Controlling the Refugee Crisis: How Population Management Strategies Derived from Humanitarians’ Limitations and Misconceptions

Portuguese authorities had registered the increasing migration of Angolans of Bakongo origin to the Congo during the 1950s. This issue, recurrent in administrative reports, reflected the abusive labor conditions in the colony’s northern districts and the ease of crossing an uncontrolled and artificial border,⁴ with movements across the border being common, especially for trading purposes and given strong kinship affinities. Legalized in 1899, forced labor legitimized the idea of work as a moral duty instituted in civilizational, humanist and economic arguments, and the use of coercive mechanisms to compel natives to produce. Although international pressures in the first half of the twentieth century conditioned the evolution of “native policies” in the colonies, leading to the outlawing of forced labor, the legislative framework of the 1950s allowed for significant arbitrariness in colonial authorities’ action, both in recruiting and distributing native labor (^{Jerónimo and Monteiro 2013}). Development plans instituted in this period aggravated the situation, specifically by encouraging European settlements in northern Angola and by imposing an exploitative “contract system,” characterized by coerced recruitment, abuse and forced dislocation of workers in response to shortages of manpower. This process created an escalating cyclical dynamic of coercion and escape: the more those of Bakongo origin fled from forced labor to the Congo, the more repressive the measures imposed by local authorities and companies in seeking to force the autochthonous population to work in coffee and cotton plantations.

While this period also saw liberalizing ideas and reforms reflecting a concern with the population’s living conditions (^{Keese 2004}; ^{Curto and Cruz 2013}), the persistence of coercive mechanisms, forced labor and even slavery was all too evident in northern Angola. The reformist impetus related more to increasing international pressure at the UN and the International Labour Organization and to the need to win over the hearts and minds of the population in response to the lack of manpower (^{Monteiro 2018}) at a time when decolonization was gaining momentum in international forums. But while the moral duty to work may have been theoretically reframed by some authorities-an often quoted example is that of Helio Esteves Felgas, the Congo district governor-into the idea of labor as a social value, discrimination through the hierarchization of different levels of civilizational evolution continued to persist. Reformist efforts to increase the population’s voluntary interest in working in agriculture were also dissociated from the latter’s way of living.

A 1960 report from an inspector at the Portuguese Overseas Ministry’s Office of Political Affairs explains that agriculture was not in the Bakongo men’s nature and that, instead, they preferred trading activities.⁵ Besides migrating to the Congo in search of better living conditions and to integrate the UPA/FNLA and other political movements, the population of northern Angola manifestly displayed resistance in 1960 to forced agricultural labor and to abuse by the “white man” (^{Keese 2013}; ^{Curto and Cruz 2013}).⁶ Deprived of rights of association, Angolans resorted to passive resistance by refusing to work or fleeing into the bush impenetrable by Portuguese authorities (^{Brinkman 2008}). Although the “reformism” of the late 1950s has given rise to the idea that criticism of the widespread dynamics of forced labor, as expressed by the International League for Human Rights, for example, may have been exaggerated due to the influence of Congolese authorities and Angolan liberation movements (^{Keese 2004}), the 1960 report referred to above alerted to precisely the same problems of forced recruitment, long-term displacement, abuse and expropriation. Feeling themselves to be treated unjustly and inhumanely, people had good reason to complain and to rejoice at the idea that, on the other side of the border, a Bakongo was in power and “blacks governed over whites.”⁷ Where forced labor persisted, the problem was non-enforcement of the law, not one of contesting the idea of labor as the “element par excellence of civilization” (^{Monteiro 2018}, 267). As we shall see later, the legislative reforms of 1961-abolishing the “Native Labor Code”-did not translate into the policies being implemented, but instead served to perpetuate a repressive developmentalism (^{Jerónimo 2017}) with an emphasis on welfare and humanist discourses.

The escalating violence in Angola’s northern districts from the end of 1960 onwards reflected both the ineffectiveness of the “late colonial reformism” and the sentiment among Angolans on both sides of the border (^{Keese 2004}) that independence and self-determination were attainable only though armed struggle-a trajectory analogous to the Algerian decolonization process (^{Byrne 2016}: 14-68)-and that passive resistance was no longer an option. The conflict was the way to politicize Angolans’ plight. The circulation of such ideas and the associating of forced labor dynamics with violence led to a situation of resistance and flight. But while the influx of thousands of Angolans into the Congo increased in cadence and changed in nature, with reports testifying to the prevalence of women, children and the elderly among the flows of refugees, they should not be dissociated from earlier dynamics.

When Angolan refugee flows were first identified, the Congolese authorities were facing multiple humanitarian emergencies related to two secession attempts in their own territory, one in Katanga, another in Kasai. The UNOC-with military and humanitarian teams-and the LRCS-with medical and operational teams supervising the work of the incipient Congolese Red Cross (CRC)-were responding on the ground when the Congolese authorities first appealed to the UNHCR in May 1961 for assistance with the Angolan refugees. At the time, the UN High Commissioner Felix Schnyder was already discussing the refugee situation with John Kelly, the UNHCR officer liaising with UNOC, and LRCS staff in Geneva. Various limitations account for the time-consuming nature of the decisions taken on UNHCR intervention. First, legal concerns arose regarding the eligibility of the refugees and Congo’s failure to ratify the 1951 Refugee Convention,⁸ with discussions revolving around the question of Angolan refugees belonging to the same “tribe” as the local population. Doubts arose about the eligibility for UNHCR protection of refugees who settled with and were assisted by relatives or acquaintances and had been granted asylum by the Congolese authorities. The UN Secretary-General, Dag Hammarskjold, ultimately resolved the matter by declaring that the criteria for eligibility depended on nationality, not on ethnic affinities. These were Portuguese nationals, according to Portuguese law, who were allegedly being persecuted by authorities in their state of origin. This decision was confirmed by interviews conducted with refugees by John Kelly in the only comprehensive report by an international humanitarian organization on the causes of refugees’ flight in the records: refugees referred to the indiscriminate shooting, arrest and deportation of the better educated, to the burning of villages by the Portuguese and also to violent actions undertaken by the UPA.⁹ Second, the costly budgetary obligations relating to the Algerian refugee crisis had depleted both the UN Agency and the LRCS, with refugees being entirely dependent on humanitarian aid from 1957 to 1962 (^{Guardião 2019}: 239-78).¹⁰ Third, there was the problem of overlapping UNOC and UNHCR responsibilities on the ground. To overcome the financial shortcomings and inter-agency discomfort, the UNHCR was relegated to overseeing integration projects designed to promote refugee self-sufficiency.

These issues were aggravated by the intricate contours of the UN intervention in the Congo (^{O’Malley 2016}), by the contours relating to the UN debates on the Angolan question and also by the extreme insecurity that obstructed information and supply channels in the Lower Congo. While concerns about refugee eligibility and operational aspects were overcome through communications between the UN Secretary-General, Schnyder and the UNOC, political obstacles and pressures delayed the intervention, while also straining relations between the UNHCR and Portugal. Portuguese authorities pressured Schnyder not to further internationalize the Angolan conflict since cooperation on refugee assistance and repatriation was already being discussed with the LRCS.¹¹ However, the UNHCR’s experience with refugees in the Algerian decolonization had proved that aligning with colonial authorities was not only unproductive for the Agency, but also dangerous. One year earlier, the UNHCR’s neutrality had come into question when the former High Commissioner August Lindt agreed with De Gaulle not to make international appeals on behalf of Algerian refugees and instead to negotiate repatriation with host states in exchange for meager funds. The issue not only caused friction between the UNHCR and LRCS, with the latter threatening to abandon the operation, but also led to severe criticism and the possibility of a formal complaint being lodged at the UN by the Afro-Asian group denouncing Lindt’s “observation of colonialist behavior” and obstruction of international assistance (^{Guardião 2019}: 247-61). Although efforts to make the Portuguese authorities comply with the UN resolutions avoided triggering total estrangement, Portugal’s refusal to grant any concessions hampered negotiations regarding refugees.¹²

The tone of the debate at the UN also reflected attempts to relate the refugees’ plight to self-determination and to condemn Portuguese colonialism, with delegates often using the refugees’ plight to substantiate Portugal’s perpetrating of human rights’ violations and to claim that the conflict represented a threat to international peace and security. In a second attempt to pass a Security Council resolution for that purpose, delegates used refugees’ testimonies to corroborate Angolans’ right to self-determination, with the Monrovia Group countries, Mali and Morocco belligerently threatening to support Angolan independence.¹³ After appealing for intervention by the UNHCR, Congolese authorities then stalled the sending of a UNHCR representative to Leopoldville by insisting on taking the Angolan refugee issue to the First Committee, where threats to international peace and security are debated, rather than the Third Committee, where human rights and UNHCR-related topics are discussed. Deputy High Commissioner Sadruddin Aga Khan and Executive Officer James Colmar were required to travel to Leopoldville to negotiate the contours of the debate and arrange for UNHCR assistance to be granted. It was not until November, therefore, that Vladimir Temnomeroff, the UNHCR representative, was able to start his mission supervising integration projects.¹⁴ Contrary to ^{Rich’s suggestion (Rich 2021}), the UNHCR’s first direct involvement in the crisis was in late 1961, when it took on responsibility for integrating refugees and resolving problems of overpopulation.

In the meantime, the UNHCR allocated USD 100,000 to the emergency relief operation coordinated by the LRCS, in cooperation with the UNOC and local voluntary agencies, to assist rising refugee numbers.¹⁵ The Lower Congo region was divided into three main assistance areas, with the LRCS directly supervising the CRC assigned to the Matadi-Songololo area, where the majority of the around 100,000 refugees were settling. Although resource distribution largely ran smoothly after a period of initial uncertainty, two associated problems soon became evident. One of these was the possibility of refugees becoming dependent on humanitarian assistance forever, while the other concerned refugee mobility.

From the beginning of the operation in June, LRCS delegates and administrative officials in Geneva agreed that emergency relief should be terminated before too long in order to avoid refugees becoming dependent on external assistance indefinitely. Angolans were consequently encouraged to produce their own food in plots assigned to them by local authorities in an effort to encourage them to quickly become self-sufficient. Although the LRCS initially scheduled 31 October as the date for terminating emergency relief, delays in production and complaints from voluntary agencies and central authorities resulted in postponement to 31 January 1962.¹⁶ Meanwhile financial burdens in North Africa, where the main issue was that of refugee dependence, made terminating the operation as soon as possible imperative. The stipulated schedule was consequently maintained despite a seemingly endless flow of reports on refugee influxes and the central authorities’ inability to deal with the situation.¹⁷ Claims that refugees were not willing to work in agriculture rose, with local authorities arguing that refugees preferred to exchange the agricultural goods given to them, which resulted in an increase in local commodities value, while resorting to prostitution and passively awaiting distributions of Red Cross resources. This in turn led to tensions with the impoverished local population.¹⁸ It then fell to overwhelmed and apprehensive local authorities and humanitarian organizations’ population management strategies and capacity to mobilize refugees to resolve these problems.

Let us now turn to the mobility question. The first refugee groups-mainly Angolans departing from the Congo and Uíge district-scattered in Congolese villages along the border in the Boma-Matadi-Songololo line, reaching north to Thysville and East to Popokabaka. One of the main concerns during the emergency operation was refugee registration because this allowed for a more accurate estimate of influxes and populations per area, and consequently the resources needed. As previously stated, refugees commonly settled among relatives or persons of their acquaintance, and Angolans of Bakongo origin had long maintained commercial routes for exchanging produce in local markets, thus strengthening ties on both sides of the border. Once they were in the Congo, refugees often moved from one village to another, looking for familiar faces to seek assistance, while others moved back and forth across the border, searching for those who had stayed in the bush or to harvest crops left behind, and defying the risk of detention and, more often than not, death (^{Brinkman 2008}: 204). The dynamics of refugee mobility made it more difficult to properly assess the influxes and to distinguish between newcomers and groups established earlier, and between Angolans and Congolese.¹⁹ This mobility also led LRCS delegates to believe that refugees were willing to return to Angola, with the result that repatriation then emerged as a central solution to the crisis.

Refugee repatriation as a means of getting the crisis under control was also connected to the idea, propagated by Portugal, that Angolans were being encouraged by the UPA/FNLA to seek refuge in the Congo and that the Portuguese counteroffensive had promoted stability and the conditions needed for their return. Red Cross personnel in Leopoldville had close ties with the Portuguese authorities and shared their conviction that liberation movements and the Congolese government were using the refugee crisis to internationalize the decolonization conflict and increase aid. Since May 1961, the Portuguese authorities in Lisbon and Leopoldville had been in contact with delegates from the LRCS and the International Committee of the Red Cross (ICRC),²⁰ both of whom regarded repatriation as a practical solution and a political weapon:

LEAGUE TRIES NOT TO DRAMATISE SITUATION STOP LEAGUE NOW NEGOTIATING WITH PORTUGUESE AUTHORITIES AND PORTUGUESE RED CROSS WITH VIEW INCREASE THEIR AID AND TRYING ESTABLISH DIRECTLY THROUGH RED CROSS ANGOLA PROTECTED RECEPTION CENTRES ON ANGOLA SIDE OF BORDER TO ENCOURAGE RETURN OF REFUGEES STOP ATTITUDE OF LEAGUE MOTIVATED BY PURPOSE OF REDUCING PROBLEM AWARE OF LOCAL TENDENCIES CONGO AND CONGO RED CROSS TO BLOW UP PLIGHT OF REFUGEES IN EFFORT TO MAKE THIS PROBLEM A COMMON CAUSE OF AFRICAN FREEDOM FIGHT QUARTO EYE [sic] HAVING THEN STATED THAT ANY FORCED REPATRIATION SHOULD BE AVOIDED LEAGUE MENTIONED THAT MAJORITY OF REFUGEES IN NO WAY INVOLVED IN POLITICAL STRUGGLE BUT VICTIMS OF TERRORIST ACTIVITIES AND OF MILITARY CLEAN-UP OPERATIONS²¹

Sending the Portuguese Red Cross (PRC) to the border was, in fact, a way to divert any intervention by the ICRC in the conflict and to show the UN that Portugal had committed itself to protecting and assisting Angolan civilians.²² The reception centers, which were suggested by LRCS delegate John Thelen and set up to receive 1000 refugees a week, served both Portuguese and LRCS aims. On the Portuguese side, they functioned as points for attracting refugees and immediately starting psychosocial action to win over the autochthonous population, as well as promoting a benevolent image of Portugal, while on the LRCS side they were seen as enabling a rapid solution and depoliticization of the crisis. Even as refugee influxes increased, thus proving the stability theory wrong, and with independent testimonies also reporting continuous air bombing in the region,²³ an LRCS official communiqué attested otherwise after a team of delegates visited the facilities.²⁴ It is evident from LRCS records that the League had placed all its bets on repatriation as a means of resolving the crisis and, as a result, self-sufficiency strategies to promote refugee integration were to be implemented only for those not willing to return.²⁵

Although regarded as a common paradigm after the Second World War (^{Chimni 2004}: 55), the focus on repatriating Angolans differed substantially from the response to another refugee crisis in the Congo.²⁶ Since the 1961 September elections in Rwanda, influxes of Watutsi refugees into the Congolese province of Kivu had increased, reaching 60,000 in January 1962. However, the possibility of repatriation was rejected in this case as, contrary to the assessment of the Angolan situation, the LRCS and UNHCR delegates concluded that conditions in Rwanda were hostile to repatriation and that local Congolese communities and authorities were either opposed to or taking advantage of the refugees. Evidence of conflicts between the Watutsi and the Congolese was verified, as was evidence of refugees being put to work in abnormal conditions. Protecting refugees in three centers led by voluntary organizations and their subsequent resettlement in Bihwe, to which they consented, were the solutions envisaged for the Rwandan refugees. It was decided that, in this case, refugees and locals should be made responsible for building infrastructure at fair pay in an effort to promote trusting relationship dynamics in the region. In these conditions, it was estimated, refugees would become self-sufficient within one year after their arrival.²⁷ The repatriation strategy adopted in the Angolan case was thus related to the assurance by Portuguese authorities, which was regarded as reliable by Red Cross staff at the time, but disregarded the specificities of a long and violent decolonization process.

The Watutsi resettlement solution was concomitant to the UNHCR’s initial attempts to implement a similar strategy for those Angolans refusing repatriation, with the UNHCR seeing this as a way of resolving overpopulation problems in urban areas and keeping the short-term emergency assistance plan afloat. By the end of 1961, the UNHCR had established partnerships with the Congolese Protestant Relief Agency (CPRA) and Caritas Congo to initiate rural development projects. One of these projects, established with CPRA, set out to allocate agricultural tools and seeds to about 7,500 refugees already installed between Kimpangu and Moerbeke and, after overcoming delays in resource distribution, was considered a success. The other envisaged relocating 6,000 refugees from the Matadi region to a settlement north of Seke-Banza, as well as building housing and community support infrastructure, such as schools and dispensaries - resorting to unpaid refugee labor - and providing seeds and tools for agricultural production.²⁸

Although they had the same objective in mind, the projects differed on several crucial points, and these, in turn, determined their success. Whereas the Watutsi project was grounded in forging better relations between refugees and local communities, with fair payment and refugee consultation and agreement to resettlement,²⁹ and the CPRA project foresaw agricultural resources being allocated to refugees who were already settled in the area, the Caritas-led project took no account of local dynamics and Angolans’ claims and expectations, and consequently failed to achieve the desired results. The UNHCR’s planning assessments for implementing resettlement strategies derived from a simplistic view of local dynamics. In Temnomeroff's appraisal, the nomadic nature of central African peoples and the predominance of agricultural activity would make it easy to persuade refugees to move, even though many of these refugees were traders and others specialized in manufacturing or worked in the tertiary sector. The temporary character of the planned resettlement also took no account of the possibility of long-term conflict. Using unpaid labor to build infrastructure saved money for the UNHCR as it meant it only had to invest in rudimentary tools for construction and agricultural production to create what were regarded as satisfactory conditions for getting refugees to relocate inland.³⁰ This evaluation, however, was utterly contrary to the considerations applied in the Watutsi project and to the dynamics of refugee settlement because it created a situation similar to the one refugees had experienced and protested against in Angola. For more than two years, therefore, Angolan refugees refused to resettle. The turmoil caused by overpopulation worsened the situation both for the refugees and the local population, while also undermining the idea that international humanitarian organizations were capable of responding to the humanitarian crisis.

In resorting to repatriation and rural development projects based on resettlement, humanitarian actors thought they had devised a solid plan for responding to the crisis within their limited capacities. However, the conditions imposed on the Angolan refugees led to the failure of the envisaged strategies, while the disregarding of local dynamics and refugee resistance led to the humanitarian projects being prematurely terminated in 1963. The following section explores these two topics and how they related to the Portuguese late-colonial state’s repressive developmentalism.

Taking Back Control? Refugee Resistance to Humanitarian and Colonial Population Management Strategies

Frustration with the population management strategies emerged almost simultaneously with their implementation, with the result that the erroneous idea that refugees would be permanently returning to Angola faded away. Refugee influxes continued increasing, thus aggravating the problems of overpopulation in rural, but also and primarily in urban areas. Refugees’ presence soon became unwanted by unemployed and impoverished locals demanding assistance and by authorities who questioned Angolans’ refugee status, given that many of them were not victims of Portuguese massacres but instead fled only due to their fear of this possibility.³¹ Moreover, refugees shared the LRCS delegates’ expectation of stability in Angola, not because they anticipated this would be ensured by the Portuguese authorities, but rather because of UPA/FNLA propaganda claiming that independence would soon materialize. According to a report sent by Temnomeroff to the UNHCR, 40% of the Angolan refugees refused to invest in agricultural projects that they would be abandoning in a few months. Their refusal prompted a firm response by local authorities, with the support of the LRCS. A pamphlet written in Portuguese and Kikongo was distributed throughout the region appealing for Angolans to work and “LISTEN TO THE WISE MEN” who “know that those who do not work do not eat.”³² Cooperation with the UPA/FNLA to this end also failed, thus demonstrating that while the liberation movement had some influence over refugees, its directives were also contested.

Resistance to resettlement was also linked to refugees’ fear of the Portuguese authorities and mistrust of those responsible for protecting and integrating them; in other words, the humanitarian agencies and local Congolese authorities. Their fear and mistrust were interconnected and cumulative and should accordingly be taken into account for a comprehensive understanding of why the population management strategies failed.

One significant element in this understanding is an awareness of what caused refugees to flee and the claims they made during inquiries. These unveil some clues as to the crisis’s intricate dynamics. Inquiries led by the UN Special Committee on Territories under Portuguese Administration (UNSCTPA) in May 1962 revealed that “most of the refugees did not want to go back to Angola until their country had attained independence. Some said they would rather die in freedom than go back to forced labour [sic] and the sufferings from which they had fled. They had heard that the Portuguese authorities were trying to get refugees to go back to Angola and were promising a new era in which things would be different. They did not believe these promises. Refugees who had recently arrived from Angola confirmed their suspicions; people were still subject to oppression, they said.” Compulsory labor was one of the main reasons prompting refugees to flee, as it had also been for other Angolans before the war. Testimonies of the use of napalm and indiscriminate attacks on civilians in northern Angola were also recorded.³³ When local Congolese authorities and humanitarians insisted that refugees should be repatriated and had to work and “listen to the wise men,” Angolans associated these tactics with a return to the previous abusive situation they had faced in Angola. Meanwhile UPA/FNLA propaganda amplified this idea by inciting refugees to stay put in areas close to the border because they would soon be able to return. Once the belief that independence was imminent started to fade, refugees willingly began to integrate into Congolese society; some were already doing so in early 1962, despite difficulties caused by poor harvests, while the majority actively displayed a willingness to integrate from 1964 onwards. Whereas humanitarians and local authorities had portrayed them as lazy and reliant on external assistance, refugees demonstrated a wish for self-determination and protection, and that these factors had been neglected.

The UNSCTPA report also sheds light on another critical factor explaining why refugees refused to move: the “new era” that Portugal was actively trying to promote in Angola. This factor was reflected not only in the failure of the repatriation strategies, but also in the failure of attempts to resettle the refugees. Trying to circumvent increasing international scrutiny - which clearly attached importance to the language of human rights, including the right of self-determination - and prevent an external intervention in Angola, Portuguese legislators swiftly promoted a structural legislative reform (^{Jerónimo and Monteiro 2020}) by abolishing the Indigenous Status and, later, the Native Labor Code, while appending this reform with rural and community developmental policies intertwined with, and dependent on, security measures. These included a full-scale reorganization of the territory, reinforcement of metropolitan settlement in autochthonous land, and the confinement of natives to regedorias, or to labor and detention camps.

We can see the intertwining of military tactics and ideas of social and economic development in the theoretical plan of General Hermes de Oliveira on counter-subversive measures to be implemented in Angola, as well as in the directives of General Venâncio Deslandes, Governor-General of the colony.³⁴ Both refer to the need to pacify the territory by resorting to psychosocial action to attract and “tame” the autochthonous populations hiding in the bush or taking refuge across the border. Besides psychosocial action within the territory, the Portuguese sent loyal Angolans to the Congo or corrupted local Congolese authorities to encourage refugees’ return.³⁵ But although several attempts at cooperation between the Congolese and Portuguese materialized during the war, refugees often escaped and returned to the Congo. The aim of this intertwinement was twofold. On the military side, it was imperative to gain the trust of the “natives” in order to get them to denounce (coercively if necessary) “terrorist elements” infiltrated among those living in the bush or returning from the Congo. On the social side, the aim was to direct the autochthonous population to see the benefits of civilization and Portuguese humanity towards them. Accordingly, efforts were made to create a dependency relationship by, for example, exaggerating the severity of illnesses and injuries inflicted on civilians or by providing limited food so as to increase contacts between the population and the Portuguese authorities. To isolate the population-again within the military strategy-recuperados (redeemed groups) who “voluntarily” presented themselves to the authorities (due to hunger and exposure to violence and persecution) would be sent to regedorias. At these sites, Oliveira stated, “something concrete, positive, real (had to be done), material acts that arouse in the population the awareness that their most instant, most distressing problems […] are being solved by us,” while the social action was to be incorporated into broader development plans in a second phase. With rural labor within regedorias increasing production and consequently these groups’ living standards, labor, masked as “voluntary,” was seen as closely connected to the desired social behavior. A “new type of society would be created (instilling) in the inhabitants’ spirit the notion of their true position within the community to which they belong.” To this end, civilians were to be removed from their “natural habitat” and confined within structures designed for rural and social development (^{Oliveira 1962}: 199-201). Population regrouping would also “facilitate psychosocial action” led by social and health agents such as the PRC.³⁶ This “rehabilitation” strategy was similar to the strategies applied in the villagization models in Kenya and Algeria, which also involved the countries’ national Red Cross Societies (^{Guardião 2019}; ^{Johnson, 2020}). This is not to say, however, that the measures resorted to by the Portuguese were exactly the same as those applied by the British and French authorities. Further research is needed for a full account of the reality in Angola, similar to that conducted in the cases of the brutal application of the Swynnerton and Constatine Plans (^{Baggallay 2011}; ^{Davis 2017}; ^{Elkins 2000}, 2005; ^{Feichtinger 2017}). The similarities are reflected in developmentalist ideas being associated with socially engineered counter-insurgency repertoires-including punishment, “rehabilitation,” social and moral re-education and community development-to forge “new societies” (^{Jerónimo 2017}). Oliveira’s aim was the “creation of a new order, understanding as such the entire transformation of the living conditions of the populations, not only economic and social but also cultural and moral.”

The policy was rapidly integrated into the five-year development plan (1959-1964), consolidating the Portuguese late-colonial state’s repressive developmentalism (^{Jerónimo and Pinto 2015}). This combination of population control and developmental policies aimed at rehabilitating the rural autochthonous population was one of the primary strategies applied by the Portuguese authorities in seeking to win the war and simultaneously promote economic stability and production in Angola. But if the Portuguese authorities were aiming to rebuild Angolan rural society through internationally disseminated developmental ideas in order to circumvent broader criticism at the UN, the reality on the ground was permeated by the perpetration of abuses contrary to the ideals of human rights.

The implementation dynamics of two regedorias in Uíge in the second half of 1961 constitute an example of this new model, with refugees returning from Congo and those hiding in the bush being the primary targets of the policies. When “appearing before” the authorities, these groups were subject to harsh interrogations, instigated by the authorities’ distrust of their intentions. They were compelled to work on the land and to build housing, administrative and communications structures within the regedoria model (^{Curto et al. 2016}: 244-6). According to the UNSCTPA report, “traditional militia” authorities, nominated by the governor or governor-general, in the regedorias sought to impose respect and to recruit labor “to public works and projects” without payment and subject to punishment. These “instruments of the authorities [did] not represent the people.” The Committee concluded that “the new decree organizing regedorias […] introduced virtually no changes and the Committee does not feel that at present the indigenous people enjoy any real measure of genuine self-government.” Moreover, rural Angolans were obliged to stay in the regedorias. Only those “who [had] ‘reached a fairly high level of development’” could be integrated into urban life. The model and legislation behind it perpetuated a stratification of the Angolan society in civilization levels, involving “cultural discrimination which is not compatible with the purposes and principles of the United Nations and the Universal Declaration of Human Rights.”³⁷ To the UN, the exacerbating of the politics of difference (^{Burbank and Cooper 2010}; ^{Jerónimo and Pinto 2015}) in Angola was all too evident.

During this period, Angolan mobility dynamics along the border areas and in the Lower Congo region facilitated the circulation of information among refugees (^{Brinkman 2008}: 210). Reports on refugees going back to Angola became less frequent, while new arrivals, mainly those hiding in the bush on their way from the distant areas of Ambriz and Ambrizete, also evidenced that refugees were trying to escape both the violence and the “new order” being implemented.

Refugees’ resistance to resettlement and integration into models such as those adopted by humanitarian organizations in the Congo is hardly surprising. Both cases involved rural Angolans being taken, without consultation, from land where they had settled and being forced to work without payment to build housing infrastructure or facilities designed for community development. According to the LRCS and local Congolese authorities, refugees claimed they should be entitled to choose where they were to be resettled and demanded to see the location before agreeing to join the Caritas project. They feared being enslaved or handed over to the Portuguese authorities. According to Art Hansen, an anthropologist who interviewed Angolan refugees in Zambia, freedom of movement was a crucial issue in refugee settlement. Most refugees chose to settle among kinship communities, even if this meant they would not receive humanitarian assistance. The preference for integrating within local communities, Hansen concluded, was mainly associated with the need to control their refuge experience after power over and control of their social reality had been taken away from them by the circumstances created in Angola (^{Hansen 1981}), and specifically the violence, cooption and punishment of local chiefs and the confining of entire communities to “open air” camps at the same time as the regedoria model was being extended to the east.³⁸ Angolans in the Congo wanted to take back control and, to a certain extent, this is exactly what they did, despite continuing to be subject to a variety of abuses in the host country (^{Guardião 2019}: 322-323).

By 1963, when both the LRCS and UNHCR terminated their operations in the Congo, only 66 refugees had agreed to be resettled.³⁹ Refugees continued instead to settle in local communities scattered across the province. The UNHCR justified the decision to terminate activities, stating that the refugee situation was now satisfactory and that the issues still to be resolved, namely those relating to integration, were host-state responsibilities. Sadruddin Aga Khan also assumed that some solutions envisaged “could not be adhered to,” attributing the lack of progress to UNHCR ineffectiveness and inexperience of humanitarian organizations in “these parts of Africa.” Examples of the “adversities” that had to be “overcome in the implementation of some of the programs” were “good or bad harvest, natural disasters [and] last, but not least, the psychology of the refugees themselves.” The UNHCR nevertheless left a representative on the ground to try to persuade refugees to work with settlement projects, especially the Caritas project, in cooperation with local organizations and the Congolese authorities. Particular attention should be paid to the continuing influx of refugees who, by that time, were settling not only in the Lower Congo, but also in Katanga and Kasai.⁴⁰ From 1964 onwards, the UNHCR dedicated its work in the Congo to supporting education, agricultural and commerce-related projects, none of which resorted to resettlement.

What is striking in Sadruddin Aga Khan’s declaration is that although the UNHCR acknowledged humanitarians’ lack of experience in dealing with refugee problems in sub-Saharan Africa, refugee psychology was identified as a major factor in the failure of resettlement strategies. It was telling that neither the UNHCR representative nor LRCS delegates sought to engage in constructive communications with the refugees. Instead, refugees’ claims were dismissed even as UN reports attested to the specificities of decolonization and refuge dynamics in this context, while LRCS delegates-in other words, those individuals specifically responsible for strategizing the operation-tended to share Portuguese views on how to manage the population. These views contrasted with those displayed in the UNHCR/LRCS relationship with Rwandan authorities in the case of the Watutsi. Some delegates, such as Norredam, even suggested with regard to the situation in the Congo that “the African mind,” specifically relating this to “tribalism,” was the cause of the deterioration experienced by African nations after becoming independent. The underlying notion for this disparity was the firm belief that African problems were “fairly simple” to resolve, and that the way to do this was through development. The frustration and abandonment felt by Angolan refugees reflected Norredam’s conviction that, contrary to what was “generally believed” (he included himself in this), “a little economic aid and a few specialists” were not sufficient to contour the mismanagement of African societies. Nor was he willing to acknowledge that the LRCS was to blame for mishandling the situation. The real reasons, Norredam stated, “lie much deeper and have their root” in what had brought the Congo back “generally to the time before the Belgians arrived […] The[y] blame ‘colonialism’ or new colonialism’ (same thing as we call ‘economic aid’!), […] and after that it will be something else. [T]he reasons why Africa deteriorates these years, and is fast on way back to the jungle, must be found in African tradition, mentality, religion.”⁴¹ While Portuguese authorities made reference to civilizational ideals in differentiating themselves from Angolans and to stratify the Angolan society, Norredam pointed to tribalism as the main factor for African societies’ underdevelopment.

Another element, in addition to the lack of communication between humanitarians and refugees, entailed the efforts by humanitarian agents to depoliticize the refugee crisis. Refugees’ sense of self-determination, allied with the presence of the UPA/FNLA in the Lower Congo, generated an environment of mistrust, in which refugees were seen as being manipulated by the liberation movement into becoming political agents. François Chassing, the UNHCR representative substituting for Temnomeroff, referred to refugees “imposing themselves politically on the local population” and stated that a “fear psychosis was voluntarily created.”⁴² This idea, disseminated among humanitarians and local authorities, contrasts with refugees’ protests against abuses by UPA/FNLA agents in various situations, specifically in the form of false promises and the appropriating of resources or taxation (^{Guardião 2019}: 322). In trying to depoliticize humanitarian action in the Congo, these agents failed to acknowledge and respect refugees’ fears, claims and action, and consequently demonstrated humanitarians’ inability to move away from their own ideas of a modernizing Africa and to see refugees not as objects of aid, but as subjects of international law. This view was shared by those in Geneva who regarded European refugees’ problems as relating to legal protection, and African refugees’ problems as relating to aid, development and labor value (^{Glasman 2017}). The “white man’s burden” and ideas of superiority were thus still very much present within the humanitarian realm-which dealt primarily with the common African-at a time of effervescing UN debates on the institutionalization of human rights, led by Global South leaders and increasingly focusing on discussions of self-determination and the condemning of racial discrimination (^{Burke 2010}; ^{Jensen 2017}; ^{Burke, Moses and Duranti 2020}; ^{Jerónimo and Monteiro 2020}).

Concluding Remarks: Refugee Management Paradigms, Decolonization and Resistance

As this article demonstrates, the paradigm of refugee management strategies established after World War II and based on voluntary repatriation and resettlement (^{Chimni 2004}) proved ineffective in the case of Angolans fleeing the violence and state-led population control policies associated with the decolonization process. This failure was related to the specificities of decolonization and, more specifically, to Angolan refugees’ resistance and to the negligence of international humanitarian agents on the ground regarding refugee agency and the humanitarians’ inability to appreciate those specificities.

From the start of the crisis, plans for repatriation and resettlement were made with two aims in mind: to end the emergency operation swiftly and to control the problems of overpopulation and refugee movement dynamics in a host state unable to respond to the crisis. An agreement between the LRCS, UNOC and UNHCR set out the responsibilities of each organization: the LRCS was responsible for emergency assistance, in cooperation with UNOC, and population management strategies, whereas the UNHCR was to conduct short- to medium-term integration projects aimed at achieving refugee self-sufficiency from November 1961. The delay in starting these activities was marked by legal, financial and operational limitations, as well as by political constraints attributable to the Congolese authorities’ efforts to portray the crisis as a matter of peace and security rather than as a humanitarian issue.

Despite the Global South countries’ efforts to denounce the perpetration of human rights abuses in Angola at the UN, negotiations between the LRCS and Portugal on repatriation began without a proper assessment of what had caused refugees to flee in the first place. Although negotiations may have been in line with the paradigm of refugee population management, the fact that humanitarians attached greater weight to colonial authorities’ guaranties than to refugees’ claims during the first year of the crisis demonstrates the unsuitability of a state-centric approach-based simultaneously on seeking to resolve host-state problems and on negotiations with “more reliable” colonial authorities-and the agencies’ lack of preparedness for dealing with the specificities of decolonization refugee crises, even though the Algerian experience had proved that repatriation was not a viable solution.

Causes of refugee flight, claims and movement dynamics were deeply associated with Portuguese population management strategies in Angola, which even before the war had been marked by persecution, forced migration and compulsory labor. These were subsequently aggravated by the conflict, thus stressing the securitized side of development policies in the colony. Repressive developmentalism, reflected in Portugal’s last stand at maintaining its empire, was characterized by critical divergencies between discourse-engendered to circumvent international criticism-and practice. The regedorias model was just one example of how development was associated with repression of the civilian population and the persistence of differentiation models among the African population. Population management strategies-repatriation and resettlement-to deal with overpopulation in the Congo were opposed by refugees in multiple ways on the grounds of their similarities with colonial strategies. This resistance ranged from the denouncing of abuses in Angola in UN inquiries to refusal of resettlement out of fear of being handed over to the Portuguese authorities or fear of enslavement and to resisting engaging in agricultural projects in the belief that independence was imminent. As well as stating that they would not return to Angola until independence was attainable, refugees also demanded the right to choose their place of settlement in order to regain control they had lost. In this sense, Angolan refugees’ agency was similar to that of Eastern Europeans who fled imperial oppressive regimes and subsequently refused repatriation or resettlement within Eastern Europe.

The unwillingness or inability of humanitarians on the ground to integrate refugee agency into their activities prompts three significant conclusions regarding such action in the early 1960s, specially in decolonization-led crises. First, the incorporation of information from UN bodies trying to assess the situation in Angola was either unproductive or neglected. While reports from the UNSCTPA used refugee claims to denounce abuses in Angola and criticize Portuguese policies, UNHCR staff insisted on a resettlement project that was similar to the modus operandi of the Portuguese regedorias model in various ways, including the idea of moving refugees from a place they had chosen to settle into an unknown and controlled location and the idea that they should provide unpaid labor to build housing and infrastructure. Second, the persistence of civilizational ideals among humanitarians served to dismiss refugee agency. Depictions of refugees as lazy and unwilling to work were at variance with claims made by refugees themselves, but resonated with the broader Portuguese view of Angolans of Bakongo origin. The link between repressive measures in Angola and refugees’ claims and resistance, and the incapacity of humanitarian agents to mobilize refugees, is unequivocal. The LRCS also insisted on negotiating repatriation with the Portuguese authorities, attesting to safe conditions in Angola, even though reports from October 1961 onwards showed that this strategy would prove ineffective. Third, an aspect related to the insistence on good conditions in Angola and on resettlement was humanitarians’ belief that (Western-directed) developmental policies were in the best interests of Angolans and the host state alike. Frustration with the failure of the developmental ideal is evident in humanitarians’ assessment of refugee behavior and the conditions in the Congo. Although Deputy High Commissioner Sadruddin Aga Khan acknowledged humanitarian organizations’ inexperience in dealing with refugee crises in sub-Saharan Africa, the perception of refugees as either victims or burdens disregarded refugee agency and refugees’ right to self-determination-which is totally different from the concept of self-governance-and contributed to the politics of difference being extended from the imperial to the humanitarian realm.

The insistence on those ideas by researchers continuing to underplay refugee agency as an element in the failure of resettlement and repatriation strategies in decolonization processes and its connection with human rights ideals must be addressed if we are to understand the interconnections between decolonization dynamics and humanitarian action in the 1960s. Refugee agency and humanitarian organizations’ failure to incorporate this into their population management strategies are the major factors accounting for these organizations’ failure to respond effectively to the Angolan refugee crisis, whereas other projects, which made proper assessments of the local context and accommodated refugees’ claims and modus vivendi, were successful. Studying the concurrent trends and intersections of the institutionalization of human rights, imperial repertoires, decolonization dynamics and humanitarian action will allow for a better understanding of how significant disparities coexisted between dealings with the common African and state elites (African and European)-but also between refugees from different origins-and how these were translated in subsequent decades into African population management strategies, with refugees increasingly being regarded either as labor value or as helpless suffering bodies.