BITs: Credibility comes at a price for attracting foreign investment

The economics community has woken up to the errors of our ways on bilateral investment treaties (BITs). In essence, it’s efficient for the Indian state to pay out a chunk of money every year in exchange for stabilising very large sums of money involved in foreign direct investment (see https://bit.ly/4aBpoX6 from 2021). It is in India’s best interests to offer deep protections to foreign investors, even greater than those promised to domestic investors. The numbers we are looking at are perhaps payments of $20 billion a year, to aggrieved foreign investors, in exchange for a stock of foreign direct investment of $2 trillion. An Indian state which pays in return for institutional failure is also one that will be more inclined to undertake policy reforms, which would yield outsized gains for society. 

With these basics in place, we should pursue two questions. How do we avoid a repeat of problems like the White Industries case? Even with better treaties, will private investors actually come? 

Why do countries sign investment treaties at all? Nobody demands a treaty before building a factory in Germany. German courts, German administration and German politics inspire confidence on their own. Investment treaties exist because developing countries have poor state institutions. The treaty is rented credibility: A way to attract capital today while the domestic institutions are still under construction. Before capital arrives, a government has every incentive to offer good terms, but there is a credible commitment problem. After the capital is sunk in a power plant or a mine, the bargaining power reverses, and the same government has every incentive to revise the terms. Investors know this in advance. A government that cannot make its promises binding will find that investors charge it for the risk, in the form of higher required returns, or stay away altogether. 

Rented credibility, however, is not a free lunch. Consider what actually happened in White Industries. An Australian firm won a routine commercial arbitration against a public-sector company. Enforcement then sat in Indian courts for nine years. An international tribunal held that this delay breached our treaty obligations, and the Indian state paid. It is tempting to read this as a treaty malfunction. It was the opposite. The tribunal did precisely what the treaty was for: It attached a price to the failure of our domestic institutions. The award was a feature, not a bug. The Indian state response was to shoot the referee. 

When we go back to treaties, we should think more about how this will work out. It will help to have a pre-funded, ring-fenced pool to honour adverse awards. This ensures prompt payment so that investors don’t need to threaten to attach Indian assets abroad in order to collect what a tribunal has awarded. And it works as a bond: Because every drawdown is visible and must be replenished from the Budget, every breach acquires an immediate fiscal cost that Parliament and the Press can see, rather than a litigation liability buried for a decade. At a conceptual level, all of us need to understand that there will be a regular flow of payments, owing to BITs, which is the insurance premium that is required to achieve the privilege of inbound foreign direct investment. 

This is useful in a steady state. If the Indian state engages in a flow of economic reforms, this insurance cost would actually go down. This requires work on taxation, contract enforcement that concludes within a commercial lifetime, and predictable regulation. None of this can be done by the finance ministry alone. Treaty promises are made by one corner of the Union government, but they are broken across the Indian state: By tax administrations, by independent regulators, by state governments whose electricity distribution companies sign the power-purchase agreements. A treaty places the Union government in the position of being an insurance provider, while using all its powers to foster an economic reforms process in parallel. 

This brings us to the second question. A treaty must convince two parties in sequence. First the counterparty government, and, second, that country’s private investors. Governments sign treaties for many reasons, including diplomatic ones; Capital does not follow signatures. A well-drafted treaty gets India past the first audience. The second audience does not read the treaty text. It reads the track record. 

The track record of the Indian state is not good. We amended tax law retrospectively, reaching back half a century, to reverse a Supreme Court judgment. We withdrew from 75 treaties after losing cases under them. We upended settled tax positions. We changed duties in ways that rewrote the economics of investment already made. Each decision was individually defensible to someone in a hurry. Together they form a pattern, and investors have priced the pattern. Credibility is cheap to keep and expensive to rebuild. 

Put yourself in the investor’s chair. A new treaty is on offer. What comfort does it give when the next revenue emergency arrives, or the next court ruling goes against the government? The investor remembers that retrospective tax was repealed in 2021 only after enforcement actions abroad made it untenable. A retraction under pressure signals much less than a promise never broken. The investor also remembers that the last set of treaties did not survive contact with adverse awards. The new treaty will be read against this history, not in place of it. 

What would change the calculus? We need to do a genuine BIT text that promises deep protections and not a niggardly text, so as to portray sophisticated policy capabilities. And we should fund the compensation pool. This machinery should hum along for a decade without friction, with a steady flow of insurance payouts that are free of nationalism or outrage. We will build credibility through deeds that match the words of the BIT, for years on end. Investors will only gain confidence slowly. It takes decades of good behaviour to reduce the Indian risk premium materially.

The authors are researchers at XKDR Forum