Search interesting materials

Tuesday, April 07, 2020

RBI vs. Covid-19: Understanding the announcements of March 27

by Rajeswari Sengupta and Josh Felman

When the first cases of Covid-19 started getting reported in India, the economy was already in a precarious situation and the space for a macroeconomic policy response was limited. Even so, the Reserve Bank of India has come up with a number of initiatives to combat the crisis. In this article, we consider the broad principles that should guide the macro policy response, summarise the RBI announcements of March 27, and assess the announcements against the principles.


The "corona crisis" consists of three interlinked problems: a health shock, an economic shock following from the lockdown, and a global economic downturn. Each one of these shocks on its own is significant. Put together, they have created considerable pressure upon policy makers to act quickly and decisively.

Coming up with an effective policy response is not an easy task. For one thing, the corona crisis poses some exceptional difficulties. It is clear that the human and economic toll will be serious, but it is unclear how long the crisis will last or how deep the damage will be. And without a clear understanding of the size and duration of the problem, it is difficult to know how to calibrate the policy response. For example, monetary easing could take a year to have a significant effect. By then the problem might be over, and inflation might have re-emerged, at which point painful measures would be required to bring it down. This is not just a theoretical possibility, it is precisely what happened in the aftermath of the Global Financial Crisis in 2009-13.

Principles of policy response

Policy making is difficult in the best of times. It is harder in exceptional times, when there is pressure for quick actions, grounded in reduced analysis. It is in exceptional times that the toolkit of good governance becomes even more important:

  • The lowest cost actions are those which are grounded in root cause analysis.
  • Each action needs to be carefully weighed in terms of the costs and benefits imposed upon society.
  • As much as possible, policy responses should be fitted into existing rules and frameworks.
  • All state actions should be preceded by public debate and consultation.

This toolkit is a valuable discipline, an institutionalised application of mind. Why is root cause analysis important? Consider the problem of weak banks lending to firms in recent years. From 2018 onwards, RBI has been trying to address this problem by injecting more and more liquidity into the banking system, in the hope that banks would deploy these resources and lend more (link, link, link). But liquidity issues were not at the root of the problem, the twin balance sheet (TBS) stresses at firms and banks were the real issue. Bank lending has also been discouraged by the government’s measures to investigate and prosecute bank officials for their lending decisions. As a result of these factors, banks have remained reluctant to lend to the private corporate sector, curtailing credit to industry to a year-on-year growth rate of just 0.67 percent in February 2020.

As an example of poor cost-benefit analysis, consider the regulatory decisions after the Global Financial Crisis. At the time, it was felt that exceptional times called for exceptional deviation from prudent financial regulation. A series of restructuring schemes followed, allowing banks to postpone NPA recognition and hide bad news. With the benefit of hindsight, we know that this restructuring worked poorly, and helped prepare the ground for the twin balance sheet crisis of 2011-2020.

As for respecting frameworks, there is a temptation during crises to abandon rules and resort to discretion. But recent experience warns us that "temporary measures" are often difficult to reverse (consider the 2010 fiscal stimulus), while inadvertent consequences (such as NPAs) are difficult to resolve. More fundamentally, temporary measures disrupt the stable configuration of expectations of economic agents, which hamper the recovery. It takes many decades of consistent behaviour in a rules-based framework to shape the rhythm of the working of state institutions, to build up policy credibility. This credibility can be rapidly dissipated.

Hence, policy makers need to proceed cautiously.

The March 27 announcements

It is in this context that we need to examine the March 27 announcements. Four bold actions were taken, following an "out of cycle" i.e., unscheduled Monetary Policy Committee (MPC) meeting:

  • The repo/reverse repo rates were cut by sizeable amounts, to 4.40/4.00 percent from 5.15/4.90 percent. The 91-day treasury bill rate, which measures the de facto stance of monetary policy, dropped to 4.31 percent from 5.09 percent on 26 March.

  • Ordinarily, banks can borrow on a short-term basis from the RBI using the repo window. To supplement this facility, a new `targeted long-term repo operations' (T-LTRO) mechanism, with a limit of Rs.1 trillion, was announced. Banks may find this attractive because they do not have to mark to market the investments made with these borrowed funds for the next three years. However, there is a condition: the money that is borrowed here must be deployed in investment-grade corporate bonds, commercial paper, and non-convertible debentures, over and above the outstanding level of their investments in these bonds as on March 27, 2020.

  • The cash reserve ratio (CRR) was reduced by 1 percentage point, bringing it down to 3% of deposits ("net demand and time liabilities"). This is the first time the CRR has been changed in the last 8 years. RBI's initiatives appear to be motivated by the desire to increase liquidity, as their statement highlights that these measures will free up Rs 3.74 trillion in banks' funds.

  • Banking regulation requires banks to recognise and provide for a loan when there is a delay in payment. According to the Prudential Framework for Resolution of Stressed Assets, banks are required to classify loan accounts in special mention categories in the event of a default. The account is to be classified as SMA-0, SMA-1 and SMA-2, depending on whether the payment is overdue for 1-30 days, 31-60 days or 61-90 days, respectively. RBI has now modified this regulation, so that banks can offer a moratorium of 90days for term loans and working capital facilities for payments falling due between March 1, 2020 and May 31, 2020. However interest on the term loans will continue to accrue during this period. If a firm applies for and receives a moratorium, the loan account in consideration will continue to be recognised as a standard asset and the SMA classifications will no longer apply. Interest on term loans will continue to accrue during this period. 

Analysing the monetary policy announcements

Monetary policy is most effective when economic agents understand and can anticipate the behaviour of the MPC. This process of learning and understanding is still underway, given that India is in the early years of building up the credibility of the inflation targeting framework and the MPC process. So, one would have expected that the MPC statement would go into great details and spell out its macroeconomic forecast, explaining why it believed the 75 basis points rate cut was consistent with its commitment to the 4 percent inflation target.

However, tt did not explain the rate decision in the context of a revised inflation forecast, or any other element of a macroeconomic forecast. It did not offer a justification for the magnitude of rate cut chosen.

Since the rate cut announcement was not couched in the standard IT framework, the public does not have the assurance that the rate cuts will be reversed when inflation begins to rise again. To remedy this problem, monetary policy actions could henceforth be couched in terms of this framework, as a way of assuring the public that the RBI is keeping its eye on this critical objective, and that the mistakes of the past will not be repeated.

Analysing the banking regulation announcements

We know that the corona crisis is a temporary shock. Standard economic theory tells us that the optimal response to a temporary shock is for (viable) firms and households to obtain financing, so that they can tide over the difficult period. Over the next few months, three categories of firms will emerge: a) firms that are able to pay their dues throughout the crisis period, b) firms that are fundamentally viable and can survive provided they are given adequate credit support, and c) firms whose business is faulty and who should become bankrupt as a result of this shock.

It will be important for the banks to distinguish among these firms. Banks should ideally do nothing with firms in category (a), extend credit support to firms in category (b), and take the firms in category (c) to the insolvency and bankruptcy courts as and when that process resumes.

Under the 27 March package, the RBI has given regulatory approval to banks and other lending institutions to decide which of their customers needs a 90-day deferral. This decision, to allow banks but not require them, to grant moratoria is a good one, as it allows banks to distinguish among the three types of firms.

However, the plan is not without drawbacks.

  • No mechanism has been created to classify the loans that will be rescheduled, so transparency has been lost. Investors – already nervous because of accounting surprises at Yes Bank and other financial institutions – will consequently provide capital only at a cost marked up to reflect this information risk premium. And this increase in banks’ costs will be passed on to the borrowing corporate sector.
  • Moratoria will create problems for pass-through certificates, i.e. loans that have been bundled as bonds and sold to mutual funds, because there are no provisions in these certificates for loan rescheduling.
  • Finally, and most importantly, there is no clarity on what happens once the moratorium period is over. How will banks clean up the mess that will be created later, as many of the firms which benefited from the moratorium end up defaulting? There will be a new wave of NPAs, which we know from experience will be difficult to resolve.

There is also a risk: now that a "temporary" moratorium has been introduced, there will be pressure for it to be extended again and again. If the RBI is unable to resist, we will quickly find ourselves back in the 'extend and pretend' era of post-2008. Banks, investors, the RBI, will all be navigating in a fog, since no one will know – and hence, be able to deal with -- the true size of the bad loan problem.

In other words, under the current design, there are risks that the costs of the moratoria could end up exceeding the benefits. Is there an alternative? In fact, two supplementary actions could reduce potential costs, while preserving the benefits.

First, RBI could announce that firms seeking a moratorium would be marked in a separate category. This would give transparency regarding the true financial situation of the banks. There will also have been a bit of a stigma for borrowers, helping to preserve debtor discipline. If a firm has no choice, it will still postpone repayment. But if a firm can afford to pay, it will do so, in order to escape the stigma.

Second, forward planning could help deal with the consequences of the inevitable surge in defaults. Even before the corona crisis, bankruptcy cases were taking far longer than what the law stipulates. Large cases were taking several years to resolve. If this situation is not addressed, there is a risk that large sections of the economy will be tied up in bankruptcy courts, making it impossible for the economy to return to normal, even after the virus abates. To make sure this does not happen, the Insolvency and Bankruptcy Code (IBC) needs to be reformed urgently in order to ensure faster and effective resolution. Such reforms would also have an immediate benefit: banks would be more confident in lending now if they knew the IBC would not be overwhelmed by cases after the crisis is over.

Reviving credit growth

The need of the hour is to revive credit to the private corporate sector. But the marginal benefit of the RBI adding more liquidity to a system that is already in a surplus mode is not clear. This strategy has already been tried, without success. It is unclear why it would work now, especially now that uncertainty about firms' prospects has only increased.

For a proper root cause analysis, let’s go back to economic fundamentals. Consider a loan decision. When a bank decides to approve a loan, it is performing two functions simultaneously: it is assuming risk, and it is allocating capital. In the current circumstances, it is still possible for banks to allocate capital. They can assess which firms are more likely to be hit badly by the crisis and which firms are going to be less affected. That is, banks can figure out the relative risk. The problem for the banks is that right now they cannot assess the absolute level of risk, because they do not have any idea about how long the crisis is going to last, or how deep the crisis is going to be. And this shock has come at a time when banks have already become risk-averse given the last few years of balance sheet problems. Hence, it is difficult for them to lend, especially to new customers.

In these circumstances, giving them liquidity, exhorting them, coming up with any number of subsidy schemes, will not work. But there is a possible solution. The government-- not the RBI -- could relieve the banks of the burden that they cannot manage: the burden of risk.

This can be done through a mechanism as follows. The government can capitalise a fund which will then give loan guarantees. The scheme would have some selection criteria, say MSMEs that have been current on their bank loans. It would also specify the maximum rupee amounts per firm, pegged say to the annual revenues of the company. Once the eligibility criteria are specified by the government, the actual selection of the firms would be done by the banks. They would identify the best firms, originate the loans, and then apply to the fund for guarantee coverage. The banks should be charged a fee for this, to discourage them from using the fund unnecessarily.

In this way, we could use the law of comparative advantage to obtain better economic outcomes: the government would do what it does best in crises, namely bearing risk, while the banks would continue to do what they do best, namely allocating capital.


The RBI’s March 27 announcements were bold and decisive. In particular, the reduction in the repo rate by 75 basis points will provide significant debt service relief to firms and households. This is a welcome measure, at a time when their cash flows are going to be seriously strained. The announcement that banks will be allowed to grant temporary debt moratoria to firms and households could also prove a major help, for exactly the same reasons.

That said, the announcements could have been better grounded in basic principles. The root causes of the banks’ reluctance to lend have not been addressed. At the same time, the way the policy actions were designed and announced run the risks of damaging confidence in the existing frameworks. The public may not be so sure that the authorities remain committed to preserving low inflation or financial stability. Nor is it clear that there is an "exit strategy", to ensure that the defaults will be resolved expeditiously, allowing the economy to return quickly to normal, once the health crisis is over.

There is still time to clear up these ambiguities, and remove any doubts. Initial actions can be followed by supplementary steps, and initial problems can always be remedied. This will take careful root cause analysis, cost-benefit calculations, and a determination to reinforce existing policy frameworks.

Josh Felman is a researcher specialising on India. Rajeswari Sengupta is a researcher at IGIDR.

Monday, April 06, 2020

Street-level officials in India's Covid-19 response

by Smriti Parsheera.

In a federal union of 28 States and 9 Union Territories, divided into over 700 Districts, translating policies into effective implementation is testing in the best of times. But when a country of 1.3 billion people comes to a near halt, as witnessed after the Prime Minister's announcement of a 21 day lockdown, the implementation challenge becomes all the more significant. The events that have unfolded over the past few weeks remind us that policy outcomes are determined not just by decision-makers sitting in the corridors of power in the Central and State Governments but also by the street-level officials who intermediate the everyday relationship between the citizen and the State.

The term "street-level bureaucrats" was coined by Michael Lipsky in 1980 to describe the host of front-line public workers who interact directly with citizens, often enjoying a substantial level of discretion in the execution of their functions (Lipsky, 2010). In the present context, this includes various district-level officials, police officers, social workers and other members of the local administration, all of whom have have become the first point of contact between the individual and the State's constantly evolving response to the Covid-19 crisis (See Shah & Misra, 2020). In Lipsky's analysis, street-level workers often lack the time, information, or other resources that are necessary to properly respond to individual cases. They therefore invent their own devices "to cope with uncertainties and work pressures, effectively become(ing) the public policies they carry out." (Lipsky, 2010). In the present situation, these pressures would include shortage of staff and protective equipment, while facing greater risks of exposure; uncertainty about the course of the virus and the policy response; and reports of attacks faced by health workers and police personnel while discharging their functions.

The Covid-19 situation has many of the ingredients of what Kelkar & Shah (2019) define as a "hard policy problem" -- problems that involve some combination of a high level of discretion, a high number of transactions, high stakes and high secrecy. Three of these four sources of complexity are present in the lockdown.

The stakes involved are high on account of the challenges posed to the life, health, livelihood and liberty of citizens.

The response involves a fairly large volume of transactions, including provision of health care services, tracing and isolating contacts, and ensuring adherence with quarantine and lockdown conditions.

In some cases, Government policies have reduced the level of discretion that might have otherwise been available to front-line actors. For instance, stating that only those who meet the specified inclusion criteria will be eligible for testing reduces the discretion in the hands of doctors and testing centres. However, as we illustrate ahead, there are many transactions where the officials on the ground enjoy a substantial degree of discretion in the implementation of the stated policy.

Role of street-level officials in the lockdown

To implement the lockdown, various States have issued regulations and orders under the Epidemic Diseases Act, 1897, imposing restrictions on the operation of establishments and movement of people. At many places, orders have also been issued under Section 144 of the Code of Criminal Procedure, 1973 (CrPC). This provision allows District Magistrates and Sub-divisional Magistrates to "direct any person to abstain from a certain act or to take certain order with respect to certain property in his possession" if this is required for reasons of health, safety, public order, etc. These measure prohibit individuals from stepping outside their homes unless it is for providing or availing an essential service, and from assembling in groups. A list of essential services, which includes groceries, hospitals, chemists, e-commerce services, etc., has been indicated in the orders.

While the letter of the orders might seem to make it clear as to what are the permitted and restricted activities, every police officer who encounters a potential detractor on the street gets to make some important choices. The officer will decide whether the person should be heard and, if appropriate, left off with a reprimand. Should legal action be initiated for violation of Section 144 of the CrPC or relevant provisions of the Epidemic Diseases Act and the Indian Penal Code? Or, as seen in several reports from the initial days of the lockdown, should violators be given an instant taste of the law through the use of brute force and humiliating physical punishments? While it may be easy, and legitimate, to classify these actions as forms of State violence, it is also important to recognise them as the actions of individual actors (acting on behalf of the State).

Visuals of overcrowded boarding points and packed buses carrying stranded migrant workers in Uttar Pradesh and Delhi have been a stark reminder of the gap between the policy messaging on mandatory social distancing and the ground reality. What role did the officials on the ground play, or should have played, in this process?

In a symbolic distancing of the policy from the implementation, the Ministry of Home Affairs reacted with the suspension of certain officials in Delhi who were held to be responsible for the lapses. The Ministry also passed an order imposing a country-wide ban on further movements of migrant workers. This order mandates State Governments to provide food and shelter to those in need and quarantine facilities for migrants arriving from outside. Again, there are bound to be many differences in the ways in which this order will be implemented by the police and local administrators in different parts of the country. For instance, some areas may still permit movements under exigent circumstances or allow relaxations for those who are already in transit. Similarly, the facilities available at the shelters, including regard for the differential sanitation and safety needs of women, will also be determined at the ground level.

Front-line officials are also playing an important role in determining the attitude towards individuals coming from outside the region or those who have had contact with coronavirus patients. Are they to be regarded as "Covid-19 suspects" who will be shunned and disinfected, as seen in the Bareilly town of Uttar Pradesh? Or are they just individuals who face a higher risk of infection and therefore need specific monitoring and care? The local administration in Himachal Pradesh has, for instance, tried to adopt the latter approach but, like in many other parts of the country, this comes with insufficient checks to safeguard individual privacy.

The author's own, admittedly privileged, interaction with the local administration in District Kullu of Himachal has been that the authorities are in regular touch with those placed under home quarantine. This includes phone calls and visits from teams consisting of doctors, the local police, ASHA workers, panchayat members and land revenue officials. Without exception, the interactions have been courteous and reflect a sense of responsibility towards ensuring the health of the individual and the community. These positive steps are, however, marred by the awareness that the personal data of the travellers, like their name, phone number, address, etc., is being freely circulated on local WhatsApp groups. Similar reports have also emerged from places like Delhi, Nagpur and Ajmer. While some States, like Karnataka, are deliberating pushing out this data, the violations in many other cases are more likely the result of indiscreet sharing by those who gained access to the data in the course of their official functions. Given that India's legal and enforcement framework around data protection remains deficient, urging local officials to prevent the dissemination of this information might be the most effective check at this point.

Finally, a number of street-level actors will also have to step in to ensure that the relief measures announced by the Finance Minister actually end up reaching the intended beneficiaries, in a fair and timely manner. To the extent that these benefits have been tied to existing schemes, functionaries like operators of fair price shops and gas distributors under the Ujjwala scheme, who discharge public official-like functions, will be responsible for intermediating access to the promised entitlements. Through the last decade, the focus of systems like Aadhaar has been on increasing accountability and reducing the discretion available to such front-line actors. The Covid-19 crisis might, however, be a time to temporarily suspend the insistence on precise targeting through technical tools, even if this leads to some increase in the discretion available to street-level officials.

The way forward

The examples given above illustrate how officials who may not traditionally be regarded as "policymakers" have a profound impact on the ways in which most Indians will navigate their lives in the coming months. How we come out of this unprecedented crisis will depend not just on the government’s policies, its available resources and citizens' cooperation but also the individual choices made by various front-line actors. Policy design for Covid-19 should take this into account, applaud the role of street-level officials, and compel them to act in ways that are compassionate and preserving of individual dignity. This perspective would be a useful addition to the Prime Minister's public addresses, the text of official orders, all the way to local coordination meetings at the district level. Traditional and social media also need to play a more responsible role in bringing out both the violence being inflicted by the authorities, as well as the positive stories of compassion and care shown by some front-line workers.

Recognising the complexities of implementation at the ground level is one element of better planning. It is a reminder that loose wording of legal instruments, of the kind that we often see in India, can end up resulting in unplanned and undesirable outcomes, from the perspective of policymakers and the public. Therefore, instead of treating the implementation stage as something that is distinct from policy planning, the planning process itself needs to account for the challenges of policy-making in high-stakes, discretionary and transaction-intensive scenarios. While this may not always be possible because, as noted by Pritchett & Woolcock (2004), practices that are discretionary and transaction-intensive in nature are by definition not amenable to be standardised and (easily) replicated, policy-makers should try to anticipate the incentives and actions of front-line agents, to the best extent possible.

Another appropriate response to the present crisis would be to initiate a systematic review of the legal framework under which the various elements of the current policy actions have been taken. The Epidemic Diseases Act, 1897, which has been the basis of the Covid-19 response by many State Governments, is a case in point. It is a skeletal legislation that consists of only five sections including one that allows the State Government to take any measures to prevent the outbreak, or spread, of a "dangerous epidemic disease", a term that is not defined under the law (Kaur, 2020). In fact, the law does very little other than specifying the broad powers of the Government, immunity for any of its actions, and legal consequences for non compliance. In particular, it does not define the scope of the measures that may be prescribed under it or lay down requirements like adherence to basic human rights, proportionality of adopted measures, transparency, accountability or redress mechanisms. A few years back the Central Government had initiated a move to replace this law with the proposed Public Health (Prevention, Control and Management of Epidemics, Bio-terrorism and Disasters) Act, 2017. Critics were, however, of the view that the proposed draft remained equally deficient in most respects (Rao, 2017). Similarly, there is a need to develop standard operating procedures on the acceptable conduct of police forces in times of public health emergencies, and incorporating these in police training programmes.

We recognise that the monopoly on violence, and not compassion, is at the essence of the State. This makes it all the more important to push for appropriate checks and balances in the exercise of any emergency powers by the State. When it comes to the role of street-level officials, their actions are motivated not only by the directions issued by the State but also by the incentives, uncertainties, pressures and threats faced by them. Better policy planning should account for these factors and address them while designing the policy response.


Kelkar & Shah, 2019: Vijay Kelkar and Ajay Shah, In Service of the Republic: The Art and Science of Economic Policy, Penguin Allen Lane, 2019.

Kaur, 2020: Harleen Kaur, Can the Indian legal framework deal with the Covid-19 pandemic? A review of the Epidemic Diseases Act, Bar and Bench, March 27, 2020.

Lipsky, 2010: Michael Lipsky, Street-Level Bureaucracy: Dilemmas of the Individual in Public Services, 30th Ed., Russell Sage Foundation, New York, 2010.

Pritchett & Woolcock: Lant Pritchett and Michael Woolcock, Solutions When the Solution is the Problem: Arraying the Disarray in Development, World Development, 32(2), 2004, 191–212.

Rao, 2017: Menka Rao, A new bill on public health emergencies allows for dubious restrictions of citizens' liberties, Scroll, March 31, 2017.

Shah & Misra, 2020: Kadambari Shah and Prakhar Misra, Covid-19: Importance of street-level bureaucracy to fight the pandemic, Moneycontrol, April 6, 2020.

Smriti Parsheera is a lawyer and technology policy researcher. She would like to thank Ajay Shah and Rudra Chaudhuri, and an anonymous referee, for valuable inputs.

Release of v2.0 of the Exchange Market Pressure dataset associated with PFM 2017.

by Joshua Felman, Madhur Mehta, Ila Patnaik, Ajay Shah.

The idea of Exchange market pressure (EMP) was introduced by Girton and Roper (1977). It suggests measurement of the total pressure on the exchange rate, some part of which is visible as the change in the exchange rate, and the remainder is resisted by currency trading of the central bank. Many researchers have worked on devising EMP estimators, the most prominent of which are Eichengreen et al. (1996), Sachs et al. (1996), and Kaminsky et al. (1998). EMP measures have been utilised in thousands of papers in international finance and macroeconomics.

In Patnaik et al. (2017) we proposed a new method for calculating EMP which attempts to overcome the well known problems of conventional EMP measures. Alongside this paper, a cross-country dataset was released, which ran from January 1996 to May 2017.

We have done a second release of this dataset, which carries these series forward to November 2018. In this dataset, which is numbered as v2.0, we have 135 countries. On the web page, we have a CSV file of the dataset, and also the few lines of R code that get you going on using it. The URL of this web page will be stable, and the next release will come out with further updation of the dataset in a few months.

In the v1.1 dataset, due to lack of annual macroeconomic data for some countries, the rho values were computed with erroneous confidence bands, which consequently affected the EMP values. We have corrected this error.

In the following paragraphs we compare version 2.0 of the new measure of EMP further for four countries, namely, India, China, Russia, and Brazil against the conventional EMP measure of Eichengreen et al. (1996). This helps give intuition about the gains from the new measure.

Example: EMP for China

In the figure above, the two grey rectangles in the conventional EMP measure plot for China are periods where the value of the EMP measure are near infinity. The new EMP measure does not have this problem.

The new EMP measure shows that in years prior to the Lehmann Brother's collapse, there was persistent appreciation pressure on the RMB. After the Lehman default, a sudden shift in the exchange market pressure can be seen. These phenomena are not present in the conventional measure.

A significant event for China occured on 12 June 2015, when a financial crisis began. The new EMP measure shows this depreciation pressure better than the conventional measure.

Example: EMP for India

In India's case, the Lehman default in 2008 brought about a sharp depreciation in the value of indian rupee. This story is nicely told in the new EMP measure. The conventional measure suggests that there was a switch from depreciation to appreciation pressure at that point.

Prior to the taper tantrum of 2013, the entire year of 2012 had high volatility in the rupee exchange rate. In the tantrum, there was high pressure on the rupee value to depreciate. These facts are well-represented in our measure of EMP, and consistent with a detailed understanding of that period, as opposed to the conventional one.

Example: EMP for Russia

In the case of Russia, the conventional measure fails to show the magnitude of the effect of the Lehman default, the taper tantrum and the Russian invasion of Ukraine in 2014. The new EMP measure has the correct features: that these events imposed depreciation pressure on the rouble.

Example: EMP for Brazil

In the case of Brazil, the Lehman default and the taper tantrum of 2013 imposed high depreciation pressure on the Brazilian real, in the new EMP measure, but not in the conventional measure.


Eichengreen, B., Rose, A., Wyplosz, C., 1996. Contagious Currency Crises , Technical Report. National Bureau of Economic Research.

Patnaik, I., Felman, J. and Shah, A., 2017. An exchange market pressure measure for cross country analysis . Journal of International Money and Finance, 73, pp.62-77.

Desai, M., Patnaik, I., Felman, J. and Shah, A., 2017. A cross-country Exchange Market Pressure (EMP) Dataset . Data in Brief.

Girton, L., and Roper, D., 1977. A monetary model of exchange market pressure applied to the postwar cnandian experience. American Economic Review, vol. 67, pp.537-538

Sachs, J., Tornell, A., Velasco, A., 1996. Financial crises in emerging markets: The lessons from 1995. National Bureau of Economic Research.

Kaminsky, G.A., Lizondo, S. and Reinhart, C.M., 1998.Leading indicators of currency crises. Staff Papers-Int. Monet. Fund (1998), pp. 1-48.

We thank Bhavyaa Sharma, Shekhar Hari Kumar and Namita Goel for their work on this release.

Friday, April 03, 2020

Holding their breath: Indian firms in an interruption of revenue

by Renuka Sane and Anjali Sharma

India is in a state of lockdown with only essential goods and services getting exemption. This has given a teachable moment on the role of liquid cash in corporate financial policy.

Most firms will see a varying degree of a revenue shock. Expenses such as salaries, rents and interest on debt will continue to be incurred in order to survive. In this article we ask: How many days of liquidity cover is there, in the large non-financial firms, to be able to meet a certain threshold of minimum expenses in the absence of any revenue?


We extract data for all non-financial firms from the CMIE ProwessDX database, updated in December 2019, for three years: 2016-17, 2017-18 and 2018-19. Data on 2018-19 is not yet available for a large proportion of the firms. For most firms, therefore, we rely on the two older years of data.

We drop firms that report negative assets, negative income and negative sales. This yields a sample of 16,300 firms which are observed in the first two, and possibly in the third, year. For each of these firms we extract information from the database on: (1) identity variables such as age, industry, listing status and ownership, (2) income and expenses from the profit and loss statement, and (3) liquid assets i.e. cash and bank balance and marketable securities from the balance sheet.

We do not know the state of the firm in March 2020, when the lockdown came. We estimate the state of the firm using the mean of two or three years seen in the data (depending on data availability). The recent three years have been a period of stability of the nominal values, i.e. low growth, thus making these simple averages more useful.


We make four admittedly extreme assumptions:

  1. A 100% sales shock for all non-financial firms. This is an extreme scenario. Most firms will face some degree of sales decline which may vary by the sector the firm is in. However, our assumption gives us a worst case picture for the hurdle that cash holdings will have to overcome.

  2. Liquid assets only include cash and marketable securities. We do not include loans from the banking sector or receivables for the following reasons:

    • Indian firms are generally credit constrained. In the months prior to the lockdown, banks and non-bank finance companies (NBFCs), which are the mainstay of lending in the economy, were already under considerable stress. Their overall lending, and especially their lending to firms had stagnated, with negative consequences for liquidity in the credit markets. Looking forward, many firms will find it difficult to obtain debt capital, given the difficulties of the financial system. Hence, it's useful to analyse how liquid assets -- alone -- may carry some firms across the shock.

    • Our rationale for not including receivables is the suddenness of the lockdown. If firms had any notice that a lockdown was imminent, they would have made efforts to accelerate their receivables collection. But since the lockdown was announced and put in force with no notice, we assume that firms were not able to engage in this strategy.

  3. A 50% realisation of the value of marketable securities in the book. Market conditions are also affected by the lockdown; the price and impact cost of marketable securities has deteriorated. The vast majority of liquid assets is cash, therefore the results are not very sensitive to this assumption.

  4. The core expenses required to stay alive. We define the core expenses as: (1) salaries for employees and outsourced staff, (2) rentals, (3) utility costs, (4) other costs such as IT and communication expenses, insurance premiums and auditors fees, and (5) interest costs on debt.
    All firms are looking at ways to reduce the wage expenditure, by cutting wages and headcounts [example]. Our calculation deals with the outer extreme: a firm that makes no adjustment to its wage expenditure.
    Similarly, some firms may be able to contractually avoid paying rentals expenses using the force majeure clause, which also entails legal risk. We assume there are no gains on this score.
    For costs such as utilities and repair and maintenance we assume the expenditure will be 60% of the regular expenses, as these expenses will be lower in the absence of regular operations.
    On interest expenses, individual banks may choose to provide relief in terms of a moratorium on interest payment after the RBI circular on the same. But since this is still a decision that banks have to take, it is useful to understand the ability of firms to cover these expenses using their liquid assets. We do two cases: including and excluding interest costs.


We compute the following variables:
  • Liquid assets (LA) = Cash and bank balance + (Marketable securities)*0.5. Liquid assets are 5% of firms' total assets in India, and cash and bank balance accounts for 99.3% of total liquid assets.
  • Minimum Costs 1 (MC1) = Employee salaries, PF and gratuity + Outsourced staff cost + (Power and water charges)*0.6 + (Repair and maintenance)*0.6 + Lease and rentals + IT and communication costs + Insurance premiums + Auditors fees + Rates and taxes + Financial services expenses including interest cost. In the dataset, the share of these elements is as follows:

    • Employee salaries: 43%
    • Outsourced staff: 12%
    • Interest cost: 20%
    • Financial service expenses: 4%
    • Rest of the expenses: 22%

    MC1 is 21% of firms' overall cash expenses excluding direct taxes (cash expenses exclude non-cash items such as depreciation and provisions).
  • Minimum Costs 2 (MC2) = MC1 - Interest costs. MC2 is 80% of MC1.
  • R1: days cover of liquid assets assuming the cost is MC1 = (LA*365)/MC1
  • R2: days cover of liquid assets assuming the cost is MC2 = (LA*365)/MC2


Figure 1: The fraction of firms that cannot hold their breath for a given number of days

Figure 1 shows the fraction of firms where the liquid assets are not able to pay for a given number of days of zero revenue, going by two definitions of minimum cost (i.e. including or excluding interest payments).

We see that 2.7% of the firms do not have liquid assets to meet MC1 for even 1 day. There are 29.8% (4,870 firms) that have 30 days or less of liquidity cover for MC1 and 54.8% (8,927 firms) that have 90 days or less of liquidity cover for MC1. These firms will be the ones most affected by the current lockdown and any future extension to it.

If we put interest expenses aside, thus assuming that all firms are able to obtain an interest payment moratorium, the results change slightly. Now, 2.3% of firms do not have liquid assets to meet MC2 even for 1 day. Now, 23.4% (3,799 firms) have 30 days or less of liquidity cover for MC2 and 47.5% (7,674 firms) have 90 days or less of liquidity cover for MC2.

Who are the firms who are vulnerable on account of low cash holdings?

Our sample of 16,300 firms has an asset base of around Rs. 164 trillion and annual sales of Rs. 107 trillion. The 8,927 firms which we identify as vulnerable (3 months or less of liquid assets for MC1) account for nearly 60% of the overall sales and total assets.

Around 19% of these are large firms (assets > Rs.5 billion), 27% are medium sized firms (assets between Rs.1 billion and Rs.5 billion) and 53% are small firms (assets less than Rs.1 billion). While the large firms may be able to solve their liquidity challenges by accessing credit, the medium and small sized firms will find it harder, given the weaknesses of the financial system.

Around 2,300 of these are listed firms, which is nearly half of all the listed firms in the country.

About 79% of these firms are old firms, that is they have been in existence for more than 10 years.

These firms are spread across many industries: manufacturing (43%), services (28%), trading (14%) and construction and infrastructure (8%).


Our calculation is admittedly based on extreme assumptions: Zero decline in wage expenditure, zero access to fresh credit, and zero revenues for a certain number of days. More than half of the Indian corporate non-financial balance sheet is unable to hold its breath for 90 days, under these assumptions. About a quarter of the firms will not be able to handle a 30 day interruption of revenues. This highlights the incompatibility of a zero decline in wage expenditure with a sustained period of zero revenue and no fresh borrowing.

The authors are with NIPFP and Finance Research Group respectively. Views are personal. We thank Josh Felman, Radhika Pandey, Ajay Shah, Pramod Sinha, and Harsh Vardhan for useful comments.

Comments on the draft Personal Data Protection Bill, 2019

by Rishab Bailey, Vrinda Bhandari, Smriti Parsheera and Faiza Rahman.

In December 2019, the Government introduced the draft Personal Data Protection Bill, 2019 (the "Bill") in the Lok Sabha. The genesis of this Bill lies in the report and draft law ("PDP Bill, 2018") prepared by a Committee of Experts headed by Justice B.N. Srikrishna. This committee was constituted by the Government in the course of the hearings before the Supreme Court in the right to privacy case (Justice K.S. Puttaswamy v. Union of India). This blog post is the first of a two part series containing our comments on the latest version of the PDP Bill, which has been referred to a Joint Parliamentary Committee for their consideration. It builds on our previous submissions on the white paper released by the Justice Srikrishna Committee and the draft PDP Bill 2018 prepared by them.

The Bill offers a fairly comprehensive set of data protection principles and rights to data subjects, particularly in relation to data processing by private entities. However, several provisions are in need of further improvements and revisions. In this piece, we focus on the provisions pertaining to the relationship between citizens and the State and the structure and independence of the proposed Data Protection Authority of India (the "DPA"). We find that by crafting a number of wide ranging exemptions for the State, providing it with various broad and ambiguous powers, and failing to check its influence over the functioning of the DPA, the Bill significantly limits the exercise of privacy rights by individuals.

Defining the State and its functions

The State is one of the biggest collectors/processors of data, and has a unique ability to impact the lives of individuals. While the Bill takes an important and essential step towards empowering the citizen vis-a-vis the State by including the State within the definition of the term "data fiduciary" (and "data processor"), it needs to do more in terms of meaningfully empowering the individual. This process has to begin with providing greater clarity on some definitional aspects.

First, the definition of "data fiduciary" in Section 3(13) and of "person" in Section 3(27) seem to refer to "the State" in its entirety, instead of regarding each of its agencies / departments as an independent data fiduciary. This would, for instance, become relevant in a context like Section 7(1)(g), which requires individuals to be given notice of any data sharing arrangements with other fiduciaries. Clearly, the State cannot be regarded as a monolith for this purpose and the requirement of notice for data sharing should also apply to any inter-departmental sharing within the Government. This is also reflected in international best practice -- the UK's Data Protection Act, 2018, for instance, applies to distinct government agencies, while the US Privacy Act, 1974 refers extensively to intra-government information sharing. To make sure that this will also be the case under Indian law, the definition of "data fiduciary" needs to clarify that any reference to the "State" means the relevant department or Government agency of the State.

Second, Section 3(39) of the Bill imports the definition of the State from Article 12 of the Constitution. Article 12 defines the State to include the Parliament and State Legislatures, Central and State Governments, and all local or other authorities that are controlled by the Government. Given its place in Part III of the Constitution (dealing with fundamental rights), Article 12 has been drafted broadly so as to impose the broadest set of "responsibilities and obligations on the 'State' vis-a-vis the individual to ensure constitutional protection of the individual's rights..." (Pradeep Kumar Biswas v. Indian Institute of Chemical Biology). Accordingly, the term "other authorities" has been read to encompass a range of bodies such as state electricity boards, research and educational institutions, regional rural banks, and statutory corporations such as the Oil and Natural Gas Commission, the Industrial Finance Corporation and the Life Insurance Corporation.

However, when this definition is used in a context that limits, rather than expands the rights of individuals, the intent of the wide definition is turned on its head. For example, Section 12 of the PDP Bill permits non-consensual processing by the State in various circumstances. Given the wide scope of the word "State", this provision could be used by a range of entities, thereby limiting individual rights. Further, terms like "function of the State" and "service or benefit" used in Section 12(a) can also be interpreted very broadly. Thus, the breadth of the word "State", together with the wide ranging nature of functions carried out by the State will imply that a whole range of entities will be permitted to exercise the option of non-consensual processing. Such a wide exemption would be against the spirit of the Bill. Further, it may also create differential regimes for private and public sector entities providing similar services (such as education and health).

Accordingly, we suggest that the references to the "State" in Section 12(1)(a) of the PDP Bill and, by extension, Section 19(2)(a) (which provides an exemption from the right to data portability), should not include any of the "other authorities" under Article 12 of the Constitution. Further, given the challenge of trying to narrowly define the scope of the State, its functions, services or benefits, we propose that the State should be required to meet requirements of "proportionality" when processing data under Section 12. This will help in safeguarding the privacy interests of the individual, in keeping with the Supreme Court's decision in the Puttaswamy right to privacy case (See Bhandari et al, 2017).

Exemption of surveillance and other agencies

Section 35 of the PDP Bill empowers the Central Government to exempt any government agency from the application of the entire Act, if it is satisfied that it is necessary or expedient to do so, subject to procedures, safeguards, and oversight mechanisms that will be prescribed by the Government. This is a very wide power, that enhances the existing asymmetry in the relationship between the citizen and the State, without increasing any corresponding accountability or transparency in the functioning of the State. This is of concern for a variety of reasons.

First, the Central Government needs only to be satisfied that the exercise of such powers is "necessary" or "expedient". The expedience test is hard to check or restrict, and can easily descend into being a mere convenience test, thereby providing an easy justification to invoke this provision.

Second, Section 35 permits the invocation of the exemption on various grounds, that go well beyond the grounds considered by the 2018 version of the PDP Bill. The earlier draft restricted its scope only to "security of the State" while the 2019 version has introduced several additional grounds that are relatable to the "reasonable restrictions" listed under Article 19(2) of the Constitution. For instance, it includes terms like "public order" that are much wider in scope and require a lower threshold for invocation as compared to "security of the State" (Ram Manohar Lohia v. State of Bihar).

Third, the provision does not require that any order that exempts the application of the legislation, and violates privacy in the process, has to be proportionate to the achievement of a stated legitimate aim. This is not compliant with the Supreme Court's judgment in the Puttaswamy right to privacy case.

Fourth, the Section allows both the scope of the provisions from which the agency would be exempted; as well as the procedures, safeguards and oversight mechanisms to be subsequently laid down by the Government. By delegating important powers to the Government, including the power to prescribe procedural safeguards, the Bill precludes the involvement of the legislature and the accompanying benefit of Parliamentary debate.

Fifth, Section 35 permits the Government to exempt the application of the entire Act. We find that there does not seem to be any discernible rationale for exempting the application of provisions like Sections 4, 5, 6, 9, 24, and Chapters I, IX-XIV of the Bill, which provide basic protections like need for fair and reasonable processing, retention norms, security safeguards, etc. Notably, jurisdictions such as the UK also do not exempt law enforcement agencies from the application of their data protection laws in their entirety (Bailey et al, 2018).

Finally, the provision lacks any independent high-level oversight mechanism or periodic review, and actions of the Central Government to exempt an agency are not subject to application of a judicial mind. The need for such oversight was recognised by the Report of the Justice Srikrishna Committee, and the Supreme Court too has repeatedly stressed on the need to implement appropriate procedural fetters on interferences with privacy rights. Further, democratic countries across the world employ multiple layers of oversight to intelligence and law enforcement agencies (See Bailey et al, 2018).

Accordingly, we believe that Section 35 of the Bill ought to be deleted in its entirety. There is an urgent need to revisit the entire legal framework pertaining to surveillance in India, and a broad exemption such as that contemplated under the Bill is undesirable. However, in the event that such a provision is to be retained, it would need to be strengthened by addressing the various concerns detailed above though appropriate safeguards to be built into the primary law. Specifically, the law must ensure that use of the provision and the actions taken pursuant to it are subject to appropriate (judicial) oversight; the grounds for invocation of the provision need to be restricted to "security of the state"; and proportionality requirements need to be specified. Further, relevant provisions of the Bill, as outlined above, must continue to apply even if an agency is being exempted from the application of other provisions of the law.

Exemption for law enforcement purposes

The PDP Bill also provides for an exception in situations where personal data is being processed in the interests of prevention, detection, investigation and prosecution of any offence or any other contravention of any law (Section 36(a)). This provision is extremely wide in its import and can thereby end up creating unreasonably restricts on the fundamental right to privacy. We list the various concerns with the provision and corresponding changes that are being suggested to limit its scope.

First, the provision does not clarify that the exemptions from the obligations under the Bill will only be available to the State. For instance, nothing currently precludes an individual from relying on this provision to set up CCTV in any public place under the guise of preventing any criminal activities. The provision could therefore encourage vigilantism or enable privatised surveillance, significantly limiting the right to privacy. Therefore the provision should clearly state that this exemption can only be availed by the Central Government or a State Government or any officer specially authorised in this behalf by the Government.

Second, the use of the phrase "any offence or any other contravention of any law" extends the exemption to include even relatively minor contraventions of law or even civil wrongs arising from a breach of contract, violations of civil laws, violations of statutory obligations, etc. The provision is therefore overboard and unworkable in its current form. One of the ways to limit its scope could be by providing that the exemption would only be applicable to the prevention, investigation or prosecution of cognizable offences and that too if they are punishable with imprisonment of 3 years and above (indicating the more serious nature of the offence). Further, the phrase "any other contravention of any law for the time being in force" should be dropped.

Third, there is no requirement that the processing has to be proportionate to, and necessary for, interests of prevention, detection, and investigation of the offence, as required under the Puttaswamy tests (Bhandari et al, 2017). There is also no requirement of prior judicial review for such collection and processing of personal data, which was one of the potential safeguards suggested by the Supreme Court in the Puttaswamy Aadhaar case. The PDP Bill should ideally ensure that any processing of personal data under this provision is subject to prior judicial review and is in compliance with the proportionality standard.

Fourth, while the Bill lists some provisions that would continue to apply despite the exemption, it is unclear why other critical user rights, like fair and reasonable processing, access, correction, retention etc., have not been included. We propose that all of these provisions should be applicable to authorities availing this exemption, subject to situations where discharging data protection obligations may actively harm or interfere with their duty of investigating or prosecuting the relevant offences. As noted previously, countries like the UK that provide similar exemptions to their investigative agencies still apply a robust set of data protection norms in such cases (Bailey et al, 2018).

Finally, the Bill removes some of the safeguards that were suggested in the PDP Bill, 2018, pertaining to the processing of personal data of victims, witnesses, informants and other such relevant persons (Section 43(3)). Under the PDP Bill, 2018, the exemptions in relation to the processing of personal data of such people were available only where following the general obligations under the law would obstruct or prejudice the investigation process. This safeguard is critical since it protects the privacy of individuals who are not suspected of having committed an offence but are involved in investigative or legal proceedings by virtue of being victims, witnesses or informants. This approach has also been adopted by the U.K Data Protection Act, 2018 (Section 38(3)). We suggest that this safeguard should be reinstated in the Bill.

Independence of the Data Protection Authority

Ensuring independence of any regulator is one of the basic pillars of good regulatory governance. As noted by the Financial Sector Legislative Reforms Commission (FSLRC), independence of regulators yields greater legal certainty, and therefore better outcomes (FSLRC, 2013). Independence also enables functioning of the regulator as an expert body, which can be particularly relevant in the context of privacy rights given their contextual and often technical nature. Independence of the regulator is even more critical in case of the proposed DPA, as unlike many existing Indian regulators, it will be charged with supervising the private sector as well almost all Government agencies. Further, the range of discretionary and enforcement powers that the DPA enjoys under the law (which range from standard setting to enforcement and adjudication) makes it vital that the body functions without favour and in an accountable and transparent manner (Parsheera, 2019). The DPA's independence therefore needs to be reflected in its composition, selection process, and functioning.

  1. Structure of the DPA: Section 42 of the Bill provides that the DPA will consist of a Chairperson and up to six whole-time members. This does not allow for the appointment of any part-time/non-executive members, who can bring in technical expertise as well as independent ideas and critiques into the functioning of the DPA. Not only is this considered good regulatory practice, it is also in line with existing Indian laws, such as those constituting regulators such as the Telecom Regulatory Authority of India (TRAI) and the Securities and Exchange Board of India (SEBI).

  2. Selection process: The Bill provides that all DPA members will be appointed on the recommendations of an executive-led selection committee comprising of the Cabinet Secretary and two other secretaries of the Central Government. This ensures that the Government will have absolute control over the DPA's selection process, thereby seriously threatening its independence. In contrast, the PDP Bill, 2018 had proposed a judiciary-led selection process. We believe that such a process should be reinstated in the Bill, in addition to which one could also learn from the Right to Information (RTI) Act, which includes the Leader of Opposition (together with representatives of the government) to select Information Commissioners. A selection committee could therefore comprise the Chief Justice of India or a nominee (as chair), and include the Cabinet Secretary, the Leader of the Opposition in Parliament, and two experts (appointed by the Chief Justice in consultation with the other two members). The Bill should also require transparency in the functioning of the selection committee, such as by making the deliberations, votes and recommendations of the committee publicly available.

  3. Terms and conditions of appointment: Section 43(2) of the Bill states that the salaries and allowances and other terms and conditions of service of the Chairperson and the Members of the DPA shall be as prescribed. Empowering the Government to vary salaries of its members, which could also be to their detriment, could end up hindering the independent functioning of the DPA. As with numerous other agencies in India such as the Information Commissioners under the RTI Act and members of the Securities Appellate Tribunal under the SEBI Act, the Bill must ensure that Government is not able to reduce salaries of the members of the DPA once they are appointed. The PDP Bill, 2018 had also suggested the inclusion of such a provision.

  4. Power of the Government to issue directions: Section 86 of the Bill allows the Government to issue any directions to the DPA on issues of public policy and also in the interests of the sovereignty and integrity of India, security of the State, friendly relations with foreign States or public order. While similarly broad provisions are contained in laws governing some other regulatory agencies - such as the TRAI, the Airports Economic Regulatory Authority and the Petroleum and Natural Gas Regulatory Board - the presence of such broad provisions has been questioned in the past, notably by the Parliamentary Standing Committee in its 93rd Report on the Competition Bill, 2001.

  5. Needless to say, conferring such a broad power on the Government can be problematic not least due to the possibility of political interference in technical and administrative functions of the DPA. Given the DPA's wide scope of authority and discretion, ensuring its functional autonomy becomes vital. Further, it must be remembered that State agencies will be among the biggest entities regulated under the law and therefore there should be no scope for such directions being issued in respect of any ongoing investigations by the DPA. In this sense, the Bill is similar to the RTI Act, which does not contain a similar provision for government directions. The Competition Act, 2002 could be seen as another example. Section 51 of that law limits government directions to matters of policy, which is admittedly a vague concept, but also clarifies that this would not include any "technical and administrative matters".

Accountability of the DPA

The Bill currently lacks sufficient mechanisms to ensure the DPA's accountability to the Government and the Parliament, as well as to stakeholders and the public. Given that the DPA is an unelected body, it is critical that appropriate accountability mechanisms be set out in the primary law itself rather than leaving this to the DPA's discretion or through rules to be framed by the Government.

  1. Meetings of the Authority: Section 46(1) of the Bill provides that the rules and procedures relating to the meetings of the DPA are to be prescribed by the Government. We believe that minimum requirements of transparency should be laid down in the primary law. For instance, the law should clearly state the time period within which the minutes of the meetings of the DPA, along with any votes cast or resolutions made, should be published.

  2. Transparency in DPA's functioning: While the Bill does contemplate a consultative and transparent process in the drafting of codes of practice to be issued by the DPA, it does not require the DPA to act transparently during the exercise of it regulation making powers, its supervisory or adjudicatory functions, or in relation to the recommendations that it will give to the Government. We recommend that the Bill ought to provide for a general obligation for the DPA to act transparently in the discharge of all its functions. Further, it should also specify what it would mean to be transparent in specific contexts.

  3. For instance, the Bill should lay out a clear process for drafting new regulations. Such a process should include the drafting of a consultation paper that lays down the problem, the possible interventions and the costs and benefits of each intervention. This should be put through an open and transparent consultation process involving all stakeholders. The DPA should then be required to provide its responses to the comments along with the proposed draft text of the regulation, and seek comments on the same, before proceeding with its final adoption. In case regulations are required to be issued urgently, DPA may issue such regulations without following the consultation process outlined above. However, such regulations should cease to operate at the expiration of six months from its notification unless the consultation process is initiated within this duration.

DPA's redress functions

The Bill creates an adjudication wing within the DPA, which would be responsible for undertaking enforcement actions against any non-compliance of the law as well as providing redress for individual complaints. Given the large number of data fiduciaries in the system and the data principals interacting with them, a large number of complaints are likely to come up before the DPA. As highlighted in our earlier submissions, expecting the same set of adjudication officers to undertake enforcement as well as redress functions, leads to a blurring of the DPA's objectives.

Housing adjudication and enforcement functions within one body could also lead to a conflict of interest within the DPA. For example, a large number of complaints on a particular issue may be due to non-compliance by the data fiduciaries or due to the DPA failing to take appropriate regulatory or supervisory actions to curb such malpractices. Embedding the redress functions within the DPA is therefore not suitable either from a design or execution perspective.

Therefore, we reiterate the need for a separate ombudsman service or a redress agency that would be responsible for adjudicating complaints raised by data principals and awarding compensation to them. By hiving-off the adjudicatory functions of the DPA from its regulatory and supervisory responsibilities, each unit can then focus on its core functions, while also acting as a check on the exercise of the other functions. A similar division of responsibilities between regulatory and redress functions was recommended by the FSLRC in the context of the financial regulatory framework (FSLRC, 2013).

The proposed redress agency should consist of specialised adjudication officers, who would function independent of the DPA, although there should be a strong feedback loop between the two bodies. The redress agency should make use of technology, such as remote participation through audio/visual means to make redress more accessible. If this proposal is accepted, the provisions regarding the terms and conditions of appointment, powers and functions, codes of practice, and other provisions in Chapter IX of the Bill could apply mutatis mutandis to the redress agency. Further, an appeal from any decision made by adjudication officers of the redress agency should lie to the Appellate Tribunal constituted under Section 67 of the Bill.

Inter-regulatory coordination

Given the cross-sectoral purview of the DPA there is significant scope for overlap between the DPA's functions and that of other sectoral regulators. Equally, there is also scope for drawing valuable synergies through cooperation between them. Section 56 of the Bill is therefore a welcome provision that requires coordination between the DPA and other regulators. The provision, however, needs to be further strengthened in order for it to be utilised effectively.

First, the current provision leaves it to the discretion of the DPA and other statutory authorities to enter into a Memorandum of Understanding (MoU) for the coordination of their activities. We propose that the MoU should instead be made mandatory along with a suggested (non-exhaustive) list of provisions that need to be covered in the MoU. This list should include items such as the process for inter-regulatory references; mechanism for cooperation in framing of regulations and codes of conduct; appointment of a nominee of one party as non-voting observer member on an action being considered by the other party; mechanism for exchange of information, subject to confidentiality obligations; and coordination in conducting awareness related activities. This list draws from the MoU between the United Kingdom's Financial Conduct Authority and the Information Commissioner's Office (FCA-ICO MoU, 2014).

Further, the agencies with whom the DPA would necessarily have to cooperate should also be set out in an Annexure to the Bill or the Central Government should be authorised to prescribe this list. For instance, this would include agencies like the Competition Commission of India, the Reserve Bank of India, SEBI, the Insurance Regulatory and Development Authority of India, the Pension Fund Regulatory and Development Authority of India, and TRAI. This will ensure that the agencies cannot subsequently deny the need or the statutory basis for such agreements.


As detailed in our response to the PDP Bill, 2018, the current state of Indian privacy law means that virtually any improvement thereon would represent a significant step towards protection of privacy rights of individuals. However, given the broad scope of the proposed law, and the significant powers given to the DPA, it becomes important to ensure that the law is properly crafted. The role of the Joint Parliamentary Committee currently examining the law becomes all the more relevant given the significant changes made to the law compared to the PDP Bill, 2018, and the absence of any explanations for the same (say in the form of an explanatory memorandum detailing why certain provisions of the law were modified). The fact that the Committee has called for and is considering public comments on the Bill is therefore a positive step.

In this post we examined the lacunae in the Bill in the context of how it delineates the State-citizen relationship, particularly in the form of the exemptions crafted for the State. We highlighted the overbroad nature of the exemptions for the State -- in the context of non-consensual processing, State surveillance and data processing for prevention and investigation of offences of offences -- and demonstrated how this may limit individual rights in a number of contexts. Carving out such broad exemptions detracts from the "horizontal" nature of the law, and also renders the Bill susceptible to constitutional challenges.

At the same time, the Bill also gives significant powers to the Central Government, not just directly (for instance through various standard setting powers), but also by limiting the independence of the proposed DPA. We point, in particular, to the problems in the structure of the DPA, provisions that enable the Government to interfere in its technical and administrative functioning and the need for greater accountability from what is likely to become one of the most powerful regulators in the country.


Bailey et al, 2018: Rishab Bailey, Vrinda Bhandari, Smriti Parsheera, Faiza Rahman, Use of Personal data by intelligence and law enforcement agencies, LEAP Blog, August 1, 2018.

Bhandari et al, 2017: Vrinda Bhandari, Amba Kak, Smriti Parsheera and Faiza Rahman, An analysis of Puttaswamy: the Supreme Court's privacy verdict, LEAP Blog, September 20, 2017.

FCA-ICO MoU, 2014: Memorandum of Understanding dated 29 September, 2014 between the United Kingdom's Financial Conduct Authority and the Information Commissioner's Office.

FSLRC, 2013: Financial Sector Legislative Reforms Commission, Volume I: Analysis and Recommendations, March, 2013.

Justice K.S. Puttaswamy v. Union of India (Right to privacy case), 2017 (10) SCC 1.

Justice K.S. Puttaswamy v. Union of India (Aadhaar case), 2019 (1) SCC 1.

Parsheera, 2019: Smriti Parsheera, Regulatory governance under the PDP Bill: A powerful ship with an unchecked captain?, Medianama, January 7, 2020.

Pradeep Kumar Biswas v. Indian Institute of Chemical Biology, 2002 (5) SCC 111.

Ram Manohar Lohia v. State of Bihar, (1966) 1 SCR 709.


Rishab Bailey, Smriti Parsheera, and Faiza Rahman are researchers in the technology policy team at the National Institute of Public Finance Policy. Vrinda Bhandari is a practicing advocate in Delhi.

Sunday, March 29, 2020

Isolation : A weak link in Indian public health

by Renuka Sane.

  1. There are four parts to the public health response to an epidemic: 1) Trace, 2) Test, 3) Isolate, and 3) Treat. Once someone is infected, tracing involves identifying, listing and following up with all the people who have come in contact with the infected person to ensure that they do not further spread the disease.

  2. While tracing is weak in most of India, some states such as Kerala and Karnataka seem to be doing a better job of tracing. This knowledge can be applied more broadly in India.

  3. India's response to testing has been slow. Only recently have private firms been allowed to do tests for Covid-19. Most of the capability in testing in India is in the private sector, but the limitations of government contracting hamper the extent to which this capability can be brought into fighting Covid-19. Once these problems are solved, and testing is done at scale, we will have a clearer picture of infection hot-spots and will be able to design a appropriate health response.

  4. Once a person has tested positive, or if a person comes in contact with someone who has tested positive, there is a need for isolation. This typically requires staying at home, in a room, separate from others in the household.

  5. Such a recipe for isolation has been designed for countries where the number of members living in a dwelling unit is low. This will be infeasible for most people in India. For example, it is estimated that over 9 million people in Mumbai live in slums. The slums in Dharavi alone are estimated to house 1 million people, with a population density of 869,565 people per square mile. It is impossible to people in low-income housing and slums to self-isolate, especially when toilets are shared in a community. Hence, self-isolation at home will not work for most people in India.

  6. Could the government run isolation facilities? A barrier here is the mistrust of the state. Public isolation facilities are generally found to be sub-standard. People have been seen to flee from public quarantine facilities. Hence, at present, the government is not useful as a mechanism for producing isolation services.

  7. All four legs of the public health response -- trace, test, isolate, treat -- are weak in India. There is considerable interest in the problems of tracing and testing. There is a need for fresh thinking on the problem of isolation also. Once millions of people in India have been tested, what are we going to do with the persons who have tested positive?

The author is a researcher at NIPFP.

Saturday, March 14, 2020

Linking welfare distribution to land records in India: Part 1

by Diya Uday and Bhargavi Zaveri.

Over a year ago, the Union Government announced the PM-KISAN scheme, which is the first centrally sponsored scheme for making direct benefit transfers to farmers. The scheme, which promises an annual transfer of Rs. 6,000 to farmers, is in line with the trend towards substituting in-kind subsidies with direct benefit transfers (DBTs) in welfare programs in India.

For example, in 2018-19, Rs. 2 trillion (around 8% of total government expenditure) was reportedly delivered through DBTs to beneficiary accounts. In the last two years alone, at least four state governments have rolled out DBTs for providing direct income support for farmers (Table 1).

Table 1: Agricultural income support schemes (2018-19)
OdishaWest Bengal
Name of schemePM KisanRythu BandhuRythu
KaliaKrishak Bandhu
Year of
Basis of
AbsolutePer acreAbsoluteAbsolutePer acre
No. of
Annual budget allocation (INR crores) 75,00012,000 8,750 5,611 3,000
Intended no. of farmers (in Mn) 120
6 farmers 4 farmers 7.5
7.2 farmers

Source: These features are extracted from the scheme documents available in public domain as on January 2020.

A common theme that runs across these agricultural income support schemes is that they utilise land records. India has had a checkered history in the field of land record administration and welfare distribution. For example, a previous attempt at linking the fertiliser subsidy to digitised land records has reportedly failed. The variation in the quality of land records across states, and of state capacity for the identification of beneficiaries, will affect the working of such schemes. We may anticipate problems such as the time-lags involved in the updation of records, the area of the land parcel where the subsidy is linked to the size of the land and the identification of possessory interests such as renters and share-croppers. For example, a ground truth study conducted across Maharashtra, Rajasthan and Himachal Pradesh, showed substantial lags in the updation of land records on account of a change in ownership (NCAER 2017). In Maharashtra, for instance, it took an average of 85 days to update the revenue records to reflect a change in the ownership of a land parcel on account of a sale transaction, and 110 days to reflect such a change on account of the death of the original owner (IGIDR 2017).

The implementation of such schemes is complicated by the fact that while the land records are generally maintained by the Revenue Department of the state government, the Agriculture Department is responsible for the implementation of the income support scheme. The efficacy of such schemes would thus depend on the extent to which these departments are able to collaborate.

A new study on the Rythu Bandhu scheme in Telangana

Telangana is an interesting place to study the implementation of an agriculture income support scheme delivered through DBT by entirely relying on land records. This is because unlike the other states that have rolled out similar schemes, the state government of Telangana, before rolling out the Rythu Bandhu scheme (RBS), undertook a state-wide land records updation drive. Referred to as the Land Records Updation Program (LRUP), the drive involved officers of the Revenue Department visiting each village for rectifying errors, updating the land records and issuing a new de-facto land record, referred to as Pattadar Passbooks, to the owners of agricultural land parcels.

While some have acclaimed the RBS as a success story for providing relief to farmers and innovating in the space of agriculture support (here and here), the scheme has been equally criticised for excluding landless and tenant farmers from its purview.

These developments bring up many interesting questions:

  • How was the LRUP done?
  • To what extent did it contribute towards the design and the implementation of the RBS?
  • What are the co-ordination and information sharing mechanisms set up between the Revenue and Agriculture Departments?
  • Has the reliance on land records for the identification of beneficiaries resulted in exclusion and inclusion errors? What is the extent of such errors? 
  • What is the impact of linking the entitlement under the RBS to the size of the land parcel?

Answering these questions will help in understanding the state capacity required for conducting a state-wide land records updation drive, the problems that arise in linking land records to welfare distribution, and the design and operationalisation of DBTs which use land records for the purpose of identification of beneficiaries.

We investigated these questions through a series of interviews, focus group discussions and matching of land records with beneficiary databases in Telangana, over a period of six months. In a two part series of articles here on The Leap Blog, we unveil our key findings. A more detailed version of this work is visible in Thomas et. al., 2020.

In this first part, we describe how the Telangana government undertook a state-wide land records updation program, the outcomes of the updation program and its potential impact on the identification of beneficiaries under the Rythu Bandhu scheme. The LRUP, which primarily aimed to cover agricultural land, achieved fairly large coverage in terms of area (86%). All districts, except Hyderabad (which does not comprise agricultural land) were covered under the LRUP. The number of disputed land parcels was surprisingly low (~5%), and a fresh revised digital land record could be issued to ~93% of the land parcels that were covered and undisputed.

We argue that the scale and efficiency of the LRUP and the rate of disputed land parcels, is attributable to two factors: the lack of a survey that measures the boundaries of land parcels; and restricting the nature of rights and interests in the updated digital land record. We further argue that these constraints have influenced the design and scope of the RBS, restricting its benefits to land owners, and excluding tenants and share-croppers whose interests are no longer recorded on the land record used for the identification of RBS beneficiaries.

In the next article, we will look at the implementation of the RBS, the payment mechanisms used, the overall satisfaction levels and the extent of errors in the identification of beneficiaries under the RBS.

Our study supplements theoretical perspectives on state capacity in India, generates fresh insights into the kind and quantum of capacity required for the upgradation of land records, a probem that is believed to be a significant challenge to India's development. It also contributes to the literature on state capacity and welfare programs in India (Muralidharan et al (2016); Totapally et al (2019)).

Our study involved interviews with government officials of departments of the State Government which were involved in the implementation of the LRUP and the RBS, namely, the Revenue Department, the Agriculture Department and the Finance Department. We interviewed officials of these Departments at the level of the state, two sample districts and sample villages in each of these districts respectively. Additionally, we interviewed officials of the Finance Department and the Integrated Tribal Development Agency (ITDA), which is the department within the government responsible for the delivery of services to tribal communities in scheduled tribal areas.

The districts of Nalgonda and Mulugu were selected as sample districts for the study, owing to the concentration of a large number of small and marginal landholdings and the population of scheduled castes and scheduled tribes in these districts. To obtain insights into the overall satisfaction levels with the RBS, we conducted focus group discussions with beneficiaries and non-beneficiaries in the two selected villages in these districts. Finally, for the purpose of estimating the extent of errors, if any, in the coverage of beneficiaries under the RBS, we compared the base land records in each of these villages with the RBS beneficiaries lists.

Scope of the LRUP and capacity deployed

The LRUP was envisaged as a three month exercise to improve land records in respect of agricultural land in Telangana through two phases:

  1. First phase : This phase covered (a) land parcels with no disputes and (b) land parcels where there were undisputed errors on the records, such as errors of names and surnames.
  2. Second phase: The second phase was designed to cover disputed land parcels, such as land parcels in respect of which disputes are pending in the revenue or civil courts, or land parcels that belonged to a Wakf (land belonging to a religious charity and administered by a statutory board) or had boundary disputes with the Forest Department. The key here was that there was a finite list of categories of "disputes" that were identified by the State Government for the purpose of excluding land parcels from Phase 1.

Table 2 gives an overview of the coverage of the LRUP. It shows that except for one district, namely, Hyderabad, which has no agricultural land, the LRUP covered all the districts aggregating to 86% of the area of the state.

Table 2: Coverage under the LRUP
Unit Coverage Total Coverage
Districts 32 33 96.77
Mandals 573 584 97.26
Revenue Villages 10,823 10,343* Unclear
Area (in acres) 2,38,18,551 2,76,94,830 86
Source: Revenue Department (June 2019) *As per Census 2011.

The coverage indicated in Table 2 was achieved in three months, with the Revenue Department officers having been divided into teams led by the Tehsildar (an official of the Revenue Department in charge of a cluster of villages). The Collector of each district was responsible for the formation of appropriately sized teams for the district, with the mandate of covering 250 acres per day per team. Table 3 gives an overview of the resources deployed for the purpose of the LRUP.

Table 3: Overview of resources deployed for the LRUP
Duration of the project 15th September, 2017-31st
December, 2017
Total number of revenue officers
involved in the project
No. of teams 1,507
No. of villages assigned to each team 9
Estimated area to be covered per day 250 acres per team
Estimated No. of days per village 10
Estimated No. of days per district100
Source: MCR Human Resource Development Institute of Telangana and interviews.

We find that four distinct records are maintained at the level of the village (Table 3a) and the village-level exercise involved updating one of the records, namely, the Record of Rights and the issuance of digital Pattadar Passbooks (explained below) to each land owner.

Table 3a: Overview of land records maintained at the village level
Sethwar A record containing details of cultivable
area on each agricultural land parcel.
Khasra Pahani A record containing the details of
crops on each agricultural parcel of land.
Pahani A record containing parcel-level information
on the ownership, non-ownership interests, area, cultivable and non-cultivable portions,
cropping patterns on each agricultural land parcel.
Record of Right (Form 1B) A record containing a
sub-set of the information from the Pahani records, with a focus
on the ownership, area, manner of acquisition of each agricultural
land parcel
Pattadar Passbook A record containing information
on the ownership of the land parcel.

Each team was part of a training conducted by the District Collectors on (a) the objectives of the exercise; (b) step by step processes involved in the exercise; (c) the formats for the collection of information; (d) the statutory processes to be followed for the correction and modification of the records; and (d) the daily progress reports to be sent to district and state. A state level control room and multiple district-level control rooms were set up to deal with questions that cropped up during the exercise. The reporting heirarchy was clear. The Tehsildars would prepare daily reports for the District Collector, who in turn, was responsible for compiling the village-wise reports in her District. Each report contains information on (a) the details of missing serial numbers for entry in the the Pahani; (b) the extent of variation in the area of the land parcel reflected in the Sethwar and the Pahani; (c) unsettled disputes and issues in the village; (d) details of the land assigned by the government; and (e) list of land parcels put to non-agricultural use in the village.

Budget deployed for the LRUP

An analysis of the financial statements of the Telangana Government shows that while the budget does not separately allocate funds for the LRUP, there is a significant increase in the budget allocation for overall 'land reforms' for the year of implementation of the LRUP, namely, financial year 2017-18 and the financial year 2018-19 (Table 4). In the absence of specific information on the extent to which this allocation was utilised for the LRUP, it is difficult to draw any insights. However, our interviews yielded more specific, although rough, information on the amounts spent for the implementation of the LRUP (Table 4a).

Table 4: Overview of budget estimates for land reform in rural areas (2015-2020)
Financial yearBudget estimate (INR lakhs) Year-on-year change (%)
2016-171256.56 -7
2017-181335.85 6.31
2018-191502.20 12.45
2019-201506.18 0.26
Source: Annual Financial Statements.

*This is the actual amount spent on land reforms in 2015-16.

Table 4a: Budget estimates for implementation of the LRUP
Aggregate budget (INR crores) 100
Amount used for printing Digital Pattadar Passbooks40-60
Amount allocated per district for title verification, etc.1-2
Source: Interviews conducted with state government officials (June 2019)

Table 4a shows that a blanket amount of Rs. 1-2 crores was allocated per district for the process of conducting the title verification, and the overall implementation of the LRUP. This was in addition to the regular budget of the Revenue Department. The allocation was uniform across all districts without regard to their size or other factors which might complicate the implementation further such as fragmented land parcels and vast forest areas. In the absence of district-level data on the usage of these additional funds, it is difficult to ascertain the sufficiency of these amounts or the extent to which a uniform allocation poses problems for the larger or more complex districts.

Digital Pattadar Passbooks

One of the key components of the LRUP was the issuance of digital Pattadar Passbooks (PPBs) to all the owners of agricultural land whose land was found to be undisputed during the LRUP. The PPB is a land record unique to the erstwhile unified state of Andhra Pradesh. Originally, it recorded the interests of owners, Pattadars (defined, under the land revenue laws, as the person who pays the land revenue), mortgagees, and tenants on the land parcel. The purpose of this document was to facilitate farmers' access to credit. For example, the AP Rights in Land and Pattadar Passbooks Act, 1971 allows farmers to apply for credit on the basis of the passbook; and empowers the Collector to recover, on behalf of the lender, an unpaid loan obtained on the basis of the passbook.

The State of Telangana adopted the AP Rights in Land and Pattadar Passbooks Act, 1971 and renamed it as the Telangana Pattadar Passbooks Act, 1971. In 2018, Telangana made two critical amendments to the Telangana Pattadar Passbooks Act, 1971: first, it dispensed with recording the interests of tenants and occupants who are not owners on the PPBs. The reason for this is unclear. However, the current law governing agricultural land tenancies in Telangana, which confers superior rights to protected tenants (namely, tenancies created in 1950) reportedly led to insecurity among owners of agricultural land parcels. It is possible that this was the motivation for restricting the set of rights that were recognised in Pattadar Passbooks.

Second, it specifically provided that loans could be made on the basis of the electronic record of rights, and that the production of the PPB should not be insisted upon for advancing a loan on the security of land, the interest of the owner in land or the crops growing on it (Table 5).

Table 5: Recording of interests in Pattadar Passbooks: pre and post 2018 amendment
Pre-2018 Post 2018
Pattadar Yes Yes
Mortgagee Yes No
Tenant Yes No

The non-inclusion of tenancies and other interests in the PPB is an important amendment as it ultimately affects the identification of beneficiaries under the RBS.

Disputes uncovered in Phase 1 of the LRUP

Contrary to the popular notion of land being a highly disputed area, nearly 95% of the area covered under the LRUP was cleared as free of disputes related to ownership during the first phase (Table 6).

Table 6: State-level outcomes under the LRUP
LRUP outcomes: in acres and gunthas
Total extent
Total extent
2,28,77,333 (94.65)
Total extent
not clear
9,75,915.34 (4.09)
LRUP outcomes: in number of land parcels
No. of Survey
Nos. verified
No. of Survey
Nos. cleared
1,87,60,272 (95.33)
No. of
Survey Nos. not cleared
9,18,572 (4.66)
Total No. of Khatas covered71,71,409
Total No. of cleared Khatas67,68,151
Agricultural Khatas60,00,509
Non-agricultural and govt. assets7,67,642
Total No. of Khatas not cleared4,03,258 (5.62)
Source: Revenue Department, Government of Telangana.
Numbers in brackets are percentages of the total of the head under which they appear.

It is possible that the speed of implementation of the LRUP, the coverage and the high clearance rate is attributable to the manner in which the verification process was conducted. Since the LRUP was not preceded by a survey, the entries in the land records with respect to the area and parcel boundaries, were verified on a self-declaration basis. In the article that follows this one, we demonstrate the discrepancies in the area recorded in the base land records and the Rythu Bandhu beneficiary lists. Second, the LRUP involved the updation of a limited number of fields of information in a PPB. As mentioned above, interests other than ownership were not recorded in the revised PPBs. Since interests such as tenancy and possession, which are inherently more difficult to record, were not within the scope of the program, this might have contributed to the wide coverage, low disputes and speed of implementation of the program.

The design of the LRUP and its bifurcation into two phases has important implications for the scale that the program could achieve. As mentioned above, all disputed land parcels were kept out of the purview of the first phase. Thus, for instance, land parcels with respect to which cases are pending in the civil courts or revenue courts or were subject to succession disputes, were kept for Phase 2 of the LRUP. This ensured that digital PPBs could be issued in respect of the bulk of the land parcels (93%) that were cleared of ownership related disputes under the LRUP (Table 7).

Table 7: Number of digital Pattadar Passbooks issued under the LRUP
No. of cleared
agricultural Khatas
No. of PPBs
55,85,396 (93.08)
No. of Khatas cleared for PPBs, but not yet digitally
4,15,113 (6.91)
AADHAAR is not
Source: Revenue Department, Government of Telangana (June
Nos. in brackets are percentages of the total of the
head under which they appear.


The LRUP was envisaged as a speedy and one-time state-wide intervention for the updation of land records and the issuance of updated digital Pattadar Passbooks in Telangana. While the scale and the speed of the program holds important lessons on the importance of planning and capacity, the design and implementation of the program have several unintended consequences affecting the utility of land records generally and the design of the RBS more specifically.

Our study reinforces the notion of the multiplicity of land records at the village level. The ubiquity of the digital Pattadar Passbooks has made them a de-facto land record for ascertaining claims to ownership in Telangana, although their original purpose was restricted to allowing easier access to agriculture credit on the security of the underlying land and crops grown on it. As demonstrated in Table 3a, information about an agricultural land parcel is now spread across the base land records - such as the Pahaani records and the record of rights - which are usually maintained in all states under the respective land revenue laws, and the digital Pattadar Passbooks. This increases the potential for inconsistency in the information across different land records, even as the records are maintained by the same department of the State Government. Given the large-scale digitisation of base land records in Telangana, and the relative ease with which they can be accessed by citizens through digital portals such as MeeSewa, the need to overlay these records with a new land record in the form of Pattadar Passbooks, remains questionable.

Second, the LRUP was not preceded by a state-wide survey. The design of the program restricting the kinds of interests required to be recorded in the updated digital Pattadar Passbooks allowed the program to proceed with relatively higher speed. Counter-intuitively, the level of disputed agricultural land parcels was relatively lower than one might have estimated. However, this seems to be in line with a previous study that sought to investigate the degree of concordance between information recorded in digital land records and reality in Maharashtra, Himachal Pradesh and Rajasthan. The survey found relatively high concordance on ownership between the digital land records and reality. For instance, the study in Maharashtra found that out of the 102 samples land parcels examined across two villages in Maharashtra, 101 of them were owned by the person reflected in the digitised land record. However, the study found higher discrepancies in recording encumbrances and the area of the land parcel as reflected in the digitised land record and the actual area occupied in reality.

In short, the LRUP could achieve this scale within such a short span of time due to the simplification and minimisation of information that was required to be recorded in the Pattadar Passbooks. However, even as the discourse on the formalisation of land records leans towards widening the range of information recorded in land records, the measure of creating a new digital land title record with a restricted set of information would have the unintended consequence of the record being of limited utility.

Finally, the failure to record interests such as tenancy and mortgages on the digital Pattadar Passbooks, might have influenced the design of the Rythu Bandhu scheme. Reliance on the digital Pattadar Passbooks for the identification of beneficiaries leads to exclusion of landless farmers, such as tenants and share croppers, who are cultivating land and incurring expenditure as tenants. To ensure that the agriculture income support is beneficial, the design of the land record that is used as the base for the identification of beneficiaries, is critical. If the land record records interests such as tenancy and occupancy, it is easier to include tenants and actual cultivators within the ambit of the DBT scheme thereby ensuring that the intended benefits under such schemes reach the tiller of the land.


Sanand, Gupta and Prabhakar. A Pilot Impact Assessment of the Digital-India Land Records Modernisation Programme, NCAER (2017).

Narayanan et al. Report on the implementation of the Digital India Land Records Modernisation Programme (DILRMP) in the state of Maharashtra, IGIDR (2017).

Muralidharan, Karthik, Paul Niehaus, and Sandip Sukhtankar. 2016. Building State Capacity: Evidence from Biometric Smartcards in India. American Economic Review, 106 (10):2895-2929.

Swetha Totapally, Petra Sonderegger, Priti Rao, Jasper Gosselt, Gaurav Gupta. State of Aadhaar Report 2019. Dalberg, 2019.

Thomas, Uday and Zaveri. Linking welfare distribution to land records: a case-study of the Rythu Bandhu Scheme (RBS) in Telangana, IGIDR (2020).

Diya Uday and Bhargavi Zaveri are researchers at the Finance
Research Group, Mumbai. The study was supported by the Omidyar Network India.