1 Introduction

There has been a dramatic increase in attention in recent years to the echoes of colonialism, imperialism, and slavery in the contemporary global financial system. There is growing attention in recent research across geography, sociology, political economy, and development studies to how present-day global finance reproduces colonial hierarchies and dynamics. A number of researchers have examined the subordinated position of formerly colonized countries in the global financial system, highlighting the constraints this poses on development and policy space (see Koddenbrock, Reference Koddenbrock2020; Tilley, Reference Tilley2021; Alami et al., Reference Alami, Alves, Bonizzi, Kaltenbrunner, Koddenbrock, Kvangraven and Powell2023). For post-colonial governments this has often meant limited control over macroeconomic and monetary policy, and persistent challenges mobilizing resources for development. Formerly colonized countries also tend to have financial systems that are dominated by foreign banks predominantly based in former colonial powers, which often mobilize very little domestic investment to a relatively narrow range of firms and relatively affluent urban residents (see Kvangraven et al., Reference Kvangraven, Koddenbrock and Sylla2021; Akolgo, Reference Akolgo2022; Bernards Reference Bernards2022a, Reference Bernards2022b; Koddenbrock, Kvangraven, and Sylla, Reference Koddenbrock, Kvangraven and Sylla2022). As Nance and Tsingou (this volume) note in their discussion of correspondent banking relationships and remittances, financial institutions in colonized countries also tend to remain subject to infrastructures largely operated from metropolitan powers and vulnerable to disruption largely beyond their control. The limited availability of credit for agricultural and industrial development in former colonies has been a persistent policy concern throughout much of the twentieth and twenty-first centuries (see Bernards, Reference Bernards2022b). Analysis of durable colonial legacies in financial systems speaks to a wider trend towards examining the ways that contemporary development interventions replicate the spatialities, power relations, and practices of colonial rule (e.g., Engel and Susilo, Reference Engel and Susilo2014; Cavanagh and Himmelfarb, Reference Cavanagh and Himmelfarb2015; Wainwright and Zempel, Reference Wainwright and Zempel2018; Enns and Bersaglio, Reference Enns and Bersaglio2020).

While previous research has shown clear echoes and parallels between the era of formal colonial rule and the present, less attention has been devoted to explaining how past and present patterns of activity are related. As Cooper (Reference Cooper2005, pp. 17–18) rightly notes, we should be wary of ‘leapfrogging legacies’ in discussing the enduring impacts of colonialism – of claiming that ‘something at time A caused something in time C without considering time B, which lies in between’. It is not enough to show a correspondence between past and present patterns without either tracing patterns of continuity and change over time, or showing how continuities are reproduced. So what is it, then, that makes colonial histories durable?

The argument in this chapter is that a lens focused on financial infrastructures – adopting what the editors of this volume call an ‘infrastructural gaze’ (see Westermeier, Campbell-Verduyn, and Brandl, this volume) is highly useful for understanding the durability of colonial patterns of uneven financial development. The argument is focused on African colonial history, with detailed illustrations primarily based on the Kenyan case, but much of what is argued is also likely of wider applicability. I develop the argument in four steps in what follows. Section 2 argues for an infrastructural lens as a means of explaining and exploring colonial continuities. Section 3 briefly describes the landscape of colonial financial systems. Sections 4 and 5 consider some of the key infrastructures of colonial financial systems – bank branches, and collateral in land – and the role that they’ve played in replicating colonial financial geographies over time.

2 How Do Colonial Histories Persist?

A number of recent analyses draw on the concept of ‘infrastructures’ to make sense of long-run colonial legacies (see Mitchell, Reference Mitchell2014; Desai, McFarlane, and Graham, Reference Desai, McFarlane and Graham2015; Davies, Reference Davies2021). Notably, many of these have focused on East Africa, predominantly on transport mega-projects (e.g., Enns and Bersaglio, Reference Enns and Bersaglio2020; Kimari and Ernstson, Reference Kimari and Ernstson2020; Aalders, Reference Aalders2021; Lesutis, Reference Lesutis2022). ‘Infrastructures’ here are understood, following Bowker and Star (Reference Bowker and Star1996), in the broader sense of backgrounded technical systems allowing basic functions and circulations to be carried out.

This kind of ‘infrastructural gaze’ (Westermeier, Campbell-Verduyn, and Brandl, this volume) offers promising means of making sense of colonial continuities in two ways. First, infrastructural systems, as a number of authors have argued, were integral to the development of colonial economies, and in the process solidified particular patterns of social relations. Notably, a number of authors have shown how the production of particular kinds of infrastructures was critical to colonial governance across multiple scales (see Kooy and Bakker, Reference Kooy and Bakker2008; Pritchard, Reference Pritchard2012; Mullenite, Reference Mullenite2019; Cowen, Reference Cowen2020; Davies, Reference Davies2021). Infrastructures embody ideas about how societies should be organized, produce, and reproduce racial hierarchies, and entrench inequalities in access. Colonial infrastructures were ‘themselves productive of race’, in Sherman’s (Reference Sherman2021) words, insofar as they codified formal racial hierarchies in a variety of different ways (I discuss one example of this, racial restrictions on land titling, later in this chapter). Davies notes – referring directly to Latin America, but making a point that could apply more broadly – that ‘It was through infrastructural networks that colonialism and imperialism operated, which made relations of economic dependency and neo-colonialism possible’ (2021, p. 741). This is no less true of financial infrastructures, although the latter haven’t been studied in historical perspective in the same depth as things like ports, rail, or water and sanitation infrastructures have.

Secondly, thinking in terms of infrastructures offers a useful perspective on the durability of colonial-era systems. Infrastructure studies in general include a perspective on patterns of continuity and change which is highly useful here. Infrastructures are rarely overhauled wholesale. Emphasizing infrastructures helps to consider how new technical systems intermingle with existing ones. As Susan Leigh Star (Reference Star1999, p. 382) puts it, ‘[i]nfrastructure does not grow de novo; it wrestles with the inertia of the installed base and inherits strengths and limitations from that base’. As a result, infrastructures pattern the spatial and functional arrangement of social practice, sometimes long after their original uses have ceased. In Marieke de Goede’s words: ‘Infrastructural technologies … inscribe specific ways of doing things – of routing flows, enabling functionalities, structuring interactions. Often, infrastructural grids – of roads, electricity networks, payment routes – sediment historical power relations and core-periphery relations’ (De Goede, Reference de Goede2021, p. 354). These durable systems form the ‘installed base’ (per Star, Reference Star1999) into which new technologies must normally be fitted, through complex processes of adjustment and retrofitting.

One reason why colonial histories cast such a long shadow on the present, in short, is that they are embedded in material networks without which (post)colonial political economies quite simply could not function. Colonial infrastructures thus represent ‘imperial remains’ in the sense highlighted by Kimari and Ernstson – they are entangled material and social networks that retain important traces of colonial social relations long after the formal end of colonial rule (2020, p. 827). Financial infrastructures are a key mechanism through which spatial patterns rooted in colonial practices of racialization and exploitation have been produced and have persisted long after the formal end of colonialism. Some existing analyses have tentatively developed arguments in this direction. De Goede (Reference de Goede2021) in particular has drawn on infrastructure studies to trace out the durable legacies of colonial financial systems – although thus far in very general terms, marking out a research agenda rather than exploring in detail how financial infrastructures enable patterns of continuity and change in post-colonial settings (see also Bernards, Reference Bernards2022a, Reference Bernards2022b). However, while the materialities of colonial infrastructures clearly matter, they are not mechanistically replicated over time. As Aalders puts it, ‘the ruins of empire – both material and metaphorical – are durable but do not determine the present’ (2021, p. 997). What’s critical here is that an attention to colonial financial infrastructures helps to reveal durable challenges with which subsequent development efforts have grappled.

3 The Colonial Financial System and Its Infrastructures

Colonial financial systems prompted uneven development on two levels. First, they were extractive in character. It’s unquestionably true that colonialism generally enriched colonizers – or, primarily, some fractions of capital in metropolitan centres – at the expense of the people and territories that were colonized. Vast fortunes were extracted from colonized territories. Utsa Patnaik (Reference Patnaik, Chakrabarti and Patnaik2017, p. 311) has recently estimated, for instance, that net transfers from India to Britain between 1765 and 1938 amounted to £9.18 trillion. Finance capital was a key beneficiary of all this – the London money market was a major source of finance for governments, mining, and infrastructure projects globally throughout the latter half of the nineteenth century. Financial institutions in colonial territories also served in important senses as ‘channels’ for the extraction of surplus (see Rodney, Reference Rodney2018). Equally, though, colonial capitalisms also worked through the production of differentiated spaces within and between colonized territories, often closely linked to the geography of state power. If the generally extraverted character of colonial financial systems is a key common trait, the differentiation of colonial territory is nonetheless also an important dynamic shaping colonial financial systems.

Colonial capitalisms generated polarization between cities and countryside, built infrastructures – transport and financial systems in particular – designed to smooth the flow of raw materials for export, and actively underdeveloped some territories to create reserves of cheap labour for large-scale mining and plantations. The financial sectors that emerged in these contexts were generally clustered around urban centres, closely linked to metropolitan capitals, and often made their profits primarily by providing remittance services or by lending for large-scale public works rather than productive credit (see Bernards, Reference Bernards2022b, Reference Bernards2023; Koddenbrock, Kvangraven, and Sylla, Reference Koddenbrock, Kvangraven and Sylla2022, for more detailed discussions). Where they did directly invest in or lend to directly productive activities, this was disproportionately steered towards large-scale plantations or mining operations. In fact, colonial banks in many territories often held the majority of their income-earning assets in metropolitan centres rather than in colonies. Colonial banks very rarely developed the infrastructures – routines, social relations, or physical structures – necessary to lend to the majority populations in colonized territories.

In Sections 4 and 5, I’ll describe some of the problems thrown up by the financial infrastructures that were developed in colonial contexts. I put particular emphasis on networks of physical branches linked by transport and telecommunications systems, and on the use of land as collateral.

4 Bank Branches

Physical bank branches, linked together by road, rail, sea transport, and telecommunications systems, were critical infrastructures in colonial financial systems. They played a key role in administering the vital functions of the financial system – notably, assessing credit risks and settling payments. Moreover, if banking operations in colonial Africa depended in large part on social relations and embedded patterns of social practice, branches were important sites where those interactions took place. Previous work has highlighted how branch networks primarily clustered around major cities and other sites of expatriate economic activity (e.g., South Africa’s mines or Kenya’s settler farms). Some previous research has traced a major expansion in branch banking across sub-Saharan Africa in the 1950s (see Engberg and Hance, Reference Engberg and Hance1969), including in some of the case study countries (Morris, Reference Morris2016; Velasco, Reference Velasco2022). This expansion was highly uneven, with new branches disproportionately built in territories with higher concentrations of older branches. There is clear evidence that certain territories – notably those with concentrations of settlers (including Kenya and South Africa) or major export industries (notably cocoa in Ghana) – were disproportionately targeted for new branches (see Table 9.1).

There were, in a number of instances, dramatic expansions of commercial banking systems in the decade before decolonization. There was, for instance, a significant expansion of branch networks across sub-Saharan Africa, particularly in British territories, in the 1950s, for instance. What’s notable, though, is how much the density of bank branches in 1950 seems to have shaped the density of branches at the end of the decade. Kenya was a major focus of this expansion, along with Ghana, Nigeria, and then-Rhodesia (see Engberg and Hance, Reference Engberg and Hance1969, p. 196). Historians have often attributed this expansion in part to efforts by banks in Kenya and elsewhere in sub-Saharan Africa to navigate the political and economic pressures created by decolonization, as well as to capitalize on business opportunities seemingly opened up by the ‘developmental’ colonialism (see Cooper, Reference Cooper1996) of the post-war period (see Engberg, Reference Engberg1965; Engberg and Hance, Reference Engberg and Hance1969; Bostock, Reference Bostock1991; Morris, Reference Morris2016; Velasco, Reference Velasco2022). While this is broadly true, we can usefully situate this development, and understand its limits, with reference to the patterns of uneven development and contestation described earlier in this chapter. If the expansion of branch banking in the 1950s was in some senses a spatial fix for British capital tentatively seeking out new spaces for accumulation in colonized territories, it was also one that was strongly shaped by the configuration of existing financial infrastructures.

This uneven development of branch banking between territories was mirrored by the uneven development within territories. To take one of the key examples with denser branch banking networks, until 1950, bank branches in Kenya were predominantly located in Mombasa and Nairobi (Engberg, Reference Engberg1965, p. 190; Bostock, Reference Bostock1991; Morris, Reference Morris2016, p. 652), and virtually all in ‘White Highland’ areas (see Section 5). As Morris notes, ‘of the 20 areas of Kenya where the three major banks … were represented in 1950, only two (Kisii and Bungoma) were not dominated by European enterprise’ (2016, p. 652). In large part because of the presence of settlers holding mortgageable property titles, Kenya had a comparatively deep financial sector in contrast to most other territories in sub-Saharan Africa (see Newlyn and Rowan, Reference Newlyn and Rowan1954, pp. 76–77). This was reinforced by the fact that Mombasa and Nairobi were sub-regional commercial centres linking export agriculture across East Africa to world markets. Advances to merchants trading elsewhere, especially in Uganda, were typically contracted in one of these two cities (Newlyn and Rowan, Reference Newlyn and Rowan1954, p. 87).

Banks sought to build physical infrastructures that would enable them to profit from the potential rise of a ‘middle class’ of Kikuyu farmers in proximity to the White Highlands (see Morris, Reference Morris2016). In practice, though, the lending operations of commercial banks were refocused on short-term commercial loans in Nairobi and Mombasa (Hyde, Reference Hyde2009, p. 86). While banks built new branches in African-dominated rural areas, these new branches engaged minimally in credit provision. For one major commercial bank, Jørgenson (1975, p. 160) reported in the mid-1970s that the median ratio of advances to deposits in rural branches was 28%, against 58% in urban branches. The result was that expanded branch networks largely channelled rural savings to European- and Indian-owned commercial firms. African borrowers accounted for less than 3% of credit from commercial banks as late as 1967 (Jørgensen, Reference Jørgensen and Widsrand1975, p. 158). The figure remained only 14% in 1973 (Jørgensen, Reference Jørgensen and Widsrand1975, p. 161).

These patterns of uneven infrastructural development appear to have had significant impacts on more recent development. The rapid development of mobile money and digital credit in Kenya is widely celebrated, but uneven. Mobile money and digital credit have had to ‘wrestle with the installed base’ (per Star, Reference Star1999, p. 381) of existing financial infrastructures, and in important respects they mirror the geography of colonial financial systems outlined earlier in this chapter. These differences are especially pronounced when looking at credit. Digital and mobile credit have proven to be more controversial than payments, with concerns about growing over-indebtedness facilitated by digital credit apps (see Donovan and Park, Reference Donovan and Park2019), even from erstwhile fintech promoters (e.g., Izaguirre, Kaffenberger, and Mazer, Reference Izaguirre, Kaffenberger and Mazer2018). The potential role of mobile money in laying the groundwork for expanded credit nonetheless remains a key claim about the long-run benefits of mobile money (see Kaffenberger, Totolo, and Soursourian, Reference Kaffenberger, Totolo and Soursourian2018). So patterns of credit access are particularly telling. And, critically, the rollout of mobile and digital credit closely mirrors the colonial financial geographies highlighted earlier in this chapter.

As shown in Table 9.2, the proportion of residents in Nairobi Metropolitan Area and Mombasa reporting past or present borrowing using both mobile money services (25%) and digital lending apps (18.2%) is more than double the respective use rates of mobile (12.3%) and digital borrowing (7.1%) elsewhere. Kenya’s fintech boom, in short, is predominantly an urban phenomenon, and especially concentrated in Mombasa and in and around Nairobi. While consumer lending, as opposed to commercial lending, has grown dramatically with the advent of the new apps, particularly in the context of widespread precarity in cities, the extension of credit has taken place much more rapidly in proximity to existing financial infrastructures.

The overarching point here is that the fintech boom reflects patterns visible across successive episodes of financial restructuring in Kenya and elsewhere. We can understand this as being the result, at least in part, of a tendency for restructured financial operations to rely on and work through existing infrastructures. The expansion of new financial services, both in the last decade of formal colonial rule and with ‘innovative’ fintech applications since around 2010, has tended to follow the established geography of branch networks. If we want to understand the ways that colonial patterns of uneven development continue to weight on changes to postcolonial financial systems, an infrastructural gaze is important.

5 Land and Collateral

Another key infrastructure of colonial financial systems was the use of land as collateral for loans. There is an extensive literature on colonial and post-colonial land issues across Africa (among others, see Manji, Reference Manji2020; Ouma, Reference Ouma2020; Dieng, Reference Dieng2022) and more widely. Several recent contributions have shown how systems of property title were central to the historical formation of racial hierarchies and colonial economies, and to the persistence of colonial legacies (Keenan, Reference Keenan2017; Bhandar, Reference Bhandar2018). References to a lack of suitable collateral as a primary reason for limited lending to Africans in colonial financial systems are commonplace from authors working in a wide range of different theoretical and methodological perspectives (e.g., Newlyn and Rowan, Reference Newlyn and Rowan1954; Cowen and Shenton, Reference Cowen and Shenton1991; Uche, Reference Uche1999). Economics and economic history perspectives on colonial legacies, as noted, often centre on the form and ‘quality’ of property rights under colonial systems.

Property titles to agricultural land were a critical financial infrastructure; indeed, transferrable property titles which could be repossessed in the event of default were arguably the main mechanism by which banks assessed and managed credit risks. In performing this function, property titles also constituted and entrenched processes of racial and spatial differentiation. In Manji’s phrase, ‘Kenyan land policy was … racialised at its inception’ (2020, p. 32). Titles to land in the White Highlands were reserved for ‘European’ settlers. Nominally ‘uninhabited’ or ‘unused’ land was claimed by the Crown and subsequently made available for purchase by settlers. Importantly, the question of what constituted ‘vacant’ land was contested – especially because the state treated fallow land and rotating pastures as ‘vacant’. The allocation of land for white settlement disrupted existing forms of agriculture and pastoral livelihoods, and was contested throughout the colonial period, particularly by Kikuyu agriculturalists living near the Highlands (Coray, Reference Coray1978). The associated policy of segregating African populations into restrictive ‘reserve’ areas based on ‘tribal’ groupings also helped to reshape and reproduce ethnic differentiation (Kanyinga, Reference Kanyinga2009, p. 328).

Alongside their role in processes of racialized dispossession, the role of land titles as specifically financial infrastructures also strongly shaped their development. Notably, land rights were initially conditional on ‘productive’ use, as the state sought to minimize land speculation. This was quickly overturned as settlers complained that such restrictions inhibited their ability to use purchased land as collateral: ‘the settlers are naturally anxious that the land on which they spend their labour should be a marketable and mortgageable security’ (East Africa Protectorate, 1908, p. 30). Officials also began to view speculation on land as a means of raising the value of farmers’ collateral, hence enabling wider access to credit (Lonsdale and Berman, Reference Lonsdale and Berman1979, p. 499). In short, while racial restrictions remained in place, other restrictions on property titles were quickly removed, specifically in order to facilitate land use as means of assessing credit risk.

Settler agriculture was highly stratified. Some large individual and corporate landholders could access relatively cheap credit in London; the bulk of settlers on smaller plots were reliant on the colonial financial system (Van Zwanenberg, Reference Van Zwanenberg1975, pp. 278–279). Many of the latter were heavily indebted, particularly because banks loaned against the market value of land rather than farm income (Van Zwanenberg, Reference Van Zwanenberg1975, p. 280). Small settlers in particular increasingly depended on access to cheap labour, secured mainly through the forcible underdevelopment of reserves and restrictive laws governing the movement of African populations (Berman and Lonsdale, Reference Berman and Lonsdale1981, p. 62). While short-term migrant labour remained important, longer-term tenant farmers (‘squatters’), governed by increasingly restrictive Resident Native Labour Ordinances gradually increased as a proportion of the labour force. The intersection of credit infrastructures with racialized structures of property relations in this sense was also crucial. It was not simply ownership over land, but also the control this granted over access to credit – and hence over inputs and machinery – that enabled settler control over migrant and tenant labour. Indeed, the importance of uneven access to formal credit is underlined by the adoption, at the behest of settlers, of increasingly severe restrictions on credit to Africans (see Jørgensen, Reference Jørgensen and Widsrand1975, p. 150).

In short, the use of mortgageable property titles to exclude some borrowers from formal financial markets was in fact a feature of colonial financial infrastructures rather than a bug. The contrast with trading economies in West Africa is instructive here. Uche (Reference Uche1999) shows particularly clearly that merchants increasingly saw the restriction of bank credit to African farmers as a crucial element of maintaining their control over cheap crops. Indeed, merchant firms often actively resisted state-backed efforts to reform agricultural finance, explicitly recognizing the vital role of credit in ensuring control over cheap crops. If the needs of colonial capital were different between colonies organized around settler-run mines or plantations and territories organized around the export of commodities produced by small farmers (needing cheap labour in the former and cheap crops in the latter), credit infrastructures designed around the collateralization of land allowed the mobilization of both.

There were tentative efforts to support African agriculture in Kenya from the 1930s, particularly in Kikuyu regions adjacent to the White Highlands, with the colonial state seeking both to expand its fiscal base and contain growing political threats. These efforts were given considerably greater emphasis by the intensification of the Mau Mau rebellion – an armed revolt led by the Kenya Land and Freedom Army, concentrated in Kikuyu-dominated regions, which was brutally repressed by the colonial state (see Anderson, Reference Anderson2005). Alongside the horrific counterinsurgency, agricultural reforms – implicitly or explicitly targeted to Kikuyu areas – formed a key part of the colonial response to the rebellion. Roger Swynnerton, then-Assistant Director of Agriculture in Kenya, was appointed to propose a strategy for agricultural development in 1953. Swynnerton’s report marked a shift towards an explicit policy encouraging the development of African agriculture (Shipton, Reference Shipton1992), intending to create a politically ‘stabilized’ middle class of property-owning Kikuyu farmers. Land titling, again, was central to these reforms. Swynnerton proposed expanding formal land titling to ‘African’ areas, facilitating the use of such titles as collateral, and as a result expanding the scope of capitalist agriculture while using minimal state resources: ‘If Africans … achieve titles to their land in economic units, much greater facilities should be made available to them for borrowing against the security of their land’ (CPK, 1954, pp. 54–55). This was, notably, explicitly presented as an alternative to providing direct state support (CPK, 1954, p. 54).

What’s important about the Swynnerton proposals is that, rather than amending financial infrastructures to address the restrictions they posed for African agricultural development, the reforms sought instead to alter the way that African communities held land. This approach ultimately threw up important political contradictions which eroded the effectiveness of credit infrastructures oriented around agricultural land.

The point here is that racialized restrictions on property ownership operated as a critical mechanism through which the colonial financial system and colonial economy more widely were produced. Land titles in this sense operated as financial infrastructures and generated important problems for subsequent development, which ironically gradually undermined the utility of land titles as credit infrastructures. The anti-colonial movement was split over whether settler land should be restored to pre-colonial inhabitants or redistributed among existing residents, as well as the nature of land rights that should be implemented (see Kanyinga, Reference Kanyinga2009; Manji, Reference Manji2020). The Kenya African National Union (KANU) – dominated by Kikuyu elites – backed the maintenance of property rights where they existed and the redistribution of settler land through purchase, while the Kenya African Democratic Union (KADU) advocated for the restoration of control over land to pre-colonial inhabitants. KADU proposals would have meant the restoration of the White Highlands primarily to collective control by predominantly Kalenjin pastoralist groups who had been displaced by the definition of ‘uninhabited’ land deployed by the colonial state. KANU ultimately won the pre-independence elections in 1963, with Jomo Kenyatta as prime minister (and subsequently president).

The KANU government under Kenyatta pursued a programme of resettlement of settler land. From 1962 to 1966, approximately 20% of settler land was purchased by the state and sold (on credit) to smallholders; by the 1970s half of former settler land had been redistributed (see Boone, Reference Boone2011, pp. 79–80). This redistribution preserved a stratified system based primarily on private ownership, albeit one in which some African (and particularly Kikuyu) elites were able to accumulate large holdings (see Manji, Reference Manji2020). It also established strong state control over the allocation of land for smallholders, many of whom were subject to de facto tenancy arrangements in which they were unable to hold a formal land title until repaying (often unpayable) debts to the government for the purchase of settler land (see Boone, Reference Boone2011). All of this reinforced the highly ethnicized nature of land conflicts, which was arguably intensified after Kenyatta’s death and succession by Vice-President Moi in 1978. Under Moi the accelerated redistribution of land in former reserves, and frequent scapegoating of Kikuyu smallholders as sources of land scarcity, was a means of developing and mobilizing a cohesive Kalenjin political identity. Large tracts of land were turned over to politically aligned elites under Moi (see Boone, Reference Boone2011, pp. 85–86).

These changes did have a dramatic effect on the composition of agricultural production. At the end of the colonial period, roughly 80% of agricultural exports came from ‘large’ farms in the White Highlands and 20% from ‘small’ farms in reserves, by the end of the 1960s the ‘large’ and ‘small’ farm sectors each produced about half (see Njonjo, Reference Njonjo1981, p. 31). These transformations further cemented the orientation of the major banks towards commercial activity in Nairobi and Mombasa, with agricultural lending restricted to large landholders. The fragmentation of land into smaller plots and the uncertain control of smallholders over land titles meant that the viability of agricultural land as security was greatly diminished (see Shipton, Reference Shipton1992). In practice, increasingly, ‘the ability to attract credit from commercial sources has been established to depend not only on the title deed per se, but rather on additional assurance of wage labour, where it is easier to attach salaries for repayment requirements’ (Gutto, Reference Gutto1981, p. 54). Agricultural borrowing was primarily restricted to the largest farmers (many of whom, as noted, had close ties to KANU), particularly those could draw on incomes from formal salaried jobs or commercial property. Survey research in the 1970s found that, while roughly 10 per cent of smallholders overall had significant sources of off-farm income, 70 per cent of farmers contracting commercial loans did (Collier and Lal, Reference Collier and Lal1980).

Observers in recent years point to parallel processes of land consolidation and fragmentation. Hakizimana et al. (Reference Hakizimana, Goldsmith, Nunow, Roba and Biashara2017, p. 564) show in a detailed analysis in Meru county that, while households with access to off-farm income have generally been able to expand landholdings, these represent a minority of rural households. The vast majority of others have only been able to acquire land through inheritance, leading to increasing fragmentation and pressure on land. Fibaek (Reference Fibaek2021) shows similar patterns nationally – with increasing rural stratification alongside falling farming incomes and productivity across strata, with the wealthiest households increasingly pursuing reinvestment in off-farm income. From the 1990s, there was also a dramatic expansion of new crops, notably horticultural exports, including cut flowers and fresh vegetables for European markets. This has been reliant on the growth of a landless or semi-landless population engaged in wage labour, often on a casualized basis, and the incorporation of smallholders into precarious outgrower schemes (Dolan, Reference Dolan2004; Hakizimana et al., Reference Hakizimana, Goldsmith, Nunow, Roba and Biashara2017).

Critically, the reforms introduced in the final years of colonial rule and first decades of independence left intact both the shape of the commercial financial sector and the privatized character of land ownership – indeed, the main thrust of colonial reforms was to introduce private land titles into reserve areas in a failed effort at increasing access to credit for African smallholders. Colonial and post-colonial states unwilling or unable to directly challenge the patterns of private landownership established under colonial rule have been faced with escalating conflict over land and access to water, exacerbated in the 1970s and 1980s and again since 2020 by severe droughts (see Mkutu Agade et al., Reference Mkutu Agade, Anderson, Lugusa and Atieno Owino2022). The point here is that the place of land as a financial infrastructure in the colonial era has had enduring impacts both for the geography of Kenya’s financial system and more widely in social and ecological terms.

6 Conclusion

In this chapter, I’ve argued that financial infrastructures offer us an important lens on the durability of colonial histories and the ways that the latter pose problems for contemporary development. Colonial financial systems were radically uneven in both social and geographical terms. This has had critical consequences for subsequent development. Looking in particular at the example of Kenya, this chapter has shown how bank branch networks and land titling operated as key infrastructures of colonial financial systems, and how the uneven development of these infrastructures mapped onto the wider uneven development of colonial financial systems and economies on one hand, and generated important problem for post-colonial development on the other. The historical details presented here are necessarily more suggestive than comprehensive. However, they do suggest the value of closer investigations of colonial financial infrastructures, particularly the development of land titling and branch networks.

Colonized territories continue to mostly occupy subordinated positions in the global financial system (see Alami et al., Reference Alami, Alves, Bonizzi, Kaltenbrunner, Koddenbrock, Kvangraven and Powell2023). Equally, financial systems within colonized territories remain highly uneven, unstable, extraverted, and unable to mobilize resources and investment in ways that will foster development (see Bernards, Reference Bernards2022b; Koddenbrock, Kvangraven, and Sylla, Reference Koddenbrock, Kvangraven and Sylla2022). An ‘infrastructural gaze’ (Westermeier, Campbell-Verduyn, and Brandl, this volume) helps towards understanding how these dynamics persist. One important feature of infrastructures is their durability over time. They condition the form and geography of subsequent developments. As such, a lens focused on financial infrastructures offers us a useful way of understanding the ways that colonial financial hierarchies persist and mutate over time. This kind of perspective usefully calls our attention to the ways that historical patterns of colonial finance remain embedded in financial infrastructures, and how the contradictory development of these infrastructures over time has shaped the development of financial systems as well as wider development outcomes.

In this sense, a focus on specifically financial infrastructures is also a useful complement to a wider debate that has emerged around the ‘coloniality of infrastructure’ in recent years (see Cowen, Reference Cowen2020; Enns and Bersaglio, Reference Enns and Bersaglio2020; Kimari and Ernstson, Reference Kimari and Ernstson2020; Davies, Reference Davies2021). There’s a tendency in these debates to focus on large-scale built systems – things like roads, railways, or water and sanitation systems. These systems were, generally, built up to support an extractive colonial economy, embody and reproduce racial hierarchies, and condition subsequent development in important ways. Insofar as finance enters into these debates, though, it is often as an accelerant of imperial extraction and uneven development (as, for instance, in the late nineteenth-century railway boom, in which British capital was exported into major rail projects, primarily in settler colonial territories, see Cowen, Reference Cowen2020). One of the reasons why it’s particularly important that we turn an infrastructural gaze on colonial finance is that colonial financial systems are no less extractive, no less racializing, no less consequential for post-colonial development – and indeed no less material and embodied – than transport or communication networks. But financial infrastructures are perhaps less obviously visible, and often deliberately opaque and obscured. Colonial financial infrastructures matter a lot, but it takes work to uncover them.