8 - General Knowledge and Particular Society: Taxation as a Way of Knowing

Summary

This chapter explores the knowledge creation aspect of contemporary tax reforms in Nigeria. It offers a historical perspective on this process which lets us see today’s reforms not only as the re-creation of long-retreated systems of state taxation-led ordering, but against the backdrop of what intervened in the meantime – a four-decade late-twentieth-century interregnum where revenue reliance on oil profits created a very different distributive system of government-as-knowledge. Today’s system of tax-and-knowledge is not just reform but an inversion of what came before.

As an effort in extending extractive control over people, property, and activities, taxation is also intrinsically the creation or extension of a system of knowledge. This chapter traces that knowledge creation aspect of contemporary tax reforms in Nigeria, framed within the long-term perspective of transforming its political economy away from a distributive system based on oil rents, and towards a present based on direct taxation revenue. These reforms require new forms of knowledge that enable both collection of surplus in the present and financialisation of the future. A historical perspective on this process lets us see today’s reforms not only as the re-creation of long-retreated systems of state taxation-led ordering, but against the backdrop of what intervened in the meantime – a four-decade late-twentieth-century interregnum where revenue reliance on oil profits created a very different distributive system of government-as-knowledge. Thus, today’s system of tax-and-knowledge is not just reform but an inversion of what came before.

Nigeria’s institutions of governance and politics reflect close to a half-century of formation around a logic of central distribution of resource rents, and the data, forms of knowledge, and even the categories in which official knowledge is ordered bear the hallmarks of that, whether it be the ‘macro’ level of how the human resources of the state are classified during recruitment (Suberu 2001) or their everyday manifestation in how paper forms which gatekeep the processes of state are structured (Mang & Ehrhardt 2019). From 2015, however, the revenue scale began to tip gradually and inexorably back towards non-natural resource revenues (Burns & Owen 2019, 2023) and with it followed a slow, emergent, but tangible reworking of the forms of knowledge and data needed to facilitate new forms of political economy. This gives an observer a window into a real-time process whereby the use of often descriptive or inflated forms of public data is transforming into a need for more precise and traceable forms of data which are tied to revenue collection imperatives. With taxes on income and property foremost among the new revenue sources, here I particularly trace the reordering of knowledge about place, property, and activity as they are reconstructed and reinscribed in official record in novel ways. In a way, this forms an extension and sequel to the story of natural resource-reliant rentier states familiar from social science literature, and shows how a careful ethnographic eye to actual revenue process can detect early and real changes in the actual examples of such archetypes.

Perhaps inevitably, this chapter draws on the insights of James C. Scott’s influential work on how states ‘see’ – how governments render things differently visible, legible, or salient, through ways of ordering and knowing (Scott 1998). It further derives lessons from Tim Mitchell’s (2002) careful study of legibility as a constructed system allowing patterned regimes of value to interact; a wonderfully full depiction of how states work on worldly realities through regimes of measuring, knowing, and registering, which move from the material to the abstract and back again in ways that endow even mundane administrative techniques with deep and powerful consequences, both intended and accidental. Broadly, these studies (and mine) sit within a framework which owes much to Foucault’s insights on how power and knowledge are manifest in combined, inseparable techniques of knowing-doing.

What this study adds is to catch one such array of state power/knowledge at a moment of its transformation. This allows us to see a logic thrown into reverse – the messy yet discernible process of a system of knowledge built to facilitate the workings of a distributional oil-funded state changing into one which can facilitate a public political economy rapidly rebuilt around a return to taxing citizens, their activities, and property. That involves sometimes discarding (but sometimes repurposing) some old practices and actors, and bringing in new ones; rediscovering lost techniques of governance and also importing new ones from elsewhere; and rebuilding not only forms of registration and the data they generate, but also making them meaningful in new ways; making numbers speak to different purposes and audiences, and with a different resonance and aesthetics.

The historicised transformations in statistical capacity of African states are equally a part of this story. Morten Jerven’s (2013) book, subtitled How We Are Misled by African Development Statistics and What to Do about It, importantly documents the decline of technical statistical capacity in African countries. Jerven shows it as coincident with the deliberate destruction of independence-era developmental states by the local–global nexus of misgovernance and forced structural adjustment. He demonstrates that this paucity of accurate statistics is not merely a technical problem. Rather, it has a huge impact on the welfare of citizens because where policy is made and targets set on the basis of faulty, incomplete, or totally invented information, it will be bad policy. Moreover, the targets will be unrealistic or impossible, leading at best to failure, and at worst, to irreparable damage to nations, places, and lives.

However, I contend that this portrayal is only half of the story, one that misses the central ‘why’. At least in Nigeria, and perhaps in many other developing nations, the decline in this capacity was not just a story about a depleted and demotivated post-structural adjustment public sector becoming increasingly unable to produce to the quality level intended. Rather, it is also about the demand for those statistical services. An impaired capacity to produce accurate data is, I argue, the absolutely predictable counterpart of a political dispensation based on the deliberately imprecise allocation of resources within a distributional logic, facilitated by a lack of transparency. It is a story of systemic intentionality rather than one of neglect, and its regrowth should be likewise seen in the light of intentionality, albeit to a different end.

At the basis of this perspective lies the realisation that tax and revenue systems do not only collect data, they create it, and do so with powerful effects in the social world far outside taxation or fiscal governance. Since the Domesday Book which calculated the surplus value that could be extracted from each landed parish of England and its free and unfree inhabitants by the colonising Normans, ways of knowing have been the inseparable counterpart of ways of profiting. Taxation and the recognition of property even intrudes, for instance, into how identity and personhood are fixed and internalised: for example, think of the gradual transformation of naming practices in Wales from patronymics to fixed family surnames under new forms of rule imposed from England from the 1540s.¹ Against traditional patterns of naming, numbering, and referencing, as noted here, taxation counterposes the requirements of bureaucratic rationality. But it is not simply redoing. The context I depict here reminds us that it is also undoing. Nigeria today is not engaged in what an overly presentist or modernist view might mistakenly see as the clearing of bush in the virgin lands of data – a rolling out of governmental bureaucratic rationality over a socially variegated ‘wild’ landscape of information, identity, and place making. Instead, it is a remaking among a longer history of continual remakings, an instalment in a dynamic story of changing political economies, in which ways of knowing have a central part. Ways of knowing and seeing peoples and places, embedding as they do philosophies of order in the service of systems of power, already capture the ebbs and flows of Nigerian history as it is lived with in the everyday.

In what follows, I explore both the present reordering of governmental knowledge about and power over publics, but also look at what it disrupts and reverses – a particular pre-existing dis-order of knowledge which may at first look like ‘no method at all’, but which I argue is in fact the information manifestation of the quintessential oil state which Nigeria has exemplified for nearly fifty years, and which is now in the process of unbecoming. So, before looking at the ways in which places and people are being reordered and reinscribed under new revenue imperatives, we need to look into the past, at how the systems previously in place came to be so. The entry point for this study is to follow the kinds of knowledge which have been selected as useful bases for revenue governance in this case. That might be knowledge about people, places, property, or activity, because each of them has something to offer in (respectively) enabling the administration of income taxes, property taxes, or taxes on transactions.

Public Revenues in the Context of Nigeria’s Political Economy

Nigeria, in common with many of the African territories taken over as non-settler colonies by European powers, has a revenue history which reflects its insertion in the international political economy of empire. This encompasses attempts to develop those spaces economically without spending metropolitan resources, and also the plurality of motives for taxation which included not only raising revenues but stimulating forms of trade, labour, and production deemed ‘useful’ to the colonial project, while deterring others (Akiga 2016; Burns 1973; Heap 2008; Lugard 1965). I explore this story in depth elsewhere (Owen 2018). Here, it is enough to say that overall there are four phases, albeit overlapping and gradual. The story starts with taxes on overseas trade in step with the mercantilist ethos of Victorian imperialism: Nigeria’s Customs service was founded in 1891 – established (like the railways) before the country even existed. Then immediately afterwards came a drive to raise domestic revenues to pay for the colonial machinery of governance, infrastructural development and services, and for the indirect effect of stimulating the development of wage labour markets. This second process of domestic resource mobilisation deepened, intensified, and became more sophisticated² along with the development of agricultural, industrial, and tertiary sectors from the 1900s through independence, until the 1970s when large quantities of oil began to be exploited. The third, late-twentieth-century part of the story, the coming of oil to dominate the Nigerian revenue base and consequent political economy, is well known to scholarship. In fact, Nigeria at its oily peak – 85 per cent reliant on hydrocarbons for export earnings and structured around the central sharing of oil revenues – is often used paradigmatically to illustrate the dynamics of natural resource rentier economies (Soares de Oliveira 2008) in scholarship, public discourse, and even film (Boyton 2013). Sociologically, its combination with the institutions of a developmental state to create distributional logics, both formal/constitutional and informal/illicit, is also well explored (particularly see Joseph 1987; Suberu 2001). And, the ramifications, reflections, and refractions of this political economy in everyday values, popular culture, practices, and aesthetics have also been the subject of some landmark works (Adebanwi 2017; Apter 1999, 2005; Smith 2007).

The debasement and re-basement of value and values in this process are usually depicted as an authorless evolution. In fact, as I explore elsewhere (Owen 2018), it is less well known that this transformation into a distributional oil-backed state evolved in lockstep with the deliberate deconstruction of previously existing direct taxation systems by political actors who considered themselves some of the country’s most progressive modernisers. Leftist progressives who ran Kaduna State in 1979 were, for instance, the first to remove unpopular cattle and head taxes amid the enabling environment of plentiful oil money (Musa 1982).

The fourth phase is the emergent present day. While most popular and scholarly understandings, and much everyday discourse and practice, still reproduce the image and superstructure of Nigeria as an oil state, I elsewhere demonstrate that since 2015, non-oil revenues once again marginally dominate (Burns & Owen 2019), and alongside that we see a return to domestic resource mobilisation efforts and debates around devolved taxation powers. This is the fiscal manifestation of the work of reconnecting state to society that has been in progress since the 1999 return to democracy.

It is, however, much more rarely noted that this transformation towards a centralised oil-extraction resource base also reversed the political economy of information accordingly, throwing the gathering of information about populations, production, activities, and places into reverse polarity around a distributional logic. Further, this is a process which we can now watch unwinding and reversing in emergent signs of political-economic logics which mirror the move away from an oil-dominated resource base. In this chapter, I use as examples the revenue allocation formulas of the Revenue Mobilisation Allocation and Fiscal Commission (RMFAC), and also the effects on crude majoritarian democracy – the evolution of a ‘game of numbers’ which permeates everything from demographic profiles to victimhood lists. I argue here that this pattern, starting in the late 1950s, devalued numbers themselves as purporting to be accurate and actionable knowledge, and revalued them as descriptive, often inflationary, and emotive expressive practices designed to effect action on public discourse, to thereby stimulate the distribution of public goods. Later, we will see how this is now being reversed in the present-day re-emergence of ways of labelling counting and measuring place, property, and activity.

The political utility of numbers, and the temptations of inflation, was clear in the political sphere even before independence,³ and before the advent of oil money and distributional logics, in the region-against-region contest for political supremacy via the ballot box wherein registered population translated directly into voter numbers and thus political power. Former colonial officer John Smith, in Nigeria working for first the British, then the Nigerian, administration of Northern Nigeria from the mid-1950s to late 1960s, remembers citizens of northern Nigeria celebrating ‘we won the census’ when the results of the 1962 census were announced showing Nigeria’s then three regions of North, East, and West having 55.1, 24, and 20 per cent of the population, respectively.⁴ Each region had densely populated urban or semi-urban zones which could justify the high claims, and equally each had grounds to cause others to cast aspersions on the numbers.

Besides, what methodological accuracy could be assured when, as Smith recalls, in many rural areas the enumerators were secondary or even primary school students, the only mass workforce with suitable literacy in English, and especially so in traditional Islamic towns where women were kept apart and male strangers like census enumerators were invited only into the Zaure?⁵ By 1962’s census attempt, which was conducted amid emergent civil–political crises in both the Western Region and the Middle-Belt minority areas of the North, the count had become such a contested free-for-all that the Federal Minister of Economic Development announced that due to inaccuracies, it ‘is useless to give out any figures to the public under the present circumstances’ (Aluko 1965) – and in the end, no figures deemed acceptable were produced. Instead, the census had to be re-run in the following year – but even the numbers that were produced then were contested in court by the government of the Eastern Region.

However, the sentiment and the utility of the ‘game of numbers’ were greatly energised by the advent of oil revenues and the importance of revenue distribution. The transition to oil reliance took place at the same time as a great post–civil war transition to political centralisation and executive-led government – the military froze the parliamentary constitution in 1966 and ruled by decree, and when the rule of law returned in 1979 it was with a powerful presidential centre; at the same time, oil revenue had moved from zero at independence to over US$16 billion (dwarfing non-oil revenue of US$4 billion) in 1979.

This institutional transformation equally encompassed revenue sourcing and distribution, as three previously semi-autonomous regions (Northern, Western, and Eastern) with great powers over taxation, and which were predominantly funded from internal revenues, transformed stage-by-stage into thirty-six states, almost all of which were outside of the commercial centre Lagos and were reliant on federal disbursement for by far the largest part of their budgets. Revenue allocation on the ‘derivation principle’ (being allocated to where it was collected) moved over time from a 50–50 split of revenue between regions and the (national) federation account in 1960, to the peak of fiscally centralised pseudo-federalism in 1992 when the Babangida military regime allocated only 3 per cent of (then mainly oil) revenues to the states where they were derived and a huge 97 per cent to the federal government (Chukwuoonte 2015 in Owen 2018).

Although some of the disbursements were budgetary and others made through grants or for statutory functions,⁶ at the heart of the logic was the formula by which monthly disbursements from the Federation Account were allocated, as overseen by the Revenue Mobilisation Allocation and Fiscal Commission (RMAFC). The actual formula covers not only what federal government, states, and local governments get, but also what states get in relation to each other – which is a potentially highly controversial and contested issue and is correspondingly almost completely unpublicised. This despite not being an official secret, it is available almost nowhere in the public domain.⁷ It is thus a piece of public data with paradoxical status – it is neither secret nor is it publicly accessible, a feature it shares with other information central to understanding ‘really existing’ governance and political economy, such as (at least in some previous years) comprehensive sets of accounts from the Customs service and Central Bank, or in a different field, the actual rules governing the everyday operations of Nigeria’s Police (see Owen 2013a, 2013b).

The formula used reflects a total of twenty indices, ranging from the size of a state (land mass), to terrain, rainfall (a proxy for agricultural productivity), revenue effort, developmental indices such as primary school enrolment, and population (from the census). Of these, most are small contributors to the formula – in a 2013 example, rainfall accounts for 15 per cent of 10 per cent of total weighting – but the largest two are equalisation (40 per cent in example used) and population (30 per cent in 2013). Whereas factors such as land area and rainfall are facts of geography, and internally generated revenue (IGR) has practical limits, population stands out as the one most vulnerable to manipulation, and which has been subject to such pressures for nearly seventy years. Therefore, the role of that category in the division of national oil spoils is magnified to become the revenue allocation counterpart of what voter registration numbers are to national politics. Once we understand that, it makes sense of otherwise surprising levels of sensitivity over population data – for instance that census figures may only be released after they have been cleared by a meeting of the National Council of State (composed of all living former heads of state), and that any government official quoting population figures other than the official statistics is committing an offence punishable by up to five years in prison.

This division of national goods between states matters all the more, and matters emotively, because regions in Nigeria are not just administrative units. They also contain and embed – increasingly so through successive rounds of state creation – units of ethnicity-based citizenship, wherein ‘indigenes’ enjoy popular notions of legitimacy which endow primary civic rights on them over and above those of ‘settlers’ from other areas. The mutually reinforcing process by which one group’s fear of domination in large administrative units propels the creation of new administrative units, which can themselves be dominated by the group in question, is well described in Suberu (2001), and remains a self-reinforcing dynamic fuelled by the continued flows of revenue which can be subdivided. This, despite a formally liberal constitution embedding generalised citizenship rights, actively promotes extremely particularistic popular ideals, incentives, practices, and logics. Through the 1970s to 2000s, it also created a contradictory kind of federalism which saw multiple constitutionally autonomous units in actual fact mainly tied to disbursements, and thus constraints, from the centre.

And its social function, in terms of people’s instinctive relationship to numbers and state knowledge gathering, was profound. Immediately after the last national census in 2006, I asked a professional colleague resident in London, but from an area of Nigeria strongly affected by the civil war in the 1960s, if his family had participated. He answered that he had, and that he and his four brothers had all been counted at their natal home along with their spouses – despite the fact that only one was living in Nigeria at the time and two were unmarried. While public debates over the census centred on whether relative numbers from north and south had been politically manipulated at a high level, the overall effects of this manipulation-from-below were rather seen in population profiles which differed markedly from typical demographic trends in comparable countries. In particular, an emphasis on male population registration echoes a popular discourse on national ‘virility’ often seen in media and political discourse.

Fiscal numbers, too, have been subject to expressive rather than accounting logics. One former commissioner for finance of a state in Nigeria’s North-Central zone described the process of budget finalisation as ‘we wait for [names two neighbouring states] to announce theirs, and then announce a figure which is N100 million more’.⁸ By stages, we can see how particular political-economic structures and practices have shaped an aesthetic of numbers as expressive claims to size, significance, and entitlement, which has been internalised not just by governments but even at the level of individuals and families. As a political-economic but equally a cultural process, it is akin to the destabilisation of the economy of signs and signifiers traced by Apter (1999, 2005) in Nigeria’s oil boom.

The Political Economy of Knowledge – The Functions of Description and Legibility

Once we understand that there is an important functional fit between the practices and habits of a distributive political economy on one hand, and a descriptive and emotive rather than accounting-based approach to numbers and statistics on the other, it is easier to see it manifested in multiple spheres. Not only population at a state level, but Nigeria’s national population is frequently misquoted or inflated in attempts to demonstrate political and economic significance. Not only voter registration figures (with a clear utility for election rigging), but also casualty figures in episodes of communal conflict are used to create narratives of victimhood. Not only years of age to impress upon one someone’s elder status, but amounts of money stolen exaggerated in police reports are used to emphasise grievance (see Owen 2013a).

Equally, this historical trend has not only been about the usefulness of smokescreens of inaccuracy to facilitate dissimulation; not only about deliberate disinvestment in exactness; and not only about imprecision as a political technology. It is also about an accompanying trend more focused than that, a reinvestment in the ultra-specific knowledge of the particular – of elite networks, social relationships, and lifetime-formed reserves of social capital – and at the same time a de-investment in the purportedly scientific standardised data on which modernist state creation was and is held to depend. It is the de-valuing of general knowledge over and about populations, and its re-valuing as a reserved marginal advantage, in creating alliances, doing deals, and/or making profits. In such a system of personal favour and leverage, constituting a logic based above all on social capital, knowing ‘who’ is a much more useful asset than knowing ‘what’. Accompanying that asymmetry, and the long apprenticeship in social networking within which it is created, is the fact that all information has an entry price and is not meant to be accessible to all, and that this is accepted as, in some way, legitimate. The circularity of what should be learned but should not be taught is replete in the Nigerian English phrase ‘you are supposed to know’. The sentiment thus implicitly also encompasses knowing who is not supposed to know – that is, those who have not been born into it, have not earned their place nor ‘paid their dues’ in experiential rites of passage.

If rationality and particularism were always covalent in the system of indirect rule, the 1980s–1990s saw the abandonment of one for the greater adoption of the other. In fact, by the turn of the millennium, the creation and organisation of knowledge about the public realm in Nigeria closely resembled the layout of a medieval city as memorably characterised by Scott (1998: 53); it was organic, specific, and particular, and with it, having a specific and differential legibility:

The fact that the layout of the city, having developed without any overall design, lacks a consistent geometric logic does not mean that it was at all confusing to its inhabitants. … For a stranger or trader arriving for the first time, however, the town was almost certainly confusing, simply because it lacked a repetitive, abstract logic that would allow a newcomer to orient herself. The cityscape of Bruges in 1500 could be said to privilege local knowledge over outside knowledge, including that of external political authorities.

This metaphor allows us to recognise the functional fit with many other aspects of the public sphere in turn-of-the-twentieth-century Nigeria. Prime among these particularistic logics has been the emphasis on indigeneship rather than common citizenship as a basis of rights and entitlements, and in the process the abundance of small and vicious wars of ‘who is who’ (Marshall-Fratani 2006) between groups contesting the identities of ‘indigene’ and ‘settler’ throughout the country – as well as smaller wars labelled ‘chieftaincy disputes’ between individuals and lineages on the even more localised specifics of who can represent, and thus access, state goods on behalf of each particular group. Related to that, the system of constituting political structures and of recruiting to public office on the basis of ‘Federal Character’ (Suberu 2001) which relies itself on such conceptions of indigeneship which are rigid in concept and changeable in content. Also related to that is a system of internal democracy in political parties which emphasised the rotational ‘zoning’ of offices to regions (and thus implicitly, certain groups inhabiting them) in such an accepted pattern of legitimacy that to attempt to upset it entailed losing national elections (Owen & Usman 2015).

Meanwhile, the hyper-particularism within which social issues, tensions, debates, and their solutions were framed introduces barriers to comprehension (and thus to access, and to participation) both internally and externally. The specificity of knowledge compartmentalises national society for both insiders and outsiders. Interestingly, at a historical global level, Scott (1998: 53) identifies that the illegibility of society or landscape created a certain relationship to the outside world:

Historically, the relative illegibility to outsiders of some urban neighbourhoods (or of their rural analogues, such as hills, marshes, and forests) has provided a vital margin of political safety from control by outside elites … the community or terrain in question enjoys at least a small measure of insulation from outside intrusion.

Nigeria’s informational illegibility, the specificity of knowledge, and the particular positionality of those who are able to use it, or at least of those who are able to leverage their social capital, experience, and skills to assemble a more complete view (because there is no fully panoptic position in this system, not even the view from government), serves a similar function in defence of patriarchal conservatism. Politico-social formations, which gatekeep both state and social processes of identity formation, are able to reproduce barriers to entry which filter and limit the participation of competitors. And since most often those formations are mediated by older men, those competitors who are filtered out are likely to be women, the young, and outsiders. This is yet more powerful, because it is not only about filtering competition, but about preserving trust and a functional society in a way which frequently uses socio-cultural markers to bridge across the unusable incompleteness of state processes.⁹

As a further effect, a similar illegibility certainly manifests in Nigeria’s relationship to the outside world, and does so in multiple ways. In one aspect, it appears to amplify the sense of risk in travelling or otherwise engaging there for the unwary, making Nigeria in the imagination of many a much more dangerous location than some which are more legibly presented but are in actual fact more violent, such as South Africa or Brazil. It has also, arguably, limited the penetration of foreign direct and portfolio investment as global financial markets can only incompletely ‘see’ into the real economy of Nigeria. This has had both positive and negative effects – on one hand limiting capital availability, but also insulating the country’s financial sector from the modern equivalent of ‘control by outside elites’, namely the speculative predation of financial markets; as seen in the interesting lack of destabilisation of Nigeria’s banks amidst the 2008 US financial crisis. So at both macro and micro scales, the decomposition of mid-twentieth-century modernist attempts to build the state of numbers and its institutions has been not only Jerven’s story of lost statistical capacity, but of real material and functional effects, both iterative and deliberate. These have reconstructed in its place a particularist edifice of knowledge which lives not in institutions but in situated actors and socio-political capital.

The Reconstruction of Governmental Knowledge

The second half of my story is the reconstruction of modernist forms of state knowledge after the destruction of Structural Adjustment Programmes and military rule, one function at a time. This is the story of governmental bodies trying to get a grip on an exploding population, sprawling urbanism, and other aspects of social life which had continued in their full dynamism across a long interlude with only limited reference to the state. And increasingly, it is the story of doing that because knowing once again matters economically; as oil becomes a steadily less dominant part of the economy, economic activity recentres on non-oil sectors and revenue effort must follow it, even if slowly and clumsily. As state functions were re-extended, each brought their own indexical ways of knowing – registration numbers, logs, receipts, databases, targets. A primary visually significant tactic for this recolonisation is inscription – all manner of things, like numbering exercises for vaccination programmes, building licences, and listing for demolition, are commonly scrawled on house walls in chalk and charcoal. They reflect that knowledge is not unified, but rather inscribed; it is a functional coming together of disaggregated state practices.

Nigeria lacks a single central means of population registration; so, when a means of knowing people with more exactitude than name and biographical detail is promised, institutions and governmental processes make use of one of the forms of partial knowledge which promises to be most adaptable to the need at hand. Ideally, tax authorities at both state and federal (national) levels would use the nationally administered Tax Identification Number (TIN) number, but that is only usable for the minority of the population which has already registered as formal taxpayers. Meanwhile, the National ID Card scheme has only reached partial coverage of the country and is not frequently updated; several people have neither a National Health Insurance card, nor a driving licence, although many have a bank-issued Bank Verification Number (BVN) to identify them within the banking system, and more recently enrolment for a National Insurance Number (NIN) has gone up when it was required for mobile phone SIM card registration. But in this realm of partial overlapping formal identifications, for instance, a programme handing out agricultural inputs to farmers requests voter cards as proof of identification, in much the same way as a driving licence might be considered a valid photo ID in many countries for numerous purposes which have nothing to do with driving.¹⁰

These state practices come together not only around the rolling out of social programmes, or the need for revenue, but around the project of importing the future into the present. Revenue flows have a function in the present – enabling the state to pay its bills – but also against the future, by providing indications of creditworthiness which then dictate access to potential credit in the form of the bonds – a process so common that it has led to a subnational public debt boom in Nigeria over the last decade. So, enlarging the tax base is also a tactic of invoking possibility, and tax numbers are also a way of triggering a productive relationship with another abstraction, that of finance, just as were the mapping surveys of Egyptian fields in Mitchell’s study. Financialisation is another significant driver of the creation of data and, just as importantly, its distribution rather than reservation. As Nigeria’s thirty-six states increasingly looked to bond markets at home (and sometimes further afield) through the 2000s, it became increasingly necessary to create, and share widely, accounts of a quality which could be defended to potential lenders.

Since 2011, the growth in legibility has also been reflected in the story of Nigeria’s National Bureau of Statistics (NBS) under the noted statistician-general Dr Yemi Kale. The NBS collects and publishes increasing amounts of social and economic data, including quarterly gross domestic product (GDP) reports which are considered to be of integrity and independent of political interference. In line with the argument of this chapter, we should see this not just as a fortunate development, but as being enabled by a historical epoch in which such information is once again needed and valued, rather than being as it was in the preceding period, an inconvenient truth standing in the way of the fluid manipulation of numbers-as-symbols in the service of distributional patronage politics.

The registrational expansion is also locked into another political economy: that of ‘exercises’ and capital disbursement. One of the aspects of many bureaucratic practices in Nigeria most salient to strangers is their reliance on ‘exercises’ – supposedly extraordinary mobilisations of effort – to fulfil basic institutional functions which might be thought of as continual and routine aspects of ordinary running, for instance, recruitment, the elimination of ghost workers, or verification of pensioners. That they are not routine reflects several factors. One is the lack of automation which means that records are hard to keep continually updated – and the advance of information technology and dematerialisation is part of the gradual and ongoing erosion of these trends. Another factor is the extremely hierarchical and executive-led nature of such state institutions, in which, for reasons both good and bad, those in charge are unwilling to relinquish too much power from their own desks, and often prefer to work by the bureaucratic equivalent of ‘commandement’ (Mbembe 2001) and mobilisation by direct hands-on executive control. And a third factor is that working by such exercises offers the opportunity to enact budgets and create capital flows attached to them, which is what the public sphere thrives on as its economic base.

Yet because this is a process commandeered from the middle, rather than organically grown from the base or effectively imposed from the top, it has tended to reproduce partially and manifests in a continual incompleteness of knowledge. One example is the incomplete issues of national ID cards and health insurance numbers, rather than their rolling capture in a routinised fashion until this was required by telecoms security policy. In fact, the only truly continually issued piece of state registration is – tellingly – the taxpayer-identifying TIN number given to businesses and registered individuals which is obtained by application.¹¹

Measuring and Counting – Place, Property, and Activity

Now, in step with the political-economic transition away from oil and back towards taxes, we also see the reversal of distributional logic, slowly and with setbacks, but a clear ongoing process nonetheless. Like everything else in this chapter, it has been a gradual and emergent process which overlaps with the opposite trend. But if we were to note some key tipping points, they would be three in number – the exhausting of Nigeria’s ‘excess crude account’ windfall from the 1990s oil boom, disbursed in ever-smaller non-dedicated packets to the thirty-six states when they were able to exert enough political pressure; the point in 2015 at which non-oil revenues overtook oil revenues for the first time since 1971 (Burns & Owen 2019); and the point in March 2020 when, amid the global COVID shutdown, the oil price temporarily entered negative figures – at this point, it actually cost more to store oil than it was being paid for on the traded market. This became a crisis which, because of the lack of other reserves, had a near-instant feed-through into Nigerian government accounts. These three events/processes together map out a moment in which IGR moved from a ‘nice-to-have’ supplement, to oil money, and then to an absolute necessity – both for the national (federal) government, which collects corporate income taxes, and state governments, which have powers to collect property and personal income taxes. Where distributional logics encouraged the descriptive, inflationary use of numbers as a symbol of powerful plenitude, revenue collection ‘money-in’ logics dictated the opposite – the precise and careful attributable use of numbers in a corresponding and traceable way. These manifest in several interesting ways. For the purposes of this study, I note three examples – of place, property, and activity.¹²

In Lokoja, Kogi State, the fading street signs of the grid-patterned colonial cantonment area display old British names such as ‘Elphinstone Street’, which nobody uses. In the century and more since their naming, residents have discarded even the idea of streets themselves as a useful locator, instead relying in everyday conversation on more deeply rooted referents of topographical location (above and below), and of notable residents and the people they attract around them. In practice, this plays out as descriptors such as ‘up garage’ or ‘Sheikh Jafaru quarters’. Traditional ways of naming and placing, based on the presence of significant persons, or types of economic activity, have regrown over colonial attempts to impose grid-patterned models of urbanisation and knowing which enable the rational and efficient location of people, whether for postal services or taxation. And yet, as property taxation, along with the placing and taxation of people, is a new necessity, in other states new orders are being imposed over this socially evolved system of urban identification – in Lafia, Nasarawa State, at the time of my fieldwork in neighbouring Benue, new street signage and numbering was being rolled out to reimpose some kind of unmediated legibility.¹³

Since property taxation itself is one of the key arenas in which states can tax wealth – as a fixed asset in a milieu where incomes and even people are hard to trace – the creation of knowledge about location and property has also been a locus of much innovation and activity. This goes beyond street signage to cadastral survey and sophisticated ways of using GIS and satellite imagery to locate and attribute properties and to rate them for tax on the basis of their size, location, and use. Schemes like Lagos State’s Land Use Charge, or Kaduna State’s KADGIS,¹⁴ have assembled sophisticated and efficient data sets to enable billing and tracing. However, this involves skill sets not available in the existing public sphere, and so for both technical and political-economic reasons, such technological advances have often been outsourced to private companies which charge a percentage on the revenue generated from these technologies. This ‘neoliberalisation on a Nigerian pattern’ introduces debates over value for money, as well as speculation over the political connectedness or otherwise of the technology partners.

More relevant for this study, however, is first, the additional layering of another type of knowledge – the combination of technocratic order and discretionary practice. The hallmark of this hybrid order as experienced by those who operate and interact with it is that it is technically transparent from above, with valuation formulas and collection rates instantly visible, satisfying the requirements of technocratic discourses of visible data and accountable governance. And yet at the same time it is utterly opaque from below, with land use rate demands often issued without explanatory breakdown and in some states subject to itemised ‘Premiums’ which can be applied or removed at the discretion of administrative or political actors. Second, the ownership of this (technical) knowledge matters: since the database is the tool, it is usually owned by the outsourcing partner, so when administrations change, as in Lagos in 2015 when the incoming administration severed the contract of the consultants who implemented the state’s new Land Use Charge property tax, the implementing corporate partners, who are often linked to political parties or networks, may be summarily changed too. This means that with them departs their whole assemblage of power/knowledge, bringing the state’s newly acquired capacity back down again. For this reason, the policy forum of the Nigeria Governor’s Forum (the body representing Nigeria’s thirty-six states) now advises that these new functions be brought in-house within state civil services for the purposes of ensuring their sustainability.

The third arena in which great effort has been put into creating new spheres of knowledge has been economic activity. Personal Income Tax (PIT) is also collected by states, but only a very few workers are formally salaried – those who mainly work for large national companies who pay all PIT in the state in which their headquarters are located. Therefore, new ways must be found to get a handle on the volume and velocity of economic activity. One crude form is the recruitment of ‘revenue consultants’ to collect fees from goods transport at roadside checkpoints. Another is the setting of ‘presumptive’ income tax rates which can be levied on the general public in the absence of any precise knowledge about what they are earning.

But there are also more sophisticated approaches aimed at panoptic knowledge and control. During my fieldwork in Benue State in 2015, the Benue Internal Revenue Service had installed a consultant-designed system of taxing markets which was operated by government employees. It involved traders at markets across the state paying an entry fee into a point of sale (POS) machine in return for a ticket receipt, with the figures transmitted in real time to a large television screen in the BIRS chair’s office – a landmark in information panopticism. Yet, at the same time, the same revenue agency was pursuing almost the opposite policy with sectors such as tradespeople – strategies of associational taxation which relied upon leveraging the strong personal networks and social knowledge of particular trade associations (see Ayee & Joshi 2008; Meagher 2013 for more on their functioning). Rather than increasing the power/knowledge complex via panoptic data like that generated on the real-time market taxes screen, the result of also deepening a reliance on associational taxation was a form of governance which therefore reinforced particularism and bound it to state functioning as a new ‘indirect rule’ in which government function is tied to working through pre-organised groups in society.

Interestingly, the one disposable form of knowledge in all this was, tellingly, the census. The decennial census calendar dictated another round of national census was due in 2016. Yet when that time came around, Nigeria was deep into a fiscal crisis, and a remarkably consistent cross-party consensus held that a census at that time, with its accompanying logistical costs, was a disposable luxury. As of 2022, the promised national census is still being discussed as a theoretical future event which would be held at the very end of President Buhari’s tenure, after elections to forestall any controversial effect on political expectations or outcomes. The lack of urgency around this supposedly crucial statistical reference point perhaps shows most tellingly that its utility has been supplanted, perhaps in part by the weakening distributional imperative, and in part because the widening cluster of forms of information about populations – from voter cards to tax identification numbers – now performs its own functions in a way which renders census knowledge obsolete or irrelevant.

Conclusion

To renumber is to elide complexity, as Scott says. New forms of ordering replace and erase specificities with somewhat fictive but functional sets of numerical data that can be used as useful abstractions to enable transaction, prediction, financialisation, and other derivative functions. Yet that underlying specificity and complexity – which encompasses those other bundles of rights or interests – does not go away, but is instead displaced to other realms where its continued social legitimacy, against the grain of official state recognition, gives it an insurgent power. It manifests as an inconvenient lump under the neat carpet of state overview and abstraction, invisible until the neat plans of bureaucratic policy implementation trip over it.¹⁵ So, instead of a one-way transformation in which a state based on particular and specific information which enables distributional politics turns into a modernist, statistics-based order, the practical logics of social ordering and rule reincorporate the particularisms.

The expansion of registration and measurement in the course of revenue and other policy rollouts also embeds the multiplicity of agencies vying for control via a multiplying constellation of ID cards, registration numbers, licences, receipt letters, and the like. But while this looks like high modernity, it also re-embeds the low-cost colonial public administration tactic known as ‘indirect rule’. That is to say, since state capacity is limited, tactics such as levying taxes via trade associations reinvent a disaggregated, franchised-out state system. So again, we have reproduced a hybrid way of knowing, in which the state’s apparent iron cage of bureaucracy rests on society’s fluid foundations, caught in multiple webs and regimes which also create a plentiful space to articulate between them, to haggle and play arbitration games.

The functions of new forms of knowing also have very different implications for those who are to be counted, and which condition the tactical choices of citizens to be visible or not in very different ways. Unlike a census, it is likely that no one subject to a tax registration drive ever inflated their presence within it (though there may still be other reasons to remain visible; see Goodfellow & Owen 2020). To contrast the changing politics of counting and being counted as part of distributional logics, today we can counterpose the tactics of boycotting voting and voter registration by separatist groups such as the Independent People of Biafra (IPOB) in exactly the same Southeastern region that my former colleague’s census inflation was taking place just fifteen years before; an epoch-marking about-turn away from tactical enhanced visibility and towards a protest of non-cooperation with the game of numbers.

So, seen from the perspective this chapter has taken, state-generated numbers and knowledge are not only the corollary of workable, plannable, evaluatable public policy, as Jerven’s functional analysis would have, nor the powerful abstraction which enables value extraction, as Mitchell’s application of Foucault unveils, but something more emergent, evolving, and yet powerfully directive. They are a political economy of knowledge facilitating certain types of extraction and distribution which co-evolves with them, both causing their transformations and being further enabled as an effect of them. This resolution sidesteps the planned/unplanned dichotomy and instead shows a circular and functional fit which connects political economy, information, cultural representations, and social capital as essentially related forms of each other. From 1960, first in becoming a distributional oil state, and now most powerfully at the present moment of structural economic transition away from that, Nigeria’s landscape of governmental knowledge and data can be seen responding to those historical forces, contesting them and transforming in real time.