(introduction...)

Daniel G. Maxwell

Among the various claims in the literature on the impact of structural adjustment programs on the urban poor and wage-earning classes in Africa is that made by Pinstrup-Andersen (1989). He refers briefly to Kampala to suggest that access to land for semisubsistence production may buffer or prevent the decline in nutritional status and household food security that is widely believed to be to the product of the urban economic crisis, or structural adjustment, or both. The source to which Pinstrup-Andersen refers is an article by Jamal (1988, p. 684), who claims that “Kampala is twice as self-sufficient in calories now as it was in 1972.” Although this statement may well be true, Jamal does not present evidence on either caloric intake or nutritional status of farming families within the city.

The depth of the economic crisis in African cities in the 1980s has been well documented (Jamal and Weeks 1987; Stren and White 1989). In the case of Kampala, the urban economic crisis began much earlier than in many other cities in the region, as a result of the Amin regime’s “economic war” in the early 1970s, which created space for an indigenous bourgeoisie, but undermined much of the formal economy (Banugire 1985; Mamdani 1990). Wage income fell precipitously in relation to the cost of living between the end of the 1970s and the present, and the major response at the household level was to diversify sources of income as a buffer against inflation and falling real wages (Bigsten and Kayizzi-Mugerwa 1992).

Only comparatively recently have donors and governments become aware of the potential risks to short- and medium-term food security and nutritional status of vulnerable groups in the process of adjustment. The conventional argument is that structural adjustment is designed to make farming a more economically attractive livelihood, and to solve the urban food problem through increased incentives for rural production. However, in the short- to medium-term, the burden on low- and middle-income urban households is increased.

The objective of this paper is to evaluate the various claims made about urban agriculture (UA) in Kampala. This includes reviewing the limited literature on the importance of UA in Kampala; attempting to assess what direct evidence is available on the question of nutritional status; examining the means of access to the critical land resource for UA; and understanding the logic of different kinds of households involved in urban food production to interpret why different groups of people engage in it.

The importance of agriculture to the city’s economy

Slightly over 50% of the land within the municipal boundaries of Kampala is used for agriculture (GTZ/DPP 1992). However, this includes areas that cannot be described as “urban,” with population densities as low as six people per hectare. Nevertheless, all the area within the municipal boundaries is subject to the same urban bylaws, which technically do not permit the use of land for agricultural purposes.

Various estimates have been quoted for the prevalence of agriculture among Kampala’s households. Our own previous work (Maxwell and Zziwa 1992) estimated that 36% of all households within a 5-km radius of the city centre engaged in some kind of agricultural production, but our sampling technique did not permit statistical extrapolation. The current study estimates 30% of households in the city are involved in agriculture. UNICEF/KCC (1981) estimated 25% of low-income households farmed and Save the Children Fund (SCF) put the estimate at 28% (Riley 1987). However, in both these cases, households surveyed included only those with small children. Taken as a whole, these studies indicate at least a general range of the proportion of Kampala households that engage in farming.

Crops grown are largely staple food crops: cassava, sweet potato, beans, maize, matooke (plantains), and cocoyams, in order of descending prevalence. Vegetable crops and fruit trees are also grown and a limited number of commercial producers grow coffee and even vanilla beans in the city. Among livestock producers, poultry raising (for both meat and eggs) is most common, but cattle, small ruminants, pigs, and rabbits and other microlivestock are raised as well (Maxwell and Zziwa 1992).

Reliable data on levels of production or total value of production are not available. Our earlier study indicated that roughly 20% of the staple foods consumed within the 5-km radius of the city centre were produced within the same area. Because this is the most built-up area of the city, the estimate would probably be higher for the whole area within the municipal boundaries. Jamal (1988) estimated that the city is 40% self-sufficient in calories, but presents no evidence to back up the claim. Government statistics indicate that some 70% of the poultry products consumed within Kampala are produced there.

Evidence on nutritional status

Several studies in the past 12 years have gathered data that can be used to assess the impact of UA on the nutritional status of children. In 1981, in the aftermath of the war with Tanzania and a politically turbulent period, the United Nations Children’s Fund (UNICEF) attempted to assess the need for supplementary feeding of children in 13 low-income Kampala neighbourhoods, and included on their brief questionnaire questions about whether the family had access to a shamba for home production of food. Their conclusion was that despite the war and the dramatic economic decline of the later Amin years, supplementary feeding was not needed. They mentioned food production within the city as a major contributing factor (UNICEF/KCC 1981), although their analysis did not include an explicit comparison of farming and nonfarming groups.

In 1987, SCF carried out a similar nutritional survey in Kawempe Division of Kampala, to determine whether their supplementary feeding program for war-displaced children should be continued. This study also concluded that supplementary feeding programs were not needed, and that urban food production was a contributing factor (Riley 1987), but again, farming and nonfarming groups were not directly compared, even though the data collected would permit such a comparison. Two other studies (Kakitahi and Zimbe 1990; Biryabarema 1994) have obtained baseline data on malnutrition on children in Kampala and have included, as background information, questions about whether food is produced by the family of the children being assessed.

The SCF data suggest that the long-term growth (height-for-age) of children differed significantly between farming and nonfarming households, with children in the former category on average half a standard deviation higher in terms of comparison with a reference median (Table 1). However, the rate of short-term growth (weight-for-height) did not differ significantly (Table 2). It should be noted that these data were collected in March, during the rainy season; thus it is to be expected that little supplementary food would be available for household consumption from the household’s own production within the city. Both farming and nonfarming households alike would depend on market sources for food during this period.

Table 1. Height-for-age Z-score comparison (indicative of stunting or long-term malnutrition).

Table 2. Weight-for-height Z-score comparison (indicative of wasting or short-term malnutrition).

The 1991 Kawempe Community Diagnosis Survey data indicate some difference in the long-term nutritional status of children in farming and nonfarming households, but not a statistically significant one. Again, because the data were collected in the rainy season, there is virtually no demonstration of the impact of farming on short-term nutritional status (Tables 3 and 4).

Table 3. Height-for-age Z-score comparison (indicative of stunting or long-term malnutrition).

Table 4. Weight-for-height Z-score comparison (indicative of wasting or short-term malnutrition).

Of course, this is only a crude comparison, and does not consider such factors as income, parental education, or household composition, nor amounts of land farmed or length of time in farming, all of which might be expected to have an impact on stunting or prevalence of long-term undernutrition. However, if there are indications of a relation between urban farming and improved nutritional status to be found in simple bivariate analysis of data from past studies in the city, it seems reasonable to conclude that the claim made by Pinstrup-Andersen (1989) is sufficiently valid to warrant further investigation.

Household logic in urban agriculture

In the view of people who farm in Kampala, the logic of UA can be described, following Mingione (1991), as a form of semiproletarianization, or relying on a measure of both labour-market participation or petty trading, as well as home-production for consumption. However, there are two distinctly different forms of agriculture within the city. The first, occurring within the central city, the older suburbs, and City Council housing estates, represents a long-term movement away from the labour market in both the formal and informal sectors of the city’s economy, with increased effort over time devoted to production for direct consumption. The other, occurring within the newly built-up areas and the pert-urban parts of the city, represents movement toward either the labour market or informal trade, but a reluctance to become entirely dependent on either.

Across the spectrum of people who farm in the city, whether commercially oriented or not, whether male or female, whether low or middle income, farming is an important spoke in the wheel of economic life. Nevertheless, the logic of how farming is incorporated into the economic strategy of the household or individual varies greatly At least four major categories of household logic emerge from an analysis of interviews: commercial production, household food self-sufficiency, a measure of food security, and a ‘no-other-means’’ logic.

Commercial Production

Although scattered examples exist of commercial production of exotic crops such as coffee or vanilla beans within Kampala, by far the largest category of commercial production in Kampala is in livestock, particularly in poultry. Although current commercial producers may have begun farming for home consumption, almost without exception, the capital invested in commercial production comes from a source outside of agriculture. Hence, commercial agricultural production is a lucrative investment of capital, rather than the evolutionary outgrowth of small-scale production.

Commercially oriented households tend to have a high level of income and access to sources of large-scale credit. Evidence from household case studies suggests that commercial producers may be either men or women, and that male and female household members may well collaborate in commercial production, treating income and expenditures as any business would. The income and expenditure pattern from the business may have little effect on other household roles.

Urban Food Self-Sufficiency

Food self-sufficiency here is taken to mean self-sufficiency in staple foodstuffs and sources of protein. Certain food items are still purchased but this kind of household is largely insulated from the high cost of food in Kampala. Even though its income may be very low in monetary terms, it is relatively food secure, except in the event of bad growing seasons or drought. These households are almost without exception well established and the family has been living in Kampala for a long time. These households must have access to a significant amount of land, including some land in swamps or low lying areas to permit year-round production of certain staple crops, such as cocoyams, that both tolerate flooding during the rains and grow on swamp land during the dry season. Almost without exception, the land used for farming is not owned by the household (if it were, it would imply a level of wealth that would permit a more leisurely lifestyle than farming). This requires well-established social relations with land owners or long-term residence in previously sparsely populated areas.

Although there is hardly a commercial logic to production, this type of farming household does sell some of its agricultural produce to generate income for other needed purchases, but the sales tend to be based on having a surplus after household needs are met or raising cash for an emergency, rather than on maximizing profit. Otherwise the selling of casual labour is a more frequent form of providing a source of cash.

With the notable exception of households in areas where the city has literally grown up around their customary land holdings, the family or the head of household reported engaging in either wage labour or some other nonagricultural occupation for some period of time after migrating to the city, before beginning to farm, and long before reaching a level of self-sufficiency in farming.

Food Security

Production for food security is by far the most common category where members of a household have gained access to some land- either within the compound on which their house or flat is located, or elsewhere-and someone within the household is producing food on it. However, the amount of food produced does not constitute the majority of what the household consumes. Almost without exception, persons within such households who are responsible for food procurement and preparation say that the market is their major source of food, with the garden or farmed plot being a secondary source, but one that provides an important buffer against a short-term shortage of food or cash.

In some cases, the land used for farming may be owned by someone in the household, and the farming may genuinely represent a “household” allocation of resources-both land and labour, and in some cases, income from other sources is used to pay for hired agricultural labour. The more common case is that of the senior woman in the household gaining access to land on her own basis through an arrangement of borrowing, renting, “squatting,” or purchase of use rights. Providing some amount of food for the household both increases the food security of the members of the family for whose welfare she is responsible, as well as permitting her to use her own cash income on items other than the purchase of food. It is particularly this group of women farmers who insist that they would never stop farming in exchange for another job that, in monetary exchange values, payed at the same rate.

Unlike the other categories, this group almost never sells any of the food produced. Other sources of income are usually present in the household-often even other sources of female income. In fact, protection of that income for nonfood uses constitutes a major part of the logic of this kind of agricultural production.

No Other Means

In a sense, the group whose logic is based on “no-other-means” is a more extreme form of the food-security group. This group often includes low-income female-headed households, widows, and families suddenly abandoned by a primary wage earner. As a group, they are very low-income, food-insecure, and land-insecure households. Although there is nothing like profit-maximizing logic, this group is often forced to sell some of what it produces to meet other expenses. In fact, it is this aspect that distinguishes this group from the former one: the former can afford to always eat the food they produce; this group is often forced to sell some, even if they don’t have enough to eat.

Such households are often not well enough connected to gain access to land through any means other than “squatting.” This means that they are often subjected to evictions without compensation. Farming, for this group, constitutes something of a “survival strategy” in the most literal sense.

Land access and tenure

Kampala is a combination of two cities: Kampala itself, the commercial capital of the Uganda Protectorate, largely planned and built by the colonial power; and Mmengo, the capital of the Kabaka of Buganda. Until 1968, they were two separate municipalities. Tenure practices in the two were quite different. In Mmengo, the Mailo system prevailed, which allowed for private ownership of land, and in which public land was held in trust by the Kabaka himself or other notables. In Kampala, other forms of freehold (including allowing land to be held by institutions such as churches or hospitals) prevailed, and much of Nakawa Division (the eastern quarter of the city) was public land held by the state (West 1972).

Amin’s expulsion of Uganda’s Asian community in 1972 radically altered urban land holding. Muench (1978) calculated that because of high transaction costs of legal sales as a result of the breakdown of fundamental institutions of law and order during the Amin era, 80% of the urban land transactions in Kampala during those years (1971-79) were “illegal.” These informal or illegal land transactions were cheaper initially, but led to numerous disputes in the long run. In an effort, ostensibly, to promote agricultural development, the Amin regime abolished (at least on paper) the Mailo system in 1975 through the Land Reform Decree, making owners into holders of long-term leases, and making kibanja tenants into sublessees at sufferance (Barrows and Kisamba-Mugerwa 1989). Kibanja simply means a plot or piece of ground; bibanja is the plural. The Land Reform Decree has never been fully implemented, and confusion over its status, as well as the level of litigation caused by informal land transactions during the period 1971-86 have led to proposals for broad reforms of tenure (Bank of Uganda 1990). With regard to urban land, the current proposals under discussion include converting all urban land to freehold.

Categories of land occupancy for UA vary from formal and legal to a variety of informal and illegal arrangements, and can be briefly outlined as follows:

Owner Occupancy

Private ownership of Mailo land actually prevails on about 45% of Kampala’s land area today (GTZ/DPP 1992). Although formally altered to long-term leaseholds, this land is treated for all intents as private land; and it is bought and sold accordingly. The most common form of agricultural land usage on private land is simply the owner of a plot engaging in some cultivation on his or her own land. If a household or individual both owns the title and occupies the land, it is a fairly sure indication of relative wealth.

Long-Term City Council Leaseholds

Over half the land within the city limits of Kampala is public land, on which an applicant can be allocated a long-term, renewable leasehold. Although few people acquire a leasehold on land for purely agricultural purposes, there is evidence of leased land being farmed. Kampala City Council will not issue a lease if the proposed land use is agricultural, but the evidence is that City Council does not interfere with people farming on leased land, so long as the formal purpose of the lease is residential or commercial.

Renewable Annual Rental Arrangements

Major areas of the eastern part of the city are devoted to City Council housing estates, in which houses are rented to both civil servants and the general public. Most of these houses have small plots, which today are covered in gardens and plantations. Formally, these are annual leases, but, in practice, people who occupied the houses 20 or 30 years ago may still be found there.

Bibanja

Under the Mailo tenure system in Buganda, where Kampala is located, kibanja meant a plot of land granted in a kind of long-term rental of use-rights to a tenant by either a private land owner, or by a chief or official of the Kabaka’s government. After the Land Reform Decree, both Mailo and “Official Mailo” bibanja holders were legally reduced to “tenants at sufferance”: the sufferance of landlords in the first case, and the sufferance of the state in the second. In both cases, bibanja holders are legally referred to as customary tenants, with rights to compensation for improvements to land. In practice, their circumstances vary.

Under the Mailo system, a landlord allowed peasants to settle on land so as to collect busuulu and envujjo rent from them-literally, ground rent and crop rent. In later years under the Mailo system, landlords actually sold long-term use rights to a tenant, and this is the more commonly implied sense of the term kibanja as it is used on Mailo land in Kampala today. In spite of the Land Reform Decree, the system of bibanja-holding of land continues. Much of the privately owned land in the parts of Kampala where Mailo land exists (or existed) is occupied by bibanja holders. Technically, the buying and selling of bibanja is not legal, but the practice continues, and in fact much agricultural land is used under this kind of tenure.

In the more pert-urban parts of Kampala, this has created a dual market in land rights. Titles to Mailo land are often bought and sold both for speculative purposes and to use as collateral to acquire bank loans, and the land owner may never even see the piece of land being bought. At the same time, kibanja rights are bought and sold, often for the purpose of cultivating, although semipermanent housing may also be constructed. The value of land in Kampala suburbs appears to be increasing at three to four times the overall rate of inflation in the economy.

A second category of kibanja rights continues to be widely recognized, but is without current legal status. In Buganda, before the Kabaka was deposed by the Obote government in 1966, much of the land was held in public trust by the Kabaka or chiefs and other notables Although individual incumbents did not get personal title to the land, rent from the land was payment for their chiefly duties, and so chiefs, ministers, and the Kabaka himself had land on which they settled bibanja-holding peasants. When the Kingdoms were abolished, ownership of this land reverted to the state, and most of the land that fell into this category in Kampala is now administered by Kampala City Council. Hence, someone who had acquired a kibanja on this so-called official Mailo land before 1966 may still be occupying such land, and is quite likely to refer to it as “his” or “her” land, even though they have no legal basis for the claim of land ownership. Of course, such people are free to apply to the City Council for a leasehold on the land, but the process is expensive and time-consuming, and few customary tenants have managed to acquire leases.

Borrowing

Land borrowing is a very common form of land access for agricultural purposes. It offers access to land with the consent of the owner or caretaker, and the assurance that even if future use-rights are withdrawn, the labour invested in a given year’s crops will not be lost through summary eviction. In some cases, a small amount of money is paid to the owner or caretaker; more commonly, some of the food harvested is given as a token of thanks or “rent.” Occasionally, the use of such land appears to be completely free, although this is mainly among close friends or relatives.

Lending land serves the interests of land owners in that there are specific short-term agreements with users-albeit often verbal- and the clear understanding that when the owner wants the land, either to sell or to build on, the users will be asked to leave. In the meantime, lending the land prevents more permanent “squatters” from taking up residence, whose removal might either be financially costly or involve litigation. Owners sometimes put caretakers on the land, and expect them to earn some portion of their living from looking after the land-hence the token payments of “rent.”

Purchasing “Use Rights”

There remain in the city large tracts of land, both public and private, that are not built up, and which the owner has reasons for not wanting to sell. In the case of Kampala City Council land, Council lacks the means to care for this land, and tolerates people’s informal use of it for agricultural purposes. Some private land owners also tolerate agricultural land users, so long as they keep their gardens weeded. In areas where such tolerance has long been practiced, an informal land market in buying and selling use-rights has emerged, but it has emerged between users, not between owners and users. Users refer to the plots as “my land” even though they know that the land owner can repossess the land, and in most cases, the issue of compensation is an unknown factor in the decision to “buy” such land. However, the prices are low, reflecting the unknown quantity of tenure security. On the other hand, the use and exclusionary rights on such land seem to be quite strong. There are unwritten rules about usage-perennial crops are not permitted, for example -and the existence of an active “land market” may also serve the interest of the owners by preventing any occupant from claiming kibanja rights.

Squatting

Informal occupancy without permission occurs on both public and private land in Kampala. On private land, the Land Reform Decree stipulated that previous bibanja holders became “tenants at sufferance” and hence there is only scant legal distinction between former legal bibanja holders and people who have occupied land without permission of the owner since the Land Reform Decree. Occupants who do have the permission of the owners and who have occupied the land since the 1975 Land Reform Decree are more aptly described in legal terms as “borrowers,” although there is quite often an exchange of money or some other form of rent. The issue of squatting on private land in Kampala is most clearly demonstrated by the issue of compensation upon eviction.

In the perception of the land owners, squatting is a strategy to make short-term use of land as well as to ensure that the squatters “get something” in return for being evicted. The perception of the kibanja holders, or the “squatters” (depending on the background of the individual case) is somewhat different. Several cases were documented in which bibanja holders, whose rights to land-use had been recognized by a previous title owner, were summarily evicted from land after a new owner had purchased the title, without compensation, notice, or even the chance to harvest the crops they had planted. Although the number of cases documented is not sufficient evidence to make generalized claims about relations between “squatters” or informal occupants of land and the land title owners, the evidence would seem to suggest that the relationship has more to do with power than with legality.

On public land, the situation is somewhat different. Here occupants generally know that the land can be reclaimed at any time by City Council, or it can be granted as a leasehold to a private buyer. At the same time, however, prior occupancy bestows use-rights until such time as City Council either repossesses the land or grants it as a leasehold. Neighbours, and in most cases Resistance Councils, will back up the claim and are usually in a position to verify who the prior occupant was.

Land access and household farming categories

In general, commercial agricultural producers operate on land that they own or lease, although several cases were noted where the farming operation was being carried out on the plots of rented houses-but here again, there was a formal rental agreement. It is fairly clear that commercial agriculture does not rely very much on informal access to land. On the other hand, much of the commercial agriculture in the city involves livestock and poultry production in confinement, and physical space requirements are fairly small. This would not be the case, however, for coffee or vegetable producers.

Households in the category “self-sufficient” rarely own their land. It requires a sizable amount of land to achieve relative food self-sufficiency, and if a family owned and occupied that much land, it would be fairly wealthy, and probably would not need to farm for a living. Such households tend to be well connected to land owners and either have kibanja rights dating from before 1975 or have long-term borrowing arrangements that amount to actual bibanja in terms of the relationship between owner and user. Many are also making use of public land.

The “measure of food security” group showed two different tendencies. In one group, the family or an individual in the family may own land (either by holding a title or a leasehold) and another member of the household farms it. The modal example here would be the husband, or the father of the husband, owning land, and the wife farming the land. However, other combinations of the provision of land and labour were noted.

The second group does not own land in the formal sense, but an individual within the household obtains land through a borrowing or “squatting” arrangement, or through purchase of use rights from someone else. Again, the modal example would be either the wife or a female head of household being the one to obtain land in one of these manners, but other cases exist.

The “no-other-means” group tended to rely either on “squatting” as a land-access strategy-including the use of land from which they were almost certain to be evicted-or used tiny remnants of bibanja holdings with which they were left when widowed or abandoned. In several cases, widows reported having been forced to sell pieces of kibanja either before their husband’s death because he was sick and could not work, or immediately after his death because they had no other way of raising money for the family. In other cases, a male head of household sold off most of a kibanja before pocketing the proceeds and leaving.

Formalization of tenure: prospects for urban agriculture

Given the confusing array of land-tenure arrangements within the city, and in particular the overlapping rights of various parties in Mailo land, urban planners have long been concerned about unplanned subdivision and fragmentation of land holdings, and with how to make sufficient land available and acquirable for urban development (Litherland 1966; KCC 1972, 1990). In Mmengo, before its incorporation into Kampala in 1968, the concern goes back even further (Gutkind 1960). In 1965, Kampala City Council recommended that broad-ranging powers be granted to ensure that sufficient land was available for “development purposes”-that is, commerce, industry, and housing (Litherland 1966, p. 21): “In order to implement any rational schemes for urban development, there must be adequate means to completely change the present land [tenure] pattern. This will entail considerable interference with the existing rights of property owners and tenants.”

The 1972 Kampala Development Plan reiterated many of these concerns, and proposed a land-exchange policy to move land out of the Mailo tenure system and into the public land category, under which owners would have been offered 199-year leases on a portion of their land that was of equal value to that of their interest in the Mailo land.

Current proposed land legislation affecting the entire country proposes that urban land be converted to freehold tenure. The presumption underlying the proposals for tenure reform (both rural and urban) in Uganda is that the ambiguity over property rights is a fetter on both urban development and rural agricultural productivity. Formalization of the rules (and the ability to enforce them) is presumed to be a necessary, though not necessarily sufficient, condition for future economic growth.

The consequences of tenure formalization for UA in Kampala will undoubtedly entail a loss of land for cultivation. It is clear from the above discussion of land access that urban cultivators have taken advantage of interstitial institutional space created in urban land both by the complexities of the Mailo tenure system and by the turmoil of the Amin and post-Amin eras. Some of this 1055 of land is already occurring: many of the access strategies of current cultivators-that is, access strategies that have worked at some time in the past-are no longer available to newcomers to the city or to current noncultivators. The perception was expressed repeatedly by interview respondents that “land is all used up.”

As the security situation has improved dramatically in the city in the late 1980s and early 1990s, land owners have begun to build rapidly on land that previously sat idle.

It is not clear yet whether the rate of increase in land values will follow a generally decreasing rate of inflation. If it does, holding land purely for speculative purposes is likely to decrease as the cost of doing so goes up. On the other hand, although economic indicators are generally improving for Kampala’s economy, formal employment has shrunk and wage remuneration has not kept pace with inflation, making reliance on economic activities such as UA more salient than ever. This situation presents several policy options regarding UA in Kampala.

A major conversion from the current variety of tenure types in Kampala to freehold will be an expensive and administratively complicated process. Strong vested interests in the Mailo system remain at many levels of government: so far, the conversion of tenure is only a proposal and a controversial one at that. At issue is whether the City Council is capable of enforcing the Town and Country Planning Act, which would be the sole instrument controlling urban development under a freehold system.

There are influential interests, planners and political leaders included, who see some value in retaining UA as a part of the city’s economy in the long term. This might entail rezoning certain areas of the city for specifically agricultural uses (on the green-belt model); or simply alter existing bylaws to permit farming in certain parts of the city - most notably in the residential suburbs and the more pert-urban areas where current bylaws make little sense and have little impact. Rezoning to create agricultural land-use areas might make sense in terms of the city’s topography; however, attempting to rezone informal economic activities is a notoriously difficult task.

Another approach might involve a two-step process, leaving the long-term use of land to be decided upon at a future date, but recognizing the current squeeze of urban residents. Urban farming could be granted short-term legitimacy in its current form, while issues of land-use planning, rezoning, and compensation, and a review of municipal bylaws could be undertaken at a specified time in the future. The evidence on the food security and nutritional status of households that have had some access to land in the city for farming should strengthen the case to be made for such a policy option.