THE MISSING PEOPLE OF STATE-SUBSIDIZED HOUSING: Lived Experiences of Non-Occupancy and Secondary Residential Mobility

Introduction

Following the slogan ‘Cities Without Slums’, initially put forward by the Millennium Development Goals (MDG) and Cities Alliance, many governments in Africa, Asia and Latin America set up new large-scale standardized housing programmes to push national development and modernization efforts (Huchzermeyer, 2011; Croese et al., 2016; Lemanski et al., 2017; Beier, 2022). Through different forms of subsidies, these programmes seek to deliver affordable and adequate housing to low-income households frequently residing in informal dwellings and shacks. Often they are designed as resettlement programmes that displace and relocate entire communities to distant peripheries, where most of the new housing stock is built from scratch and without much attention to livelihood disruptions (Coelho, 2016; Koenig, 2018). Even if people are not forcibly resettled but deliberately place themselves on waiting lists to receive state housing, this mobility still may be understood as de facto resettlement, or as a form of ‘disrupted re-placement’ (Meth et al., 2023), because of a lack of choice regarding the location and type of dwelling, as well as the likely disruption of social and economic relations due to relocation (Beier et al., 2022).

Since Turner's (1968) ground-breaking work there have been repeated concerns that such large-scale standardized housing and (de facto) resettlement programmes miss their target groups (Buckley et al., 2016; Potts, 2020). Unaffordable schemes and maintenance, underserviced sites in inappropriate locations, as well as gentrification would imply that much of the new, state-subsidized housing stock is actually not occupied by those who policy stakeholders call the ‘intended beneficiaries’. These ‘missing people’ of state-subsidized housing and resettlement programmes have either never inhabited their units or vacated their property after a period of residence—but why do they leave, and where do they go?

Mainly due to the methodological challenges of finding such resettled dwellers outside of their resettlement sites (Beier, 2023a), these questions have largely remained unanswered to date, causing accusatory speculations and divergent assumptions. In India and South Africa some policy stakeholders have lamented the deliberate return of missing people to informal settlements—the very places that were intended to be replaced by the new housing estates (Anand and Rademacher, 2011; Charlton, 2018a)—while in Morocco, politicians accused ‘intended beneficiaries’ of being profiteers who were primarily interested in quick cash-ins (Zaki, 2005: 65f). In contrast, South African housing policy has viewed the disposal of new units as a possible sign of successful enabling policy, integrating the original beneficiaries into formal housing markets (Department of Human Settlements [DHS], 2004). However, most scholars have assumed gentrification and the displacement of the ‘missing people’ to worse living conditions, with little conceptual scope for people's own agency (Lemanski, 2011; Koster and Nuijten, 2012: 193; Choi, 2015: 586).

Following these questions, this article sheds light on the ‘missing people’ of large-scale housing and resettlement programmes, conceptualizing their departure from state-subsidized housing within a people-centred, antithetic continuum of constraint and agency. Based on a reconstructive analytical approach to the housing pathways of these ‘missing people’, I argue that non-occupancy and secondary residential mobility can be understood as demand-driven practices related to coping with, resisting and adapting to the shortcomings of supply-driven housing policy. They represent restricted but mindful and people-led reconfigurations of standardized policy, reflecting both the people's own agency and severe policy constraints in terms of affordability, uniformity and displacement.

The article is comparative in nature, focusing on three of the most significant affordable housing programmes in Africa and their implementation in the capital regions of Ethiopia, Morocco and South Africa. Despite considerable differences in terms of their allocation processes, the types of dwelling and their funding schemes, all three programmes share a common agenda to fight informal housing. Taking the MDGs and the slogan ‘Cities Without Slums’ as common points of reference, they all intend to provide homeownership to low-income households residing largely outside formal markets. Field research took place in Gauteng (October 2020– April 2021), Rabat-Salé (and additionally in Casablanca) (September 2021) and Addis Ababa (March–April 2023). I conducted 101 biographic interviews with recipients of state-subsidized housing who had either never inhabited their new units or who had vacated them after some time, either to sell them or rent them out. Using Clapham's (2005) housing pathway approach as an analytical framework, interlocutors narrated their own housing trajectory and offered comprehensive insights into the residential decision-making underlying non-occupancy and secondary residential mobility.

In the following, I will first sketch out the significance of secondary residential mobility within large-scale state-subsidized housing and resettlement programmes. After briefly summarizing the state of the art, I will give particular attention to the agency of the ‘missing people’. Thereafter I present the case studies, and discuss my comparative methodological approach in more detail. Then I introduce a typology of non-occupation and secondary residential mobility, which I developed based on my empirical work and which uses the interplay of agency and constraint as a guiding continuum. Subsequently, I discuss four different types of departure from state-subsidized housing, ranging from severely constrained moves to more deliberate relocation.

Large-scale housing programmes and secondary residential mobility

The renaissance of state-subsidized, standardized and supply-driven housing programmes may be seen as grounded in both neo-paternalistic welfarism and globalized developmentalism (Civelek, 2017; Beier, 2022; Meth et al., 2023). Although being clearly linked to globally inspired, exclusionary and neoliberal urban renewal agendas (Huchzermeyer, 2011), these programmes are framed as pro-poor, promising to offer ‘better’ housing futures (more modern, more comfortable, more secure, formal, etc.) despite the imposition of relocation and a narrow focus on shelter that ignores wider functions of adequate housing (Coelho, 2016; Beier, 2022). Conceptually, the ways in which relocated citizens are conceived either as ‘beneficiaries’ (cf. Turok, 2016; Hernández Bertone et al., 2022) or as ‘displaced/project-affected persons’ (cf. Koenig, 2018; Nikuze et al., 2022) may reflect this ambiguity (Beier et al., 2022). It is also evident in divergent conceptual perspectives on social housing estate regeneration in Europe (cf. Kleinhans and Kearns, 2013; Watt and Morris, 2024). Indeed, recent studies of state-subsidized housing and resettlement programmes stress the paradoxical simultaneities and uncertain temporalities of gain (e.g. shelter quality, property titles) and loss (e.g. livelihoods, social networks), which are perceived very unevenly across resettled dwellers (Doshi, 2013; Hammar, 2017; Leitner et al., 2022; Meth et al., 2023; Beier, 2024a). Questioning the ‘settling’ character of resettlement (Fernández Arrigoitia, 2017), such time-sensitive and people-centred perspectives urge scholars to pay more attention to people's long-term, pre- and post-resettlement lived experiences (Sakizlioğlu, 2014; Wang, 2020; Williams et al., 2022; Spire and Beier, 2024) as well as different scales of place attachment (Watt, 2022). Thus, there is a need to expand analysis of housing and resettlement programmes beyond both precise localities (e.g. resettlement sites) and project frameworks that assume a clear ‘before and after’ and a singular direction of moving.

Against this background, many studies of state-subsidized housing and resettlement programmes worldwide have taken (a side) note of secondary residential mobility and the non-occupancy of new housing units by ‘intended beneficiaries’.2 In francophone research contexts the phenomenon is known as the ‘dropout’ (glissement) of intended beneficiaries from housing programmes (Le Tellier and Guérin, 2009: 662; Navez-Bouchanine, 2012: 170). In this article I suggest a more neutral approach, referring to the non-occupancy of state-subsidized housing units by their recipients at a certain moment in time after housing allocation. Since recipients are usually expected, required or even forced to move out of their previous dwellings to inhabit their allocated units, non-occupants must have found alternative accommodation at dispersed locations elsewhere. This is what I will call secondary residential mobility, which can take place either after relocation to the allocated housing unit or without any period of habitation. Secondary residential mobility can be temporary, with a potential later move or return of (parts of) the household to the allocated unit in the future (Charlton, 2018b), and it can also be time-lagged. In Ethiopia, some absentee owners of subsidized condominiums initially continued living in their original dwellings, before they were subsequently evicted and moved to a third place.

Several governments running large-scale housing programmes have regarded non-occupancy and secondary residential mobility as undesirable phenomena and as a diversion of public funds, notwithstanding some policy documents framing resales as a sign of successful upward mobility (Anand and Rademacher, 2011; Lemanski, 2011). Typically, policy stakeholders are concerned that people will return to informal settlements and contribute to a proliferation of the very neighbourhoods that the housing programmes were intended to replace (Charlton, 2018a: 2172). Politicians and administrative officials, but also ordinary people, have condemned resales as illegitimate profit-seeking or as a waste of a once-in-a-lifetime opportunity (Zaki, 2005: 65; Lemanski, 2014: 2947; Hernández Bertone et al., 2022: 77). Hence, several governments have passed clauses that prohibit the transfer of subsidized housing units within a certain period of time, although these have had limited success. As a consequence, most secondary residential development takes place in a clandestine manner, which tends to lower prices and increase risks for both purchasers and sellers (Charlton, 2018a; Garmany and Burdick, 2021; Mbatha, 2022).

However, to date there is only a limited body of research dealing explicitly with lived experiences of secondary residential mobility (Charlton, 2018b; Garmany and Burdick, 2021; Beier, 2023b; 2024b), making it difficult to evaluate such concerns and policy responses. This is mainly because of its illicit nature, and the methodological challenges discussed below (cf. Beier, 2023a). Thus, there is a significant need for a comprehensive, and especially people-centred, conceptualization of the phenomenon. So far most authors have tried to understand secondary residential mobility as an outcome of (hybrid) gentrification and downward-raiding, whereby more affluent population groups purchase or rent state-subsidized housing from the original ‘beneficiaries’ (Payne, 1997; Navez-Bouchanine, 2012: 170; Lemanski, 2014; Choi, 2015: 586; Leeruttanawisut and Yap, 2016; Civelek, 2017: 421; Croese and Pitcher, 2019: 414; Debnath et al., 2019; Garmany and Burdick, 2021; Williams et al., 2022). These authors suggest that state subsidies end up benefitting the middle class, assuming a further exclusion of ‘intended beneficiaries’ who are pressured to move (back) into worse housing conditions, typically informal settlements. Concerning Cape Town, Lemanski (2014: 2955) argues that ‘the poor are not simply sacrificing their current accommodation but potentially their only chance for decent housing.’

Regarding the reasons noted in the literature as to why people leave their subsidized units, one may differentiate between two logics, namely affordability constraints and opportunism. Regarding affordability, Pankhurst et al. (2022: 252) note that in Addis Ababa ‘some poorer households could not afford the mortgage payments and sold their apartments. … Other families rented out their apartments to cover the monthly payments.’ Furthermore, residence in new housing could itself become unaffordable over time (Restrepo Cadavid, 2010; Lemanski, 2014; Debnath et al., 2019). In Rio de Janeiro, Garmany and Burdick (2021: 2807) indicate that ‘new homes … come with new expenses … that make costs of living unaffordable. Not surprisingly, many of these families sell their new homes.’ In contrast, other studies have found that people resell their new properties for more opportunistic reasons, even if they cannot afford to stay in the same area afterwards (Payne, 1997: 21; Civelek, 2017). In Mumbai, for example, ‘cash-ins’ of subsidized housing are common: ‘With real estate values at astronomical highs for much of 2008 in Mumbai, settlers stood to gain approximately 10–20 years of their annual income by … selling their [new] flats’ (Anand and Rademacher, 2011: 1765).

Notwithstanding the validity of a ‘gentrification and downward raiding’ lens, my concern is that such a conceptual framing of secondary residential mobility plays down the (restricted) agency and lived experiences of ‘out-movers’. Lemanski (2014: 2946) rightfully remarks that gentrification and downward raiding both prioritize in-movers, while out-movers are considered ‘displaced’ and eventually disappear from both the sites and the focus of research. There is little concern about where displaced people decide to move afterwards, for which reason they choose a particular location, and how they experience their mobility—long-standing concerns with displacement research also in the global North (Atkinson 2000; Elliott-Cooper et al., 2020).3 When scholars  mention possible directions of secondary residential mobility, they frequently report assumptions that people return (close) to their former places of residence (Choi, 2015: 587; Garmany and Burdick, 2021: 2806; Hernández Bertone et al., 2022: 77). Concerning a housing and resettlement project in Luanda, Croese and Pitcher (2019: 414) report testimonies that ‘a shortage of employment, infrastructure and services initially drove many residents … to sell or rent out their [new] houses … and return to the city core’. The question of whether the returning households themselves see this move as a step backwards, or whether they feel they have made progress in life, remains unanswered.

In fact, residential decisions behind secondary residential mobility appear to be more complex than gentrification-induced displacement would assume. If one accepts that people merely ‘end up handing [new housing] over to the more affluent population’ (Choi, 2015: 587, own emphasis), there is little analytical scope for the more purposeful, strategic, and emotional dimensions of residential decision making. Therefore, I join recent calls to value the experiential dimension of displacement and resettlement (Elliott-Cooper et al., 2020; Hirsh et al., 2020; Beier et al., 2022), foregrounding people's constraints and agency behind secondary residential mobility. In doing so, I build on Clapham's (2005) analytical framework of a housing pathway to understand how people perceive and navigate changing housing contexts, and how they practice housing and attach meaning to it over time. The housing pathway approach helps to bridge structural constraints and individual agency, and can also frame residential mobility as a biographic experience of making and being at home under conditions of change (cf. Blunt et al., 2021). Here, a time-sensitive notion of home enables us to stress the more subjective, emotional dimensions that shape experiences of (partially) enforced mobility (Brickell et al., 2017), which are too often ignored by shelter-centric housing programmes. Consequently, I conceive not only primary resettlement (Beier et al., 2022) but also the experience of secondary mobility as being shaped by a simultaneity of displacement and relocation, understood as the ‘intensely felt and experiential process of un-homing’ (Elliott-Cooper et al., 2020: 498, own emphasis), and the complementary (im)possibilities of multi-scalar re-homing (Deboulet and Lafaye, 2018; Watt, 2022). From here we may begin to enquire about the reasons for non-occupation, and whether households themselves consider their secondary residential mobility to be progressive or regressive.

Case studies and methodology

Following Robinson (2022: 125), I designed the present study in a generative comparative manner treating non-occupation and secondary residential mobility as ‘shared features’ of state-subsidized housing programmes. In line with Ren's (2022) call for a more global approach to urban theory-building, I intended comparison to enable a conceptualization of non-occupation as a structural phenomenon of large-scale, supply-driven programmes beyond (but, of course, not detached from) local particularities. I selected three of the most significant housing programmes on the African continent, embodying a recent turn towards supply-driven housing policy (Croese et al., 2016). The Integrated Housing Development Programme (IHDP) in Ethiopia, South Africa's post-apartheid housing programmes, and Villes Sans Bidonvilles (VSB, Cities Without Slums) in Morocco all give priority to high-quantity, top-down production of state-subsidized, standardized housing at the urban peripheries. This ‘affordable housing’ is intended to facilitate access to homeownership for low-income households who do not own formal property yet.

The programmes’ common points of reference are the MDGs and the related international policy slogan to achieve ‘Cities Without Slums’ (Huchzermeyer, 2011: 115; MHUPV, 2012: 14; Imam and Yonas, 2018: 132). Thus, they all target a physical reduction (if not an eradication) of so-called slums and shacks, which are perceived by policy stakeholders as incompatible with national development and modernization agendas (Meth, 2013; Beier, 2022; Di Nunzio, 2022). While providing different kinds of housing—ranging from condominiums in high-rise housing estates in Addis Ababa to the provision of plots for self-construction in Rabat-Salé and basic free-standing houses in Gauteng—all three programmes are emblematic of shelter-centrism that prioritizes ‘material decency’ (Beier, 2022), while ignoring wider socio-economic dimensions of adequate housing (Huchzermeyer, 2010; Planel and Bridonneau, 2017; Imam and Yonas, 2018; Williams et al., 2022; Beier, 2024a).

Despite striking similarities among the three programmes, there are important differences concerning the precise modalities of housing provision (see Table 1). In Morocco, housing provision through the VSB programme is effectuated by direct resettlement alone (World Bank, 2006; Bogaert, 2018; Beier, 2024a; 2024b). The VSB programme exclusively targets households who reside in bidonvilles (shantytowns), irrespective of their income. They are asked to demolish their original dwelling and then enter a relocation scheme that resembles sites-and-services approaches. In the observed project in Salé, households pay for half a subsidized plot on which they must then construct a three-storey house together with another relocated household. Given that state-backed mortgages remain largely inaccessible (Le Tellier and Guérin, 2009), households have to bear the costs of construction themselves.4 To diversify the sample, I also observed a case in Casablanca, where resettled households received the right to purchase a subsidized apartment.

In Ethiopia's IHDP programme (UN-Habitat, 2011; Planel and Bridonneau, 2017; Pankhurst et al., 2022), housing provision in Addis Ababa takes place through direct resettlement of shack dwellers and also via a lottery system to which citizens may apply if they do not own property. Depending on their income level and capacity to save up, households may choose to register for condominiums of different sizes ranging from studios to three-bedroom units. Once their name is drawn, they have to use their savings for an initial down payment (minimum 10% of the total price of the condominium) and then pay monthly mortgage instalments for 15 to 20 years.

Among the three programmes under investigation, South Africa's housing scheme is the only one that provides housing units to eligible persons free of charge (Meth, 2013; Huchzermeyer and Karam, 2016). Households who do not own property and earn less than ZAR 3,500 per month (approximately EUR 175) may sign up to so-called ‘waiting lists’, through which they might receive – often after years of waiting – a basic, free-standing, often one-roomed top structure colloquially known as a ‘RDP house’, alluding to the post-apartheid Reconstruction and Development Programme. While allocation through waiting lists is common, in Gauteng the direct resettlement of entire communities as well as the in-situ provision of RDP houses take place as well.

In all three programmes, non-occupancy by recipients and secondary residential mobility are significant phenomena that have featured prominently in housing policy debates. Whereas government actors in South Africa have rather shown a hostile attitude towards any kind of non-occupation (Charlton, 2018a; Beier, 2023b), the Ethiopian government has seemed more tolerant towards people renting out state housing units (UN-Habitat, 2011: 38). In Morocco, high levels of glissement have led to policy change, partially towards more repressive modes of implementation ensuring the definite demolition of shacks, and temporarily towards more inclusive financing schemes (Le Tellier and Guérin, 2009; Beier, 2021; 2024b). In fact, all the programmes include clauses that prohibit the resale of received housing units within a certain period of time after allocation (five to eight years).

A typical challenge for the study of non-occupancy and secondary residential mobility is locating, contacting and interviewing former dwellers. As an ‘unconsciously hidden population’ (Beier, 2023a), the ‘missing people’ of state-subsidized housing programmes live at dispersed locations across the entire metropolitan area. Moreover, even if it is possible to locate them, they might be hesitant to talk due to the (perceived) illicit nature of their non-occupancy (Lemanski, 2011; Garmany and Burdick, 2021). For this research, therefore, I developed and applied insider-driven snowball sampling techniques that could effectively counteract these challenges for sampling in terms of both hidden location and reluctance to talk.5 Thanks to my research assistants and our context-specific adaptations of sampling procedures, it was possible to realize a total number of 105 narrative biographic interviews, of which 101 fulfilled all the criteria to be considered ‘missing people’ for this study.6 Of these interviews, 35 took place in Addis Ababa (March–April 2023) and 27 in the Gauteng City Region (GCR) (October 2020–April 2021). Of the 39 interviews conducted in Morocco in September 2021, 29 took place in Salé, mostly with interlocutors who had sold their resettlement plots. To increase diversity my colleague Manon Troux conducted ten additional interviews with people in Casablanca, who had received apartments as resettlement compensation. Table 2 offers an overview of all the interviews, showing that the case studies offer a diversity of allocation practices and forms of (de facto) resettlement. Before embarking on a more detailed discussion below, it is clear from Tables 1 and 2 together that specific modes of implementation (e.g. access free of charge) influence the type of departure (renting out vs. selling) and whether recipients have previously occupied their units.

The narrative interviews were designed to incite respondents to talk freely and to narrate their personal ‘housing pathway’ (Clapham, 2005). If an interlocutor was hesitant about talking freely, we used basic follow-up questions to explore their past (and future) reasoning behind moving house. This holistic approach to enquiry acknowledges the complexity of moving decisions, reflecting biographic experiences (including aspects of family, gender and age) and wider socio-economic contexts (including employment and social conflicts). Except for six interviews in the GCR, one in Addis Ababa and those that took place in Casablanca, the author was present for all interviews, which lasted around one hour on average. They were conducted in several local languages as selected by the interviewed person, with the assistance of experienced research assistants fluent in these languages. The interviews were transcribed and translated into English (ETH, RSA) and French (MAR) and analysed thematically using MAXQDA. Schematic representations of each elicited housing pathway allowed for a sequential and reconstructive comparative analysis, which was helpful in developing a typology across the diverse case studies.

The ‘missing people’ between constraint and choice

From the comparative analysis of the interviews, I derived a typology organized around the extent to which agency and constraint shape people's decisions not to inhabit state-subsidized housing units. Recognizing ambiguous conceptualizations of housing and resettlement programmes in both the global North and South (Kleinhans and Kearns, 2013; Deboulet and Lafaye, 2018; Hammar, 2017; Beier et al., 2022; Meth et al., 2023), the selection of agency and constraint as key variables goes beyond a mere displacement lens, which tends to foreground involuntary experiences of ‘un-homing’ (Elliot-Cooper et al., 2020; Watt and Morris, 2024). Acknowledging people's more or less limited power and desire to shape their own housing futures, this typology is also able to account for people's attempts at and practices of ‘re-homing’ (Deboulet and Lafaye, 2018; Watt, 2022; Meth et al., 2023). In this sense, agency and constraint are seen as two inversely related variables that shape each moving decision to a varying degree. Along the heuristic continuum shown in Figure 1, one may distinguish four positions representing different categories of types of departure: 1) unwanted departure, 2) departure as temporary setback, 3) departure as restricted preference, and 4) departure to ‘move up’. It is important to note that these are not rigid categories; rather, they are overlapping and fluid. For example, over time a particular respondent's decision-making may align with more than one type. Nevertheless, they are helpful tools to explore the spectrum of non-occupation and secondary residential mobility.

Conclusion

Non-occupation of state-subsidized housing and secondary residential mobility are important phenomena—both in terms of quantity and quality—affecting large-scale housing and resettlement programmes in Africa and beyond. However, common conceptual lenses rooted in gentrification theory or downward-raiding are biased towards the idea of displacement. They tend to restrict reasons for non-occupation to notions of coercion and opportunism, while struggling to account for the agency of people who are ‘missing’ from resettlement sites. Moreover, they are unable to account for the diverse destinations of secondary residential mobility, and the complex decision-making behind it. Given the wide range of literature concerned with the deficiencies of the peripherally located and under-serviced sites chosen for standardized housing programmes (Buckley et al., 2016; Koenig, 2018; Williams et al., 2022), it seems unsatisfactory to reduce explanations for departure from state-subsidized housing to households’ mere inability to finance long-term residency. Instead, this phenomenon should be conceived as a multi-dimensional demand-driven response.

Because of the outlined shortcomings and based on a generative comparison of biographic narratives of non-occupants, I have suggested a people-centred typology differentiating four positions along an antithetic continuum of agency and constraint (see Figure 1). However, this is not intended to downplay the relevance of downward-raiding and displacement. Indeed, for several respondents non-occupation predominantly represents an experience of un-homing. Nevertheless, other respondents deliberately choose not to occupy their allocated state-subsidized housing in order to re-home at places of belonging and/or desire. Therefore, the suggested typology goes beyond a singular emphasis on displacement, offering a much-needed complementary perspective that combines a conceptual emphasis on experiences of un-homing and re-homing (Beier et al., 2022; Meth et al., 2023). Thus, in line with literature exploring life after displacement and housing provision (Kleinhans and Kearns, 2013; Lemanski et al., 2017; Deboulet and Lafaye, 2018; Charlton, 2018b; Wang, 2020; Watt, 2022), it enlarges the temporal frame of analysis and looks beyond a single site. Most significantly, it helps us to conceive of people who depart from state-subsidized housing as active subjects who reconfigure supply-driven, shelter-centric housing policy according to their own, future-oriented demands, while at the same time being affected by different degrees of financial constraints.

These financial constraints often relate to the direct costs of housing acquisition, which appear to have the most significant impact on the extent of non-occupation in a particular setting. While Morocco's non-inclusive resettlement scheme produced a high number of ‘unwanted departures’, these were less common in South Africa where houses were provided for free. At the same time, the findings stress the importance of decent labour (Potts, 2020), clearly showing that a higher degree of agency—i.e. a higher capacity to direct one's own housing pathway—is strongly linked to more secure forms of employment. Thus, the conceptual emphasis on restricted but mindful and people-driven reconfigurations of standardized housing policy intends to balance the structural and individual factors driving non-occupation. As such, some forms of non-occupation appear as individual strategies to resolve structural constraints (e.g. location, mortgage costs) and to sustain future residency within state-subsidized housing—most commonly in Addis Ababa and Gauteng (Charlton, 2018b; Pankhurst et al., 2022).

In terms of the destinations of secondary residential mobility, my findings may partially confirm concerns about downward-raiding and returns to inadequate housing. However, from the perspective of the ‘missing people’ this should not automatically be seen as regressive, or solely as an enforced form of residential mobility. Rather, most respondents aligning with types two and three make better use of state-subsidized housing if they do not occupy it, instead renting or selling it. They tend to be better off in terms of living conditions than before they received state-subsidized housing because of new rental income, more security, and/or better housing comfort. Geographically, it is also too simplistic to assume displacement towards the urban peripheries. Paradoxically, some destitute respondents who could not access state housing squeezed into shared accommodation in central locations, while some respondents from the lower middle-class moved even further towards the periphery after they left state-subsidized housing. Thus, a close analysis of secondary residential mobility may suggest a conceptualization of departure from state-subsidized housing not only as a form of displacement, but also as a possibility for people to introduce elements of personal choice into ill-designed and inflexible supply-driven housing policy.

Consequently, I argue that the sale and letting of received units should not be sanctioned. Temporary resale bans or occupancy audits impede people's own pragmatic efforts to overcome the unaffordability of state-subsidized housing, and prevent homeowners from deciding for themselves how to use their assets in their best way under given conditions. Here, my argument aligns with Standing's (2011) critique of conditional cash transfer, as well as Mbatha's (2022) concerns with black-market sales of state housing. In fact, the findings show that non-occupation tends to represent the outcome of a careful, often strategic consideration of different options under considerable financial and structural constraints. Thus, I suggest that people should be free to make their own informed decisions about whether they prefer to occupy allocated units or not.

However, as noted by Audycka (2023), such a ‘right to decide’ should be inextricably linked to the aim of housing programmes to keep the extent of non-occupation as low as possible through inclusive and affordable schemes, limited geographical and social disruption, and opportunities for spatial and economic appropriation at the new sites. Nevertheless, I argue that a recent (re)strengthening of authoritarian and destructive urban policies in Morocco and Ethiopia (Di Nunzio, 2022; Beier, 2024b) is likely to impair secondary residential mobility, raising significant doubts about whether authorities are actually interested in facilitating progressive housing pathways for low-income citizens.