure in a given season if it has enough food to provide its members all the usualmeals in a day, for the entire season. Otherwise, the household is food insecure.

8

4. Research Findings

The research findings are presented in two sections. The first section gives qualitativefindings supported by simple statistical analysis. The second section gives the empiricalresults and discussion of the logistic regressions.

4.1. Food Security Status at National Level

In order to understand the food security situation in Uganda, one should not just lookthe general food status of the country, but also at the specificity that is brought about bydifferences in geographical location and agro-climatic conditions. And for policy, it is thespecificity of various parts of the country that is of interest in order to design policies andprograms that are tailored to a specific region or district. In general, the food securitystatus of rural households changes from one season to another. During the period July-December 1997, 52% of households were food secure, while 48% were food insecure. Inthe following season, (January-June 1998), there was an improvement in the food securitysituation, with 59% of households reporting that they were food secure, compared to41% that were food insecure. Despite the improvement, these survey results indicate thatduring both seasons, spanning a period of one year, at least 40% of households inUganda were food insecure. Table 1 gives a summary of the national food security statusduring the period of July 1997-June 1998.

Table 1: Household Food Security Status for Uganda, July 1997-June 1998

Status/Season July-December 1997Count Percentage

January-June 1998Count Percentage

Food Secure 596 48.3 688 59Food Insecure 557 51.7 479 41Total 1153 100 1167 100Source: Survey data.

4.2. Food Security Status by Region

On a regional basis, the Western region was food secure in both seasons, with morehouseholds having enough food to feed themselves in the second season. The Centralregion had 40% of its households food secure in the first season while in the secondseason the proportion of the food secure people increased to 77%. In the Eastern regionthe proportion of the food secure and food insecure households was about the same inthe first season, with a slight worsening of the situation in the second season. About halfthe households in the Northern region were food insecure in the first season, but thesituation got worse in the second season, with 74% of the households reporting that theydid not have enough to feed themselves. Table 2 summaries the regional findings.

9

4.3. Food Security Status by Agro-ecological Zone

Table 3 gives the food security status in Uganda by agro-ecological zones (FarmingSystems). Agro-ecological zones are demarcated by the type of soils, rainfall patterns andtype of farming characteristics, that is, the major crops/animals that can be supported bythe agro-climatic factors in that part of the country. Appendix 3 gives the differentagro-ecological zones and the districts that were surveyed. Except for the Montanesystem and the Pastoral System which were food secure and food insecure in bothseasons, respectively, the other zones had fluctuating food security situations, withvarying degrees from one season to the next.

10

Table 2: Household Food Security Status by Region, July 1997 – June 1998

Region/SeasonJuly – December 1997FS FI Total

January – June 1998FS FI Total

Western Count% within region% within Uganda

19265.532.2

10134.518.1

29310025.4

22776.233.0

7123.814.8

29810025.5

Central Count% within region% within Uganda

11740.919.6

16959.130.3

28610024

22376.932.4

6723.114.0

29010024.9

Eastern Count% within region% within Uganda

19550.632.7

19049.434.1

38510033.4

18948.927.5

20051.441.8

38910033.3

Northern Count% within region% within Uganda

9248.715.4

9751.317.4

18910016.4

4925.87.1

14174.229.4

19010016.3

TotalUganda

Count% within Uganda

59651.7

55748.3

1153100

68859

47941

1167100

Source: Survey Data, July 1997-June 1998. FS = Food Secure; FI = Food Insecure

Table 3: Household Food Security Status by Agro-ecological Zone, July 1997 – June 1998Agro-Ecological Zone (AEZ)/Season July – December 1997

FS FI TotalJanuary – June 1998FS FI Total

Teso system Count% within AEZ% within Uganda

12061.920.1

7438.113.3

19410016.8

9649.214.0

9950.820.7

19510016.7

Plantain/Robusta system Count% within AEZ% within Uganda

14451.424/2

13648.624.4

28010024.3

22980.133.3

5719.911.9

28610024.5

Plantain/Millet/Cotton Count% within AEZ% within Uganda

7530.712.6

16969.330.3

24410021.2

12450.418.0

12249.611.9

28610024.5

Northern system Count% within AEZ% within Uganda

7071.411.7

2828.65.0

981008.5

4546.46.5

5253.610.9

971008.3

Montane system Count% within AEZ% within Uganda

16567.127.7

8132.914.5

24610021.3

19076.027.6

6024.012.5

25010021.4

Pastoral system Count% within AEZ% within Uganda

22224.23.7

6975.812.4

2911007.9

44.3.6

8995.718.6

931008.0

Total Count% within AEZ

59651.7

55748.3

1153100

68859.0

47941.0

1167100

Source: Survey Data, July 1997-June 1998. FS = Food Secure; FI = Food Insecure

11

Table 4: House Food Security Status by District July 1997 – June 1998

District/Season July – December 1997FS FI Total

January – June 1998FS FI Total

Mbarara Count%within District% within Uganda

6767.711.2

3232.35.7

991008.6

8989.012.9

1111.02.3

1001008.6

Kumi Count%within District% within Uganda

5657.79.4

4142.37.4

971004.2

4848.57.0

5151.510.6

991008.5

Busia Count%within District% within Uganda

1939.63.2

2960.45.2

481004.2

1938.02.8

3162.06.5

501004.3

Tororo Count%within District% within Uganda

1530.62.5

3469.46.1

491004.2

612.2.9

4387.89.0

491004.2

Luwero Count%within District% within Uganda

2657.84.4

1942.23.4

451003.9

4185.46.0

714.61.5

481004.1

Apac Count%within District% within Uganda

7071.411.7

2828.65.0

981008.5

4546.46.5

5253.610.9

971008.3

Bushenyi Count%within District% within Uganda

7072.911.7

2627.14.7

961008.3

7374.510.6

2525.55.2

981008.4

Kotido Count%within District% within Uganda

2224.23.7

6975.812.4

911007.9

44.30.6

8995.718.6

931008.0

Mpigi Count%within District% within Uganda

5053.28.4

4446.87.9

941008.2

8387.412.1

1212.62.5

951008.1

Kiboga Count%within District% within Uganda

3132.35.2

6567.711.7

961008.3

6668.09.6

3132.06.5

971008.3

Nakasongola Count%within District% within Uganda

1019.61.7

4180.47.4

511004.4

3366.04.8

1734.03.5

501004.3

Kabarole Count%within District% within Uganda

5556.19.2

4343.97.7

981008.5

6565.09.4

3535.07.3

1001008.6

Soroti Count%within District% within Uganda

6466.010.7

3334.05.9

971008.2

4850.07.0

4850.010.0

961008.2

Iganga Count%within District% within Uganda

4143.66.9

5356.49.5

941008.2

6871.69.9

2728.45.6

951008.1

Total Count%within District

59651.7

55748.3

1153100

68859.0

47941.0

1167100

Source: Survey Data, July 1997-June 1998. FS = Food Secure; FI = Food Insecure

12

4.4 Food Security Status by District

Table 4 gives the food security status by district. Four categories emerged from thefindings. The first category consists of districts that were food secure during bothperiods. The second category consists of districts that were food secure during the firstperiod, but food insecure during the second period. The third category consists ofdistricts that were food insecure during the first period, but food secure during thesecond period. The fourth category consists of districts that were food insecure duringboth periods. Table 5 gives a summary of the four categories. While no clear patternemerges, some observation can be made. First, all districts that were food secure duringboth periods were in the Western and Central regions. Second, none of the districts inthe Eastern and Northern regions were food secure in both seasons. Third, all districtsthat were food insecure during both seasons were in the Eastern and Northern regions.Fourth, Montane and Plantain-Robusta agro-ecological zones tend to be more foodsecure than the other agro-ecological zones. Fifth, the pastoral system alone seems to bethe most vulnerable of all agro-ecological zones.

Table 5: Summary of Food Security Status by District, Region and Agro- Ecological zone(s)

Status, Period District Region Agro-ecological zone(s)Food secure, both periods Mbarara

BushenyiKabaroleLuweroMpigi

WesternWesternWesternCentralCentral

Pastoral /Montane systemsMontane systemMontane/Plantain-RobustaPlantain-Millet-CottonPlantain-Robusta

Food secure, first period only KumiSorotiApac

EasternEasternNorthern

Teso systemTeso systemNorthern system

Food secure, second period only KibogaNakasongolaIganga

CentralCentralEastern

Plantain-Millet-CottonPlantain-Millet-CottonPlantain-Millet-Cotton

Food insecure, both periods BusiaTororoKotido

EasternEasternNorthern

Plantain-Millet-CottonPlantain-Millet-CottonPastoral System

Source: Survey Data, July 1997-June 1998

4.5. Source of Household Food

The majority of households surveyed ranked own production as the main source ofhousehold food consumption. The market was ranked second to own production as asource of food for households. The third was inter-household transfers, and other lesssignificant sources included community reserves, relatives and exchange of labor forfood. These findings are summarized in Table 6 below.

13

4.6 Causes of Food Insecurity

Households gave weather-related problems (little or too much rain) as the mostimportant reason for food insecurity during both seasons. The second most importantcause of food insecurity in both periods was pests and diseases. Other significantcontributors to household food insecurity were inadequate labor, inadequate land, notgrowing enough food during the season and soil infertility, with varying importance inboth seasons. Table 7 gives a summary of the causes of food insecurity in Uganda.

Table 6: Sources of Household Food, National LevelRank Source Count Percentage123456

Own productionMarket purchaseInter-household transfersCommunity reservesRelativesExchange labor for food

112894522838187

95.479.919.3 3.2 1.5 0.6

Source: Survey data. Number of households = 1183.

Table 7: Causes of Rural Household Food Insecurity, National LevelReason July-Dec. 1997 Jan.-June 1998

Count Percent Count PercentLittle rain during seasonPests and diseasesToo much rainLand not enoughDid not grow enough foodLabor not enoughLand infertilePoor healthLack of planting materialsLack of oxen for ploughingUse of hand hoe for cultivationOthers

34517211296866656661

29.214.59.58.17.30.55.50.50.50.50.1

2591748410379107-1098-2

21.914.77.18.76.79.0-0.80.80.7-0.2

Source: Survey data. Number of households =1183

4.7. Coping Strategies

Although 95% of rural households derive their food consumption from own production,during times of food scarcity most households depend on the market as the major sourceof food. Other coping strategies include donations from relatives and neighbors,reducing the number of meals or ration, sale of livestock and exchange of labor for food.However, survey data indicate that in both periods, the market is by far the largest sourceof household food in times of food shortages. Table 8 gives summary of copingstrategies during both periods covered by the study.

14

Table 8: Short-term Coping Strategies to Food Insecurity, National LevelCoping Strategy July-Dec 1997 Jan-June 1998

Count Percent Count PercentBuy food from marketDonation from relatives/neighborsReduce number of meals/rationSale of livestockExchange labor for foodOther

53130492129

44.92.54.11.80.20.8

4293142373

36.32.63.60.30.60.3

Source: Survey data. Number of households =1183.

To ensure long term availability of food, households had two main strategies: growingmore food and expanding on acreage. Other strategies included buying more land,rearing of improved livestock breeds, using pesticides, growing diseases/pest resistantcrop varieties, hiring out labor, and depending on the market. Table 9 gives a summaryof the findings.

Table 9: Household Long-term Strategies for Ensuring Food Security, National Level

Long term strategy Count PercentGrow more foodExpand acreageBuy more landUse pesticidesGrow disease/pest resistant cropsRear improved livestock breedsRely on the marketHire out laborOthers

6581416716934820

55.611.95.71.40.80.30.30.71.7

Source: Survey data. Number of households = 1183.

It is important to note that long-term household strategies for ensuring food security areproduction-based and not income-based, at least not directly. This is perhaps becausetheir incomes are derived from production.

4.8. Role of Local Government in Food Security

Households were asked what they thought should be the role of local government inensuring food security. Three major roles were identified namely: providing informationthrough seminars and extension services, provision of credit and supply of improvedseed. Other roles expected of local governments are summarized in Table 10 below.

15

Table 10: Households expectations of the role of local governmentRole Count Percent of

responsesProviding information (seminar/extension)Provide loans/creditSupply improved seedConstruct marketsSupply pesticidesProvide tractor for hire purchaseImpose by-laws on food productionOrganize people in groupsProvide infrastructureAvail landOthers

856965322420161612857

211716865443214

Source: Survey data. Total responses = 405.

4.9. Ensuring National Food Security

One of the key questions in the research was about households’ views of what needs tobe done to ensure that Uganda is a food secure country. For the key strategies, the resultsalmost tally with those for household strategies for ensuring food security. Farmereducation, availability of improved crop varieties and provision of credit were given asthe three major strategies for ensuring food security in Uganda. Other strategiessuggested are given in Table 11 below.

Table 11: Strategies to Ensure that Uganda is a Food Secure Country, National Level

Strategy Count Percent ofresponses

Educate farmers (extension information)Avail improved crop varietiesProvide loans/creditMechanize productionGovernment provide storage facilitiesControl pests and diseasesGrow more food cropsResettle people to places with productive landProvide fertilizersOthers

48530229921720618111910383586

18.811.711.68.48.07.04.64.03.222.7

Source: Survey data. Total responses = 2581.

The most critical factors for ensuring both household and national food security can besummarized as extension, improved technology, access to credit and markets.

4.10 . Gender and Food Security

In the first and second periods, 67.4% and 68.6% of respondents were household headsrespectively. In the first period, 11.8% of the household heads were female and in the

16

second period female household heads comprised 9.9%. Within gender comparisons,female-headed households were more food secure than male-headed households, duringboth periods. In the first period, 58% of female-headed households were food secure,compared to 52% of male-headed households. In the second period, only 28% offemale-headed households were food insecure, compared to 41% of male-headedhouseholds. These findings are summarized in Table 12 below.

Table 12: Food Security Status by Gender of Household Head, National LevelPeriod Status Gender of household head

Male Female TotalFS Count

% within status% within gender

37288.252.2

5011.857.5

42210052.9

July-Dec 1997

FI Count% within status% within gender

33890.147.6

379.942.5

37210047.1

FS Count% within status% within gender

42886.859.4

6513.272.2

49310060.8

Jan-June 1998

FI Count% within status% within gender

29392.140.6

257.927.8

31810039.2

Source: Survey Data. FS = Food secure, FI = Food insecure.

4.11. Education and Food Security

Most household heads in the survey had primary education, followed by ordinary leveleducation. There does not seem to be a specific pattern that indicates that the higher thelevel of education of the household head, the more food secure a household will be. It isironical that households headed by people with A-level of education were morevulnerable to food insecurity than those with primary and O-level education during bothperiods. However, households headed by people with no formal education at all werefood insecure in both periods. This may possibly mean that while some level ofeducation is important to household food security, its marginal contribution beyondprimary education is very small. Details are summarized in Table 13.

Table 13: Household food security by level of education of household head July-December 1997 January-June 1998

Level of Education FS F1 Total FS F1 TotalNone Count % within level

5844.6

7255.4

130100

5440.9

7859.1

132100

Primary Count % within level

19152.2

17547.8

366100

23261.9

14338.1

375100

O-Level Count % within level

9851.6

9248.4

190100

12465.3

6634.7

190100

A-Level Count % within level

1139.3

1760.7

28100

1653.3

1446.7

30100

Tertiary Count % within level

5673.7

2026.3

76100

5672.7

2127.3

77100

Source: Survey data.

17

5. Conclusions and Policy Implications

This study had two main objectives. The first objective was to determine the foodsecurity status of rural households during the period July 1997-June 1998. The secondobjective was to identify the main determinants of household food security. The studywas conducted in 14 districts selected from the four geographical regions, covering six ofthe seven agro-ecological zones (farming systems) in the country. The survey wasconducted during March-June 1998, but households were asked about their food securitystatus during two agricultural seasons: July-December 1997 and January-June 1998. Thereason for covering two seasons was to assess if there are seasonal variations inhousehold food security.

The study finds that during the first period (July-December 1997), 48 percent and 52percent of households in Uganda were food secure and food insecure, respectively.During the second period (January-June 1998), 59 percent of households in Uganda werefood secure, while 41 percent did not have enough food to feed themselves. The studyestablished that food security varies from one season to the next, depending mainly onthe weather pattern. However, despite the variation, the study finds that at any one point,at least 40 percent of households in Uganda do not have enough food to feedthemselves. Indeed weather fluctuations that result in either little or too much rain, wereidentified as the major cause of food insecurity. In fact, during July-December 1997 therains were not sufficient for a good harvest. However, due to El Nino, the rains weremuch better during January-June 1998, although excessive in some instance, but thatexplains the improved food security situation during that period.

Household food security also varies across regions, agro-ecological zones and districts.The western region was found to be food secure in both periods. The central region wasfood insecure in the first season, but food secure in the second season. The situation inthe eastern region was about the same, roughly half the population were food insecure inboth seasons. The northern region had 51 percent of the population food insecure in thefirst season, but that figure rose to 74% during January-June 1998. Overall, the northernregion was the most food insecure. District level analysis revealed four categories:districts that were food secure during both seasons; districts that were food secure in thefirst period, but food insecure in the second period; districts that were food insecure inthe first period, but food secure in the second period; and, districts that were foodinsecure during both periods.

In general, the three main causes of household food insecurity in both periods wereinadequate rainfall, pests and diseases, and excessive rain, in that order. The incidence ofpests and diseases is likely to be influenced by the weather fluctuations. While 95 percentof households depend on own production as the main source of food, in times of foodshortages, survey results reveal that the market is the main source of food. Dependencyon the market is by far the main coping strategy for food insecure households.

Households indicated that to ensure household food security, they would expect threemain roles of the local government: providing information through extension andseminars, improving access to credit, and supply of improved seed. These results almosttally with those household suggestions for ensuring national food security. Farmereducation (extension/information), availability of improved seed, and access to creditwere given as the most important interventions, in that order.

18

These results are expected to help policy makers in selecting interventions that will leadto improved household food security. For example, the study findings suggest that inselecting priority intervention areas, the Plan for Modernization of Agriculture mightconsider agricultural extension, rural credit and access to improved technologies as thethree most important areas.

19

References

1. Van Braun, Joachim, Howarth Bouis, Shubh Kumar and Rajul Pandya-Lorch.1996. Improving Food Security of the Poor: Concept, Policy and Programs. InternationalFood Policy Research Institute, Washington, D.C.

2. Opio Fred. 1997. “ Structural Adjustment, Growth and Poverty in Uganda” inStructural Adjustment and Poverty in Uganda. Economic Policy Research Bulletin,Vol.3. No.1 (April 1997). EPRC, Kampala, Uganda.

3. Government of Uganda (GOU). 1998. Poverty Trends in Uganda 1992-1996.Ministry of Finance, Planning and Economic Development. Discussion PaperNo.2 Kampala, Uganda.

4. Food and Agriculture Organization of the United Nation (FAO) 1996. TechnicalBackground Documents. World Food Summit 1996 Rome.

20

Appendix 1: Food Crop Production in Uganda

Year Total Output Per Capita Population Indices (1970=100)(000'MT) (Mt/Capita) (Millions) Total

OutputPer Capita Population

1970 14076 1439 9.78 100 100 1001971 13295 1325 10.03 94 92 1031972 13499 1311 10.30 96 91 1051973 13319 1263 10.55 95 88 1081974 14781 1364 10.84 105 95 1111975 16294 1466 11.11 116 102 1141976 15130 1326 11.41 107 92 1171977 15123 1292 11.71 107 90 1201978 15022 1251 12.01 107 87 1231979 11191 909 12.31 80 63 1261980 10490 830 12.64 75 58 1291981 11985 925 12.96 85 64 1321982 12925 974 13.27 92 68 1361983 13717 1008 13.61 97 70 1391984 12446 892 13.95 88 62 1431985 12643 884 14.30 90 61 1461986 12995 887 14.65 92 62 1501987 13756 916 15.02 98 64 1541988 14482 941 15.39 103 65 1571989 15271 968 15.78 108 67 1611990 15514 959 16.18 110 67 1651991 15676 940 16.68 111 65 1701992 15399 879 17.52 109 61 1791993 16354 903 18.11 116 63 1851994 15795 845 18.69 112 59 1911995 16994 867 19.60 121 60 2001996 15484 780 19.85 110 54 2031997 16494 807 20.44 117 56 209Source: Uganda Agriculture, World Bank (1993); Statistical Abstract, 1998 (MFPED)

21

Appendix 2:Survey Areas for the Food Security Study in Uganda

Region District Coutry Sub-countiesNorthern Apac Maruzi

Kole

AkokoroCegereAkaloAboke

Kotido LabworDodoth

AbimKaabongKarenga

Eastern Soroti Soroti

Kaberamaido

ArapaiAsuretKaberamaidoOchero

Kumi Bukede

Busia

Bukedea

Samia-Bugwe

TororoKwapaMoloBusitemaDabani

Iganga Bunya

Luuka

BuwayaImmanyiroWaibugaIrongo

Central Luwero Wabusana KikyusaKalagalaKakoogeNakasongolaWabinyonyi

Mpigi Busiro

Mpigi

WakisoNamayumbaMpigiBuwama

Kiboga E.Consitituency

W. Constituency

BukomeroLwamataNtwetweKyankwanzi

Western Kabarole Kyaka

Burahya

KasuleMparaRuteeteHakibaale

Bushenyi Bunyaruguru

Sheema

KichewambaRyeruKagangoKitagata

Mbarara Kashari

Rwampara

KashareRubayaBugambaNyakayojo

22

Appendix 3: Agro-ecological Zones (Farming Systems) in Uganda

Agro-ecological Zone Major Crops/Livestock District Surveyed

Teso System Cattle Cassava, Millet, Cotton Soroti, Kumi

Plantain/Robusta System Bananas, Robusta Coffee Iganga, Mpigi, Kabarole

Plantain/Millet/CottonSystem

Bananas, Millet, Cotton Tororo, Iganga, Luwero,Kiboga, Nakasongola

Northern System Millet. Cotton, MaizeCassava

Apac

Montane System Bananas, Coffee, Cattle BushenyiMbararaKabarole

Pastoral System Cattle Mbarara, Kotido

West Nile System Tobacco, Bananas, Coffee None