Introduction
In India, trillions of rupees, collected by the state through hypothecated taxation and ringfenced in special-purpose administrative vehicles, remain untouched.Footnote 1 Why? Conventional explanations point to institutional failures. The state might lack adequate rules, officials, systems, and know-how to get the money out of the door and into the pockets of those who need it – a result, perhaps, of limited political will to build administrative capacity to spend on vulnerable groups with little political clout; or weak social pressure to tackle their plight owing to their relative social invisibility.
Beneath these explanations is a symptomatic approach to underspending. I say ‘symptomatic’ as these explanations take as their backdrop the modern fiscal state: a specific institutional edifice emerging from, and embedded in, the social and political pressures that drive revenue collection and expenditure over space and time. Against this backdrop, taxation and expenditure are sites where the state, and its relationships with citizens and society are constituted and contested. By contrast, underspending is simply a symptom of failures of the institutional edifice, and/or a particular and unusual arrangement of social and political pressures. In other words, underspending is understood an effect of power that is already legally or politically constituted – meaning that underspending is not of state theoretic concern, and is at best an interesting entry-point for a discussion of the flaws and features of any given fiscal state, in its social and political context.
In this paper I argue that, when approached with a different set of conceptual and methodological tools, underspending might be understood as producing distinctive institutional forms and logics of the state, just as taxation and expenditure do. The paper works through an example of such an argument. It focuses on the ‘cess’ in India, a taxation instrument that levies for a specific purpose – such as the welfare of particular categories of labourer, or for the construction of roads – and is ringfenced and non-lapsable (ie unspent money accumulates over time).
On a symptomatic view, the cess is doubly failed. By earmarking or hypothecating, it deviates from an ideal form of an effective fiscal state, in which ‘all sources of revenue should be paid into a single, consolidated fund[, since] the danger of hypothecation [is] that spending would always rise to the maximum permissible level of the specified [ie earmarked] revenue’.Footnote 2 And it fails in outcome also, given the significant levels of ongoing cess underspending and concomitant accumulation, which run contrary to the notion that revenue surpluses should be avoided or released, so they ‘could be more efficiently invested elsewhere’.Footnote 3 This view is premised on an understanding of the cess as an element of an already-constituted fiscal state – an institutional ‘machine’ governed by a logic of fiscal efficacy, or the efficient collection, transfer, and spending of revenue.Footnote 4 In other words, cesses and their underspending are evidence of sand in the fiscal state’s gears.
I offer a different view. Rather than a failure of fiscal efficacy, I argue that underspending is part of a mode of discretionary management of ad hoc pressures on rule. Cess underspending leads to the accumulation of liquidity in its ringfenced pool, which can then be remobilised to tackle any phenomenon identified as such a pressure. Crucially, the cess has always been collected as an addendum to ordinary taxation, whether direct or indirect, meaning cess monies are not gathered at the expense of the ordinary revenue. My claim is thus that cess underspending, and its relationship to ordinary revenue, describes a particular structure of a bifurcated state – a state, characteristic of colonial indirect rule, defined by its particular combination of a set of formalised and rule-bound institutions (fiscal, political, legal, or otherwise) for some segment of the population or activities, and administrative institutions with high degrees of discretion for another segment.Footnote 5
In my case, we have on the one hand a familiar approach to the fisc and fiscal surplus. Here, budgets set out the ‘business program of the government’, inviting scrutiny and authorisation of that program and appropriations in its name – for example from the legislature.Footnote 6 As a corollary, in the ordinary course of things, states reabsorb budgetary underspends into the general funds and pay down debt. This allows for fiscal headroom and trust in fiscal stability. Through this approach to fiscal surplus, ‘barriers [a]re constructed to the expansion of the state. These technical accounting procedures and the annual votes of the Commons [a]re erected into matters of high constitutional principle…’Footnote 7 And on the other hand, we have the routinised hypothecation, underspending, and potential consequent reallocation of unspent cess funds to respond in an ad hoc fashion to contextual pressures on rule, based on a minimally sufficient articulation of the particular pressure.Footnote 8 This, in turn, allows the state to respond to evolving and unforeseen pressures without disrupting ordinary budgetary order.
Cesses and their underspending, then, are not simply a failure or quirk of the Indian taxation system. Their relationship to ordinary taxation describes a state form and logic. And unlike other studies of bifurcated states, this form is grounded not in the arrangement of legal institutions,Footnote 9 nor political ones such as citizenship,Footnote 10 but in revenue ones. This has important consequences for the nature of fundamental constitutional questions, such as bringing into their ambit the practices, institutions, norms, and accountability structures for the discretionary remobilisation of pooled cess funds.Footnote 11
This argument is of substantive value as it opens up new spaces of inquiry for those interested in the relationship between tax and the state. But it is also in the service of a broader conceptual move: reinterpreting something that might conventionally be understood as a marker of state dysfunction - here, underspending - as in fact constitutive of the state. This move is increasingly commonplace in a range of literatures, including post-colonial studies, development studies, and critical international relations. It is particularly concerned with refuting what these literatures see as the pathologising tendencies of disciplinary orthodoxies when describing the state in the Global South.Footnote 12 Finally, the argument also has a methodological dimension. In order to execute this redescription, I draw on historical institutionalist approaches to tracking state change – drawing on parliamentary archives, administrative manuals, and other public documents – and combine it with analysis of case law. I do so as this is well suited to my argument, and also to exemplify the methodological consequences of the broader conceptual move. Tying taxation and constitution together from a legal perspective requires a multi-disciplinary methodological apparatus: ‘To see taxation through public law lenses is therefore to increase the permeability of legal scholarship to the insights of other disciplines, and potentially also to increase our own relevance to these other fields.’Footnote 13 This, I suggest, is also true of efforts to reinterpret seeming failures of the revenue system.
The paper makes its contributions as follows. Section 1 sets out the conceptual and methodological frameworks for relating underspending, hypothecated taxation and constitutions. Section 2 introduces the cess, and how existing jurisprudence treats the cess as a form of constituted taxation power. This is a symptomatic account of underspending, and serves as a foil for the subsequent sections. Section 3 conducts a historical institutionalist tracing of the cess in India, turning to its colonial-era roots to show that it produced high levels of discretion for fund administrators to meet ‘local’ (ie ad hoc and contextual) demands on the colonial state. Section 4 traces this logic forward to the contemporary Indian cess, where underspent cess funds form a reserve of liquidity through which state administrators can deal with the exigencies of rule. It argues that this produces a bifurcated state form, demonstrating how to draw out a broader constitutional argument from the case. Section 5 maps this argument onto the contemporary European context, where it identifies a similar bifurcation of revenue practices and institutions, with attendant constitutional consequences. The paper ends with a brief conclusion.
1. Concepts and methods
(a) From revenue and the state, to tax law and the constitution
In this section, I sketch out the broader literatures that intersect with the relationships between tax law and constitutions – fiscal sociology and political institutionalism in particular. These literatures have a lot to say about historical, comparative, and contemporary processes of state formation, the centrality of the fisc to them, and the political constitution more broadly. I set them out here, and develop an argument for the relevance of law to them. I conclude by situating hypothecated welfare underspending as an instance of this relevance.
I begin, as do many, with Schumpeter’s seminal work.Footnote 14 He traces the historical shift in Europe from a demesne state (where resources were gathered from the king’s domain) to the tax state (where elites and the broader population provided the state with funds). The tax state created an architecture for an increasingly ‘impersonal’ mode of rule by producing a distinction between ‘public and private social realms … out of an historical fiscal crisis’ (ie the need to fund wars).Footnote 15 The state, in other words, was produced through taxation of the estates in the name of a common response to a ‘common exigency’.Footnote 16
The need for expanded revenue collection and expenditure led to various forms of institutionalised settlements amongst elites that shaped (in the sense of both enabling and limiting) the exercise of prerogative power. Thus, ‘[w]ith the separation of the public purse from the prince’s private household, “the tax state had arrived – its idea and its machinery”’.Footnote 17 The state could thus be understood as an efficient institutional ‘machine for certain fairly narrowly circumscribed ends’ –the management of fiscal affairs – and with it, a set of institutional mechanisms to facilitate both the smooth and accountable running of the machine.Footnote 18
Sociological accounts of the institutional patterns of the fisc-state relationship go on to explore its social contexts and effects. Here, ‘welfare and tax policy are mutually constitutive’ of the state and its social order.Footnote 19 This literature is concerned with ‘the organizational forms and patterns which emerge as different revenue-increasing [and revenue-spending] measures are designed and implemented, and with the effect of such measures on a society’s economic institutions, behavioral attitudes and institutional landscapes’.Footnote 20 Thus, fiscal systems are not only forms given to political bargains amongst elites, but also ‘forms of social contract that influence and specify relationships between individuals and their government and society’.Footnote 21 Taxation and spending, then, might also be understood in the context of social pressures.Footnote 22 From this perspective, the fisc ‘is not merely the means to pay for state provisions but an existential imperative shaping state–society relations’.Footnote 23
These literatures articulate the politically and socially – but not legally – constitutive aspects of the fisc. Yet ‘tax law plays a part in the wider sweep of history, politics and constitution-building’.Footnote 24 That is, law’s agency in this process of fiscal state-building is worth unpicking, for both lawyers and non-lawyers alike. From a legal perspective, as Dominic de Cogan argues,
it is useful to think of taxes as being both constituted and constitutive. They are constituted in the sense that they are subject to the legal system and constitutional regime of the relevant state, and, besides, cannot but reflect their political, economic, social and cultural context … Yet taxes are also constitutive, imposing their own imprint on law and society more broadly.Footnote 25
For de Cogan, the former entails studies of administrative and tax law, along with political and sociological studies of tax and welfare bureaucracies and the related public/private divides they enact. The latter is of greater intellectual concern to him, and involves not only the fiscal drivers and consequences of major constitutional moments, such as crises (so central to Schumpeter’s account); it also entails the role that tax and its administration plays in the ordinary patterning and repatterning of states. Taxation, its law, and its administration should ‘be of everyday constitutional importance’.Footnote 26 It articulates, over time, ‘revenue-raising, institution-building, regulation, redistribution, and the structuring of society’.Footnote 27
(b) Hypothecated welfare underspending and constitutions
I situate my intervention here, as a contribution to understanding the everyday constitutional importance of tax law and administration. My intervention focuses on hypothecated welfare underspending. Let me briefly justify my focus on hypothecated welfare: it is central to the fiscal activity that de Cogan sees as constitutionally important. As O’Brien notes, ‘[i]nvestigations into and comparisons across the fiscal, financial, and monetary institutions sustained by states will display complex inter- connections to their political forms and arrangements for making and implementing decisions. Connections r[u]n both ways…’.Footnote 28 Thus, while ‘[t]he study of the fiscal state is often narrowed to taxation’Footnote 29 – the sociology and politics of the formation and effects of tax policy; taxpayer consent; public budgeting; the relationship between poverty and tax policy – many studies go beyond taxation to encompass the relationship between state revenue and expenditures.Footnote 30
A fortiori for hypothecation. Although, as noted, it is understood as inefficacious in the literature on public financing,Footnote 31 it is reasonably widespread: for example, 80 countries use earmarking for their health systems.Footnote 32 And it is methodologically useful in demonstrating clearly how taxation enacts specific issues of political and social salience, backed by the state’s power to raise revenue: a practice that reveals political accountability, the limits of the state, and the politics of social (dis)advantage at work.Footnote 33
Turning to my second term, it is not so easy to insert underspending into the constitutional aspects of ‘tax’ (incorporating welfare spending). Underspending, of course, is constituted in some sense – it occurs against a constitutional backdrop. Underspending might be a means of renegotiating elements of constituted orders, for example (such as the deliberate withholding of fiscal transfers from the centre to subnational entities, as a means of tilting the federal compact),Footnote 34 or a marker of the failure of a constitution to provide for an effective state, and for an equitable state.Footnote 35 However, as noted, I also reinterpret underspending as producing a state form and logic. To do so, I develop an extreme case of underspending: India. India casts my argument in its sharpest light – although it also means that I am circumspect in my claims to the broader validity of my approach. I study the constitutional effects of underspending with an eye to both contemporary Indian jurisprudence and administrative systems; and the historical evolution of taxation institutions and instruments.Footnote 36 Combining law and historical institutionalism is valuable in general when studying the constitutions of states through the fisc, since such an enquiry requires examining ‘the mechanics of rule and workings of power through such apparently mundane state activities as the collection of taxes [or] the distribution of subsidized food to the poor’.Footnote 37 A fortiori for India, where Parekh reminds us of the importance of placing the relationships between state, society and the economy in their historical context.Footnote 38
In the next section, I introduce the cess, explain its features, demonstrate its underspending –tied to its non-lapsability – and describe how that underspending has been interpreted by the courts as a failing of tax administration. This sets up an orthodox ‘constituted’ understanding of welfare underspending, which the subsequent sections challenge.
2. Cesses in India
Cesses contributed around 10% of all gross tax revenues in India in the financial year (FY) 2016–17 (up from around 6.9% in FY 2012–13).Footnote 39 The central government can levy them broadly; state governments can levy them on certain goods and activities that fall within their taxation purview (eg liquor); and state governments can receive a mandate from central legislation to collect a cess.Footnote 40 Yet the total number and value of cesses is hard to determine. In her historical study of cesses from 1950 to 2016, Kotha notes that ‘it is difficult to arrive at the exact number’, as many cess revenues were simply folded into an accounting head ‘titled “cess under other accounts”’.Footnote 41
In spite of their significance in tax collection, cesses receive only one mention in the fiscal chapters of the Indian Constitution. Article 277 incorporates by reference all colonial-era revenue-raising measures by states and local authorities, and ‘cesses’ are listed as one such measure, alongside taxes and fees. Cesses were thus retained at independence in the Constitution as a revenue instrument, but are otherwise seemingly unaffected by the elaborate fiscal federal relationships between the centre and the states inaugurated by the Constitution’s fiscal provisions.
A core feature of cesses in India is their specific or special purpose. A cess is ‘levied for a specific object’,Footnote 42 and ‘generates revenue which is utilised for [that] specific purpose’.Footnote 43 It ‘is generally used when the levy is for some special administrative expense which the name (health cess, education cess, road cess etc) indicates’.Footnote 44 The Andhra Pradesh High Court noted, as dicta, that ‘cess is a tax confined to local area for a particular purpose’,Footnote 45 although what precisely this means is unclear. At the very least, there is agreement that cesses are not for general purposes. The latter is the purview of general taxation and surcharges. Rather, cesses are for specific purposes. In India, and based on themes emerging from certain legislation, these ‘can be categorised as follows: (a) development of a particular industry, (b) … welfare and (c) general and broad-based causes’ such as road and infrastructure development.Footnote 46 And cess funds tend to be held in non-lapsable institutional vehicles, such that they accumulate year on year.
The nature of the taxation power underpinning the cess is ambiguous. It is neither self-evidently a matter of general taxation, nor a fee paid by an individual or entity for a service rendered by the state.Footnote 47 Vijayalashmi Rice Mill, a 2006 Supreme Court case, attempted to clarify the matter. It dealt with the central government’s Andhra Pradesh Rural Development Act. This Act levied a cess of 5% on the purchase of certain agricultural goods, in order ‘to provide financial assistance for the development of rural areas in the State [Andhra Pradesh] by creating infrastructure facilities’.Footnote 48
Does the special purpose of the cess, the Court asked, mean that it appears as ‘a tax [or] compulsory exaction of money by the State or a public authority for public purposes’ – albeit a special or targeted purpose – or is it ‘a payment for some specific services rendered’?Footnote 49 The answer, the Court held, was to observe the cess in its context. The Court built on its 2004 decision in Kesoram Industries:
Depending on the context and purpose of levy, cess may not be a tax; it may be a fee or fee as well. It is not necessary that the services rendered from out of the fee collected should be directly in proportion with the amount of fee collected. It is equally not necessary that the services rendered by the fee collected should remain confined to the persons from whom the fee has been collected. Availability of indirect benefit and a general nexus between the persons bearing the burden of levy of fee and the services rendered out of the fee collected is enough to uphold the validity of the fee charged.Footnote 50
In other words, whether the cess is a tax for the general good, or a fee for a specific service, is highly context dependent. The court then held that, for the cess to be treated as a fee, there need not be a ‘quid pro quo in the strict sense … All that is necessary is that there should be a reasonable relationship between the levy of fee and the services rendered.’Footnote 51 In this instance, the Court went on to treat the rural development cess as a fee, even though the proceeds were being applied towards general infrastructure.
The cess thus produces a fiscal domain that exists in between the collection of revenue for general purpose or interest on the one hand, and for specific purpose or interest, on the other. This space is related to the nature of the cess as a special purpose levy – a purpose that is by definition narrower than a general public purpose, but not wholly reducible to a transaction that requires a strict quid pro quo.
Both the special purpose and the unclear fiscal status of the cess make it complex to administer through the public finances. Both the Constitution and court judgments provide that ‘cess tax proceeds must be credited to the Consolidated Fund of India’ or of the relevant state, and earmarked ‘to ensure that the money does not get merged into the general tax revenues’.Footnote 52 Cess fees, by contrast, ‘must go into a special fund and not the Consolidated Fund of India. The accounts of such cess fees are maintained in the public accounts of the … government.’Footnote 53 This assumes that the status of the cess as tax or fee is already known; however, this is ordinarily not specified in the legislation, and even if it were, the courts have made clear that the status of the cess is to be interpreted by the context, function, and implementation of the cess.
Once initially ringfenced, the subsequent transfer of cess proceeds produces further complexity.
Cesses … need additional machinery for administration … Some cess laws have set up or make reference to existing machinery such as boards, committees, councils and commissioners for administering the cesses … [And] depending on the subject matter of the cess, the machinery [may be] governed by different … departments and/or ministries.Footnote 54
As a result of this complexity, ‘cess proceeds held in the [Consolidated Fund] often remain unutilised’.Footnote 55 For example, in its report on central government finances for FY 2018–19, the Comptroller and Auditor General of India (CAG) examined central government cess collection. Referring to cess collection by the central government – ie not including state-level cess collection – the CAG found that
out of the Rs 2,74,592 crore [INR 2.74 trillion] received… in 2018–19, only Rs 1,64,322 crore [INR 1.64 trillion] had been transferred to Reserve Funds/ Boards during the year and the rest was retained in the CFI [Consolidated Fund of India]. This included collections amounting to Rs 382 crore [INR 3.82 billion] on account of 17 cesses abolished … with effect from 1 July 2017 [ie these cesses were collected even after their legal authority had lapsed], which were retained in the CFI.Footnote 56
This does not even begin to itemise the underspending of cess funds once they have been remitted to the relevant central or state-level special purpose vehicle for their expenditure – which has not been systematically analysed.
The very same administrative complexity that leads to underspending also diminishes the effectiveness of efforts to pursue accountability for that underspending. Take the Building and Other Construction Workers’ (BOCW) Cess, a cess collected, held, and spent to provide for various lifetime welfare expenses of the 74 million labourers in the construction sector.Footnote 57 The funds are ringfenced and deposited in BOCW Boards, or special purpose vehicles to hold the funds. Each Indian state is responsible for the collection, management, and expenditure of the funds, pursuant to legislation and general guidelines at the central level. Construction workers have to pay a nominal registration fee to any one Board to be eligible for benefits – ideally the Board in the state where they have been working, although this is a challenge in the context of the high proportion of construction workers who are internal circular migrants.
Collected since 1996, by early 2020 only 25% of collected funds had been spent, with approximately INR 310 billion lying dormant.Footnote 58 This led to lawsuits before the Indian Supreme Court, which explains why I focus on this particular cess. Here, I contrast two BOCW cases, to demonstrate how efforts to pursue accountability for underspending, and trigger greater spending, suffer from a lack of clarity over the public dimensions of the BOCW’s institutional arrangements, and the nature of construction workers as a general or special interest group more broadly. In doing so, the cases reveal ambiguities around the legal ordering of society through the fisc – and with it, forms of state accountability that individuals and groups can pursue.
First, take a series of Supreme Court cases brought by the National Campaign Committee for Central Legislation on Construction Labour (NCC-CL), a civil society group, between 2006 and 2018. The NCC-CL sought a legal compulsion to spend BOCW funds. It did so through a range of public and administrative law arguments: in brief, the levels of underspending were so egregious as to defeat the purpose of the legislation – which in turn, it was argued, was to realise constitutional provisions guaranteeing the dignity and fundamental rights of construction labourers. As noted above, the Supreme Court issued a series of directions to the central and state governments from 2006 onwards to fully constitute the Boards (and surrounding administrative apparatus for expenditure) as required by the enabling Act.
Importantly for my argument, in proceedings before the Supreme Court, the central government claimed that the institutional structure of the BOCW was too decentralised, fragmented, multi-level, and multi-actor for it to ensure spending. It was a struggle to ‘ensure that all the concerned authorities sit together for the benefit of these building and construction workers, on whose account thousands of crores of rupees are being collected under the Cess Act’.Footnote 59 The central government claimed that it had tried to get each state to constitute boards and spend monies, issuing nine directions between 2010 and 2016 ‘to the State Governments … with regard to implementation of the BOCW Act and the Cess Act’Footnote 60 – all to no avail.Footnote 61
More specifically, the states failed to constitute the relevant bodies to hold the cess monies and put them to use. And even if the bodies were to be constituted, it remained unclear on what basis would they make decisions, since, as the Court put it, ‘even the CAG does not have accurate figures from the State Governments’ on collection and expenditure, and ‘whatever figures are available, may not be reliable’.Footnote 62 Thus, ‘the affidavits filed [by the central and state governments] did not take the matter of utilization of funds any further, in the sense that the State Governments … had no clue on how to spend the cess that had been collected’Footnote 63 owing to lack of information on monies.
In the foregoing, the central government – and, in their turn, the states – argued that they could not be held responsible or accountable for non-spending owing to decentralisation, the fragmentation of administrative authority, and their inability to subject the BOCW Boards to administrative oversight. The Court did not reject this argument. Instead, it continued to issue repeat directions for the proper implementation of the Acts, while, as a matter of deference due to the political functioning of centre-state relationships, leaving it to the central government ‘to discuss and decide on the modalities and methodologies for ensuring that directions issued under laws enacted by Parliament are given due respect by the State[s] … and directions issued thereunder for the implementation of the laws in letter and spirit are acted upon with due dispatch and promptitude’.Footnote 64
In 2020, a different Supreme Court case, Kesar Lal, saw the plaintiff use consumer protection legislation to sue for disbursal of BOCW funds. He claimed that he was unfairly denied INR 51000 to support his daughter’s wedding costs, which was an eligible expense in the state of Rajasthan at the time. He had properly registered and paid dues with the Rajasthan BOCW Board. He was denied on the grounds of ‘technical defects’ in his request (including an incomplete form). The plaintiff argued that the BOCW registration fee was a contribution for a service, giving rise to a consumer contract under which he was entitled to the monies. Here, he was a consumer, with a concomitant stake in the Board’s functioning. The Board, he argued, should be compelled to expend its money on that basis, irrespective of the technical defects in his request.
The Supreme Court responded by explicitly stating that it was trying to find some way of holding BOCW Boards accountable for their non-expenditure, in light of the problems faced in the NCC-CL cases of enforcing expenditure under public law principles.Footnote 65 Having prefaced its findings with an extensive citation from NCC-CL about the difficulties in compelling state governments to spend BOCW monies, the court continued:
The … vulnerabilities [of building workers] have been attempted to be safeguarded by a law which unfortunately has not been implemented either in letter, or in spirit. Yet, we have in the present case, the spectacle of a statutory welfare board seeking to exempt itself from being held accountable to the remedies provided under the Consumer Protection Act 1986. The submission which has been urged before the Court, simply put, boils down to this: the beneficiaries of the service pay such a meagre amount as contributions that they cannot be regarded as ‘consumers’ within the meaning of Section 2(d) of the Consumer Protection Act 1986.Footnote 66
In Kesar Lal, the government (here, the central government and state of Rajasthan pleaded together) pursued the argument that the BOCW was an ordinary welfare scheme: ‘In the present case, though a service is rendered by the Board, the expenditure on the welfare scheme is defrayed from the cess which is collected and hence, is not a “service”’. In other words, the BOCW should be understood as the ‘State commit[ting] itself to [a] welfare scheme’ that is tax-funded.Footnote 67 In this argument, Boards cannot be understood as service providers. They are instead part of state administration and its sovereign welfare function.
Thus, in NCC-CL, when asked to safeguard and realise the general entitlements of vulnerable workers by spending BOCW funds, the central and state governments argued that they could not do so and sought to pass the buck. They pleaded inability due to capacity and structural issues. This included the challenges of setting up BOCW Boards. In Kesar Lal, the plaintiff pursued a different tack: the BOCW has a contributory requirement, and BOCW Boards should be sued accordingly. The plaintiff framed his entitlement as a consumer one and sought relief. In response, the government did not plead lack of capacity (a plea irrelevant to a consumer contract claim). Instead, it asserted that the Boards are straightforwardly public bodies. This was eventually unsuccessful, the Court finding for the plaintiff based on a ‘need to ensure a sense of public accountability by allowing consumers a redressal in the context of the discharge of non-sovereign functions which are not rendered free of charge’.Footnote 68 Yet, as a form of accountability, this simply provided a remedy of specific performance in favour of a specific individual with a special or particular interest.
Taking the two cases together, in NCC-CL, according to the government, the Board is institutionally complex, and on this basis central and state governments claim that they cannot be held accountable for non-spending under public law. And in Kesar Lal, according to the government, the Board has a simple and clear public function, which the government embraces … and on the basis of which it cannot be held accountable for non-spending on construction workers understood as a group of individuals.
To summarise, in the Indian context, cesses are special purpose revenue institutions. As a result of this, they can be either or both a tax and a fee – and thus blur the divide between the two. So framed, this view takes a constituted perspective on the cess. Cesses, their ambiguous status as tax or fee, and their underspending, might reveal deficiencies in the constitution of India’s fiscal state (if assessed against an ideal type). Alternatively, they might describe its particular post-colonial condition: their seeming ambiguity is an enactment of the broader specificity and complexity of combining the ‘general’ and ‘special’ interest into a particular vision of Indian (or perhaps post-colonial) ‘society’, and the state’s concomitant relationships and obligations to it.Footnote 69 Either way, cess underspending is a constitutional effect, not a cause.
In the next sections, I redescribe cess underspending through a historical institutionalist account of the evolution of the cess during under British company-colonial and Imperial rule in India. In doing so, I point to the ways that cesses, their ambiguous status, and their underspending, also constitute forms of rule – that translate into post-independence state forms, as Section 4 then shows.
3. The cess as constitutive: an historical perspective
The story of the cess in India begins in 1793, when the East India Company instituted the Permanent Settlement. Replacing previous tax farming arrangements, the Settlement was a revenue management arrangement that installed a permanent Indian landed class, zamindars, in large parts of India. In return for security of their particular tenure – the alienable right to manage the land and its productivity as they saw fit – zamindars would pay a fixed rent to the Company in perpetuity.
The Settlement soon proved to be insufficient, both in terms of its administrability and its revenue-raising capacity.Footnote 70 It was in this context that Biswas recounts the initial use of a ‘cess’ as a label for a specific type of revenue-raising instrument implemented during Company, and then Imperial, rule. It began with Archibald Campbell, the Acting Collector of Shahabad, who noted a series of ‘local’ levies imposed by zamindars for the upkeep of roads in his district – particularly those roads that facilitated commerce between settlements (as opposed to those that facilitated the circulation of goods within Company circuits of commerce).Footnote 71 ‘After 1813, the collection started to be mentioned as “cess” in the Revenue records’, and other Company administrators sought to scale it up for road construction and upkeep across Bengal.Footnote 72
The levy of this ‘cess’ had to be justified through its relationship to actual or potential future use for a specified local purpose (in this case, a network of roads)Footnote 73 – even as such prospects might be disclaimed by other local bureaucrats as unimplementable. Sinha recounts a member of the Bengal Presidency council dismissing ‘the silly young nobleman [Lord Elphinstone]’, since he ‘actually talk[ed] of making roads’ with cesses,Footnote 74 covering a distance ‘equal to that from London to Vienna’.Footnote 75 Here we see an initial indication that the cess might be targeted at specified local purposes, with a high tolerance for speculation and improvisation – albeit not disconnected from that initial purpose.
The cess was not limited to the Bengal Presidency.Footnote 76 In the Madras Presidency, the Collector of South Arcot (and future Governor of Madras) Edward Maltby noted that in his 6,000 square mile district, there was no road for the specific purposes of the district itself. He requested that the Presidency Board of Revenue allow him to retain a fraction of the land tax in a special district land fund, for the specific purpose of building and improving such roads. In 1855, the Board approved the ‘land cess’. The ‘land cess’ was a levy not to be merged in the general revenues, but instead earmarked under a special head of ‘Local Funds’ and disbursed in an ad hoc fashion, on approval by the District Collector and Engineer (the latter reflecting the infrastructural nature of the envisaged projects). Over time, the funds came to be managed by ‘Local Boards’, whose committees – a mix of locals and British administrators – would request the works, and whose activities and fund management were subject to annual audit from 1882.
The special purpose of the various funds was rather elastic: it emerged in response to requests from specific Boards, but also as part of broader top-down projects of change imagined by the Imperial government. For example, in the 1880s, the funds were mobilised for the provision of local primary education. Funds were also spent on local health and sanitation, disaster relief, and administrative expenses among others. Indeed, in recognition of this, an 1884 Act reduced the minimum proportion of Local Board funds dedicated to road construction from 66% to 50%.
The Madras Presidency thus institutionalised the cess as a highly discretionary revenue instrument, operating for ‘local’ purposes, organised through a ‘Board’ whose form was administrative and subject to audit-based accountability, but with no political and legal accountability as we might otherwise recognise it. The cess functioned to tackle ad hoc local issues; it also reflected a strategy of decentralising expenditure on various policy issues in pursuit of savings for the Imperial exchequer: ‘the maintenance of highways, which was always an imperial charge, is now almost completely thrown upon local funds, and the total saving to the State by the creation of local funds and the transfer thereto of the services formerly rendered at the cost of the imperial exchequer was calculated in 1876 to be 21.5 lakhs of rupees per annum’.Footnote 77
After concerns in Westminster that ‘arbitrary cesses’, as a mark of post-Permanent Settlement fiscal reforms, had begun to disturb the legal certainty of the Settlement,Footnote 78 this form of cess was regularised in the 1880 Cess Act.Footnote 79 Although regularised, the precise nature of the cess proved elusive, as a 1935 parliamentary debate on what would become the Government of India Act demonstrates. Referring to a provision restricting tariff and non-tariff barriers in the internal Indian market (essentially a precursor to the fiscal federalism provisions of the eventual Constitution), one member asked the Attorney-General ‘exactly what is the difference between a cess … a tax which may be imposed by a Provincial Government, and the terminal tax [imposed by the centre]’, when imposed for the benefit of a locality? ‘One of the difficulties we have had in these discussions’, he continued, ‘is that none of us knows what exactly the word “cess” means’.Footnote 80 Was it a tax or not? Was it an instrument to do with trade, or land? Was it levied by a locality, a province, the centre, or otherwise? And for whose or which benefit?Footnote 81 The response was brief: ‘It does not matter… what [the provision] says about terminal taxes and assessing’ as long as the broader principle of internal free trade was enshrined in the Act.Footnote 82
The cess continued as a revenue instrument in post-independence India. As noted, the Constitution does not define cesses and the sole reference to them is in section 277. Nevertheless, they proliferated: Kotha tracks 35 post-independence central government cesses alone that support the development of industries (such as salt extraction or agricultural production), provide welfare (from cinema to beedi workers), and pursue other specified goods (such as tackling water pollution or building infrastructure).Footnote 83
The cess also appeared in other colonial-era British territories, from Ceylon,Footnote 84 to Uganda,Footnote 85 to Sarawak and Tanganikya.Footnote 86 A case of a cess in colonial Kenya provides some insight into the fact and dynamics of underspending which emerge from the historical context of cesses described in colonial India. Here, District Councils levied cesses on various agricultural products, which funded ‘Agricultural Betterment Funds’ that they managed. By the end of 1951, they had accumulated an aggregate unspent balance of over £560,000 – a figure which led to a parliamentary debate in Westminster. The government justified the figure as follows: ‘the balance in these funds is large because, as revenues from cesses are precarious, it is necessary to maintain a relatively large balance to ensure continuity of agricultural development projects and services’.Footnote 87
Here, we see both institutionalised underspending and a political logic to it: the accumulation of a pool of funds to tackle ‘local’ issues as they arise over time, whether in sectors that are of specific interest to the colonial state (as is the case in the Kenyan funds), or in response to political and fiscal pressures on colonial rule. The two also overlap, as in the case of the Indian road cess. These funds are more than merely slush funds – they must have some purpose, and in the instances mentioned here, they should also be audited (albeit this was also observed in the breach). But within that context, they are spent with high degrees of discretion, in a responsive fashion, by special-purpose administrative bodies such as Boards.
The cess, then, was not just an instrument through which the British state consolidated the separation between the public and private spheres, as Schumpeter and others wrote of the effects of the emergence of modern taxation.Footnote 88 It was a means of carving out revenue, and attendant policies and institutions, for the administrative management of the ‘local’ in colonial contexts. In other words, the cess drew a divide not between public and private, but between imperial and non-imperial revenue – and relatedly, between imperial and local concerns.
This feature of the cess, I argue, was a means of managing contextual and ad hoc pressures on colonial control. The cess was a flexible instrument, allowing for high levels of administrative discretion regarding its objectives, collection, and expenditure. Its boundaries were found in its orientation towards the ‘local’. Wherever the ‘local’ was, whatever its issues were, and however they were configured, it had two features: first, the ‘local’ was always defined in contradistinction to the ‘Imperial’ or ‘colonial’; second, the ‘local’ referred to the necessary special purpose legitimating cess collection. This purpose was the basis for the high discretion the administrator (ordinarily the District Collector) enjoyed, and was also the only basis against which collection and expenditure was accountable (in both the political and accounting senses). The cess and its underspending, then, can be understood as a constitutional feature of colonial Indian rule, institutionalised around constant administrative adjustment to evolving pressures, that is minimally justified, and that leaves regular central revenue untouched.
In the next section, I turn to patterns of cess underspending in contemporary India, and the remobilisation of these underspent funds to tackle COVID-19 and the effects of lockdown. Returning to the BOCW cess, I show that the money that pools in the BOCW Boards can be remobilised for purposes other than the special ones for which it was collected – such as general budgetary expenditure – when pressures on rule are high. This demonstrates a continuity in the forms, logic, and institutions of rule from colonial times, embedded in cess structures, even as ‘colonial’ rule has transformed into the ‘post-colonial’ state.
4. Unspent cess funds in contemporary India: the BOCW cess and COVID-19
The scale of underspending of BOCW funds is striking. By September 2018, only 39% of funds collected since 1996 had been disbursed. In the same year, and in response to similar figures contained in a civil society lawsuit, the Indian Supreme Court ordered the central and state governments to increase disbursal rates.Footnote 89 Yet, as already noted, by early 2020 only 25% of collected funds had been spent – a decrease in expended proportion.Footnote 90 These dormant funds – approximately INR 310 billion – would subsequently form almost a sixth of the total value of the Indian Government’s Spring 2020 emergency welfare package following the COVID-19 lockdowns, aimed at the most vulnerable citizens whose lives and livelihoods were affected.Footnote 91
In some instances, underspending is compounded year on year, as yearly Board cess collection outstrips expenditure.Footnote 92 This increasing amount of unspent money can further accumulate interest. In October 2021, for example, the Karnataka Board invited quotations for one-year fixed deposits for INR 33.5 billion in unutilised funds.Footnote 93 Yet in some instances, the unspent money is itself underutilised as basic capital. Many Boards have failed to register as legal bodies ‘established for charitable purposes’ for the purpose of income tax exemptions.Footnote 94 In Karnataka, the CAG noted that a failure to register meant that the Board had to pay to the central government an ‘income tax expense of 351.12 crore [INR 3.5 billion] during 2018–19’ alone – thereby remitting to the general central budget some of the funds collected for construction workers in the state.Footnote 95 Funds are also placed in accounts which provide lower-than-market interest rates – which in Karnataka amounted to a loss of INR 20.5 million on investments of INR 4.3 billion.Footnote 96 And some funds are altogether moribund, being neither expended nor earning any interest. In Karnataka again, the CAG noted that the Board had parked INR 1 billion in a bank account ‘for the period from 22.03.2017 to 22.02.2018 for a duration of 11 months … [yet] bank documents showed that no interest was credited’.Footnote 97
The onset of the COVID-19 pandemic appeared to change the political dynamics around these unspent funds: the central government claimed that it would mobilise these dormant funds to support construction workers, as a vulnerable group whose livelihoods were particularly affected by lockdowns.Footnote 98 Yet, as outlined in a response by the central government’s Ministry of Labour and Employment to a parliamentary question in 2021:
In order to mitigate the financial crisis of the construction workers and protect them against the economic disruptions due to the outbreak of COVID-19 pandemic and Country-wide lockdown the Central Government had taken numerous measures to provide them financial assistance, food packages and other benefits … under Section 60 of the Building and Other Construction Workers Act, 1996 [ie a residual and emergency powers provision] … In response, during the first wave of COVID-19 the State Welfare Boards have cumulatively disbursed more than Rs 5618/- Crores [ie INR 56.18 billion], through DBT [direct bank transfers] … [and] during second wave of COVID-19 an amount of Rs 1704.302 crores [ie INR 17.04 billion] have been disbursed.Footnote 99
Thus, over a year after the initial announcement, only approximately INR 73 billion, or approximately a quarter of the unspent balance of the BOCW fund, had actually been spent.
Moreover, the Punjab BOCW Board interpreted the central government direction as giving them latitude to purchase COVID-19 vaccines (at INR 315 per dose) and administer them – to registered BOCW beneficiaries, as well as to their family members, who were not registered Board members.Footnote 100 This was at a cost of around INR 70 million.Footnote 101 The central government initiated a free vaccination scheme for all adults shortly thereafter, and demanded that the Board repay this expended amount back into the state BOCW fund – thereby increasing the unspent amount again. By December 2022, the unspent balance of BOCW monies was INR 382 billion – a 23% increase on the amount unspent prior to the pandemic, and which amounted to 43% of the total cess collected since 1996.Footnote 102 This was particularly striking given the dire situation that workers found themselves in during the months of the pandemic – with one survey indicating that 78% of construction workers saw their income drop to zero.Footnote 103
I draw two conclusions from this pattern of spending, which subsequently translated into continuing underspending. First, as noted, there were debates over the legality and legitimacy of remobilising special purpose funds for general welfare. But underspending might not only rehearse the ambiguities around the status of construction workers as a group of public concern, and the cess as a tax or fee. It could also be understood in the context of construction workers’ welfare as a contingent and ‘local’ pressure on, or an issue of concern to, central rule – ‘local’ referring to an administratively defined, non-centralised, special purpose.
In turn, under-expenditure is not just the marker of a flawed state that cannot execute a welfare programme; it also marks a state that administers potential sources of discontent in an ad hoc manner. Unspent funds were here remobilised without the need for, and scrutiny of, legislative appropriations. And while this particular instance was routed through the emergency provisions of the relevant Act, it was one of various approaches to remobilise unspent balances of different cess funds during COVID-19 without scrutiny, such as states borrowing those balances at below-market rates to fund general welfare expenditure.Footnote 104 Underspending and remobilisation are thus distinct from conventional mechanisms of targeted budgeting and planning. In the state-society relationship forged in the BOCW example, construction workers are not an object of welfarist public concern. Rather, they are a group that the state has structured such that their welfare demands should not and will not pressure central revenue and its attendant politics to the extent possible.
Secondly, why, then, do the relevant local entities not spend the BOCW funds and meet construction workers’ basic needs? Recall that there was no strong impetus to expend cesses in the colonial context. The purpose of the cess needed to be sufficiently plausible, collection roughly proportionate to it, and there needed to be some vision for spending. At the same time, that vision could be vague, and cesses could accumulate in anticipation of higher future local pressures on colonial rule (as in the Kenyan Agricultural Betterment funds). This limited impetus to spend persists in the BOCW case: there is no clear timetable nor programme for implementation and expenditure, and accountability for underspending is weak. Moreover, the absence of legislative scrutiny implied here provides the administration with latitude to respond to pressures.
This continues the colonial institutions and logics of rule sketched out in the previous section, marked by high discretion, low accountability, and the operative distinction between ‘colonial’ and ‘local’. However, we see a distinction between the ‘centre’ and ‘local’ take the place of the latter, where the local is the domain of discretionary administrative rule, and the centre is the inverse. This tracks Mamdani’s account of a ‘bifurcated’ state, whose structures continued from colonial to post-independence contexts.Footnote 105 For Mamdani, the structures of legal pluralism in general constitute that state. Here, I suggest that tax law and administration play an importantrole in constituting it – and they do so not through general institutional structures but through the specific institutions of the central revenue, their relationship to the institutional frameworks that allow for the accumulation of unspent funds, the continuing redescription of the ‘local’ or special purpose, and the resultant remobilisation of unspent funds.
Unspent cesses thus offer a vantage point on the post-colonial state that is distinct from more familiar accounts of bifurcation – such as the rule of law and citizenship rights for some, and administrative or emergency rule for others.Footnote 106 They draw our attention to the fisc, its institutionalisation (for example, its accountability structures), and its subjunctive concepts (for example, fiscal federalism, as distinct from concepts of political federalism that relate to citizenship rights). In doing so, they also foreground distinctive mechanisms whose institutionalisation constitutes the state: in this case, cess accumulation and remobilisation, rather than, say, security and emergency.
The differences between these mechanisms are analytically useful in understanding that state. For example, from the perspective of security and emergency, even if emergency is routinised, its temporality is temporary, in the sense that it is in the service of the restoration of order.Footnote 107 The temporality of cess underspending, by contrast, is indeterminate, given that the funds are both ringfenced and non-lapsable – enabling continual adjustments to pressures on rule, rather than immediate responses to ruptures in order. Or, consider the source of political authority for the discretionary and flexible elements of rule in a bifurcated state. Whereas for Mamdani the invented notion of customary tradition – existing from time immemorial, and anchored in the notion that it mapped onto ethnic difference – expressly underwrote the authority of the chief,Footnote 108 the remobilisation of cess funds is authorised by the expression of a minimally plausible special purpose or policy goal itself, to be reiterated over an indeterminate time.
In other words, cess underspending and subsequent remobilisation take their place in a bifurcated state by constituting a forward-looking, policy-oriented domain of discretion, authorised by its functional ability to respond to pressures on rule – quite distinct from, say, discretion premised on backwards-looking appeals to historic custom, authorising the production and management of ethnic difference à la Mamdani. This, in turn, produces a distinctive set of constitutional inquiries. For example: what is the political status of the person or process who articulates the evolving policy domain? On what grounds and by whom can accountability be pursued for the exercise of discretion through underspending and remobilisation? And through what legal mechanisms are funds such as cesses connected to the political and legal accountability structures of formal revenue, budgeting, and expenditure?
5. Underspending and remobilisation beyond India
This argument, I suggest, has broader relevance. First, it travels beyond India. Cesses, with attendant special-purpose bodies, persist in many former British colonies, from post-independence Pakistan’s cess on spindles and looms, to ‘produce cesses’ in Tanzania’s Local Government Finance Act 1982 (also present in Zambia and Uganda). And as recently as 2018, the Kenyan Meru County Tea Cess Act provided for a cess on tea sales, to be held in a fund and managed by Tea Cess Committees, for ‘the infrastructure of the tea growing areas’ as well as ‘any other lawful purpose as would be decided by the Committee … provided that [it] would be for the benefit of the tea farmer in general’.Footnote 109
More broadly, underspent non-lapsable welfare funds persist in countries across the Global North and South, all with patterns of administrative discretion to remobilise funds to handle pressures on rule.Footnote 110 This offers an opportunity to reinterpret these funds in light of my focus on accumulation and remobilisation, and their consequences for the constitutional order. Take, for example, recent European Union (EU) efforts to roll over and remobilise its funds for COVID-19 response. The European Central Bank (ECB) has noted how Member States have long struggled to absorb and spend their allocation of a range of EU funds.Footnote 111 Cohesion policy funds, for example, amounted to a little under €400 billion from 2021–2027.Footnote 112 They are held in ringfenced budget lines, and are lapsable (usually by the end of a seven-year programming cycle). However, states’ inability to spend them means that they have frequently been the subject of ad hoc grace periods for further spending.Footnote 113
At the immediate onset of the pandemic, the European Commission decided to allow Member States to flexibly remobilise their unspent balance of cohesion policy funds to tackle the effects of the pandemic.Footnote 114 Subsequently, the EU raised €750 billion on the capital markets to further support pandemic recovery. Of these, approximately €672.6 billion was ringfenced in an instrument called the ‘Recovery and Resilience Facility’ (RRF) for a series of grants and loans to Member States. The funds lapse at the end of 2026 (although MEPs voted for an 18-month extension in June 2025, the European Commission has so far resisted any change to the deadline).Footnote 115
The legal authority for the RRF was grounded in the principle of ‘promot[ing] the Union’s economic, social and territorial cohesion’ following the shock of the pandemic.Footnote 116 The initial use of ‘cohesion’ as a justification for raising COVID-19 relief funds was expedient, since it would facilitate fundraising under the relevant provisions of the EU Treaties without requiring a treaty amendment.Footnote 117 However, in light of the significant underspending of these funds – only around one-seventh disbursed halfway through the life of the RRFFootnote 118 – the EU has amended or interpreted the regulations underpinning the funds to allow for their use in everything from transforming energy systems across the EU, to bolstering defence manufacturing.Footnote 119
Here, we see an emerging practice of continually evolving the initial policy purpose of ringfenced funds to deploy unspent balances to changing pressures on rule. At the moment, these funds are lapsable. However, the policy debate over the design and use of existing and new ringfenced EU funds (eg for defence) points towards future ‘flexibility’ in both purpose and time horizon.Footnote 120 As Polverari notes,
[i]t should not be forgotten that the RRF was created as an emergency tool to respond to a major, unforeseen crisis. Such a type of emergency tool may no longer be needed if, in reforming the ordinary policies of the EU (and cohesion policy among them), more attention was paid to making them more forward-looking and more flexible to respond to unforeseeable events.Footnote 121
Formulating this argument in my terms, we see not just a one-off fiscal response to an unprecedented crisis, but the evolution of an ongoing practice and institutionalisation of fund accumulation and remobilisation. This, in turn, promises an administrative and discretionary fiscal regime that exists in some relationship to formal budgetary processes at the EU and Member State levels. And the relationship between these two is of significant constitutional import: for example in placing the locus of negotiation between Member States and the EU over the nature of integration at the juncture between these two processes – a locus which, I note, is quite different from a view of the EU constitution that places that locus in public debate over political ‘ideas of demos, civitas, and ius’.Footnote 122
This observation leads to a final suggestive reflection. While what I have described bears at least a family resemblance to the colonial and post-colonial logics and forms of the Indian state canvassed above, it may not be happenstance to draw on state theory developed from the colonial context to understand state bifurcation in the Global North. At a generic level, this might reflect some version of the argument that the Global South is the ‘vanguard’ of Global North – ie where conditions of late capitalism mean that the Global North is increasingly unable to resist the spread of patterns of exploitation and extraction that so structure states in the Global South.Footnote 123 In other words, the Global South got there first, and has adapted in ways that the Global North may also have to, or draw inspiration from.
But the relevance of building insights from the Global South may be more direct. Some scholars are tracing how European states, organised around imperial statehood, had to reimagine their self-understanding of statehood during the era of decolonisation – and how the EU was a response to that.Footnote 124 It represented a pooling of sovereignty in order to manage the loss of political, social, and economic order that colonialism afforded. And it may be that, as Europe has to respond to intensifying pressures on rule, it can no longer solely rely on neo-colonial circuits of extraction, cheap labour, and the externalisation of social harms. And thus old colonial techniques of pooling and remobilisation may be increasingly relevant, allowing for ordinary fiscal order to prevail while managing and defusing potential disruptions to it.
While the foregoing is suggestive, it is offered to demonstrate three things. First, it shows how the notion of a bifurcated fisc – grounded in part in the hypothecation, underspending, pooling, and remobilisation of funds – might emerge from the Indian context, but can be formulated as a concept that can travel across space and time. Secondly, this offers a reinterpretation of certain underspending, away from state failures and towards novel state forms. Thirdly, it demonstrates how the concept may productively produce ‘friction’ as it moves from India. That is, the concept may provide explanatory purchase on empirical phenomena, but it also and importantly acts as a vehicle for critical reformulation of the broader context to which it moves.Footnote 125
Conclusion
In this paper, I have argued that underspending is not solely something to be explained in terms of its flaws, but also to be engaged with as constitutive of the state. I have exemplified this conceptual argument through a substantive intervention (on the nature of the cess in India, and its role in constituting a particular type of bifurcated state), a conceptual one (on developing state concepts by reinterpreting seeming flaws like underspending), and a methodological one (tying together historical institutionalism and analysis of court decisions in service of the argument). Taken together, I suggest that underspending – particularly the pooling and remobilisation of funds – might be a vantage-point from which to understand some of the constitutional features of India, other post-colonial states, and perhaps beyond.
Open access
