4.1. Infrastructural context

As in many colonial contexts across Africa, Beira (Figure 2) was also divided between a core developed city (cidade de cimento), reserved for the white population, and the surrounding indigenous quarters, called ‘reed city’ (cidade de caniço) in Mozambique (Morton 2019; Noormahomed 2022; Sumich & Nielsen 2020). The legacies of this kind of socio-spatial stratification, or ‘dualist structure’ (Barros et al. 2014), which not only defined the quality and style of housing but also the infrastructure and services available, shape the functioning of the city to this day. In the colonial developed core of the city, there are (largely) tarmac roads and a separated sewage and drainage system, surrounded by extensive expansion zones, both planned and unplanned, where water management is a much more improvised affair.

In either situation, the maintenance of this drainage and sewage system is a constant preoccupation. The city’s sanitation service was transformed in 2008 from a branch of municipal administration to a state-held company, the Beira Autonomous Sanitation Service, or SASB by its Portuguese acronym (Muradás et al. 2021). The sewage system (wastewater) was upgraded between 2008 and 2012 with EU funding, which included the installation of electric pumps at 11 pumping substations – the city is so low-lying that there’s not enough depth to move water by gravity, like in other Mozambican cities. The wastewater treatment plant was also damaged by Cyclone Idai and is being rehabilitated between 2023 and 2026 with funding from the French Development Agency (Agence Française de Développement n.d.).

The rainwater drainage system was also upgraded (Figure 3), with the first phase of rehabilitation completed just before Cyclone Idai hit. And, still, the system was overwhelmed, with power cuts meaning that the tide locks had to be operated by hand – and this, as residents and sanitation engineers would convey, was at low tide. If the cyclone had hit at high tide, the catastrophe would have been much, much worse. Yet, while the second phase of rehabilitation was slated to start at the end of 2023, with US$120 million of joint funding by the Dutch government and the World Bank (Henriques 2023), the reality of trying to keep that system functional is insufficiently expressed by the mapping and planning exercises of civil engineers and donors.

Rather, it looks like a typical work day for the SASB cleaning brigades, who try to make the most of the mornings, before the temperatures get too hot, to manually clean out the gutters and manholes – work that is, in theory, planned in advance and according to seasonal rainfalls, but in practice looks ever so much like unblocking the pipes ad hoc wherever a blockage occurs.

Or it looks like life in the informal, central neighbourhood of Goto, where Luciano, in his early twenties and out of school and out of a job, semi-regularly cleans out the drainage canals with three or four other youths to try to prevent worse floodings. Some of the smaller canals have been cased in concrete by the municipality, with more or less regularly placed slab covers, but the casing often runs out at the point where the budget ran out, and hand-dug canals take over.

More problematically, the lanes here are too narrow for vehicle access, and the collection points at the perimeter of Goto often too sparse and far away for the everyday convenience of residents, and, although people often dig their own rubbish pit in the yards of their self-built cement brick houses, the canals are clogged.

There is stagnant water everywhere, sometimes turning green with mosquitos buzzing above the surface. In most places, the air is stagnant and the stench of black water is overpowering. The incidence of waterborne diarrhoeal diseases is very high here.

The irony of this is that Goto is literally abutting a section of Beira’s ‘Green Infrastructures’ – a showcase project of mangrove reforestation along the River Chiveve, complete with publicly accessible, but fenced off, leisure facilities (Figure 4). Publicly accessible, that is, for the symbolic entrance fee of 20 mt (about GB£0.15), which, I was told, ‘already keeps a certain category of people out’ and prevents the installations from degrading too quickly.

For many I spoke to, these Infra-Estruturas Verdes, a showcase project of the late and venerated former mayor Daviz Simango, are the only visible resilience enhancement in Beira – and that was planned and completed just before Cyclone Idai. The mangroves help absorb tidal water and rainwater, and the landscape park serves to teach city residents about the importance of nature conservancy but, with the linkages from hand-dug and rubbish-clogged tertiary valas to a main drainage outlet like the Chiveve so precarious, the spatial and social inequality is indexed quite visibly by the levels of flooding and the speed of drainage during the rainy season.

4.2. Infrastructural interdependencies

In addition to the challenges of managing pluvial and wastewater, Beira is marked by infrastructural interdependencies that decisively delimit its economic horizons, and thus the possibilities of building resilience in the city. Focusing, as development practitioners tend to do, on the multiple and real existing challenges of unplanned (or incoherently planned) urban extension into flood-prone areas will reveal a cluster of issues, from drainage and sanitation to waste management, to building materials and standards, and general socio-economic deprivation (see the previous section). It is worthwhile highlighting the role and weight of the port and of extractivism in shaping the development trajectories of Beira.

The Port of Beira was historically, and still is, the main driver of urbanisation in the city. Owned by the Mozambican Railway Company (Caminhos de Ferro de Moçambique, CFM), its cargo terminals are operated by Cornelder de Moçambique S.A., a consortium formed between CFM and the Dutch Cornelder Holding B.V., which in 1998 obtained a 25-year operating concession.

With the concession up for renewal in 2024, the port has ambitious expansion plans, including the construction of a new access road that would bifurcate off from the main entrance road to Beira (the EN6) and help decongest this existing road, which is currently beset by long lines of lorries waiting to use the existing turnoff to the port at Munhava. Along the planned new road, an industrial zone of logistics warehouses would be built, making this a prime real estate development area.

One of the questions this raises is about the synergies between the port development plans and Masterplan Beira (van Weelden 2013), elaborated by a consortium of Dutch consultancy firms. Should these synergies be viewed as instrumental? Was the port expansion and new access road recommended in the masterplan to adapt the city to changing weather patterns or rather to serve Dutch commercial interests (via Cornelder de Moçambique)? Is this a case of disinterested, technical ‘best practice’ seeking to ‘avoid the human influence’, or should this be viewed as a ‘trans-local initiative aimed at securing Dutch influence abroad’ (Shannon 2019: 397)? While deliberately leaving this question open here, it should be noted that the continuities between contemporary ‘master planning’ and colonial urban and corridor development plans (Harrison & Croese 2023; Kirshner & Baptista 2023; Lazzarini 2023) are equally manifest in Beira.

The creation of a land development company, suggested by the masterplan, was put into practice in 2018, when the Beira Urban Development Company was constituted. SDUBeira (Sociedade de Desenvolvimento Urbano da Beira) was created with the support of the Netherlands Enterprise Agency (Rijksdienst voor Ondernemend Nederland, RVO) to attract investors and channel these funds into ordered urban development. Fully owned by the Beira Municipal Administration, SDUBeira is in charge of developing the Munhava Industrial and Logistics Park (along with the planned new port access road, a 900 ha concession) and the Maraza Residential Area, a 400 ha development zone wedged between the airport and the EN6 (SDUB n.d.). Yet, even at that institutionalised level, the idea of incremental building for adaptation prevails (see also the next section). SDUBeira is seeking investors for both projects but, so far, none has materialised. In the Maraza project in the Vaz area, a first plot has been prepared for the building of six apartment blocks of 10 flats each. Yet, upon visiting the site with two SDUB functionaries, not much beyond a terraced plot with minor, tertiary drainage canals is visible, with a lone guard in a tent preventing further encroachment on the area, and subterranean drainage and sewers laid, as well as piped water and electricity. As the SDUBeira website proclaims, the project ‘is being implemented in phases, allowing for short investment cycles. The pilot project is currently underway, covering an area of 14 hectares.’¹

But all these best-laid plans are moot without an influx of investment. The port is operating reasonably well, as almost all fuel imports to, and marble exports from, Zimbabwe and central Mozambique are done via Beira. But the city remains extremely vulnerable to economic downturns. This helps direct attention across scales – in the case of Beira, for example, thought needs to be given to how coal consumption in East Asia drives the boom-and-bust cycles of coal production in the inland province of Tete, directly influencing the availability of jobs and money in Beira, and thus shaping the patterns of house-building and ownership in the new expansion zones of the city. Though that is not the whole story – a new railway line was also built from Tete to the northern port of Nacala, diverting coal exports away from Beira and that, as Beirenses will tell you, is part of central government schemes to economically choke their opposition-held city.

Given the constraints detailed here, many of my interlocutors dreamt of a new economic upturn for the city of Beira and projected their hopes onto the rumoured start of gas exploration off the coast of Búzi municipality, which is just a bit further south of the mouth of the Búzi River. This, people were sure, would create a new influx of money and more jobs, and allow them to pursue their plans of socio-economic advancement, including the expansion or completion of their construction projects.

5.1. Political conflicts

Politically, the city is one of two cities governed by the opposition, which is a thorn in the side of the ruling Frelimo Party, which has governed Mozambique since independence. A brief background: Frelimo, the Mozambican Liberation Front (Frente de Libertação de Moçambique) ‘won’ independence from Portugal in 1975 after its liberation struggle (1964–1974) and installed a socialist, single-party regime. Yet, only two years after independence, the country was plunged into civil war, when the Mozambican National Resistance (Resistência Nacional Moçambicana, Renamo) took up arms against the Frelimo state. Renamo was funded by first the Rhodesian, then the South African Apartheid regime as a means to destabilise its Marxist neighbour. Yet despite portrayals as apolitical ‘bandits’ Renamo did in fact tap into and articulate political grievances of a part of the rural population especially in central Mozambique that felt alienated by Frelimo’s urban, socialist, high-modernist and ‘anti-tribalist’ vision of independent Mozambique (Pearce 2020; Wiegink 2020). The civil war also brought about mass displacement of people from the countryside to the cities, which grew rapidly and in an unplanned way (Lubkemann 2005).

After the end of the war in 1992 and the transition to multiparty democracy, Renamo’s share of the national vote continuously decreased, partly also due to Frelimo’s abuse of the powers of incumbency (Pitcher 2020). Yet, in Sofala, Renamo always had its best results at national elections, and decentralisation, though flawed, opened spaces for the opposition to govern at city level. In the case of Beira, Daviz Simango was elected mayor for Renamo in 2003, and re-elected as an independent candidate in 2008, after he was excluded from Renamo by its then leader, Afonso Dhlakama. Daviz then founded the Mozambican Democractic Movement (Movimento Democrático Moçambicano, MDM), which sought to position itself as a third force in Mozambican politics beyond the old civil war divide but failed to establish itself beyond its stronghold, Beira (Chichava 2010; Pearce 2020).

Accordingly, many of my conversations were shaded by the notion that the municipal and the national government were jockeying for money and influence in the reconstruction process, with urban residents (and especially public functionaries at the city administration) accusing Frelimo of wanting to ‘strangle’ the city. More recent administrative reforms have further curtailed the autonomy of the city government, by setting up parallel provincial-level structures that are appointed by the central government and effectively double existing functions, leading to further confusion about how government competencies are distributed between the municipal (city) and provincial levels (Forquilha 2020).

5.2. Institutional priorities

This struggle for influence and control directly shapes the set-up of reconstruction funding, as the national government created an Office for Post-Cyclone Reconstruction (known as GREPOC, by the acronym for its Portuguese name, Gabinete de Reconstrução Pós-Ciclone) to coordinate reconstruction efforts and money. GREPOC was initially planned as my entry point, to look at the effects of their programmes, especially in the area of housing. Yet, four years after Idai, there were very limited visible results of reconstruction.

Unsurprisingly, an often-heard comment on the situation was that:

All the money for Idai reconstruction has been stolen. Look around you. You don’t need a study for this. When Idai hit, they all descended on Beira, all those ministers who had never set a foot here. All the donors came, and TV crews, and the helpers in their helicopters. All the hotels were full, it seemed like a different city. Daviz Simango worked like crazy until it killed him; he also stole, but only a little, to make us believe. But those guys? They stole everything. They come and say they’ll give food to the poor, but as soon as the donor has turned his back they’ll take even that food. To steal from the poor? You really have to have an evil heart…

Yet the institutional set-up makes for a more complex (and interesting) story than the often-heard tale of corruption and inefficiency. Indeed, an assemblage of actors is engaged in remaking the city along standards of resilience, some more, some less adapted to the local context. The gist of Engineer Gonçalo’s comment, cited earlier, was that Beira was being maintained ‘against nature’, and that it was well and right for the Dutch to build their cities below sea level, as they had the money and resources to do so, but that in the case of Beira this battle against the elements was ultimately doomed to fail. Yet money was poured in à fonds perdus because there was a clear interest to maintain Beira as a functioning city governed by an opposition capable of delivering outcomes to continue presenting Mozambique as a de facto – and not only on paper – multiparty democracy.

Resilience, as Lorraine Dongo argues, serves as a ‘boundary object’ here, a malleable concept ‘that enables collaboration and dialogue between different actors whilst allowing for differences in use and perception’ (Dongo 2021: 54). Most donors and implementing agencies were following UN-Habitat’s models of resilient housing and shelter schools, though the humanist, Buddhist, Taiwanese charity Tzu Chi, for example, insisted on adorning its shelter schools with pagoda-like rooftops. Such more humorous details aside, it was probably as much institutional logics and ideology than ‘corruption’ that led to delays. The Dutch built drainage canals and coastal protection walls, because, as I was told, ‘that’s what the Dutch do’.

The World Bank, the main funder and actor of reconstruction, for example, had – unsurprisingly – opted for an ‘owner-driven reconstruction’ process. However, the bank also did not trust the same houseowners to use the cash or vouchers they would have received under such a scheme ‘correctly’, and so a long-winded public tendering selection and training process of implementing partner organisations was devised, which was still ongoing at the time of research.

More crucially, the World Bank had opted for reconstruction, not the construction of resettlement housing, and to exclude areas designated as prone to flooding in a risk map drawn up by consultants for the Japanese International Cooperation Agency (JICA), thereby excluding significant parts of the city and those residents who did not own their houses. An estimated 290,000 families were affected as per the Post-Disaster Needs Assessment (GREPOC 2019). The budget was sufficient for the rebuilding of only 15,000 houses. Accordingly, choosing the beneficiaries required a relatively laborious survey process to determine the absolutely most vulnerable of the vulnerable among the potential beneficiaries – female-headed households or those headed by orphan children, the chronically ill or living with deficiencies, etc. Such a targeted selection did, however, not make needs for support less pressing in the areas excluded by donor criteria.

The central neighbourhood of Ponta Gêa, for example, counts as a ‘noble’ area of town, where Beira’s iconic Grand Hotel stands: a looming, derelict hulk that has been occupied by people displaced by the war – who by now form an authentic ‘republic’, as people say – and a popular motif for any foreign photojournalist coming to town (Gillis 2021; Lazzarini 2023). Beyond that, the building stock of the zone is largely colonial, from airy, ‘Indian’ merchant houses and Portuguese vivendas to late colonial apartment blocks.

This is an area considered squarely within the developed city, both by donors and the city government. Yet, four years after Cyclone Idai, many roofs are still caved in, windowpanes missing or shuttered with plywood, walls pockmarked by water stains. On an exploratory walk through Ponta Gêa, I spoke with residents of those, still visibly damaged, houses, and none of them had received any assistance. All of them had comparatively secure rights, either owning their house/flat outright, either by inheritance or having bought it shortly after independence, when the state nationalised, then sold off the colonial housing stock at relatively affordable prices to the occupiers, or as durable tenancy rights in housing reserved for state employees (see also Roque et al. 2020; Sumich 2018). Yet in all of the cases household incomes were dependent on either meagre widow’s/old-age pensions or ad hoc, informal commerce and services, and depended on extra support to partially, if at all, rehabilitate their houses.

Dona Marisa and Senhor Jacinto, for example, were a retired couple who had survived Idai by sheltering with one of their children in a different neighbourhood. The cyclone lifted off the roof and flooded their top-floor apartment. Four of the eight parties pooled money to pay a contractor to redo the roof, but this ate up all of the couple’s savings. Accordingly, they now rented out their flat to a tenant, while they had moved into their lower-ground-level garage that they had rearranged into a very modest, makeshift studio. This was a windowless sleeping area (the actual garage) fronted by the access ramp, now covered with a corrugated iron sheet roof that held a cooking, bathing and dining area, together with one huge, old, buckling wooden wardrobe. They were just about surviving from the rental income from their flat, but their cooking gas had ended the previous week and they had not yet been able to buy a new bottle. While they considered themselves lucky to at least have this garage to live in, they were bitter that any money from the outside had been distributed elsewhere, as Ponta Gêa had been counted with the rich, not the needy.

That’s not to say that everyone in Ponta Gêa is destitute – I also talked to people who had the capital or income to rehabilitate or even expand their houses – but it shows how the seemingly apolitical, technical criteria of vulnerability laid down by donors can lead to exclusion of central, no less vulnerable parts of the urban population.

5.3. Who does not have a voice?

In a way, such stories of not being heard in the reconstruction process became a constant refrain of fieldwork, across different zones of the city, ranging from formal cement city neighbourhoods like Ponta Gêa and Pioneiros to ordered expansion zones like Manganhe, to more precarious informal outer bairros like Munhava Matope or Ndunda 2. Here, in addition to my own walkthroughs and encounters, I benefited from input from third-year anthropology students at Universidade do Licungo, with whom I held a half-semester practical fieldwork course and who went around interviewing residents in selected neighbourhoods of the city, helping me extend my range of coverage.

In some neighbourhoods, a few of the people we spoke to told us they had met with someone who came to survey the damages, but in most cases no levantamento (assessment, survey) had been done. Time and again, I heard that ‘the government does not see us’, which stood in direct contradiction to how my interlocutor at the National Statistics Institute (Instituto Nacional de Estatística, INE) talked about the post-disaster survey they carried out:

We wanted to make a contribution for the country. We were seeing the suffering of the povo and wanted to contribute with our work. […] It was pretty challenging […] the World Bank demanded we do not include the so-called risk areas, or protected areas. [In the survey area] we had 100% of coverage, it was phenomenal. […] after tidying up the data, we looked at eligibility. That was the difficult part. The bank had budgeted for 15,000 houses, and from the survey we saw that three to four times as many needed help. We had to redefine the most vulnerable, but even with that we were way above.

Despite these limitations, Senhor Júlio felt the survey was a success: the INE learnt a lot in the process, and it was a ‘beautiful, patriotic exercise’. Yet, when I spoke to GREPOC in April 2023, I was told the database had to be updated again, because the donors wanted to avoid ‘human influence’ in the selection of beneficiaries. When I tried to insist on finding out who were the households already identified for reconstruction as per the existing map, I met with some resistance:

I would not [try to speak to them]. People have been waiting for so long already that now one has to arrive with the project ready. Otherwise, there will be some turbulences. Only in the city of Beira, 200,000 families were surveyed, and 6,000 e tal will be considered. [I try to insist] We would not recommend it. Even for the (implementing) NGO it will be a difficult task.

This also shows how the logics of inclusion and exclusion imposed by the major donors conditioned the access to post-disaster reconstruction: this city maintained ‘against nature’ is, per the Japanese consultants’ map, considered almost in its entirety as at risk of flooding. This, according to my interlocutors at the municipal administration, misrecognises the nature of the threat: ‘in Beira, no one dies of flooding’. Flooding was minimal, compared to the rural municipalities where residents had to be evacuated by boat; the main cyclone damage was from the wind gusts and rains. As a consequence, cyclone-proofing for the disasters to come will yield very selective results, as such risk assessments not only misread risk but also misrecognise context-specific patterns of dwelling and making a living as risky.

The bairro of Praia Nova (New Beach) is a central area of the city that grew on reclaimed land that the Portuguese colonial city government had filled in in front of the central administrative and business area to protect it against spring flood. This is nowadays a fishing community that, because of its exposure to the sea, is considered one of the city’s most vulnerable areas. Here, some non-World Bank, non-GREPOC resettlement projects are under way, and some residents have been resettled to safer areas further inlands as part of earlier projects. But, while these residents may be happy to have received a new house somewhere else, their source of income is fishing, which they evidently cannot do from an interior location, so cycles of resettlement go hand in hand with cycles of precarious reconstruction in Praia Nova. However, these shacks are, from the perspective of their users, only temporary shelters for the days they are out fishing from Praia Nova, while the actual home then becomes the new place in the resettlement area.