You are on page 1of 13

World Devehpmenr, Vol. 23, No. 10, pp.

1669-1681, 1995
Copyright 0 1995 Elsevier Science Ltd
Pergamon
Printed in Great Britain. All rights reserved
0305-750x/95 $9.50 + 0.00
0305-750x(95)ooo73-9

Alternative Food Security Strategy: A Household


Analysis of Urban Agriculture in Kampala

DANIEL G. MAXWELL*
University of Wisconsin-Madison, U.S.A.

Summary. - Farming within African cities has become an increasingly important source of food for
urban populations. Yet little is understood about the forces behind urban farming or its impact at the
household level. lntrahousehold dynamics and gender relations, as well as declining wages and economic
informalization are all important to an understanding of urban farming. Access to land is a major con-
straint, and only a small fraction of urban farmers own their land. For those with access to land, urban
farming is associated with higher levels of household food security and child nutrition. This paper outlines
the linkages between economic strategies. access to land, and food security, and discusses the policy
implications of urban farming.

I. INTRODUCTION urban poor.“’ Review articles in the late 1980s made


similar observations about the potential of urban agri-
Farming within African cities has become an culture (Hussain and Lundven, 1988; Sachs and Silk,
important activity over the past two decades, yet it 1987). Yet recent policy studies have taken a some-
remains largely misunderstood. This paper attempts what more equivocal view, noting equity and envi-
to address the multiple questions that arise about ronmental concerns (Atkinson, 1992) and the lack of
semisubsistence farming in the city of Kampala, conclusive evidence about the impact of urban farm-
Uganda. What are the forces driving the increase in ing. A 1993 study of urban food security by the
urban farming? How is the incorporation of farming International Food Policy Research Institute noted
into urban economic strategies to be understood? In that the information base about urban agriculture is
the heavily built-up areas of the city, how do urban “so limited that policy and program designs are often
residents acquire land for cultivation, particularly based on speculation.“2
low-income residents who do not own land? What is In many cities in Eastern and Southern Africa
the impact of urban agriculture at the household level farming is technically illegal, yet it is pervasive.
in terms of food security, nutritional status, and Urban authorities are often opposed to the practice for
income? What are the implications of urban farming several reasons. First, it is perceived as an artifact
for policy makers? of rural life that simply does not belong in cities.
During the late 198Os, a number of research studies Second, it is perceived as a public health nuisance.
in different disciplines and different countries noted Third, it is perceived as being of marginal importance
the increasing importance of farming within African
cities. While the first research was concerned primar-
ily with urban planning (Sanyal, 1985), other studies * I am extremely grateful for the excellent research assis-
focused on the alleviation of poverty and the impact tance of Gertrude Atukunda of Makerere Institute of Social
of urban farming on food supply (Lee-Smith et al., Research; and for comments on earlier drafts of this paper
1987; Freeman, 199 1; Drakakis-Smith, 1992; from Aili Tripp, John Bruce, Christopher Barren and
Maxwell and Zziwa, 1992), the gender aspects of Laurence Becker of the University of Wisconsin, Joanne
urban farming (Rakodi, 1988,199l; Memon and Lee- Csete of UNICEF, the late Emmanuel Nabuguzi of
Makerere Institute of Social Research, and two anonymous
Smith, 1993), or land use and environmental consid-
reviewers. I am also grateful for the institutional support of
erations (Sawio, 1993; Smit and Nasr, 1992). These Makerere Institute of Social Research during fieldwork.
studies raised the issue of urban farming for policy Funding for this research came from a Fulbright-Hays
makers at both local and international levels. The Doctoral Dissertation Research Abroad Fellowship, and
Brundtland Commission noted that “urban agriculture from the National Science Foundation, USA. All views
could become an important component of urban expressed, and all errors and omissions are mine alone. Final
development and make more food available to the revision accepted: April 26, 1995.

1669
1670 WORLD DEVELOPMENT

to the urban economy (Sanyal, 1985; Freeman, 1991; dence of urban farming during that period. The third
Maxwell and Zziwa, 1992). Previous studies have section presents an overview of the household analy-
attempted to dispel some of these perceptions. They sis of urban farming in Kampala, including a discus-
have noted that the crisis of the formal urban econ- sion of intrahousehold dynamics, access to land, and a
omy and the impacts of structural adjustment are comparison of food security and nutritional status in
forces driving the trend of increased food production farming and non-farming urban households in Kampala.
within urban areas, and most of them suggest that The final section discusses the implications of urban
urban agriculture has become one of the critical farming, and suggests possible policy alternatives.
means of “feeding the urban masses” in Africa
(Sawio, 1993).3Urban agriculture is usually portrayed
as a household survival strategy, one of the means 2. THE URBAN ECONOMY OF KAMPALA
urban households have devised to adapt to an increas-
ingly hostile economic environment (Sanyal, 1985; (a) The decline ofthe formal economy
Freeman, 1991; Mvena, Lupanga and Mlozi, 1991).
Most studies note that the primary motive for urban The formal economy G. Kampala, the capital and
farming is food for household consumption and that largest city of Uganda, was severely damaged by the
urban farmers are mostly women. Yet of studies “war of economic independence” of the Amin regime
carried out to date, only Rakodi (1988, 1991) (1971-79) that was initiated with the expulsion of the
and Memon and Lee-Smith (1992) examine gender Indian minority in Uganda and rapidly gave rise to a
dynamics in detail, and only Lee-Smith er al. (1987) highly informal magendoS economy dominated by
estimate the household impact of urban farming. Most smuggling, illegal currency transactions and state
studies presume a positive impact on household food appropriation of private property (Banugire, 1985). In
security but provide no conclusive evidence, hence the early to mid-1980s, a guerilla war was centered on
the skepticism of some local authorities and policy the outskirts of the city. These combined factors had a
makers toward this potentially important strategy for devastating impact on the urban economy (Bigsten
urban food security. and Kayizzi-Mugerwa, 1992; Jamal, 1985; Jamal and
This paper is based on field research carried out Weeks, 1988, 1993), and contributed to a substan-
in Kampala, Uganda, between November 1992 and tially lower rate of urban population growth (Table 1).
October 1993. The research included a series of 40 Jamal and Weeks (1993) note that wages in Kampala
comparative household case studies, a two-round sur- had been among the highest in East Africa in the
vey of 360 households selected into a multistage ran- 196Os, but fell dramatically in the 1970s. By the
dom sample, and focus groups as follow-up to both 198Os, the urban wage-earning class had shrunk
the case studies and the survey. The case study com- markedly as city residents were forced to diversify
parison spanned the entire year; the two rounds of the their sources of income to compensate for the drastic
survey were carried out during the rainy and dry sea- drop in wage earnings (Bigsten and Kayizzi-
sons.4 The next section of this paper briefly sketches Mugerwa, 1992). This resulted in the massive infor-
changes in the urban economy of Kampala during the malization of the city’s economy, and an increase in
1970s and 1980s and describes the increased inci- urban agriculture.

Table 1. Descriptive information. Kampala

Census years 1969 1980 1991

Population 330,700 458,500 774,250


Growth rate 9.3% 3.8% 4.9%
Mean household size - - 4.0
Employment*
Manufacturing 6.6%
Formal services 21.4%
Informal services 46.3%
Government 17.1%
Agriculture 4.5%
Domestic labor 4.1%

*Expressed as percentage of employed labor force. Complete unemployment is not counted. In the
absence of social safety nets, it is very difficult to measure, and most of the “unemployed” engage in
some kind of self-employment in informal services. These figures enumerate only one form of occupation
per household - an obvious underenumeration of total employment. Data not available for earlier
periods.
Sources: van Nostrand (1993). Ministry of Planning and Economic Development (1992).
FOOD SECURITY IN KAMPALA 1671

Table 2. Cost offood as a percentage of minimum wage, Kampala

Year Minimum wage Consumer Real Wage Percentage of


(Shs./Month) price index Index minimum wage to
(1972 = 100) (1972 = 100) purchase food*

1967 150 75 108 49


I972 185 100 100 60
1984 6,000 35,000 9 450
1988 55,OOW 496,000+ 6 600

*Based on 9,ooO calories per day for a household of four using the price of matooke ZISthe staple, prices
from Kampala’s central Owino market.
tThe Uganda Shilling was demonetized in 1987; one new Shilling was worth 100 old ones. Figures
shown here are 1972 Shillings.
Source: Jamal and Weeks (1993)

Jamal and Weeks depict in graphic terms the farming is low. While attempts at farming are indica-
decline in the value of urban wages between the tive of people’s “survival instincts,” they suggest that
1960s and the 1980s. Whereas in 1972, 60% of the farming is marginal to the urban household economy.
minimum wage was sufficient to provide food for a Two different major trends can be noted regarding
household of four, by 1988 the entire monthly mini- urban agriculture in Kampala, resulting from different
mum wage purchased enough food to last the same historical origins. The municipal boundaries were
household only about four or five days, as depicted by expanded in 1968, incorporating into the city peri-
Table 2. urban areas where farming had always been the
prevalent economic activity. Farming is still the dom-
inant land use and economic practice in some of these
(b) The rise of urban farming areas today, even though the areas described fall
under municipal administration and bylaws. In the
Urban agriculture is not a new phenomenon in inner city and the older suburbs, farming is a rela-
Kampala. Southall and Gutkind (1957) made various tively new phenomenon that, by most respondents’
references to its occurrence. But the common percep- accounts, arose in the mid-1970s as the impact of
tion among current residents is that farming within the Amin’s “economic war” began to be felt by the urban
built-up areas of the city really began in the mid- middle and lower classes. In either case, urban farm-
1970s. Both studies mentioned above (Jamal and ing is a form of semiproletarianism, or reliance on
Weeks, 1993; Bigsten and Kayizzi-Mugerwa, 1992) both cash and subsistence income. In the older parts
note the increase in urban farming, but interpret it dif- of the city it represents a movement away from sole
ferently. Jamal (1985) suggested that urban farming reliance on cash income earned from wage labor and
had been a critical factor stabilizing household food informal trade; in the peri-urban areas, it represents a
security and preventing the massive malnutrition that reluctance to become totally reliant on cash income.
such a devasting economic decline might have About 35% of the households in the city engage in
caused. Bigsten and Kayizzi-Mugerwa (1992) inter- some form of agricultural production (Table 3). Given
pret urban farming almost entirely as a market- that farming households are considerably larger than
oriented activity but note that income from urban the overall mean for the city, this means that the diet

Table 3. Frequency qf urban agricuirure in Kampala

Proportion Enumeration area*


1 2 3 Total
n= I17 fl= 115 n= 116 n = 348

Keeping livestock 20.5% 5.2% 2.6% 9.5%


Cultivating 53.0% 27.0% 21.6% 33.9%
Farming (total) 54.7% 28.7% 21.6% 34.8%

*Enumeration areas:
I Luwafu, a rapidly growing low to middle-income peri-urban area; population density less than 30
people/ha in the enumeration area.
2. Kiswa, a low-income, densely populated inner city slum; population density around 190 people/ha in
the enumeration area.
3. Najjanankumbi, an older, mixed-income residential suburb; population density around 70 people/ha in
the enumeration area.
Source: Author’s household survey, 1993.
1612 WORLD DEVELOPMENT

and/or livelihood of about half the city’s population is no statistically significant association with farming.
directly affected by urban farming. Staple foods, Only the size of household and the length of time the
notably cassava, sweet potatoes, cocoyams, and plan- head of household has been resident in Kampala sig-
tains are the most prevalent crops, but seasonal crops nificantly affect the probability of farming. Length of
such as maize and beans, as well as vegetables and residence is positively associated, meaning that the
fruit tree production are all common. Poultry is the longer someone has been in the city, the more likely
most frequently kept livestock; pigs, small ruminants he or she is to be engaged in farming. This point con-
and cattle are somewhat less frequent, though not tradicts one of the common beliefs about urban agri-
uncommon. Table 3 presents data about the frequency culture - that those farming in the city are the most
of farming in Kampala. recent in-migrants who have yet to integrate them-
Land-use mapping shows that in 1992, 56.1% of selves into the urban economy - and suggests
the land in the city is used for agriculture (Department instead that it takes an extended period of time in the
of Physical Planning, 1992). But less than 20% of city to gain access to land for farming, a point born
urban farmers own the land they farm. Well over half out by interview data.
rely on informal means of access that afford them Studies on the nature of the household (Guyer,
little security of tenure. 1986; Guyer and Peters, 1987; Dwyer and Bruce,
With the exception of a small group of commercial 1988; Folbre, 1986) suggest that intrahousehold
farmers, the primary motive of urban farmers is to processes should also be examined to understand the
secure a nonmarket source of food for the household forces driving urban farming. Recent research sug-
or family of the farmer. Urban farmers in Kampala are gests that what have been labeled household survival
overwhelmingly women; men who farm are mostly strategies may in fact be “an uneasy aggregate of indi-
single and either live alone or live with other men. In vidual strategies.“6 Two major issues are suggested
other words, in conjugal households in the city where for analysis: the allocation of labor and the control of
someone is farming, that person is almost without resources, particularly cash income. In over 80% of
exception a woman. the farming households surveyed, labor invested in
urban agriculture in Kampala is exclusively that of
women; in another lo%, labor was shared between
3. A HOUSEHOLD ANALYSIS OF URBAN men and women. In only about 15% of the households
FARMING IN KAMPALA do adults pool their income and spend jointly. In
the vast majority, therefore, incomes and financial
(a) A household strategy or a woman’s strategy? responsibilities are to some extent kept separate.
Analysis of case studies yielded four main cate-
Urban farming has most often been described as a gories of urban farmers, as depicted by Table 5. The
household survival strategy and the drastic decline in relative frequency of each category in the survey is
the value of wage labor depicted in Table 2 is evi- shown in the second column of Table 5. One group
dence that Kampala households had to alter their eco- produces almost entirely for the urban market, and so
nomic strategies to survive. Table 4 estimates the are categorized as commercial farmers. Numerically
probability of urban farming based on a variety of small, this group is nevertheless economically impor-
household factors. Income level, as well as the age, tant: poultry production is the most common form of
sex, and education level of household head all show commercial urban agriculture, and an estimated 70%

Table 4. Logistic regression, household correlates of urban agriculture

Variable Estimate SE. p-value


Household size 0.155 0.057 0.007
Number of children 0.173 0.158 0.274
Sex of head of household 0.166 0.319 0.602
Age of head of household 0.024 0.014 0.082
Education of head of household 0.010 0.034 0.769
Time in Kampala 0.031 0.011 0.006
Very low income* 0.900 0.642 0.161
Low income* 0.713 0.559 0.202
Lower middle income* 0.65 1 0.595 0.273
Income pooling 0.234 0.357 0.512

*Income groups in the regression analysis were dummy variables, with the highest income group the
reference. The proportion of households engaging in urban farming was virtually identical across all
income groups.
Source: Author’s household survey, 1993.
FOOD SECURITY IN KAMPALA 1673

Table 5. Descriptive categories of urban farming*

Category % of Primary Primary use of Land tenure+ Household Model case


total reason produce income

Commercial 2.5% Income Sales Own or Middle, Poultry, run as a


lease high commercial business by
men or women

Self-sufficiency 5.8% Food Basis of diet Customary Low Long-term customary


tenancy tenants, men or women;
city grew up around
their holdings

Measure of 81.0% Food a) Supplementary All categories$ AI1 levels$ Low or lower-middle
food security b) Reserve income conjugal
household, woman
farms

No other means 10.7% Food a) Basis of diet “Squatting” Low or Very low-income female
b) Often forced to or very low household head, often
sell some food to borrowing recently widowed or
meet other expenses abandoned by husband

*Farming households only, proportions in column 2 are based on the number of farming households in the survey (N = 121).
*See following section for discussion of land tenure.
$Strong statistical association noted between income group and tenure security of within the “means of food security”
category.
Source: Author’s case studies and household survey, 1993.

of all poultry products consumed in Kampala are pro- holds, women view farming very differently from
duced in the city. A second group is comprised mostly men: women view urban farming in very pragmatic
of households in the peri-urban areas who retain terms as a source of food; men describeit in terms of
enough of their customary land holdings to be charac- women’s cultural expectations of themselves, but
terized in terms of self-sufficiency, though in practice, view the practice as being of only marginal
this refers to self-sufficiency in staple foods. Though importance. Married women who farm noted several
cash-poor, these households tend to be somewhat reasons for doing so, beyond the provision of supple-
insulated from the high cost of living in Kampala. The mental food. First, farming meshes well with other
largest group farming in the city does so for some expected household duties, particularly cooking and
measure of food security. Farming for this group is child care. Second, some women have businesses that
likely to be both a secondary form of employment and rely on the produce of farming, selling prepared foods
a secondary source of food. The last group consists is the most common. Third, not only does farming
largely of the households of very low-income women, provide a source of food for their families, it can also
often recently widowed or abandoned by their hus- be a means of protecting other sources of women’s
bands, who face very limited economic options. income in the event that the money allocated to
Members of this group farm because they often have household needs from their husband’s income is
no other means by which to acquire food. Whereas the insufficient. Women rarely challenged their hus-
“measure of food security” group can usually con- bands’ perceptions of urban farming as a marginal
sume all the food they produce, this last group is often activity. In fact, under circumstances characterized by
forced to sell food, even if food for the household is intrahousehold conflict over control of resources, par-
insufficient. ticularly cash, women have good reason for keeping
Within the third and largest category, the use of their economic activities marginal in appearance.
food from farming varies. In some cases it is simply a Should urban farming be understood as a “house-
supplement to other sources of food to be used when hold strategy?’ In terms of the imperative to provide a
it is available. In other cases it is a strategic reserve secure source of food, the household is the social unit
for times when there is no money for buying food, from which the imperative springs. Nevertheless, the
either because of erratic household income or analysis here has suggested that in conjugal house-
because, in conjugal households, the wife who is holds where incomes are not pooled, farming is
responsible for food is not allocated a sufficient specifically a strategy of women to protect or supple-
amount of money by her husband. In such house- ment their other sources of cash income and to assert
1674 WORLD DEVELOPMENT

some control over a source of food for their families to land and stopped. The value of land in Kampala has
that is not dependent on either the urban food market increased rapidly over the past decade,* but privately
or their husband’s income.’ held land in the city is not taxed and public land has
been ineffectively managed for much of the past two
decades. Under these circumstances, urban and peri-
(b) Access to urban land for farming urban land parcels that are privately held or occupied
by customary tenants have been indiscriminantly sub-
The land tenure system prevailing in Kampala is divided and sold, leading to the crowded, sprawling
highly complex. Kampala was two separate cities appearance of low-income neighborhoods. Both pub-
until 1968 - Mengo was the capital of the Kingdom lic and private land has been occupied by “squat-
of Buganda, Kampala was the European and Indian ters.“9 All of these factors help to explain how people
settlement that grew up alongside Mengo. In Mengo, have obtained access to land for farming. Table 6 pre-
land was held under the mailo tenure system - a sents data on the manner in which farming households
form of freehold ownership under which tenants in the survey had obtained their land. While over 20
could acquire long-term occupancy rights from a different categories of land access were noted in the
landlord. In Kampala itself, land was held by the state case studies, Table 6 is simplified to show only aggre-
and administered as public land which could be gate groupings.
acquired on the basis of leasehold. Freehold land Formal ownership of land is strongly associated
grants were made to religious institutions in both with higher income. The lower income groups had
cities in the early colonial era. Customary tenancies mostly gained access to land through one of the latter
were granted under all three tenure systems. As rents four categories, and all of these are subject to change
on mailo land were limited by law, various other as both national and municipal land laws are revised.
forms of occupancy also arose in mailo areas, but the Customary tenancy has effectively split ownership
Amin government abolished this whole category of and occupancy rights between the title holder and the
land in 1975, and vested all land in the state. In prac- tenant: the owner cannot take occupancy without
tice, however, private ownership continued in pre- buying out the tenant, and the tenant cannot legally
viously mailo areas. During the Amin era, the engage in any transaction on the land without the
administrative machinery of local government began consent of the title holder. This set of institutional
to break down and municipal authorities were no arrangements effectively prevents the development of
longer able to adequately manage public land. a market for tenanted land, and land is therefore used
Legal status notwithstanding, the major constraint for farming by the tenant when an economic analysis
to farming in the city is access to land. Of nonfarming might suggest that its value would compel more inten-
households, inability to access land in the city was the sive uses. Such land is often subdivided and sold
major reason given for not farming; 10% of this group without permission of the title holder, and hence is a
had previously farmed within Kampala but lost access major source of land in the third category on Table 6.

Table 6. Frequency of land access/tenure category*

Access/tenurecategory Number of parcels


Enumeration area
1 2 3 Total % Median size (M’)t

1. Formal ownership+ 12 2 12 26 19.1% 300


2. Customary tenancy$ 16 3 19 13.1% 1300
3. Unauthorized subdivision’j 21 I 22 16.2% 635
4. Borrowing 23 8 3 34 25.0% 180
5. Informal access” 7 23 5 35 25.8% 125
6. Total 79 33 24 136 100.0% 380

Median size of farmed plot (M2)t 820 125 260 380


*Units of measure are land parcels, not households, Some households reported farming more than one
parcel.
tOn1y land used for farming; other land uses occur on the same parcel. All figures are medians; means are
considerably larger in all cases.
*Titled or leased land, or contractual rental agreement.
$Long-term occupancy and use rights granted by private or institutional landlord, or by the King prior to
1966.
ISubdivided and sold without the authority of the owner or administrator of the land.
Ykluatting,” purchase of use-rights (“squatters’ rights”).
Source: Author’s household survey, 1993.
FOOD SECURITY IN KAMPALA 1675

Unauthorized subdivision is the most common on extent, maize) are grown on such land. Largely for
land that is not effectively managed by either a private this reason, theft of food from distant land parcels is
owner or a public authority and is a common practice no more prevalent a problem than theft of higher
throughout the city, though is most commonly a value crops closer to the household.
source of land for housing rather than agriculture. This discussion highlights the fact that urban farm-
Over half the land parcels measured in the survey ers have taken advantage of interstitial institutional
fall into the categories of borrowing, “squatting,” or space in urban land created partly by the complexities
the purchase of use-rights from a previous user. of land tenure in the city, and partly by the adminis-
Cultivation of unused land that municipal authorities trative turmoil wrought by the Amin period. While
were no longer adequately managing was a common both the tenure system and the impact of the Amin
means of access in the mid-1970s when farming regime are unique to Uganda, similar processes to
became more common in the inner city. In the para- those described here are occurring in urban land
phrased words of numerous respondents who began across the continent (Aronson, 1978; Mabogunje,
farming during that era, “I just saw a bushy place and 1992; Lado, 1991; Sawio, 1993). Major institutional
started cultivating it.” While some of this land is still reforms, often donor initiatives linked to structural
being farmed today, finding unused land in the city adjustment lending, are now under consideration that
today is virtually impossible, but in areas where would introduce major changes in urban land owner-
people have been cultivating for years without being ship and administration. While helping to sustain
harassed by either municipal authorities or private growth in the commercial sector, these initiatives
land owners, a secondary land market has emerged could have a major impact on the urban poor - a
for the buying and selling use-rights in informally point discussed further in the conclusion of this
occupied land. paper.lO
Borrowing land from a private landowner is the
single most prevalent means of access to land for
people currently farming in the city. Large areas of (c) The impact offarming: income, food
urban land are unbuilt and are held by private land- and nutrition
lords essentially for speculative purposes or as a
hedge against the high inflation that characterized The impact of urban farming at the household level
the Ugandan economy until very recently. Land is is a combination of three factors: income, food secu-
not taxed but structures are, so for purely speculative rity and nutritional status. Income, in this case, refers
purposes, land is often kept undeveloped. Vacant less to cash income than to income in-kind or fungible
land that is not effectively managed, however, often income through savings in food expenditure. The
attracts squatters who erect semipermanent housing major measurable outcome here is nutritional status,
and who often must be compensated when the land is particularly in young children. Table 7 presents a
sold. Owners therefore often have a caretaker who bivariate comparison of the nutritional status of
lends land on a verbal contract basis to women wish- children in farming and nonfarming households
ing to cultivate, a system that serves the interests of all in Kampala while controlling for the influence of
parties: the land-owner reaps speculative profits while income. In the lowest two income groups, which
keeping the land free of uninvited squatters, and can make up nearly 80% of the total sample, there is
terminate the cultivators’ verbal agreements when a strong and statistically significant association
needed; the caretaker earns his or her living through between farming in the city and improved child nutri-
token rents paid by the cultivators; and urban farmers tional status. The same relationship is noted for the
get access to land for cultivation at submarket rents lower-middle income group, although the low num-
and with the assurance that even if told to quit the ber of children weakens the result’s statistical signifi-
land, they will at least be able to harvest their crops cance. The relationship appears reversed in the upper
Iirst. “Rents” usually amount to a token payment of income groups, but the number of children is very
cash at the beginning of the year and offerings of small and the difference is not significant. While there
some of the produce throughout the year at the discre- appears to be a general relationship between income
tion of the cultivator. and nutritional status, among children in farming
Some 65% of the land parcels farmed by house- households there is no statistically significant differ-
holds surveyed are close by the house of the cultiva- ence in nutritional status between the lowest and high-
tor, the remaining 35% vary from 100 meters to est income groups. In nonfarming households, the
several kilometers away. Much of the land acquired difference is two and a half times as large and statisti-
by the latter two means in Table 6 is not close by the cally significant.
household. Because of both the distance from the The comparison in Table 7 is height for age (HAZ)
household of the cultivator and the risk of losing or stunting, a long-term process that has occurred
access to such land, mostly lower value, annual crops sometime in the past. A similar comparison of weight
(cassava, sweet potatoes, cocoyams and to some for height or wasting, a measure of current or acute
1676 WORLD DEVELOPMENT

Table 7. Nutritional status, heightfor age by income group andfarming*

Income groupt RDS Farming Nonfarming Diff t-test p-value Group means
n HAZ$ n HAZ$ N
VL 1 29 -0.710 26 -2.054 55 1.344 0.001 -1.345
2 26 -0.738 24-1.471 50 0.732 0.061 -1.090

L 1 71 -0.610 100-1.026 171 0.416 0.013 -0.853


2 71 -0.514 97-1.002 168 0.488 0.007 -0.796

LM 1 21 -0.310 23-0.861 44 0.551 0.080 -0.598


2 22 -0.527 244.888 46 0.366 0.227 -0.715

UM/H 1 7 0.400 16 0.863 23 a.463 0.503 0.722


2 8 0.525 12 0.967 20 -0.442 0.517 0.790

Group means 1 128 -0.528 165 -0.982 293 0.454 0.003 -0.780
2 127 -0.497 157 -0.906 284 0.409 0.007 -0.720

ANOVA’J[ F = 1.490 F=21.50 F = 16.28


p = 0.208 p < 0.001 p < 0.001
*Confirmed birth dates only.
tVL = Very Low
L=Low
LM = Lower Middle
UMH = Upper Middle and High
*Survey Round. Round 1 during rainy season (April, 1993); Round 2 after harvest of seasonal crops
(July-August, 1993).
$HAZ refers to Height for Age z-scores (standard deviations above or below World Health Organization
reference median), a measure of stunting. All scores age-adjusted for recumbent measurement.
IAnalysis of variance shown for Round 1 only. Similar results were found for Round 2.
sour&: Author’s household survey, 1993.

malnutrition, showed virtually no differences to as the “working poor” that make up the majority of
between the groups depicted in Table 7. Little sea- the city’s residents, indicators of short-term food suf-
sonal variation in nutritional status is evident in Table ficiency at the household level did not differ signifi-
7, and a similar lack of seasonal variation was noted cantly between farming and nonfarming groups, but
for wasting. Both these points imply that the associa- the amount of money spent on food per person per
tion of urban agriculture and nutritional status is a day is significantly less in farming households. This
long-term rather than short-term phenomenon. indicates that for this group, farming is a major
While Table 7 is a simple bivariate comparison source of fungible income in terms of saving on
controlling only for income, the same relationship is food expenditure. In higher income groups, these
found in multivariate analysis when other factors trends were less uniform and the differences were
associated with nutritional status are included: health not significant.13
and maternal care factors, food and dietary factors, A comparison of the mean amount of money spent
and a variety of household factors in addition to on food per adult equivalent by farming and nonfarm-
income. Three main factors mediate the relationship ing households in the low-income group during both
between urban farming and nutritional status of chil- rounds of the survey provides a rough estimate of the
dren: food sufficiency, dietary adequacy, and the value of the urban subsistence production. On aver-
amount of time mothers can directly care for their age, nonfarming households report spending one-and-
children.” a-half to two times as much on food as farming
In the very low-income group, persons responsible households, or $35-45 more per month. While farm-
for the provision and preparation of food in both ing may involve some small cash payments for land
farming and nonfarming households reported spend- access or inputs, and while there is an implicit oppor-
ing roughly the same amount of money on food per tunity cost in the labor time in farming, this is a siz-
person per day. I* In farming households, however, able proportion of total household income in an
unpurchased food was available from farming, result- economy where $125-130 per month is considered by
ing in significantly higher levels of short-term food economic planners to be the cut-off between low and
sufficiency in these households during both rounds of lower-middle incomes.14 This also underlines why
the survey. For the low-income group, often referred many married women have reason to keep their farm-
FOOD SECURITY IN KAMPALA 1677

ing activities marginal in appearance: married women food, either through exchange or production. Security
respondents in focus groups repeatedly insisted that if of access is the converse of the risk of entitlement fail-
their husbands knew the real value of their economic ure; the higher the share of a household’s resources
activities, the result would be a lower fmancial contri- devoted to the acquisition of food, the greater the risk
bution on the part of the husband to the costs of main- of failure (Maxwell and Frankenberger, 1992).
taining the household, which would increase the Such a definition highlights the nature of urban
financial strain on women and reduce their options for farming: it is the deliberate usage of labor and land for
maintaining food security. subsistence production (direct entitlements according
The amount of time that mothers are able to care to Sen’s definition) to reduce dependence on financial
directly for their children is also positively associated or exchange entitlements in the provision of food.16
with child nutritional status. A comparison of farming To the extent that urban farmers are successful in
women with women engaging in wage labor or infor- making maximum use of land and labor while main-
mal trade indicates that farming women have signifi- taining access to cash from other sources of income
cantly more time to devote to childcare, bearing out within their households, they do achieve higher levels
almost precisely the logic expressed by many women of both food security and nutritional status. By engag-
farmers. ing in direct production, urban farmers accept the
inherent risk that drought or pests may reduce yields,
which in turn would reduce food availability from
4. DISCUSSION AND POLICY IMPLICATIONS farming, but interview data suggest that this occurs
infrequently in Kampala.
(a) Urban farming as a food securie strategy The much larger risk to food security is the risk of
losing access to land. The actual amount of land does
To summarize, urban farming in Kampala is not appear to be of major importance: the size of land
largely a long-term adaptive strategy of women to parcels for farming in Kampala ranged from as little
protect the food security of the persons for whom they as nine square meters to as much as several hectares,
are responsible, either through the direct provision of yet the size of the parcel being farmed correlated only
supplemental source of food, a food reserve, or as a very weakly and insignificantly with nutritional sta-
means of stretching other sources of income. Urban tus, and not at all with food sufficiency measuresI
farming has been demonstrated to be significantly This is not to imply that giving every household in
associated with improved food security and child the city nine square meters would result in major
nutritional status, particularly among lower-income improvements in the level of urban malnutrition, but
groups. In the majority of cases, urban farming is only it does imply that physical limitations in urban
one of several strategies employed in the provision of space are only a partial constraint to making use of
food, but it is a special case for several reasons. First, urban agriculture as a policy tool or programmatic
farming is almost completely under the control of intervention.
women, who bear primary responsibility for provi- The crucial factor is the way in which land is man-
sion of food. Second, farming provides a source of aged, even though to the casual observer, farmed land
food that is dependent on access to land rather than appears to be managed in a haphazard way. The key
access to cash income. to relying on urban farming for food security is two-
A recent review defined food security as “secure fold. First, the period of time during which food can
access at all times to sufficient food.“i5 Four basic be harvested must be maximized, meaning that plant-
concepts are stressed in this definition: sufficiency, ing must be staggered throughout the year, and root
temporal considerations, access, and security crops that can be harvested piecemeal are favored
(Maxwell and Frankenberger, 1992). Sufficiency of over seasonal crops that must be harvested immedi-
food is often a subjective measure; in this study it was ately when they mature. Second and more impor-
a measure of the frequency and severity of means for tantly, extreme care must be taken not to lose access
coping with food insufficiency. Temporal considera- to land. Both of these imperatives contribute to the
tions can either be cyclical or secular; in Kampala haphazard appearance of agricultural practices: plant-
measures of food sufficiency did improve in the dry ing root crops at suboptimal times in order to ensure
season compared to the rainy season, but there were year-round occupancy of land, which is the only way
few differences between farming and nonfarming to maintain a claim to land accessed through borrow-
households in the level of improvement. Long-term ing or squatting; intercropping mixes that do not
trends in Kampala indicate secular changes in diet and include legumes; or planting seasonal crops in the
in food access strategies. Access to food is deter- shade of root crops or perennials. These apparently
mined by food entitlements (Sen, 1981), the sum of haphazard farming practices inadvertently help to
assets - human, physical or financial - which an maintain the marginal appearance of urban farming,
individual or household can use to acquire food, and but also earn the scorn of some agricultural profes-
the rate at which those assets can be converted into sionals who view yield per unit of cultivated area as
1678 WORLD DEVELOPMENT

the variable to maximize. Clearly, technical improve- wetlands or the unprotected cultivation of steep hill-
ments could boost productivity, but the key con- sides. Unfortunately, changing the legal status is not a
straints to urban farming are first and foremost matter of simply deleting a single urban bylaw.
institutional, not technical. These constraints go back Rather than a single bylaw, in Kampala there is a
to the maintenance of control over human and physi- three-decade long series of verbal decrees, unwritten
cal resources. Framing this discussion in terms of the rules and administrative practices. These can be
definition of food security noted earlier highlights changed, and indeed there is evidence that official
these constraints. attitudes are changing, but the changes cannot be
To evaluate food security implications fully, two implemented by administrative fiat or a political vote.
further points must be made. The first concerns the
physical and financial aspects of food entitlements. (ii) Access to land
The comparison between farming and nonfarming The land issue highlights both a physical resource
households should be broadened to include three and an institutional constraint. The use of land for
groups: those without land or sufficient exchange farming in Kampala has to some extent relied on
entitlements (low-income groups that do not farm); niches created by the complexity of tenure rules in the
those with secure rights in their land resource or with city and by administrative turmoil. Urban planners
secure exchange entitlements (low-income landown- question whether the use of land for farming, particu-
ing farmers, or middle- and upper-income groups irre- larly the kind of farming practiced in Kampala,
spective of whether they farm); and those who farm reflects the value of the land, and suggest that tax
but who lack secure rights in the land they use (low- rules and confusion over property rights act as a fetter
income farmers who accessed land informally). on other forms of investment in the land which would
Survey data indicate that about 52% of the population preclude its use for farming. Institutional reforms in
of Kampala would fall into the first category, 21% urban land are being implemented as part of structural
would fall into the second, and 27% into the third. adjustment lending programs, not only in Kampala
Protecting the access of the third group while enhanc- but elsewhere on the continent (Nsamba-Gayiiya,
ing the access of the first group would be one objec- 1993; Mabogunje, 1992). These reforms are aimed
tive of any programmatic intervention in urban at increasing the tax base, streamlining urban land
farming. administration, or introducing major changes in the
The second point about food security regards con- system of land ownership. All of these changes would
trol over human resources - in other words, control represent a substantial threat to continued informal
over labor. At the household level, the retention of access to urban land for farming. But part of the con-
control over food by those who produce it is impera- cern in these reforms is maintaining the access of the
tive not only from an equity perspective, but also from urban poor to land, and even if the proposed reforms
a food security perspective. Both these points have were adopted immediately, they would take some
important implications for policy toward urban time to implement. Medium-term planning in
agriculture. Kampala is presently based on the presumption that
major changes in land tenure rules will not take effect
in the coming decade (van Nostrand, 1993).
(b) Policy implications for urban farming Given that a large proportion of the city’s popula-
in Kampala tion directly obtains some part of its food needs from
urban farming, and given the relationship between
Given increased interest in urban agriculture as a farming and food security at the household level as
tool to alleviate urban poverty and some of the outlined above, there will naturally arise some con-
adverse impacts of structural adjustment programs on flict of interest between land-use planning aimed at
urban populations in Africa, several categories of pol- economic growth at the level of the city as a whole
icy recommendations follow. The first concerns the versus economic survival at the level of the household.
legal status of urban farming. The second concerns In the abstract sense, planning for overall economic
the land issue. The third concerns the issue of labor. growth would call for a land policy that promotes
“best and highest use” - a free and open market in
(i) Legal status urban land, secure and recognized ownership rights,
Clearly, if urban farming remains technically ille- and value-based taxation. On the other hand, a more
gal, little can be done in the way of programmatic politically pragmatic approach to planning would rec-
intervention even if farming in the city is a pervasive ognize that one of the reasons Kampala residents have
practice. Not only does the legal issue prevent small- been able to withstand both the economic crisis of the
scale projects from being undertaken, it also prevents 1970s and 1980s and the impact of structural adjust-
the rational implementation of programs to control ment policies in the 1980s and 1990s is that a sizable
some of the environmentally damaging aspects of proportion of them had informal access to urban land
urban agriculture, such as the unplanned draining of for both housing and subsistence production.
FOOD SECURITY IN KAMPALA 1679

Beyond a general commitment to maintain the in terms of food security, and intrahousehold conflict
access of the urban poor to land, several specific sug- over benefits is somewhat less likely. Programs could
gestions follow the above discussion. The first would be established through women’s organizations such
be the alteration of zoning ordinances to permit mixed as informal savings and credit groups. Such groups
land use in general, and possible specific designation are often numerous in urban and peri-urban areas, and
of certain areas for farming. The second point is the their informality is one of their strengths. Lastly, the
promotion of very localized control over land use, a impact of any intervention should be. closely moni-
phenomenon that is already occurring to some extent tored, both in terms of the direct effect on women’s
in Kampala. This permits conflicts arising over squat- income, and in terms of food security and child nutri-
ting and other forms of informal land access to be tional status.
resolved by local leaders who have greater knowledge Farming in the city is a strategy of urban women
about the circumstances and greater sympathy with to protect the food security of their families and
local residents who may be dependent on the land. households. Programmatic intervention into urban
The third recommendation about land regards the agriculture should, as a first priority, protect and
compelling case to be made for compensation for the enhance the food security strategies of vulnerable
loss of access to farm land. Cities inevitably grow and urban groups.
displace farmers from agricultural land, a trend ulti-
mately unchanged by the fact that it was temporarily
reversed for a decade and half in Kampala. Current (c) Broader implications
practice is variable, but evictees are often compen-
sated only for improvements on the land, not to rights This analysis suggests that semisubsistence pro-
in the land itself. The value of one season’s crops is duction in the city is not marginal, but in fact critical
tiny compared to the value of continued access. The to urban household food security, and is an important
extent of compensation often depends more on the aspect of women’s economic strategies. While spe-
good graces of the owner or on power relations cific statistical findings here cannot be generalized
between owner and occupant than on accepted rules, beyond Kampala, several general points should be
written or unwritten. considered, both in other African cities and in densely
populated periurban areas. The first is the growing
(iii) Conrrol over labor importance of semiproletarian economic strategies,
Labor in urban farming is largely that of women, or reliance on both subsistence production and
and it is reasonable to suggest that technical services cash income from wage labor or informal trade.
could be offered to support women’s efforts in farm- Apparently marginal economic activities may, in fact,
ing. The above analysis suggests, however, that farm- be critical to economic survival. The second is the
ing is a successful food security strategy of urban importance of understanding of intrahousehold
women at least in part because the real value of food conflict in the production and/or purchase of food.
produced is known only to the women who produce it. Traditional patterns in food procurement are rapidly
Research from elsewhere in Africa suggests that high- changing in both rural and urban areas, paralleled by
visibility interventions can be coopted by men to changes in gender relations and responsibilities. The
capture women’s labor or rights in land and natural third is the retreat of the state in the face of fiscal
resources in circumstances where women had constraints and structural adjustment, the resulting
previously asserted some autonomy over their eco- changes in both markets and safety nets, and the
nomic activities (Schroeder, 1993). increased importance of self-reliant strategies at the
Such an observation would imply several caveats: household level. Nevertheless, state action can either
Programs promoting urban farming for equity reasons enhance or hinder equity and effectiveness of such
should give priority to low-income, female-headed strategies - through legal statutes, policy, and direct
households. These are likely to be the most vulnerable program or project interventions.

NOTES

I. World Commission on Environment and Development Land, Food and Farming: a Household Analysis of Urban
(1987) p. 254. Agriculture in Kampala, Uganda.” University of Wisconsin
(1995).
2 Von Braun et crl. (1993). p. 38.
5. Magendois usually translated as “black market.”
3. “Feedmg the Urban Masses” is the title of Sawio’s
study. 6. Dwyer and Bruce (1988). p. 8.

4. The full study is the author’s Ph.D. dissertation, “Labor, 7. For a more complete discussion of gender relations and
1680 WORLD DEVELOPMENT

intrahousehold dynamics in urban farming, see Maxwell of domestic foodstuffs has not dropped. The shilling traded
(1994). at 1,200 to the dollar at the time of the survey. In early 1995,
the rate was closer to 950 to the dollar. In other words, mea-
8. The limited information available suggests a rate of sured in dollars, the value of subsistence production has
increase of 50-759’0 per year above the general rate of infla- probably risen dramatically since these data were collected.
tion. See Nordberg and Nsamba-Gayiiya (1991).
15. Maxwell and Frankenberger (1992) p. 8.
9. The term is used in various ways in Kampala, but is used
here to mean only a land user who does not have the permis- 16. Sen’s famous analysis, of course, did not apply to urban
sion of the owner or administrator. farmers or consumers, but his example - a comparison of
the entitlement risks of sharecroppers and fixed-wage rural
10. For a more complete discussion of the land issue in agricultural laborers under conditions of sharp changes in
urban fanning, see Maxwell (1993) and (1995). food prices and/or food availability (i.e. famines) - is
analogous to urban residents who get some of their food
I 1. Food sufficiency here is measured by the frequency and from their own farming in the city and those who are com-
severity of measures for coping with insufficient food, such pletely dependent on market sources under circumstances
as eating foods that are less preferred, borrowing food or depicted in Table 1.
money, limiting portion size, skipping meals, and maternal
buffering. Dietary adequacy is a composite measure of 17. The main reason for this seemingly counter intuitive
major food groups following the methodology of Guthrie finding is that for most urban farmers, subsistence produc-
and Sheer (1981). tion is a secondary or fall-back source of food, as was noted
in Table 5. Clearly, having this fall-back makes a difference
12. Again, with the exception of single men, these persons in food security, but beyond some threshold level, having
were almost always women. access to more land probably means that labor is devoted to
farming rather than to other sources of income, and losses in
13. This section summarizes a complex analysis. For a more exchange entitlements roughly cancel out gains in direct
complete analysis, see Maxwell (1995). entitlements. The reason farming makes an overall differ-
ence is that cash income (exchange entitlements) may be
14. Putting these figures in dollars at the 1993 exchange rate irregular, or, as noted above, may be the object of intra-
is somewhat misleading, since the shilling has strengthened household conflict, and farming protects food security in
significantly against the dollar in the meantime, but the price interim periods when exchange entitlements fail.

REFERENCES

Aronson, Dan, “Capitalism and culture in Ibadan urban Freeman, Donald, A City of Farmers: Informal Urban
development,” Urban Anthropology, Vol. 7 (1978), pp. Agriculture in the Open Spaces of Nairobi, Kenya”
253-267. (Toronto: McGill University Press, 1991).
Atkinson, Sarah, “Food for the cities: Urban nutrition policy Guthrie, Helen and James Sheer, “Validity of dietary score
in developing countries,” Urban Health Program, Health for assessing nutrient adequacy,” Journal of the American
Policy Unit. Department of Public Health and Policy Diereric Association, Vol. 78 (March, 1981). pp.
(London: London School of Hygiene and Tropical 24&244.
Medicine, 1992). Guyer, Jane, “Intra-household processes and farming sys-
Banugire, Firimoni, “Class struggle, clan politics and the terns research: Perspectives from anthropology,” in Joyce
‘Magendo’ economy,” Mawazo, Vol. 6, No. 2 (1985). pp. Moock (Ed.), Understanding Africa’s Rural Households
52-66. and Farming Systems (Boulder, CO: Westview Press,
Bigsten, Ame and Steve Kayizzi-Mugerwa, “Adaption and 1986).
distress in the urban economy: A study of Kampala Guyer, Jane and Pauline Peters, “Introduction:
households,” World Developmeni, Vol. 20, No. 10 Conceptualizing the household,” Development and
(1992). pp. 1423-1441. Change, Vol. 9, No. 4 (1987), pp. 197-214.
Department of Physical Planning, Cir)’ of Kampala: Revision Hussain. A. M. and Paul Lundven. “Urbanization and
of Structure Plan (Kampala: Ministry of Land, Housing hunger in the cities,” Food and Nutrition Bulletin, Vol. 9,
and Urban Development, 1992). No. 4 (l988), pp. 5@6l.
Drakakis-Smith, David, “Strategies for meeting basic food Jamal, Vali, “Structural adjustment and food security in
needs in Harare,” J. Baker and P. Pedersen @is.), The Uganda,” World Employment Research Paper (Geneva:
Rural-Urban Interface in Africa. (Uppsala: The International Labour Organization, 1985).
Scandinavian Institute of African Studies, 1992). Jamal, Vali and John Weeks, Africa Misunderstood
Dwyer, Daisy and Judith Bruce, Home Divided (Boulder, (London: Macmillan, 1993).
CO: Westview, 1988). Jamal, Vali and John Weeks, “The vanishing rural-urban gap
Folbre, Nancy, “Cleaning house: New perspectives on in sub-Saharan Africa,” International Labour Review,
households and economic development,” Journal of Vol. 127, No. 3 (1988). pp. 271-291.
Development Economics, Vol. 22 (1986). pp. 5-40. Lado, Cleophas, “Informal urban agriculture in Nairobi,
FOOD SECURITY 1N KAMPALA 1681

Kenya,” Land Use Policy, Vol. 7 (1990), pp. 257-266. on Urban Land Policy and Management” (Kampala:
Lee-Smith, Diana, Mutsembi Manundu, Davinder Lamba Ministry of Lands, Housing and Urban Development,
and Kuria Gathuru, Urban Food Production and the 1993).
Cooking Fuel Situation in Urban Kenya (Nairobi: Rakodi, Carole, “Women’s work or household strategies?,”
Mazingira Institute, 1987). Environment and Urbanization, Vol. 3, No. 2 (1991). pp.
Mabogunje, Akin, “Perspectives on urban land and urban 39-45.
management policies in sub-Saharan Africa,” World Rakodi, &role, “Urban agriculture: Research questions and
Bank Technical Paper Number 196 (Washington, DC: the Zambian evidence,” Journal of Modern African
World Bank 1992). Studies, Vol. 26, No. 3 (1988), pp. 495-515.
Maxwell, Daniel, “Labor, land, food and farming: A house- Sachs, lgnacy and Dana Silk, “Introduction: Urban agricul-
hold analysis of urban agriculture in Kampala, Uganda,” ture and self-reliance,” Food and Nutrition Bulletin, Vol.
Ph.D. Dissertation, (Madison, WI: University of 9, No. 2 (1987). pp. 2-4.
Wisconsin, 1995). Sanyal, Bishwapriya, “Urban agriculture: Who cultivates
Maxwell, Daniel, “Unplanned responses to the economic and why?,” Food and Nutrition Bulletin, Vol. 7, No. 3
crisis? Urban agriculture in Kampala,” Paper presented to (1985). pp. 15-24.
the Workshop on “Developing Uganda,” Centre of Sawio, Camillus, Feeding the Urban Masses.? Towards an
African Studies, Univrrsity of Copenhagen, held at Understunding of the Dynamics of Urban Agriculture in
Lyngby Landbrusskole (Roskilde, Denmark: June 2-5, Dar es Salaam, Tanzania (Ph.D. Dissertation, Worcester,
1994). MA: Clark University, 1993).
Maxwell, Daniel, “Land access and household logic: Urban Schroeder, Richard, “Shady practice: Gender and the
farming in Kampala,” MISR Research Paper (Kampala: political ecology of resource stabilization in Gambian
Makerere Institute of Social Research, 1993). garden/orchards,” Economic Geography, Vol. 69,
Maxwell, Daniel and Samuel Zziwa, Urban Farming in No. 4 (1993), pp. 349-365.
Africa: The Case of Kampala, Uganda (Nairobi: ACTS Sen, A. K.. Poverty and Famines: An Essay on Entitlements
Press, 1992). and Deprivation (Oxford: Clarendon Press, 1981).
Maxwell, Simon and Timothy Frankenberger, Household Smit, JX and Joe Nasr, “Urban agriculture for sustainable
Food Security: Concepts, indicators, Measurements. A cities: Using waste and idle land and water bodies as
Technical Review (New York and Rome: UNICEF and resources,” Environment and Urbanization, Vol. 4, No. 2
IFAD, 1992). (1992). pp. 141-151.
Memon, Pyar Ali and Dianna Lee-Smith, “Urban agriculture Southall, Aidan and Peter Gutkind. Townsmen in the
in Kenya,” Canadian Journal of African Studies, Vol. 27, Making: Kampala and its Suburbs (Kampala: East
No. 1 (1993), pp. 25-42. African Institute of Social Research, 1957).
Ministry of Planning and Economic Planning, “The 1991 van Nostrand, John, Kampala Urban Study Phase I Report
population and housing census, Kampala District” (Toronto: John van Nostrand Associates, 1993).
(Entebbe: Department of Statistics, 1992). von Braun, Joachim, John McComb, Ben Fred-Mensah
Mvena, Z. S. K., I. J. Lupanga and M. R. Mlozi. Urban and Rajul Pandya-Larch, “Urban food insecurity and
Agriculture in Tanzania: A Study of Six Towns malnutrition in developing countries: Trends,
(Morogoro, Tanzania: Sokoine University of Agriculture, policies, and research implications” (Washington,
1991). DC: International Food Policy Research Institute,
Nordberg, Lauri and Eddie Nsamba-Gayiya, “Report on 1993).
land delivery” (Kampala: Ministry of Lands, Housing and World Commission on Environment and Development, Our
Urban Development, 1991). Common Future (Oxford: Oxford University Press,
Nsamba-Gayiiya, Eddie, “Report of the National Workshop 1987).

You might also like