
Sharmila Chavaly, a former civil servant who held key roles in the railways and finance ministries, specialises in infrastructure, project finance, and PPPs.
November 18, 2026 at 12:15 PM IST
Andhra Pradesh has proposed an Investment Protection Act which seeks to protect committed investments from policy reversals by successor governments. The trigger was the experience of the current governing alliance - when it lost an earlier election, it watched its successor undo the decision to build a new capital city at Amaravati, well past the land acquisition and preliminary design phases. The reversal was widely seen as political vendetta and was especially damaging to financial investors who had priced the project on continuity.
Investors look for policy consistency, certainty and predictability. Can another law, that too at the sub-national level, provide this in a country with many good laws, even while enforcement is weak and sometimes biased? What would such a law add to the domestic contract law and international investment framework?
The Government of India’s 2016 Model Bilateral Investment Treaty (BIT), which governs new investment negotiations, is seen as having shrunk the treaty backstop which older BITs had once provided. Is an Act the right instrument to address this or would a market mechanism do the job better? And can a state-level statute provide the domestic credibility that the treaty architecture no longer has?
The short answer in this piece: the state itself is only signalling, the escrow is the real instrument, and there IS a better substitute.
The Structure
The proposed statute seeks to shield industry approvals, land allotments, pricing of utilities, announced non-fiscal incentives and subsidies, and shielding of infrastructure from possible policy reversal. This is coupled with an escrow which will draw from state taxes the amount equivalent to the announced incentives, with milestone-completion as the trigger for the immediate release.
The most significant part of the proposed structure is the escrow, since recalling cash parked in a third party account cannot be done easily. If the mechanism also provides for an independent party to verify the milestone-completion, the instrument becomes a pre-funded, priced commitment, which is genuine de-risking, and worthy of maximum design attention.
The statute also provides unmistakeable policy signalling, which is worth something, however modest. As the state would actually legislate the protection structure, this carries more weight than, say, a well-drafted MOU, though it is not a shield in and of itself.
And it attempts to fill the treaty vacuum - as BITs protections have narrowed, local resolution of disputes before they escalate into international arbitration would be useful to the Centre and investors. This is important also because state-level decisions are internationally attributable to the Union and carry attendant reputational and legal costs the Centre does not control.
Will it Work?
The statute promises more than it can deliver. While it would have to make it clear that contracts or approvals based upon illegal action, corruption, or environmental violations are not protected, beyond that, no Act can be made “un-repealable” - a government cannot permanently freeze successor governments’ law-making powers (including those to repeal the Act) because that would be constitutionally void.
The core point is that an election mandate to reverse even a project decision is about as strong a public-interest claim as an elected government can make! The legal “shield”, therefore, becomes more of an escalated political cost than a guaranteed protection. It may be best, therefore, to draft this as a standard piece of legislation, albeit one which escalates the cost of reversal, not a purported entrenchment.
Next, a Treaty-grade guarantee or commitment made by the state cannot create international arbitration rights or tie the federal government into treaty obligations, and this needs to be made explicit in the provisions of the legislation.
This is linked to Presidential assent, which would entail detailed deliberation. The Bill is likely to be reserved, given the overlap with subjects on the Union and Concurrent Lists: contracts, corporations, treaties. Assent of the Centre would be predicated upon the implications of its own exposure to liability, not just the investor’s, in the case of breach by the state. The Centre’s experience with the counter-guarantee it provided in the Enron case would weigh in on such a decision. And even a recovery mechanism which has prior consent of the state (far preferable to an open-ended guarantee) would require the Centre to underwrite a risk it does not control, and the state could withhold such consent if it has already defaulted.
Is there an Alternative?
Given the goal of the proposed Act, the instrument chosen may not be the right one. A law commits to non-reversal, but the Enron-trap means the Centre would be unlikely to sign off into becoming a passive guarantor of state actions which are beyond its control.
What could be considered, instead, is pricing of political risk instead of promising against it. Insurance can do this, has precedents and works. Sub-sovereign cover is already provided by specialised agencies and the cost of delinquency carries directly to a risky counterparty for defined actions (breach of contract, transfer restrictions, expropriation, etc).
Insurance avoids the Enron-trap, though the Centre has a vital role: it would need to oversee the terms and conditions, a common framework for premium, disclosure, etc, not in underwriting. Insurance would also force disclosure and due diligence, which a statutory guarantee will make skip-able. And it places the incentive where it should be, with the state.
The proposed Act is actually a political risk-insurance, a signalling of sorts, whose instrument or tool is the escrow. The escrow can remain in the Insurance option, to pre-fund incentives which the insurance covers. Pricing the political risk instead of legislating it away makes it harder to repeal or reverse in monetary and political terms, and is more difficult to undo quietly.
How do Other Federal Systems Tackle This?
It is worth asking if any other federal countries have tackled what must be a fairly common political risk for infrastructure projects. China has SEZ statutes operate within a unitary system where policy continuity is not legislated at the subnational level but enforced from Beijing. Canadian provinces have their own securities regulators and investment promotion frameworks but no sub-sovereign binding successor governments in a federal democratic set-up.
A better comparison would perhaps be with the stabilization clauses in mining contracts or the sovereign guarantees India offered in the 1990s. Both have had a mixed record. Modern mining contracts have been increasingly pairing stabilization clauses with political risk insurance as the former have not been enough. This is why it would be prudent to deliberate on the design details, rather than rush into a done-first-in-India race.