Two constitutional amendments created roughly three million elected representatives, more than any comparable act of institutional design anywhere, and then declined to say what any of them were entitled to decide. That sentence is the whole implementation record of the 73rd and 74th Amendments, and every subsequent argument about devolution, weak mayors, ghost panchayats and unspent grants is a consequence of it. The amendments made local elections compulsory and local power optional, and the outcome was written into the text before a single panchayat was constituted.

The gap is not a story about administrative incapacity, corruption or state indifference, though all three exist and all three make it worse. It is a story about two verbs. Part IX says that there shall be constituted panchayats in every state, that elections shall be held before the expiry of a five-year term, and that a State Election Commission shall be established. Those are commands. It then says that the legislature of a state may by law endow the panchayats with such powers and authority as may be necessary to enable them to function as institutions of self-government. That is a permission. A state that holds elections on time and devolves nothing is in full compliance with the Constitution, and a great many states have chosen exactly that position for decades. The purpose of this article is to show how a reader can diagnose any state’s devolution record from that single asymmetry, and to be honest about what the amendments did achieve, because the reflexive verdict that they failed is as wrong as the official claim that they succeeded.

This article is about the record, not the content. The design of Part IX and the structure of panchayati raj as the 73rd Amendment set it up are treated in the article on the 73rd Amendment and panchayati raj, and the corresponding treatment of municipalities sits in the article on the 74th Amendment and urban local bodies. What follows assumes the reader knows roughly what the amendments contain and wants to know what happened afterwards.

The two amendments, and the bargain inside them

The Constitution (Seventy-third Amendment) Act, 1992 inserted Part IX and the Eleventh Schedule and came into force on 24 April 1993. The Constitution (Seventy-fourth Amendment) Act, 1992 inserted Part IXA and the Twelfth Schedule and came into force on 1 June 1993. Both were passed by Parliament in December 1992 and received assent on 20 April 1993, having been ratified by more than half the state legislatures, which they required because they altered the distribution of legislative and executive authority. The ratification requirement, and which categories of amendment attract it, is set out in the guide to the constitutional amendment process.

The passage record explains the bargain. Two earlier attempts had failed. The Constitution (Sixty-fourth Amendment) Bill, which sought to give panchayats constitutional status, was defeated in the Rajya Sabha in 1989, and the companion bill on municipalities did not survive either. A further attempt in 1990 lapsed. Opposition to those bills was not primarily about local government; it was about the Union writing rules for a subject that sat squarely on the State List and about the prospect of central schemes flowing directly to districts over the heads of state governments. The bills that became the 73rd and 74th Amendments succeeded because they were redrafted to remove precisely that threat.

The redrafting produced the asymmetry. Structure was made uniform and compulsory: three tiers, five-year terms, elections conducted by an independent commission, reserved seats, a finance commission every five years. Substance was left entirely to the states: what a panchayat or a municipality is actually permitted to decide, what staff it controls, and what money it commands. A state legislature could satisfy every mandatory requirement in Part IX and Part IXA while leaving its panchayats with no function beyond issuing certificates and its municipalities with no authority over the officer who runs them.

That was not an oversight. It was the price of getting the amendments through a Parliament in which state interests are represented, and the drafters understood what they were conceding. The uniform structure across India and the enormous variation in substance from state to state both follow directly from it.

What did the 73rd and 74th Amendments actually make compulsory?

They made structure compulsory and substance optional. Constituting panchayats and municipalities, holding elections before the five-year term expires, establishing a State Election Commission and a State Finance Commission, and reserving seats for Scheduled Castes, Scheduled Tribes and women are obligations. Endowing local bodies with powers, functions, staff and taxes is left to state discretion.

The devolution gap table

The table below is the diagnostic instrument this article offers. It takes the three Fs of devolution, which is the standard vocabulary in this field, and sets against each the constitutional provision that addresses it, whether that provision commands or permits, the mechanism through which it was meant to operate, and the failure mode that follows. A reader who holds this table can assess any state’s devolution record from a small number of questions.

The three Fs	Constitutional provisions	Command or permission	Intended mechanism	Typical failure mode
Functions	Article 243G for panchayats and Article 243W for municipalities, read with the Eleventh and Twelfth Schedules	Permission: the state legislature “may by law endow”	A state law assigning subjects from the schedules, followed by activity mapping that splits each subject between the tiers	Formal transfer of many subjects on paper with no activity mapping, so no tier knows which specific task it owns and the line department continues to perform all of them
Funds	Article 243H and Article 243X on taxes, tolls and fees; Article 243I and Article 243Y on the State Finance Commission	Permission on taxation, command to constitute the finance commission	Own-source revenue authorised by state law, plus a share of state revenues recommended by the finance commission and grants from central awards	Own-source powers withheld or set at unusable rates, finance commissions constituted late, reports delayed, recommendations not accepted, and untied funds replaced by tied scheme money
Functionaries	No provision addresses staff at all	Neither: the Constitution is silent	State law and executive order placing line-department staff under the control of the elected body	Staff remain on state department cadres, are appraised and transferred by the department, and treat the elected body as a stakeholder rather than an employer
The enforceable residue	Article 243B on constitution, Article 243E on terms and elections, Article 243K and Article 243ZA on the State Election Commission, Article 243D on reservation	Command	Direct constitutional obligation enforceable by the courts through the State Election Commission and by writ petition	Delay attempts by states, generally defeated in litigation, so this is the one column where compliance is high

Read the “command or permission” column downward. Everything a local body needs in order to govern sits in a permission or in a silence. Everything that determines whether a local body exists and who sits in it sits in a command. That single alignment predicts almost the entire implementation record, and it predicted it from 1993.

This is the permissive-verb failure, and it is worth stating as a general rule because it applies far beyond local government. Where a constitutional or statutory scheme makes existence mandatory and authority discretionary, the body will exist and will not have authority, because the discretion sits with the institution that stands to lose from exercising it. The state legislature that must decide whether to endow a panchayat with power over primary schools is composed of members whose own influence over primary schools in their constituencies depends on the line department remaining in charge. Asking that body to devolve is asking it to reduce the value of its own office.

What was built: the mandatory architecture

The parts of the amendments that used the imperative have a strong compliance record, and any honest implementation account has to begin there rather than with the failures.

Elections happen, and they happen on time more often than not

Before 1993, local elections in most states were held when the state government found it convenient, which in several states meant not for a decade or more. Supersession of an elected panchayat followed by prolonged administration by a state-appointed officer was routine, and there was no constitutional obstacle to it.

Article 243E changed that by fixing a five-year term, by requiring an election to be completed before the expiry of the term, and by requiring that a body dissolved early be reconstituted within six months. Article 243K vested the superintendence, direction and control of the electoral process in a State Election Commission whose commissioner can be removed only in the manner and on the grounds prescribed for a High Court judge. Article 243ZA does the same for municipalities.

The courts have taken those provisions seriously. The Supreme Court has held that state governments cannot postpone local elections on grounds of administrative convenience, incomplete delimitation of wards, or pending decisions on reservation, and that the State Election Commission’s constitutional position is analogous to that of the Election Commission of India within its sphere. The practical effect is that a state which wishes to delay must litigate and will usually lose. Delays still occur, and some have run for long periods where a state has manufactured a legal obstacle, but the default has reversed: postponement is now the exception that requires justification rather than the norm that required nothing.

That is a genuine and large achievement, and it deserves to be stated without qualification before the failures are catalogued. Regular local elections at a scale involving hundreds of thousands of bodies and millions of candidates are conducted across India, and they were not conducted before these amendments.

Reserved seats brought new people into elected office

Article 243D requires reservation of seats for Scheduled Castes and Scheduled Tribes in proportion to their share of the population in the area, and reservation of not less than one-third of all seats for women, including one-third of the seats reserved for Scheduled Castes and Scheduled Tribes. It applies the same requirement to the offices of chairperson, with rotation among constituencies. Article 243T does the same for municipalities.

The one-third figure is a floor and has been treated as one. Bihar legislated half the seats for women in panchayats, and a large number of states have since done the same, so in much of the country the reservation for women in local bodies is at or above half rather than at the constitutional minimum. Anyone describing one-third as the current position is describing the constitutional floor rather than the law in most states.

The composition of local bodies changed in a way that no other Indian institutional reform has matched. Whether that change in composition translated into a change in who exercises power is a separate question, treated below, and the honest answer is mixed and contested.

The State Election Commissions exist and function

Every state has a State Election Commission. Their independence is uneven, their staffing is often borrowed from state departments, and several have had to litigate against their own state governments to conduct elections on schedule. But they exist, they conduct elections, and they have a constitutional position that a state government cannot simply override. Compared with the position before 1993, when the conduct of local elections was an executive function of the state department of panchayati raj or urban development, this is a structural change of the first order.

The State Finance Commissions exist on paper

Article 243I requires the Governor to constitute a Finance Commission every five years to review the financial position of the panchayats and make recommendations on the distribution of taxes and duties between the state and the local bodies, the assignment of taxes to local bodies, grants-in-aid, and measures to improve their financial position. Article 243Y extends the same body’s mandate to municipalities. The Governor is required to lay every recommendation before the state legislature together with an explanatory memorandum on the action taken.

The obligation to constitute is a command and has been largely complied with, in the sense that states do eventually appoint these commissions. Everything after that point is where the mechanism breaks, and it is dealt with in detail below.

What was not built: the discretionary substance

Functions were transferred on paper and not in practice

The commonest form of non-implementation is not refusal but formal compliance. A state passes a conformity Act listing many or all of the twenty-nine subjects in the Eleventh Schedule as devolved to panchayats. The state then publishes the number of subjects transferred as evidence of its devolution record. Nothing changes on the ground, because a subject is not a task.

“Primary education” is a subject. Recruiting teachers, transferring teachers, approving school building works, sanctioning maintenance expenditure, monitoring attendance, procuring textbooks, and running the mid-day meal are tasks, and each of them can sit at the state, district, block or village level independently of where the others sit. Transferring the subject without allocating the tasks leaves everything where it was, because the line department continues to perform every task and no elected body has been given any of them.

The corrective is activity mapping, which means breaking each schedule subject into discrete activities and assigning each activity to a specified tier with the matching funds and staff. Activity mapping has been attempted seriously by a small number of states and either not attempted or attempted and abandoned by most. Where it has been done, devolution becomes measurable, because there is a document that says which body decides what. Where it has not, the reported number of transferred subjects carries no information at all.

This is the single most important thing a reader can know about assessing devolution claims. When a state says it has devolved twenty-nine subjects, the correct next question is whether it has published an activity map and whether the map assigns any decision that costs money.

Functionaries were never addressed by the Constitution at all

Part IX and Part IXA say nothing about staff. There is no article requiring that any officer be placed under the control of a panchayat or a municipality, no article about deputation, and no article about who writes an officer’s performance appraisal. The omission is complete, and it is arguably the most consequential silence in either amendment.

The result is that the person who runs a block development office, a primary health centre or a municipal engineering wing is a state government employee whose recruitment, promotion, transfer and appraisal are controlled by a state department. An elected panchayat president can ask that officer to do something. The officer’s career depends on someone else. Where an elected body has no control over the people who execute its decisions, its decisions are requests.

This is why the phrase “the three Fs” is slightly misleading if it suggests three parallel problems. Funds and functions are addressed by permissive provisions that a state may or may not use. Functionaries are addressed by nothing.

Municipalities got the same treatment with an extra layer

Article 243W is the municipal counterpart of Article 243G and uses the same permissive construction. But the urban position has a second structural problem that has no rural equivalent.

In most states the executive head of a municipal corporation is a commissioner appointed by the state government from the state civil service or the all-India services, who reports to the state urban development department and whose tenure is at the state’s disposal. The elected head, the mayor, is in most states elected indirectly by the councillors rather than directly by voters, and in several states holds office for a term substantially shorter than the council’s five years, sometimes with the post rotating among reserved categories within a single council term. A mayor with a one-year or two-and-a-half-year term, no control over the commissioner, and no independent revenue is an office with ceremonial content and very little executive content.

Why does a mayor often have less real authority than a sarpanch?

Because the municipal executive is a state-appointed commissioner while the panchayat has no equivalent parallel executive, and because many states give mayors short and rotating terms while panchayat presidents serve the full five years. In states that have devolved seriously to panchayats, a sarpanch may control more real decisions than the mayor of a large city.

That comparison surprises people and is worth stating carefully, because it is not universally true. In states with weak rural devolution, a sarpanch controls very little. In states with directly elected mayors and longer terms, the urban position is better. The general pattern holds because the structural obstacle is different in kind: rural devolution is blocked by a state legislature declining to use a permission, while urban devolution is blocked by that same non-use plus an executive officer who does not report to the elected body at all.

The money: why the fiscal channel is the weakest of the three

Of the three Fs, funds is the one where the constitutional text tried hardest and achieved least, and understanding why requires separating four different money flows.

Own-source revenue

Article 243H permits a state legislature to authorise a panchayat to levy, collect and appropriate taxes, duties, tolls and fees, to assign to a panchayat taxes collected by the state, and to provide for grants-in-aid from the consolidated fund of the state. Article 243X does the same for municipalities. Every verb in both articles is permissive.

States have used the permission unevenly and usually cautiously. Property tax is the principal own-source revenue available to municipalities, and its yield depends on assessment practice, valuation cycles, exemption policy and collection efficiency, all of which are typically constrained by state law or state instruction. Panchayats have a narrower base still, generally covering taxes on buildings, professions, markets, fairs, water and lighting, at rates that are either fixed by the state or capped by it. The result across most of India is that own-source revenue covers a small fraction of expenditure, and in many panchayats covers almost none of it.

There is a political reason for the caution that deserves to be stated fairly rather than treated as bad faith. Local taxation is unpopular, and an elected representative at any level who authorises a local body to raise a tax has created a grievance that voters will bring to them. States that have pushed local bodies to raise property tax realistically have generally had to absorb political cost to do it. The incentive against devolving taxing power is therefore not merely a desire to retain control; it is a desire to avoid being blamed.

The State Finance Commission channel

This is the mechanism the Constitution designed to solve the problem, and it is where the record is worst.

The design is sound. An independent commission constituted every five years reviews the financial position of local bodies and recommends how state revenues should be shared with them. The Governor must lay the report before the legislature with a memorandum on the action taken. The intent was to create for the state-to-local relationship the same disciplined, periodic, evidence-based transfer mechanism that the Finance Commission under Article 280 creates for the Union-to-state relationship.

Four failure modes recur, and any reader assessing a state’s record should test for all four.

The first is delay in constitution. Commissions are appointed late, sometimes years after the five-year cycle required, which means the award period is compressed or begins retrospectively.

The second is delay in reporting, often caused by inadequate secretariat support, absent or unreliable local body accounts, and terms of reference that arrive late. A commission asked to recommend a share of state revenues cannot do the work without data on local body expenditure, and in many states that data does not exist in usable form.

The third is non-acceptance. The recommendations are not binding on the state government, which may accept, modify or reject them. Rejection or partial acceptance is common, and the pattern across states is that the recommendations most likely to be accepted are those that cost the state least.

The fourth is the action-taken memorandum, which is a constitutional obligation and is treated in practice as a formality. Where the memorandum is laid late, or is laid in a form that does not engage with individual recommendations, the accountability the article was meant to create does not operate.

What happens to a State Finance Commission report after it is submitted?

Often not. The commissions are constituted, sometimes late; their reports are frequently delayed; their recommendations are not binding and are commonly accepted only in part; and the action-taken memorandum required to be laid before the legislature is often perfunctory. The constitutional obligation runs to constituting the body and laying the report, not to accepting what it says.

That last sentence is the whole problem in one line. The Constitution commanded the process and left the outcome to discretion, which is the permissive-verb failure operating in the fiscal register rather than the functional one.

The central transfer channel

The one flow that has grown reliably is the one the amendments did not create. Successive central Finance Commissions have recommended grants for local bodies out of the divisible pool, and those grants have increased substantially over successive award periods and have generally been routed to local bodies with conditions attached, such as the maintenance of accounts, their audit and publication, and in later awards the earmarking of a share for specified basic services.

This channel matters for three reasons. It is predictable in a way that state transfers frequently are not. It is untied or lightly tied compared with scheme money, which gives a local body something to actually decide about. And its conditions have done more to improve local body accounting practice than any state instruction, because the money stops if the accounts are not produced.

It also creates an oddity worth noticing. The Union, which has no constitutional relationship with panchayats and municipalities at all, has become a more reliable source of untied local finance than the state governments that hold the constitutional obligation. That is not what the design contemplated, and it has fuelled the recurring political objection that central transfers to local bodies bypass the state.

The scheme channel, which is the largest and the least useful

Most money that reaches a panchayat or a municipality arrives as a centrally sponsored or state scheme with its own guidelines, its own implementation society, its own reporting format and its own permitted expenditure heads. A body that spends large sums under scheme guidelines is an implementing agency, not a government. It has no discretion over allocation between sectors, no ability to move money from a purpose the guideline permits to a purpose the village actually needs, and no ability to build the local judgment that self-government is supposed to develop.

The distinction between money spent and money decided is therefore the single most useful correction to devolution statistics. A state can report large per-panchayat expenditure and have devolved nothing, because the expenditure is scheme money the panchayat was required to spend in a prescribed way. The question that separates the two is simple: how much of the money reaching this body can it allocate as it chooses?

Parallel bodies: the bypass that hollows out the schedules

If the permissive verb is the structural cause of the devolution gap, parallel bodies are the mechanism through which the gap is maintained in practice, and they deserve their own treatment because they are the least understood part of the picture.

A parallel body is an institution created outside Part IX or Part IXA which performs a function listed in the Eleventh or Twelfth Schedule. The category includes district rural development agencies, registered societies and missions created to deliver national programmes in education, health, sanitation, water supply, skills and livelihoods, urban development authorities that control land use and large infrastructure inside municipal limits, water and sewerage boards that operate at state level, and special purpose vehicles created for particular projects.

Their common features are instructive. They are usually registered societies or statutory authorities rather than constitutional bodies, so Part IX has nothing to say about them. They are usually chaired by an officer or a minister rather than by an elected local representative. They usually have their own funds, their own staff and their own reporting line to a state department or to the Union. And they usually perform precisely the functions that the Eleventh and Twelfth Schedules list.

The effect is that a state can devolve a subject on paper while the actual delivery of that subject runs through a body the panchayat does not control. Rural water supply may be a devolved subject and be delivered by a state water board. Primary education may be a devolved subject and be delivered through a mission with its own district and block structures. Urban planning may be a Twelfth Schedule function and be exercised by a development authority whose board the municipality does not control and whose plans the municipality cannot alter.

Parallel bodies are not always indefensible. Some exist because a function genuinely requires a scale larger than a panchayat or a municipality, such as a regional water source or a metropolitan transport network. Some exist because a national programme needs uniform delivery and financial accountability that local bodies could not initially provide. The honest criticism is not that they exist but that they were allowed to occupy the constitutional space that Part IX assigned to elected bodies, and that no mechanism was ever built to fold them back in as local capacity improved.

What counts as a parallel body in local government?

They are societies, missions, authorities and boards created outside Part IX which perform Eleventh or Twelfth Schedule functions with their own funds, staff and reporting lines. Because the Constitution says nothing about them, a state can list a subject as devolved while a parallel body continues to deliver it, leaving the elected body with the label and none of the authority.

The planning committees, and the most complete non-implementation in either amendment

Article 243ZD requires that a District Planning Committee be constituted in every district to consolidate the plans prepared by the panchayats and the municipalities in the district and to prepare a draft development plan for the district as a whole, with not less than four-fifths of its members elected by and from among the elected members of the district panchayat and the municipalities in the district, in proportion to the ratio of rural to urban population. Article 243ZE requires a Metropolitan Planning Committee for every metropolitan area, with the same four-fifths requirement.

These are commands, not permissions, and they are the strongest devolutionary provisions in either amendment, because they place spatial planning for an entire district in a body dominated by elected local representatives rather than in a state department.

They are also the least implemented provisions in either amendment. District Planning Committees have been constituted in most states, in the sense that a notification exists, but in many they do not meet, do not receive plans from the constituent bodies, do not prepare a consolidated district plan, or prepare one that has no relationship to the budgets that actually get spent. Metropolitan Planning Committees are worse: several of India’s largest metropolitan areas have had no functioning committee for long periods, and metropolitan planning is instead done by development authorities whose composition owes nothing to Article 243ZE.

The reason for this failure is different from the reason for the functional failure, and the difference matters. District and metropolitan planning committees do not depend on a state legislature exercising a permission. They depend on a state government constituting a body and then routing planning through it, which requires the state to accept that a committee of elected local representatives will decide the spatial and sectoral shape of district investment. Very few state governments have been willing to do that, and because the obligation is not attached to any consequence, nothing follows from declining.

This is the clearest available evidence for a point that runs through the whole implementation record. Even the mandatory provisions of Part IX have been complied with only where non-compliance had a visible and enforceable consequence. Elections have a consequence, because an election petition or a writ petition produces a court order and a date. Constituting a planning committee has no equivalent consequence, so the compliance is nominal.

Diagnosing the gap: four kinds of failure, only one of which is administrative

Implementation writing tends to collapse every shortfall into “poor implementation”, which explains nothing and suggests no remedy. The four categories below are distinguishable by evidence, and the remedy differs for each.

Drafting failure

This is the largest category here and the one the record supports most strongly. The devolution of functions was written as a permission addressed to the body least likely to use it. The devolution of staff was not written at all. The finance commission mechanism was written as a process obligation with no obligation as to outcome. Parallel bodies were not addressed. The Eleventh and Twelfth Schedules were drafted as lists of subjects rather than as allocations of tasks, so even a willing state had no template for what transfer meant.

The test for drafting failure is whether a state acting in complete good faith would still have been able to under-deliver without breaching anything. Here the answer is plainly yes, and that is what makes this a drafting failure rather than an implementation one.

Deliberate non-implementation

This category exists and should be named, but it is smaller than the rhetoric suggests and larger than official accounts admit. Deliberate non-implementation shows up as formal compliance designed to defeat the purpose: transferring subjects without activity mapping, constituting a finance commission and then rejecting its recommendations, creating a parallel body to deliver a function immediately after devolving it, or setting mayoral terms short enough that no mayor can build an independent base.

The evidence for deliberateness is the pattern rather than any single act. Where a state has done all four of those things, the most economical explanation is a policy of retaining control while satisfying the text. Where a state has done one of them, incapacity is at least as good an explanation.

Administrative and capacity failure

This is real and is often described dismissively by people who have not tried to build the capacity in question. A village panchayat with a part-time secretary, no accountant, no engineer, and no computerised accounting system cannot manage a substantial budget, cannot produce audited accounts, and cannot prepare a plan that a district committee could consolidate. The absence of that capacity is a genuine obstacle to devolution and not merely an excuse for withholding it.

But the capacity argument has a circularity that must be named. Capacity is built by exercising responsibility, and a body that is never given responsibility never builds capacity, which is then cited as the reason for not giving it responsibility. States that devolved seriously found that capacity followed, unevenly and with failures, but it followed. The capacity argument is a reason for sequencing devolution and building support systems, not a reason for indefinite postponement.

Resource failure

Local bodies are under-resourced, but this is largely a consequence of the other three failures rather than an independent cause. A body with no assigned taxes, no accepted finance commission award and no untied grant is not short of money because money is scarce in the system; it is short of money because the channels that were supposed to carry money to it were left to discretion and the discretion was exercised the other way. The central Finance Commission grants demonstrate the point, since they show that when a reliable untied channel is created, money does reach local bodies and is spent.

Where exactly did the devolution promise break down?

Partly, and unevenly. They devolved the right to elect a local government, which is real and now universal. They did not devolve the power to decide, because the provisions dealing with functions and finance were written as permissions to state legislatures and the provision dealing with staff was not written at all. A small number of states have devolved substantially; most have devolved formally.

Where it works, and why

The variation across states is the most useful natural experiment in this whole area, because the constitutional framework is identical everywhere and the outcomes are not. Three patterns explain most of the difference.

The first is prior commitment. States that had built functioning local government before 1993 generally continued to have it afterwards. Maharashtra and Gujarat had operated district-level rural bodies with substantial staff and budget since the 1960s. Karnataka legislated an ambitious panchayat system in the mid-1980s that transferred real authority to district bodies. West Bengal had been holding regular panchayat elections on a party basis since 1978 and had built a working relationship between panchayats and the state administration. In each case, the amendments consolidated something already present rather than creating it. Where nothing existed before 1993, the amendments created the structure and the substance largely did not follow.

The second is a deliberate fiscal decision. Kerala’s approach from the mid-1990s, in which a large share of state plan funds was devolved to local bodies as untied money accompanied by a participatory planning process and a substantial training effort, is the clearest demonstration available that the constitutional framework permits real devolution if a state chooses it. The essential move was not a change in the constitutional text, which is the same in Kerala as everywhere else, but a decision to convert the permission in Article 243G into an actual transfer of money that local bodies could allocate themselves. The programme has been criticised on several grounds, including the quality of local planning and the burden placed on untrained bodies, and those criticisms are worth taking seriously. What is not seriously disputed is that the transfer happened and that it was possible under the existing text.

The third is activity mapping. States that broke schedule subjects into discrete activities and assigned each one, with matching funds and staff, produced measurable devolution. States that transferred subjects without mapping produced none. This is the practical lesson most easily transferred to any other state, and it requires no constitutional change at all.

The general conclusion is uncomfortable for both sides of the reform argument. The framework does not compel devolution, which is the critics’ point and is correct. But the framework does not prevent it either, which is the defenders’ point and is also correct. Every state that wanted to devolve has been able to, and the number that wanted to is small.

Reservation for women: the change that did happen, and the argument about it

The reservation of at least one-third of seats and chairperson posts for women, raised to half in a large number of states, is the most visible consequence of these amendments and the most argued about.

The case that it produced real change rests on several observations. The number of women holding elected office in India is dominated by local bodies, and it rose from a small base to a very large one directly because of Article 243D. Reserved chairperson posts have placed women in positions that carry executive authority rather than only representative status. And the rotation of reserved constituencies means that over successive terms a large number of constituencies have had a woman as chairperson at some point, which changes what voters have seen.

The case that it produced less change than the numbers suggest rests on the proxy phenomenon, which is real and well documented in qualitative research and in journalism. In many places an elected woman’s husband or another male relative attends meetings, signs papers in practice if not in law, deals with officials and is treated by the administration as the effective office-holder. The vocabulary that has grown up around this, in which the husband of a village head is addressed by a title derived from the office, exists because the phenomenon is common enough to need a word.

The honest position holds both of these at once and adds a third observation that is frequently omitted. Proxy representation is most pronounced in a first term and in places with low female literacy and strong social restriction, and it declines with experience, training and repeat candidacy. Women who have served a term have contested and won unreserved seats, have contested at higher tiers, and have in some cases built independent political careers. The mechanism by which reservation works is therefore slower than its advocates claimed and faster than its critics allow: it does not convert a reserved seat into real authority immediately, and it does create a pool of women with electoral experience where none existed.

What cannot responsibly be asserted is a number. Published estimates of the prevalence of proxy representation vary enormously depending on the state, the tier, the method and the period studied, and any single figure quoted for India as a whole should be treated with suspicion. The pattern and its direction are what the evidence supports.

Did reserved seats put women in charge or only in office?

It has changed who holds office decisively and who exercises authority more slowly. Reserved seats and reserved chairperson posts placed very large numbers of women in local government, and proxy exercise of that office by male relatives is common, particularly in first terms. The proxy pattern weakens with experience and repeat candidacy, so the effect accumulates across terms rather than appearing immediately.

The areas Part IX does not reach

An implementation account that ignores the exclusions overstates the coverage. Article 243M provides that Part IX does not apply to the Scheduled Areas and tribal areas referred to in Article 244, to Nagaland, Meghalaya and Mizoram, to the hill areas of Manipur for which district councils exist, and to certain other specified areas. Comparable exclusions apply on the urban side.

Parliament extended Part IX to the Scheduled Areas by a separate statute in 1996, with substantial modifications designed to protect customary practice and to give the gram sabha powers over land, minor forest produce, consultation before acquisition and other matters that ordinary panchayats do not possess. That statute has its own implementation record, which is in several respects worse than the record of the amendments themselves, and it is treated in the article on the extension of panchayat provisions to Scheduled Areas. The autonomous district councils in the north-eastern states operate under a different constitutional scheme altogether.

The practical consequence for a reader assessing national coverage is that statements about universal local self-government in India need qualification, and that the areas excluded from Part IX are disproportionately those where the relationship between the state and the community is most contested.

A comparative check: what a differently drafted amendment achieved

The most instructive comparison is not with another country but with another Indian constitutional amendment that faced a structurally similar problem and solved it differently.

The 101st Amendment created the Goods and Services Tax Council, a body in which the Union and every state are represented and which recommends the design of an indirect tax that all of them levy. Like the 73rd and 74th Amendments, it required states to give something up. Unlike them, it specified the institution’s composition, its voting rule and the majority required for a decision in the constitutional text itself, and it paired the surrender with a compensation guarantee enacted in the same instrument. The result is an institution whose authority does not depend on any state choosing to exercise a permission, and whose decisions are made by a defined procedure rather than by discretion. The mechanics of that design are set out in the article on the 101st Amendment and the GST Council.

The comparison is not an argument that local government should have been designed the same way, because the two problems differ in important respects. It is an argument about what constitutional drafting can achieve when it chooses to. A constitutional amendment can specify a decision rule, a composition, and a consequence for non-compliance. Part IX specified composition and terms, and left the decision rule about power to the body with the incentive to withhold it.

The same lesson appears in statutory form elsewhere in this series. Transparency legislation works, where it works, because it creates a specific duty on a named officer, a deadline, an appeal to an independent authority, and a penalty. A reader who wants to know why a right becomes real rather than notional should compare the machinery, not the language of the entitlement.

The election record examined more closely

Because timely elections are the one area of high compliance, it is worth understanding exactly what the courts have had to enforce and where the pressure points remain.

Three recurring devices have been used by state governments to delay local elections, and all three have been litigated.

The first is incomplete delimitation of wards or constituencies. Because ward boundaries must be redrawn after a census or after municipal boundary changes, a state can defer an election by leaving the delimitation exercise unfinished. The judicial answer has been that the constitutional obligation to complete an election before the expiry of the term is not conditional on the state completing an administrative exercise it controls, and that the State Election Commission may proceed on the existing position if necessary.

The second is a pending decision on reservation, particularly the reservation of seats for backward classes, which several state laws provide for in addition to the constitutional reservations for Scheduled Castes, Scheduled Tribes and women. Because backward-class reservation in local bodies is a matter of state law rather than a constitutional command, disputes about its extent have repeatedly delayed elections. The courts have developed a demanding standard for such reservation, requiring a state to establish a dedicated commission, to gather empirical data on backwardness in local bodies specifically rather than relying on general lists, and to keep the aggregate reservation within the overall ceiling. States that have not done that work have found their reservation notifications set aside, and elections have then been held without backward-class reservation or delayed further while the exercise is carried out. This has become the largest single source of local election delay.

The third is dissolution followed by administration. A state that dissolves a local body must reconstitute it within six months, and the six-month rule has generally been enforced. Attempts to hold bodies under administrators for extended periods have not survived challenge where they have been challenged, though the qualification matters: enforcement depends on somebody bringing the challenge, and in smaller bodies with weaker political stakes nobody always does.

Two structural features shape all of this. Article 243-O bars courts from interfering with the delimitation of constituencies or the allotment of seats, and provides that an election to a panchayat may be questioned only by an election petition presented to the authority and in the manner provided by state law. That bar limits pre-election litigation and channels post-election disputes into a statutory forum, which is a design borrowed from Article 329 for parliamentary and assembly elections. And the State Election Commission, though constitutionally protected in the commissioner’s tenure, usually depends on the state government for staff, budget and the machinery of election administration, which limits how far it can act against the state’s wishes without a court order.

The overall picture is therefore not that elections happen automatically. It is that a state which wants to delay must construct a legal obstacle, and that constructing one is possible but costly and usually temporary. That is a meaningful constitutional achievement, and it is the difference between an enforceable command and a permission.

The urban problem, which is not the same problem

Municipal government is failing in a way that shares the constitutional cause but has three additional features, and treating the urban and rural stories as one obscures the diagnosis.

Functional fragmentation

In a large Indian city, the functions listed in the Twelfth Schedule are typically distributed across a municipal corporation, a development authority responsible for land use and major infrastructure, a water and sewerage board, a state transport undertaking or metro corporation, a state pollution control body, a police force under the state government, and one or more special purpose vehicles created for particular projects. The elected municipal body is one actor among many, usually the one with the weakest revenue base and no authority over the others.

This fragmentation is not a failure of the 74th Amendment so much as a condition the amendment did not address. The Twelfth Schedule lists urban planning including town planning, regulation of land use, roads and bridges, water supply, public health and sanitation, fire services, urban forestry, and several others. It says nothing about what happens to a function that a state statutory authority is already performing, and Article 243W’s permissive construction meant no state had to resolve the conflict.

Executive control

The commissioner problem described earlier is the most consequential single feature of Indian urban government. An elected council that cannot appoint, direct, appraise or remove its chief executive is not the government of the city in any operational sense. In the states that have experimented with strengthening the elected side, the reforms that mattered were the ones that changed this relationship, either by making the commissioner accountable to the elected head or by giving the elected head a longer term and direct election.

Revenue

Municipal own-source revenue depends overwhelmingly on property tax, and property tax in India yields far less than in comparable systems because of a combination of outdated valuations, wide exemptions, poor coverage of the property base, and rate-setting that is often controlled by the state rather than the municipality. Every serious municipal finance analysis in India reaches the same conclusion about the direction of the problem, though estimates of magnitude vary by source and method. A municipality that cannot set or effectively collect its principal tax cannot fund its own priorities, which returns it to dependence on state transfers and scheme money.

Why are Indian cities described as ungoverned?

Because authority over a city is divided among a municipal corporation, state development authorities, state utility boards and state departments, none of which answers to the elected council, while the council’s own executive head is a state appointee and its principal tax is constrained by state law. The 74th Amendment created an elected body without consolidating city government into it.

The reform proposals that follow from this diagnosis are contested rather than obvious, and the article returns to them below.

Accountability from below, which the amendments did address

One part of the design has been under-used rather than blocked, and it deserves attention because it is the only mechanism in Part IX that does not depend on a state legislature.

Article 243A provides that a gram sabha, meaning the body of persons registered in the electoral rolls of a village, may exercise such powers and perform such functions at the village level as the legislature of a state may provide. The verb is permissive again, but the institution is constitutional and exists everywhere, and states have generally given it at least a formal role in approving plans, identifying beneficiaries and reviewing accounts.

Where the gram sabha has been given a defined function with a consequence attached, it has worked better than most other accountability mechanisms in Indian local government. Statutory social audit requirements in employment and food programmes, which route verification through the gram sabha rather than through an administrative inspectorate, have produced findings that internal audit did not. The reason is structural: the gram sabha contains the people who know whether a road was built, which is information no auditor can obtain from a file.

The limits are equally structural. Attendance is often low, meetings are frequently held without genuine notice, resolutions are recorded that were never passed, and dominant local interests can capture the proceedings as easily as they can capture an elected body. The gram sabha is a mechanism, not a guarantee, and its performance depends on whether anything follows from what it finds.

The urban counterpart is weaker. Article 243S requires ward committees in municipalities above a specified population, and the composition and functions of those committees are left to state law. In many states a ward committee covers several wards and includes a number of councillors rather than residents, which makes it a smaller council rather than a neighbourhood body. Where cities have created genuine area sabhas or ward committees with resident membership, the results have been better, but the constitutional obligation does not require it.

The reform debate, stated fairly on both sides

There are two serious reform positions and they are not compatible.

The constitutional amendment position

This position holds that the permissive verb is the disease and that only a constitutional amendment can cure it. In its strongest form it proposes making the endowment of functions in Article 243G and Article 243W mandatory, specifying a minimum set of functions that must be devolved, requiring that staff performing devolved functions be placed under the control of the elected body, and making the State Finance Commission’s recommendations binding or at least requiring reasoned rejection with legislative approval.

The case for it is the whole implementation record set out above. Nothing in seventeen years of trying has changed the fact that a state may decline to devolve and suffer no consequence, and the states that devolved did so for their own reasons rather than because the framework pushed them. If discretion is the cause, removing discretion is the remedy.

The case against it has three parts, and each deserves to be stated properly. The first is that it would require ratification by at least half the state legislatures, and the states are precisely the parties whose discretion is being removed, so the arithmetic is unpromising. The second is that a mandatory devolution list drafted centrally would flatten genuine variation: a state with three thousand large village panchayats and a state with two hundred thousand small ones do not need the same allocation, and a uniform mandate would create bodies too small to perform what they had been given. The third is the objection of principle, that local government is a state subject and that the Union prescribing what states must hand to their own local bodies is a further centralisation dressed as decentralisation.

The fiscal position

This position holds that the functional argument is a distraction and that money determines everything. It proposes strengthening the State Finance Commission mechanism, standardising its timing and terms of reference, requiring the action-taken memorandum to address recommendations individually, expanding untied central transfers to local bodies with accounting conditions, and reforming property tax so that municipalities have a real own-source base.

The case for it is that it requires no constitutional amendment, that the one channel which has demonstrably strengthened local bodies is the untied central grant, and that a body with money to allocate acquires functions in practice regardless of what a state statute says, because departments will negotiate with a body that can co-fund.

The case against it is that money without authority produces implementation agencies rather than governments, that untied central transfers deepen the fiscal dependence of local bodies on a level of government with no constitutional relationship to them, and that property tax reform is politically hard for exactly the reasons local taxation has always been politically hard.

The middle position

A third position, less frequently articulated but arguably better supported by the record, holds that the highest-return interventions require neither a constitutional amendment nor a fiscal transformation. Mandatory published activity mapping, enforcement of the existing obligation to constitute and operate district and metropolitan planning committees, placing staff performing devolved functions under local appraisal even while they remain on state cadres, and fixed longer terms for elected municipal heads are all achievable under the existing text. The evidence for this position is that every state which has devolved substantially did so under the same constitutional provisions everyone else has.

This article does not resolve the disagreement between the three positions, because it turns partly on how much uniformity a federation should impose and that is a question about values rather than evidence. What the record does establish is narrower and worth stating plainly: the failure is not primarily one of capacity or resources, it was predictable from the drafting, and it has been overcome by states that chose to overcome it.

The three Fs, and the fourth that nobody names

The vocabulary of funds, functions and functionaries is the standard shorthand in Indian devolution debate, and it is useful because it separates three things that are genuinely separate. A body can have a function and no money, money and no staff, or staff and no function, and each combination produces a different kind of paralysis.

The shorthand is incomplete in one respect that matters. There is a fourth condition without which the other three do not add up to self-government, and it has no agreed name. Call it discretion. A body may have a function assigned to it, money allocated for that function, and staff to execute it, and still not be governing, if the function is defined by a scheme guideline, the money is tied to that guideline, and the staff report to the department that wrote it. That describes a large share of local bodies in India accurately. They are administering, and administering is not deciding.

Discretion is the thing the Eleventh and Twelfth Schedules were understood to be about and the thing the constitutional text never secured. The schedules list subjects because the drafters were thinking in terms of the legislative lists, where an entry does confer a power to decide. But a schedule entry in Part IX is not a legislative entry. It confers nothing by itself; it merely names a subject that a state legislature may choose to devolve. The similarity in form to the Seventh Schedule has misled a great many readers, including some who write about the subject professionally.

Do the Eleventh and Twelfth Schedules transfer subjects by themselves?

No. Unlike the Seventh Schedule, which distributes legislative competence, the Eleventh and Twelfth Schedules are lists of subjects that a state legislature may devolve to panchayats and municipalities under Article 243G and Article 243W. Neither schedule transfers anything by its own force, and a state that devolves none of the listed subjects breaches no constitutional obligation.

The fourth F also explains why devolution statistics are so unreliable. A state reporting the number of subjects devolved is reporting on the second column of the table. A state reporting the volume of funds flowing through local bodies is reporting on money that is mostly tied. Neither reports on discretion, and discretion is what the reader wants to know about.

A diagnostic protocol for any state’s record

The devolution gap table becomes a working instrument when it is turned into a sequence of questions. Six questions will place any state accurately, and all six can be answered from public documents.

The first question is whether the state has published an activity map, and whether that map assigns discrete activities rather than subjects. If there is no activity map, the number of subjects devolved carries no information and should be disregarded.

The second is whether any assigned activity involves a decision that costs money. Issuing certificates, maintaining registers and certifying beneficiary lists are administrative tasks. Deciding which road is built, which repair is funded, or which of two competing local demands is met is governing. If no assigned activity involves the second kind of decision, the state has devolved nothing of substance regardless of how many subjects appear in its conformity Act.

The third is whether the staff performing devolved activities are appraised by the elected body or by the line department. This single question separates real devolution from formal devolution more reliably than any other, and the answer is usually available from state service rules.

The fourth is what proportion of the money reaching local bodies is untied. Untied money is money the body may allocate between purposes. Scheme money is not. A state whose local bodies receive large sums of which almost none is untied has built an implementation network, not a system of local government.

The fifth is whether the State Finance Commission has been constituted on time, whether its report was laid, and whether the action-taken memorandum engages with individual recommendations. A state that has skipped a cycle or laid a memorandum that accepts recommendations in principle without specifying which has told the reader what its intentions are.

The sixth is whether the district planning committee actually meets and produces a consolidated plan that relates to the budget. Since this obligation is mandatory rather than permissive, non-compliance here is the strongest available evidence of deliberate rather than incidental under-devolution.

The state-by-state detail behind those six questions, including the comparative devolution record and the index exercises that attempt to score it, is taken further in the article on the devolution of funds, functions and functionaries. A state that answers well on all six has devolved. A state that answers well on the first two and badly on the rest has devolved on paper. A state that answers badly on all six has complied with Part IX by holding elections, which is all Part IX required of it.

What the amendments got right, and are rarely credited for

An implementation article that ends in a catalogue of failures misdescribes the record, so it is worth listing what would look different if the amendments had never been passed.

Local bodies would be creatures of state statute, abolishable and supersedable at will, and the historical record before 1993 shows that the will was frequently exercised. Elections would be held when convenient. There would be no independent authority conducting them. There would be no constitutional floor on the representation of Scheduled Castes, Scheduled Tribes and women, and the number of women in elected office in India would be a small fraction of what it is. There would be no periodic institutional review of local finances at all, however imperfectly the reviews are acted upon. And there would be no constitutional standing from which a local body or a citizen could litigate against a state that wished to postpone an election indefinitely.

Every one of those is a real change and none of them is reversible without a further constitutional amendment. The amendments made local government permanent, elected and representative. What they did not do is make it powerful. Those are different achievements, and the second failure does not cancel the first.

The distinction also explains why the two most common verdicts on these amendments are both defensible and both incomplete. A person who says they succeeded is describing the mandatory column. A person who says they failed is describing the permissive column. The verdict that fits the whole record is that they succeeded completely at what they commanded and failed almost completely at what they merely permitted, which is exactly what a reader should expect and is the reason the permissive-verb framing is worth carrying to other statutes.

What the record establishes as a general rule

Three propositions generalise beyond local government, and they are the portable content of this article.

The first is that in a constitutional or statutory scheme, the verb is the design. A provision that says “shall” creates an obligation someone can enforce. A provision that says “may” creates an option whose exercise depends entirely on the incentives of the body holding it. A drafter who wants an outcome must either command it or attach a consequence to its absence, and a reader assessing any statute should read the verbs before reading anything else.

The second is that discretion should never be given to the party that loses by exercising it. Article 243G gives the power to devolve to the state legislature, which is composed of members whose constituency influence depends on line departments remaining in charge. That is not a criticism of any state; it is a prediction that holds across parties, regions and decades, and it has held.

The third is that mandatory institutional obligations are complied with in proportion to the visibility and enforceability of the consequence of breach. Elections are held because a court will order one on a date. Planning committees are notified but do not function because nothing follows from their not functioning. Finance commissions are constituted but their reports are shelved because the obligation runs to the process and not the outcome. If a reform wants an obligation honoured, it must specify what happens when it is not.

Those three rules can be applied to almost any statute in this series, and the local government record is the cleanest available demonstration of all three operating at once on the same instrument.

What would count as success, and why the question is rarely answered

A verdict on implementation depends on a measure, and one reason this debate never resolves is that the participants use incompatible measures without saying so.

The official measure is compliance. Counted this way, implementation is close to complete: panchayats and municipalities exist everywhere Part IX applies, elections are held, commissions exist, seats are reserved, conformity legislation has been passed. On this measure the amendments succeeded and the argument is over.

The devolution index measure counts subjects transferred, funds routed and staff placed, usually weighted and aggregated into a score by state. This is better, because it tries to measure substance, but it inherits the paper-transfer problem: a state that lists subjects without activity mapping scores well, and a state that has devolved three functions completely may score worse than a state that has devolved twenty-nine nominally. Weighting choices drive the rankings, and different exercises produce different orders.

The expenditure measure counts money spent by or through local bodies. This is the least informative of the three, because it does not distinguish between money a body allocates and money it disburses under a guideline.

The measure that would answer the question is a discretion measure: what proportion of the decisions that affect a citizen in a given area are taken by a body that citizen elects locally. Nobody publishes that, because it is hard to construct and harder to standardise across states, and its absence is why the debate runs on anecdote and index scores.

For a reader who wants a rough version of it, the six diagnostic questions above are a serviceable substitute. They do not produce a number, and their advantage is precisely that they do not, because a number in this field invites the false precision that has made much of the published assessment unreliable.

There is a second and deeper problem with measurement here that is worth naming. Devolution is not obviously good in every dimension, and an honest assessment has to accept that. Devolved decision-making is more responsive to local preference and more exposed to local capture. A village body may allocate a road better than a state department because it knows which road is used, and may allocate a benefit worse because it knows which family is politically inconvenient. Reservation brings excluded groups into office and does not by itself protect them from the social power that excluded them. The case for devolution rests on the judgment that responsiveness and participation are worth their costs, and that capture is better addressed by transparency and by higher-tier appeal than by centralising the decision. That judgment is defensible and is a judgment, not a finding.

The verdict

The 73rd and 74th Amendments did what they commanded and did not do what they permitted, and the distribution between those two categories was fixed by the drafting rather than by anything that happened afterwards.

They created a permanent, elected, representative tier of government across almost the whole country and made its existence and its renewal non-negotiable. That is a structural achievement of the first order and it is irreversible without a further amendment. Before 1993 the existence of an elected local body in most of India was a matter of state convenience; after 1993 it is a constitutional obligation enforced by courts. The reservation provisions changed the composition of elected office in India more than any other single measure, and while the exercise of that office lagged the holding of it, the lag has narrowed with each electoral cycle.

They did not create local self-government, because self-government requires the power to decide and the power to decide was left in the hands of state legislatures under a permissive verb. The result across most of India is elected bodies with real legitimacy, real terms, real reservations and very little to determine. The subjects listed in the Eleventh and Twelfth Schedules are mostly delivered by line departments, state authorities and programme societies over which the elected bodies have no control. The staff who deliver them answer elsewhere. The money that funds them arrives tied. And the two mandatory provisions that could have changed this, the district and metropolitan planning committees, were complied with formally and hollowed out in practice, because no consequence attaches to their not working.

The single most useful conclusion is that this outcome was predictable in 1992 and was predicted at the time by people who read the text carefully. The permissive verb in Article 243G was not an accident of drafting; it was the concession that allowed the amendments to pass after two earlier attempts had failed on precisely this ground. The 73rd and 74th Amendments bought uniform structure at the price of discretionary substance, and the price turned out to be almost the entire value of the transaction from the perspective of local authority, while the structure turned out to be worth considerably more than critics allowed.

Whether that trade was worth making depends on the counterfactual. If the alternative was a stronger amendment that failed in the Rajya Sabha as its two predecessors had, then a weak amendment that passed was the right choice and the subsequent record is the price of getting anything at all. If the alternative was a stronger amendment that could have passed with more negotiation, the concession was premature. The passage record supports the first reading more than the second, but not overwhelmingly, and reasonable people who have studied the same record disagree about it.

What is not in reasonable dispute is the diagnosis. The failure is in the verb, not in the capacity of local bodies, not in the scarcity of resources, and not in the character of local politics. Every state that decided to devolve was able to devolve under the same text, which disposes of the argument that the framework prevents it, and almost no state decided to, which disposes of the argument that the framework compels it.

What a resident can actually do when a local body does not function

The implementation record is usually written for policy readers, but the most common way people encounter it is personal: a ward that has had no election for years, a gram sabha that is minuted but never held, a municipal budget nobody has seen, a finance commission report that exists and has vanished. It is worth setting out what the constitutional structure makes available in each case, because the answer differs sharply depending on which column of the devolution gap table the complaint falls into.

Where the complaint concerns an election that has not been held, or a body dissolved and not reconstituted within six months, the obligation is a command and the route is a writ petition, usually against the state government and the State Election Commission together. This is the strongest position a complainant can be in under Part IX, and the case law supports it.

Where the complaint concerns reservation, whether the rotation of reserved chairperson posts or the extent of reservation for backward classes under state law, the position is more complex, because part of the scheme is constitutional and part is statutory, and because Article 243-O restricts pre-election challenges to delimitation and seat allotment and channels election disputes into a statutory election petition.

Where the complaint concerns the failure to constitute or operate a district planning committee, or the failure to lay a finance commission report and an action-taken memorandum before the legislature, the obligation is again a command, and these are the least litigated mandatory obligations in Part IX. That is an opportunity rather than an obstacle.

Where the complaint concerns the substance of devolution, that a panchayat or a municipality has been given nothing to decide, there is no constitutional remedy at all, because the provision is permissive. The available routes are political and informational rather than judicial: transparency requests for the conformity Act, the activity mapping notification, the budget and the audited accounts, followed by argument in the state legislature and in public. The gram sabha or ward committee is the forum the Constitution supplies for the informational part, and where it functions it is more effective than any external inspection, because the people in the room know what was actually built.

The distinction between those four situations is the practical payoff of the whole analysis. A complaint that sits on a command has a forum and a remedy. A complaint that sits on a permission has neither, and treating the second as though it were the first wastes time and money that could be spent on the routes that work.

Using this article

Three practical notes for readers who intend to work with this material.

Anyone assessing a particular state should treat published devolution rankings as a starting point and run the six diagnostic questions against that state’s conformity Act, its activity mapping notification if one exists, its most recent State Finance Commission report and action-taken memorandum, and its district planning committee notification. Those five documents answer the questions, and they are the documents that most secondary commentary has not read.

Anyone dealing with a live dispute about a local body’s powers, an election that has been postponed, or a reservation notification should be aware that the position varies substantially by state statute and that election disputes are channelled into a statutory election petition with short and strictly enforced limitation periods; the general position described here does not substitute for advice on an individual matter, and a missed limitation period in this field is usually fatal.

For readers building a working file on this subject, the devolution gap table is designed to be extended state by state, and you can keep a private, encrypted set of state-wise devolution notes and finance commission summaries on VaultBook as you work through the underlying documents. Aspirants preparing for the civil services and state services papers should note that this cluster is examined heavily on the mandatory-versus-permissive distinction and on the three Fs rather than on article numbers alone, and can test that distinction against previous-year questions on the ReportMedic explorer before moving on. The wider statutory architecture of representation and inclusion that these reservation provisions sit inside is treated in the guide to social justice legislation.

The shortest statement of the finding, for anyone who needs one line: the Constitution required India to elect local governments and left it to the states to decide whether those governments would govern.

Six common claims about local government, tested against the record

The public argument about panchayats and municipalities runs on a handful of claims that are stated as facts. Each is worth testing, because the accurate version of each is more useful than the slogan.

The claim that panchayats are patronage machines is partly supported and badly framed. Local bodies distribute benefits, and the distribution of benefits generates patronage at every level of government everywhere. What the criticism usually means is that local distribution is captured by dominant social groups, and that is well documented in qualitative research. The framing is bad because it treats capture as an argument against devolution when it is at least as strong an argument for the accountability mechanisms that devolution enables, since a distant department is not less captured, only less visible. The serious version of this criticism concerns the design of grievance and appeal routes rather than the existence of elected local bodies.

The claim that local bodies lack capacity and therefore should not be given money is true as to the first half and does not follow to the second. Capacity is built by exercising responsibility with support, and states that devolved with training, handholding and accounting support built capacity. The claim is used far more often as a reason for postponement than as a programme for building what is missing.

The claim that the Eleventh Schedule transferred twenty-nine subjects to panchayats is simply wrong as a matter of law, and it appears in an enormous quantity of secondary material. The schedule lists subjects that a state legislature may devolve. It transfers nothing.

The claim that central grants to local bodies undermine federalism is contested and deserves both sides. The objection is that a channel running from the Union to bodies constituted under state law weakens the state’s constitutional position and creates a direct fiscal relationship the Constitution did not contemplate. The response is that the constitutional obligation to finance local bodies sits with states which have not discharged it, that the central awards are recommendations of a constitutional commission rather than executive discretion, and that a local body starved by its own state has no other route. This is a genuine federalism disagreement and not a factual dispute.

The claim that the amendments failed is the one this article has spent most of its length qualifying. They achieved what they made mandatory and not what they permitted, and describing that as failure discards the mandatory column entirely.

The claim that a further constitutional amendment is the obvious fix is the one most often stated as though nobody could disagree. The arithmetic alone makes it unlikely, since it would require ratification by the state legislatures whose discretion is being removed, and the design objections about uniformity across states of wildly different local structures are serious. It is a defensible reform position, not a self-evident one, and the alternative positions set out earlier deserve to be argued against rather than ignored.

The first decade: what states did with the transition period

Article 243N gave states a year from the commencement of Part IX to bring existing panchayat laws into conformity with it, failing which those laws would cease to be in force, and Article 243ZF made equivalent provision for municipal law. That transition clause is the most underrated provision in either amendment, because it forced every state to legislate within a fixed period and because the shape of what each state legislated in that window has largely determined its trajectory since.

The conformity legislation of the mid-1990s falls into three broad patterns.

Some states rewrote their panchayat and municipal law substantially, using the occasion to allocate functions, create district planning machinery that met rather than merely existed, and give local bodies a defined share of state revenues. These are largely the states that had operated real local government before 1993, and the conformity exercise deepened an existing arrangement rather than creating a new one.

Some states passed minimal conformity amendments to existing statutes, inserting the mandatory provisions about terms, elections, reservation and commissions into legislation that otherwise continued to treat panchayats as agencies of the state administration. This was the most common approach. It satisfied Article 243N completely, because nothing in Part IX required anything more, and it left the substantive relationship between the local body and the line department untouched.

A third group passed a new statute that listed all or most of the schedule subjects as devolved, and then did not follow it with activity mapping, staff transfer or fund devolution. This produced the paper-transfer position that dominates published devolution statistics, and it is the reason those statistics have to be read against the activity-mapping question rather than at face value.

The first decade also established two habits that have proved difficult to reverse. The first is the practice of routing new programmes through societies and missions rather than through the constitutional bodies. When large national programmes in education, health, water and sanitation were designed in the late 1990s and the 2000s, most were built with their own district and block implementation structures because those structures could be created quickly, staffed on contract and held to central reporting requirements in a way that panchayats and municipalities could not. Each such decision was individually defensible on delivery grounds and collectively hollowed out the schedules.

The second habit is the treatment of the State Finance Commission as a formality. Where a state’s first commission was appointed late, reported late and had its recommendations largely rejected, the pattern tended to repeat across subsequent cycles, because the institutional memory that formed was that the exercise did not change anything. Where a state’s first commission was taken seriously, later commissions had data to work with and a precedent for acceptance. The path dependence in this mechanism is unusually strong.

The recommendations that were made and not adopted

The implementation gap has not gone unexamined. Successive official review exercises have looked at local government and have converged on a remarkably consistent set of prescriptions, and the fact that the prescriptions recur is itself evidence about why nothing changes.

The Second Administrative Reforms Commission devoted a full report to local governance and argued for treating local bodies as the primary unit of governance for the functions assigned to them rather than as one delivery agency among several. Its recommendations included clear delineation of functions across the tiers through activity mapping, placing the personnel who perform devolved functions under the effective control of the elected body, strengthening the finance commission mechanism and the follow-through on its recommendations, giving municipal government a directly elected executive head with a full term, and folding parallel bodies back into the constitutional structure where the function properly belongs to it. The commission on Centre-state relations that reported at the end of the 2000s reached compatible conclusions from the federal side, emphasising the fiscal position of local bodies and the need for states to honour the finance commission mechanism they are obliged to operate.

The consistency of these prescriptions across bodies with different mandates and different political contexts is striking, and so is the near-absence of implementation. That combination points to a conclusion that no report can state as bluntly as an analytical article can: the obstacle is not that the remedy is unknown or contested at a technical level. Activity mapping is not a difficult idea. Placing staff under local appraisal is not a difficult idea. Laying a proper action-taken memorandum is not a difficult idea, and it is already a constitutional obligation. The obstacle is that each of these steps transfers something of value from a state department and a state legislature to a local body, and no mechanism exists to make that transfer happen against the wishes of the party that holds the thing being transferred.

That is why the reform debate keeps returning to the constitutional amendment position despite its arithmetic difficulties. It is the only proposal that addresses the incentive rather than the technique. And it is why the middle position set out earlier, which relies on enforcing the mandatory obligations that already exist, is more promising than it first appears: the district planning committee obligation, the finance commission obligation and the reconstitution obligation are commands, not permissions, and a jurisprudence has already developed around the election obligations showing that commands in Part IX are enforceable when somebody brings them to a court.

The practical implication for anyone working in this field is that litigation over the mandatory provisions has more leverage than advocacy over the permissive ones. A petition seeking to compel the constitution and functioning of a district planning committee, or the laying of an action-taken memorandum that actually addresses recommendations, rests on a constitutional obligation. A petition seeking to compel devolution under Article 243G rests on a permission and will fail.

That asymmetry, which is the same asymmetry the whole article has traced, is the reason the permissive-verb framing is worth carrying beyond this subject. It tells a reader where pressure can be applied and where it cannot, which is more useful than a verdict on whether a reform succeeded.

Reading the record without either illusion

A reader who has followed this far may want a way of holding the mandatory achievements and the permissive failures together without letting either dissolve the other, and the cleanest way is to separate three different questions that the phrase “did it work” runs together.

Did the amendments create what they said they would create? Almost entirely, yes. The bodies exist, the elections happen, the commissions are appointed, the seats are reserved. Measured against its own mandatory text, Part IX has one of the better compliance records of any constitutional obligation in India.

Did the amendments produce local self-government? Mostly not, because self-government means deciding and the deciding was left elsewhere. Measured against the phrase “institutions of self-government” that appears in Article 243G itself, the record is poor in most states and good in a few.

Did the amendments make local self-government possible? Yes, and this is the answer that matters most for anyone thinking about reform. Every state that decided to devolve did so under this text, without waiting for an amendment, using instruments that were available to all of them. That fact simultaneously indicts the states that did not and complicates the argument that the text must be changed before anything can improve.

Those three answers are all true at once, and the disagreement in public debate is very largely a disagreement about which of the three questions is being asked. Naming the question first turns an unproductive argument about success and failure into a productive one about what should happen next.

The habit of asking which question is being asked also disciplines the language people use about this subject. “Devolution has failed” is a claim about the second question presented as though it answered the first and third. “India has the largest experiment in local democracy in the world” is a claim about the first presented as though it answered the second. Both statements are true of what they describe and misleading about everything else, and a reader who separates the three questions will stop being persuaded by either.

Frequently Asked Questions

Q: Have the 73rd and 74th Amendments actually devolved power to local bodies?

Partly, and very unevenly. They devolved the right to elect a local government, which is now universal and enforceable, and they changed the composition of local bodies through reservation. They did not devolve the power to decide, because Article 243G and Article 243W say a state legislature may endow local bodies with powers rather than that it must, and no provision addresses control over staff at all. A small number of states have devolved substantially by their own choice, converting the permission into a real transfer of untied funds and mapped activities. Most have transferred subjects on paper while line departments, state authorities and programme societies continue to perform the work.

Q: What are the three Fs of devolution?

Funds, functions and functionaries, meaning money the local body can allocate, tasks it is responsible for deciding, and staff who answer to it. The shorthand is useful because the three can be separated: a body may have a function and no money, or money and no staff, and each combination paralyses it differently. In the Indian constitutional scheme, functions are addressed by a permissive provision, funds partly by a permissive provision and partly by a finance commission whose recommendations do not bind, and functionaries are not addressed at all. A fourth condition, discretion over how assigned money is used, is not part of the standard vocabulary but determines whether a body is governing or administering.

Q: Why is Article 243G considered the weak link in panchayati raj?

Because it uses a permissive verb where the rest of Part IX uses commands. It provides that the legislature of a state may by law endow panchayats with such powers and authority as may be necessary to enable them to function as institutions of self-government, together with responsibility for preparing plans and implementing schemes for economic development and social justice, including in relation to the matters listed in the Eleventh Schedule. Because the choice sits with a state legislature whose members’ influence depends on line departments retaining control, the incentive runs against devolution, and a state that devolves nothing breaches nothing. Article 243W does the same for municipalities.

Q: Do State Finance Commission reports get acted on?

Frequently not in any substantial way. The constitutional obligation is to constitute a commission every five years, to lay its report before the state legislature, and to lay a memorandum explaining the action taken. It is not an obligation to accept the recommendations. Four failure modes recur: late constitution, delayed reports caused partly by the absence of usable local body accounts, partial or nil acceptance of the recommendations that cost the state money, and action-taken memoranda that are laid late or drafted so generally that they do not engage with individual recommendations. The design mirrored the Article 280 mechanism for Union to state transfers without giving it comparable practical weight.

Q: Why are Indian mayors weaker than panchayat presidents in some states?

Because the municipal executive is a state appointee and the mayoralty is often structurally weakened in ways a panchayat chairperson’s office is not. In most states the commissioner who runs a municipal corporation is appointed by the state government, reports to a state department, and cannot be appointed, directed or removed by the elected council. Many states also elect mayors indirectly through councillors and give them terms much shorter than the council’s five years, sometimes rotating the post among reserved categories within a single term. In states that have devolved seriously to rural bodies, a sarpanch serving a full five-year term may control more real decisions than the mayor of a large city.

Q: What are parallel bodies and how do they bypass panchayats?

They are institutions created outside Part IX and Part IXA that perform functions listed in the Eleventh or Twelfth Schedule, including district rural development agencies, registered societies and missions set up to deliver national programmes, urban development authorities, water and sewerage boards, and project-specific vehicles. They typically have their own funds, staff and reporting lines to a state department or to the Union, and are usually headed by an officer rather than an elected representative. Because the Constitution says nothing about them, a state can list a subject as devolved while a parallel body continues to deliver it, leaving the elected body with the label and none of the authority.

Q: Are the Eleventh and Twelfth Schedule subjects binding on states?

No, and this is the single most widespread error in writing on the subject. Unlike the Seventh Schedule, which distributes legislative competence between the Union and the states and operates by its own force, the Eleventh and Twelfth Schedules are lists of subjects that a state legislature may choose to devolve under Article 243G and Article 243W respectively. Neither schedule transfers anything. A state that devolves none of the listed subjects breaches no constitutional obligation, and a state that lists all of them in a conformity Act without allocating discrete activities has changed nothing on the ground. The resemblance in form to the Seventh Schedule misleads many readers.

Q: Has reservation for women in local bodies changed who governs?

It changed who holds office immediately and who exercises authority gradually. The requirement of not less than one-third of seats and chairperson posts, raised to half by a large number of states, placed a very large number of women in elected local office where almost none had held it before. Exercise of that office by a husband or other male relative is common, particularly in first terms and in areas with strong social restriction, and the phenomenon is well enough established to have generated its own vocabulary. It weakens with experience, training and repeat candidacy, and women who have served a term have gone on to contest unreserved seats and higher tiers, so the effect accumulates across electoral cycles rather than appearing at once.

Q: Why did the 73rd and 74th Amendments come into force on different dates?

They are separate Acts dealing with separate Parts of the Constitution, and each was brought into force by its own commencement provision. The Constitution (Seventy-third Amendment) Act, 1992, which inserted Part IX and the Eleventh Schedule for panchayats, came into force on 24 April 1993. The Constitution (Seventy-fourth Amendment) Act, 1992, which inserted Part IXA and the Twelfth Schedule for municipalities, came into force on 1 June 1993. Both were passed by Parliament in December 1992 and received presidential assent on 20 April 1993 after ratification by more than half the state legislatures. Conflating the two commencement dates is one of the most common factual slips in secondary accounts.

Q: Can a state government postpone panchayat or municipal elections?

Not lawfully on grounds of administrative convenience. Article 243E requires that an election to constitute a panchayat be completed before the expiry of its five-year term, and that a body dissolved early be reconstituted within six months, with Article 243U making equivalent provision for municipalities. The Supreme Court has held that these obligations are not conditional on the state completing exercises it controls, such as ward delimitation, and that the State Election Commission occupies a position analogous to the Election Commission of India within its sphere. Delays still occur, most often through disputes about reservation for backward classes, but a state that wishes to delay must construct a legal obstacle and will usually be required to justify it in court.

Q: Why do reported devolution figures overstate what panchayats control?

Because the figures usually count subjects listed as transferred rather than tasks actually assigned. A subject such as primary education contains recruitment, transfer, building works, maintenance sanction, procurement, monitoring and meal delivery, and each of those can sit at a different level independently of the others. Transferring a subject without mapping those activities to a tier, with matching funds and staff, changes nothing, because the line department continues to perform every task. Published devolution counts and index scores therefore reward formal transfer and cannot distinguish it from real transfer, which is why the presence or absence of a published activity map is a better indicator than any reported number.

Q: Do panchayats have the power to levy taxes?

Only to the extent a state legislature authorises. Article 243H permits a state legislature to authorise a panchayat to levy, collect and appropriate taxes, duties, tolls and fees, to assign to it taxes collected by the state, and to provide grants-in-aid, and Article 243X makes equivalent provision for municipalities. Every verb is permissive. In practice the base is narrow, typically covering taxes on buildings, professions, markets, fairs, water and lighting, often at rates fixed or capped by the state. Own-source revenue therefore funds a small share of expenditure in most panchayats. Property tax is the principal municipal source and is constrained by valuation practice, exemptions and state control over rates.

Q: Why do District Planning Committees rarely function?

Because the obligation to constitute them carries no consequence for their not working. Article 243ZD requires every district to have a committee, with at least four-fifths of its members elected from among the elected members of the district panchayat and the municipalities in the district, to consolidate local plans into a draft district development plan. Most states have issued notifications constituting these committees. In many, the committees meet rarely, receive no plans from the constituent bodies, and produce either no consolidated plan or one unrelated to the budgets actually spent. Because nothing follows from that, formal compliance has substituted for the function, and the same is true of metropolitan planning committees under Article 243ZE.

Q: Does Part IX apply to Scheduled Areas and the north-eastern states?

Not automatically. Article 243M excludes the Scheduled Areas and tribal areas referred to in Article 244, the states of Nagaland, Meghalaya and Mizoram, the hill areas of Manipur for which district councils exist, and certain other specified areas, with comparable exclusions on the municipal side. Parliament extended Part IX to the Scheduled Areas by a separate statute in 1996 with substantial modifications designed to protect customary practice and to give the gram sabha powers over land, minor forest produce and consultation before acquisition that ordinary panchayats do not possess. The autonomous district councils in several north-eastern states operate under a different constitutional scheme entirely.

Q: Is the one-third reservation for women still the rule?

One-third is the constitutional floor, not the position in most of the country. Article 243D requires that not less than one-third of the total seats, and not less than one-third of the chairperson posts, be reserved for women, and Article 243T does the same for municipalities. Because the provision sets a minimum rather than a ceiling, states are free to legislate higher, and a large number have done so at half. Bihar was the first state to legislate half the seats for women in panchayats and many others followed. Any account that presents one-third as the operative figure across India is describing the constitutional minimum rather than the applicable state law.

Q: Who controls the staff working for a panchayat or municipality?

In most states, the state government does, and this is the largest silence in both amendments. Neither Part IX nor Part IXA contains any provision about staff: nothing about deputation, nothing about control, and nothing about who writes an officer’s performance appraisal. The officers who deliver devolved functions are therefore usually employees of state line departments whose recruitment, promotion, transfer and appraisal are controlled by those departments. An elected local body can request action from such an officer but cannot make its instruction consequential for the officer’s career, which is why decisions of local bodies frequently function as requests rather than directions.

Q: Would making Article 243G mandatory fix devolution?

It would remove the principal legal obstacle and create three new problems. Removing the discretion would require a constitutional amendment ratified by at least half the state legislatures, and those are the bodies whose discretion is being removed, so the arithmetic is difficult. A centrally drafted mandatory list would flatten genuine variation between states whose local bodies differ enormously in size and number, potentially assigning functions to bodies too small to perform them. And there is an objection of principle that the Union prescribing what states must give their own local bodies is a further centralisation. The reform is defensible but is not self-evident, and the fiscal and administrative alternatives deserve to be argued against rather than dismissed.

Q: Which states have devolved most power to panchayats, and why?

Three patterns explain most of the variation. States that had functioning local government before 1993, including those that had operated district bodies with real budgets since the 1960s or legislated ambitious panchayat systems in the 1980s, generally continued to have it afterwards, because the amendments consolidated something already present. States that made a deliberate fiscal decision to route a large share of plan funds to local bodies as untied money, accompanied by participatory planning and substantial training, produced real devolution under the same constitutional text everyone else has. And states that carried out genuine activity mapping produced measurable transfer, while those that transferred subjects without mapping produced none. Prior commitment, untied money and activity mapping are the three predictors.