One of the outstanding features of Africa's recent past has been its transformation from an overwhelmingly rural continent to one that has become increasingly urbanized.1 This has had diverse sociocultural, economic, and political consequences. Social and economic differentiation, in particular, has had a significant impact on the organization and administration of urban Africa.2 In the postcolonial period tidy-minded elite cultures emerged that supported and shaped the modernizing and developmental ideals of municipal governments.3 The manner in which these ideals were frequently out of touch with social realities was ignored or concealed, as was the fact that the adoption of such policies often resulted in the marginalization of large sections of the urban populations.4 A key aspect of the problem was the growth of an informal sector, towards which urban Africa's administrators and emergent elites formed a distinct antipathy. Early theoretical literature on the informal sector tended to be economistic in its focus.5 Here I seek to explore the socioculturel and ideological place of the informal sector in a rapidly expanding African urban center.6 Informal sector activity in Dar es Salaam posed a challenge to a municipal order that administrators and an urban "middle class" were anxious to impose. Rather, participation in the informal sector represented what de Certeau celebrates as the "proliferating illegitimacy" of everyday urban life, in which the tidy developmental vision of planners and administrators unraveled.7 Moreover, by implicitly contesting official policy, such activity represented a form of "noncompliance."8 In response, the state resorted to campaigns targeting the urban poor. However, these haphazard purges merely represented pyrrhic victories in an ongoing struggle over the legitimate occupation and utilization of urban space.
In contrast to the contemporary city, in which the informal is in the ascendant (from unlicensed trade to unplanned suburbs), Dar es Salaam's "parallel" economy in the 1960s and 1970s was small. Aided by a healthier formal sector (at least up to the late-1970s), and by TANU's ideological strength as nationalist and socialist vanguard, the postcolonial state achieved some success in its suppression, and the developmental ideal of a modern, planned city whose working population was restricted to those engaged in waged employment, was sustained into the 1980s. However, the official view of the city- informed not only by modernization theory, but also well entrenched historical discourses on the negative consequences of urbanization in African and earlier Western society9represented a simplification. Tt attempted to restrict a complex and unpredictable social organism to the role of planned capital in a developmental state.10 This official view may have had some ideological success in delegitimizing the urban presence of the un- and under-employed. However, the ultimate goal of an "orderly" metropolis proved unsustainable in the face of a demographic revolution that resulted in the unprecedented growth of an informally employed, housed and serviced urban population. Moreover, by the mid-1980s, with increasing numbers of Dar es Salaam's formally employed residents engaging in secondary economic activities to supplement shrinking salaries, the political necessity of a more liberal approach towards the informal sector was clear.11
Rapid Urbanization in Dar es Salaam
In the mid-1950s the British colonial administration in Dar es Salaam experienced diverse pressures. The most overt was that represented by a burgeoning nationalist movement whose growing support in the capital was evidenced by the large crowds attending TANU rallies. However, perhaps the most pressing-certainly the longest standing-concern of colonial administrators was not that represented by the political aspirations drawing the masses to the rallies, but their simple presence in Dar es Salaam in the first place. The urban population had grown at a prodigious rate over the preceding decade or so. Massive rural-urban migration was overwhelming the colonial ideal of an orderly town.12 Raids targeting the apprehension of so-called "undesirables" and their removal from the town were stepped up, becoming, by 1958, a daily occurrence. Provoked by the raids, and emboldened by nationalist rhetoric, those on the margins of urban society displayed a growing disrespect for colonial law and order.13 In TANU they saw the hope of release from a colonial regime that in Dar es Salaam had become ever more coercive in the course of the decade. TANU politicians occasionally articulated these concerns.14 As it turned out, however, town purges represented an enduring feature of Dar es Salaam life between the 1950s and 1980s, as urban policy after independence bore distinct similarities to that implemented by the British.
In discussing continuities of urban policy in the colonial and postcolonial period it is important to set policy decisions made at the time in their context. The most noteworthy aspect of Dar es Salaam during these decades was the town's inexorable expansion. It is this background feature, above all, to which continuities in urban policy may primarily be attributed. In 1948 the total urban population was 69,200; by 1978 it had grown to 769,445.15 Between 1948 and 1967 the town grew at an average annual rate of 7.5 percent, which meant that in 1950 approximately 5,500 people either moved to or were born in the Tanganyikan capital (see Figure 1); by 1967 this figure approached 20.000.16 Over the next ten years urban growth actually accelerated (to 9.8 percent),17 and by 1978 the annual increase had reached almost 70,000.
Urban expansion is even more striking when placed alongside rates of increase in employment and housing construction. Between 1962-78, when the urban population grew by almost 600,000, just 117,657 enumerated jobs were created in the town's formal sector. Booms in particular industries, such as occurred in the construction industry in 1973, could occasionally result in the number of jobs created matching (or even exceeding) the expansion of population in any one year, but a substantial shortfall was more usual (see Figure 1). In some years employment actually contracted. By 1978, about 470,000 people for whom formal work was unavailable had found their way to Dar es Salaam. While a substantial proportion may have been women or dependants for whom employment may not have been necessary, many were not.18 Without an increase in formal employment these extra people could only be fed, clothed, and housed through a substantial increase in informal economic activity. Statistics on Dar es Salaam's informal sector are sparse for the period under consideration. However, evidence from the final decade of colonial rule indicates a substantial increase in petty trading and other such activities at a time when rapid urbanization had begun and urban unemployment was mounting.19 This trend continued into the 1960s and 1970s, by which time informal operators such as the mama ntilie and wamagenge (food vendors)-both of whom operated cooked food stalls-had become well established urban institutions.20
Just as the formal economy was swamped by rapid urbanization, so too was the town's planned housing. Between 1948 and 1962, when Dar es Salaam's population grew by 121,000, only 4,839 houses were built by the government.24 New areas were also surveyed where Africans could build their own houses. However, the number of surveyed plots and houses were outstripped by rapid urban growth. In the post-independence period the position deteriorated further. With the population growing at an annual rate of between 13,000 and 39,000 in the decade 1962-72, an average of 821 houses were built annually by the National Housing Corporation in Dar es Salaam; between 1973 and 1979 this declined to just 437.25 The TANU government continued to provide land for planned development by individuals, and in the 1970s "sites and services" schemes were implemented. However, as in the colonial period, even if such land was accessible to impoverished urban immigrants, the number of available plots was once again dwarfed by the expansion of the urban population.26 With planned land and houses in short supply, Dar es Salaam's inhabitants increasingly turned-in local parlance-to "building in a random fashion" (kujenga ovyo ovyo).21 "Shanty" settlements in Dar es Salaam emerged for the first time in the late 1940s at Makaburi and Chang'ombe, and at Mikoroshoni in the 1950s.28 Meanwhile, unplanned urban "village" communities-notably Buguruni-were incorporated as the town expanded. In 1960 there were an estimated 5,000 "squatter" houses in Dar es Salaam.29 As urban growth accelerated, unplanned housing grew apace (see Figure 2). By 1979 there were 43,501 unplanned houses, at which point informal settlements provided a home to a majority of the urban population478,489 out of 769,445.
Urban Policy and Social Differentiation
Up to WWII, officials had devoted little attention to incipient urbanization. However, in response to rapid post-war urban growth a mature colonial policy emerged, the ultimate goal of which was a planned city containing a stabilized, respectable, African working class enjoying modern infrastructure and amenities.30 It was not long before such plans went awry. However, officials did not discard their neat urban vision. Instead they struggled to maintain its coherence by vainly attempting to remove those urban elements whose presence contradicted their hopes for a model town. The spiraling population simply fuelled anxieties about a collapse of civic order. Accelerating urbanization after independence placed African administrators under even greater pressure. Officials were only too aware of the massive discrepancy between urban growth and the expansion of employment and housing. With little sympathy towards peoples' self-help survival strategies, TANU officials, like the British before them, resorted to the attempted removal of the urban un- and under-employed. While the new regime may have displayed greater empathy with the plight of the urban poor, officials remained committed to the colonial vision of the city, and vainly attempted its realization through coercion.
The continued commitment to this urban vision reflected TANU's acceptance of the colonial program of modernization.31 Continuity in urban policy represented a successful outcome for one aspect of late colonial policy. An aim of urban labor stabilization in the 1950s had been the creation of a restricted urban workforce, receiving substantially higher wages, who would not only provide a revenue base for the development of the town, but would also produce a class of Africans with a vested interest in socioeconomic and political stability. This privileged class formed the most affluent and influential section of the urban population after independence. An International Labor Organization (ILO) report in 1967 observed a shrinking, but better remunerated, formal workforce post1961. It noted that while generally Tanzanian living standards had remained stagnant since independence, wages had risen by over 80 percent and "wage-earners were 65 percent better off in real terms than they were in 1962."32 Rising wages were in part paid for by increased productivity and by mechanization. However, the main cost was borne by smallholders due to "a change in the terms of trade between the industrial and rural sectors."33 Gross domestic product per capita in urban districts throughout Tanzania was significantly higher than the surrounding regions in which they were located. In Dar es Salaam this was particularly marked, where the GDP of the urban area in the early 1970s was TShs.4,152/- and that for Coast Region just TShs.427/-.34 According to annual surveys of employment, wages earned in Dar es Salaam formed on average almost 35 percent of the national wage bill between 1966 and 1978. Moreover, those people in enumerated employment receiving these wages constituted just 20 percent of the town's population.35 The 1967 ILO mission highlighted not only a widening gap between workers and smallholders, but also "considerable inequalities between employees themselves."36 Manual laborers (skilled and unskilled) received poorer returns than those in white-collar employment, as did those engaged by smaller employers. "[A] group of people are emerging," it observed, "who are fortunate enough to be employed by a handful of big monopolistic concerns and public services, to whom most of the benefits of recent economic development appear to have gone."37
Other aspects of policy served to further the interests of an emerging urban elite. In the case of housing, the colonial African Urban Housing Loan Scheme was replaced by postcolonial mortgaging systems both of which directly benefited the well-connected.38 A "Revolving Loan Fund," established in 1963 for senior government employees, had by 1968 disbursed TShs. 16.3 billion to just 230 borrowers. In 1974, the average loan of the Tanzania Housing Bank (which provided loans to individuals from TShs. 1,000/- to 40,000/-) was TShs.32,300/-, three times the amount required to build a six-room house in the "Swahili" style favored by the bulk of the urban population. Meager public resources, noted Bienefeld in 1970, were used "to provide housing for those in a position to provide themselves with housing through the market."39 Access to such resources enabled a minority to gain a foothold in some of Dar es Salaam's more exclusive areas. Kironde discusses the acquisition of property by well-connected Africans, first in Magomeni (as beneficiaries of late colonial housing policy), and then after independence in what had been predominantly European suburbs such as Oyster Bay. Mascarenhas dates the move to Oyster Bay even earlier, to 1959, when the first African ministers were appointed to the government.40 He observes a separation between African leaders and the masses. A view shared by a correspondent to the Daily News in 1973: "Many areas continue to be called Uzunguni [the popular name for the predominantly European areas in the colonial period] ... connoting a 'disgusting' sense of superiority. Most Tanzanians still think that people who live there are superhuman and they fear to pass through such areas (even though most residents are now Africans)."41
The Party Line and Its Critics
Having secured a house in a planned neighborhood and remunerative employment in the formal sector, an individual who had benefited from late colonial urban policy was inclined to pull up the drawbridge.42 The expansion of shanties represented a threat to public health, to property, and to their peace of mind. Repatriation campaigns in the postcolonial era were organized by and on behalf of this class. Mutasingwa observed how in Magomeni in the mid-1970s "party cadres and peoples' militia" enforced the absence of "lumpen-proletarian" elements in the area. Before a person's identity card was "considered valid by the local authorities" they had to prove their "monthly income allows him/her to meet the Magomeni standard of living." As a result:
Small-time businesses such as hawking groundnuts, peddling cigarettes, selling ice-water, is strictly forbidden and local mg'ambos (peoples' militia) are always on the look out for those small-time business men. The local government, and the police station adjacent to it, act as constant reminders to the unwanted guests of Magomeni to stay clear of the area, or else....43
Politically correct arguments were used to disguise self-interested action against the urban poor. In justifying urban purges, paternalistic colonial administrators could frankly state that the unemployed were "undesirables" who were either vulnerable to corruption or already up to no good. This provided sufficient motive for an individual's forced removal from the town. African officials after independence could not afford to be so cavalier. While they continued to rail against the consequences of unemployment, joblessness was as likely to elicit sympathy as hostility among the urban population. Moreover, officials could not be seen to be simply shifting an urban problem to the countryside, as had occurred in the 1950s, when large numbers of Africans were repatriated to the rural areas without any attempt to alter the conditions that played a part in their move to town. In addition, privileged employees in Dar es Salaam's formal sector-whose support was crucial in legitimizing TANU policy-displayed a distinct ambivalence towards informal economic activity. Those engaged in the informal sector were often condemned by their more selfconsciously respectable fellow townsmen for their slothfulness, unsightliness, or their perceived criminality. At the same time, however-as we shall see in the case of the magenge food vendors and the shanty village of Kisutu44-the informal sector provided goods and services to all sections of the urban population. Therefore action taken against the urban poor required not only a greater sin than joblessness, but also a policy to ameliorate the problem-this in order to recommend itself both to the general population and to officials themselves, whose nation building philosophy, while often self-interested, was just as often firmly held.
The ideological apparatus that accompanied this process of nation building provided a new discourse with which to stigmatize the urban poor. At a time when the nation was, to use a term from neighboring Kenya, "pulling together," to reside in a town without formal employment could be portrayed as irresponsible. Not only were urban idlers failing to contribute to national goals; they were also living off the backs of hard working rural peasants. So colonial "undesirables" became in the post-independence period "unproductive": parasitic elements undermining efforts at national development.45 Such attitudes arose from local suspicion towards urbanization. They were reinforced by "African socialist" ideology. The urban un- and under-employed were no more in the vanguard of Ujamaa, than the Victorian lumpenproletariat were in the communist program of Marx and Engels. The question was, as a Daily News byline put it in 1974, "how street hawkers can fit in a socialist Tanzania?"46
The rhetoric employed against the urban poor found many forms of expression. Repeatedly, and from early on, the laziness of those surviving in the city without formal employment was singled out. In October 1964, for example, the streets of the capital were "combed for unemployed people ... idling their time away." Later the same day, according to the Nationalist, "[n]o one was seen playing 'bao,' chatting uselessly or playing cards in the streets." 47 Three years later, women in the town were condemned for their sloth. After "Debe men" (water-carriers) had been rounded up in another campaign against "idlers," the new Regional Commissioner, Mustafa Songambele, observed that "[i]t would give the city women a chance to go and draw their own water instead of remaining at home."48 In 1970 Dar es Salaam Area Commissioner A.N. Lyander observed that "[wjandering in towns was not a good thing," and recommended a return to the land for those without formal employment. To encourage this he announced that "surprise round-ups of the unemployed" were to "continue forever."49 Eternal vigilance was just what was required to combat people later characterized as congenitally slothful:
There is an induspitable fact that there are people who have been taken to laziness from their birth. They are, so to speak, idlers by nature. Laziness is part of their traditional set up-a characteristic which is deeply embedded in a number of tribes in developing countries.50
Indolence was not an isolated evil, it was associated with or responsible for other forms of disreputable behavior. In 1964 the Commissioner of Coast Region, Commissioner Kitundu, observed that "[c]rimes of all sorts originate from idling."51 Similarly, in lamenting the ineffectiveness of periodic urban purges the following decade, Julius Nyerere wrote of "criminals and idle parasites" hiding "in their houses to evade urban raids."52 The link between parasitism and sloth was also frequently made. One urban campaign conducted in 1973 was actually named after the Swahili word for a parasitic tick: Operation Kupe. Under section 176 of the penal code, idleness and disorderliness were actually linked-although this was a hold-over from the colonial period, it had revealingly not been changed.53
Indeed, alongside idleness, it was the disorderliness of the urban poor that was most commonly invoked to justify action taken against them. The growth of shanties and proliferation of itinerant traders were a blot on the landscape, contaminating officials' ideal vision of the modern urban center. In some respects the emergence of such a town was indeed occurring. A correspondent for the Weekly News described a changing Dar es Salaam in 1967:
New buildings ... are transforming what used to be little more than a shanty town, into a smart and modern capital. Wide dual carriage-ways are being driven through an area which was formerly a sprawl of teetering hovels built of cans, mud wattle and hope. On the outskirts of the city, spruce new housing estates are going up... .54
However, while in some areas they may have been replaced by gleaming modernist offices and metalled roads, to the chagrin of many officials "teetering hovels" were mushrooming elsewhere. Simultaneously, a burgeoning informal economy supposedly brought chaos to city streets.
Action against these manifestations of urban poverty was often discussed in terms of purification. In 1967, the Weekly News referred to a "city purge."55 "Uchafu wote utafagiliwa" (All the dirt will be swept away) announced the Swahili paper Ngurumo, the following year, in connection with "Operation Vijana," a campaign aimed at disorderly urban youth.56 In 1973, dawn raids were described as "part of a campaign to clean the city of criminals, vagabonds and unemployed."57 A year on, vendors of handicrafts and foodstuffs were ordered to quit the streets "in order," according to llala District Development Director A.M. Tesha, "to maintain the high standard of city cleanliness." "The measure has been taken," announced the Daily News, "with the aim of keeping the city clean, especially during the forthcoming celebrations of the 20"1 Anniversary of TANU."58 Urban order was essential at such an auspicious moment. Threats to health were also often invoked. Beggars with "severe diseases," complained a correspondent to the Daily News in 1974, were "roaming about spreading it [sic] everywhere in the city."59 Public health concerns were used to justify action against petty traders, as attempts were made to enforce unrealistic standards of public hygiene that punished the poor (for whom informal food retailing provided both a source of cheap food and a potential income) more than the threat of disease.60
Condemnation of the moral and hygienic shortcomings of the urban poor was all very well, but for a progressive African socialist government it was also necessary to put forward policies of improvement. In 1970, after action was announced against some "15,000 jobless people who spent nights in abandoned fuel drums in the capital," a Regional Office spokesmen claimed the move to "seek their departure from Dar es Salaam was not designed to torture them." It was motivated instead by their "unnecessary suffering" arising from a lack of "fixed abodes ... [and] proper food" as well as the fact that through "joining other Tanzanians in building the nation through ujamaa villages in their home districts, they can boost not only their living conditions but also the country's foreign currency earning."61 The need to re-educate repatriatees was often stressed. In 1971 the Minister of Health and Social Welfare, in announcing campaigns against indigency, argued that most of the beggars were "in fact able, fit and healthy citizens who could be made to become self-reliant."62 For many commentators, the lack of action in this direction was one of the problems with the urban campaigns. A Daily News editorial on Operation Kupe, while agreeing that the presence of "unproductive" elements was to be deplored, argued that the reasons for these "mostly highly schooled" migrants coming to the city had to be understood-they were seeking to improve themselves. Operation Kupe, the editorial concluded:
cannot merely be a question of "catching" the jobless youths and "sending them" to the rural areas ... It has to be a question of political mobilisation and political education, it has to be a question of what the youths are going to do and where....63
Action taken against the "idle and disorderly" frequently received public endorsement, for all of the reasons discussed above. A Nationalist editorial in 1964, for example, declared round-ups "a step in the right direction" which promised to be of "immense benefit to the state":
First, apart from removing loafers from the big towns, the process will, without doubt, reduce the crime wave which these unemployed people create when they fail to find the El Dorado which they think exists in the towns and cities. second, this new step should discourage many others in the rural areas from being attracted to the bright lights of the big towns.64
Regional Commissioner Songambele "and all those involved in the clearing of these parasites" were singled out in a 1971 letter to the Daily News for "high praise and credit."65 In September 1983, government rhetoric was echoed in comments by white collar workers interviewed by the Sunday News. The presence of a "class" of urban loiterers-as articulated in TANU propaganda-was accepted unproblematically by most of those interviewed. New legislation aimed at their removal was discussed in abstract terms of what was good for the country and its "development."66
The policy of urban purges also had its critics, who took the ideological high ground. Naijuka Kasiwaki, in the Standard, in 1970, condemned action taken against "the scapegoats of society" as "colonialist and unsocialist."67 Similarly, a "tersely worded" Radio Tanzania commentary the following year "dubbed all those persons in positions of authority who initiate and carry out such operations anti-TANU and anti-socialist elements." The author of an editorial in the Daily News concurred. He had hoped that "this barbaric method of 'repatriating' people had been buried once and for all."68 In a suspension of action against the unemployed, which occurred prior to elections in 1971, he detected political motives. Indeed, he hinted that the negative radio commentary could also have been broadcast for electoral purposes. "Since the election is over," he continued,
and its results are favourable too, the campaign to rid the city of the poor must resume so as to let the Wabenzi69 lead their flamboyant lives without being bothered by beggars, thieves, loiterers, prostitutes (excluding the "birds"70 of the Wabenzi), etc.
The same writer pointed out the absurdity of sending the unemployed to rural areas when no adequate structures were in place to keep them there. Many letters supported this analysis. If the criminal tendency of the unemployed was a justification for repatriations, observed one, then it was surely counter-productive to be sending potential criminals to disturb the peace of areas lacking police posts.71 In a particularly scathing critique of Operation Kupe, Philip Ochieng, a Daily News columnist, observed that the "parasites" singled out in the raids were the wrong ones:
We are told kupe are the hundreds of young men and women who revile our holy city by doing such sinful things as looking for jobs.
Why is it wrong for any intruder to look for a job in our city? Because what we have is ours and we can't share it with any urbonauts from the Planet Rural.72
...[W]ho in our society is "kupe" and who is cattle ... to see this is, in a significant measure, to anticipate the answer. The answer is that we really need Operation Kupe!
Kupe are quite aware of this necessity. That is why they are trying to divert our attention from the real problems by saying that the cattle that bear the burden of our growth are the kupe.73
Rural Development and Urban Purges
Accusations that the TANU administration evolved no policy to deal with individuals apprehended in the urban campaigns are somewhat harsh. Colonial repatriation campaigns had been accompanied by no substantive schemes to encourage people to stay in the rural areas. African politicians, by contrast, lost no time in planning rural development projects. Just one month after achieving internal self-rule they announced a number of schemes aimed at drawing surplus labor away from the capital. A land clearing and resettlement scheme in the Kilombero valley to which 500 urban unemployed were sent in early 1960 was the centerpiece scheme,74 which according to the minutes of a cabinet meeting aimed at "offering counter-attractions to the excitements which led people to endure the discomforts of unemployment in large towns."75 Alongside activities in the Kilombero valley, in 1962 a 3,000 acre scheme designed for 200 unemployed was opened by the Regional Commissioner,76 and another a resettlement scheme was announced at Mapinga the following year.77
Measures taken upcountry to settle numbers of urban unemployed were in line with the emphasis on rural development that emerged in Tanzania, particularly after the Arusha Declaration. In the Declaration, Nyerere condemned the fact that whilst the bulk of development loans would be used in the urban areas, the repayment of those same loans would occur thanks to the efforts of rural farmers. In its wake a pro-rural strategy was adopted that, it was hoped, would improve rural conditions and discourage urban migration. One important means of redressing the rural-urban imbalance was to reverse the trend since independence-noted in the 1967 ILO report-of wage earners benefiting from higher pay at the expense of farmers. Moves to limit wage increases were included in the Second Five Year Plan, 1969-74. Changes in schooling at this time also reflected the pro-rural strategy: the policy of education for self-reliance explicitly aimed at equipping young Tanzanians with the skills required of peasant farmers rather than wage earners.78 Speaking to an assembly of primary school teachers and pupils in Nachingwea in 1974, Prime Minister Kawawa stressed "[i]t was the duty of every young man and woman to be actively engaged in agricultural activities."79 Moreover, according to Armstrong, "minimal resource allocations of public investment for urban infrastructure were made during the 1970s, further reinforced by the rural bias of the major and substantial foreign aid investments."80
The political commitment to attempt to shift resources away from the urban areas, and in particular Dar es Salaam,81 did little to stem rural-urban flow. Like the urban raids themselves, rural initiatives were sporadic and, when they did occur, inadequate. Resettlement schemes accompanied urban control campaigns in 1967, when regional authorities responsible for the villagization program established communities for the urban unemployed.82 Two years later, Nyerere stressed the need to improve conditions in the countryside, and the Social Welfare Division began educational programs in Coast Region to discourage rural-urban migration.83 In the same year a temporary destitutes camp was established at Kipawa (presumably to augment the one established there by the colonial government84), and in 1970 a rehabilitation center for destitutes was opened twelve miles from the capital, at Yombo.85 These camps catered only for the disabled poor, however. While repatriation campaigns occurred between 1969 and 1975, it appears that little was done to make the rural areas more attractive for either able-bodied repatriatees or potential migrants. Operation Kupe, the biggest campaign of the period, attracted much negative comment from both press and public for this very reason.
Perhaps officials were chastened by this criticism. The next urban campaign, in 1976, Operation kila mtu afanye kazi (everyone must work), or Operation Rwegasira as it was more popularly known (after the then Regional Commissioner), was accompanied by the most ambitious plans yet, with ten million shillings set aside for resettlement. It formed a part of the national villagization program, Operation Tanzania, which included a pilot program to resettle 15,000 unemployed from the capital.86 In due course, thirty Ujamaa villages were established in the rural parts of llala, Temeke, and Kinondoni districts.87 Eleven thousand jobless residents were identified by September 1976. However, two months later fewer than 150 people had voluntarily moved to the villages. Compulsion was then employed, on the orders of Commissioner Rwegasira. One thousand people were to be moved at 10-day intervals. The resulting raids indiscriminately rounded-up people from Dar es Salaam, who were deposited in the villages. However, the organization extended only to the establishment of the settlements and to the raids. "After moving them out," Rwegasira himself complained, "very rarely did the initiators of the whole thing interest themselves with what the people starting a new life in their respective villages were actually doing."88 Within days the majority drifted back to town.89
Despite the dismal results of the Rwegasira campaign, few lessons were learned. Officials proceeded to repeat the same errors in the final, and most notorious, large-scale campaign: Nguvu Kazi, conducted in 1983-84. Once again land was provided: 6,500 acres were allocated in the vicinity of the city solely for unemployed persons who originated from Dar es Salaam.90 As before, little thought went into the needs of the people moved there. Lack of transportation and water made farming untenable.91 Unemployed persons who originally came from outside of Dar es Salaam met a different fate. For "upcountry" Africans, officials turned to the time honored tactic of carting repatriatees back to either their home district (if they could pay for transport) or to nationalized sisal plantations (at the government's expense).92 However, repatriatees soon made their way back to the city. Of those removed in October and November 1983, only ten percent reached their assigned destinations.93
Explaining Urban Purges
Campaigns against the urban poor can be interpreted as having overtly political origins. Prior to independence, repatriation campaigns were intensified after outbreaks of urban violence had posed a challenge to the colonial administration. Large-scale raids occurred in the wake of the violence associated with the strikes of 1947 and 1950. Similarly, campaigns were stepped up from the mid-1950s, a time when more minor but frequent outbreaks of violence in the town occurred as British authority was undermined by nationalist agitation. Unsurprisingly perhaps, the largest operation to occur took place after the 1959 Buguruni riot, a particularly serious outbreak of violence that resulted in the death of a policeman.94 In the postcolonial period too there is evidence that raids occurred in response to events that alerted politicians to the potential threat posed by the urban poor. After just two months in office as Chief Minister, Nyerere voiced anxiety over "the political dangers inherent in the number of unemployed in Dar es Salaam who were easily susceptible to the suggestion that the new Government was doing nothing for them."95 Such fears appeared to be confirmed by rioting that occurred in the wake of the army mutiny in 1964, and it is no coincidence that the first concerted urban purge occurred shortly afterwards. While later campaigns do not appear to have been preceded by any serious violence in the town, the presence of growing numbers of un- and under-employed no doubt concentrated the minds of urban officials. Moreover, Nguvu Kazi happened at a time when not only the national economy was descending into crisis, but the Tanzanian state had also just been rocked by an unsuccessful coup attempt and by the hijacking of an Air Tanzania plane (one of the demands of the hijackers being the resignation of a government accused of economic mismanagement).96
Highlighting political roots for the purges, however, lends them a rationality they do not deserve. Given the striking lack of success that accompanied the campaigns it is difficult to account for the stubborn persistence of both colonial and postcolonial officials in adopting tactics that had little or no impact on urban problems as they defined them. The raids were a disaster in every conceivable way. They did nothing to arrest rural-urban migration. Both unemployment and crime rates in the town increased despite attempts to remove both criminals and "idlers." A growing percentage of the urban population lived in unplanned housing and subsisted thanks to a burgeoning informal sector. When people were repatriated, without controls or incentives to make them stay in the rural areas, it was simply a matter of time before most of them drifted back to the capital. Moreover, in the course of campaigns that (both before and after independence), were frequently indiscriminately applied, the colonial and postcolonial administrations risked alienating the urban poor.97
The most obvious explanation for this seemingly irrational action is simply panic in the face of rapid urbanization. Without some form of highly coercive regime restricting people's freedom of movement, attempting to stem the rural-urban flow was like placing the proverbial finger in the dyke. Even if they wished to do so, neither the colonial administration, accountable to the United Nations, nor the TANU government, accountable to the wananchi, could afford to implement such restrictions. In the colonial period British officials repeatedly expressed their desire to introduce some form of pass system.98 Such proposals were knocked back time after time for economic reasons and, more significantly, because of the political implications. Denied the means of controlling rural-urban movement, officials resorted to alternative measures to remove those they deemed had no place in the town. After independence, restrictions on movement reminiscent of the colonial period, or of the White regimes of southern Africa, were unthinkable.99 Systematic daily raids such as those conducted by the British in the late-1950s were also politically unacceptable. Constrained by public opinion, the postcolonial administration could only afford to unleash sporadic reprisals against the "unproductive" poor. Their ferocity when they did occur was an indication of the frustration engendered by a town slipping out of official control. This policy reached its apotheosis in Nguvu Kazi. At the time of its implementation the Tanzanian state was in crisis.100 In Dar es Salaam, the situation was as far beyond the government's control as it had ever been. So it was at this point that the most desperate, haphazard measures were resorted to, effectively criminalizing a large proportion of the urban population.101
The manner in which Nguvu Kazi and other urban campaigns criminalized urban dwellers is deeply ironic, for the panic forming in the minds of officials as the urban population swelled was informed by fears of increasing lawlessness. From the early colonial period it was common practice to associate joblessness with crime.102 In 1926, Tanganyikan Labor Commissioner Orde-Browne complained of Africans drifting into the town:
[where] they find some sort of casual work, but probably fail to get steady employment; intervals of idleness between jobs tend to increase, until the individual drifts gradually into the class of unemployable loafer, from which stage it is fatally easy to join the definitely criminal class.103
The rationale was substantially unchanged in the 1980s. Orde-Browne's "unemployable loafers" were Nyerere's "unproductive loiterers." The assumption was made that, with the growing discrepancy between the numbers coming to Dar es Salaam and the capacity of the urban economy to generate employment, the jobless would inevitably turn to crime. To the extent that many people found work in the informal sector, these fears were well-founded, as the majority of informal economic activities remained illegal. However, the disorderly presence of unlicensed petty traders, prostitutes, or beggars, was not the principal criminal menace. Rather, it was a growing incidence of property crime that was portrayed as the greatest danger of rapid urbanization. These fears appear to have been exaggerated. While urban crime rates did increase in the late colonial period, per capita rates were actually more erratic (see Figure 3). Postcolonial crime statistics are hard to come by. However, judging by evidence from the 1950s, it would be wrong to assume that accelerating urbanization after independence necessarily resulted in a growing tendency towards criminality.104 This was, nevertheless, an assumption that informed policy. Fear of crime motivated the heavy-handed urban purges. At the very least, the frequently raised threat of crime provided a convenient excuse for action against the urban poor. It also helped legitimize raids among crime-fearing "respectable" urban residents, whose complicity-if not active support-was essential in enabling the state to employ such coercive measures.
While property crime may have been uppermost in officials' minds, informal sector operators were also targeted.105 In schemes devised by colonial urban and economic planners, there was no acknowledgment of the role that informal sector activities could play in providing a livelihood. Although social surveys, such as Leslie's in 1956,106 brought to light the existence of a flourishing informal sector, to administrators this represented a loss of control over the urban arena, not the reassertion of government influence over the course of urban development they attempted to achieve. Postcolonial officials inherited this approach. Indeed it was not until long after the informal sector had been christened by Hart in 1970 that more sympathetic policies were adopted in Tanzania, as elsewhere in Africa. In the 1960s and 1970s, as in the colonial period, informal economic activity was interpreted as an indication of backwardness, having no part in government plans for economic development and modernization.107
By engaging in informal economic activities urban dwellers evaded state supervision. Hyden has described a process of de-commodification and the growth of a black economy that resulted from the withdrawal of peasants from official markets in the rural areas of Tanzania during the Ujamaa period.108 Urban petty traders, beer-brewers, smugglers and black-marketeers likewise withdrew from the formal economy (though they often had no choice in the matter). Intolerance of the informal sector can in part be explained by the fiscal consequences of these untaxed pursuits for a cash-strapped administration. According to Maliyamkono and Bagachwa such activities posed "an economic challenge to the official establishment." State legitimacy was "threatened in the sense that the challenge to the official economy makes official policies and distributive channels ineffective by diverting the flow of goods and services and by creating its own rent levels."109 Officials appeared particularly hostile towards those activities that earned their practitioners more than a basic subsistence. For example, a round-up of water-sellers in 1967 targeted a section of the urban population prospering thanks in part to the inadequate service infrastructure provided by the state. In Arusha, in 1974, newspaper vendors were removed from the streets because, according to the vendors testimony, "they could make up to 1,500/- a month."110 As Cooper observes, "the hawker or workshop owner who ignores regulations flaunts a framework of legal and bureaucratic supervision of accumulation that fledgling states are trying anxiously to impose."111
The political repercussions of engagement in the informal economy also aroused official ire. Action taken against the informal sector represented an attempt by the Tanzanian state to assert its own urban order in what Cooper evocatively termed the "struggle for the city."112 This struggle was over the form Dar es Salaam would take: the predictable, disciplined, above all "modern" town envisaged by urban and development planners, or the unpredictable, supposedly disorderly, organic networks of the "shanties." In the 1950s, the new urban policy was motivated by a colonial desire to shape the town's future. Repatriation campaigns occurred in response to accelerating rural-urban migration that was derailing these objectives. The postcolonial state was more ambitious than its colonial predecessor. With the legitimacy derived from its role as nationalist liberator, and endowed with a persuasive ideology of African socialism that supported greater intervention, it could afford to be. The TANU administration retained the colonial goal of an urban population predominantly engaged in formal employment, and this provided one motivation for ongoing urban campaigns. However, the campaigns also formed an integral part of the national development drive. Informal sector operators and the unemployed in Dar es Salaam were the urban equivalent of Goran Hyden's "uncaptured peasantry," and in the course of the campaigns many "unproductive" residents were physically captured. They were captured only temporarily, however, and the heavyhandedness of the purges can also be explained by official frustration over development plans going awry, over the diminishing capacity of the state to determine the course of events. Nguvu Kazi occurred, as has been noted, at a time when the post-Ujamaa state was in almost terminal crisis. In the event, it represented the last throw of the dice by the old regime.113 Simultaneously, an alternative urban order was emerging.
Dar es Salaam's Ujamaa village
While TANU officials, and the more self-consciously respectable elements among the urban population, might have gladly wished it away, the informal sector provided a large proportion of Dar es Salaam's residents-a majority by the 1980s, if not earlier-with incomes and services. It also played an important social role. Medical treatment and spiritual solace was provided by traditional healers; unlicensed buses transported commuters; illegal bars provided a place to escape the daily grind; itinerant musicians offered entertainment. Informal networks emerged that outperformed the state's ineffective attempts at economic and community development. In contradiction to official rhetoric, and in the absence of wealth and job creation resulting from an expansion of the formal sector, such networks played a central part in the lives of Dar es Salaam's inhabitants, rich and poor alike.114 Two examples, the magenge pavement restaurants and the informal settlement of Kisutu, are discussed below.
By the 1970s magenge (sing, genge) stalls had become an important institution for poor urban residents in particular,115 providing an income for their owners and cheap food for their customers. Mesaki noted in 1976 that to enterprising "unemployed women, widows or spinsters, and even married women increasing the family budget" it was an important source of income. "Since one need not be trained," he observed, "nor require large capital many women find the genge a solution to their problems."116 Indeed their commercial potential was such that "homes ha[d] been built on the profits of such a business." Customers were highly appreciative of their activities. "[F]or those of us ... in difficulty making ends meet," observed one Daily News interviewee in 1973, magenge "certainly do provide a solution to the high prices of food."117 Magenge, though, did more than provide just sustenance. The original meaning of the word was "a place or a group of people found together talking, gossiping, playing cards, bao etc."118 "To the poorly paid worker," according to Muhondwa, the genge was "both an eating place and an informal social club. ... the genge operator is the centre of the social setting and confidante for the current gossips." At a genge near the Sunday News office in 1973 a journalist observed the following:
There was a lot of talking going on. Almost everybody was taking part in discussions about the weather, politics, commodity prices, queues, the match between Young Africans and Mereeikli-so many issues.
An elderly woman, known as Mama Asha, the owner of the stall, was busy serving people as she took part in the discussions. She kept everybody laughing-this was part of her business promotion technique....119
Both the social and economic roles played by the magenge were invariably ignored by local politicians, who, decrying the disorderly and unhygienic nature of the food stalls, intermittently pressed for their removal.120 The perceived legitimacy of such campaigns was often bolstered by attempts to provide officially allocated and supervised space for some of the informal food retailers whose stalls were closed.121 However, magenge operators and other traders usually showed a marked reluctance to move-and for good reason. Many returned to their old haunts after finding that the space promised them in other areas was in fact not available.122 Even when alternative trading plots existed, the locations were more often than not poorly sited for potential custom, and traders no doubt resented increased official supervision.123 The most ambitious scheme of this kind was the Mbaruku Project, in which permanent accommodation was built for magenge operators working along Mbaruku Street, whose stalls were earmarked for demolition in 1971. The project formed part of the wider initiatives of the Tanzanian state. While traders could continue to operate their own stalls they were forced to engage in cooperative ventures simultaneously.124 According to Mzee Simba, the TANU chairman at Mbaruku, the whole project was "viewed as an aspect of Ujamaa." "The fact that members will continue to run their own businesses individually," he observed, "is just like what happens in an Ujamaa village where although members live and work together, they still maintain household shambas."ni After opening, Mbaruku performed poorly. Customers complained the food was expensive compared to the informal magenge, while stall owners bemoaned heightened running costs and a serious decline in business.126 The Mbaruku Project was, Lucas concluded, simply another example of "development from above," in which ordinary people's needs were misinterpreted or ignored by Tanzanian officialdom.127
The notorious shanty settlement of Kisutu played a similar role to magenge in meeting both economic and social needs of urban residents. Favorably located on the fringes of the town center, Kisutu was long a thorn in the government's side. In the colonial period its position in Zone II, the predominantly Indian commercial area, led to attempts to eradicate African settlement there.128 However, by the 1950s Kisutu appeared to be thriving as the town's principle red-light area, and a reputed den of criminality.129 After independence its reputation as Dar es Salaam's most infamous center of illicit trade and leisure continued to flourish. By the early-1970s this "kupe's paradise"130 was known ironically as Dodoma (an ironic reference to the planned Tanzanian capital), or as Dar es Salaam's very own Ujamaa village.131 The latter appellation testified to Kisutu's ability to meet peoples' needs in a way that state initiatives frequently failed to. At a time of widespread shortages, revelers in Kisutu, interviewed by Robert Rweyemamu in 1973, observed that "in our special Ujamaa village everything is gettable. Everything you can think of-beer, meat, food, gongo (maziwa wa simba)...."132 "[W]e all welcome you and your friends," exhorted one Kisutu resident in a letter to the Sunday News the following year, "(cjome one come all to do shopping, touring, gambling etc."133 Admittedly, Kisutu was somewhat rough and ready. "UJn 'Dodoma' [i.e., Kisutu]," wrote Rweyemamu, "you find the ugliest, messiest and shoddiest compound you could think of right in the heart of a thriving, fast-developing modern city." This incongruity, however, was perhaps more apparent to some than others. Rweyemamu noted:
[An] impressive aspect of this "micro-city within a city" is the conviviality noticed all round. Men and women mix so freely here in a way that would make one wonder if they are not one massive family. People talk and joke with each other so garrulously that one wonders what makes the public consider Kisutu a home of the frustrated, jobless people and a hideout for crime. All around the place I saw people who seemed to be enjoying heavenly bliss apparently in utter disregard for the glaring poverty in which they were so deeply immersed.... [A] sea of humanity-men, women, and children, ageing people and teenagers, representing virtually all the main tribes of Tanzania.134
For all its conviviality, the social order of such a settlement was not the social order of state hegemony.135 In June 1974 Kisutu was earmarked for demolition by Dar es Salaam Commissioner Rwegasira.136 For respectable townspeople this was long overdue. In a letter to the Daily News Lucas Maziku endorsed the move wholeheartedly. Kisutu's inhabitants, he wrote, "live in terribly dirty huts beyond human habitation which spoil the good view of the city. Hence the fall of Kisutu is exclusively welcome."137 The following month bulldozers and the police moved in. The shanty was flattened and set ablaze (see Fig. 4). "As the 93 huts that made Kisutu a world of its own succumbed," the Daily News reported, "the last batch of its residents rushed here and there in uncertainty." "People were very angry," recalled Saidi Amiri Mwinyi, who worked in Kisutu at the time, "because that was their own area; the place where they could meet their needs cheaply."138 Government had other plans for Kisutu. In place of the insalubrious social networks of the shanty it was hoped that a musical conservatoire, planned for construction by the Ministry of National Culture and Youth, would have an elevating influence on the vicinity.139
Rapid urbanization after WWIl flew in the face of the orderly urban vision of colonial officials. In nurturing an urban elite from the 1950s, however, the British succeeded in forming an influential section of the population sympathetic to this vision. Continuity in urban policy was achieved even while its shortcomings were demonstrated with each addition to the burgeoning shanties. An African elite took the place of colonial officials in a renewed "struggle for the city" enjoined after independence, with the combatants representing the winners and losers from the shift in late colonial urban policy. Legal mechanisms used by the state to punish informal economic activity or to harass inhabitants of unregulated housing "underlined the hegemony of the dominant class, whose orderly ways were made to appear legitimate, while the illegitimacy of lower-class life was rubbed in by the humiliation of police raids."140 An official policy emphasizing rural development sought to keep peasants on the land, while at the same time concealing a continuing allocation of resources that overwhelmingly favored the urban areas, and in particular Dar es Salaam.141 Although the accelerating number of migrants entering the city may not have been conscious of this allocation of resources-nor in most cases were they direct beneficiaries-the concentration of wealth presented formal and informal income-earning opportunities absent in the rural areas. Internal migration was a fundamentally rational choice, despite the opprobrium heaped upon urban immigrants.
As growing numbers found a home and living in the unregulated city, the "modern" city became increasingly marginal to the existence of many urban dwellers.142 However, to "respectable" urban Africans and municipal officials subscribing to the developmental modernization of African society, the informal sector formed an unwelcome reminder of lingering backwardness.143 So one "Honest Citizen," in a letter to the Daily News in 1974, lamented the decline since independence of what "used to be a beautiful and hygienic city." Outside of the upmarket areas such as Upanga, Oyster Bay, and Msasani, urban locations had descended into "rotten and decaying places where the whole nation should feel ashamed indeed."144 Meanwhile, informal sector activities contaminated "proper" work practices. In a letter to the Daily News in 1974, Amon Anasa condemned the work of car washers operating at Mnazi Mmoja as '"childish1 as it can be done by even a child aged 4 and not by 30 year-olds." "This," Anasa concluded, "is one way of cheating on work."145 Elite concern arose over the unmediated influence of modernity on previously subordinate sections of the population. For example, the urban arena presented opportunities for youth and for women to establish autonomous identities-which were frequently adopted with an alacrity and ingenuity beyond the more conformist member of the "petty" or "bureaucratic" bourgeoisie.146 For those in positions of social, political, or economic power these trends were to be resisted, and arguments surrounding both the impact of urbanization147 and the connected issue of the legitimacy of urban residence formed means of attempting to suppress such challenges. These struggles had their origins in the colonial period, a time when generational, gender, and class conflicts were slowly emerging in Dar es Salaam.148 Shifts in late colonial policy exacerbated these conflicts. From being centers of European predominance where Africans had restricted rights of residence, towns became the focus for the modernization of African society. The beneficiaries of this shift in policy became the advocates of modernization.
While there was a marked continuity in policy towards the urban poor, both the quantity and quality of the rhetoric employed to justify "wahuni raids" and their postcolonial equivalent changed significantly. In the 1950s, the little press coverage these campaigns received took the form of blunt denunciations of shiftless undesirables that carried overtones of racial stereotyping. In the postcolonial media, campaigns were preceded and accompanied by numerous official denouncements of the "idle" poor, their position and role in the wider nation. What is more, they prompted a vigorous public debate. The degree to which TANU politicians engaged with the public formed one stark contrast with the colonial period. It reflected both officials' perceived need to persuade Tanzanians of the correctness of their actions, and at the same time demonstrated their confidence that a policy of repatriations had at least some support among the population.149 In their attempts at suasion, officials and their supporters invoked socialism, nationalism, and (paradoxically) tradition. At the same time such rhetoric helped conceal the interests served by such a policy: of municipal administrators desperate to somehow turn the urban tide; of TANU officials anxious to exert greater economic and spatial control over a frequently uncooperative Tanzanian populace; and of a privileged urban class keen to preserve their "Haven of Peace."
Andrew Burton is an Honorary Research Associate of the British Institute in Eastern Africa, and is currently based in Addis Ababa. He has published widely on East African urban history in Africanist and non- Africanist journals. His books include African Underclass: Urbanisation, Crime and Colonial Order in Dar es Salaam (Athens, Ohio/ Oxford/ Dar es Salaam: Ohio University Press/ James Currey/ Mkuki na Nyota, 2005); and the forthcoming collection (co-edited with James R. Brennan and Yusufu Q. Lawi), Dar es Salaam: Histories from an Emerging African Metropolis, (Dar es Salaam/ Nairobi: Mkuki na Nyota/ British Institute in Eastern Africa).