Real Estate and Late Urbanization in Kigali and Addis Ababa

Urban Fortunes and Skeleton Cityscapes: Real Estate and Late Urbanization in Kigali and Addis Ababa

First published: September 2017   Full publication history
DOI: 10.1111/1468-2427.12550    View/save citation
Cited by (CrossRef): 0 articles Check for updates

I am grateful to John Henneberry and three anonymous IJURR reviewers for the helpful comments and suggestions on the text.


In many parts of Africa, societies that remain primarily rural are experiencing accelerated urban growth and highly visible booms in property development. In the absence of significant industrialization, investment is pouring directly into what Lefebvre and Harvey characterized as the ‘secondary circuit’ of capital. Debates about the drivers of investment in real estate are longstanding in relation to the global North, but have given little consideration to low-income and late-urbanizing countries in Africa. Yet such contexts offer important opportunities to reflect on existing theory. Focusing on Kigali and Addis Ababa (both transformed virtually beyond recognition over the past two decades), this article explores the drivers and consequences of investment in urban real estate in countries striving to structurally transform their economies. It argues that a range of formal and informal incentives and constraints have led to high-end real estate being viewed as the ‘safest bet’ for those with resources to invest, even where demand is limited and governments are promoting other kinds of investment. While some people are reaping urban fortunes in largely untaxed rents, much of the construction is purely speculative, creating landscapes of unused and underused high-end properties in contexts where investment is desperately needed elsewhere.


(onlinelibrary) –Eastern Africa is the last world region to undergo the urban transition. Nowhere else is there a cluster of countries with such low urbanization levels––notably Burundi, Rwanda, Uganda, Ethiopia and South Sudan, all of which are still less than 20% urbanized.1 These societies are, however, currently experiencing accelerated urbanization and urban growth, accompanied by highly visible booms in property development and construction. The notion that real estate should be both a driver and symbol of urban development seems obvious today; yet little attention has been devoted to the dynamics of investment in urban land and property that are shaping urban transitions in some of the world’s poorest and latest-urbanizing countries. At a time when visible (and especially vertical) physical transformation is often equated with modernization and development, these processes demand interrogation.

Countries in Africa comprising the ‘final frontier’ of global urbanization and property development (Watson, 2014) constitute important sites in which to explore the limits and contemporary relevance of existing urban theory. The extensive literature on ‘late development’, and what this means for countries’ potential to economically specialize and prosper in a globalized world, has not been matched by sufficient attention to the implications of ‘late urbanization’ in this same global context. There is an urgent need to better understand the drivers and characteristics of urban property development in countries where real estate is a relatively new sector but is rapidly remaking cities with long-term implications.

It is now widely known that urbanization and industrialization do not correlate in the contemporary global South as they did in the period of Northern industrialization (Fox, 2012; Gollin et al., 2016). Recent research shows that in many developing countries urbanization is characterized by the rise of ‘consumption cities’, with different characteristics from ‘production cities’. While consumption cities are richer than production cities overall (particularly where consumption is driven by natural resource exports), they have higher slum populations and their relative wealth does not translate into improved wellbeing ‘to the same degree [as] an income boost through industrialization’ (Gollin et al., 2016: 37).2 This may be partly because the service-based economic strategies that many late-developing countries are turning to––given that they are ‘running out of industrialization opportunities’ (Rodrik, 2015: 24) faster than their predecessors––do not ‘have the capacity to absorb––as manufacturing did––the type of labor that low- and middle-income economies have in abundance’ (ibid.). Urbanization without industrialization thus raises important questions about the very nature of cities now and in the future.

Having been debated for decades by urbanists working on the global North and some highly urbanized parts of the South, the pertinence of debates between Henri Lefebvre, David Harvey and Manuel Castells for poor but rapidly urbanizing parts of the world has been largely overlooked. Yet these contexts breathe new relevance into the debate. Lefebvre (1970: 15) believed that after a certain point in history, urbanism itself ‘becomes a productive force, like science’ and that the city ‘is a place for creation and not simply a result, the simple spatial effect of a creative act that occurred elsewhere’ (ibid.: 28). The central critique levelled against him by Castells and (more sympathetically) by Harvey was that the urban was just the spatial embodiment of capitalist industrial development and therefore epiphenomenal to industrialization. This line, maintained for decades by Harvey, was persuasive with respect to the developed world and some large emerging economies such as those of the BRICS countries (Smith, 2003: xx). Yet it neglects to consider dynamics unfolding in some the latest-developing and fastest-urbanizing parts of the world––precisely those areas that Lefebvre’s (1970) speculations in The Urban Revolution implicitly urge us to consider. Empirically, Lefebvre is being vindicated by developments in places such as Eastern and Central Africa where urbanization and urban property development have gathered pace in the virtual absence of industrial development, and are themselves shaping economies and societies.

This article explores these issues by examining two African countries becoming renowned for their ambitious developmental visions as well as their transforming urban skylines: Rwanda and Ethiopia. These countries are not yet developed enough to even be considered ‘emerging’ or ‘frontier’ market economies in the literature on real estate (Abdulai et al., 2015: 4). They also fall within the 5% of countries globally that had an urbanization level below 20% in 2011. In both cases, however, this is changing rapidly: according to UN figures, Rwanda had the world’s fastest urbanization rate from 2010 to 2015,3 with Ethiopia not far behind in joint-seventh place (UNDESA, 2014).4 There is little by way of accompanying industrialization in either, though efforts in this regard are accelerating. Instead, much of the capital flowing into these countries, often in the form of international aid and remittances (neither have major natural resources to export), is coalescing around urban land and real estate: what Lefebvre (1970), Harvey (19781985) and others have termed the ‘secondary circuit of capital’. This is reflected by the fact that, in both cases, growth in real estate and construction significantly outstrips overall economic growth.5 Given the concerns raised by these theorists about over-reliance on this type of investment, and the questionable prospects for equality and inclusion in ‘consumption cities’ (Gollin et al., 2016), it is important to understand what lies behind this headlong leap into the ‘secondary circuit’.

Rwanda and Ethiopia provide something of a natural experiment. In the former, real estate has been actively stimulated; in the latter, the reverse is true and property developers have had a troubled relationship with the state. This article explores how the real estate sector has been built almost ‘from scratch’ in both countries over the past two decades (given the legacy of the Rwandan genocide in 1994 and communism followed by civil war in Ethiopia pre-1991). In both capital cities (Kigali and Addis Ababa), real estate has rapidly become a central pillar of the urban economy, albeit in rather different ways. I argue that, despite the Ethiopian government’s stated objections to private real estate-led development and its commitment to manufacturing industry, informal incentives alongside obstacles to other investment render real estate equally (if not even more) magnetic than in Rwanda, where the government has encouraged it. This has important implications, suggesting a very powerful propensity under current global conditions for capital to coalesce specifically around high-end real estate, even where this becomes divorced from the level of demand in low-income settings that urgently require more productive forms of investment.

This article is based on interviews with stakeholders in Kigali and Addis Ababa from June 2014 to January 2015, including property developers, house owners and renters, estate agents, government officials, international advisors, construction firms and architects. This is complemented by analysis of government documents together with investment and taxation data. The article proceeds by exploring the relevance of theoretical debates on real estate investment for low-income countries today. This is followed by an overview of the structural origins of the property boom in both cities, leading into a discussion of the incentives and constraints (both formal and informal) in the real estate sector in each. We then turn to the outcomes of these incentive structures, which have produced substantial real estate investment (and wealth) in both cases. The character of investment differs in important ways between the two cities, yet in both it is out of step with demand and highly problematic in terms of meeting the populations’ needs both now and in the future. The article concludes with a discussion of why property investment is seen as the ‘safest bet’, reflecting on the implications not only for these cities but broader understandings of ‘late urbanization’. The article thus contributes to the broader literature on African, Southern and comparative urbanism (Pieterse, 2010; Myers, 2011; Roy, 2011; Robinson, 2015), as well as literature more specifically concerned with dynamics of real estate in African cities (Murray, 2011; Obeng-Odoom, 2015).

The ‘secondary circuit’ of capital and property development in Africa

The contemporary development of countries like Rwanda and Ethiopia poses a paradox with respect to Marxist urban theories. For Lefebvre, the flow of capital into urban land and real estate was a symptom of advanced forms of urbanization: societies that have been through the ‘urban revolution’, where cities no longer appear as islands ‘in a rural ocean’ (Lefebvre, 1970: 11). In such societies: ‘Real estate functions as a second sector, a circuit that runs parallel to that of industrial production … This second sector serves as a buffer. It is where capital flows in the event of a depression’ (ibid.: 159). Harvey (1978: 107) further developed these ideas, arguing that overaccumulation in the primary (industrial) sector leads to the ‘switching’ of capital to the secondary sector. The paradox is that, even though this advanced industrial urbanization has not been reached in societies such as Ethiopia and Rwanda (where most people are still rural farmers), capital is rapidly coalescing around urban land and real estate as if they were advanced urban societies.

Globalization is clearly relevant here; the flow of capital straight into the secondary circuit in the South is predicated on a history of industrial growth in the North and can be linked to what Harvey (2001) terms the ‘spatial fix’, whereby over-accumulated capital seeks new geographical terrain. Lefebvre (1970: 160), meanwhile, was prescient enough to note that ‘It can even happen that real-estate speculation becomes the principal source for the formation of capital … The second circuit supplants the first, becomes essential’. What he may not have considered is that cities which were still effectively ‘islands in a rural ocean’ might become focal points for real estate investment, as capital ‘hops’ rather than flows into strategic sites in less-developed parts of Africa (Ferguson, 2006). Moreover, as global growth slows in the contemporary setting, returns on capital are likely to further increase (Piketty, 2014) and developing-country cities offer interesting opportunities for those ‘who are looking to add some risk to their portfolios’, due to the potential rewards that such risky investments offer (Abdulai et al., 2015: 3). In this context ‘geographical switching’ as well as ‘sectoral switching’ is likely to increase (Kutz, 2016).

The specificity of cities experiencing major real estate development without a backdrop of industrial urbanism nevertheless poses challenges for dominant paradigms in urban studies. Much late-twentieth-century urban theory is predicated on assumptions about the shift from industrial to post-industrial cities, and how the latter have both been shaped by and destroyed the former. The spectre of an industrial past is hardwired into much urban theory. More recently, however, scholars have questioned Harvey and Lefebvre’s assumption that investment in the secondary circuit is necessarily linked to crises and overaccumulation in the primary circuit (Charney, 2001; Gotham, 2006; Aalbers, 2007). These authors instead see the real estate sector as ‘a conceptually separate and analytically distinct circuit of capital investment’ (Gotham, 2009: 359). Aalbers (2007: 174) notes that the flow of capital into the secondary circuit is not only a consequence of economic crisis but ‘partly a sign of the intrinsic opportunities that the built environment provides’. It has intrinsic appeal because it can be a superior hedge against inflation, diversifies investment portfolios away from stocks and bonds, and offers greater tax benefits compared to other kinds of investments (Charney, 2001: 742).

In most developing countries this relative appeal is even greater, for foreign and domestic investors alike. As well as the difficulty of making the primary sector profitable (Ghani and O’Connell, 2014; Rodrik, 2015), inflation rates tend to be high, while stock and bond investment options are limited or non-existent. Thus recent research in Uganda has shown that, ‘instead of depositing money in bank accounts or investing in instruments such as shares and bonds, most individuals invest in physical structures, such as land and buildings’ (Kangave et al., 2016: 11). Evidence suggests this investment is disproportionate to the use value of such structures (ibid.). Moreover, as will be explored below, tax benefits also play a significant role––though less through formal tax incentives than through the weak capacity of existing taxation systems. There are thus reasons to consider an intrinsic link between conditions of late development and the enhanced pull of the secondary circuit, though this is rarely examined explicitly in the literature.

The global North focus of mainstream property development literature also involves assumptions that do not travel well about the availability of information on real estate values and transactions (see e.g. Archer and Ling, 1997). In the literature on ‘property market maturity’, even markets in Italy and Spain are considered ‘immature’ or ‘emerging’ by conventional criteria (Keogh and D’Arcy, 1994). Yet the very foundations on which these criteria are based are of limited relevance in low-income countries with nascent real estate sectors, where most property transactions are unregistered and construction often informal (Kusiluka et al., 2012). Most property research involves taxonomies of agents with clear-cut roles such as landowner, speculator, investor, developer, valuer and so on, and established systems of interaction between them (Haila, 1991). These institutionalized networks of actors are often absent or highly underdeveloped in late-urbanizing low-income countries. This is not to say that agency is insignificant in the property sector. As elsewhere, the interaction between agents’ strategies and structural incentives is crucial to understanding urban property dynamics (Healey and Barrett, 1990; Gottdiener, 1994; Soares de Magalhãs, 1999). Exploring contextual variables that influence property investment in these places is crucial; as Beauregard (2005) has pointed out, market logic is ‘thin’ but property development itself is ‘thick’––concrete, textured and place-specific. An important general point is that, in late-urbanizing countries with nascent property sectors, the roles of agents may be blurred and property relationships characterized to a significant degree by informal socially embedded brokerage (Obeng-Odoom, 2015).

The sphere of urban property development is thus guided by a different range of institutions and ‘place entrepreneurs’ (Molotch, 1993) in contexts where the urban institutional environment is relatively highly informalized (Simone, 2004; Roy, 2011; Kusiluka et al., 2012; Obeng-Odoom, 2015). The shaping of urban fortunes everywhere is crucially affected by government regulations and institutions (Logan and Molotch, 2007). Yet in late-urbanizing low-income countries with relatively weak state implementation capabilities, it may not be the formal regulations themselves but the very fact that they are weak and compromised that most shapes the investment environment. Investors can therefore realize substantial profits ‘in places that have weak or non-existent state requirements for developers to contribute to infrastructure costs or to share in the land value increases resulting from real estate development’ (Palmer and Berrisford, 2015: 5). Under current global conditions, real estate can thus become especially appealing in places where economic growth is creating new property markets, yet the institutions that might share the benefits of property investments (thereby making them less profitable) are still weak. This is what makes tax such a significant consideration. Hence in a country like Myanmar, opening up suddenly to international capital, property prices are soaring to such an extent that ‘anyone with the funds to do so is investing in real estate’ and the government is now considering a range of property-related taxes for fear of the impact of this on other forms of saving and investment.6

The supposed incentive effects of particular property-related taxes have been extensively theorized by economists regarding their possible effects on the efficiency of land use and decisions about where to locate buildings (Bahl and Linn, 1992; Netzer, 2003). However, the broader question of how property taxes or their absence might affect the overall pattern of investment, and decisions to invest in property vis-à-vis other sectors, is largely overlooked. There is reason to believe (as the government of Myanmar evidently does) that the presence or absence of effective property taxation plays a role in the overall incentive structure for investment. So too do many other factors beyond the official policies specifically intended to guide investment in particular strategic directions. In order to understand why investment is concentrating so intensely on urban real estate in Rwanda and Ethiopia, but also why this investment takes different forms in each, close attention needs to be paid to the informal as well as formal incentive structures, and how these interact with the practices of agents involved in creating real estate.

Building a real estate sector from scratch

Beginning with Rwanda, this section briefly discusses the background to the property boom in each case and how it relates to the current context for real estate development. The extent of civil war devastation, a rapid inflow of international resources and various waves of refugee influx post-1994 are critical to understanding the investment context in Kigali.7 In the final years of the last millennium, Kigali’s population grew at an astonishing 18% per annum as millions of returnees poured into the city and the Rwandan Patriotic Front (RPF) consolidated security in the city (Goodfellow and Smith, 2013). Large amounts of foreign aid were disbursed, coming on top of a US $2 billion debt cancellation in 2002 (former government official, interviewed 5 June 2014). Meanwhile, a new generation of (often foreign-born) Rwandans were increasingly interested in owning their own homes and, in a decimated economy, the channelling of resources into the built environment was inevitable. There was ‘aggressive infrastructure rollout’ by the state, supported by donor funds, ‘opening new areas of the city for development’ (estate agent, interviewed 9 June 2014).

It is important to note that, at this time, there was virtually nothing by way of a property development sector; the only agencies engaging in real estate development at scale by 2005 were one state-owned enterprise and the Rwanda Social Security Fund (discussed further below). For the most part, houses were developed by individuals who contracted small building firms. This included a significant number of ‘high-end’ residential properties, financed by elites, which were usually built to be rented out. In the context of heightened international donor and NGO presence, there was by the mid-2000s a booming rental real estate market largely driven by international personnel. Much of this was in the ‘new areas’ of the city (opening up at the north and northeast urban periphery).

Addis Ababa has a distinctive history due to Ethiopia’s experience of imperial (rather than European colonial) rule until 1974. A real estate market of sorts operated under the emperors, whereby dignitaries possessing large land concessions gradually sold parts of these off to the growing administrative and commercial classes, who in turn constructed at the rear of their lots to offer buildings for rent to new urban dwellers (Duroyaume, 2015: 398). Hence the city became characterized by an unusual pattern of ‘sumptuous dwellings alongside hovels of wattle and daub’ (ibid.: 398). This real estate market was, however, effectively terminated when the communist Derg regime took power in 1974. As well as all land, most housing was nationalized and the imperial urban structure was ‘frozen’ (ibid.: 400). On overthrowing the Derg in 1991, the Ethiopian People’s Revolutionary Democratic Front (EPRDF) government was initially uninterested in cities. Existing housing was neglected and, as the state retained land ownership, there was no formal property market to speak of. In the 2000s, however, this began to change as economic growth accelerated and homeownership became an important government objective.

As the EPRDF developed a land-leasing policy, initiated in 1993 but altered over time (see Adam, 2014), private property developers came to play an increasingly important role. In the early 2000s, loans were made available for the purpose of real estate development, as was cheap land (property developer, interviewed 30 September 2014). Ayat Real Estate was the first firm to offer suburban villa complexes in the eastern suburbs of Addis Ababa, sold at prices affordable for the upper and middle classes. Other firms followed suit. Meanwhile, as the focal city for the African Union and other international agencies, Addis Ababa’s international community grew to sizeable proportions, necessitating both office space in central areas and higher-end housing in the new suburbs. New forms of socio-economic segregation thus began to emerge which were new in Ethiopian history. Finally, the role of the diaspora has been crucial. Ethiopia’s diaspora numbers an estimated 1.5–2 million people (mostly living in the USA), who provide an estimated gross annual income of US $20 billion to the country (Lefort, 2015: 368). After the 2008 financial crisis, diaspora Ethiopians were even more likely to invest in property development ‘back home’ in Addis Ababa (construction firm representative, interviewed 29 September 2014).

Despite these important driving factors in the wake of conflict and post-socialist transition, there are clear reasons to believe that property development has taken on dynamics in both countries that are divorced from the level and structure of demand. As demonstrated below, the incentive structures in both countries have (whether intentionally or not) led to real estate absorbing a disproportionate amount of resources in both Kigali and Addis Ababa, to the detriment of other possibilities for productive investment.

Incentives for (and constraints upon) property development

Rwanda has various formal incentives pertaining to real estate and construction, in line with the prioritization of urban infrastructure development in its Vision 2020 strategy, which also sets out the country’s objective to be 35% urbanized by 2020. One way in which investment in real estate and infrastructure has been encouraged is through generous incentives in the construction sector, including a flat tax of just 5% levied on the import of building and finishing materials for projects worth US $1.8 million or more (investment official, interviewed 3 June 2014). Kigali’s city government has also made concerted efforts to encourage the sector, particularly since the establishment of a ‘One Stop Centre’ for would-be property investors in 2010. This centre (inspired by Singapore) aims to ensure that investors in large real estate projects obtain all necessary documentation within 30 days.8

Meanwhile, the Rwandan Development Bank has a mandate to facilitate financial tools for real estate development (Republic of Rwanda, 2015a) and several banks now offer commercial mortgage schemes. These are only accessible to a small minority of the population; an estimated 19% of Kigali’s population earn enough to qualify for a mortgage, while in Rwanda as a whole the percentage is ‘unknown but far lower’ (Republic of Rwanda, 2015b: 18). Nevertheless, for those who are eligible, mortgages have become popular, both among people seeking to build their own houses and those acquiring existing ones. Up to 70% of the loans made by some banks are mortgages (military bank representative, interviewed 19 January 2015). For many among the upper and middle classes, obtaining a mortgage is much easier than other forms of finance, creating incentives to invest in building additional houses for rental. As one individual pointed out, ‘I invested in housing because I had few options. I can access a mortgage, but not a loan’ (civil servant, interviewed 4 June 2014).

Alongside these various measures to stimulate the sector, the government has been directly involved in real estate development through the Rwanda Social Security Board (RSSB), which has a strategic aim of investing 25% of its funds in property. To date, however, this has been extremely high-end in focus. Most striking of all is the current Vision City mega-project (reputedly the largest real estate project in East Africa) comprising 500 variable-sized units, none of which can be considered anything but luxurious in the Rwandan context. The most expensive units are expected to sell for US $560,000, with more modest ones priced at US $370,000. The project, which builds on some smaller RSSB-funded real estate initiatives ostensibly aimed at ‘middle-income’ groups but only affordable to the rich, was viewed with almost universal scepticism by all non-RSSB stakeholders consulted. For the RPF-led government, however, the intention is for RSSB to lead the way in positioning Kigali as a prime site for real estate development: ‘Vision City is intended as a landmark––to provide an inspiration, a model to stimulate more high-quality real estate investment’ (RSSB representative, interviewed 3 June 2014). As with many other cities worldwide, the influence of specific economic and aesthetic conceptions of the ‘global city’ is clearly discernible in this project.

Arguably even more important than these formal efforts to incentivize and ‘inspire’ real estate are other policies and practices which act as informal incentives to build. Among the most significant are those relating to property-related taxation. Of particular importance here is the situation regarding rental income tax. In a very unusual move, the Rwandan government decided in 2001 that, rather than taxing rental incomes through the Rwanda Revenue Authority (RRA), rental income tax would be decentralized to districts, leaving the RRA to concentrate on business incomes. This shocked observers, given that Rwanda’s district governments severely lacked the institutional and technical capacity to collect such taxes. The tax was subsequently re-centralized in 2014. However, given the lack of district capacity, this created a 13-year window of opportunity at a crucial time in Rwanda’s development for people to reap rewards from soaring and largely untaxed rents. In many cases, property rentals (as well as sales) take place through informal brokers and middlemen, forms of ‘place entrepreneur’ that have inevitably flourished given the profits in the sector but which further exacerbate the opacity of property transactions, rendering taxation extremely challenging. By 2010, a donor report noted that ‘the most important single source of untapped revenues in the country is rental incomes from ownership of urban commercial and residential property, above all in Kigali’ (DFID, 2010: 15). It is doubtful that the rush to build high-end properties would have been so intense had those rents been taxed by the relatively effective RRA, rather than left virtually untouched at local level.

In addition, the existing system of ‘fixed asset’ (property) taxation itself is exceedingly weak (see Goodfellow, 2017), reinforcing the appeal of developing high-end property. Many people are not even aware of property tax, and it barely figures in the thinking of the Rwandan Development Board (investment officials, property developers, estate agents and property owners, interviewed various dates, 2014–15). Despite the kinds of rational-choice-based theorizing discussed by Bahl and Linn (1992), it is difficult to predict the effect a tax would have on the property market in practice. When asked about the possible impact of more efficient property taxation, most sources involved in real estate believed it would have a significant effect of some kind. In terms of the lucrative rental market (discussed below), some observed that the tax would probably cause rents to increase, thus pushing renters down into less valuable properties and creating a squeeze on the housing market (architect, interviewed 3 June 2014). It ‘might affect the incentives for people to construct at all, as there are so many costs’ (estate agent, interviewed 9 June 2014). An alternative potential ramification was that developers would ‘obviously’ have to start building to a ‘lower standard’ (property developer, interviewed 20 January 2015). How this would affect rents (and housing affordability) is difficult to say in the abstract, because it would depend on the profit margins developers are willing to accept and the nature of demand. Unless a more effective and progressive property tax system is developed, one can only speculate on its potential effects; yet the continuing absence of such taxation is another clear manifestation of the lack of disincentives to invest in high-end real estate, even if that is not where capital investments are most needed.

Despite all these factors incentivizing real estate investment, there are enormous constraints upon construction in Rwanda. Materials are very expensive (with 60% imported from abroad). Moreover, despite the nascent mortgage market, access to finance is highly constrained as interest rates are so high. One individual described having ‘stacks of land all around Kigali’ but not being able to use much of it due to prohibitive 18–20% interest rates (property developer, interviewed 21 January 2015). The huge costs faced by developers mean that they expect very high profit margins of at least 20–40% (urban planner, interviewed 8 June 2014). Ironically, the costs of construction further push people towards the top end of the market, where big profits are more likely. The price of ‘premium’ property is determined by location, taste and style as well as construction cost, and the renters targeted are often not paying out of their own pockets but spending international organizations’ generous allowances. In a new market such as Rwanda, this often means the higher that developers go in terms of producing properties perceived as high-end and ‘international standard’, the greater the potential profit: ‘the only way you can battle the huge cost of finance is to build more expensively’ (architect, interviewed 5 June 2014). Meanwhile, the fact that 81% of the population do not earn enough to be able to access any kind of mortgage means there is no effective demand from low- and middle-income groups to incentivize developers to build for them. The constraints upon property development and the incentives to construct high-end properties are thus two sides of the same coin.

In marked contrast to Rwanda, there are no formal incentives for real estate developers in Ethiopia. As noted above, the government initially provided cheap land, but following increased commitments to prioritize other sectors (alongside concerns about corruption and speculation in real estate) this is no longer the case (property developer, interviewed 30 September 2014). The government is suspicious of service sectors generally, viewing them as having rent-seeking rather than value-creating tendencies (international adviser, interviewed 2 October 2014). A series of events, including one particularly high-profile scandal involving the firm Access Real Estate, heightened distrust between developers and the government from the late 2000s. A regime that was already suspicious of the sector thus became actively hostile, especially towards those who, like Access, aim to sell properties ‘off plan’, taking a large percentage of payment prior to actual construction. This lack of trust and the policies it generated have had some unintended consequences. One is that people are so distrustful of developers that they are desperate for their own land, contributing to a scramble that has pushed up land prices and made the production of housing even more expensive (property developer, interviewed 30 September 2014). Partly due to the Access scandal, a series of complex policies implemented from around 2010 sought to prevent land speculation. These introduced a limited six-month window during which individuals who have acquired land may sell it on undeveloped. Thereafter, for residential properties owners need to have built 50% of the structure outlined in their approved plans, or 30% for commercial property, before they are allowed to pass on the title deed (property developer, interviewed 30 September 2014). Moreover, if developers want to subdivide land into individual parcels then they have to have developed 70% of the structure before they can acquire individual title deeds (civil engineer, interviewed 30 September 2014).

Despite the more hostile formal institutional environment, as in Kigali there are numerous powerful informal incentives to invest in property. Inflation plays a role; in the words of one source, ‘if you’re a wealthy Ethiopian, what do you do with your money? You get 5% at the bank but inflation is 10–15%’ (foreign adviser, interviewed 2 October 2014). There is also no stock market; Ethiopia is the world’s largest country in population terms not to have a stock exchange. Meanwhile, as in Kigali, property tax is virtually non-existent (Goodfellow, 2017). In addition (and contrasting somewhat with Kigali), despite the above measures to prevent land speculation, the planning and construction regulations themselves are relatively lax. Building permits are required, but the system of occupation permits, whereby a construction has to be approved for use after being constructed, is barely functional (tax adviser, interviewed 26 September 2014), removing a further barrier to the use of property as a relatively easy source of economic returns.

Alongside these unintended incentives to invest in buildings, there are enormous constraints. There is no mortgage system and, with the banks tightly controlled by the state, it is very difficult for developers to obtain loans. The state-owned banks prioritize loans for national development projects, such as the textile and leather industries. Moreover, private banks are constrained because, for every US $1 they lend, they are obliged to invest 27 cents into bonds issued by the national development bank to help fund Ethiopia’s flagship energy project, the National Renaissance Dam (World Bank official, interviewed 3 October 2014). This discourages banks from granting long-term commercial loans, as they need to retain sufficient capital to purchase these bonds. In addition, real estate developers are only able to obtain loans from commercial banks if they have title deeds for individual plots, which (as noted above) requires them to have already developed 70% of the property (civil engineer, interviewed 30 September 2014).

Further constraining development is the fact that land is extremely expensive, as the government only releases limited amounts of land for leasehold sale, and for commercial projects this is mostly allocated through competitive bidding. Construction costs are inflated by lack of infrastructure, transport charges and delays caused by anti-corruption measures (property developers, architects, civil engineers and construction firm representative, interviewed various dates, September–October 2014). In the absence of mortgages, and with new constraints upon developers’ capacity to sell properties ‘off plan’, most property development now comes either in the form of mega-investments by investors with sufficient upfront cash (including the state itself) or as incremental building by individuals and organizations as they acquire money gradually. Diaspora Ethiopians fall into both categories. For them, the incentive to invest in property is augmented by the fact that they are prohibited from investing in key sectors reserved exclusively for domestic Ethiopians: insurance, banking, telecommunications, media and air transport services.

The constraints upon the property market and the lure of investing in real estate are again two sides of the same coin. Indeed, the additional squeeze on the sector after the Access scandal was described by one firm as ‘both a problem and an opportunity’, because now if you finish the job (rather than selling off plan) people will pay much more (property developer, interviewed 29 September 2014). There is sufficient demand from Ethiopia’s wealthy elites for properties, to either live in or rent out, meaning that any developer who can deliver the finished product is well placed to make astonishing profits (local researcher, interviewed 23 September 2014). It is to the urban outcomes of the incentive structures discussed here, and the income generation linked to them, that we now turn.

Urban fortunes and skeleton cityscapes

Returning again to Kigali, the effect of the incentives discussed above on the high-end residential property market is palpable, particularly in respect of rental properties (see Figure 1). Rentals predominate in the business of the few formal estate agents operating in the city. Few if any houses are available to rent below US $500 per month (this would only obtain a cheap apartment far from the city centre). Top-end properties rent for US $8,000–10,000 per month, and most estate agents reportedly even have properties on their books asking as much as US $12,000 per month. More commonly, houses in ‘new’ areas of the city rent for around US $2,500 per month, though around US $4,000 per month would be ‘typical’ for a diplomat or someone working for the EU estate agent, interviewed 6 June 2014). Clearly, in a country with annual per capita GDP of US $697,9 ‘not many Rwandese are renting those houses at those prices!’ (architect, interviewed 3 June 2014).

Prime residential real estate in Kigali at various stages of construction (photo by author, 2011)

Given the limitations of property-related taxation, the rental income accruing to owners of these properties is very substantial, further incentivizing the construction of houses at the top end of the market by both individuals and property development companies. Thus ‘all firms are fighting for 20% of the market’ when 80% of the demand is for low-income housing (property developer, interviewed 21 January 2015). The RSSB projects are exacerbating the problem, developing the kinds of property that primarily serve to make the rich richer. Rwanda is the most unequal country in the East African region, and among the 10% most unequal countries in the world by Gini coefficient;10 these real estate dynamics can have done little to alleviate this. One source linked to RSSB itself claimed that most people who were able to buy houses in their projects already own other houses, so they just rent the new ones out (former RSSB official, interviewed 3 June 2014). Some observers lament that the actors in the real estate sector are ‘creating a bubble. We have too much office space, too many hotels, and too much “high end”’ (foreign adviser, interviewed 8 June 2014). In the commercial real estate sector some sources claim that ‘most office blocks are empty’ (infrastructure firm representative, interviewed 19 January 2015) and that, despite the apparent scramble to build more, commercial areas are ‘failing’ (architect, interviewed 3 June 2014).

By 2014, it was apparent that the bubble had already started to burst and rents were falling due to an oversupply of high-end housing. One source noted that ‘you can’t really charge US $4,000 anymore, though landlords still want to’ (architect, interviewed 3 June 2014). Meanwhile, although most high-end houses were built for rental, ‘many of them end up on sale as [the owners] can’t get the rent’ (estate agent, interviewed 6 June 2014). Others likewise noted that ‘many people have failed to get tenants. My rents have dropped from US $1,500 to US $1,000’ (former government official, interviewed 5 June 2014). While the rental market was considered stable in 2011, continued supply at the high end combined with international aid cuts and budget cuts meant that by 2014 there was ‘less appetite for humungous villas’ (estate agent, interviewed 9 June 2014). Meanwhile, prospective renters of luxury properties can negotiate prices down with relative ease, for example from US $8,000 to US $6,000 (real estate broker, interviewed 6 June 2014).

Despite this, actors in the sector noted that ‘These days … people also want a bit of finesse’ and ‘you have to up your game today’ in terms of quality fixtures and fittings. Consequently, while profit margins may be dropping, there is still an ongoing effort to create properties that appeal to the high-income group in terms of quality relative to existing properties. Because the effective demand from the low-income majority of the population is so constrained, developers continue to build at the high end in the hope of future rental and sale profits––and with little by way of tax disincentives to steer their investments elsewhere. Moreover, while rents may be dipping, anyone who has channelled great sums into property is unlikely to lower prices very substantially, so rents remain far beyond the reach of the majority. A bubble economy serving a tiny fraction of the country’s property needs, promoted by the circulation of a certain number of international personnel and the growing wealth of a small economic elite, thus prevails.

The one major institutional investor, RSSB, is meanwhile continuing to build houses that are clearly unaffordable even for relatively well-off civil servants, which is ironic as they are the ones paying into the pension scheme. The disconnect is striking: as one individual rhetorically asked, ‘am I buying a house in New York?’ (civil servant, interviewed 4 June 2014). RSSB representatives suggested the houses are aimed at diaspora Ethiopians. In opposition to this, a source with close knowledge of diaspora profiles claimed: ‘we don’t have a diaspora that rich! They don’t know who our diaspora are’ (former government official, interviewed 5 June 2014).

In stark contrast to Kigali, those involved in real estate in Addis Ababa believe that there is an enormous amount of unmet demand for residential housing even at the high end, due to the extreme constraints discussed above. One claimed that there is currently unmet demand for 50,000 housing units in the US $300,000–400,000 range. All of the real estate companies together only build an estimated 1,000–1,500 units per year, so this demand is nowhere near being met (construction firm representative, interviewed 29 September 2014). Unlike in Kigali, Ethiopia’s (much larger) diaspora undoubtedly play a critical role here. One source affirmed that ‘75% of our buyers are Ethiopians living in foreign countries, usually the US’ (property developer, interviewed 30 September 2014). The ongoing demand for high-end residential property is thus much higher in Addis Ababa, both because there is a much larger wealthy diaspora looking to own homes in the city and because the squeeze on developers means that supply is more constrained.

All of this means that, relative to other sectors, real estate is extremely profitable for those who have the finances to engage in development. It is perfectly feasible to make 100% profit on the construction and sale of a house (property developer, interviewed 30 September 2014). Finished houses can also command twice the price or more of similar houses sold ‘off plan’ (construction firm representative, interviewed 29 September 2014). Undersupply is such that the first houses built by Ayat around the turn of the millennium, which at the time sold for 187,000 Ethiopian Birr, were in 2014 (having undergone no improvements) being sold for 4 million Ethiopian Birr––in other words their value increased over twentyfold in under 15 years (property developer, interviewed 3 October 2014). Other developers said that houses sold for 3 million Ethiopian Birr in 2009 were worth 12 million Ethiopian Birr by 2014 (property developer, interviewed 3 October 2014). Those who purchase houses from developers and sell them on relatively soon are thus able to make enormous profits. Meanwhile, people lucky enough to be allocated (by lottery) an apartment through the government’s condominium programme11 can also either obtain a very good income through rent12 or sell the apartment on (after a five-year period during which selling is prohibited) at a profit. While the purpose of the condominium programme was to house low-income communities, the ability to sell the units has led to sale prices of around 1.5 million Ethiopian Birr (around US $72,000 at September 2015 exchange rates) and rising for an apartment. This is far beyond the capacity of low- and even most middle-income groups to purchase, especially given the absence of mortgages (construction engineers, housing policy advisers and property developers, interviewed September–October 2014). At US $619 per annum, Ethiopia’s annual per capita GDP is very similar to Rwanda’s.13

Another striking feature of Addis Ababa’s emerging landscape, which dwarfs that occurring in Kigali, is the extent of commercial real estate development and in particular the mushrooming of high-rise towers (see Figure 2). The fact that the construction sector is second only to the government as an employer in Addis Ababa (civil engineer, interviewed 29 September 2014) cannot be adequately grasped without considering the scale of commercial real estate development being undertaken in the city. Opinions differ regarding the actual extent of demand for commercial property. The presence of a large number of international organizations in a city that fulfils the role of diplomatic ‘capital of Africa’ is clearly significant in accounting for the explosion of commercial development. Yet there is also speculation on the perceived potential for future profits in commercial real estate and again the lack of any tax disincentive for unused or under-utilized commercial structures is important here. those who own land are often keen to ‘gamble on the strong demand for office and commercial space’ (Duroyaume, 2015: 405) by constructing, rather than sell it or use it in any other way.

Unfinished commercial structures in Addis Ababa (photo by author, 2014)

For diaspora Ethiopians, who are prohibited from investing in several key sectors open to other Ethiopians, building office or retail space for rent is especially appealing. This is one group that does have much easier access to finance. Since 2005, when Ethiopia adopted a more market-driven growth model, the opportunity to engage in major real estate ventures relatively unhindered by tax has helped to keep the potentially hostile diaspora Ethiopians (many of whom do not support the ruling regime) content. The deal between the government and the country’s ‘new entrepreneurs’ (many of whom were diaspora Ethiopians) was implicit: ‘stop politicking and we’ll help you get rich’ (Lefort, 2015: 365). It is clear which prospective route to ‘getting rich’ most wealthy diaspora Ethiopians have picked; despite numerous formal government incentives to invest in industry and none for real estate, the reality is that many diaspora Ethiopians have become de facto ‘place entrepreneurs’ in the city. Sixty percent of all investment by diaspora Ethiopians during the period 1994–2014 was in real estate and related services (four times more than the amount they invested in manufacturing), and 91% of all diaspora Ethiopians’ investment was located in Addis Ababa.14 It therefore comes as little surprise that even in Ethiopia, where the government has concertedly pursued a strategy of industrialization, manufacturing still constitutes barely 5% of GDP (ibid.: 369).

A significant degree of informality characterizes such property investments and the channels through which these resources travel to fund the escalating landscape are difficult to identify. As one source noted, ‘the question we are all asking is: where is all the money for construction coming from?’ (local researcher, interviewed 24 September 2014). One study by a local research organization reportedly showed that just 10–15% of the money ploughed into construction in a five-year period had passed through banks (local researcher, interviewed 24 September 2014). What is clear is that the current incentive structure, and consequent behaviour of those who invest in and develop property, leads to a very large number of incomplete structures. This is partly because of the perceived appeal of real estate as a source of profit for anyone able to get involved, even those without the expertise or capacity to finish the job. These are often not property developers in the sense assumed in the conventional real estate literature. One industry professional noted that many people in other businesses, with no knowledge whatsoever of the sector, effectively become entrepreneurs dealing in unfinished places: they ‘see that it’s so lucrative so try to invest their profits in real estate. When they run out of funds they re-invest what they have in their [other] business, hoping to make more profit to finish the job, but this rarely happens’ (construction engineer, interviewed 29 September 2014).

Further feeding this are the aforementioned rules regarding how much must be built on a plot of land before title deeds can be transferred, which creates incentives to part-construct buildings so the land is saleable. In reality, of course, many plots do exchange hands without the requisite percentage of construction having been reached, through systems of informal brokerage and exchange. However, the new ‘owner’ remains in a position of insecurity until they have undertaken sufficient construction to be able to officially acquire the deed (civil engineer, interviewed 30 September 2014). This regulatory framework combines with the lack of savings and investment opportunities to encourage people to channel resources into buildings, even if they cannot be finished when easy finance runs out. The combination of incentives and constraints in place is thus producing ‘skeleton cityscapes’ of speculative and unfinished construction.

Concluding reflections: the safest bet?

Putting up a building is the best bet now … Manufacturing is too risky and complicated. We lack the risk-taking entrepreneurs and it is very hard to compete in the global market … Putting up a building is risky, but it’s less risky than anything else. It’s very much seen as a ‘safe bet’ (local researcher, interviewed 24 September 2014).

As the above quotation from a source in Addis Ababa illustrates, the situation in which late-developing countries find themselves today is one where competing in manufacturing export markets is extremely difficult, and highly skilled service sectors even more so. Appealing investment options are few and far between. In Rwanda, with even fewer raw materials and a much smaller market, things are even more acute. What profits can be made in other businesses are perceived as best reinvested not in that business or in a bank, but in property. As a Kigali source put it, ‘If I’m running a business whose absorption is only US $5,000, the rest I invest in a risk-free asset: housing. It will appreciate. It’s a safe bet … for those with cash to spare, they invest in housing. It’s all about managing risk’ (property owner, interviewed 4 June 2014). While the buildings that appear through such strategies are not necessarily profitable or even usable in the short term, in the medium to long term they are seen as more likely to yield profits than anything else. In the absence of disincentives, such investment carries relatively little risk.

This lack of measures to effectively deter property speculation is cause for serious concern. The development strategies of most African states seek to attract large-scale investment into productive sectors and encourage savings, but a situation where major de facto tax breaks accrue to real estate may be impeding this by increasing the magnetism of construction of (potentially) high-return property. The property tax literature suggests that higher taxes on land (as opposed to structures) reduce land speculation, because they incentivize people to build and make the land more profitable, while higher taxes on structures discourage development of the land (Bahl and Linn, 1992: 173–4). However, built into this analysis is an assumption that the ‘intensive’ and ‘efficient’ use of land through high-rise development of any kind is necessarily a positive outcome in any given setting. What this ignores is the possibility not of speculation on land but speculation on buildings themselves. If, as in Addis Ababa, land is very expensive but buildings effectively untaxed, then the incentive is to build in the hope that the building will generate profits––and the bigger the building the better. Under such conditions and in the context of rapid economic growth, people with resources may be more inclined to speculate on buildings than on land itself. This is potentially even more damaging if it results in an oversupply of buildings that does not match demand, a reduction in the availability of land for different uses and the sucking away of resources from productive investment.

Moreover, once the buildings are there, the city has been fundamentally shaped by them in ways that are difficult to reverse. As Lefebvre noted, urbanism becomes a force in its own right. The strategies and tactics of people involved in property––which normalize incremental development, cross-financing from other businesses and the leaving of buildings unfinished for long periods of time––are not irrational given particular land regulations, high inflation, constrained finance, absence of property taxation and anticipation of future economic growth. These conditions and behaviours are not adequately accounted for in conventional accounts of actors and institutions in the real estate sector, but they are the conditions and behaviours shaping many cities today. Situations of late development are thus often also characterized by a condition of ‘late urbanization’, which is undertheorized and often overlooked. These countries are shaped by urbanization occurring at a time when it is very hard to industrialize due to the strength of international competition for manufactures, yet economic growth is taking place internally, diaspora finance is flowing in quest of high returns and international personnel on big salaries are helping to buoy the urban economy. All of this combines powerfully with ideas of ‘modern’ urban landscapes befitting future ‘global cities’ to draw capital into the secondary circuit. Such forces can even thwart concerted efforts to channel resources into manufacturing industry, as in Ethiopia, especially when bolstered by the informal incentives created by state failure to tax property and related incomes.

It is clear that there are important differences between the countries considered here in terms of the kinds of real estate that capital is flowing into most vigorously: much more commercial property is being developed in Addis Ababa and oversupply of high-end residential property is more evident in Kigali. Nevertheless, just as there are generic features of property development in advanced capitalist societies that real estate researchers have sought to specify, so too are there general trends in situations of ‘late urbanization’ that these cases highlight. Speculative urbanism, rather than the force of past or present industrial development, is a major (arguably the major) driving force in the urban landscapes discussed here. This is clearly not unique to the two countries discussed above. In Tanzania, for example, an oversupply of both high-end residential and commercial property is now widely acknowledged (government officials and property developers in Tanzania, interviewed various dates, May–June 2016), yet construction for these chimerical markets continues. One Tanzanian source spoke of being approached in all seriousness by a major investor regarding a real estate project worth US $1 billion in Dar es Salaam; upon asking incredulously ‘where is the demand?’, he was met with the response that it might appear, ‘maybe in 20 years’ (property developer, interviewed 1 June 2016).

The problems with this mode of urban development are potentially severe for several reasons. A first relates to the glaring disjuncture between what is being built and the needs of the majority of the population, particularly with regard to affordable housing. While the Ethiopian government has attempted to innovate in this respect with its condominium programme, the shortcomings of this scheme in practice are significant. While some low-income groups can access these units through the lottery system, the monthly payments required are still too much, so those who acquire them usually continue living in a slum while renting them out to middle-income groups in order to cover the cost. This is creating new forms of spatial segregation, albeit ones that diverge markedly from what would be expected based on the programme’s stated intentions. In Kigali, meanwhile, virtually nothing has been done to produce affordable housing despite the government’s awareness of the urgency of the issue (Republic of Rwanda, 2015b), while the production of Vision City and pursuit of a ‘modern’ skyline continues apace, squeezing the poor into an ever smaller proportion of urban space. Interestingly, in both Addis Aababa and Kigali this segregation is not of the highly fortified variety associated with some other African cities, but there are grounds to believe that such fortification will follow not long behind when space is partitioned in this way (Murray, 2011).

Secondly, as scholars exploring the potential and limits of the ‘secondary circuit’ have argued, a fixation on real estate may present problems for long-term capital accumulation itself. As Gotham (2009: 359) notes, ‘real estate is by definition illiquid, spatially fixed and immobile, relatively durable and costly’, while capital is ‘abstract, nomadic and placeless’; this duality between immobile properties and mobile capital represents an ‘inherent contradiction’. In this view, real estate can itself ‘be a barrier to capital accumulation, when its enduring qualities render it outdated and anachronistic, or when the financing needed to construct, sell, and rehabilitate it is unavailable’ (ibid.). Although addressing the subprime mortgage crisis in the global North, this is very pertinent to the problems discussed here. Few in either Kigali or Addis Ababa doubt that many of the residential or commercial properties being erected at huge cost will lie partially or wholly unused, with their relevance to future development an open question. Taking up an increasing proportion of the urban surface in situations where land is under huge pressure, the channelling of resources into these structures results in decreasing space (as well as a relatively stagnant industrial base) into which capital can more productively flow. As such, it is not only social equity but structural economic transformation itself under threat from these cityscapes. The inability of existing theories of urban real estate investment to adequately predict and explain these developments provides fertile ground for further theoretical reflection and innovation.

  1. Due to the paucity of data and the simultaneous use of conflicting definitions of urban (even within countries), these figures are inevitably contested, even without entering into current debates on what constitutes the urban (Brenner and Schmid, 2014). This is particularly notable in the case of Rwanda, where the UN population data (which places all the other countries mentioned below 20%) places the urbanization level at 28% (UNDESA, 2014), but the government’s own census data records a figure of 16.5% (Republic of Rwanda, 2015b). See Potts ( ) for an extensive discussion.
  2. See also Obeng-Odoom (2014) on the effects of oil-driven urban development on inequality in the context of Ghana.
  3. There are important debates over the validity of this data (see Potts, 2017).
  4. Again, these figures are contested due to the paucity of data.
  5. Various government datasets acquired in Rwanda and Ethiopia, 2014–15.
  6. International Business Times 26 August 2013; see (accessed 27 July 2016).
  7. For a detailed discussion of these issues, see Goodfellow and Smith (2013).
  8. In reality, however, the process can take years.
  9. World Bank 2015 estimate; see (accessed 27 July 2016).
  10. 2013 data from the United Nations Development Programme; see (accessed 1 June 2016).
  11. See Duroyaume (2015), Keller and Mukudi-Omwami (2017) and Planel and Bridonneau (2017) for discussions of this scheme.
  12. The very desirable condominium apartments were commonly renting for 3,000–5,000 Ethiopian Birr (US$ 144––240 at September 2015 exchange rates) per month, largely pricing out the poor.
  13. World Bank 2015 estimate; S See (accessed 27 July 2016).
  14. Figures acquired from the Ethiopian Investment Authority, October 2014