Kautilya as Constitutional Theorist

The Arthashastra is conventionally filed in the corner of the syllabus reserved for political theory. The reader meets it after Plato and before Hobbes, learns that Kautilya was a clever and ruthless adviser to a fourth-century-BCE emperor, and is told that the text is a manual of statecraft. The reader then closes the book and proceeds to the more serious business of studying constitutional law and administrative law, both of which arrived in India, as we are told, with the East India Company and the colonial courts. This sequence is widely taught, mostly unchallenged, and substantially wrong. The Arthashastra is not principally a manual of how to be a ruler. It is a sustained working-out of how a centralised state should constitute, audit, and discipline its own officers. On that reading, much of what modern Indian administrative law has had to reinvent under twentieth-century constitutional pressure was already there, sketched out in granular institutional detail, in a text that has been waiting two and a half millennia for the right kind of reader.¹

The mis-filing problem

The Arthashastra reaches the modern Indian student as one of three things: a Sanskrit antiquity, a Machiavellian curiosity, or a piece of cultural nationalism. Each of these framings is a partial misreading, and the cumulative effect is to keep the text out of the conversation in which it would do most work — administrative law.

The Sanskrit-antiquity framing treats the Arthashastra as a primary text in the history of ideas, to be read for what it reveals about classical Indian thought. This is true as far as it goes; the difficulty is that it stops there. The reader is not encouraged to ask what would happen if the institutional design of the Arthashastra were applied to a problem the modern Indian administrative state actually faces.

The Machiavellian framing treats Kautilya as a clever ancestor of Realpolitik. The pairing is forced by chronology (Kautilya predates Machiavelli by nearly two thousand years), by intention (Machiavelli writes about the conduct of a single ruler, Kautilya writes about the structure of a permanent administrative state), and by content (Machiavelli has almost nothing to say about audit, record-keeping, or the discipline of public officers; the Arthashastra has chapters on each). The pairing has stuck because it is convenient for syllabus-writers, not because it is accurate.

The cultural-nationalist framing — popular in some quarters since the 1990s — treats the Arthashastra as a trophy of indigenous achievement, mostly without engaging with its institutional content. The result is, paradoxically, the same as the Sanskrit-antiquity framing: the text becomes a possession rather than a working resource. It is admired and not used.

What is needed is the fourth framing: read the Arthashastra as a treatise on the structure of administrative authority and on the constitutional limits that such authority must observe. The text richly rewards that reading. The reading is not a stretch; it is closer to what the text is actually doing.

The Saptanga as a constitutional analytic

The first place where the Arthashastra reveals its administrative-law character is the Saptanga doctrine — the analysis of the state into seven elements: Swami (the king), Amatya (the ministers), Janapada (the people and the territory they occupy), Durga (the fortified capital), Kosha (the treasury), Danda (the coercive arm), and Mitra (the allies). The list has been taught for generations as a list of the components a kingdom needs in order to be a kingdom. It is more usefully read as a typology of the offices and resources that constitute the state as a juristic person.²

Read that way, Saptanga begins to look like an early constitutional analytic. The state is not a personal possession of the ruler; it is a composite of seven institutional elements, each with its own internal structure and discipline. Swami is one element among seven, not the source of all the others. Amatya is a separate organ, with its own rules of selection, advisory function, and accountability. Kosha is a treasury with rules of receipt, expenditure, and audit. Danda is the coercive arm, regulated by Dandaniti — the science of law enforcement — which has its own internal limits. The Saptanga is not a list of resources at the king's disposal; it is a list of the elements whose proper functioning is the king's duty.

The contrast with the conventional picture of pre-modern monarchy is sharp. The conventional picture has the king as sovereign, with the apparatus of the state radiating outwards as an extension of his will. The Arthashastra picture is closer to the modern administrative state: a system of offices each with their own brief, the king's role being to staff them well, supervise them, and discipline them when they fail. The king is not the source of the system; he is the supervisor of the system, and the system has the right to expect that supervision to be diligent.

The hierarchy of sources of law

The Arthashastra's account of the sources of law is the second piece of evidence for the administrative-law reading. Kautilya lists four sources — Dharma (sacred law), Vyavahara (the facts established by evidence), Charitra (social convention and community practice), and Rajashasana (the royal edict).³ The four are arranged in a hierarchy, with a conflict-of-laws rule: where Charitra or Vyavahara conflicts with Dharma, Dharma prevails; where Dharma conflicts with the rational law (Dharmanyaya, often taken as the king's law), the rational law prevails, on the ground that the original justification of the Smriti rule is no longer accessible.

The hierarchy is a striking piece of jurisprudential architecture. It carries within it the recognition that legal sources can be plural, that they can conflict, and that a system needs a rule for resolving conflicts among them. The conflict-of-laws rule favours codified, rational, sovereign law over inherited religious authority where the latter cannot be defended on accessible grounds. The structural similarity to the modern Indian recognition that statutory law prevails over personal-law custom where the two are inconsistent is not a coincidence. Kautilya is doing the same kind of work, and doing it with more openness than the modern Indian Supreme Court has typically managed.

The Rajashasana element of the hierarchy is what makes the architecture administrative-law and not merely jurisprudential. The royal edict — the king's written order — is the operative source of law in the Arthashastra system, and Kautilya is explicit that it must be in writing. "The rule of kings depends primarily on written orders; even peace and war have their roots in them," he writes at 2.10.⁴ The writing-down rule is the foundational move of the Arthashastra's administrative-law system. The contemporary administrative-law principle that an administrative decision must be recorded in writing, and that a record-less decision is presumptively suspect, has its conceptual ancestor here.

The audit chapter

The third piece of evidence is Book 2, Chapter 7 — the chapter on the Office of Accounts. The Superintendent of Accounts is required to maintain a well-organised accounts office and to keep precise records of every department's activities: profits, losses, wages, taxes, prices, trade details both domestic and foreign. Monthly and annual audits are mandatory. Accounts are verified through inspection, precedent, arithmetic checks, and espionage. Officers who increase revenue are rewarded; officers who falsify, omit, or duplicate entries are fined, with the fine scaled to the severity of the offence and to the loss caused.⁵

This is not a generic exhortation to keep the books straight. It is a specification of the audit function as an institutional check on administrative officers — with timelines (monthly, annual, on the Vyushta day), methods (inspection, precedent, arithmetic verification, intelligence-gathering), and sanctions (graduated fines). The structural similarity to the C&AG audit mandate under Article 148 of the Indian Constitution is direct. Article 148 institutionalises a permanent audit office staffed by personnel insulated from the executive whose audit reports go to the legislature. The Arthashastra had the institutional idea, if not the constitutional bifurcation that allows audit to police executive without depending on the executive's goodwill.

The Arthashastra goes further. It identifies thirteen categories of "undesirable persons" — including "corrupt judges and magistrates, heads of villages or departments who extort money from the public, perjurers and procurers of perjury" — who are to be exposed by secret agents and either exiled or made to pay compensation proportionate to the offence.⁷ The category 'corrupt judges' is on the same list as the category 'counterfeiters and poisoners'. The point is not that the punishments were the same; the point is that the public officer is recognised as a distinct category of potential wrongdoer who must be policed by the state on its own account. The intellectual ancestor of the Prevention of Corruption Act regime is recognisable here, two and a half thousand years before the Act itself.

The discipline of the king

The most striking administrative-law feature of the Arthashastra is the discipline imposed on the king himself. The king is required, daily, to hear petitioners; he is forbidden from making himself inaccessible; he is told that delegation of his hearing function to immediate officers engenders confusion and public disaffection.¹⁰ The king is required to clean up the administration before he can demand good conduct from the people of the towns and countryside.⁶ The king is required to administer justice in the Dharmasthiya court daily and is told that a failure to do so is a grave sin. The constraints on the king are administrative-law constraints in everything but name.

Compare this to the Valmiki Ramayana's account of Rama in the Dharmasana, sitting on the seat of justice flanked by ministers, Dharmashastra scholars, and Sabhasads who function as a kind of jury.¹³ The administrative-law structure is the same: the highest office-holder is required to perform his function in public, in the presence of advisers, and according to rules that are not of his own making. Rama in the Ramayana is even described as Dharmapala (guardian of Dharma) and Dandadhara (executor of punishment); the descriptions are of an officer of the constitutional order, not of an autocrat.

The Arthashastra also makes the king's discretion legally consequential in a way that the literature has under-noticed. The king has discretion in sentencing, but the discretion is bounded: punishment must be impartial (regardless of whether the offender is the king's son or an enemy), proportionate (Sam Dand, neither too severe nor too lenient — Ati Dand and Nyoon Dand are both treated as defects), and certain (delays and arbitrary remissions are themselves treated as wrongs). The triple test for administrative discretion — impartial, proportionate, certain — that the Arthashastra applies to the highest office-holder in the state is the same test that modern Indian administrative law applies to executive decisions through judicial review.

It is worth taking the proportionality requirement seriously. The Arthashastra is not satisfied with the rule that punishment should fit the offence in some general sense; it works out the operational consequences. Excessive punishment (Ati Dand) is treated as a wrong in itself, capable of producing public disaffection and disorder; deficient punishment (Nyoon Dand) is treated as a different but equally serious wrong, capable of producing impunity and the gradual breakdown of the legal order. The intermediate position (Sam Dand) is the standard, and the standard is justiciable in the sense that the failure to meet it is itself a ground for criticism of the king. Modern Indian proportionality doctrine, developed in Om Kumar v Union of India (2001) and refined in Modern Dental College v State of Madhya Pradesh (2016), reaches a structurally similar position. The Indian Court did not invent the underlying intuition in 2001; it rediscovered it.

The procedural detail of the Arthashastra is the fifth piece of evidence. Book 3 lays out eighteen Tirthas (titles of law) — non-payment of debts, deposits and pledges, rescission of sale, ownership and division of property, breach of contract, resumption of gifts, non-performance of agreement, disputes between partners, boundary disputes, theft, robbery and violence, assault, defamation, adultery, duties of man and wife, inheritance, and gambling and betting.¹¹ The catalogue is recognisably the syllabus of a modern civil-and-criminal-law course. The categories are not arbitrary; they are the categories an administrative system would need if it were trying to regulate the most common transactions of an urbanising, commercial society.

The procedural rules are equally striking. A trial begins with the registration of the case — year, season, month, fortnight, date, place, nature of the deed, the amount in dispute, the country and residence and caste and gotra and name and occupation of both parties. The statements of the parties are taken in the order required by the case and then scrutinised.⁸ The judge — the dharmastha — has the power to initiate proceedings on his own motion when the affected persons are temples, Brahmins, ascetics, women, minors, the elderly, the sick, or orphans; the categories of suo motu jurisdiction cannot be ousted by lack of jurisdiction (deśacchala), by delay (kālacchala), or by adverse possession (bhogacchala).⁹

The structural similarity to the post-1979 Indian public-interest-litigation jurisprudence is direct. The Indian Supreme Court has read into Article 32 a jurisdiction to entertain petitions from persons not aggrieved on behalf of categories of disadvantaged petitioners — women, prisoners, children, the disabled, victims of inhuman conditions in custody. The categories overlap with Kautilya's list with notable precision. The procedural innovation is the same: the court is allowed to do justice for those who cannot themselves come to court. The Arthashastra had reached the institutional answer in the fourth century BCE; modern Indian administrative law has had to argue the same answer afresh, against the resistance of an English-derived doctrine of locus standi, in the late twentieth.

A second procedural feature deserves notice. The Arthashastra distinguishes the case-categories it calls vyavahara — transactions properly so called, with two parties and a mutual agreement — from the categories it calls sahasa (assault or violence), kalaha (quarrel), and parushya (abuse), which involve two parties but are not transactions in the contractual sense. The distinction is structural; it allocates different procedural treatment to different kinds of dispute, and it does so on a principled ground rather than by ad hoc enumeration. The instinct is the same as the modern Indian distinction between civil and criminal procedure, with vyavahara mapping broadly onto civil and sahasa-kalaha-parushya onto criminal. That the distinction was made on this conceptual ground in the fourth century BCE, and that the Arthashastra worked out the consequences of the distinction across its procedural chapters, is itself worth taking seriously.

The substantive contract law of the Arthashastra is the sixth piece of evidence. Contracts, oral or written, are legally binding. Their validity depends on the competence of the parties, their physical presence at the time of agreement, the absence of secrecy or fraud, and the suitability of the time and place. Contracts entered into in secret, by night, or by parties unfit to contract are void. Fraud and misrepresentation void a transaction; suit on a void transaction attracts not only loss of suit but a heavy additional fine for relying on an inadmissible transaction; even witnesses to such a transaction can be fined.¹²

The doctrinal architecture is identifiable. Capacity, free consent, lawful object, lawful consideration — the four planks of the Indian Contract Act 1872, sections 10, 11, 13-22, 23 and 25 — are all present, in substance, in Book 3 of the Arthashastra. The Arthashastra adds a procedural penalty for litigation founded on a void transaction that modern Indian contract law does not have. It also has a developed law of suretyship, a rule of restitution on rescission, a recognisable concept of fiduciary duty, and an explicit force-majeure rule for cases where performance becomes impossible owing to natural calamity or factors beyond the party's control. The treatment is not less developed than the early common-law treatment of these doctrines; it is, in important respects, more systematic.

What the Arthashastra was not, and what it was

The Arthashastra was not a constitution, in the modern sense of a single written instrument that establishes the legal order and limits the highest power within it. The text does not purport to be that. It also was not a code of administrative law in the modern bureaucratic sense; the administrative apparatus it describes lacks some features — career civil services with security of tenure, a clear separation of judicial and executive functions, a written constitution that creates the highest offices — that modern administrative-law systems typically assume.

But the Arthashastra was a systematic working-out of administrative-law problems before the vocabulary of administrative law was available to express them. It identified the public officer as a category to be policed by the state on its own account. It distinguished personal will from official function. It required written orders. It required record-keeping. It required audit. It required the suo motu protection of those who could not come to court themselves. It required the highest office-holder to perform his function in public, in the presence of advisers, daily, without delegation. It required proportionate, certain, impartial discretion in punishment.

These are not the concerns of pre-modern political philosophy; they are the concerns of administrative law. To file the Arthashastra under the former is to leave the second category empty of its earliest sophisticated Indian source. The Indian administrative-law tradition that emerged in the late twentieth century — through R.D. Shetty, through Maneka Gandhi, through S.R. Bommai, through Vishaka, through the public-interest-litigation jurisprudence under Article 32 — has been built largely without acknowledging that the institutional materials were already there in a Sanskrit text written in the fourth century BCE.

The mis-filing has persisted for three reasons. The first is the colonial inheritance of legal scholarship. The English-language jurisprudence syllabus inherited from the late nineteenth and early twentieth centuries was organised around the European canon and the early-modern common law; non-European pre-modern legal texts were placed in a separate, ornamental compartment labelled "comparative" or "historical" jurisprudence, and the placement was not revisited after independence. The Arthashastra was a casualty of this placement.

The second is the difficulty of the source. Kautilya wrote in compressed classical Sanskrit; the English translations of Shamasastry (1915), Kangle (1969) and Rangarajan (1992) are excellent but require sustained reading; the institutional content of the text is dispersed across fifteen books and 150 chapters and does not present itself as administrative law without some interpretive work. Lawyers trained to read judgments are not, on the whole, trained to do this interpretive work on a primary source from the fourth century BCE.

The third is the genuine and reasonable suspicion of using ancient texts for modern argument. The history of constitutional law in the last fifty years carries cautionary examples of selective ancient-text invocations made to do political work. The administrative-law reading of the Arthashastra proposed here is not that kind of invocation. The claim is not that the Arthashastra should be cited as authority for a modern administrative-law principle. The claim is that it should be read alongside modern administrative-law materials, as one source among several, for the institutional insight it carries — and that the failure to do so has been a loss to Indian administrative-law thought.

What the reading changes

A serious administrative-law reading of the Arthashastra would change three things about the present landscape. First, it would shift the location of the text in the Indian legal-studies syllabus — out of the political-philosophy chapter and into the administrative-law chapter, where its content is more useful and its institutional detail more legible. Second, it would supply the Indian administrative-law tradition with an indigenous intellectual genealogy that the present tradition lacks; the standard story has Indian administrative law beginning with Government of India Acts and ending with the present Article 32 jurisprudence, and the entire pre-modern period is mostly absent.

The exercise has an institutional precedent in Indian administrative thought that has been quietly underway for a generation. The committees of the 1970s and 1980s that re-thought public administration in India — from the Administrative Reforms Commissions onward — have, in their better moments, drawn on Kautilyan categories, even where they did not cite the Arthashastra by name. The Saptanga structure is recognisable behind some of the institutional-design proposals of the second Administrative Reforms Commission; the audit-and-discipline architecture of Book 2 informs the design of the Lokpal and Lokayukta legislation as it was eventually enacted. The lineage is there; what is missing is the open acknowledgement that the lineage is what it is.

Third, and most importantly, it would give Indian administrative-law thought a vocabulary for institutional design that is not borrowed from Dicey, from the American administrative-procedure tradition, or from the French Conseil d'État. The Saptanga as a typology of the state, the four-tiered hierarchy of sources of law, the writing-down rule for executive decisions, the audit function as an institutional check, the suo motu protective jurisdiction over disadvantaged categories, the discipline of the highest office-holder by way of daily justice-administration — these are concepts that the Arthashastra worked out in detail. Modern Indian administrative law has reached comparable answers, but slowly, through litigation, and without the institutional resource that an early indigenous source would have provided. The reading proposed here is not a substitute for the modern materials. It is a recovery of a source that should have been part of the conversation all along.

Get the week's biggest decisions, distilled.

A weekly digest for lawyers, students and the legally curious. Every Friday morning. Free.