## Tuesday, February 28, 2017

### Author: Smriti Sharma

Smriti Sharma is a researcher at the National Institute for Public Finance and Policy.

### Sustainable strategy to eliminate vector-borne diseases

by Shubho Roy and Smriti Sharma.

In his budget speech of 2017, the Finance Minister Arun Jaitley announced that the government has prepared an action plan to eliminate two vector-borne diseases by the end of this year (Paragraph 64):

"Poverty is usually associated with poor health. It is the poor who suffer the maximum from various chronic diseases. Government has therefore prepared an action plan to eliminate Kala-Azar and Filariasis by 2017..."

Kala Azar, also known as Visceral leishmaniasis, is caused by a protozoan parasite of the Leishmania genus. It is carried by an insect vector: Sandfly. If left untreated, Kala Azar can lead to the death of the patient.

Filariasis is a painful disfiguring disease which is caused by roundworms of the Filarioidea superfamily. It is carried by a mosquito vector: Culex quinquefasciatus. The most disturbing symptom is elephantiasis where the patients body parts swell to massive proportions. The infection generally occurs during childhood but manifests itself later in life and can lead to permanent disability.

In this article, we analyse this announcement. We argue that eliminating vector-borne diseases is a good cause for health policy to pursue. However, there is a need to place these actions within a larger strategy on communicable and vector-borne diseases. The critical component of that, which is at present lacking in India, is a sound surveillance system.

### Communicable diseases in health policy

A lot of what governments do in the field of health is of dubious value. Tackling communicable diseases, like Kala-Azar and Filariasis, however passes the basic tests of public economics (Hammer, 2015). Communicable diseases involve a market failure, an externality. When one person gets infected, not only does that person suffer, there is the possibility of others getting infected. This is a negative externality. Each person will under-spend on preventing or curing the disease as the individual does not price the adverse impact upon others. This creates a market failure and justifies a role for the State.

In the extreme, when we get to eradication, we get to a public good. When a communicable disease is eliminated, everyone is protected, even if they did not pay for it. It fits both the tests of a public good:

1. non-excludable: it is not possible to prevent consumers who have not paid for it from having access to it. There is no way to ensure that only the people who paid to eliminate Kala Azar don't get Kala Azar and others are still exposed to it.
2. non-rivalrous: it may be consumed by one consumer without preventing simultaneous consumption by others. When an infectious disease is eliminated, it cannot come back. Enjoying good health by some persons does not reduce the supply of good health, i.e. absence of the disease.

### Why Kala Azar and Filariasis?

But we must ask: Why were Kala Azar and Filariasis prioritised for elimination in the Budget speech?

It is not because they are most widespread diseases in the country. Statistics from the Directorate of National Vector Borne Disease Control Programme (NVBDCP) (Table 1) show that here are other vector-borne diseases like Malaria, Dengue and Chikungunya which are more important.

Table 1: Magnitude of vector-borne diseases in India (2014)
DiseaseCases
in 2014
Deaths
in 2014
Malaria
1,102,205 562
Dengue
40,571 137
Chikungunya
16,049 Unavailable
Encephalitis
(Japanese and Acute)

12,528 2,084
Kala Azar
9,241 11
Filariasis
Unavailable Unavailable

Perhaps it is felt that these diseases are low hanging fruits. Kala Azar is restricted to four states while other vector borne diseases are everywhere. In 2014, a new drug liposomal amphotericin was found to cure Kala Azar with a single dose. Previously, a course of 28 daily doses with hospitalisation was required.In the case of Filariasis, India has reduced the national microfilaria rate (based on sample blood tests) from 1.24% in 2004 to 0.44% in 2014. In Goa, Daman & Diu and Pondicherry, the rates have fallen low enough that mass drug administration (the standard treatment methodology) was stopped in 2012. The government has also achieved good coverage under Mass Drug Administration with 85.6% of the population covered in 2014.

Choosing low hanging fruits has its virtues. It allows for quick results and can potentially help create State capacity. There is a unique charm to eradication. Once a communicable disease is eradicated, you do not have to work on it again (apart from some low-level continued surveillance to check for recurrence). This frees up resources for other purposes.

### A story of failure

In the recent past, India has seen many outbreaks of re-emerging infections, which were claimed to be eliminated. Kala Azar itself reemerged after near eradication in the 1960s. DDT was used for controlling malaria between 1953-64. This helped to decimate the sand-flies that cause Kala Azar. But the pathogen continued to reside in humans. In 1977, when the sand-flies resurged, Kala Azar resurfaced. The disease again resurged in 1983 and 2003 (Muniaraj, 2014). The National Health Policy of 2002 had envisaged elimination of Kala Azar and Filariasis by 2010. This was postponed to 2015.The plague in Surat in 1994 was followed by an outbreak of pneumonic plague in Himachal Pradesh in 2002 (Joshi et al., 2009). There was another bout of plague in Uttarkashi in 2004 (Mittal et. al., 2004). The first outbreak of Chikungunya in India was reported in 1963 but it resurged after three decades in 2006.

### What does it take to finish the job?

It is important to ensure that an elimination drive is sustained beyond its stated date. Vector-borne diseases will not cease to exist with the administration of mass dosage of drugs alone. We need continued disease surveillance and epidemiological investigation post-2017 too. A Kala Azar patient can relapse after six months after the end of treatment. Similarly, Filariasis does not manifest itself in early stages in any outward symptoms. The infected person can thus continue to host the disease for several years.

India is at the cusp of eliminating Kala Azar and Filariasis. It must ensure that the eliminated diseases do not re-emerge. This requires a surveillance system. Surveillance can provide information on entomological data, which can help in creating clusters of diseases that can be targeted by similar vector management measures. Surveillance for epidemiological data can help in calculating the disease burden attributable to each disease. Lastly, surveillance can provide information on implementation. This can help in monitoring and evaluation of disease control and prevention.

### Prevention, control and surveillance: current scenario

NVBDCP's present strategy concentrates on vector management by using indoor residual spraying, nets impregnated with insecticides and anti-larval measures. In addition, early diagnosis and complete treatment is provided to afflicted patients. NVBDCP frames technical guidelines and policies to guide States for implementation. Health departments are responsible for the prevention and control of vector-borne diseases at the State level. But their approach to vector-borne diseases continues to be ancillary and ad hoc in absence of entomological and epidemiological data, and data dissemination:

• Poor entomological surveillance: The government established 72 zonal Malaria offices to conduct entomological surveillance but only 50% of these are functional. State health departments conduct larval surveys but they do not count adult mosquito populations. The field staff uses the outdated ladle and dip method for studying the vector population, despite the availability of new and improved devices like ovitraps. Vector management efforts employ fogging and anti-larval activities to contain mosquito populations. But, this is done without adequate evidence on vector populations.
• Poor epidemiological surveillance: IDSP has a laboratory network of 117 district labs in 28 States/UTs to perform tests for epidemic-prone diseases. Out of the 117 laboratories, 44% of the laboratories do not conduct all the tests recommended under IDSP. This means that some diseases cannot be confirmed and therefore they remain unreported. Private healthcare providers do not report diseases and that leads to under-estimation of the disease incidence. The data for dengue, Kala Azar and Chikungunya on NVBDCP's website is outdated enough to render it useful for public health management. Case and death data for filariasis is unavailable.
• Poor data collection and dissemination: The reports from the rural reporting units to District Surveillance Units of IDSP are often delayed. While 85% districts communicate surveillance data through emails, 67% report data through the portal. This leads to delays in collating and analysing data. State health departments campaign, educate and inform the public about the diseases. But, none of the measures taken by State health departments are ever evaluated for their impact.

The information from NVDCP has three groups of problems:

1. Missing data: The epidemiological data for some diseases is either incomplete or missing. The missing data on deaths cannot be construed as zero deaths.
2. Policy decisions without evidence: India hopes to eliminate Filariasis by the end of 2017. But there is no data on the number of cases and deaths resulting from Filariasis.
3. Under-estimation of incidence: The burden of disease is likely to be highly under-stated by the official statistics. As an example, Dhingra et. al., 2010 estimate that Malaria kills between 125,000 and 277,000 persons in India every year. This is vastly unlike the official statistics. Similarly, Haanshus et. al., 2016 find that in the class of hospitalised patients with undifferentiated fever in India, malaria prevalence is as high as 19%. Similarly, Shepard et. al., 2014 estimate there were 5.8 million cases of dengue per year.

### Building a surveillance system

A sustainable strategy should bring down the infectious diseases in the short term and avoid resurgence later on. The US Centers for Disease Control and Prevention (CDC) has developed a framework for preventing infectious diseases which focuses on:

• Continued surveillance of infectious diseases, laboratory detection and epidemiological investigation;
• Reducing diseases by developing vaccines, preparing strategies for infection control and treatment;
• Using scientific data to inform health policies to prevent and control infectious diseases.

India requires the development of similar principles. We need a robust surveillance system that measures all vector-borne and communicable diseases. This system should generate a constant stream of good quality data which can feed back into management decisions in public health.

The policy agenda for vector borne diseases involves the following components:

• Building entomological surveillance capabilities: State health departments should invest in their entomological surveillance capabilities. They should systematically collect and document changes in vector occurrence, abundance and infection rate for the entire country. State health departments should complement larval surveys with adult surveys. In order to estimate and monitor adult mosquito prevalence, health departments should procure mosquito traps like ovitraps and BG-Sentinel traps (Sivagnaname and Gunasekaran, 2012). State health departments should make the entomological surveys public. For example, the Health Department of New York has launched interactive maps. These maps show the progress made on mosquito surveillance and control operations. The City of Chicago publishes maps with a list of locations and mosquito test results. The CDC displays information on vector borne diseases in maps which also give information on the vectors for each disease.
• Improving epidemiological surveillance: The government should assess the disease burden for each vector-borne disease. State health departments collect information on disease incidence and mortality. The epidemiological surveillance should also include information on geographical distribution of the disease and sub-populations affected. This information should be compiled and made publicly available. Better laboratories are required that conduct tests for all the vector-borne diseases.
• Improving data collection and dissemination: The government should strengthen data management. The health staff responsible for collecting entomological and epidemiological data should be given electronic devices like tablets or mobile phones. The field staff should enter surveillance data through these digital devices. For example, Kenya moved away from manual data reporting to electronic data reporting for its National Tuberculosis, Leprosy and Lung Disease Programme with an Android based application called TIBU. Florida's Department of Health puts out weekly and annual reports on surveillance. State health departments should also conduct impact analysis on vector control measures. NVBDCP and IDSP should make these studies publicly available to all the stake-holders including other State governments, the private sector and consumers.

Building a generalised and integrated communicable disease management system is laying infrastructure. It can be used for tackling different problems from year to year. Our objective should be to lay this general infrastructure, and not narrowly run campaigns targeting one disease or another.

For an analogy, consider Aadhar. Aadhar is just an identity platform. However, it has been built on robust technology using sound processes. In itself, Aadhar does not do much. However, Aadhar can act as a backbone for multiple initiatives ranging from financial inclusion, rationalising subsidies, targeting delivery of public services, national security, preventing corruption and leakage, etc. It constitutes a general purpose infrastructure which builds a platform on which many specific public services can run.

In similar fashion, a well designed communicable disease management platform which leverages technology can be used to deal with kala-azar and filariasis this year, but can be used for malaria, chikungunya and dengue the next year. The same surveillance, monitoring and data dissemination systems will work for multiple diseases. So far, the government has integrated the disease surveillance programme, i.e. brought multiple programmes under one umbrella but it has not re-imagined the way it should be carried out.

### References

The case for Universal Healthcare is weak by Jeffrey S. Hammer, July 2015, Ajay Shah's blog

Epidemiological features of pneumonic plague outbreak in Himachal Pradesh, India by Joshi K et al., May 2009, Transactions of the Royal Society of Tropical Medicine & Hygiene

Quick control of bubonic plague outbreak in Uttar Kashi, India by Mittal V et al., December 2004, The Journal of Communicable Diseases

Report of the Working Group on Disease Burden for the 12th Five Year Plan by Planning Commission, Govt. of India, May 2011

Adult and child malaria mortality in India: a nationally representative mortality survey by Dhingra Neeraj et al., October 2010, The Lancet

A high malaria prevalence identified by PCR among patients with acute undifferentiated fever in India by Haanshuus CG et al., July 2016, PLOS ONE

Economic and Disease Burden of Dengue Illness in India by Shepard Donald S. et al., December 2014, The American Journal of Tropical Medicine and Hygiene

Need for an efficient adult trap for the surveillance of dengue vectors by Sivagnaname N. and Gunasekaran K., Indian Journal of Medical Research, November 2012

Handbook for integrated vector management by World Health Organization, 2012

Smriti Sharma and Shubho Roy are researchers at the National Institute of Public Finance and Policy, New Delhi.

## Monday, February 27, 2017

### Protecting retail investors' interest in the bond market

by Ritika Chhabra and Anjali Sharma.

### Context

On January 04, 2017, SEBI released a discussion paper titled "Public Issuance of Non-Convertible Debentures having credit rating below Investment Grade" (hereafter, discussion paper) for public comments. The proposals in the discussion paper focus on retail investors' participation in public issuances of sub-investment grade non-convertible debentures (NCDs). Each of these terms is technical and its useful to understand what they mean before analysing SEBIs proposals:

• While protecting retail investors' interest in securities markets is one of SEBIs stated objective, there is no standard definition of a retail investor which is consistent across securities. For example: for equity public issuances, a retail investor is one whose bid does not exceed Rs. 2 lakhs. For issuance of tax free bonds, the IT Act, 1961 defines retail investor as one whose bid does not exceed Rs. 10 lakh.

• A public issuance is defined as per Section 42 of Companies Act, 2013 as one where securities are issued to more than 200 investors. Typically, public issuance of corporate debt securities is in the form of NCDs. The issuance process is laid down by the SEBI Issuance and Listing of Debt Securities (ILDS) Regulation, 2008. As per ILDS Regulation 4(2)(c), for public issuance of debt securities, the issuer has to get at least one credit rating.

• A sub-investment grade rating is one that is below BBB-. Debt securities with this rating are commonly referred to as junk bonds.

In the discussion paper, SEBI makes two proposals with regard to such issuances. First, the introduction of a pictograph, known as the risk-o-meter, to disclose credit rating in the offer document. Currently, credit rating is disclosed on the first page of the offer document in text form. Second, the introduction of a two level restriction on retail investor participation - an investor level restriction: Rs. 2 lakh; and an aggregate restriction on all retail investors: 5% or 10% of the issue size.

In this article, we evaluate SEBIs proposals and offer our analysis of key issues that this discussion paper highlights.

### Analysis of SEBIs proposals

1. The proposals are relevant for a very small part of the corporate bond issuance market: The market for public issuance of sub-investment grade NCDs is small, both in absolute terms and relative to the corporate bond market (Table 1). In 2015-16, 2 of the 20 public issuances were sub-investment grade and these accounted for only 1.3% of the public issuance volume. As a proportion of the total corporate bond issuance volume, these were less than 0.1%.

When we look at the identity of issuers and the rating profile of their issuances, we find that the public issuance market is dominated by highly rated issuers. These include large public sector entities like NTPC, NHAI, NABARD, IRFCL and 3-4 private non-bank finance companies (NBFCs) who are regular issuers. There are only 2-3 NBFCs which issue in the sub-investment grade category and these are repeat issuers.

2. Table 1: Corporate bond issuance market in India

Year Issuance value Private placement Public issuance Public issuance - sub-investment grade1
(Rs. trillion) (%) (%) (%)

2014-15 4.1 97.7 2.3 0.34
2015-16 4.9 93.2 6.8 0.09
2016-172 5.1 94.3 5.7 0.09

Source: SEBI
1 Sub-investment grade represents a credit rating below BBB-
2 Data till December, 2016

3. Retail investors, for whom investment limits are proposed, are the main investors in publicly issued sub-investment grade NCDs: We find that retail investors are the primary investors in these issuances. For example: 16 unique ISINs were generated as part of 2 issuances in 2015-16. An analysis of the ownership pattern of these ISINs, using data from NSDL, shows that in 15 ISINs, retail investors' subscription was more than 90% of the issue size. In the remaining one ISIN, it was more than 70%.

4. SEBIs problem identification is unclear and not supported by data: In the discussion paper, SEBI states:

Unlike private placement of debt securities, retail investors participate in the public issues. These public issues which are rated below investment grade (i.e. below BBB-) give high coupon rate. The advertisements of such issues also focus mainly on the coupon which lures the retail investors to invest.

Further, certain issuers with credit rating below investment grade, have issued both secured and another unsecured NCD through same offer document with different credit ratings. Thus, for a retail investor to differentiate between secured and unsecured tranches within the same offer document and with different credit ratings may be a complex task and may affect their investment decision.

It is felt that there needs to be an additional layer of protection for the retail investors, who get attracted towards such debt securities which though on one side pay higher coupon but on the other side have a below investment grade credit rating.

From these statements, it appears that SEBI identifies three problems: (1) the manner of disclosure of credit rating in sub-investment grade issue advertisements; (2) issuers offering secured and unsecured tranches within the same offer document, which makes in complex for retail investors to understand credit rating; and (3) retail investors getting attracted to the return and investing in risky securities.

There are two issues with SEBIs problem identification process. First, it is not supported by any data or analysis. For example in case of problem (1) and (2), there could be two issues: with the visibility of the text based disclosure, or with investors' understanding of the meaning of a credit rating or both of these. It could also be that there are more problems with disclosure quality than just credit rating or that disclosure quality is not a problem at all. Retail investors are attracted by high returns, are aware of the risk, yet invest because of other considerations, such as fact that these are repeat issuers and their past performance is known. Second, the problem identification does not evaluate whether a regulatory intervention is needed. For example: in case of problem (3) if retail investors are attracted to risky securities, it is unclear why the regulator should step in.

5. Cost benefit analysis is missing: Proposal 1 -- introduction of a risk-o-meter, seeks to address problem (1). Credit rating is currently displayed in text form on the first page of the offer document. The table that explains the implications of the rating is placed inside the offer document (Figure 1).

#### Figure 1: Current disclosure of credit rating in offer document

The proposed NCD risk-o-meter makes the credit rating more clearly visible than a text based disclosure (Figure 1 and Figure 2). However, it does not necessarily improve investor understanding of the credit risk. In case of the mutual fund risk-o-meter, from which this proposal is inspired, the risk of the offer can be clearly understood from the risk-o-meter (Figure 3). As mentioned earlier, its possible that investors do not understand credit risk or that there is no problem with the current disclosure format. In both these cases, the inclusion of a risk-o-meter adds no value. The modification of the offer document format may create costs for issuers. However, no cost-benefit evaluation is offered in this regard.

For problem (2), where an unsecured and secured tranche is issued through the same offer document, SEBI offers no proposals. For example: it does not state whether such offer documents will have two risk-o-meters.

#### Figure 3: Mutual fund risk-o-meter

In Proposal 2 -- introduction of a investor level and aggregate level investment limit for retail investors, SEBIs approach is contrary to its approach to retail investors in equity markets. An equity security is far riskier than a sub-investment grade bond, yet SEBI actively encourages retail investor participation in equity issuance. Further both the limits proposed: Rs. 200,000 per investor and 5% or 10% aggregate limits are arbitrary. The benefits of this proposal from an investor protection perspective are unclear. Yet the costs are obvious. Given that retail investors are the main investors in these issuances, these limits may cause the sub-investment grade issuance market to dry up completely. Also, regulatory interventions that seek to limit investor choice are a firm step away from SEBIs stated objective of being a disclosure based regulator and a step towards being a merit based regulator.

### A small survey: can key disclosures be easily found in offer documents?

Given that SEBI has identified the visibility of credit rating disclosure in offer document as a problem, we conducted a small survey to understand whether other key disclosures can be easily found from an NCD offer document. We gave respondents an offer document for a public NCD issuance, and asked them to find three details about the issue from it: (1) whether the issue is secured or unsecured, (2) what is the rate of return, and (3) what is the tenure. We asked them to measure the time it took them to find this information. 15 personnel from Indira Gandhi Institute of Development Research (IGIDR) and the National Institute of Public Finance and Policy participated in the survey. Table 2 provides a profile of the respondents and Table 3 shows they survey results.

Table 2: Profile of survey respondents

No. of respondents 15
Respondents with demat account 53%
Respondents who invested in equity securities in the last 18 months 40%
Respondents who invested in debt securities in the last 18 months 20%
Average time taken to respond to the survey 15-30 minutes

Table 3: Survey results

Responses
Questions Correct (%) Incorrect (%) Can't say (%)

Is the issue secured or unsecured? 67% 20% 13%
What is the size of the issue? 47% 40% 13%
What is the minimum tenure of the NCD? 27% 53% 20%
What is the yield for minimum tenure NCD? 40% 40% 20%
What is the maximum tenure of the NCD? 33% 40% 27%
What is the yield for maximum tenure NCD? 20% 53% 27%

The survey results point to a deeper problem in offer document disclosures than just the credit rating. Even locating basic information like yield and tenure is not easy.

As a separate exercise, we reviewed 5 abridged prospectuses from public issuances done in 2015-16 and 2016-17. An abridged prospectus accompanies the application form, and SEBI has mandated that the issue term sheet be disclosed on the first page. Even here, we found that in 1 case the terms of the issue were not on the first page. This was a case where issuance was done as part of a shelf prospectus. A typical abridged prospectus is around 40 to 50 pages in length, and even here it is not easy to locate key disclosures within it.

### Conclusion: way forward for SEBI

SEBIs proposals in the discussion paper focus on retail investor protection. Globally, regulators follow a two pronged disclosure based approach to do this. First, reducing information asymmetry between issuers and investors by regulating the quantum and quality of disclosures. Second, enforcing disclosure standards and penalising violations of disclosure norms. Securities regulators, even in Asian countries, have moved away from the merit based approach of imposing investment limits.

Given SEBIs stated objective of being a disclosure based regulator, it needs to focus on improving disclosure quality, not on constraining retail investor participation. However, this requires adopting a systematic and research based approach to addressing the disclosure problem. Our survey findings suggest that SEBI needs to do the hard work of evaluating disclosure effectiveness more comprehensively. In undertaking this exercise, SEBI can take cues from the recommendations of the Sumit Bose Committee (2015) on improving disclosures and curbing mis-selling for financial products. Simply, introducing a risk-o-meter will be insufficient and ineffective.

Other measures, such as stepping up investor education efforts in the debt securities space, may also be relevant. In case of equity securities, SEBIs efforts in this area along with the efforts of a growing industry of equity analysts and proxy advisory firms has contributed to improving investors' understanding of risks and return. A similar effort for the corporate debt market may have greater long term benefits, for both investors and issuers, than introducing any investment constraints.

Finally, sound regulatory governance is critical to building participants' confidence in the financial system and fostering certainty. Regulatory capacity is finite and regulatory interventions not cost-less. Hence, regulation making needs to follow the discipline of a robust process of identifying a problem, proposing an appropriate solution based on research and analysis, evaluating costs and benefits and carrying out a public consultation process. This will ensure that only meaningful interventions, for which benefits exceed the costs, get enacted. SEBI has already started on this path by regularly seeking public feedback on proposed interventions. However, its proposals are not supported by robust problem identification. There is no cost benefit analysis, research or empirical evidence in support of the proposals made. It shows that SEBI has some way to go in meeting global standards of regulatory governance.

The Finance Research Group has submitted its comments on this discussion paper to SEBI.

Ritika Chhabra and Anjali Sharma are researchers at Finance Research Group. We thank all the FRG-IGIDR and NIPFP team members who took time out from their busy schedules to respond to our survey. We also thank Renuka Sane and Bhargavi Zaveri for useful discussions and suggestions.

## Tuesday, February 21, 2017

The agenda at SEBI by Ajay Shah in Business Standard, February 20, 2017.

Containing Trump by Jonathan Rauch in The Atlantic, March, 2017.

Budget and agri-commodity trading: Searching for a spot in the future by Pravesh Sharma and Raghav Raghunathan in The Indian Express, February 16, 2017.

Half-hearted FDI reform by Bhargavi Zaveri & Radhika Pandey in Business Standard, February 16, 2017.

American Institutions Are Pushing Back Against Trump by Peter Beinart in The Atlantic, February 15, 2017.

Importance of goods and services by Bhanu Joshi and Neelanjan Sircar in The Hindu, February 15, 2017.

Stringent data protection law is the need of the hour by O N Ravi in The Economic Times, February 15, 2017.

ISRO PSLV-C37 Launch with Onboard Camera and Stage/Satellite Separation events by Wayne Tsai, February 15, 2017.

Tech and the Fake Market tactic by Anil Dash in Medium, February 10, 2017.

Big Brother is winning by Pratap Bhanu Mehta in The Indian Express, February 8, 2017.

PM and the UP crucible by Kumar Ketkar in The Indian Express, February 8, 2017.

Budgeting For Democracy by Chakshu Roy in The Indian Express, February 7, 2017.

Why algorithmic trading is unpopular among finance practitioners: As Goldman Embraces Automation, Even the Masters of the Universe Are Threatened by Nanette Byrnes in MIT Technology Review, February 7, 2017.

Recipe for unfettered raid raj by Saumitra Dasgupta in The Telegraph, Febraury 6, 2017.

Ur-Fascism by Umberto Eco in The New York Review of Books, June 22, 1995.

### Workshops at IGIDR, 2017

The Finance Research Group at Indira Gandhi Institute of Development Research is inviting submission of papers for Field Workshops during 2017. Each field workshop will discuss six papers in a day, dedicating a full hour to each paper for presentation, discussions by an expert in the field and the general audience. Details are as follows:

Support: IGIDR will provide accomodation at the workshop venue for 2 nights around the dates of the workshop.

Contact details: Submissions must be sent as PDF files, to Jyoti Manke at jyoti@igidr.ac.in, +91-22-28416592 (office) or +91-98205-20180 (cellphone).

## Friday, February 17, 2017

### Financial Sector Reforms: A status report, 2017

In this article, I piece together information in the 2017 budget to write a status report on financial sector reforms.

1. Consumer protection: Three important initiatives are in progress: -

Financial Redress Agency (FRA): The FRA is expected to provide a unified, speedy and convenient complaint settlement mechanism to retail financial consumers. In the Budget speech of 2015, the Finance Minister had proposed to create a Task Force to establish a sector-neutral FRA. In June 2016, the Task Force recommended enacting a financial sector consumer protection law and proposed an implementation blueprint. It also addressed concerns expressed by RBI and SEBI. RBI has supported the idea of a single grievance redress agency. One example is their submission in the context of regulating deposit taking activities across regulators and State Governments (Twenty-first Standing Committee on Finance, 2015-16, Para 59). The government had recently invited public comments on the Task Force's report. The next steps in this journey consist of draft Bill on the proposed law, and the establishment of the FRA.

Curbing illicit deposit taking schemes: There is considerable understanding of the problem of ponzi schemes. The twenty-first report of the Standing Committee on Finance (referred above) had recommended various measures to plug the regulatory gaps and overlaps related to deposit taking activities. This year, the Finance Minister has promised to introduce the Banning of Unregulated Deposit Schemes and Protection of Depositors' Interests Bill soon. This is the second version of the draft law and was released in November 2016 for comments. The Government had promised this law in last year's budget as well as in the ATR on the above report. The proposed law aims to empower the State Governments to regulate deposit schemes which today slip through the regulatory cracks. It allows the designated courts to impose significant penalties (Jail up to 10 years and fine of up to twice the amount of total funds collected). It proposes significant powers to the State Governments and the police.

The Government should use concepts from the draft Indian Financial Code to strengthen the regulating of deposit taking activities by the State Governments. For example, provisions related to accountability (Chapter 14), regulation making (Chapter 17), show cause notices and orders (chapter 25) could be adapted to strengthen the proposed Bill. The clause on warrant-less searches should consider the provisions related to investigation in the draft IFC (chapter 22). These would help improve the balance of regulatory powers with accountability. This year, the Finance Minister has also promised to plug the regulatory gaps in the Multi State Cooperative Societies Act, 2002, as part of the follow up to the recommendations of the above Standing Committee. As part of the clean up, the Government is also expected to soon introduce a Bill to amend the Chit Funds Act, 1982, though that has not been mentioned in the budget documents. Given the extensive nature of changes, it would be doubly useful to harness the work embedded in the draft IFC.

Securities Appellate Tribunal (SAT): SAT's jurisdiction today covers SEBI, IRDAI and PFRDA. Orders by RBI are still not appealable at a tribunal. Other than that, the scope of SAT now approaches that of the Financial Sector Appellate Tribunal (FSAT), proposed in 2014. In light of the expanded jurisdiction, the Finance Bill, 2017 (S.145) proposes to provide for more members and benches of the SAT, including benches outside Mumbai. This would be achieved by amending the Securities and Exchange Board of India Act, 1992.

The Government should consider modernisation of SAT's processes and administrative functions. Tribunal members should be able to focus on their case load instead of the day to day aspects of administrative functions like finance, human resource and information technology (Datta, 2016, Towards a Tribunal Services Agency). This would help reduce the time taken to dispose off cases.

2. Systemic Risk Regulation: Financial Data Management Centre (FDMC) is expected to provide for a nation-wide integrated repository of information relating to the financial sector. It would be used to study systemic risk, system-wide trends and facilitate a discussion about policy alternatives. Para 90(iii) of Budget document on implementation of last year's announcements (Implementation Document) notes that a draft Bill to set up the FDMC is proposed to be placed for public consultation. Last year, a draft Cabinet Note on setting up of a non-statutory FDMC was circulated. Based on the feedback received, the Finance Minister had approved creation of a statutory FDMC. Thereafter, a Committee to suggest a draft law was set up. In October 2016, this Committee submitted its report which included a draft Bill titled Financial Data Management Centre Bill 2016.

The Government should focus on ensuring that FDMC has the statutory ability to: (i) create a truly integrated repository; (ii) develop capacity to provide research and analysis support to the Government and (iii) ensure that the Data Centre has the ability to evolve with the changing requirements over a period of time. It would be a sub-optimal if the FDMC would need to rely on the executive powers of the Government/ FSDC or the willingness of the regulators to achieve its objectives.

3. Digital Payments: The Committee to review framework related to digital payments has, as part of its report, suggested that regulation of payments should be separated from the central banking function of the RBI. This year, the Finance Minister has proposed creation of a Payments Regulatory Board (PRB) within the overall framework of RBI to regulate payments. The proposed PRB has equal representation from RBI and nominees of the Central Government, with the RBI Governor being the chair (S.148, The Finance Bill, 2017). In the proposed PRB design, no decision can be taken unless RBI agrees. However, this is an improvement over the present regime where a sub-committee of the Board of RBI is regulating payments.

The Finance Minister has said in his budget speech that the Government will undertake a comprehensive review of the Payment and Settlement Systems Act, 2007 and bring about appropriate changes. This would be a major reform in this field.

4. Monetary Policy: In March 2015, as part of its monetary policy framework agreement, India had established inflation targeting as a goal for RBI. In June 2016, the Parliament amended the RBI Act to require the Government to set a CPI based inflation target once in every five years (S.45ZA). A six member Monetary Policy Committee (MPC) was designed to determine the Policy Rate required to achieve this target (S.45ZB). RBI and Central Government have three votes each on this Committee. In case of a tie, the RBI Governor has a second vote. In addition to creating institutional capacity, this reform brings transparency to the decision making process. Section 45ZI(11) requires each member of the Committee to record the reasons for voting in favour or against the resolution. Section 45ZL requires the RBI to publish the minutes of the meeting with details of vote of each member and the reasons recorded by them. Para 90(ii) of this year's Implementation Document notes that action on this reform has been completed.

The Government should re-visit the design of the MPC in the future. In the current design, the RBI does not need to convince even one non-RBI Committee member on its policy stance for the decision to go through.

The Government should use the MPC meeting process to develop a cookie cutter approach to improve working of regulatory forums. The effect of these amendments is immediately visible. Contrast the detailed minutes of the MPC with the brief disclosures of the meeting of the Central Board of RBI, held around the same time (Minutes of MPC meeting, December 6, 2016 vs. Release on the 562nd meeting of the Central Board, December 15, 2016). The meeting of the Central Board is summarised in 150 characters, excluding details of attendance etc. The tweet style disclosure of the Central Board meeting is not a one-off. The immediately previous meeting details are also 150 characters long and are, co-incidently, exactly the same in content. The gaps in the quality of disclosures are glaring when we make international comparisons (Patnaik and Roy, 2017, The RBI board: Comparison against international benchmarks).

5. Capital Controls: This year's budget speech proposes the abolition of the Foreign Investment Promotion Board (FIPB) in 2017-18. It notes that the FIPB has successfully implemented e-filing and online processing of the FDI applications. It proposes that the road map for abolition of FIPB would be announced over the next few months. The Government has also indicated its desire to further liberalise the FDI policy. The abolition of FIPB would be a significant step if abolition is achieved in substance.

Government needs to ensure that important reform steps do not slip through or get stalled. Control on all capital flows is exercised by the RBI, in consultation with the Government. The Finance Act of 2015 (S.139) had amended Section-6 of the Foreign Exchange Management Act, 1999 (FEMA) to provide that control on non-debt capital flows would be exercised by the Government, in consultation with the RBI. This amendment has not yet been notified. This requires the Government to first issue a notification distinguishing debt and non-debt instruments. One would have assumed that once the Parliament has amended a law, the government would be able to notify the same in a reasonable time.

6. Government debt management and its bond market: Two important initiatives are in progress: -

Public Debt Management Agency (PDMA): In October 2016, the Government took first step by setting up an advisory Public Debt Management Cell (PDMC). The Office Memorandum notes the functions of PDMC and mentions a two-year (October 2018) journey towards a statutory PDMA. The Finance Bill, 2015 (Chapter VII) had proposed setting up a PDMA but the move was rolled back in April that year. This reform appears to be back on track. However, the first real progress would be to introduce a law that would establish the agency (Pandey and Patnaik, 2016, Legislative strategy for PDMA).

Unified market for government securities: This reform appears to have lost traction after the 2015 roll back of the move to shift regulation of bond market from RBI to SEBI. The Finance Bill, 2015 (S.157) had proposed to create a unified market for government securities. There is some action in this Budget aimed at improving retail participation in government securities. It also captures some initiatives aimed at deepening of the corporate bond market.

7. Recovery of debt: DRTs deal with cases related to debt due to banks and financial institutions. Last year's budget had said that the Government would focus to strengthen the DRTs through computerised processing of court cases. This year, the budget notes that the Government is providing appropriate infrastructure, filling up vacancies and providing training to DRT staff. Recent reports put the Debt Recovery Tribunals (DRTs) case backlog at over 95,000 cases. This year's Budget further note that along with the SARFAESI Act, the Recovery of Debts Due to Banks and Financial Institutions Act, 1993 (RDDB & FI Act) has been amended through The Enforcement of Security Interest and Recovery of Debts Laws and Miscellaneous Provisions (Amendment) Act, 2016. This is expected to facilitate expeditious disposal of recovery applications. Harmonisation of provisions of IBC, 2016 with the provisions of the SARFAESI and the RDDB & FI Act is expected to help to improve the credit and recovery environment.

The Government should prioritise the capacity building efforts at the DRTs. Going forward, based on the Insolvency and bankruptcy Code, 2016 (IBC, 2016), the corporate cases are expected to shift to NCLT. However, the rules and judicial procedures need to be redesigned for both NCLT and DRT to deliver the expected outcomes (Understanding judicial delays in India: Evidence from Debt Recovery Tribunals).

8. Market for stressed assets: A well function market for stressed assets should encourage a lender sell its NPAs to an Asset Reconstruction Company (ARC). An ARC should be able to raise funds by issuing NPAs backed Security Receipts (SRs). It should then be able to sell the NPAs to redeem the SRs at a profit. This year, the Finance Minister has permitted listing and trading of SRs issued by a securitisation company or a reconstruction company in a registered stock exchange. This is aimed at enhancing capital flows into the securitisation industry and help deal with bank NPAs. This should help efforts to develop the market. Last year's budget proposal to enable the sponsor of an ARC to hold up to 100% stake in the ARC and permit non-institutional investors to invest in SRs have been effected. These have been achieved by amending the SARFAESI Act through The Enforcement of Security Interest and Recovery of Debt Laws and Miscellaneous Provisions (Amendment) Act, 2016 in August 2016.

9. Mechanism to close down failed financial firms: This year, the Finance Minister has promised to introduce the Bill relating to resolution of financial firms in the current Budget Session. He noted that together with the IBC, 2016, a resolution mechanism for financial firms would ensure comprehensiveness of the resolution system in our country. Last year, the budget had recognised the need for a specialised resolution mechanism to deal with bankruptcy situations in banks, insurance companies and financial sector entities so that losses are minimised. This year's budget notes that the draft law -- The Financial Resolution and Deposit Insurance Bill, 2016 is under vetting by Ministry of Law and Justice (See: Report of Committee to Draft Code on Resolution of Financial Firms). The Insolvency and Bankruptcy Board of India has already been constituted in October 2016. With the above work, the stage is now set for a resolution law. This would lead to creation of a Resolution Corporation (RC).

10. Taxation: Taxation of finance contains serious problems. The Government needs to re-engage with the task of enabling a simpler direct tax law. The Direct Tax Code Bill of 2009, which promised this, went through several amendments before being finally shelved (Para 129, Budget speech, 2015). The said speech made a case that the Income Tax Act had incorporated most of the suggestions. This assertion is a subject of debate (India still needs the Direct Tax Code, Requiem for a Code). The proposed amendments in this year's Financial Bill related tax administration are retrograde and need to be reconsidered (See: Rai, 2017, Notes on Union Budget 2017-18)

Outside of the above ten areas, the bad loans of the Public Sector Banks (PSBs) present a major concern for the Government. Based on RBI's data table, the net NPAs (gross bad loans less provisions) of nationalised banks and SBI stood at over Rs 3.2 trillion or about 2.4% of GDP at the end of March 2016. The problem of PSBs is an outcome of Government ownership. Unlike PSBs, the NPAs of private sector banks are relatively small at Rs 266.7 billion (7% of the PSBs' figure). RBI's recent Financial Stability Report (FSR, December 2016) forecasts that PSB category may continue to register the highest GNPA ratio (ratio of gross NPA to gross advances). Attempts at reforms are moving rather slowly. The road map for consolidation of banks is being drawn up. However, no specific progress is listed in this year's Budget. The Banking Bureau's job list is expanding but it seems to be struggling with relatively small battles. Last year, the Government had allocated Rs 229 billion for recapitalisation to 13 PSBs. This year, another Rs 100 billion has been budgeted with promise of more. However, over the years, infusing capital has not yielded the desired results.

There is much to look forward to in 2017-18. Most of the reforms seems to be headed in the correct direction. Direct tax administration and the NPA problem of PSBs seem to be the exceptions.

Ashish Aggarwal is a researcher at National Institute of Public Finance and Policy.

## Thursday, February 16, 2017

### Monetary policy strategy for 2017

by Ila Patnaik and Ajay Shah.

India now has an inflation targeting central bank and a monetary policy committee. The first three monetary policy committee meetings have taken place. The first meeting cut the de jure policy rate, and the next two meetings chose to hold.

Winston Churchill once said If you put two economists in a room, you get two opinions, unless one of them is Lord Keynes, in which case you get three opinions. However, all the three meetings of the MPC featured six economists with one opinion.

In this article, we argue that conditions in the economy suggest that it is time to worry about forecasted inflation going closer to the low end of the target range.

Let's start at the measure of inflation that is used in defining RBI's objective, i.e. the year-on-year change of CPI:

 Headline inflation, i.e. year-on-year CPI inflation

Y-o-y CPI inflation breached 5% in February 2006. After that, we had a long and painful bout of inflation. A recession began in India in 2012, and by mid-2013, inflation was on the decline. The latest value, for January 2017, shows 3.17%. This is benign when compared against the range from 2 to 6 per cent, which is coded into the RBI Act.

Each reading of year-on-year inflation is the average of twelve changes for the latest twelve months. To understand what is going on in the economy in recent days, it's useful to look at month-on-month changes. This requires seasonal adjustment. We have developed the models for seasonal adjustment at NIPFP, and will use this ahead.

Roughly half the CPI basket is food and food inflation is thus critical for the overall CPI. What is going on with food inflation? We use the WPI Food to look at this:

 Month-on-month WPI Food inflation (SA, Annualised)

The values above are annualised month-on-month changes of seasonally adjusted WPI Food. This shows that from July onwards, we have had remarkably low food inflation. The CPI inflation that we have got stems from non-food inflation. Looking forward, the outlook for non-food inflation is limited because of softness in global prices of tradeables.

The poor man's statistical model of y-o-y CPI inflation is to forecast the m-o-m values using univariate time-series methods, and add up the latest 11 facts with 1 forecast to get a one-month ahead forecast. When we do this, the forecasts for February, March and April work out to 3.32%, 3.65% and 3.43%. These benign forecasts use no economic knowledge - they only reflect the time series structure of month-on-month inflation. These should be treated as the baseline on top of which we layer on economic thinking.

What about pressures on aggregate demand? There are four perspectives which suggest that the demand side will be weak in 2017 and 2018.

1. Exports growth is faring badly, partly owing to the difficulties of the global economy. The outlook for the global economy is poor, given the difficulties in China, Europe and the US.
2. From November 2016, we have been adversely affected by the demonetisation shock. We estimate that demonetisation induced a median -0.45 sigma shock to month-on-month seasonally adjusted changes in 27 macroeconomic series for November 2016, and a -0.15 sigma shock for 24 macroeconomic series in December 2016. For a comparison, when our surprise measurement methods are applied to 2008, we estimate there was a -0.25 sigma shock in September 2008 and a -0.44 sigma shock in October 2008. Demonetisation has adversely affected optimism of households. We expect that demonetisation will exert a sustained negative impact upon the economy through 2017.
3. Investment in India is faring poorly. The best measure of investment activity is the stock of projects classified as being under implementation' in the CMIE Capex database. This stalled -- in nominal rupees! -- in 2012 and has not grown for five years. Things are likely to worsen on this front in the aftermath of demonetisation.
4. We are in the midst of a banking crisis. In December 2016, non-food credit grew by 5.32% nominal when compared with December 2015, which is 1.91% in real terms. The last time we saw lower values was at the time of the Lehman crisis in late 2008.

These four problems are, of course, inter-related. We overstate the gloom when we think of them as four orthogonal issues. Each of the four is a difficult problem which resists quick solutions. As an example, consider the time series of cash in circulation:

 Cash in circulation (Trillion rupees)

If you extrapolate the straight line at the end, it will be many months before cash is back to pre-shock conditions. Similarly, consider the year-on-year changes of imports by the US from China:

 Imports by the US from China

It is remarkable to see that the recent low value was as bad as that seen in the 2008 crisis. The sluggish values here bode ill for global demand for Indian exports.

These four difficulties suggest that output and inflation will evolve in a more negative way as compared with the baseline statistical forecasts described above. In this case, CPI inflation outcomes could be knocking on the lower end of the target range.

We feel that these issues will weigh on monetary policy in 2017 and 2018. Monetary policy acts with a long lag, so we have to look ahead when thinking about policy changes today. Further, monetary policy in India is relatively ineffectual, as the monetary policy transmission is weak. Mere 25 bps changes have little impact. When monetary policy in India has to move, large moves are required. We feel that substantial reductions of the short rate are required in 2017 in order to stay at the inflation target of 4%.

The authors are researchers at the National Institute for Public Finance and Policy.

### RBI reforms in the light of the demonetisation episode

by Ajay Shah.

The demonetisation episode has shed new light on the issues of RBI reforms. The long-standing deficiencies of the institution made this event possible. We have a group of articles on RBI as an institution:

1. Ila Patnaik in the Indian Express, on transparency and the RBI board, 14 January.
2. Ajay Shah in the Business Standard, on clear thinking about independence of a central bank, 23 January.
3. Ila Patnaik and Shubho Roy on this blog, comparing the governance of the RBI against the processes seen in the US Federal Reserve and the Bank of England, 24 January.
4. Bhargavi Zaveri on this blog, war-gaming how demonetisation would have shaped up in a reformed RBI, 25 January.

## Sunday, February 12, 2017

### Universal basic income in India: An idea whose time has not come

by Ajay Shah.

#### Cash transfers to poor people

For many years, economists have advocated cash transfers to poor people as the best tool for poverty elimination. This avoids the inefficiency of in-kind transfers such as those undertaken by the Indian Public Distribution System' which is beset by operational difficulties, and transfers to the non-poor.

The basic arithmetic works out as follows:

If we deliver \$0.5/day to the bottom 20% of society, this is an expenditure of \$48 billion/year at the current Indian population of 1.3 billion. This is 2.4% of GDP/year.

This calculation is all nominal. When we say \$0.5/day this is Rs.35 per day or Rs.12,775 per person per year. This a decent scale of transfer that eliminates poverty in India today. The debate around this lies on the problem of spending 2.4% of GDP. This is a very large number, when compared with the small size of the Indian State. As an example, the total revenue of the central government is just 9% of GDP. We can debate whether spending 2.4% of GDP on this one project is a good idea. I personally think it is, provided it is part of an overall (daunting) policy project: 1. It should be accompanied by closing down all the existing poverty programs; 2. We'd need the State capacity to identify the poorest 20% of the population. 3. Even under perfect execution on the above two issues, this will increase fiscal stress. We'd need to accompany this with fundamental reform of the tax system, so as to reduce the tax-related distortions which are hampering GDP. 4. We'd need the political maturity where the government does not use this cash transfer highway to give out dole on a large scale when elections are approaching. I feel we would be playing with fire if these kinds of systems are built, without a corresponding Constitutional amendment that requires fiscal responsibility. #### Cash transfers to all This benign discussion does not carry through when you switch to Universal Basic Income'. This is a transfer to everyone. This is attractive because it removes element 2 of the work program above -- you no longer need the State capacity to identify the poorest 20% of the population. But now the basic arithmetic jumps up by a factor of 5x: If we deliver \$0.5/day to everyone in India, this is an expenditure of \$238 billion/year at the current Indian population of 1.3 billion. This is 11.9% of GDP/year. For a country where the total revenue of the central government is 9% of GDP, this is completely out of reach. There are countries where certain Universal Basic Income proposals are within reach. As an example, consider Sweden. Their population is 0.01 billion people. If they paid \$0.5/day to everyone, this would cost \$1.8 billion per year. Their GDP is pretty large -- it is \$600 billion USD per year. Hence, such a program is a cost to them of 0.3% of GDP. They might be able to afford numbers bigger than \$0.5/day also. It makes sense for them to discuss a small Universal Basic Income. #### Conclusion For rich countries, it is feasible to pay out \$0.5 per person per day to all. This is the fashionable Universal Basic Income' proposal that is being talked about in the West. In the West, the left believes this is a good idea, while others think it is a bad idea. That is a reasonable discussion.

In India, we need to first get up to a per capita GDP of nominal \\$10,000 per person per year before we start talking about this.

Like many other themes of the economic policy discourse in the West, Universal Basic Income is a discussion that belongs in the West which is inappropriately transplanted into India.We need to be grounded in our backyard, and have common sense about the numbers. We should understand our backyard, and  have an authentic sense of the important issues which should be the Indian policy discussion.

## Thursday, February 09, 2017

### Notes on Union Budget 2017-18

by Suyash Rai.

This budget was unique in the scale and intensity of anticipatory anxiety. With prior assumptions about possibilities of policymaking in India unsettled, many were worried about government's next move. Could the government add two-three percentage points to the fiscal deficit to launch a spending spree? Could there be a loan waiver, Universal Basic Income, or Massive tax cuts? It turned out to be a textbook case of workmanlike budget-making - not dazzling, but reasonably prudent. It is heartening to see that our politics can produce this budget in such a situation. One hopes that the government works consistently to reduce policy uncertainty. The budget is a step in that direction.

A budget should be judged primarily on fiscal management, and how it links to larger policy priorities. Each budget tells a story about government's priorities. However, since most of budgeting is not zero-based, important reforms tend to span several budgets. It is important to consider the budget in the context of medium-term fiscal strategy of the government, and the fiscal issues that need to be addressed in medium term. Changes to the systems of raising, allocating and spending resources are also relevant for evaluating a budget. Further, since the budget is made within a fiscal responsibility framework, changes to the framework are also pertinent. Finally, changes to the formats of reporting and accounting are also relevant.

The budget scores reasonably well on fiscal prudence, changes in the reporting formats, and reform of the budget process. It signals good fiscal marksmanship, but we cannot know this for sure until the data on actuals becomes available. As budgets in India go, this is a good housekeeping performance. However, if we take a medium-term perspective, we see that the budget indicates progress on some fronts, but does not do much to address most of the persistent fiscal issues in India. Further, some of the proposals in the Finance Bill raise worries about the future of tax administration.

### Reform of reporting formats

This is the first budget without the plan-non plan expenditure distinction. The reporting formats no longer include this distinction. The government has used this opportunity to change the reporting formats. New statements have been added, annexes have been done away with (most of them are included as statements now), some statements have been omitted, and the formats of certain statements have been changed. Here is a summary of the key changes in the reporting format of Volume I (now-called "Expenditure Profile") of the Expenditure Budget:

• Changes in reported categories of expenditure: at summary level (Statement 1), the expenditure is now disaggregated into six categories:
1. Establishment expenditures of the Centre: this category includes salaries, medical expenses, wages, allowances, travel expenses, office expenses, training, professional services, rent paid, taxes, pensions, etc. This is expenditure that is incurred for maintaining the administrative entity, as opposed to expenditure incurred on programme and schemes.
2. Central Sector Schemes: these are schemes for which the central government provides the entire budgetary support, and most of them are implemented by the central government.
3. Transfers under centrally sponsored schemes: for these schemes, the central government shares the budgetary support with State or Union Territory government (based on a sharing pattern determined by the central government). These schemes are implemented by the State/UT governments.
4. Other central expenditure: this category includes expenditure on CPSEs and Autonomous Bodies.
5. Finance Commission Transfers: these are grants given under Article 275(1) of the Constitution to urban and rural local bodies, grant-in-aid to State Disaster Response Funds (SDRF), and post-devolution revenue deficit grant. The revenue deficit grant is meant to cover gap in revenue expenditure after taking into account all the sources of revenue for states. Based on 14th Finance Commission's recommendations, 11 states receive these grants, and about a third of the grant goes to Jammu and Kashmir.
6. Other transfers (to States): this mainly includes additional central assistance for externally aided projects (given as grants or block loans), and special assistance to states.

In the summary statement, these six categories replace the categories of "plan", "non-plan" and "central assistance for state/UT plans". Continuing with the previous budgets, the "resources of public enterprises" are also reported. These include internal resources (accruals), bonds/debentures, external commercial borrowings/suppliers credit, and other resources, but do not include budgetary support to these enterprises.

• Statements based on categories of expenditures: probably the most useful inclusions in the new format are separate statements on centrally sponsored schemes (Statement 4A) and central sector schemes (Statement 4B) under various ministries, as well as a statement summarising the scheme category-wise expenditure for each Ministry (with aggregates given separately for centrally sponsored and central sector schemes). There is also a statement on allocation for important schemes (Statement 4C), which includes the major allocations under all the expenditure categories - this information was earlier scattered across various budget statements. Further, multiple statements (statements 4, 5 and 6) from the old format have been consolidated into one statement on subsidies and subsidy-related schemes (Statement 7).

• Statements on transfers to States/Union Territories: there is a statement on Transfers to Union Territories with legislatures (Statement 5), whihc includes Ministry/Department-wise information on transfers to these UTs. Earlier, this information was spread across multiple statements. Statement 18 provides information on "Transfer of Resources to States and Union Territories with Legislatures". These statements are consolidated. Earlier, this information was also spread across several statements covering plan and non-plan expenditure/outlay.

• Statements on public sector enterprises, autonomous bodies, and departmental commercial undertakings: the statements on "Assistance given to Autonomous/grantee bodies", "Resources of Public Enterprises", and "Investment in Public Enterprises" have been retained, as statements 30, 31, and 32, respectively, without change in format. The statement on "Grants-in-aid Salaries", which shows salary grants for autonomous bodies and schemes, has been made into Statement 29. This was an annex in the old format. The statement on departmental commerical undertakings, which was Statement 7 in the old format, is now Statement 8. It gives information on net budgetary support for revenue expenditure in these undertakings, after deducting the receipts of these undertakings.

• Statements on allocations for certain beneficiaries: These statements mostly remain the same, but some of them have been renamed:
• The statement on Budget Provisions for Schemes for the Welfare of Children (Statement 22 in the old format) has been renamed "Allocation for the Welfare of Children" (Statement 12 in the new format).
• The Gender Budget is Statement 13 in the new format.
• In the old format, Statement 21 and 21A provided information about schemes under scheduled castes sub-plan and tribal sub-plan, respectively. These have been renamed as "Allocation for Welfare of Scheduled Castes" (Statement 10A) and "Allocation for Welfare of Scheduled Tribes" (Statement 10B)
• Statement on "Budget Allocated by Ministries/Departments for the North Eastern Region" has been renamed "Allocation for the North Eastern Region" (Statement 11).
• The opportunity of removing plan-non plan distinction should be used to make these statements more comprehensive. Earlier, only plan expenditure towards welfare of these beneficiaries was captured in these statements, and it seems that the same expenditure is being captured in the renamed statements. Even the expenditure that was earlier classified as non-plan had components that benefited these beneficiary groups. Those allocations should also be included in these statements. For instance, the share of subsidies going towards these groups should be included in these statements. A beginning in this regard seems to have been made, as the interest subsidy, which was a non-plan expenditure in the earlier scheme of things, has been included in these statements. It wasn’t included in the statements till last year.
• Changes in positions of annexes: the new format contains no annexes. Most of the annexes in the older format have now been converted into statements. Annex 1 (Budget provisions by Heads of Accounts) is now Statement 16. Annex 2(Reconciliation between Expenditure shown in Demands for Grants, Annual Financial Statement and Budget Provisions by Heads of Accounts) is now Statement 17. Annex 4 (Contributions to International Bodies) is now Statement 21. Annex 5 (Grants in Aid to Private Institutions/Organisations/Individuals) is now Statement 9. Annex 6 (grants for creation of capital assets) has been slightly modified and is now called "Allocation under the object head Grants for creation of Capital Assets" (Statement 6). It provides information about grants given to state and UT government for creation of capital assets, and goes into calculating the effective revenue deficit (revenue deficit minus these grants). Annex 7 (Estimated strength of Establishment and provisions therefor) is now Statement 22. Annex 7A (Budget Provisions under "Grants-in-aid Salaries") is now Statement 23.
• Inclusion of statement on "Expenditure charged on the Consolidated Fund of India": a statement on all expenditures charged on the Consolidated Fund has been included. Earlier, this information was not included in the Expenditure Budget, but was provided in the Annual Financial Statement.
• Inclusion of Railways Statements: Five statements from railway budget have been appended to the volume. These include statements on: Overview of Receipts and Expenditure; Railway Expenditure; Railway Receipts; Investment (Part A: Financials; Part B: Physical Targets); and Railways Reserve Funds.
• Omissions: the annex on reliefs provided to CPSEs in the form of waiver, write-off, etc (Annex 2A) has been done away with. It used to give this information disaggregated by type of relief and the name of CPSE. The aggregate number is provided in Statement 17 (Reconciliation between Expenditure shown in Demands for Grants, Annual Financial Statement and Budget Provisions by Heads of Accounts). The decision to discontinue the annex may have been taken because this is a relatively small item of expenditure (budget for 2017-18 is Rs. 255 crore), but this is important for accountability for a type of expenditure that may be a sub-optimal use of public funds. In my view, the annex should have been continued as a statement. The annex on trends in expenditure (Annex 3) has also been discontinued. It used to provide ten-year trends for the major categories of expenditure. It was useful information for expenditure analysis. The statement could have been continued without the plan-non plan distinction.

In my view, the new formats are easier to read and understand. They are more informative. An opportunity that has been missed, and should be considered in subsequent years, is to report consolidated spending on activities, which are spread over various schemes. For example, it will be useful to have a statement that gives information about spending in different areas of activity, such as health, education, skill development. Scheme-wise information is useful, but common citizens will be better able to understand the budget if the information is given in terms of areas of activities. With the removal of plan-non plan distinction this has become easier to do.

### Reforms of the budget process

Advancing the budget day and merger of the Railway Budget and Union Budget are significant improvements over practices prevalent earlier. Advancing the budget day would help ensure that implementation of the new schemes can begin as soon as the financial year begins. It gives time to the departments and ministries to prepare for implementation and plan they spending. This is consistent with best practices in other countries.

The practice of presenting the Railway Budget separately was little more than a long-standing legacy. Although it is much bigger than other such enterprises, the Indian Railways is just one of the ten departmentally run commerical undertakings. Now, a single Appropriation Bill, including the estimates of Railways, will be prepared, instead of a separate Bill for Railways. Railways will get exemption from payment of dividend to General Revenues, and its Capital-at-charge would be wiped off. For the rest, things will remain the same. Ministry of Finance will continue to provide Gross Budgetary Support to Ministry of Railways towards meeting part of its capital expenditure, and Railways will continue to raise resources from market through Extra-Budgetary Resources to finance its capital expenditure.

### Profile of expenditure and receipts

In 2016-17, government is budgeted to spend Rs. 19.78 lakh crore. Major components of the budget, which comprise about 75 percent of total expenditure budgeted for 2016-17 are (BE: Budget Estimate; RE: Revised Estimate):

Component
Expenditure (2016-1, BE) (in Rs. lakh crore) Share in total (2016-17,BE) (in percent) Expenditure (2016-17, RE) (in Rs. lakh crore) Share in total (2016-17,RE) (in percent) Expenditure (2017-18, BE) (in Rs. lakh crore) Share in total (2017-18,BE) (in percent)
Interest payments
4.93 24.8 4.83 23.98 5.32 24.3
Defence, including defence pensions
3.4 17.2 3.45 17.13 3.6 16.76
Food subsidy
1.34 6.8 1.35 6.7 1.45 6.75
Finance Commission transfers to states and local bodies
1 5.1 0.99 4.9 1.03 4.8
Fertilizer subsidy
0.7 3.5 0.7 3.47 0.7 3.26
0.58 2.9 0.52 2.6 0.65 3.02
Central armed police forces
0.5 2.5 0.52 2.6 0.55 2.56
Railways
0.45 2.3 0.46 2.29 0.55 2.56
MGNREGS
0.38 1.9 0.47 2.36 0.48 2.24
Petroleum Subsidy
0.29 1.5 0.27 1.37 0.25 1.16
Pensions
0.29 1.47 0.3 1.5 0.32 1.49
National Education Mission
0.27 1.36 0.27 1.34 0.29 1.37
National Health Mission
0.21 1.06 0.23 1.14 0.27 1.26
Pradhan Mantri Awas Yojana
0.20 1.01 0.21 1.04 0.29 1.35
Total Expenditure
19.78 100 20.14 100 21.47 100

For most of these, the revised estimate of expenditure during 2016-17 is quite close to the budgeted expenditure. For roads and highways, the revised estimates are about ten percent lower than the budgeted expenditure. For MGNREGS, the revised estimate is about 23 percent higher than the budgeted expenditure. Being a demand-driven scheme, this is not unusual for MGNREGS.

For 5 of these, the shares in budgeted expenditure for 2017-18 are stable. For defence, fertilizer subsidy, MGNREGS and petroleum subsidy, the share of expenditure is significantly lower. The share is significantly higher for Pradhan Mantri Awas Yojana, National Health Mission, Railways as well as Roads and Highways.

Fiscal marksmanship on expenditure side significantly depends on receipts. If receipts fall short, expenditures are cut or fiscal deficit target is not achieved. In 2016-17, the expenditure is budgeted to be financed by the following receipts (numbers in brackets are percentages of total receipts:

1. Tax revenues, net of transfers to states: Rs. 10.54 lakh crore (53.31 percent)
2. Non-tax revenues: Rs. 3.23 lakh crore of (16.34 percent), about two-thirds of which were budgeted to be from proceeds of spectrum auctions, and dividends from PSUs, banks, and the RBI)
3. Non-debt capital receipts: Rs. 0.67 lakh crore (3.39 percent), which include disinvestments of shares and recovery of loans.
4. Borrowings from various sources: Rs. 5.33 lakh crore (26.96 percent of expenditure). This is the fiscal deficit, which is budgeted to be 3.54 percent of the estimated GDP for 2016-17.

The revised estimates suggest that the government is likely to collect about 3 percent higher tax revenues than it had budgeted. While the revised estimates of direct tax collections are very close to the budget estimates, those for the indirect tax collections are different. Customs collections are estimated to fall short by 5.6 percent; excise duty collections are estimated to be 21.6 percent higher than budget estimates; and service tax collections 7.1 percent higher. The government may have set a modest target for growth in collection of excise duty, in anticipation of increase in crude oil prices. If crude oil prices had indeed risen sharply, government would have had to cut the excise duty on petroleum products, and that would have led to a smaller increase in collections. Fortunately for the government, this did not happen.

The non-tax revenue collections are estimated to be 3.7 percent higher than budgeted. This is primarily on account of 43 percent higher collection of dividends from CPSEs. This is estimated to more than make up for the shortfall in collections from spectrum auctions and interest receipts. In non-debt capital receipts, while the overall receipts are estimated to be close to the budgeted amount, proceeds from disinvestments are expected to fall short by about 20 percent. So, while a lot is going on in the components, the overall receipts are better than budgeted.

On fiscal marksmanship, three significant caveats are in order. First, the government's reported numbers sometimes turn out to be quite inaccurate. Recently, a CAG audit concluded that the fiscal deficit in 2015-16 was 4.31 percent of GDP, and not 3.9 percent, as was reported by the government. It is a cause for concern that the government's reporting of actuals was off by 0.41 percent of GDP (Rs. 53,146 crore). Second, due to advancement of the budget day, this year's revised estimates were prepared using lesser amount of data on expenditure/receipts, because of which the probability of actual expenditure/receipts being different from revised estimates is higher. Third, due to uncertainty created by demonetisation, it is difficult to make good GDP and revenue estimates for this year. The budget has taken the GDP number from the economic survey, which differs considerably from the advance estimates put out by the CSO in January. Any numbers reported as percentage of GDP are subject to changes in GDP estimates.

### Fiscal prudence

According to the revised estimates, the government is expected to achieve its fiscal deficit target (3.5 percent of GDP) for 2016-17, and has set a fiscal deficit target for 2017-18 (3.24 percent) that is close to the roadmap given in the Medium Term Fiscal Policy (MTFP) statement two years ago (3 percent). The fiscal deficit target for 2018-19 and 2019-20 is 3 percent. In a rare instance, government is expected to do better than its target for revenue deficit (difference between revenue expenditure and revenue receipts). The target was set at 2.3 percent of GDP, but the revised estimates suggest that the revenue deficit this year will be 2.1 percent. This is because of higher tax and non-tax revenue collections, while the revenue expenditure is estimated to be along the budgeted lines. For 2017-18, the target is set at 1.9 percent, which is slightly higher than 1.8 percent target laid down in last year's MTFP statement. The primary deficit (fiscal deficit minus interest payments - shows whether we are borrowing to pay interest on borrowings) is estimated to be the same as budgeted (0.3 percent of GDP), and is budgeted at 0.1 percent in 2017-18.

The GDP estimates suggest that government expenditure is the main driver of growth in 2016-17, while growth in other types of expenditure is likely to be sluggish (private investment is estimated to fall). The strategy of using public investment to crowd in private investment was launched about two years ago, and it seems to have yielded underwhelming results. Perhaps the government did not consider it wise to continue down this path for more time, and is now keen to use other instruments to encourage private investments. It will have to undertake a sustained reform programme to boost private investments in the next few years.

There are several pathways to fiscal consolidation. Fiscal consolidation may involve a combination of: cutting expenditure, increasing tax revenues, increasing non-debt capital receipts (especially disinvestment and privatisation), and raising non-tax revenues (especially through user charges). The fiscal consolidation budgeted for 2017-18 is 0.3 percent of the GDP projected for the year, or about 0.5 lakh crore. How is this being achieved?

On the receipts side, the budgeted increases in net tax and disinvestment receipts are far smaller than the budgeted fall in non-tax revenues. Non-tax revenues are budgeted to fall because of lower collections from spectrum sale, and because Railways is no longer required to pay interest to government (since the budgets have been merged). So, in 2017-18, the receipts (excluding borrowings) are budgeted to be 9.5 percent of GDP - lower than they were in 2016-17 (9.82 percent). With these budgeted receipts, if expenditure grows at the rate at which GDP is projected to grow, the fiscal deficit in 2017-18 would be about 3.9 percent.

The government has bet on cuts in expenditure to achieve fiscal consolidation. This means that central government expenditure, as a percentage of GDP, is budgeted to shrink from 13.3 percent in 2016-17 (revised estimate) to 12.7 percent (budget estimate). Some of the areas where expenditure, in terms of percentage of GDP, has been cut are: defence (0.15), MGNREGS (0.04), fertilizer subsidy (0.04), food subsidy (0.04), petroleum subsidy (0.03), agriculture (0.02), and Pradhan Mantri Gram Sadak Yojana (0.02).

If most of these cuts were coming from expenditure reforms that improve efficiency of expenditure, i.e. get the same or better outcomes for smaller expenditure, they would hurt less. However, it is not clear if that is the case. Moreover, there is no significant change in the budgeted revenue-capital ratio of expenditure (from 86.1:13.9 to 85.6:14.4).

Since the economy seems to be in doldrums, a less contractionary consolidation pathway would have been more appropriate. The strategy should have comprised of substantive subsidy reforms (discussed later), an aggressive privatisation/disinvestment programme, raising non-tax revenues through user charges, and, to a lesser extent, other expenditure cuts. The budget targets for disinvestment are aggressive, but the targets for privatisation are lower than they were in 2016-17. This year, the government should have built the systems and processes for privatisation transactions, and reaped much higher receipts in 2017-18.

The deficit targets for 2017-18 must be considered in the context of fiscal uncertainties. The uncertainties of GST rollout, consequences of demonetisation, and external circumstances make it difficult to project macro indicators for 2017-18, and to achieve the targets. Government may need to review its strategy during the course of the year.

In summary, while the deficit targets are prudent, the strategy for achieving them seems sub-optimal, and due to uncertainties, it will take considerable dexterity to achieve them.

### Medium-term fiscal issues

Much has been written about the specific expenditure decisions in this budget. Except in a few areas, there is not much change in allocations this year. There are certain fiscal issues that need to be addressed in medium to long-term. Let us consider some of them and what this recent budgets has done about them:

1. Declining share of capital expenditure in defence budget: a problematic trend in defence expenditure in India has been the declining share of capital expenditure. Capital expenditur is incurred on building the "material" component of India's defence capabilites. The share of "Capital Outlay" in the total defence budget has fallen from about 33 percent in 2006-07 to 20.8 percent in 2016-17 (RE). This year also seems to have followed the trend, and the share fell from 24.36 percent in 2015-16. The FM has announced a 20.6 percent increase in 2017-18 over revised estimates for 2016-17, to take the share of capital outlay to 24.03 percent. However, about half of this increase is because capital outlays on "research and development" and "Defence Ordinance Factories" have been moved from the demand titled "Ministry of Defence (Misc)" to the demand titled "capital outlay on defence services". Without these, the increase in capital outlay is just 9 percent, which is quite normal, and would take the share of capital outlay to 21.7 percent of the total defence budget. A problem in recent years has been that capital outlays have only been partially utilised, and a significant part of the allocation lapses.

The One-Rank-One-Pension (OROP) decision has exacerbated the trend towards more revenue expenditure. The decision is quite consequential, and in my view, it was not a wise decision from a public finance and pension policy perspective. It increased the pension outlay, and because of the way it is designed, it has also introduced considerable uncertainty in budgeting for pensions (see my column on some of the problems with the OROP decision).

Most of the modern restructuring of defence organisations in other countries has focused on trimming the forces of personnel, while building up and modernising the weapon system. China has reportedly completed an exercise that left its armed forces with 300,000 fewer personnel. The expenditure pattern in India may point at larger problems of procurement systems, policy priorities, and even our grand strategy. Since more than 70 percent of revenue expenditure in defence is incurred on pensions, pay and allowances, changing the pattern of expenditure will require some difficult strategic decisions that will have human resource consequences, which no government appears keen to take.
2. Poor outcomes of social sector schemes and the shrinking role of central government: Since the 14th Finance Commission recommended sharp increase in sharing of central taxes with states, the allocations to several schemes had to be cut. Also, the sharing patterns for centrally sponsored has been changed to reduce central government’s share in expenditure on these schemes. The role that the central government plays in designing the schemes now appears anachronous.

The biggest challenge across social sector schemes has been: how to shift away from a focus on inputs, and (to a lesser extent) outputs, and focus on achieving outcomes. Take the example of school education. While we have done reasonable progress on improving inputs (building schools, hiring teachers, etc) and outputs (enrolments, access to schools, etc), India's performance on learning outcomes, as measured through learning tests, has been abysmal. In school education, central government spends just about 15 percent of the total expenditure (with sub-national government putting in the rest). It is now a marginal player in financing the sector, but continues to occupy the commanding heights on scheme planning and design. The challenge of improving outcomes varies from one context to another. Central government will need to rethink the way it uses its funds to drive change towards better outcome. States need to be given much more flexibility to innovate than they presently enjoy in practice.

Although the FM did touch upon the issue of outcomes in education, a concrete proposal has not been forthcoming. This is the situation across various social sector schemes. Government seems intent on continuing with the set ways, without doing the needful to reorient the programmes towards achieving outcomes. Each sector poses its own unique challenges, and will have to find innovative ways to deal with this challenge.
3. Distortions in major subsidies: In 2004-05, subsidies were 12.56 percent of non-plan expenditure, and 9.22 percent of total expenditure. In 2013-14, subsidies were 23 percent of non-plan expenditure and 16.3 percent of total expenditure. In last three years, there has been some decline in the share of subsidies in expenditure (estimated to be 12.9 percent in 2016-17). This is mainly because of the favorable effect of benign crude oil prices, and savings from the direct benefit transfer programme. However, most of the substantive issues of subsidy reform remain. Let us consider the top three subsidies.

Food subsidy:Food subsidy is the difference between the economic cost of food grains and the price that government charges for them. Economic cost includes the cost of procurement, transportation, storage, etc. Till 2001-02, the issue price at which food grains were sold to those above the poverty line was close to the economic cost. The price for households below the poverty line was about half of the economic costs, and Antyodaya households (poorest of the poor) were charged a nominal price (less than a quarter of the economic cost). Since then, the subsidy regime has changed. In 2002-2003, the price for grains supplied to households above the poverty line was reduced (from Rs. 8 to Rs. 6.1 per kg for wheat; from Rs. 11 to Rs. 7.95 per kg for common paddy), while prices for Antyodaya and below poverty line households were not changed. The prices for all categories of beneficiaries have remained the same since then. In these 15 years, the economic cost for wheat has increased by 163 percent, and for common paddy by 190 percent. This has led to a massive increase in food subsidy bill. The Food Security Act had frozen the issue price for food grains for certain beneficiaries for three years, but that window is now open. It is time the government reviewed the rationale for keeping issue prices frozen for so long. Subsidy should ideally be set as a percentage of economic cost, and therefore, the price should be revised annually to track the economic cost. At the same time, reforms should be undertaken to improve efficiency to keep economic costs in check.

Fertilizer subsidy: since 2010, the gap between the subsidy for urea and that for other fertilizers has widened significantly. This is because urea was not included in the nutriend-based subsidy scheme that started in 2010. There is evidence to suggest that this distortion has led to excessive use of urea, which has hurt the nutrient balance of fertilizers being used. The proportion of nutrients in actual usage is now far from the ideal proportion (see Chapter 2 of the Economic Survey, 2013-14 for a discussion on this issue). Further, the subsidy regime in urea does not discourage inefficiency, as the subsidy amount varies from one manufacturer to another. These and other problems need to be addressed to develop a reasonable fertilizer subsidy regime. Many people have proposed good ideas, such as bringing urea into the nutrient-based subsidy regime, increasing the price of urea, moving towards direct transfer of subsidy, changing the urea subsidy regime to encourage efficiency, and so on.

LPG subsidy: Although the direct benefit transfer programme is reported to have reduced the leakages from this scheme, the substantive issue of the reducing the amount of subsidised LPG sremains. There is significant evidence to show that most of the LPG subsidy goes to the non-poor. The poor use smaller amount of subsidised LPG, and therefore avail of smaller share of subsidy. Therefore, this is appropriately called a "middle class subsidy". The government had tried introducing a cap of 6 subsidised cylinders (about 85 kg of LPG) per annum, but this was later withdrawn, and the cap of 12 subsidised cylinders was restored. A good step taken last year was that the government has capped the per kg subsidy at a nominal amount, and over time, if this cap is not raised, the subsidy's salience will fall automatically. Government has also taken steps to expand access of LPG to poor households. Now, the government should consider reducing the cap of subsidised cylinders to 6 or 8.
4. Freeing up resources locked up in low-priority public sector enterprises: According to the public enterprise survey conducted by the Department of Public Enterprises, there are 298 Central Public Sector Enterprises - 235 active and 63 yet to commence commercial operations (as on March 31, 2015). A larger number of these are in sectors where there is a vibrant private sector, and there is no longer a need for public sector enterprises. However, the agenda of privatising public enterprises has been on the back burner since 2003, and the pace at which sick CPSEs are being closed is very slow. Although shares have been regularly disinvested, there have been no exits from enterprises in almost 14 years. This has meant that a large amount of resources, especially capital and land, are locked up in enterprises that should not be in the public sector at all. These resources could be freed up and deployed in higher priority areas. In each budget, a few thousand crores are allocated for these enterprises, and this money could also be used elsewhere. This is an unfinished agenda of the old industrial policy in India, and it also points at a significant allocative efficiency problem in India's fiscal management.

In the budget speech of 2016-17, the FM had announced a plan for strategic disinvestment (aka privatisation) from certain CPSEs, and set a target of Rs. 20,500 crore. The revised estimates suggest that while the government is likely to overshoot the target for disinvestment by about 11 percent, it will fall short of the strategic disinvestment by almost 75 percent. The target for strategic disinvestment proceeds in 2017-18 has been set at Rs. 15,000 crore. This year, Government also plans to list certain insurance companies, and collect Rs. 11,000 crore from the listing. Further, it has announced "a revised mechanism and procedure to ensure time bound listing of identified CPSEs on stock exchanges". These are steps in the right direction. The system of disinvestment is a well-oiled machinery. However, there is a need to expedite the agenda of closing sick and lossmaking CPSEs, and privatising CPSEs that are in sectors where government ownership is not justified.
5. Over-reliance on petroleum products for collection of indirect taxes: in 2015-16, about 68 percent of total collection of excise duty was from petroleum products. This was about 27 percent of total indirect tax collection. In 2013-14, these were 55 percent and 19 percent, respectively. Since the last round of increases in duties on petroleum products happened in late 2015-16, the contribution of petroleum products is likely to have increased in 2016-17. Since the budget seems to have largely postponed the indirect tax decisions, one can only hope that the GST rollout will be such that this risky fiscal strategy of relying on a small number of commodities for so much of tax collection is discontinued, and we are able to build a broad tax base.
6. Shrinking sharable pool: States get a share of the central government's tax collection, based on Finance Commission recommendation. This is a share of the sharable pool, which is gross tax revenue minus cesses and surcharges. Between 2011-12 and 2015-16, the sharable pool as a percentage of the Gross Tax Revenues shrunk from 89.8 percent to 82.8 percent. As cesses and surcharges came to comprise a larger portion of tax collections, the amount States received as devolution from the centre was lower than it would have been otherwise. To consider a counterfactual, had the portion of sharable pool in 2015-16 remained the same as it was 2011-12, States would have received Rs. 42 thousand crore more in devolution from the Centre in 2015-16. It is too early to say, but this trend may be halting. In 2016-17 (revised estimate) and 2017-18 (budget estimate), sharable pool as percentage of gross tax collection is expected to be 83.1 percent and 84 percent, respectively. Hopefully, with GST, the cesses and surcharges will become less prominent.
7. Medium-term approach in budgeting: the Planning Commission used to make the five-year plans, which used to be the anchors for budgeting decisions regarding a number of areas of expenditure. This brought a medium-term perspective to budgeting. The process had its flaws, and its excesses have fueled urban legends in central Delhi. For better or for worse, the system has been dismantled. The twelfth and last five-year plan ran its course from 2012 to 2017. What we have now is the absence of any clear, publicly available medium-term perspective in budgeting. This has consequences for fiscal management, as many important priorities need to be pursued over the medium-term. Although there are talks about NITI Aayog coming up with Vision and Strategy documents, so far, there is no indication of the government moving towards a formal and comprehensive medium-term fiscal management framework.

The FRBM-mandated Medium Term Fiscal Policy Statement serves only as a basic ingredient for fiscal discipline over medium-term. It includes top-down estimates. Since we no longer have any other medium-term anchor for budgeting, it is important for India to move towards a medium-term budget framework, which would help the government make better forward estimates and think about strategies across areas of expenditure, so that annual budgetary decisions for various schemes and programmes can be reconciled with the medium-term framework. This would require combining a top-down approach and a ground-up, negotiated approach to medium-term fiscal management.

The Finance Bill proposes certain amendments to the Income Tax Act to change the powers that tax authorities enjoy:

• Under Section 132(1), the tax authorities have the power to conduct search and seizure, if they have reason to believe that the person has not disclosed the information asked for, is not likely be submit the required information, or is in possession of valuables that may have been accumulated from income on which tax was not paid. The proposed amendment says that the "reason to believe" need not be disclosed to anyone, including to any authority of Appellate Tribunal. This amendment is proposed to take effect retrospectively from April 1, 1962, which is the date when the original provision was enacted.
• Section 132(1)(A) empowers the authorities to expand the search and seizure to include locations that are not included in the authorisation for search and seizure, as long as they have reason to suspect that this would yield useful information. This section is also being sought to be amended to include an explanation that the "reason to suspect" will not be disclosed to anyone, including to any authority of Appellate Tribunal. This amendment is proposed to take effect retrospectively from October 1, 1975, which is the date when the original provision was enacted.
• Section 132 is also proposed to be amended to insert sub-sections that will give powers to the authorised officer conducting search and seizure to provisionally attach, for a period of up to six months, property that they find during the course of a search and seizure. This would be done with the prior approval of of senior officers.
• Section 133 empowers income-tax authorities to call for information for the purpose of any inquiry or proceeding under the Income Tax Act. At present, if there is no proceeding pending against a person, this power can only be exercised by senior officers above a certain rank. This section is being sought to be amended to give this power to junior-ranking officers as well.
• Section 133A empowers income-tax authority to conduct a survey at a place where a business or profession is carried on or a place where documents or property relating to the business or profession are kept. This section is proposed to be amended to include places of charitable activities as well.
• Section 133C empowers certain income tax authorities to issue notice calling for information and documents for verification of information in its possession. The proposed amendment would empower the Central Board of Direct Taxes to make a scheme for centralised issuance of these notices.

The amendment to 133C is potentially an improvement, as it might reduce arbitrariness in the issuing of notices. However, the other amendments mentioned above may have unintended negative consequences. There may be arguments in favour of these amendments. For example, it would be easier to protect the identity of whistleblowers if reasons to suspect are not disclosed. Provisional attachment during search and seizure could make it easier for tax authorities to extract revenues from tax evaders. However, the powers being given through these amendments can also be misued to conduct arbitrary searches and seizures, provisionally attach properties, and disrupt people’s lives and businesses, all without having to explain the reasons behind the entire process. Important checks and balances are being proposed to be diluted.

These amendments can be seen in the context of the 25 percent increase targeted for personal income tax collection in 2017-18. Government has proposed changes to tax rates that would lead to Rs. 15,500 crore lower personal income tax collection. Accounting for this, the targeted increase in income tax collection is about 29.2 percent. Between 2010-11 and 2015-16, the average rate of growth in income tax collection was about 15 percent. In 2016-17, because of the one-time collection under the income disclosure scheme, the rate of increase is estimated to be 23.35 percent. An increase of, say, 15 percent can be considered to be normal, and the additional 14.2 percent (about Rs. 50,000 crore) would have to be mobilised through special measures. Given the state of the economy, the only way to get a 29.2 percent increase is to expand the tax base by getting more people to pay taxes, and by making further demands from those who may be under-paying the taxes.

As the Economic Survey, 2015-16 (see page 109 onwards), pointed out, given our level of economic development, India's income tax collection compares favorably with other countries. In fact, the Survey found that income tax collection is significantly better than expected at our level of economic development. For example, India's income tax to GDP ratio is 2.1 percent, while the ratio for Brazil is 2.3 percent. To account for the theory that democracies tend to tax and spend more, the Survey controled for democracy as a variable, and the finding on personal income tax holds, albeit the overall tax to GDP ratio is lower than it should be. While the percentage individuals paying taxes is much smaller than expected, the amount of personal income tax collected is actually better than one would expect at this per capita income. This mismatch between satisfactory income tax collection and low number of income tax payers may be because income is concentrated in a smaller number of individuals, but this requires further research.

There is tax evasion and tax avoidance, but there may not be enough "low hanging fruits" that can be plucked to yield Rs. 50,000 crore of additional income tax collection over and above the normal increase in collections. The important issue is that expanding fiscal capacity is a long-term task that requires building capabilites in the tax administration, while upholding the rule of law and the basic principles of government accountability. In a context where income tax collections are good for the level of development, a target to deliver a huge increase in collections, may tempt the tax administration to use their expanded powers to take draconian measures to extract taxes. In the process, innocent people will get hurt. For example, the tax administration would cast a much wider net to go after those who deposited cash after demonetisation than they would normally have. This is not a good way to build fiscal capacity. Rule of law and accountability of government are as important, if not more important, than collecting more income tax.

### Conclusion

This reading of the budget suggests that while the budget has got the basic housekeeping of fiscal management right, it is a middling performance on addressing important fiscal issues that need to be addressed in medium term. Further, the pathway chosen for fiscal consolidation, although not necessarily bad, is sub-optimal because of the state of the economy. Finally, the amendments to the Income Tax Act proposed in the Finance Bill should be reconsidered, because they may harm basic principles of rule of law and government accountability.

The author is a researcher at National Institute of Public Finance and Policy. Views expressed here are personal.