Welcome to the eighteenth issue of Ideas and Institutions – a fortnightly newsletter from Carnegie India’s political economy team.

This issue includes an essay on the geopolitics of competition and cooperation in technology and an essay reflecting on a paper by Douglas Diamond and Philip Dybvig on banking crises.

US-China tech rivalry will have spillover effects. India must hedge itself

Even as the Indian government is in the process of overhauling its legal framework for regulating technology, the geopolitical environment is becoming complex. On the one hand, the fight over leadership of high-end technology is becoming increasingly contentious and competitive. There is increasing competition for dominance between the US and China that will have spillover effects for countries like India. On the other, countries like India, parts of Africa and regional blocs like the EU are increasingly recognising areas of mutual cooperation, of learning from and adopting scalable and useful technology in cross-border payments, public health systems and e-commerce.

A concurrent trend is the increasing adoption of digital privacy norms that are solidifying around a core set of privacy principles such as privacy-by-design, purpose limitations and user rights to their personal data. Though some room for debate on these issues remain, discussions on privacy principles have increasingly moved on to the specific designs of data protection laws, and to implementation mechanisms. However, other issues related to the transfer of personal data across nation-states, and national security access to personal data remain contested.

Countries like India will not only have to manage the challenges arising from increasing competition, they will also have to make sure the spillover effects of such competition do not hinder cooperation. Lastly, both competition and cooperation can potentially increase or decrease domestic regulatory autonomy in areas like personal data protection.

Competition

On 7 October, the US introduced export controls designed to thwart China’s efforts to acquire or build advanced computer chips. Last month, US National Security Adviser Jake Sullivan  anticipated this shift in policy by identifying three key technologies – “computing-related technologies, biotech, and clean tech” – as “force multipliers”, and stated that leadership in each is a US national security imperative. A recent piece in Foreign Policy has argued that the move to limit semiconductor exports to China “guarantees a continued march toward broad-based technological decoupling.”

High-end technologies, as the US’s move highlights, will be subjects of increasingly strategic competition, and affect the choices India and other countries make with regard to them. Technology-importing countries will increasingly have to strike a balance between continuing their dependence on countries that possess these technologies, and developing a degree of self-reliance. From the perspective of the global south, it should be wary of the spill-over effects of this competition into areas of cooperation.

Cooperation

Co-existing with strategic competition, there is the trend of increasing cooperation on digital technologies. One such area has been health. The Covid-19 pandemic saw the increased use of technologically-enabled identification systems used for preventing transmission, and for vaccine delivery. India’s COWIN platform was a forerunner in the latter. This experience has accelerated discussions on adoption of such systems for future pandemic prevention and also for public health systems in general.

A high degree of convergence, diffusion and adoption is possible in the areas of digital infrastructure that support data-enabled services like public health, payments, and e-commerce. In India, this infrastructure was created on the back of a digital ID system (Aadhaar). This enabled subsequent transformations in payment systems, vaccine delivery and so on.

Even before the pandemic, the UN’s report on digital cooperation had mooted the idea of universalising access to digital networks to achieve the UN’s Sustainable Development Goals. One of the outcomes of the report has been the idea of “digital public goods” or DPGs. DPGs are open-source software, data, standards and models that incorporate privacy by design and help implement the UN’s SDGs. The Digital Public Goods Alliance registers DPGs that are then promoted by governments and multilateral organisations.

India has been keen to promote the global adoption of its own public digital infrastructure. Aadhaar, IndiaStack, and UPI are being showcased as examples of successful DPGs whose underlying infrastructure can also be adopted by other jurisdictions. At the same time, the mechanisms through which such technologies are deployed, their use cases, and their costs and benefits in different contexts need to be carefully tested and understood. DPGs may have the character of public goods, but their adoption should still be based on market preferences and incentives rather than government fiat.

Data regulation

Discussions on global data regulation increasingly focus on questions of free flow of data vis-a-vis localisation and issues of data access for law enforcement. Countries are divided on whether or not to impose restrictions on the free flow of data. More than 60 countries now have some form of restriction on data flows. India has such restrictions in telecom and finance and is exploring additional localisation requirements.

The country has the leverage to affect the international approach to the free flow of data because it is a large data market. It may be ceding this leverage by not articulating a clear position while the opportunity exists.

However, the focus on the free flow of data assumes that most other parts of personal data regulation are a given – the focus on consumer consent, purpose limitations, security and storage requirements, etc. It will become harder to create domestic exemptions and carve-outs as these principles solidify, especially as global cooperation and competition both assume countries should make similar design choices in their domestic frameworks. For example, the DPG registry requires privacy by design as a basic requirement before it agrees to register a software, product or design as a DPG. This may be a good thing, but it adopts a standard that is hard to deviate from. The absence of a well-formulated and articulated domestic position on many contentious issues in global data regulation can potentially hurt India as more such norms and standards get created.

As the politics of global technology policy become more complex, India will have to do both — manage any negative spillover effects of the US-China rivalry in technology, and take advantage of the increasing possibilities such competition provides. In the field of technology cooperation, India has created standards and products that it can leverage, provided that questions about use cases and deployment mechanisms are answered. Lastly, India should articulate a well-formulated position on contentious issues surrounding data regulation to take advantage of the evolving situation.

This was originally published in ThePrint on 19 October, 2022.

—By Anirudh Burman

Reading Diamond and Dybvig in India

Economists Douglas W. Diamond and Philip H. Dybvig shared this year’s Nobel Prize with Ben Bernanke. The most seminal work by Diamond and Dybvig is on the economics of banking crises. In this essay, I consider one of their classic papers published in 1983 in the Journal of Political Economy, “Bank Runs, Deposit Insurance, and Liquidity” and reflect upon it in the context of India’s banking system.

Bank deposits are callable at par. Depositors expect them to be returned immediately and fully whenever they make the demand. Bank assets are, however, mostly in the form of loans with long maturity periods (in India, about two-thirds of bank assets are in the form of loans) and only a fraction is in liquid instruments, such as government securities, well-trade corporate bonds, and currency. Banks work in spite of this mismatch between callable deposits and illiquid assets because only a small number of depositors usually withdraw their deposits on any given day. They are also able to make profit by maintaining a spread between the interest they pay to the depositors and the one they charge from borrowers (and the return they make on their holding of securities).

If a large number of depositors go to a bank to withdraw their deposits at the same time, the bank would be unable to meet all the demands because it cannot immediately convert its assets into cash, and if it tries to do so, it would incur losses that may lead to its failure. When do depositors rush to get their deposits from banks? When they perceive the bank to be failing. But the key is that it does not matter whether the bank was indeed failing. Since bank balance sheets are opaque, it is difficult for third parties to be sure about their health at any given time. Once panic spreads, a bank run becomes a self-fulfilling prophecy, making even a previously healthy bank fail in the attempt to meet the depositors’ demands. Faced with such a situation, the only options for a bank are—failure or suspension of withdrawal. Banks often choose the latter.

Diamond and Dybvig begin the paper by observing this problem with banks. Written at the time when the savings and loans crisis was beginning and banking was being deregulated in the U.S., the paper presents an elegant model to understand how banks add value to the economy, why bank runs happen, and what could help avoid them. Their model demonstrated that banks can do better allocation than exchange markets between savers and borrowers by “providing better risk sharing among people who need to consume at different random times.” However, the demand deposit contract providing this improvement has multiple equilibria with different levels of confidence in the bank. One of these equilibria is undesirable—a bank run in which all depositors panic and look to withdraw their deposits immediately, including even those who would prefer to leave their deposits in if they were not concerned about the bank failing. Diamond and Dybvig also show that bank runs cause real economic problems because even "healthy" banks can fail, which has negative effects on the economy. They then go on to show that a deposit insurance mechanism offered by the government or central bank can, in certain circumstances, work better than the alternative—suspension of withdrawals.

Deposit insurance mechanism typically involves payment of a premium by the banks (and eventually by the depositors), recognition of the failure of a bank, and payouts to all eligible depositors. Deposit insurance can reduce the chances of bank runs by assuring the depositors about the safety of their deposits with banks. The main benefit of such a mechanism is to prevent failure of healthy banks on account of bank runs motivated by false information about the bank’s health. Functionally, deposit insurance is similar to the “lender of the last resort” activity that central banks sometimes undertake during crises, when the interbank markets are not able to provide liquidity. Both these mechanisms protect banks that are solvent but are facing a liquidity crisis. Deposit insurance is not free. Even if it is efficiently priced, it creates a moral hazard—the incentive of depositors to exercise prudence while choosing a bank is considerably weakened. So, it should be accompanied by regulatory oversight.

In the paper, Diamond and Dybvig gave elegant explanations for phenomena like banking crises and deposit insurance, which had existed for long. Their achievement was in explaining something that already existed. As Raghuram Rajan noted in a recent tribute to Diamond and Dybvig, a major contribution need not be complicated. It could be simple. Similarly, a major contribution need not propose something new in the world but may offer a better explanation for something that already exists. Most social science research is about explaining phenomena that already exist.

Now, let’s consider India’s banking system. In India, among the scheduled commercial banks, government-owned banks account for about two-thirds in terms of share of bank deposits. These banks run maturity mismatches, but the depositors do not worry about the bank failing. They see these banks as extensions of the government’s balance sheet. The role that Diamond and Dybvig suggest deposit insurance should play is played by the implicit guarantee by the government to save these banks, as the depositors expect government-owned banks to be recapitalized before they fail. This has many consequences.

First, there is a difference between the backstop available to private banks and that available to government banks, which becomes a source of fragility for the former. Deposit insurance covers all banks, but government banks have a stronger backstop—the government’s fiscal resources. So, in a crisis, the depositors having more deposits with private banks than are covered by deposit insurance have an incentive to shift their deposits to public sector banks in search of safety. Viral Acharya and co-authors show this effect in a recent paper. They also argue that this leads to poor resource allocation in the aggregate.

Second, as Urjit Patel has argued, the problem of moral hazard is aggravated in public sector banks due to the enhanced backstop that they enjoy. While deposit insurance creates moral hazard for all except those who place deposits larger than the deposit insurance limit still have an incentive to make a careful choice, with government banks, no such incentive exists as there is no cap on the backstop as such.

Third, since the basic rationale for prudential regulation of banks—assuring the depositors that the bank is healthy—does not really exist for government banks, there is little incentive to allocate regulatory resources to these banks. Regulatory authorities, like all organizations, face resource constraints, which necessitate allocational choices. For the Reserve Bank of India (RBI) as the banking regulator, it does not make sense to give equal priority to regulating government banks and private banks. Further, the special backstop available to government banks also encourages the regulator to delay the recognition of balance sheet problems. Timely recognition of problems and corrective action can prevent bank runs triggered by suspicion of bank failure and help with quick resolution when a bank cannot recover. Neither of these is a factor when it comes to government banks. So, when the problems are eventually recognized, there is some exchange of tough words between the government that foots the recapitalization bill and the RBI, but things again go back to how they were earlier.

Fourth, some inefficiency is created by covering government banks in the deposit insurance mechanism even though they are not expected to use it. Government banks pay premia for deposit insurance. The deposit insurance fund now has about Rs. 1.5 trillion, which is about 1.8 percent of the insured deposits. However, since the deposits with government banks are not expected to use this mechanism, the fund is excessive. This is an inefficient allocation of resources.

A key insight from the Diamond-Dybvig paper is that efficiency gains and crises are two sides of the same coin. Banks improve efficiency of resource allocation by running maturity mismatches, but this can sometimes get them into a crisis. Government interventions should seek to mitigate the latter while doing as little harm to the banks’ contribution as efficient resource allocators as possible. Government ownership of banks obviates bank runs but creates significant distortions in resource allocation done by banks in the economy. So, there may be a silent crisis of inefficiency. It may be better to count on a combination of regulation, resolution, and deposit insurance to ensure safety and soundness in the banking system.

—By Suyash Rai

Carnegie India is organizing the seventh iteration of the Global Technology Summit (GTS) 2022: Geopolitics of Technology, co-hosted with the Ministry of External Affairs from November 29 to December 01.

For more information on the agenda and to attend the discussions virtually, please register here.