For most of the second half of the twentieth century, electricity law in India meant three statutes working as one regime, and the centre of that regime was not a market, a contract or a regulator but a statutory corporation: the State Electricity Board. The Board generated the power, carried it across the state, sold it to the consumer, fixed the price at which it was sold, borrowed the money with which the system was built, and answered for all of it to the government that appointed its members. Understanding Indian electricity, including the financial condition of distribution that every later reform has inherited, begins with understanding what those three statutes asked of that single institution, because the statutes asked for two things that could not be done at once.

The three statutes were the Indian Electricity Act, 1910, the Electricity (Supply) Act, 1948 and the Electricity Regulatory Commissions Act, 1998. The first was a works and safety statute from the colonial period, concerned with licences, lines and the protection of the public. The second, passed in the first year of the Republic’s approach, built the public supply system and wrote into law both a commercial obligation and a duty of obedience to government direction. The third, arriving half a century later, began the separation of the regulator from the operator that the first two had fused together. In 2003 all three were repealed by a single section of a consolidating Act, and the manner of their repeal, which preserved far more than it destroyed, is part of the story. This article sets out the regime the three statutes made, explains the contradiction at its centre with the attribution the evidence requires, and states precisely what survived the repeal. The successor statute has its own pillar in this series, the Electricity Act 2003 explained, and the field around both is mapped in the infrastructure laws guide; the questions taken here are the predecessor regime’s own.

Three statutes, one regime

The layering of the three Acts was a product of the constitutional place of electricity. Electricity is Entry 38 of the Concurrent List, so Parliament and the state legislatures could both legislate, and the statutory structure that grew up divided the work accordingly: a central Act of general application for the physical and licensing framework, a central Act that organised the public supply industry through state-level institutions, and state legislation filling the spaces the central Acts left. No single statute contained the whole law of electricity at any point in the regime’s life. A practitioner advising a licensee in, say, 1975 worked simultaneously with the licensing and works provisions of the 1910 Act, the institutional and financial provisions of the 1948 Act, the financial principles for licensees set out in the Sixth Schedule to the 1948 Act, and the rules made under both.

That division explains the feature of the regime that most often misleads later readers. The 1910 Act is sometimes described as though it were the economic constitution of the early industry, and the 1948 Act as though it merely administered a system the earlier Act designed. The truth runs the other way. The 1910 Act contained almost no economic regulation at all. It decided who might supply energy in an area and on what physical conditions, how works could be laid, what duties were owed to the public and to the telegraph administration, and what offences attached to interference with lines and meters. The price of electricity, the return a supplier might earn, the obligation to supply, and the planning of generation were all governed elsewhere: for private licensees, principally by the Sixth Schedule to the 1948 Act and the conditions of their licences; for the Boards, by the financial chapter of the 1948 Act itself. When the 2003 Act repealed all three statutes together, it was acknowledging what practice had always known, that they formed one regime with three layers and could only be replaced as a set.

The regime also had a direction of travel built into it. The 1910 Act assumed a world of private and municipal licensees supplying defined areas. The 1948 Act assumed a world in which the state, acting through a Board in each state, would progressively take over the industry, and it gave the Board the legal instruments to do so, including the vesting in the Board of rights to purchase licensees’ undertakings. The 1998 Act assumed, against the premise of 1948, that the operator should no longer judge its own prices. Each layer therefore qualified the one beneath it, and the repeals of 2003 were the final qualification: the licensing world of 1910, the Board world of 1948 and the transitional regulator of 1998 were replaced by a single statute that treated generation, transmission, distribution and trading as separable licensed or regulated activities.

The 1910 Act: a works and safety statute

The Indian Electricity Act, 1910 (Act 9 of 1910) was the second Indian statute on the subject, consolidating and replacing an Act of 1903, and it remained the licensing backbone of the industry for ninety-three years. Its scheme was simple. The state government granted licences authorising a person to supply energy within a specified area. The licensee acquired, in exchange, a set of statutory powers and duties: powers to break up streets and lay lines, duties to supply on request within the area, duties of safety toward the public, and obligations toward the owners of telegraph lines whose circuits the licensee’s works might disturb. The Act’s middle sections, sections 12 to 18, dealt with the licensee’s works, the opening of streets and railways, notice to telegraph authorities and the settlement of the physical conflicts that arise when one network must cross another. Its later parts created offences, including the dishonest abstraction of energy, and machinery for inspection and inquiry.

What did the Indian Electricity Act, 1910 leave outside its reach?

The 1910 Act left the economics of supply almost wholly outside its reach. It did not fix tariffs, prescribe a rate of return, plan generation, or create any institution charged with the industry’s finances. Those matters sat in licence conditions, in the Sixth Schedule to the 1948 Act for licensees, and in the 1948 Act’s financial chapter for the Boards.

What the 1910 Act regulated, it regulated in the idiom of the nineteenth-century public utilities statutes from which it descended: the idiom of wayleaves, notice, compensation and safety. A licensee’s right to place works in a street, the protection of pipes and telegraph lines, the duty to restore surfaces, the compensation payable for damage, the penalties for theft of energy and for interference with meters, the powers of the Electric Inspector: these were its subjects. It is a measure of the Act’s character that when the 2003 Act repealed it, two of its components were thought too useful to let lapse even for a day. Sections 12 to 18, the works provisions, were kept in force until rules under the new Act’s corresponding sections could be made, and the Indian Electricity Rules, 1956, made under section 37 of the 1910 Act and containing the detailed safety code for installations, were continued until regulations under the new Act replaced them. A statute remembered for its age was, in its operative parts, a safety code with a licensing front end.

The licensees who operated under the 1910 Act were a mixed population. There were companies supplying the great cities, of which the Calcutta Electric Supply Corporation, supplying Calcutta under licences reaching back to the end of the nineteenth century, and the Tata enterprises supplying Bombay were the largest; there were municipal undertakings, of which the Bombay Electric Supply and Transport undertaking was the best known; and there were small town licensees of every degree of competence. Their commercial terms were not set by the 1910 Act but by the Sixth Schedule to the 1948 Act, which prescribed how a licensee’s capital base was to be computed, what counted as clear profit, and by how much clear profit might exceed a reasonable return before the excess had to be shared with consumers or carried to reserve. The result was a regulated private sector of a distinctive kind: entry by licence, conduct governed by a works code, and profit disciplined by a statutory formula, all supervised not by a specialised regulator but by the state government and its inspectors.

The industry the 1948 Act inherited

To see why Parliament legislated as it did in 1948, the industry of that moment has to be pictured accurately. Electricity supply in India at Independence was small, urban and fragmented. Generating capacity for the whole country stood at about 1,700 megawatts. Supply undertakings numbered in the hundreds, most of them tiny: licensees serving a single town, municipal departments running a powerhouse for street lighting and a few thousand connections, and a handful of substantial enterprises in the presidency cities. Alongside the licensees stood the works of governments themselves: the hydroelectric stations built by provincial governments, of which the Pykara project in Madras Presidency, commissioned in the early nineteen thirties, and the Mettur project that followed it were the notable examples, and the electricity departments through which several provinces and princely states supplied their capitals. There was no grid in any modern sense; systems were islanded, frequencies and voltages varied, and the industry’s growth was constrained less by demand than by the capital a licensee could raise against a regulated return in a limited area for a limited term.

The case for reorganisation, as the new Union government saw it, had three strands. The first was scale: power stations and transmission lines of economic size crossed the boundaries of licence areas and often of provinces, and a licensing map drawn for town lighting could not organise a national programme of river valley development and industrial supply. The second was capital: the investment the plans contemplated could not be raised by small licensees on Sixth Schedule returns, and the state would have to borrow, which meant the state would have to own. The third was uniformity: a national policy for the development of electricity, covering tariffs, grid standards and the orderly absorption of existing undertakings, required institutions that did not exist. The British parallel stood close at hand and was consciously used: the United Kingdom had nationalised its own fragmented industry by the Electricity Act, 1947, creating a central authority and area boards that took over generation and distribution from companies and municipalities alike. India’s 1948 Act adapted that template to a federal constitution, replacing the single nationalised industry with a central authority for coordination and a board in each state as the owner-operator, a difference that flowed from electricity’s place in the Concurrent List and from the political fact that the states, not the Centre, were to run supply. The adaptation mattered for everything that followed: a British area board answered to one national authority and one minister, whereas an Indian Board answered to a state government whose electoral incentives on tariff were immediate and local.

The inherited licensees and departments were not swept away on commencement. The 1948 Act worked by superimposition: the Board was constituted, the existing undertakings continued, and the transfer of the industry to the Boards proceeded by purchase, by expiry and by the Board’s own extension into unserved territory. In the early years some states were slow to constitute Boards at all, and government electricity departments continued to operate supply in several states well into the nineteen fifties. The regime’s mature form, in which the Board was the industry in most states, was the product of the nineteen fifties and nineteen sixties rather than of the commencement date, and its last pieces, in the states that retained departmental supply longest, were still being assembled when the financial difficulties described below were already visible.

The Board at work: planning, bulk supply and the grid

The Board in operation was a planner as much as a utility. Load forecasts were prepared state by state, generating schemes were framed and sanctioned through the Central Electricity Authority’s appraisal, and the state’s five-year plan allocation for power, typically the largest single head of plan expenditure, was in substance the Board’s investment programme financed by state loans, market borrowing guaranteed by the state, and whatever internal surplus section 59 produced. The Board sold electricity in bulk to the licensees remaining in its territory and to large industrial consumers connected at high voltage, and at retail to everyone else. Its engineers constituted one of the largest technical cadres in state government service, and its chairman and members, appointed by the state government, were drawn heavily from the engineering services and the civil service, a staffing pattern that shaped the institution’s culture: strong on project execution, hierarchical in management, and oriented toward the government of the day as the source of both capital and instruction.

Two developments of the middle decades changed the Board’s environment without changing its statute. The first was interconnection. As state systems grew, regional grids linked them, and Regional Electricity Boards were constituted in 1964 by government resolution to coordinate the operation of the state systems within each region and to advise on the scheduling of generation and the sharing of surpluses. They were advisory bodies without statutory ownership, and their existence testified to a gap in the 1948 scheme, which had organised the industry state by state and had no institution, beyond the Central Electricity Authority’s coordinating role, for the regional operation that physical interconnection made necessary. The second development was the entry of the Centre as a generator. The Boards’ investment capacity, dependent on state finances and their own elusive surpluses, fell behind the growth of demand through the nineteen seventies, and the Union government created central generating companies, the National Thermal Power Corporation and the National Hydroelectric Power Corporation, both in 1975, to build large stations whose output was allocated among the states of a region and sold to the Boards. The central sector’s share of generation grew steadily thereafter. The arrangement relieved the capacity constraint and created a new financial nexus: the Boards became the central stations’ principal debtors, and unpaid dues to the central generating companies became, by the nineteen nineties, one of the standard measures of Board distress and eventually the subject of a one-time settlement under which states took over and securitised a large part of the accumulated dues on the recommendation of an expert group chaired by Montek Singh Ahluwalia that reported in 2001.

The operational record across these decades was therefore mixed in a specific pattern. The system grew at rates few developing countries matched, and the lights, in the cities and in an expanding share of villages, stayed on for most purposes most of the time. But shortage was chronic: demand exceeded supply in nearly every year, load shedding was an administrative routine rather than an emergency, and the agricultural supply that governments promised was rationed by hours, delivered in rotating blocks, often at night, at voltages that damaged the pumpsets it served. The Board administered scarcity as its core function, and the politics of scarcity, who was cut, for how long, and at whose complaint relief came, ran through the same channel as the politics of tariff: the government that directed the Board. An institution designed in 1948 to develop abundance had become, by the nineteen eighties, an instrument for distributing shortage under political direction, and its accounts recorded the cost of both mandates.

The 1948 Act and the invention of the Board

The Electricity (Supply) Act, 1948 (Act 54 of 1948) changed the industry’s ownership premise. Its long title described an Act to provide for the rationalisation of the production and supply of electricity, and its institutional scheme had three tiers. At the top, section 3 provided for the constitution of a Central Electricity Authority, the planning and coordinating body for the sector as a whole, charged among other things with developing a sound, adequate and uniform national power policy and with arbitrating in matters the Act referred to it. In each state, section 5 imposed a duty on the state government: the State Government shall, as soon as may be after the Act’s commencement notification, constitute by notification in the Official Gazette a State Electricity Board under such name as the notification specified. The Board was to consist of not less than three and not more than seven members appointed by the state government, and the Act prescribed the expertise the membership had to include: a person of commercial and administrative capacity, an electrical engineer of wide experience, and a person experienced in accounting and financial matters in a public utility. Section 12 made the Board a body corporate with perpetual succession, able to hold property, to contract and to sue in its notified name.

The Board’s functions swallowed the industry. Under the 1948 Act’s scheme the Board was charged with promoting the co-ordinated development of the generation, supply and distribution of electricity within the state, with arranging the supply of electricity throughout the state, with constructing and operating generating stations and transmission lines, and with supplying electricity to licensees and consumers. Existing licensees did not vanish; they continued in their areas, and the Board supplied many of them in bulk. But the Act’s expectations ran against them. Section 71 provided that the rights and options of purchase of a licensee’s undertaking conferred by the 1910 Act should vest in the Board rather than in the local authority or the state government, so that when a licence reached the point at which the undertaking could be compulsorily purchased, it was the Board that stood ready to absorb it. Over the following decades, as licences expired or were revoked or purchased, supply areas passed to the Boards, and in most states the Board became in substance the state’s electricity industry: generator, transmission operator, distributor and, through its bulk supply, the supplier behind most remaining licensees as well.

Between the state government and the Board the Act built a relationship of a particular and fateful kind. The Board was autonomous in form: a statutory corporation with its own funds, its own borrowing powers and its own accounts, audited and published, with an annual financial statement to be laid before the state legislature. The state government appointed the members, could remove them, provided the Board’s capital and guaranteed its loans. And by section 78A, inserted into the Act by the amending Act 101 of 1956, the relationship was given a legal channel that ran one way. In the discharge of its functions, the Board was to be guided by such directions on questions of policy as the state government might give it; and if a dispute arose between the Board and the government whether a question was or was not a question of policy, the question went to the Central Electricity Authority, whose decision was final. The provision did not make the Board a department. It did something subtler: it preserved the Board’s corporate separateness for purposes of liability, employment and accounting, while subordinating its judgment on every matter that could be characterised as policy, and tariff was the most consequential question of policy there was.

The Central Electricity Authority’s place in this structure deserves a sentence of its own, because later readers often mistake it for a regulator. It was not one. It was a planning and technical authority: it collected statistics, it advised governments, it carried out the functions the Act assigned in respect of schemes and arbitration, and after 1956 it held the casting vote on what counted as policy between a Board and its government. It determined no tariffs and licensed no one. The regime of the 1948 Act had no regulator because, on its own premises, it needed none: the operator was a public body, its prices were set under statutory financial principles, and its owner was the state. The absence reads as a gap only from the far side of 1998.

The supporting institutions of the 1948 Act

Around the Board the 1948 Act placed a set of lesser institutions whose fortunes track the regime’s assumptions. The State Electricity Consultative Council, provided for by section 16, was to advise the Board on major questions of policy and schemes, and Local Advisory Committees under section 17 were to represent consumer and commercial interests in particular areas. Both were advisory, both were appointed, and neither developed into a channel through which consumers could contest the decisions that governed them; they belonged to a conception of consultation in which the governed are heard by courtesy of the governor, and their weakness is the institutional counterpart of the consumer position described later in this article. The generating company, given statutory form by section 15A as inserted by amendment in the nineteen seventies, was a more consequential innovation: it allowed the Central Government, a state government, or governments jointly to own a company whose object was to establish and operate generating stations and sell the power to the Boards. The central generating companies of 1975 and the state generation corporations that some states later formed were built on this device, and with it the Board ceased, in those states and at the margin everywhere, to be the sole public generator its designers had imagined. The Board’s capital structure received its own provision in section 12A, under which a state government could notify that the Board would have a capital structure, converting the state’s accumulated loans and grants into something resembling equity, an attempt to give the Board’s balance sheet the shape of the corporation the Act said it was.

Between the Board and the licensees, and between the Board and its consumers, the Act’s dispute machinery was thin: arbitration under section 76 for the matters the Act referred to it, the Central Electricity Authority’s decisions on referred questions, and the ordinary courts. There was no forum in which the pattern of a Board’s conduct could be reviewed as a pattern. Each failure of the regime therefore presented itself to the law as an isolated grievance, an employee’s promotion, a consumer’s bill, a licensee’s bulk supply rate, a generator’s unpaid dues, and was decided as such, correctly on its own terms and without effect on the structure that produced it. It is a recurring observation in this series that a legal system gets the litigation its institutions make possible; the 1948 Act made possible every kind of electricity litigation except the kind that mattered, a challenge to the allocation of functions itself, and that challenge, when it came, came through politics and legislation rather than through the courts.

The financial constitution of the Board

Chapter VI of the 1948 Act, headed as the Board’s finance, accounts and audit, was the closest the regime came to a commercial code for the Boards, and its centre was section 59. In the form the section took for most of the Boards’ lives, after its substitution by the Electricity (Supply) Amendment Act, 1978 (Act 23 of 1978), section 59(1) required the Board, after taking credit for any subvention from the state government under section 63, to carry on its operations and adjust its tariffs so as to ensure that its total revenues in any year of account, after meeting all expenses properly chargeable to revenue, including operating, maintenance and management expenses, taxes, depreciation and interest payable on all debentures, bonds and loans, left a surplus of not less than three per cent, or such higher percentage as the state government might specify by notification, of the value of the fixed assets of the Board in service at the beginning of that year. An explanation defined that value: the original cost of the fixed assets in service, reduced by cumulative depreciation and by consumers’ contributions for service lines.

How was the three per cent figure in Section 59 worked out?

The percentage applied to the value of the Board’s fixed assets in service at the start of the year of account, computed as original cost less cumulative depreciation and less consumers’ contributions for service lines. The surplus had to remain after operating expenses, taxes, depreciation and interest had all been met, and any subvention received from the state government under section 63 was credited before the test was applied.

The architecture around section 59 shows what the surplus was for. The Board’s expansion was to be financed substantially from its own resources and from borrowing raised against them: section 65 empowered the Board to borrow, section 66 authorised the state government to guarantee repayment, and section 64 provided for loans from the state government itself, while section 63 authorised the state government to make subventions to the Board, grants intended to bridge the gap between what tariffs yielded and what the statutory duties cost. A Board earning its three per cent generated an internal surplus that funded part of each year’s capital works and demonstrated the creditworthiness on which the rest depended. A Board that did not was not merely untidy; it was consuming the financing mechanism of the next decade’s investment. The financial principles were not decoration. They were the load-bearing assumption of a system in which the state built power stations and networks on a scale no annual budget could have carried, financed by an institution whose tariffs were supposed to service the debt.

Two features of the drafting repay attention, because both mattered later. The first is that section 59 directed the Board to adjust its tariffs to achieve the surplus. Tariff-setting was thus not, in the statute’s design, a political gift to be distributed; it was the instrument by which a statutory financial target was to be met, and the Board held the instrument. Section 49 of the Act, under which the Board fixed the terms on which it sold electricity to persons other than licensees, was the operative power, and the Sixth Schedule performed the same office for licensees. The second feature is the role reserved to the state government inside the financial chapter itself: the government could specify a higher percentage than three, it was the source of subventions and of loan capital, and under section 67A the interest on its own loans ranked behind the Board’s other expenses. The state was simultaneously the Board’s owner, banker, guarantor and, through section 78A, its policy master. Section 59 addressed its command to the Board, but nearly every lever that determined whether the command could be obeyed sat in the government’s hands.

The direction power and the impossible pair

Set section 59 beside section 78A and the regime’s central defect stands in the open. The same statute that required every Board to leave a surplus of not less than three per cent on its fixed assets also required every Board to be guided by its state government’s directions on questions of policy. Tariff levels for the largest classes of consumers were, in every state, treated as questions of policy, and the policy of successive governments, of every political colour, was that electricity for agriculture and for domestic consumers should be cheap: below the cost of supply, and in the extreme cases free. A direction, formal or understood, to hold farm tariffs at a fraction of cost made the section 59 surplus arithmetically unreachable unless other consumers paid far more than cost or the government paid the difference as subvention. The governments that issued the directions controlled the subvention too, and the subventions voted in state budgets were, year after year, smaller than the gap the directions had opened, or were not released in full, or were treated in the accounts as loans rather than grants, adding interest to the very expenses section 59 required the surplus to survive.

Could a Board refuse a direction from its State Government?

Within the statute, in substance no. Section 78A obliged the Board to be guided by directions on questions of policy, and the only appeal the Act offered was a reference to the Central Electricity Authority on the anterior question whether the matter was one of policy at all. A Board could argue characterisation; it could not disobey a direction that survived the argument.

The consequences are a matter of record rather than conjecture, provided the record is cited for what it is. The Planning Commission published an annual report on the working of the State Electricity Boards and Electricity Departments throughout this period, and its successive volumes chart the same progression: the three per cent return was earned, in the Boards’ early decades, by some Boards in some years, and then by almost none. The Observer Research Foundation’s retrospective account of the sector records that the Boards began to falter financially in the late nineteen sixties and to depend on subsidies and handouts from their governments, and that by the nineteen seventies their liabilities had grown into a standing feature of state finances. The National Development Council’s committee on power, appointed in June 1993 under the chairmanship of Sharad Pawar, reported in September 1994 with recommendations that read as a diagnosis of the statutory defect: restore to the Boards the autonomy the statute envisaged, make tariff policies transparent, and let the Boards function on commercial lines with the ability to mobilise their own resources. A committee of chief ministers does not recommend restoring a statutory autonomy unless that autonomy has been lost in practice to the very governments the statute made the Boards obey.

It is worth pausing on the word “policy”, because the legal and the political defects were the same defect seen from two sides. A direction on policy, in a well-ordered version of the 1948 scheme, would settle the ends: universal supply, priority to irrigation, protection of the poor consumer. The Board would then choose the means, including the tariff structure, and if the ends cost more than the tariffs raised, the state would fund the difference openly as subvention, and the size of the subsidy would appear in the state’s budget where legislatures could see it. What happened instead was that the direction settled the means as well, fixing the tariff itself, while the funding decision was deferred, disputed or disguised. The subsidy therefore never appeared as a subsidy. It appeared, at the end of each year, as a Board loss, and was read by commentators as evidence of the Board’s inefficiency, when a large and unmeasurable part of it was the price of a policy the Board had been directed to execute and the state had declined to fund. The statute had built a machine for converting political decisions into accounting losses and attributing them to the wrong institution.

The subsidy that could not be measured

One practice made the defect impossible even to quantify, and it deserves separate treatment because it explains why arguments about the scale of the farm power subsidy ran for decades without resolution. Agricultural supply in most states was unmetered. The pumpset consumer was billed, where billed at all, at a flat rate related to the horsepower of the pump, and the energy actually consumed was not measured at the point of use. A Board’s accounts therefore had to estimate agricultural consumption, and the estimate was derived as a residual: total generation, less measured sales to metered consumers, less an allowance for transmission and distribution losses, was booked as farm consumption. Every error in the system collected in that residual. Losses in the network, theft of energy, unbilled and unmetered urban consumption, and genuine farm use were all inside the same number, and by moving the assumed loss percentage a Board could make agricultural consumption, and with it the apparent size of the farm subsidy, larger or smaller within a wide band.

The measurement failure had three consequences that outlasted the regime. First, it protected the subsidy politically: a benefit whose cost cannot be stated cannot easily be debated, and governments that might have had to defend a figure were spared the necessity. Second, it corrupted the loss figures on which all reform programmes depended, because transmission and distribution loss, the standard measure of a network’s technical health and its exposure to theft, could not be separated from the consumption estimate; the category later formalised as aggregate technical and commercial loss was an attempt to rebuild a measurement the old regime had never made. The Ministry of Power’s published figures put aggregate technical and commercial losses at 36.64 per cent for the year 2002-03, a figure that captures in one number the accumulated measurement and collection failure of the system at the moment of its replacement. Third, unmetered supply destroyed the price signal at the point of use: a farmer paying a flat charge per horsepower faced no marginal cost for an additional hour of pumping, with consequences for groundwater extraction that belonged to water policy but were manufactured by electricity law’s tolerance of unmetered supply.

The cross-subsidy that was supposed to close the gap had its own limit, and the Boards reached it. As farm and domestic tariffs were held down, industrial and commercial tariffs were pushed up, in state after state, to levels well above the cost of supply. Industry responded rationally: large consumers built captive generating plants, removing the most creditworthy load from the Board’s system and leaving the Board with a consumer mix progressively weighted toward those who paid least. Commentators associated with the Delhi Science Forum made this point at the time and made it sharply, arguing that high-cost power imposed on industry, the mainstay of Board finances, was driving that mainstay off the grid and deepening the crisis it was meant to relieve. The statutory design had no answer to the spiral, because each element of it, the directed tariff, the unfunded shortfall, the industrial surcharge and the captive exit, was a lawful product of the scheme itself.

Anatomy of a Board loss

A Board’s annual loss, the figure around which so much argument circled, repays decomposition, because its components had different authors and different remedies. Begin with the revenue side. The average realisation per unit sold was set by the tariff schedule, and the schedule’s structure was political in the way described: agriculture at a fraction of cost or free, domestic consumers below cost, industry and commerce above cost. The mix of sales then determined the average, and the mix worsened steadily for the Board, because farm consumption, the lowest-yielding category, was also the fastest growing, while industrial sales, the highest-yielding, leaked to captive plants. Costs, on the other side, were largely outside the Board’s control in the short run: fuel prices and freight for coal, the interest on accumulated borrowing, wages settled in pay commissions whose awards the Board applied as a state instrumentality, and the price of power purchased from central stations at tariffs fixed by the Centre. Between a revenue line the government compressed and a cost line the Board could not compress stood the commercial loss, and section 59’s three per cent surplus was the first casualty, followed by maintenance expenditure, followed by the deferral of the investment programme, each deferral raising future costs and future losses.

Accounting practice then determined how much of the truth the loss figure told. Where a state government announced a subsidy for free or cheap farm power, the amount was frequently booked by the Board as receivable from the government, and where the money did not arrive the receivable accumulated as an asset of doubtful value, flattering the accounts while the cash position rotted. Regulatory practice under the later commissions would name a version of this the regulatory asset and quarrel over its legitimacy; under the Boards it had no name and no quarrel, because no institution outside the Board and its owner examined the treatment. The Planning Commission’s annual reports, working through the accounts, repeatedly drew the distinction between losses with subsidy counted and losses without it, and the second figure, the honest one, was always the larger. By the end of the century the aggregate commercial losses of the Boards, measured without subvention, were of an order that finance commissions treated as a first charge on the states’ fiscal health, and the accumulated position had begun to threaten the credit of the sector as a whole: banks and financial institutions that had lent to Boards on state guarantees grew unwilling to extend the exposure, and the central generating companies’ unpaid bills, discussed above, were settled by state borrowing precisely because the operating companies could no longer carry them.

One further component of the loss belonged neither to tariff politics nor to accounting but to the network itself. Technical losses in dilapidated distribution systems were high, and commercial losses, energy taken without payment through tampered meters, direct tapping and unbilled connections, were higher still in the worst-run territories, reaching levels that no tariff structure could have financed. The two were habitually reported as one figure, transmission and distribution loss, and the habit served everyone except the reader of the accounts: a government could attribute to physics what belonged to enforcement, and a Board could attribute to theft what belonged to its own failure to meter and bill. The reform programme’s insistence on feeder metering, energy audit and the separation of technical from commercial loss was an insistence on rebuilding distinctions the old regime’s reporting had collapsed. It is one measure of the 1948 design’s optimism that it assumed a statutory corporation staffed by engineers, publishing audited accounts under a finance chapter that prescribed depreciation and interest discipline, would produce accounts from which the system’s condition could be read. The accounts were produced. The condition could not be read from them, because the categories that mattered, subsidy, theft, technical loss and directed underpricing, were merged in the figures the statute required.

What the Board model achieved

The case against the Board model is easily made and was made, in the end, by the structure of the reforms that replaced it. The case for it must be put at full strength as well, because it is substantial, and because the reformers of the nineteen nineties, to their credit, rarely denied it. Under the Board model, and under no other institutional arrangement that was on offer in the period, India built a national electricity system. Installed generating capacity rose from about 1,700 megawatts in 1950 to about 100,000 megawatts by the end of the century, on the Central Electricity Authority’s published counts. Village electrification rose from roughly three thousand villages in 1950 to more than four hundred thousand villages by the century’s end on the same source. The pumpset population that made the Green Revolution’s irrigation possible was energised by the Boards. The heavy industrial load of the planned economy was carried by the Boards. A private licensing model of the 1910 kind would not have extended supply to loss-making rural areas at anything like that pace, because no licence condition the state could have enforced would have made rural extension profitable at the tariffs rural consumers could pay; the extension happened because a public institution could be directed to do it and could be given capital with which to obey.

The criticism, properly stated, is therefore not that the Boards existed or that they were public. It is the narrower and more damaging criticism that the statute conflated two different obligations and funded neither transparently. The developmental obligation, to extend supply regardless of return, and the commercial obligation, to earn a return on assets, were imposed on the same body by the same Act, and the Act provided no mechanism by which the cost of the first was identified, measured and paid for by the government that imposed it. Section 63’s subvention was such a mechanism in embryo, but it was discretionary, it depended on annual budgetary grace, and it was habitually used at a fraction of the required scale or not used at all. Other public utility systems solved this problem, where they solved it, by explicit community service obligation payments or by licence conditions with identified funding. The 1948 Act left it to goodwill between a Board and the government that appointed it, and goodwill was the first casualty of electoral competition over tariffs. The deficit that resulted was real money: investment deferred, maintenance starved, dues to central generators unpaid, and a habit of treating the power sector’s balance sheet as the shock absorber of state politics, a habit the successor regime inherited along with the assets.

The licence in practice: term, renewal and the shadow of purchase

The 1910 Act licence had a temporal structure that shaped the industry’s investment behaviour for decades, and it deserves attention because its effects are a standing lesson in how time limits in infrastructure law consume the assets they govern. A licence to supply energy was granted for a defined term, twenty years in the first instance under the original scheme, renewable for further periods of ten, and the licensing framework carried with it the correlative right of the public authority to purchase the undertaking when the licence ran out, on terms the Acts prescribed. For the licensee the arrangement created a horizon problem of a classic kind. Investment in generation and distribution assets pays for itself over thirty or forty years; a licence with twenty years to run, and a purchase option at the end of it exercisable by an authority that would value the undertaking under statutory rules, made investment in the licence’s second half a speculation on the purchaser’s generosity. The predictable results followed: licensees sweated their assets as terms ran down, extensions of area and capacity lagged behind the growth of the towns they served, and the visible decay of service in the older licensed areas became one of the standing arguments for the 1948 reorganisation, deployed by the state against the companies whose underinvestment the state’s own purchase rights had encouraged.

The 1948 Act sharpened the dilemma by moving the purchase right to the Board under section 71, thereby uniting in one institution the roles of competitor, bulk supplier and eventual purchaser. A licensee negotiating renewal or extension with its state government after 1948 knew that refusal left the field to a Board that could wait for the term to expire and take the undertaking at the statutory value, and the knowledge disciplined licensees quite as effectively as any direction could have done. The lengthening of licence terms in the 1991 amendments, to thirty years with renewal for twenty, was a deliberate correction of the horizon problem in the generation context, an acknowledgment that the old terms had been set for an industry of town lighting and could not finance an industry of power stations. It arrived too late to revive the distribution licensee as a serious investment vehicle, since by then the Boards occupied the field, but as a piece of legal diagnosis it was exact: the reformers had identified the term structure of the licence itself as one of the old regime’s self-defeating features, half a century after the structure was set.

The purchase machinery also generated a quieter legal legacy. Valuation disputes on the taking over of undertakings, over the treatment of service lines paid for by consumers, over assets added in a licence’s final years and over the going-concern value of a supply business, recur in the reports across the regime’s life, and the principles they generated about statutory valuation coloured the drafting of every later vesting provision, including the transfer schemes of the reform era, whose schedules are drafted with a particularity that only makes sense as a response to generations of litigation over what, exactly, had been taken and what it was worth.

The Sixth Schedule: the licensee’s financial discipline

The financial regulation of the surviving licensees deserves a section of its own, because the Sixth Schedule to the 1948 Act was the most elaborate piece of economic regulation in the entire regime and because its concepts outlived it. Where section 59 stated a single target for the Boards, the Sixth Schedule built a complete accounting machine for licensees. It defined the capital base, the sum on which return was to be computed, built from the original cost of the licensee’s works less depreciation and consumers’ contributions and intangibles. It defined clear profit, the licensee’s net earnings after working expenses, depreciation, taxes and interest. And it defined the reasonable return to which the licensee was entitled on its capital base, comprising principally a return at a standard rate linked to the Reserve Bank’s rate, together with specified additions for reserves and borrowings. The regulatory teeth were in the sharing rule: clear profit was permitted to exceed the reasonable return only within a margin, and where the excess passed the prescribed limit, set at fifteen per cent of the amount of reasonable return, the surplus beyond it was divided, one part retained by the licensee within limits, one part returned to consumers by way of rebate in the following period, and one part carried to a tariffs and dividends control reserve. The licensee thus had its profit computed by the statute, capped by the statute and shared out by the statute, in a regime of cost-based regulation more prescriptive than anything the later commissions would attempt.

The Schedule’s virtues and defects were those of cost-plus regulation everywhere. It gave the licensee security: capital honestly invested earned a defined return, and the formula, administered by state governments through the accounts the licensees were required to render, kept supply companies financeable in an era when electricity was a speculative novelty in most of the country. It also dulled every incentive to economy, since costs honestly incurred passed into the computation and profits were capped just beyond the point at which efficiency would have enlarged them, while the division of the excess among licensee, consumers and reserve ensured that no one’s gain from a well-run year was large enough to be worth a quarrel with the auditor. And it depended, as the Boards’ regime depended, on the integrity of the accounts: capital base could be swollen by the valuation of works, expenses could migrate between heads, and the supervision available to a state electricity department was a fraction of what the formula’s complexity required. When the regulatory commissions took over tariff determination after 1998, they abandoned the Sixth Schedule’s machinery while retaining its philosophy, that a monopoly’s price should be built from its audited costs plus a fair return, and the abandonment is usually described as modernisation. In substance it was a change of machinery in the service of the same idea, and the commissions’ revenue requirement and return on equity are the Sixth Schedule’s clear profit and reasonable return in contemporary dress.

The private licensees’ twilight

The private and municipal licensees did not disappear when the Boards arrived; they entered a long twilight whose terms the 1948 Act dictated. Their licences continued, their areas were protected while the licences ran, and the Sixth Schedule continued to discipline their finances through the concepts of capital base, clear profit and reasonable return. But the centre of gravity had moved. The Board decided whether to supply them in bulk and on what terms under the Act’s scheme for sales to licensees; the Board held, under section 71, the purchase rights that the 1910 Act had given to local authorities and governments; and the Board, backed by state capital, undertook the extensions into new load that a licensee financed from a regulated return could not match. When licences expired, areas were taken over; when licensees failed, Boards stepped in. In a few places the licensee was strong enough to hold its ground indefinitely, and the Calcutta licensee and the Bombay undertakings remained substantial operators to the end of the regime, supplying some of the densest and most valuable load in the country under statutes that contemplated their absorption.

The twilight matters to the legal history for one reason: it demonstrates that the 1948 Act was not nationalisation by proclamation but nationalisation by machinery. No provision expropriated the licensees. Instead the Act rearranged the incidents of their position, the source of their bulk supply, the identity of the authority holding the option to buy them out, the financial formula under which they earned, until the private sector’s continuation depended entirely on the forbearance of the public institution the Act had placed around it. When policy turned in the nineteen nineties and private participation was invited back, it returned in generation first precisely because generation was the segment where the Board’s machinery had left a space: a generating company sold to the Board under contract and needed no licence area, no street works and no retail relationship with the public.

The opening of 1991

The amendment of the 1948 Act in 1991 was the statutory expression of the wider economic policy change of that year, and in this sector, as in industry generally, the pattern of the licence raj’s dismantling was repeated: policy moved first and the statute was adjusted to accommodate it. The Centre and the states adopted a common minimum national action plan for power in 1991, recording that public resources could not bridge the gap between demand and supply and that private investment in generation was to be encouraged. The 1948 Act was amended accordingly: the provisions governing generating companies, in section 15A and its companions, were recast so that companies other than the central and state generating companies could establish and operate generating stations and sell the output to the Board; the licence terms available to private licensees were lengthened, to thirty years in the first instance with renewal for twenty, in place of the earlier twenty and ten; and the provisions of sections 49, 57 and related sections and of the Sixth Schedule were adjusted to the new dispensation. A private generator’s statutory counterparty remained what it had been since 1948: the Board, which bought the power and sold it on. Nothing in the 1991 amendments touched the Board’s finances, its tariff subordination, or the unmetered subsidy that consumed its revenues.

Why did private generators in the nineteen nineties insist on guarantees?

Because their only customer was a Board whose statutory finances were already broken. A generator selling its entire output to a single buyer of doubtful creditworthiness could not raise project finance on the buyer’s promise alone, so the promise was wrapped: escrow cover over the Board’s receivables, a state guarantee behind the Board, and for the fast-track projects a counter-guarantee from the Centre behind the state.

The guarantee architecture that resulted became the defining artefact of the decade’s power policy. A power purchase agreement committed the Board to buy the plant’s output, typically on take-or-pay terms under which payment followed availability rather than despatch, for twenty years or more. Because lenders would not accept the Board’s covenant, the agreement was supported by an escrow arrangement over identified receivables of the Board, by a guarantee from the state government, and in the case of the eight fast-track projects to which the Centre extended the facility, by a counter-guarantee from the Government of India capped at a specified amount. Credit risk that in a commercial system would have been priced was instead transferred, layer by layer, to public balance sheets, while the underlying reason for the risk, a buyer directed to sell below cost, was left untouched. The Supreme Court, narrating this history in litigation arising from the Andhra Pradesh arrangements, recorded both the enthusiasm of the response, with states issuing large numbers of letters of comfort and memoranda for private projects, and the Centre’s later conclusion that the results were not encouraging despite the amendments, a conclusion that fed directly into the decision to create independent regulators in 1998.

The Dabhol project in Maharashtra became the episode through which the whole structure was publicly understood, and it should be described with precision because it is usually invoked as a slogan. The Dabhol Power Company, promoted by Enron with General Electric and Bechtel as co-sponsors, signed a power purchase agreement with the Maharashtra State Electricity Board in December 1993 for a two-phase gas-fired plant of about 2,000 megawatts on the Konkan coast. The Board was the sole buyer. The state government guaranteed the Board’s payment obligations and the Centre counter-guaranteed them, subject to a cap of roughly 300 million dollars. The incoming state government of a different political combination cancelled the project in 1995; after review it was renegotiated and revived, with revised costs and a part of the equity passing to interests associated with the Board. Phase I, of 740 megawatts, began producing power in May 1999. The tariff the Board had to pay for Dabhol power was several times the average cost of its existing supply, and from 2000 the Board fell behind on its bills; the state invoked and then contested its guarantee, the Centre’s counter-guarantee was called, a committee under Madhav Godbole recommended renegotiation in April 2001, arbitration followed, and the plant ceased production in 2001, its second phase nearly complete. Enron’s collapse into bankruptcy at the end of 2001 finished the episode as a commercial matter, though the asset was revived years later under public ownership.

Two lessons were drawn, and they point in opposite directions, which is why the episode still divides opinion. The first, drawn by the reformers, is that the failure was structural: a reform that admitted private capital at the generation end of a chain whose retail end remained politically priced and whose sole buyer was insolvent by design could only produce contracts that socialised the risk, and the wonder is not that one project failed but that the template was applied at all. The second, drawn by critics of the opening such as those writing in the Delhi Science Forum’s journal, is that high-cost private power with guaranteed off-take could only worsen the Boards’ position, since every unit bought dear and sold cheap enlarged the loss the Board was already directed to incur. Both lessons are true of the same facts, and the 2003 settlement can be read as an attempt to answer both: keep private generation, but separate the tariff decision from the government and the operator alike, so that the buyer’s creditworthiness would rest on a regulator’s cost-based tariff rather than on a guarantee.

Transmission in 1998 and the first separation

Two enactments of 1998 carried the reform from generation to the rest of the chain. The first was an amendment to the old statutes themselves. The Electricity Laws (Amendment) Act, 1998 (Act 22 of 1998), effective from 31 December 1998, recast the law of transmission: it recognised transmission as an activity distinct from generation and supply, provided for a Central Transmission Utility and State Transmission Utilities, empowered those utilities to use the surplus capacity in other licensees’ lines on payment of charges, and opened the construction and operation of transmission lines to transmission licensees from the private sector. Section 41 of the 1948 Act was substituted to give the new scheme its operative core. The conceptual importance of the amendment exceeded its immediate commercial effect, which was modest: for the first time in the regime’s history, a function inside the Board’s vertically integrated monopoly was named in the statute book as a separable business, with its own licensees and its own charges. Unbundling began as a definition before it became a restructuring.

The designations under the amendment showed how far the idea had travelled. The Power Grid Corporation of India, the central public sector company that had taken over the transmission assets of the central generating companies at the start of the decade, was notified as the Central Transmission Utility, charged with planning and coordinating the inter-state system and with operating the regional load despatch centres through which the grids were run. State Transmission Utilities were designated within the states, in most cases the Board itself wearing a second hat, which exposed the amendment’s transitional character: the function had been named and separated in law while remaining, for the time, inside the institution it would eventually leave. Even that half-separation had practical consequences, because transmission charges and losses could for the first time be identified as quantities distinct from the Board’s general finances, and a generator or licensee seeking to move power across another’s lines had, for the first time, a statutory scheme of charges and a utility obliged to deal rather than a monopoly entitled to refuse.

The second enactment was the Electricity Regulatory Commissions Act, 1998 (Act 14 of 1998), and it made the break that mattered most. It required the Central Government to establish a Central Electricity Regulatory Commission and empowered state governments to establish State Electricity Regulatory Commissions, and it transferred to those commissions the function that the 1948 Act had left with the Boards and their governments: the determination of tariffs. Where a state commission was constituted, the tariff for supply within the state, including the bulk supply tariff between a Board and its licensees and the retail tariff to consumers, was to be determined by the commission under the 1998 Act’s principles, guided by the financial principles inherited from the 1948 Act but answerable to its own statutory duty to safeguard consumers and ensure a fair return. For the first time since 1948, the price of electricity was to be set by a body that neither generated, transmitted, distributed nor sold it, and which no government could direct on a question of policy in the manner of section 78A.

The transfer was deliberately incomplete, and its limits explain the transitional character of the 1998 Act. Establishing a state commission was, in the Act’s first design, a state choice; the central commission was mandatory but the state commissions were enabling, so the separation arrived state by state over several years. The commissions inherited tariff principles that still pointed back to the 1948 Act’s schedules. And the commissions regulated Boards that remained vertically integrated and government-directed in their ownership, so the regulator’s independence was formal while the operator’s subordination continued. Even so, the institutional fact had been created, and it could not be uncreated: tariff orders began to be issued after public hearings, with reasons, by commissions whose proceedings generated the record of costs and losses that the old regime had kept inside the Boards’ accounts. Much of what the country learned about the true size of the farm subsidy and the true level of distribution losses, it learned from regulatory proceedings of the years after 1998, conducted on tariff petitions, cost statements and audited accounts placed on public record, records that survive in the document and spreadsheet files of the period, of the kind that a viewer for opening Excel, Word and presentation files exists to make readable. That last disclosure function was, in the event, as consequential as the tariff function itself.

Orissa goes first

The complete version of reform, in which a state restructured its own industry rather than waiting for the Centre, came from Orissa, and it came before the 1998 central Acts. Orissa’s circumstances made it the plausible pioneer: the Orissa State Electricity Board was in acute financial distress, the state’s agricultural share of consumption was small by national standards, which reduced the political cost of tariff rationalisation, and the World Bank, already engaged on a hydro project in the state, made a substantial loan, of 350 million dollars, conditional on restructuring and phased against its milestones.

What did Orissa do in 1996 that no other State had done before?

Orissa enacted its own reform statute, the Orissa Electricity Reform Act, 1995, and from 1 April 1996 unbundled its Board by law into separate generation, transmission and distribution businesses, vested hydro generation in a new corporation, created an independent state regulatory commission, and put distribution on a path to private ownership, years before the central statute made the model general.

The sequence deserves its detail. The Orissa Electricity Reform Act, 1995 (Orissa Act 2 of 1996) took effect on 1 April 1996. The Orissa State Electricity Board’s hydro generating stations passed to a new Orissa Hydro Power Corporation; its thermal stations were already with the Orissa Power Generation Corporation; and its transmission and distribution functions passed to the Grid Corporation of Orissa, GRIDCO. The Orissa Electricity Regulatory Commission was constituted in 1996 among the first state commissions in the country, predating the central Act that would have authorised it. GRIDCO’s distribution business was divided into four zones, corporatised as four distribution companies in 1997, and in 1999 fifty-one per cent of the equity in each was sold: three companies to Bombay Suburban Electric Supply and the central company, CESCO, to the AES Corporation. The Board, as an integrated utility, had ceased to exist in the state that reformed first, seven years before the national repeal.

The Orissa experiment’s afterlife was mixed in ways that both camps in the reform debate cite, and honesty requires both halves. The restructuring created the template: separate the businesses, corporatise them, regulate them independently, and admit private capital where it can be made to bear risk. It also produced an early warning: the private distribution companies’ performance on losses and collections disappointed the expectations set for them, relations between GRIDCO and the distribution companies over bulk supply dues soured into dispute, and AES eventually exited the central company, whose licence was revoked and whose operations passed to a utility under regulatory administration in 2006. Haryana followed Orissa with a reform Act in 1997, Andhra Pradesh in 1998, and further states, including Uttar Pradesh, Karnataka, Rajasthan and Delhi, enacted reform legislation in the years around the turn of the century, each adapting the template with variations: unbundling everywhere, privatisation of distribution only in some. The Schedule to the 2003 Act, discussed below, is in part a roll of these state enactments, preserved because they had created institutions and vested property in ways the repealing statute chose to respect rather than override.

The regime’s uneven geography

The Board model was general but never universal, and its exceptions are instructive. Section 5 obliged every state government to constitute a Board, but in several smaller states and in the Union territories the supply of electricity continued to be run by a government electricity department rather than by a Board, under arrangements the Act tolerated in practice through the state government’s own position as supplier and through the licensing of the department’s operations. Departmental supply differed from Board supply in ways that cut both ways. A department had no separate balance sheet, no section 59 to fail, and no fiction of commercial autonomy to maintain; its losses were simply the government’s losses, visible in the budget. It also had no borrowing capacity of its own, no corporate continuity, and no insulation at all between the minister and the switchyard. The comparison suggests that the Board form was neither the cause of the sector’s troubles nor an irrelevance: the same tariff politics operated in the departmental states with the same fiscal results, while the Board states added to those results the particular opacity of a corporation that could be made to carry, off budget, costs that belonged to the state.

Performance among the Boards themselves varied widely, and the variation tracks the structural variables rather than the quality of managements. Boards serving states with a high share of industrial load and a low share of agricultural load found the arithmetic of cross-subsidy survivable for longer; boards in the great agricultural states, where farm consumption was the largest and fastest-growing category and where the political commitment to cheap or free farm power was most absolute, reached insolvency first and deepest. Hydrology mattered: a state with a large share of depreciated hydroelectric capacity enjoyed a cost base that coal-dependent states could not match, and the accident of river geography appears in Board accounts as apparent managerial virtue or failure. The point matters for the assessment of the model, because the national debate habitually treated the Boards as a single class of institution succeeding or failing together, when the statute’s design in fact guaranteed a distribution of outcomes: identical legal obligations, radically different loads, resources and politics. A design that can only be complied with where geography and politics are kind is not complied with by virtue where it is complied with at all.

Delhi furnished the limiting case of a different kind. The capital’s supply was run for most of the period not by a Board under the 1948 Act but by a municipal undertaking, the Delhi Electric Supply Undertaking, succeeded in 1997 by the Delhi Vidyut Board as the national capital’s restructuring began. Delhi’s losses, driven overwhelmingly by theft and non-collection in a dense urban network rather than by agricultural subsidy, reached levels, reported at more than half of all energy supplied in the worst years, that no Indian state approached, and they demonstrated that the distribution disease had more than one pathology: where the farm subsidy was absent, collection failure and theft could do the same work alone. The Delhi reform, legislated by the Delhi Electricity Reform Act, 2000 and carried into privatisation in 2002, belonged to the same family as Orissa’s and is part of the prehistory the 2003 Act’s Schedule preserved.

The consumer under the old regime

The consumer’s legal position under the three statutes was that of a recipient at the end of a chain of public decisions, and the texts gave the consumer few rights that could be asserted against the chain. A licensee under the 1910 Act owed duties to supply within its area on the terms of its licence, and the Act and the Indian Electricity Rules regulated the safety and technical quality of the connection, the meter and the installation. The terms of supply, the tariff and the conditions of disconnection were set by the licensee within its licence or by the Board under its conditions of supply, documents of a contractual appearance whose content the consumer had no part in making and no forum in which to contest as a class. An individual dispute over a meter’s accuracy or a bill’s correctness could be taken to the Electric Inspector on the technical questions the Rules assigned, or to the civil courts, at a cost in time and money that confined the remedy to substantial consumers. Collective redress there was none: the tariff that determined every bill was fixed without hearing, notice or published reasons, and the first Indian electricity consumers to be heard on the price of their supply were heard by the regulatory commissions after 1998, in proceedings whose novelty contemporaries remarked on precisely because the old regime had never provided them.

The criminal law of the regime touched the consumer more readily than its civil law empowered the consumer. The 1910 Act made the dishonest abstraction, consumption or use of energy an offence under section 39, deeming it theft within the meaning of the Indian Penal Code, and created companion offences for interference with meters and with the licensee’s works. In the Boards’ later decades these provisions, supplemented by the Boards’ powers to disconnect and to assess, became instruments of revenue enforcement on a large scale, and their administration generated its own jurisprudence on the proof of dishonest intention, the evidentiary value of a tampered seal, and the distinction between a defective meter and a bypassed one. The asymmetry is worth recording without exaggeration: the old statutes specified the consumer’s offences in detail and the consumer’s remedies hardly at all, an allocation that reflected the regime’s founding assumption, that supply was a public service administered by a public body whose benevolence was structural and whose decisions therefore needed no adversarial testing. The regulatory era’s apparatus of standards of performance, compensation for failure, ombudsmen and consumer advocacy is unintelligible except as a reaction to that assumption’s long failure in practice.

One more feature of the consumer’s position deserves mention because it explains a persistent feature of Indian tariff politics. Most consumers, most of the time, had no metered relationship with price at all. The agricultural consumer paid by the horsepower; the domestic consumer in much of the country paid on meters that were defective, unread or estimated in a large minority of cases; the industrial consumer, alone, faced an accurate meter and a price above cost, and responded, as described, by leaving. A tariff system in which the price signal reaches only the consumer the system can least afford to lose is not merely inequitable in the small; it is a machine for selecting the worst possible revenue base, and it ran for decades inside a statutory framework whose finance chapter assumed the opposite. The metering and billing reforms that every post-2003 programme has pursued are, seen from this angle, attempts to build for the first time the informational foundation that section 59’s arithmetic had presupposed in 1948 and that the regime never in fact possessed.

Electricity in a federal constitution

The regime’s federal texture explains several of its oddities. Because electricity stood in the Concurrent List, the three central statutes coexisted with state law at every point: state governments granted the 1910 Act licences, constituted the Boards, made the rules under both central Acts for their territories, and in the reform decade legislated over the same field under their own competence, subject to the repugnancy discipline of Article 254. The Damodar Valley Corporation, created by a central Act of 1948 (Act 14 of 1948) on the model of the Tennessee Valley Authority, operated a multipurpose river valley project across Bihar and West Bengal outside the Board system altogether, generating and transmitting power as a central statutory corporation in a field the 1948 supply statute was simultaneously organising state by state. The Bhakra Beas Management Board, constituted under the Punjab Reorganisation Act, 1966, administered the great northern dams and their power houses for the successor states of a divided Punjab. Inter-state river projects generally produced bodies of this kind, creatures of reorganisation statutes and inter-state agreements rather than of the electricity Acts, so that the map of Indian electricity law in, say, 1990 showed the three-statute regime in the foreground and, behind it, a sediment of special statutes and agreements governing some of the largest assets in the system.

The federal structure also determined where the reform pressure could lawfully be applied. The Centre could amend the three Acts, as it did in 1991 and 1998, and could create a central commission and central utilities; it could not unbundle a state Board, set a state’s retail tariff, or direct a state government’s power policy, because the Board was the state’s creature and distribution was administered by the states. Hence the characteristic instruments of the nineteen nineties: model reform legislation commended to the states, conferences of chief ministers and power ministers that adopted common minimum programmes, fiscal incentives and memoranda by which states bound themselves voluntarily to milestones, and finally a consolidating central Act that set the framework for the whole country while expressly saving, in its Schedule, the divergent state settlements already made. Readers of this series will recognise the pattern from other concurrent fields, and electricity states it with unusual clarity because the assets were immovable, the employees numerous and the voters all consumers: in a concurrent field, the Union legislates the grammar and the states write most of the sentences, and a repeal in New Delhi is only ever half the event.

Repair by amendment, and why it could not hold

The repeal of 2003 was not Parliament’s first response to the regime’s difficulties, and the sequence of earlier repairs is the strongest evidence for the structural reading of the failure, because each repair addressed the symptom it could reach and the defect survived them all. Parliament stiffened the financial duty in 1978, substituting section 59 in the precise three per cent form described above, with its defined asset base, at a time when the Boards’ failure to earn the earlier and vaguer standard was already documented; precision did not produce compliance, because the obstacle was never the duty’s vagueness. The Centre exhorted through plan documents and the National Development Council; the Pawar Committee of 1993 and 1994 recommended restoring the Boards’ statutory autonomy, a recommendation addressed to the state governments whose conduct was the problem and dependent on those governments for its execution. The opening of 1991 added private generation to a system whose buyer could not pay. The transmission amendment of 1998 separated a function on paper within institutions that remained integrated in fact. The Regulatory Commissions Act of 1998 at last touched the real defect, the fusion of operator and tariff setter, and its partial character, optional for states, principles inherited from the old schedules, showed how difficult the political system found it to surrender the tariff even to its own appointees.

There is a general lesson here that this series has met in other fields and that the electricity case states in its purest form. Where a statute’s failure flows from the incentives of the actors it creates, amendment that leaves the actors and their incentives in place will be absorbed by them. The Boards did not disobey section 59 out of perversity; they complied with the more powerful instruction, section 78A and the political reality behind it, and reported the result. The governments did not refuse subvention out of forgetfulness; an unfunded promise costs nothing at announcement and its price falls on a corporation’s balance sheet rather than a budget’s. The industrial consumer did not build captive capacity out of disloyalty; the tariff made it rational. Reform worked, when it worked, by changing who decided: a commission for tariffs, companies for operations, a statute for the subsidy that required it to be paid in cash and shown in the open. The 2003 Act’s significance lies less in any single provision than in its acceptance of that logic across the whole field at once, which is also why it repealed rather than amended: the three statutes were not repaired because the thing wrong with them was their joint allocation of functions, and a joint allocation can only be replaced jointly.

The regime on one page

The table below is the findable artifact of this article. It takes the functions an electricity law must assign somewhere, and shows which of the three statutes dealt with each and how, with a final column recording what the consolidating Act of 2003 did with the function. Read across a row and the regime’s evolution is visible; read down the column of the 1910 Act and the narrowness of the original statute is visible, which is the point the text above has laboured.

Function	Indian Electricity Act, 1910	Electricity (Supply) Act, 1948	Electricity Regulatory Commissions Act, 1998	Electricity Act, 2003
Licensing of supply	Licences granted by the state government for supply in a specified area	Licences continued; Boards supplied licensees in bulk and held purchase rights over their undertakings	Licensing left with the existing authorities; commissions given tariff and regulatory functions	Unified licensing for transmission, distribution and trading; generation delicensed
Ownership of generation	Private and municipal licensees and government departments	State Electricity Boards as principal generators, with central and state generating companies added later	No change to ownership; commissions regulated whoever generated	Generation opened generally; Boards’ generation passed to successor companies under transfer schemes
Transmission	Incidental to licensees’ works; sections 12 to 18 governed lines and street works	A Board function within the integrated state system	Activity separately recognised by the 1998 amendment; Central and State Transmission Utilities designated	Separate transmission licensees and utilities; open access to networks
Distribution	By licensees within licensed areas	By Boards across the state, absorbing licensees as licences ended	Regulated by state commissions where constituted	Distribution licensees, unbundled from Boards in most states, with open access phased in
Tariff setting	Not addressed; left to licence conditions	By the Board under section 49, subject to the section 59 financial principles; licensees under the Sixth Schedule	Transferred to central and state regulatory commissions	Determined by the appropriate commission under the 2003 Act’s own principles
Subsidy	Not addressed	State subventions to Boards under section 63, discretionary and budget-dependent	Commissions directed to reflect subsidy in tariff where the state paid it	State may grant subsidy, which must be paid to the licensee in advance and shown in the tariff order
Dispute resolution	Arbitration and the Electric Inspector for works and safety matters	Central Electricity Authority arbitrated referred matters, including whether a question was policy under section 78A	Commissions adjudicated disputes within their regulatory sphere	Commissions adjudicate; an appellate tribunal hears appeals from their orders

The courts and the Boards

The Boards generated a distinctive body of constitutional litigation, and its leading case is among the foundational authorities of Indian administrative law. In Rajasthan State Electricity Board v. Mohan Lal, decided on 3 April 1967, a Board employee’s claim raised the question whether a State Electricity Board was “State” within Article 12 of the Constitution, so that its service decisions were answerable to the fundamental rights. The Supreme Court held that it was: the expression “other authorities” in Article 12 covered every authority created by statute and functioning within the territory of India, and the circumstance that the Board carried on activities of a commercial nature did not exclude it. The reasoning fastened on the very features this article has described, a body created by statute, armed with statutory powers including powers of direction enforceable by penalty, and performing a function the state would otherwise perform itself. Sukhdev Singh v. Bhagatram Sardar Singh Raghuvanshi in 1975 extended the instrumentality doctrine across the statutory corporations of the period. The consequence for the Boards was permanent: their employments, their contracts at the public law margin and their treatment of consumers were all exposed to writ jurisdiction, so that the institution the 1948 Act designed as a commercial corporation spent much of its court life being reminded that it was also the state.

On the financial constitution, the courts’ role was of a different kind: conspicuous restraint. Section 59 was litigated around rather than through. Tariff revisions by Boards were challenged by industrial consumers and defended by Boards as necessary to statutory duty and rising costs, and the courts treated tariff fixation as a legislative and policy function into which they would not substitute their own judgment, intervening on grounds of arbitrariness or want of authority rather than of arithmetic. No line of authority converted the three per cent clause into an enforceable entitlement, whether of a consumer to resist an increase or of a creditor to demand one. The clause did its work, such as it was, inside public finance: finance commissions examining state finances, committees of the kind the National Development Council appointed in 1993, and the Planning Commission’s annual reports invoked it as the benchmark against which Board performance was to be judged, and recorded, year after year, that the benchmark was missed. A statutory obligation can fail in two ways, by being defied or by being unenforceable against the party whose conduct determines compliance. Section 59 failed in the second way: the only institution that could have compelled the tariff or funded the gap was the government, and the Act gave no one a remedy against the government for the Board’s shortfall.

Transfer schemes and their litigation

Where states reformed before 2003, and in every state after it, the Board’s assets, liabilities and personnel had to be moved to successor companies, and the instrument was the transfer scheme. Under the state reform Acts, and later under the 2003 Act’s own scheme provisions, a state government notified a scheme vesting identified assets, rights, liabilities and proceedings in the new entities, and providing for the transfer of the Board’s officers and employees. The schemes were drafted to carry two assurances that the litigation then tested: that employees’ terms and conditions would be no less favourable in the successor than in the Board, and that pension and retirement benefits would be protected, commonly through a funded pension trust or an express assumption of the liability by a named successor or the state.

The disputes that followed were substantial and long-lived, and practitioners still meet their residue. Employees contested allocation between successors, claiming that options promised under a scheme had not been honoured or that inter se seniority had been disturbed by the vesting. Pensioners and serving staff litigated the adequacy of pension funding where a scheme divided a single pension liability among several successors whose fortunes differed. Creditors and contracting parties contested which successor had inherited a liability the scheme’s schedules had described imperfectly. Consumers and licensees contested the continuity of proceedings begun against the Board. The courts’ general approach, across these categories, was to treat the scheme as the governing instrument, to hold successors to the assurances the scheme contained, and to construe vesting provisions so that no liability fell into the gap between a dissolved Board and its successors. The deeper point for the legal historian is that the transfer scheme became the characteristic legal technology of Indian power reform: because the Board was a statutory corporation and not a company, it could not simply be restructured by resolution or merger, and every step of unbundling required statutory authority and a notified scheme, which is why the litigation shadow of the Board era is measured in decades rather than years.

Pension: the longest tail of the Board era

Of all the obligations the transfer schemes had to allocate, the pension liability proved the longest-lived, and it concentrates the era’s themes in a single problem. Board employees were, in the main, entitled to pension on the pattern of government service, a defined benefit promise whose cost fell due decades after the service that earned it. While the Board existed as a single employer, the promise was carried on the Board’s and behind it the state’s credit, largely unfunded in any actuarial sense. Unbundling divided the promise’s beneficiaries among successor companies of unequal strength and divided the corresponding liability according to the scheme’s schedules, commonly with a pension trust interposed as the vehicle that was to hold the funds and pay the pensions. Where the trust was funded on assumptions that events falsified, or where a successor’s contribution obligations were disputed, pensioners who had served an institution that no longer existed found their security dependent on the construction of a scheme and the solvency of entities they had never chosen to serve. The litigation and negotiation that followed in several states, over the state’s residual responsibility for the trust’s deficit and over the treatment of employees who had retired before the transfer date, continued well into the successor regime’s life, and the questions it generated, about the extent to which a statutory promise of pension can be restructured by the same authority that made it, belong to the permanent jurisprudence of Indian public employment. For the drafter of any future restructuring, the electricity experience fixes the lesson: the pension schedule is not an annexure to a transfer scheme but its longest obligation, and it is priced, if it is priced honestly, at the beginning or litigated, at far greater cost, at the end.

The repeal of 2003 and what Section 185 saved

The Electricity Act, 2003 repealed the three statutes in a single provision, section 185(1): save as otherwise provided in the Act, the Indian Electricity Act, 1910 (9 of 1910), the Electricity (Supply) Act, 1948 (54 of 1948) and the Electricity Regulatory Commissions Act, 1998 (14 of 1998) are repealed. The remainder of section 185 is a savings scheme of unusual care, and stating it exactly matters, because the afterlife of the old regime in litigation and in practice is governed by it.

First, the general saving in section 185(2)(a): anything done or any action taken under the repealed laws, including any rule, notification, order or notice, any appointment or declaration, and any licence, permission, authorisation or exemption granted, is deemed to have been done under the corresponding provisions of the 2003 Act, in so far as it is not inconsistent with that Act. Licences granted under the 1910 Act therefore continued as licences under the new dispensation, and the accumulated administrative output of ninety-three years was carried across rather than annulled. Second, two specific continuities from the oldest statute: sections 12 to 18 of the 1910 Act, the works provisions, were to have effect until rules under sections 67 to 69 of the 2003 Act were made, and the Indian Electricity Rules, 1956 were to continue in force until regulations under section 53 of the 2003 Act replaced them. Third, rules made under section 69(1) of the 1948 Act were continued until rescinded or modified. Fourth, and most significant for federal relations, section 185(2)(e) provided that directives issued by a state government under the enactments specified in the Schedule to the 2003 Act should continue to apply for the period for which they were issued, and section 185(3) provided that the provisions of the enactments specified in the Schedule, in so far as not inconsistent with the 2003 Act, should apply to the states in which those enactments were applicable. The Schedule was the roll of state reform Acts: the Orissa, Haryana, Andhra Pradesh, Uttar Pradesh, Karnataka, Rajasthan, Delhi and Madhya Pradesh enactments, among others, by which states had unbundled their Boards and created their commissions before the Centre legislated. The central Act thus repealed the central statutes outright but took the state reforms as it found them, preserving their schemes, their vestings and their commissions, the last reinforced by the provision in the 2003 Act itself under which a state commission constituted under the 1998 Act or a reform Act continued as the State Commission under the new Act. Section 185(4) empowered the Central Government to amend the Schedule, and section 185(5) preserved the general application of section 6 of the General Clauses Act, 1897 to the repeal, so that rights accrued and liabilities incurred under the old law survived by the ordinary rule as well.

Why did the 2003 Act keep the Indian Electricity Rules, 1956 in force?

Because the Rules were the operative safety code for electrical installations, and a gap in them would have left installations, inspectors and licensees without governing standards from the day of repeal. Section 185(2)(c) continued them until the new Act’s regulations under section 53 could be framed, trading a clean break for continuity in the one area where a break was dangerous.

The treatment of the Boards themselves under the repeal followed the same philosophy of continuity through transformation. The 2003 Act did not dissolve the Boards by its own force in the states that had not reformed. Its transitional provisions allowed a Board to continue for a period as the State Transmission Utility and as a licensee, and contemplated the eventual transfer of its functions and property under schemes framed by the state government, while the reform states’ schemes were already in place and were left standing. A Board thus ended, where it ended, by scheme and vesting rather than by repeal, and in a few states the Board form persisted for years within the new Act’s tolerance. The institution that section 5 of the 1948 Act had required every state to create was retired the way it had been created: by notification, state by state, at different speeds.

Why the regime endured

A design this defective invites the question why it lasted for half a century, and the answer completes the analysis, because the regime’s durability had the same source as its defect. The Board system distributed benefits that were visible, concentrated and electorally timed: a village connected, a pumpset energised, a tariff held down through an election season, employment in a large and unionised public workforce, contracts and postings within the gift of the institution and its government. Its costs were distributed invisibly, across future budgets, deferred maintenance and the balance sheets of a corporation few voters read about. Every actor with the power to change the design drew a benefit from its continuation, and the actors who bore its costs, future consumers, industrial users who could exit to captive generation, and the state’s own credit, were diffuse or could protect themselves. Reform came when the costs found a constituency that could not be ignored: the Union government, whose central generating companies were not being paid and whose fiscal exposure through guarantees and plan support had become undeniable, and the lending institutions, Indian and multilateral, whose continued finance the sector required and who could make restructuring a condition. The chronology fits the explanation. The serious repair attempts cluster in the decade after 1991, when the balance of payments crisis had weakened the old coalition’s claim on resources and when the Centre’s own dues and guarantees were at stake, and they succeeded in the states, Orissa first among them, where a lender’s condition met a government willing to spend political capital. Where neither condition held, the regime continued, and in departmental and Board form alike it continued in places long after 2003, because the repeal changed the law faster than it changed the coalition the law had served.

The eve of repeal

By the turn of the century the old regime’s continuation had become untenable in a form its participants could measure. The one-time settlement of the Boards’ dues to the central public sector undertakings, framed on the Ahluwalia expert group’s recommendations of 2001 and executed in 2002 and 2003, required the states to take over dues the Boards could not pay and to issue bonds against them, converting an operating failure into state debt on a scale that made the sector’s condition a direct charge on public credit. Reform programmes tied central assistance to state commitments: memoranda of understanding and then of agreement between the Centre and individual states traded financial support under programmes such as the Accelerated Power Development Programme for milestones on metering, energy audit, tariff petition filing and loss reduction, and by then nearly every state had either enacted a reform statute, constituted a commission, or committed itself to both. The Electricity Bill that became the 2003 Act was debated against that background: a sector in which the states had already diverged, some reformed on the Orissa template, some mid-transition, some barely begun, and in which a central consolidating statute had to gather the divergence without disowning it. The drafting of section 185, with its Schedule of protected state enactments and its deemed continuation of licences and actions, is the trace of that negotiation inside the Act, and it is why the repeal, alone among the great Indian repeals this series covers, reads less like a demolition than like a treaty between the old order’s several successors.

Assessment: a duty designed to be broken

The namable claim of this article can be stated in its full form. The Electricity (Supply) Act, 1948 imposed on every State Electricity Board a defined financial duty, to conduct its operations and adjust its tariffs so as to leave a surplus of not less than three per cent on the value of its fixed assets in service, and the same Act imposed on the same Board a duty to be guided by its state government’s directions on questions of policy, a category that in practice included the tariff itself. The governments’ settled policy was to price the largest categories of consumption below cost. The Act provided one mechanism, the section 63 subvention, by which the state could fund the difference its directions created, and left that mechanism to annual discretion. The financial duty was therefore unmeetable by design in every state where the direction was given and the subvention withheld, which is to say in most states for most of the regime’s life. The failure of the Boards to earn three per cent is the most consistently documented fact in Indian power sector history, recorded year after year in the Planning Commission’s annual reports on their working, and it is misread whenever it is read as a fact about management.

The two readings of that fact should be given their strongest forms, as this series’ charter requires, because the choice between them governed the reforms. On the first reading, associated with the reform programme of the nineteen nineties and with the committees that prepared it, the failure was one of institutional design and political economy: merge the operator, the tariff setter and the policy instrument in one body answerable to an elected government, and the tariff will be set electorally, the accounts will absorb the cost, and investment will be crowded out until the system decays. The remedy follows: separate the functions, put the tariff in the hands of a commission that must hear and reason, make subsidy an explicit payment from the budget, and let companies be companies. On the second reading, pressed by the critics of unbundling in journals such as the Delhi Science Forum’s and by the engineers’ and employees’ organisations that resisted restructuring, the failure was one of public finance and political choice for which the Board was the scapegoat: the state used its utility to deliver welfare and development it was unwilling to tax for, and then cited the resulting losses as proof that public ownership had failed, while the private alternative on offer in the nineteen nineties consisted of guaranteed contracts that socialised risk more completely than the Board ever had. On this reading, unbundling changed the accounting location of the subsidy without confronting the refusal to fund it, and distribution’s finances after reform are the evidence: the losses migrated to successor companies and continued.

The evidence assembled in this article supports a verdict, and the verdict is asymmetrical. On the mechanism, the first reading is right, and it is not close: the contradiction was inside the statute, the direction power was used as the critics of design said it would be, the subvention mechanism failed as a matter of routine, and no Board under sustained direction escaped the arithmetic. A legal design that requires an institution to disobey either its finance chapter or its political master has failed as law, whatever the merits of its intentions, and the 1948 Act’s defenders have never answered that point on the statute’s own terms. On the prognosis, the second reading has worn better: separating the functions proved necessary and not sufficient, because the funding question the 1948 Act evaded, who pays for electricity sold below cost and through what visible instrument, survived the repeal and confronts the successor regime in every tariff cycle. The honest summary is that 2003 fixed the address of the problem. Under the old regime the subsidy hid in a Board’s loss, unattributed and unmeasured; under the new it must at least be named, quantified in a tariff order and either paid or recorded as a default by an identified government. That is a large change in the quality of public argument, purchased by half a century of the alternative. But the change is in the argument, not yet in the arithmetic, and a reader who understands why section 59 failed understands both halves of that sentence.

Two subsidiary judgments follow for the practitioner and the student. The first concerns the 1910 Act, which the repeal buried with the others and which did not deserve the company it was buried in: as a works and safety code it functioned for ninety-three years, and the 2003 Act’s decision to keep its works sections and its 1956 Rules alive until replacements were ready was a tribute exacted by necessity to a statute that had done its limited job. The common error of treating the 1910 Act as an economic regulation statute should be retired; the economics were never in it. The second concerns the 1998 Act, which did its work and was repealed within five years, absorbed into the consolidating statute. Its brevity of life conceals its importance. The separation of the tariff from the operator was the single structural idea on which everything after 1998 depends, and it arrived not in the celebrated consolidation of 2003 but in the transitional statute of 1998, tested first in the states that adopted commissions under it and in Orissa, which had anticipated it under its own Act two years earlier. Legal history that dates Indian electricity regulation from 2003 misdates the turn by five years and misunderstands the 2003 Act, which consolidated a revolution already in progress rather than beginning one.

What the successor inherited

A repeal and replacement article owes the reader a closing account of continuity, because the regime’s questions did not end with its statutes. The Boards’ employees serve in the successor companies, and their service conditions, seniority and pension expectations remain governed by the transfer schemes under which they moved, documents framed under the reform Acts and the 2003 Act that practitioners will consult for as long as the transferred generation remains in service or in receipt of pension. The Boards’ contracts, guarantees and proceedings continued under the savings of section 185 and the schemes, and the escrow arrangements and guarantees of the nineteen nineties structured settlements that occupied the following decade. The Indian Electricity Rules, 1956 governed installations until the Central Electricity Authority’s regulations on safety and electric supply replaced them under the new Act, and prosecutions and assessments for periods under the old law continued to be governed by it through the ordinary saving of accrued liability. Even the three per cent idea survived in transmuted form: the regulatory commissions’ tariff practice under the 2003 Act, allowing a return on equity to licensees and building the utility’s revenue requirement from audited costs, is the intellectual descendant of the Sixth Schedule and of section 59, with the difference that the return is determined after hearing by a body the government does not direct, and the shortfall, where a government directs cheap power, is supposed to be made good by that government in cash. The machinery changed; the question, who pays the difference between the cost of electricity and the price politics assigns to it, is the oldest in Indian electricity law, and the three statutes repealed in 2003 are where it was first posed in statutory form and first answered by evasion.

A final word on method, owed to the student who will use this article in an examination hall and to the researcher who will use it in a footnote. The regime’s history is unusually well documented because its central failure was statistical: the Planning Commission’s annual reports on the working of the Boards, the finance commissions’ examinations of state finances, the committees of 1993 and 2001 and the regulatory proceedings after 1998 together constitute a record in which the same quantities are measured, year after year, by institutions with no agreed interest in the answer. Where this article has given a figure, it has tied the figure to that record; where the record itself was corrupted, as it was by unmetered consumption and merged loss categories, the article has said so rather than repeat a number whose foundations it had just criticised. That discipline is the difference between the history of a failure and one more telling of it.

Frequently Asked Questions

Q: Which laws governed electricity before 2003?

Three central statutes formed one regime. The Indian Electricity Act, 1910 (Act 9 of 1910) governed licences to supply energy in a specified area, the laying of works and lines, safety and offences. The Electricity (Supply) Act, 1948 (Act 54 of 1948) created the Central Electricity Authority and required each state to constitute a State Electricity Board, and it contained the financial principles for Boards and, in its Sixth Schedule, for licensees. The Electricity Regulatory Commissions Act, 1998 (Act 14 of 1998) created regulatory commissions and transferred tariff determination to them. State rules and, in the reform decade, state reform Acts completed the structure until the Electricity Act, 2003 repealed all three central statutes.

Q: What were State Electricity Boards and how did they work?

A State Electricity Board was a statutory corporation constituted by the state government under section 5 of the Electricity (Supply) Act, 1948, with three to seven members appointed by the government. It was a body corporate with its own funds, borrowing powers and audited accounts. In most states the Board became the vertically integrated electricity industry: it generated power, transmitted it across the state, distributed and sold it to consumers, supplied surviving licensees in bulk, fixed tariffs under the Act, and executed the state’s power investment programme with capital provided or guaranteed by the state government. Under section 78A it was bound to be guided by the state government’s directions on questions of policy.

Q: What return did the 1948 Act require of State Electricity Boards?

Section 59 of the Electricity (Supply) Act, 1948, in the form substituted by the amending Act 23 of 1978, required each Board, after crediting any subvention received from the state government under section 63, to carry on its operations and adjust its tariffs so that revenues, after meeting operating expenses, taxes, depreciation and interest, left a surplus of not less than three per cent, or a higher percentage the state government specified, of the value of the Board’s fixed assets in service at the beginning of the year. That value was the original cost of the assets less cumulative depreciation and consumers’ contributions for service lines.

Q: When was private generation first allowed?

Private participation in generation was opened by policy and amendment in 1991. The Centre and the states adopted a common minimum national action plan for power in that year, and the Electricity (Supply) Act, 1948 was amended so that generating companies outside the public sector could set up stations and sell their output to the State Electricity Boards; licence terms were lengthened to thirty years with renewal for twenty. The opening built on the generating company device in section 15A, under which public generating companies had operated since the nineteen seventies. The private generator’s buyer remained the Board, which is why the deals of the decade rested on escrow cover, state guarantees and, for eight fast-track projects, central counter-guarantees.

Q: What did the 1998 regulatory commissions Act transfer?

The Electricity Regulatory Commissions Act, 1998 transferred the determination of electricity tariffs from the State Electricity Boards and state governments to independent regulatory commissions. It required the Central Government to constitute a Central Electricity Regulatory Commission and empowered state governments to constitute State Electricity Regulatory Commissions. Where a state commission was formed, bulk supply and retail tariffs within the state were thereafter determined by the commission, after proceedings on the record, rather than by the Board that sold the power or the government that owned the Board. The transfer was the first structural separation of the regulatory function from the operating function in Indian electricity law, ending the position in which the seller of electricity also set its price.

Q: Which state reformed its power sector first?

Orissa. The Orissa Electricity Reform Act, 1995 (Orissa Act 2 of 1996) took effect on 1 April 1996, before any central reform statute. It unbundled the Orissa State Electricity Board, vesting hydro generation in the Orissa Hydro Power Corporation and transmission and distribution in the Grid Corporation of Orissa, constituted the Orissa Electricity Regulatory Commission among the first state commissions in the country, and divided distribution into four companies in which fifty-one per cent of the equity was sold in 1999. The reform was supported by a World Bank loan of 350 million dollars phased against restructuring milestones. Haryana followed in 1997 and Andhra Pradesh in 1998, and the 2003 Act’s Schedule preserved the state reform enactments.

Q: Why were the three earlier electricity statutes replaced?

Because their joint allocation of functions had failed. The 1910 Act was a licensing and safety code with no economic regulation. The 1948 Act fused generation, transmission, distribution and tariff setting in Boards that were simultaneously required to earn a statutory surplus and to obey government directions that set tariffs below cost, a contradiction that produced chronic losses, unmeasured subsidy and starved investment. The 1998 Act separated regulation from operation but was transitional and optional for states. The Electricity Act, 2003 replaced the set with a single framework: delicensed generation, separately licensed transmission and distribution, tariff determination by commissions, explicit subsidy payment by governments, and transfer schemes to unbundle the Boards.

Q: What happened to the assets and staff of the State Electricity Boards?

They moved by transfer scheme, not by the repeal itself. Under the state reform Acts in the states that reformed early, and under schemes framed by state governments in the remaining states within the 2003 Act’s framework, a notified scheme vested the Board’s assets, rights, liabilities and pending proceedings in successor generation, transmission and distribution companies, and transferred the Board’s officers and employees to those companies on terms the schemes required to be no less favourable, with pension and retirement benefits protected, commonly through a pension trust. In states that had not reformed, the 2003 Act’s transitional provisions allowed the Board to continue for a period as State Transmission Utility and licensee until schemes were framed.

Q: How were a licensee’s profits controlled under the Sixth Schedule?

By formula. The Sixth Schedule to the Electricity (Supply) Act, 1948 defined the licensee’s capital base, the original cost of its works less depreciation, consumers’ contributions and intangibles, and its clear profit, the net earnings after working expenses, depreciation, taxes and interest. The licensee was entitled to a reasonable return on the capital base, computed principally at a standard rate linked to the Reserve Bank’s rate with specified additions. Clear profit could exceed the reasonable return only within a margin: where the excess passed fifteen per cent of the amount of reasonable return, the surplus beyond it was divided between the licensee within limits, a rebate to consumers in the following period, and a tariffs and dividends control reserve. Profit was thus computed, capped and shared by statute.

Q: What happened to the private licensees once the Boards were formed?

They continued, but inside a structure arranged for their gradual absorption. Their licences ran their terms, their finances remained governed by the Sixth Schedule to the 1948 Act, and they bought bulk power from the Board. Section 71 of the 1948 Act vested the rights to purchase a licensee’s undertaking, which the 1910 Act had given to local authorities and governments, in the Board, so the institution that competed with the licensee and supplied it also held the option to buy it out when the licence expired. As licences ended or undertakings were purchased, areas passed to the Boards. Strong metropolitan licensees, in Calcutta and Bombay among others, held their areas to the end of the regime.

Q: Who fixed electricity tariffs before the regulatory commissions?

The seller did, under the owner’s supervision. Under section 49 of the Electricity (Supply) Act, 1948 the State Electricity Board fixed the tariffs at which it sold electricity, subject to the financial principles of section 59, while licensees’ tariffs were governed by their licences and the Sixth Schedule. In practice the state government, which appointed the Board and could direct it on questions of policy under section 78A, had the decisive voice, and tariff revisions were infrequent, political and unexplained by any published reasoning. No hearing, notice or reasoned order stood between the decision and the consumer. Determination by an independent commission, after a public proceeding, arrived only with the Electricity Regulatory Commissions Act, 1998 and the state reform Acts.

Q: Why did unmetered farm supply make the subsidy impossible to measure?

Because consumption that is never measured can only be estimated, and the estimate was constructed as a residual. Agricultural pumpsets were mostly unmetered and billed at flat rates related to horsepower. A Board therefore derived farm consumption by taking total generation, subtracting metered sales and subtracting an assumed figure for transmission and distribution losses. Everything unmeasured, network losses, theft, unbilled supply and genuine farm use, collected inside that one residual, and by varying the assumed loss percentage the apparent farm consumption, and with it the apparent size of the subsidy, could be moved within a wide band. The figure at the centre of the largest subsidy in the sector was thus an artefact of an assumption, which protected the subsidy from scrutiny and corrupted the loss statistics on which reform depended.

Q: What was a subvention, and why did subventions not repair Board finances?

A subvention was a grant from the state government to its Board under section 63 of the Electricity (Supply) Act, 1948, intended to bridge the gap between what the Board’s tariffs yielded and what its statutory duties cost; section 59 credited subventions before applying the three per cent test. Subventions failed for three reasons. They were discretionary, depending on annual budgetary decisions by the same governments whose tariff directions had created the gap. They were habitually smaller than the gap, or sanctioned and not released in full, or booked as loans rather than grants, adding interest to the expenses the surplus had to survive. And because they were negotiated rather than computed from a published measure of the subsidy, they never acquired the character of an entitlement the Board could plan upon.

Q: What did the Central Electricity Authority do under the 1948 Act?

The Central Electricity Authority, constituted under section 3 of the Electricity (Supply) Act, 1948, was the sector’s planning and technical authority. It developed national power policy, collected and published the statistics on which planning rested, appraised and advised on generating schemes, coordinated the development of the industry across states, and exercised the specific functions the Act assigned to it, including deciding, under section 78A, whether a disputed question between a Board and its state government was a question of policy, its decision being final. It was not a regulator: it licensed no one and determined no tariffs. Its role was coordination and expertise within a system whose operative decisions were taken by Boards and state governments.

Q: What changed in transmission law in 1998?

The Electricity Laws (Amendment) Act, 1998 (Act 22 of 1998), effective from 31 December 1998, recognised transmission as an activity distinct from generation and supply for the first time. It substituted section 41 of the Electricity (Supply) Act, 1948, provided for a Central Transmission Utility and State Transmission Utilities, empowered those utilities to use surplus capacity in the transmission lines of generating companies and licensees on payment of charges determined under the Act, and opened the construction and operation of transmission lines to private transmission licensees. The Power Grid Corporation of India was notified as the Central Transmission Utility. Unbundling began here, as a statutory definition, before it became a programme of restructuring.

Q: What were escrow cover and counter-guarantees in the power deals of the nineteen nineties?

They were the credit supports that substituted for the buyer’s creditworthiness. A private generator sold its entire output to a State Electricity Board whose finances were broken, so lenders required more than the Board’s promise: an escrow arrangement charged identified receivables of the Board to the project’s payments, a guarantee from the state government stood behind the Board, and for eight fast-track projects a counter-guarantee from the Government of India, capped in amount, stood behind the state. The Dabhol project in Maharashtra carried all three layers. The structure transferred the payment risk from the project to public balance sheets without curing its cause, a buyer directed to sell power below the price at which it bought, and the defaults and arbitrations of 2000 and 2001 were the predictable result.

Q: Do licences granted under the repealed Acts still have legal effect?

Yes, through the savings in section 185 of the Electricity Act, 2003. Section 185(2)(a) provides that anything done or any action taken under the repealed laws, including any licence, permission, authorisation or exemption granted, is deemed to have been done or taken under the corresponding provisions of the 2003 Act, in so far as it is not inconsistent with that Act. Licences under the Indian Electricity Act, 1910 therefore continued as licences in the new framework. Specific continuities reinforce the general one: sections 12 to 18 of the 1910 Act remained effective until rules under sections 67 to 69 of the 2003 Act were made, and the Indian Electricity Rules, 1956 continued until regulations under section 53 replaced them.

Q: Did the Supreme Court treat a State Electricity Board as part of the State?

Yes. In Rajasthan State Electricity Board v. Mohan Lal, decided on 3 April 1967, the Supreme Court held that a State Electricity Board is an authority within Article 12 of the Constitution, so that its actions are answerable to the fundamental rights. The Court reasoned that “other authorities” covers every authority created by statute and functioning within India, that the Board was armed with statutory powers including directions enforceable by penalty, and that the commercial character of its activities made no difference. Sukhdev Singh v. Bhagatram Sardar Singh Raghuvanshi in 1975 extended the instrumentality doctrine to statutory corporations generally. Board employees and consumers could therefore invoke writ jurisdiction against the Boards throughout the regime.

Q: Why did industrial consumers pay more than farmers under the Board system?

Because the tariff structure was built as a cross-subsidy and the arithmetic forced it upward. Governments directed or induced Boards to price agricultural supply, and much domestic supply, below the cost of serving it, and the shortfall had to be found somewhere if the Board’s revenues were to approach its costs at all. Industrial and commercial consumers, metered accurately and unable to bargain, were charged above cost, at levels that rose as the subsidised share of consumption grew. The structure defeated itself over time: the consumers charged the most were those best able to leave, and large industry built captive generating plants, removing high-yielding load from the Board and leaving a sales mix weighted still more heavily toward the categories that paid least, which pressed industrial tariffs higher again.

Q: What happened to the Boards in states that never passed a reform Act?

They were not dissolved by the repeal. The Electricity Act, 2003 repealed the Electricity (Supply) Act, 1948 under which the Boards existed, but its transitional provisions allowed a State Electricity Board to continue, for a period the state government could extend, discharging the functions of the State Transmission Utility and holding a licence, while the state framed transfer schemes to vest its undertaking in successor companies. States proceeded at very different speeds: some unbundled within a few years, others allowed the Board to function within the new Act’s tolerance for an extended period, and the institution that section 5 of the 1948 Act had required every state to create was retired state by state, by notification and scheme, rather than on a single appointed day.