One of the proposed solutions to India’s public procurement problems is new legislation to govern how the government buys goods and services from the private sector. Will a law help India? We connect two data sources to test this idea. Instead of a new law, monitoring public procurement, identifying failures, and then building state capacity may be a better solution. Legislation may not be the silver bullet for our problems.
In India, legislation is often viewed as a panacea when faced with policy problems. Whether it is bankruptcy, privacy, warehousing, or medical testing in private laboratories; the government is quick to propose a new law to solve problems. The same approach has been attempted to address the issues of public procurement. In 2012, the government introduced the Public Procurement Bill with the stated reason as:
“Major countries of the world have well codified legal provisions governing public procurement.” (Statement of Objects and Reasons).
An international organisation also prescribed this solution for India. In 2013, the United Nations Office on Drugs and Crime recommended that India should enact the Public Procurement Bill. According to the UN agency, the bill would improve public procurement and reduce corruption. The bill lapsed, and the government changed. However, the idea that a law is needed persists. The present government had plans for introducing similar legislation. In the 2015-16 budget speech, the finance minister stated:
“Malfeasance in public procurement can perhaps be contained by having a procurement law and an institutional structure consistent with the UNCITRAL model. I believe Parliament needs to take a view soon on whether we need a procurement law, and if so, what shape it should take.” (Paragraph 72)
The present government is yet to introduce a bill.
It seems intuitive that a better law should improve public procurement. More transparent systems that make procurement information widely accessible and encourage more firms to participate, deter kickbacks and other forms of fraud and corruption (Ware et al.). Countries with legal provisions which discourage governments from closing bids to select vendors or establish an independent dispute resolution mechanism seem to have less bribery of public officials (Knack et al.). However, better laws may not necessarily result in better outcomes (Sukhtankar and Vaishnav and Bosio et al.). In this article, we look at the correlation between the state of the procurement law in a country and the outcomes from public procurement.
Parliamentary laws and corruption outcomes
The first step towards measuring the outcome is to agree on metrics of the quality of public procurement. The quality of a procurement law/system may be determined by multiple variables such as the conservation of public resources, purchase of better products, timely payment to vendors and integrity. However, we do not have data to measure these. We suggest an interesting proxy that we do observe: corruption perception. The predominant form of corruption, in most countries, is corruption in public procurement. Therefore, one of the primary objectives of making a public procurement law is to reduce corruption. We hypothesise:
If adopting a law improves public procurement, we should see lower corruption in those countries.
World Bank’s 2017 Benchmarking Public Procurement Database(BPP). This is a comparative evaluation of the legal systems governing public procurement in 180 countries (World Bank BPP, 2017). Experts analyse the laws governing public procurement on eight criteria. The criteria start from the preparation before a tender is published and extend to dispute resolution and complaint management systems. Economies with more extensive legal frameworks score higher on the BPP than countries with less comprehensive legal frameworks for public procurement. In this sense, the BPP measures the extent to which a country has accepted and implemented the idea that a better law for public procurement is desirable.
Transparency International’s Corruption Perception Index (CPI). Transparency International scores jurisdictions based on the perception of corruption in a country’s public sector. It is based on opinion polls and surveys across countries. Low scores mean higher corruption and higher scores imply high government integrity.
We look at the correlation between the World Bank’s BPP score and the Corruption Perception Index. We collected BPP data for 2017 and the CPI data for 2019 (latest years). We narrowed down the countries present in both databases, which yields information about 163 of 180 countries (91.12% of the datasets).
Findings
As Figure 1 shows, We find no correlation between the BPP scores and the CPI scores of countries. It is particularly interesting to look at the countries where the two run in different directions. Italy and Kazakhstan have very similar BPP Scores (79.33 and 79.50) but very different CPI Scores (34 and 53). China has a much higher BPP score than Hong Kong (74.66 against 48.66), but in CPI scores, China does significantly worse than Hong Kong (41 and 76). India (61.50), Australia (60.83), and Singapore (60.50) have very similar BPP Scores, but very different CPI scores (41, 77, and 85, respectively). Russia is 14 points ahead of the United Kingdom in the BPP but significantly behind on the CPI by 49 points.
Figure 1:Quality of Law and Corruption
Similarly, as Table 1 shows, the Bahamas, Hong Kong and Barbados rank quite high on the CPI (little corruption) but do quite poorly on BPP ranks. On the other hand, Kazakhstan, Congo and Yemen have high corruption (low CPI score) but score higher on the BPP.
Table 1: Comparing Rankings
Country
CPI score
BPP score
CPI Rank
BPP Rank
Barbados
62
40.20
30
157
Hong Kong
76
48.66
16
141
Bahamas
64
44.66
29
151
Kazakhstan
34
79.50
104
2
Congo
18
64.33
155
43
Yemen
15
64.66
162
41
This evidence is consistent with the arguments by Sukhtankar and Vaishnav and Bosio et al. that better laws do not correlate with better outcomes in public procurement.
What might be going on?
Why is there no correlation between corruption and quality of public procurement laws? Two reasons may explain our observations: isomorphic mimicry or imperfect measurement.
Isomorphic mimicry: ‘Isomorphic mimicry’ is the ability of organisations to sustain legitimacy through the imitation of the forms of modern institutions, but without functionality (Andrews et al.). Countries may adopt laws and institutions which are considered global best practices. However, the laws are not enforced, and the institutions are ineffective. One of the reasons for the observed results could be that countries are adopting law intending to score high on an international indicator without the requisite state capacity or active institutions to implement such a law. While this creates the facade of a sound legal system, the on-ground reality is quite different. International aid agencies sometimes require that a country have a sound legal system for public procurement, where superficial measures such as passing a law are considered sufficient. A government trying to attract international donors might pass `modern’ legislation to showcase or appeal to donors, foreign academics, journalists or NGOs. However, the government may have no intention or capacity to implement the law.
Imperfections in the BPP: The BPP as a measure appears to have a sensitivity problem. The OECD has overarching public procurement guidelines with which all members have to comply. We should, therefore, see OECD countries cluster towards the higher end of the CPI and BPP scores. While this holds for CPI scores, it does not, for BPP. BPP scores of OECD show much more variance than their CPI scores. The fact that OECD countries have adopted a common framework on public procurement appears to be not captured by the BPP measurement system.
The BPP may fail in measuring the quality of procurement laws in a country because of invisible infrastructure. Invisible infrastructure is the superset of general laws, institutions and accountability arrangements in the country which are crucial for determining the success of specific policy intervention (Kelkar and Shah).A common law country like the UK may have binding precedents setting transparency and accountability standards but may not have legislation. Constitutional provisions governing equality before the law or requiring due process apply to government procurement. Freedom of information laws may bring about transparency generally and may apply to procurements. Governments may have general laws which require government agencies to appoint an ombudsman or inspector general. Such offices may take active steps to reduce corruption and settle procurement disputes. However, such rules are not captured in a measurement system like the BPP as it is limited to government procurement legislation (Bosio et al.) The elements of invisible infrastructure may suffice, in itself, to generate high-quality procurement absent a law, and invisible infrastructure may matter in shaping the consequences of any procurement law. In either event, by focusing on the procurement law we tend to not notice the binding constraint, the invisible infrastructure.
Looking ahead
Before making laws, we need to identify the causes of the poor performance of public procurement in India. We have a history of failing in implementation and monitoring in India. Both require robust, invisible infrastructure which is missing. The first step is to build the load-bearing capacity of the procurement system. Pritchett et al. point out that premature load-bearing arising from unrealistic expectations about the level and rate of improvement of the ability of a state lead to stresses and demands on systems that cause capability to weaken if not collapse.
Two websites which aggregate procurement across government departments may provide clues on how to improve state capacity. The Government E-Marketplace (GEM) and the Central Procurement Portal (CPPP), operated by the central government, aggregate and standardise procurement notices across various government bodies. These websites aid the procurement process in many ways. Tenders are made public on a common portal instead of being scattered across multiple publication sources. This increases competition as bidders are less likely to miss a tender because they do not buy a specific newspaper. The method of tender publications is standardised, and this helps bidders apply for tenders with lesser effort. Moving away from paper-based systems reduces the chance of bids getting lost.
The more significant benefit from these websites is that they allow the government to measure/monitor the quality of the procurement process (outcome measurement) across multiple variables. This is better than measuring the quality of some legislation (input measurement) of BPP. The CPPP website publishes 16 performance indicators derived from the transactions carried out on the site. For instance, in 2019-20, 23% of the open tenders were not awarded within the bid-validity period. i.e. the buyer did not finalise the transaction in time. Sadly, most of the performance indicators tracked by the CPPP website, since 2016, show no discernable trends that procurement performance is improving.
Other jurisdictions have implemented interventions, similar to the performance indicators in the CPPP website, to improve public procurement system. The Government Accountability Office of the U.S. publishes performance reports on government procurement (which does worse than Kazakhstan on the BPP Score). Instead of legislating, India may benefit from looking at the performance indicators on the CPPP website and working on improving them every year.
We should not be lured by silver bullets, such as enacting legislation. While legislation has a role to play in governance, the evidence indicates that it is not a panacea for our problems. Some countries with good outcomes do not necessarily have an extensive legal framework for public procurement. Some nations with comprehensive laws continue to demonstrate poor results. The pathway to a better procurement system perhaps lies in detailed research that integrates public administration, law and public economics.
Shubho Roy is a researcher at the University of Chicago. Diya Uday is a senior researcher at the Finance Research Group, Mumbai and visiting faculty at the Tata Institute of Social Science, Mumbai.
The Covid-19 pandemic has provided us with fresh insights on health policy in India. One key element of this thinking lies in a careful understanding of what elements of public health are best done at the city/district level, at the state level or at the union government. The Constitution of India has allocated the tasks in some detail. Considerable policy research work is now required, to bring life to the Constitutional scheme, based on a first principles understanding of the work that is required in public health, drawing on our experiences of 2020.
Market failure in health policy
There are great insights that can be obtained in the field of health policy by applying the toolkit of market failure. It is best to define the task of government as addressing market failure, and market failure comes in four categories: concentration of market power, presence of positive or negative externalities, presence of information asymmetry, and the need to provide public goods. There is a neat split in the field of health: public health is about public goods and externalities, while health care may contain market power and asymmetric information.
Public goods are a compelling example where the government is central, and the things that are not done by the government are hard to achieve through purely private initiatives other than pure philanthropy. Knowledge is the ultimate public good -- once a research paper is released on a website it is non-rival and non-excludable -- and we need public funding for research. When one person coughs and communicates Covid-19 to another, this is a negative externality, and there is some role for the government in reducing this externality. The main task of health policy thinking lies in analysing the landscape of public health, identifying the market failures (public goods and externalities), defining the tasks of the government, and finding a path to achieving state capacity on these functions.
Where should each function be placed?
Once we have a picture of the various functions which have to be performed in public health, we come to the question of the best place where it should be performed: the union government or the state government or the local government. The famous `Subsidiarity principle' of public economics asserts that every function should be placed at the lowest level of government where it can possibly be performed.
As an example, Amy Harman and Farah Stockman have an article in the New York Times which describes the treatment of travellers from China into the US. The federal government (which we in India call the union government) is the right agency to track flights and obtain lists of passengers. After this, there is a handover of information, that person x flew in from China, to the local government where that person resides. At this point, the local government is the one best equipped to work on contact tracing, testing, and isolation. This is an optimal allocation of the two tasks. It is hard for a local government to keep track of who flew in from China. It is hard for the union government to manage front line staff in a city or a district.
It is interesting and important to think about the elements of a public health system, and to think about the optimal placement of each of these elements, between the union, state and local governments. However, we do not engage in policy thinking on a tabula rasa. We do policy thinking in India where the Constitution of India has a well-developed point of view on these questions, and amendments to the Constitution on this aspect are rare. Hence, our puzzle in thinking about public health in India lies in taking full cognisance of the Constitutional scheme and best adapting it for our present understanding.
Health in the Indian Constitution
The distribution of subjects in the Constitution is reasonably elaborate. It sets up a division of labour between different levels of government, viz, the union, state, panchayat (rural local bodies), and municipalities through a list of subjects which are enumerated in its schedules VII, XI, and XII.
The Seventh Schedule of the Constitution lists the distribution of the subjects between the union and the states, while the eleventh and twelfth schedules deal with the distribution of responsibilities at the local level, i.e., panchayats and municipalities. Every policy thinker in India needs to fully understand these three schedules. Table 1 summarises the distribution of subjects in the domain of public health.
Government
Subject
Reference
Union
Port Quarantine
Schedule VII, List I, Item 28
Union
Union agencies and institutions for professional, vocational or
technical training, etc.
Schedule VII, List I, Item 65
Union
Co-ordination and determination of standards in institutions
for higher education or research and scientific and technical institutions
Schedule VII, List I, Item 66
Union
Inter-state migration and inter-state quarantine
Schedule VII, List I, Item 81
State
Public health and sanitation; hospitals and
dispensaries
Schedule VII, List II, Item 6
Concurrent (both union and state subjects)
Lunacy and mental deficiency, including places for reception
or treatment of lunatics and mental deficients
Schedule VII, List III, Item 16
Concurrent
Medical education and profession
Schedule VII, List III, Items 25 and 26
Concurrent
Prevention of the extension from one State to another of
infectious or contagious diseases
Schedule VII, List III, Item 29
Panchayat
Health and sanitation, including hospitals, primary health
centres and dispensaries
Schedule XI, Item 23
Panchayat
Family welfare, women and child development
Schedule XI, Items 24 and 25
Panchayat
Social welfare, including welfare of the handicapped and
mentally retarded
Schedule XI, Item 26
Municipality
Public health, sanitation conservancy and solid waste management
Schedule XII, Item 6
Municipality
Safeguarding the interests of weaker sections of society,
including the handicapped and mentally retarded
Schedule XII, Item 9
Table 1: Distribution of 'health' related subjects in the Indian Constitution
There is a significant role of union government in subjects relating to contagious diseases and pandemics. It is also responsible for setting standards of medical education and profession along with the state government. On the other hand, state and local bodies are responsible for most public health functions such as sanitisation and family welfare.
A simple reading of the distribution of functions induces many questions. For instance, vaccination is a public health function which is a part of state list under the Constitution. This is logical, given that immunisation programs require a large front-line workforce that interacts with the population. However, the design of the standard package of vaccinations for all kids, and envisioning ambitious projects like the eradication of smallpox or polio, require thinking and coordinating by the union government.
Similarly, in a public health crisis such as COVID-19 all levels of government are required to perform their specific functions that are elements of the overall public health response. These elements include tasks such as planning, funding, managing and on-ground implementation. These elements are not described in detail in the Constitution but are an important part of the legal and policy mechanisms adopted by the government.
There is at present relatively little in place, in India, by way of Parliamentary law which shapes and circumscribes the work of public health. The British-era Epidemic Diseases Act, 1897, has many problems. The legal framework under which India is responding to the COVID-19 crisis is the Disaster Management Act, 2005 which sets up a National Authority whose role is briefly discussed below.
The role of the National Authority
The Disaster Management Act, 2005 is the union law that was used by the union government in its Covid-19 response. In this Act, a disaster is defined to be:
a catastrophe, mishap, calamity or grave occurrence in any area, arising from natural or man-made causes, or by accident or negligence which results in substantial loss of life or human suffering or damage to, and destruction of, property, or damage to, or degradation of, environment, and is of such a nature or magnitude as to be beyond the coping capacity of the community of the affected area;
Under this law, the National Authority is responsible for drawing a national plan for disaster mitigation, prevention, and preparedness. This plan is to be reviewed and updated periodically. The law also recognises the role of multi-level governments as it sets up the national, state and district level authorities which are responsible to follow the guidelines of the National Authority.
The National Disaster Management Plan in India was last updated in November 2019, its only revision after the first plan was released in 2016. While the plan deals with Biological and Public Health Emergencies (BPHE), it does not provide detailed guidelines on the structural frameworks required for dealing with a global pandemic at the scale of COVID-19. In this sense, India does not have a national plan to deal with the COVID-19 crisis as of now. It would be useful to design a national plan which guides the government in undertaking a well-coordinated action to deal with the crisis. The national plan should be mindful of the spatial element of the public health interventions in COVID-19 such as:
Inter-state migrations, operations of flights require intervention by the union government.
Hospital preparedness, such as the presence of an adequate number of hospital beds, medical equipment such as ventilators and oxygen etc. require intervention at the state level.
Contact tracing and quarantine enforcement require intervention at the municipal or local level.
A guidance document by the National Authority with conceptual clarity about the elements of public health will be useful to minimise policy failures in COVID-19 management. At present, some clear policy failures in COVID-19 management are being observed. These failures are at all levels of the government, the union, state, and local levels. Some of them are described below as illustrations:
Union-state coordination
Actions taken by the government during a pandemic have political repercussions and therefore, a tension between the state and union government priorities can exist. For instance, in Delhi, the elected government and the Lieutenant governor had disagreed on the conditions being imposed on businesses during the lockdown period leading to uncertainty for the public.
Varying state priorities
Border state conflicts relating to inter-state travel of persons became common in the early period of the COVID-19 pandemic. In the first week of April, Karnataka state sought intervention of the Supreme Court to resolve a dispute regarding border movement with the neighbouring state of Kerala during lockdown imposed due to COVID-19. This was after the Kerala High Court passed a verdict asking Karnataka to allow movement of persons between the states. Eventually, the union government was involved in reaching an amicable settlement between the states regarding conditions of movement of persons during the lockdown.
Varying priorities of local bodies
The local bodies are empowered to take action in public interest under the Disaster Management Act. During the COVID-19 crisis, it was observed that local bodies failed to take into consideration the impact of their decision on neighbouring districts. For instance, the Noida district administration barred entry of persons from the Delhi border without a pass issued by them. This caused trouble to essential workers such as doctors and nurses who worked across the district border who would be left stuck at the border without knowledge of requirements for such a pass.
Heterogeneity within the vast country
There is great heterogeneity within the 3.3 million square kilometres of India, in the state of the epidemic, in trade-offs between mobility and disease control, and in state capacity. There is great value in having democratic legitimacy in each city or each district in choosing the optimal path.
While working through the Disaster Management Act was expedient when faced with the pandemic, as the dust settles, there is a need for health policy thinkers to envision a public health system for India. It is important to, lay this on sound legal foundations, whereby the Disaster Management Act is ultimately focused on natural disasters like earthquakes, and public health has its own legal and institutional architecture that is fit for this purpose.
Conclusion
There is a need to bring greater coherence to all the elements of state power that are in play in the response to Covid-19. This has led to twin challenges of a) micromanagement by the union bodies, and b) excessive delegation of powers to the state and local governments without adequate checks and balances. For instance, approval for Covid-19 testing labs throughout the country is done by a single body, the ICMR, an approach that has difficulties. Similarly, certain orders by district and state authorities have also been criticised during the course of the pandemic for being arbitrary.
We should utilise our fresh understanding of the present problems, to build a body of knowledge on (a) What are the tasks of public health in India (b) What is the role of the union / state / local government in each of these and (c) How to achieve state capacity on each of these components?
K. P. Krishnan is Professor at National Council of Applied Economic Research (NCAER).
The banking sector is the most important financial intermediary in India's debt market. Over the last few years the bond market has emerged as an alternative to the banking sector especially for the top rated firms. This trend has been pronounced ever since the banking sector started reporting high levels of non performing assets. Figure 1 below shows the flow of commercial credit in India from various sources and highlights the growing relative importance of bond issuance especially from 2015 onwards.
The bond market has faced two big shocks in recent years: (i) the default by IL&FS (Infrastructure Leasing and Financial Services Limited) in September 2018, followed by other relatively low-impact shocks due to problems in companies such as DHFL (Dewan Housing and Finance Limited) and IndiaBulls Housing Finance as well as Yes Bank, and (ii) the outbreak of the Covid-19 pandemic in India since March 2020. As a result of these shocks the risk perceptions in the bond market have gone up. In this article, we take a look at changes in the risk perceptions in the corporate bond market especially in the ongoing context of the pandemic and ensuing economic slowdown. We also highlight the asymmetry in the risk perceptions of the markets towards private sector corporate bonds vis-a-vis public sector unit (PSU) bonds and discuss the likely implications of changes in the risk perceptions, for the future funding model of non-banking finance companies (NBFCs).
Figure 1: Flow of Commercial Credit in India (Source: RBI)
Measuring risk perception
The most important metric for assessing risk perception in the bond market is the credit spread which is the difference between the yield of a corporate bond and of a government security of comparable maturity. Highly rated bonds (with ratings of AAA and AA) are traded relatively actively and their yields reflect changing perceptions of investors regarding the riskiness of these bonds. Movement over time of credit spreads on corporate bonds is therefore a good indicator of the bond market's perception of risk.
We look at the credit spreads of AAA rated bonds of 3 years and 5 years maturity from April 2018 to June 2020. The data is sourced from Bloomberg. The bonds in our data are separated into 3 categories - NBFCs (non-banking finance companies) and HFCs (housing finance companies), private corporations and public sector undertakings (PSUs), which may include public sector NBFCs such as Power Finance Corporation (PFC) and Rural Electrification Corporation (REC). The figures 2 and 3 below show the evolution of credit spreads for these three categories of bonds for the two specific maturities.
The IL&FS default
Figure 2: Credit Spreads on 5 Year AAA Paper (Source: Bloomberg)
As we see from figure 2 above, prior to September 2018, the credit spreads on the NBFC, private corporate and PSU bonds were fairly stable, between 50 and 100 basis points for the 3 year paper and between 40 and 60 basis points for the 5 year paper. In the rest of our discussion we focus on the credit spreads on the 5 year paper. The pattern is more or less the same for the 3 year paper, only the absolute levels of credit spreads are different.
Figure 2 shows that credit spreads on NBFC AAA paper of 5 year maturity nearly doubled between September 2018 and November 2018 and reached 160 basis points by February 2019. This shows that the IL&FS episode that unfolded in the 3rd week of September significantly enhanced the risk perception of the bond market regarding all top rated NBFCs.
After a small dip, the spreads went back to around 140-150 basis points by July 2019 and stayed at this high level, with some fluctuations, till November 2019. During this period, crisis in other NBFCs (such as the Dewan Housing and Finance Limited (DHFL)) as well as in Yes bank, added to the overall risk perception of the bond market. This is reflected in the credit spreads remaining high one year after the IL&FS default.
Private corporate and PSU bonds' credit spreads also widened in the aftermath of the IL&FS default, but not by the same magnitude as the NBFCs. The IL&FS default triggered a liquidity crunch primarily for the NBFC sector. The corporate sector experienced spill over effects owing to a rise in risk aversion in the bond market.
While in the pre IL&FS default period the spreads of all three categories of bonds were closely bunched together, the difference between them began increasing from October 2018 onwards. The difference was particularly acute between the NBFC and private corporate bond spreads on one hand and the PSU bond spreads on the other hand especially in the second half of 2019. This is despite the fact that these bonds were all rated AAA. This reflects the implicit government guarantee enjoyed by the PSU bonds.
The government and the RBI took several actions to deal with the ensuing crisis in the NBFC sector. Government appointed a new Board for IL&FS. RBI took several steps including open market operations to inject liquidity into the system, reducing the risk weights on bank lending to NBFCs, instructing banks to disburse sanctioned but undisbursed credit to NBFCs etc.
These eventually resulted in enhanced credit flow to the NBFCs which reduced the credit spreads in the later part of 2019. For both NBFCs and private corporate sector, the spreads declined by about 50 basis points to settle at about 100 and 50 basis points respectively. These spreads, especially for the NBFCs, were still higher than pre-IL&FS episode but much lower than their peak. We see a similar dynamic with the 3 year maturity bonds as well as shown in figure 3 below, except the absolute levels of the spreads were different.
Figure 3: Credit Spreads on 3 Year AAA Paper (Source: Bloomberg)
The Covid-19 outbreak
Just as the bond market was recovering from the shock of IL&FS default followed by crises in DHFL and Yes bank, the Indian economy got hit by another massive shock in the form of the ongoing Covid-19 pandemic. Credit spreads in the bond market began rising sharply from the middle of March once again reflecting growing risk perceptions. Figure 2 shows the increase in the spreads around the time when the nationwide lockdown was announced on 24 March.
For both NBFC and corporate bonds, the spreads rose by about 30-40 basis points between February 2020 and April 2020. For both categories of bonds the credit spreads reached their peak in the first half of May, close to 180 basis points for NBFCs and 170 basis points for the corporate bonds. The peak of the credit spreads during the pandemic has so far been higher than the peak reached in the aftermath of the IL&FS default episode.
Spreads on PSU paper also went up, but by a smaller amount. The average spread on these bonds in March and April was only 30-35 basis points. The difference between the credit spreads on NBFC and corporate bonds on one hand and PSU bonds on the other widened significantly to about 100 basis points. The large gap in spreads for bonds of the same ratings is worth noting. Similar to the post-IL&FS period, this too is a reflection of the market's perception of implicit government guarantee to the public sector units.
The impact of policy actions on credit spreads
The sharp rise in credit spreads of NBFC and corporate bonds in April 2020 could be attributed to the announcement by the RBI to grant moratorium on loan repayments for all borrowers in order to alleviate the financial stress triggered by the pandemic and the lockdown. Following this announcement, NBFCs had to offer moratorium to their borrowers but at the time it was not clear whether they themselves would also receive a moratorium from banks on their repayment obligations.
In the second half of May, the government announced a package to boost the economy. This included Rs 20 lakh crore of 'benefits' and effectively entailed an outlay of around Rs 3 lakh crore for 2020-21. RBI also adopted several policy initiatives such as cutting the policy interest rates aggressively and establishing new long term targeted repo operations (T-LTRO) that would provide 3 year funding to banks under a repo arrangement. RBI made the repo arrangement `targeted' so as to ensure that the funds raised by the banks were made available to the NBFCs.
These policy actions increased the credit supply to all issuers. Consequently, by the 3rd week of June, the credit spreads on both NBFC and corporate bonds came down from their respective peak levels of mid May by about 50 basis points.
However, the RBI and government actions notwithstanding, the credit spreads for NBFCs and private corporate sector continue to be substantially high. In fact the spreads in June 2020 were similar to the spreads in December 2018 in the aftermath of the IL&FS default. For PSUs the spreads have come down to around the same levels that prevailed before the IL&FS crisis.
This shows that the bond market remains concerned about the riskiness of the corporate sector and the NBFCs. PSUs on the other hand, benefit from implicit government guarantee. The significantly lower credit spreads they are experiencing in the time of the pandemic reflect a `flight to safety' by the bond investors.
Credit spreads and funding costs
As we interpret the bond market data, it is important to understand the difference between credit spreads and funding costs. Credit spreads going up does not necessarily mean that the cost of funding for the issuer is going up. Cost of funding for a company that raises capital in the debt market depends on the market determined yield on the security it issues This yield on debt consists of two components: risk free rate and credit spreads. RBI's monetary policy impacts the risk free rate but not the credit spreads. Credit spreads reflect the premium that the investor charges over and above the risk free rate, taking into account the inherent riskiness of the underlying bond.
Since the IL&FS episode, the risk free rate has been coming down steadily due to the actions by the RBI such as reduction in the policy interest rates (repo and reverse repo rate) and large scale open market operations to inject liquidity in the financial system. Figure 4 below depicts the yield on 5 year and 3 year government securities from the April 2018 to June 2020 period.
Figure 4: Government Securities Yield
The 5 year risk free interest rate has come down from about 8.4% in September 2018 (before the IL&FS episode) to about 5.5% in June 2020 indicating a decline of 300 basis points. The 3 year risk free interest rate has declined even more to about 4.5% over this period, a decline of nearly 350 basis points.
Since RBI's monetary policy does not affect the credit spreads, the impact of policy action on the actual cost of funding will not be the same as the reduction in the risk free rate. If risk aversion in the market goes up, then investors will demand higher price for the credit risk which will result in rising credit spreads. Thus, the net cost of funding for an issuer may decline to a lower extent compared to the reduction in the policy rates.
This is what has been happening since the IL&FS episode. Risk free rate has been declining but owing to high risk aversion, credit spreads have remained elevated. As a result, funding costs of companies have not come down by as much as the risk free rate. This implies that in an environment of high and rising risk perception such as the ongoing Covid-19 period, the effectiveness of policy rate cuts will be constrained.
The widening gap between the credit spreads on PSU debt versus private sector points to lower risk perception for PSU entities which are perceived to have implicit sovereign guarantees. The combined effects of rising risk perception, widening gap between credit spreads of identically rated issuances and reduction in the policy interest rates would mean that the debt market will skew towards government owned issuers who might experience the greatest reduction in funding cost.
Conclusion
Bond market credit spreads provide important information about the risk perception of an important class of investors. Sustained high credit spreads (compared to long term average levels) suggest elevated risk perception and imply heightened risk aversion. Specifically, it also points to the role that individual episodes of corporate defaults and the associated policy responses (or lack thereof) play in shaping risk perceptions.
Wide spreads between bonds of the same ratings issued by private companies and those owned by the government clearly indicates a strong perception of the implicit government guarantee enjoyed by public sector companies. This raises important questions as to whether the debt of government owned companies should be treated as a part of government's debt.
Finally, economic recovery in India in the post Covid-19 period will depend crucially on the flow of credit in the economy. The economic package recently announced by the government depends largely on the financial sector. Nearly 70% of the 'benefits' of Rs 20 lakh crore in the package are expected to be routed through the financial sector. In a recent article we discussed the rise in risk aversion in the banking sector. With both the banks and the bonds markets showing high levels of risk aversion, growth of credit may be less than envisaged in the package. This may dilute the overall effectiveness of government's monetary and fiscal policy actions.
Harsh Vardhan is an Executive-in-Residence at the Center for Financial Studies and an Adjunct Faculty at the SP Jain Institute of Management and Research, Mumbai. Rajeswari Sengupta is an Assistant Professor of Economics at IGIDR, Mumbai.
1. All JuliaCon Events Are Free and Online:Click here now to register for JuliaCon. JuliaCon continues from now until Friday July 31. The full JuliaCon schedule is available here, and includes 10 workshops and over 100 talks.
3. Julia Computing Virtual Booth: Connect
with Julia Computing live by visiting our Virtual Booth Wed July 29 -
Fri July 31 from 12:30 pm to 7:30 pm UTC. To be connected, click here and select the Julia Computing Virtual Booth link when it is available during these hours.
School shutdowns across the country have sparked disagreements between parents and schools about fees. Parents filed a plea in the Supreme Court seeking more time to pay schools due to COVID-19. The Supreme Court refused to hear the petition, arguing that it had to be tackled by the executive first. Schools need money to pay their staff, including teachers, and are threatening to cut off access to online classes in case of non-payment (here, here and here). Both parties have approached High Courts in at least 15 states for a ruling (for example, in West Bengal, Madhya Pradesh, and Gujarat).
COVID-19 has only exacerbated an old fight. School fees have been an oft litigated issue in India. Courts have pronounced judgements against capitation fees, profiteering and fee hikes for over two decades (See 1992, 1993, 2002, 2004 and 2019). Several states also enacted laws to regulate school fees. Such price regulations are a poor way of addressing the underlying issue of market failure in school education. The primary reason for fee disputes with parents is that the entry of new schools in India is severely restricted due to a cumbersome regulatory environment.
In this article, we discuss the fee regulation architecture across India. In particular, we focus on the legislative drafting and implementation of the UP Self-Financed Independent Schools (Fee Regulation) Act, 2018.
Regulation of school fees in India
Nine states and union territories inIndia have stand-alone Acts regulating the collection of fees. These Acts were passed between 2009 and 2019 and mandate a Fee Regulatory Committee to hear fee-related complaints and proposals.
Biharlimits fee hikes to 7% over the previous year’s fee. Schools which wish to increase charges beyond 7%, need to seek approval from a divisional fee regulatory committee. These committees are usually composed of parents, private school representatives, and government officials.
Gujaratlaw empowers the district fee regulatory committees to determine fees for schools. However, schools charging fees below a specified amount are exempt from the regulation.
Rajasthan and Maharashtrarequire a school-level committee to approve fee hikes. The school-level is composed of representatives from school management, teachers and parents. If the school-level committee fails to agree on the increase, the school can approach a divisional fee regulatory committee.
The Maharashtra law adds a nuance, missing in Rajasthan. Schools can choose between a block declaration or capped revision. During admission, schools can declare fees for a block of classes (for example, grade 1 to 5). Or, a school can revise fees subject to a cap. Under this option, costs cannot be raised more than 15%, once every two years. In case of unforeseen circumstances, schools may increase fees beyond this cap. But, such increases, above the cap, has to be approved by either 76% of the parents, or the school-level committee. The Act also allows for management and parents (not less than 25% of parents in the affected standard/school) aggrieved by the decision of the school-level Committee to approach the divisional fee regulatory committee.
Other states pass orders and notifications to regulate fees but struggle with implementation. In Delhi, districts are supposed to set up Fee Anomaly Committees. But these are either not constituted or defunct, so parents raise their complaints to the Directorate of Education (Agarwal et al. 2019).
Uttar Pradesh
Uttar Pradesh enacted the UP Self-Financed Independent Schools (Fee Regulation) Act, 2018 to control fees for schools. Sadly, the law suffers from two drafting problems: a contradiction in fee fixation provisions, and the lack of clear instructions on the price index to use.
Section 3(1) of the law lays down the heads which the school can take into consideration. It is the governing principle of the law on fixing school fees. It reads:
"A recognised School shall determine its fee structure under subsections (1) and (2) of Section 4 … commensurate to, inter alia, meeting its operational expenses, providing for augmentation of facilities and expansion of infrastructure and for providing facilities to the students, to generate reasonable surplus to be utilised for development of educational purposes including establishment of a new branch or a new school under the management of the same eligible educational entity;"
The law is refreshingly pro-school in this provision. It recognises that a fee increase is not just to meet operational expenses. Schools have to augment facilities, expand infrastructure (build new classrooms, maybe a swimming pool), provide facilities to students. The law also recognises that a school must be allowed to generate some reasonable surplus and may wish to expand by opening new branches or schools. It is a surprisingly frank and forward-looking recognition of the myriad expenses that a school faces. In India, where price controls laws rarely recognise the costs that the provider has to undertake, the provision stands out as one recognising the genuine needs of the school.
Section 3(1) states that the process of determining fees is laid down in Sections 4(1) and 4(2).
Section 4(1) completely undermines the approach of Section 3(1). It reads:
“A recognised school may revise its fee annually for its existing students by itself for each grade/class/level of school equivalent to average percentage per capita increase of monthly salary of teaching staff of previous year, but the fee increase shall not exceed latest available yearly percentage increase in consumer price index [CPI] + five per cent of the fee realised from the student;”
(emphasis added)
Gone are the grounds recognised in Section 3. Section 4(1) reduces all those grounds to only one: teacher salaries. No more can schools increase fees to pay for expanding infrastructure, providing facilities, opening branches or generating a surplus. The only amount that the school can raise fees is the increase in salaries. If a school does not increase teacher salaries but wants to build a new auditorium, it is out of luck. Buying a new computer lab? Section 4(1) will not allow you to raise money for it. School’s financial reserves are low? Section 4(1) has no solution for management. The only criteria for school fees increase are teacher salaries. All the good ideas in Section 3 have been washed away by the restrictions in Section 4.
The pegging of fee increases to teacher salary is indicative of a deeper problem in Indian education: teacher interest domination. Too often, laws designed for education end up protecting teachers. For example, the only performance measure that the Right to Education Act enforces in the parent legislation itself is the teacher-to-student ratio, even when the evidence of the effectiveness of this measure is weak. This one measure is baked into the Parliamentary law. All other performance measures are left to be decided by the government through subordinate legislation.
The second problem in the UP Act arises from the formula under Section 4(1). The law provides a cap on the fee hike. Fee hikes have to be less than CPI + 5% per year. The drafters have left out defining CPI. In India, there are two bodies which provide five types of consumer price indexes. The Labour Bureau publishes two indexes, (CPI industrial workers and rural workers) and the Ministry of Statistics publishes three indexes (Rural, Urban, and Combined). Narrowing down to the applicable index is not the end of your problems. These indexes are published monthly, while the fee increase is supposed to happen once a year. The law is silent about which CPI to use and how to convert the monthly numbers into a yearly value. Predictably, different district fee regulatory committees have come up with different values for the maximum fee hike. Gautam Budh Nagar calculated the maximum fee hike as 7.88% (5+2.88), and Varanasi calculated the same as 8.71% (5+3.71).
This problem could have been solved by clearly cross-referencing to the specific index that the schools should use. Since the value of the index does not change across districts, there is no need for each district committee to decide the CPI. This function could have been done at the state level itself and saved schools from the confusion.
Conclusion
Why is the UP law drafted so poorly? The underlying reason is that the legislators have misidentified the problem. The best way to regulate prices is through a market mechanism. Parents should have a wide choice of schools at different price points. Sadly, we do not have that in India. Regulatory burdens imposed on schools reduce the supply of private schools. State governments control the availability of land in urban areas, mandate minimum salaries for teachers, and impose many requirements on schools through state school laws and the Right to Education Act. The consequence of these laws is two-fold: it raises the costs of running a school and makes it difficult to set up new schools. In turn, the existing schools raise fees. The entry barrier to new schools ensures that they do not face any competitive pressure to reduce fees.
Instead of encouraging competition in private schooling, the laws put administrative controls over the fee setting mechanism. An administrative price-setting usually misprices the fees. A government committee is in no better position in deciding what the price of education in a school should be. Even the legislature is unable to articulate any principles by which such committees should determine fees. The U.P. law starts with a wide range of costs that a school may incur. But when it comes to the implementation clause, it narrows down to just teacher salaries.
The price control laws are trying to solve a problem which should not exist in the first place. It would be much better if we tried to identify and dismantle the entry barriers to setting up low-cost private schools in the first place and encourage competition.
References
Does Class Size Matter?, Ronald G. Ehrenberg, Dominic J. Brewer, Adam Gamoran and J. Douglas Willms, Scientific American, November 2001.
How are private school fees regulated?, Ritika Agarwal, Atreyi Bhaumik, Adit Shankar and Anindya Tomar, in Anatomy of K-12 Governance in India, Centre for Civil Society, October 2019.
Bhuvana Anand is a researcher at Centre for Civil Society and Shubho Roy is a researcher at the University of Chicago. The authors thank Tarini Sudhakar at Centre for Civil Society for research support.
The Ministry of Electronics and IT, Government of India (MeitY) had sought public comments on a Consultation Whitepaper (CW) titled a "Strategy for National Open Digital Ecosystems (NODEs)" earlier this year. NODEs are defined as:
open and secure delivery platforms, anchored by transparent governance mechanisms, which enable a community of partners to unlock solutions and thereby transform social outcomes.
The NODES framework will allow the opening up, and sharing of personal and non-personal data held in various sectors (such as healthcare, agriculture, and skills development). Each NODE will consist of infrastructure developed and operated by the government. The private sector will utilise the common infrastructure and data to provide solutions to the public. Per the CW, this will enable greater intra-government and public-private coordination and create efficiency gains. This framework will promote access to innovative e-governance and other services for citizens while enabling robust governance processes to be implemented.
We wrote a detailed response to MeitY. In our submission, we make suggestions on four key issues with the CW:
Role of the state: The CW needs to demonstrate clarity on the need for government intervention on the scale proposed. The market failures that require State intervention must be identified on a sectoral basis.
Centralisation of governance and technical systems: The CW envisages establishing monolithic, stack-based digital systems in a variety of sectors. The government would be responsible for establishing and operating the technology infrastructure as well as the governance of such systems. However, excessive centralisation can reduce competition, innovation, and produce unsecure systems.
Alignment with the existing government policies: The CW needs to consider existing government policies on the adoption of open source software (OSS), open APIs and open standards. Further, the CW needs to account for existing open data and e-governance related initiatives in the identified sectors, and how these would interact with the NODEs framework.
Preserving and protecting Constitutional norms: The CW needs to ensure the protection of fundamental rights, democratic accountability, and transparency in the creation and regulation of NODEs. Further, it also needs to account for the federal division of competencies enshrined in the Constitution.
This article summarises our comments and suggestions on the above mentioned issues.
Role of the State
The CW adopts a 'solutionist' approach, in that it does not undertake sufficient analysis of the circumstances and problems in each sector. For instance, the CW identifies two market failure in the skills sector: (i) information asymmetry amongst the stakeholders, and (ii) a lack of trust in the information that is available. It proposes a Talent (Skilling and Job) NODE as a one-stop solution to connect employers, job seekers, counsellors and skilling institutes. Instead of the approach undertaken by the CW, one should consider if private entities can or are already innovating to bridge the information asymmetry and trust issues in the sector, and what policies could provide an environment where such information asymmetry may be reduced. If the problem in the skills sector is a lack of trust, it is unclear why this cannot be solved by interventions such as certification standards.
As a general rule, the State should be involved with building technological systems only for essential state activities (Kelkar and Shah, 2019). It is therefore critical to differentiate between sectors where the State has a legitimate role (say in the provision of its welfare and statutory functions), from sectors where private sector solutions could suffice. For example, the State could have a role in providing access to Public Distribution System (PDS), but need not be a player in building a platform for access to rail reservations.
The responsible ministry should analyse if the NODE is serving welfare or other essential function of government. In case there is no such element, the government should not use its finances on creating infrastructure for such a NODE. Such an approach would promote innovation, prevent the emergence of a state-centric technological mono-culture, and allow the private sector to respond appropriately to requirements of any particular sector. Entities would not be forced to build on top of state-mandated infrastructure, which may not always be necessary or appropriate.
In the context of the NODEs framework, the State should primarily have three roles:
Open up data: The government must focus on building databases and providing access to the public, in a non-discriminatory manner. The benefits of enabling free flows of information are well known. That said, it is important to keep in mind the need to ensure non-discriminatory access to ensure data quality, and to prevent against privacy and other downstream harms. For instance, the Delhi government recently shared locations for COVID-19 relief centers on Google maps, thereby giving Google a competitive edge over other mapping solutions. We believe that an appropriate approach would involve the Delhi government making the relevant information open. This can be done by providing the geo-tagged locations on its open data governance website. Methods to embed this data in third-party apps and services could be provided to enable non-discriminatory access. Similarly, the benefits of opening up railways related data, which is currently monopolised by the IRCTC can enable the provision of customised travel solutions. Greater linkages could be formed with private players in the hospitality and tourism sectors, leading to mutual benefits to the railways as well as the private sector and consumers.
Implement regulatory frameworks: The government should institute regulatory processes and norms based on the need to protect and promote fundamental rights and correct market failures. Interventions must be designed to (a) promote effective competition and the maintenance of a level playing field, (b) avoid function creep, (c) protect and promote fundamental rights, (d) ensure appropriate apportioning of functions, obligations and responsibilities/liabilities.
Ensure democratic accountability: It is now well-established that "code is law" (Lessig, 1999). This makes it imperative for the government to establish systems of democratic accountability, transparency and openness in the creation and regulation of public digital systems. Transparency and accountability measures should be implemented both at the conceptualisation stage as well as thereafter. This should involve:
A cost-benefit analysis that takes into account the economic costs and benefits of operationalising a NODE within a sector. This would also allow for suitable alternative approaches to be explored.
Integration of principles of participatory and democratic governance into the implementation and operation phases. This would promote citizen-centric governance, particularly in the context of privatisation of regulatory functions. For example, the National Payments Corporation of India (NPCI) functions as a quasi-regulatory agency due to the scope of its powers, functions, and de-facto regulatory monopoly. However, being a private entity, it has not been brought under the purview of the Right to Information Act, 2005. This limits citizen engagement with governance processes.
Mechanisms to enable allocation of responsibilities and coordination between government entities at different levels (local, state, and central). This is especially important when dealing with common issues (such as tagging of data sets, instituting grievance redress mechanisms, etc.) without usurping constitutional and statutory functions.
Centralisation of governance and technical systems
Enabling the government to pick technological winners and losers or enabling a technical monoculture would decrease innovation and competition. It is well-recognised that centralisation can lead to increased security concerns. One must also be wary of unintended consequences of even the best planned regulation in the technology space. Technology moves too fast and has multiple possible future use cases. Over-regulation or excessive centralisation could have negative effects on expected outcomes.
In cases where the government is required to create digital systems, these must be federated and decentralised to the extent possible. The creation of monolithic technical architectures, which are often de facto mandatory, must be avoided. For instance, the creation of a centralised identification system -Aadhaar- which was thereafter mandated for use across different sectors has caused various problems ranging from exclusions, intrusions into privacy rights of citizens and inhibiting innovation (i.e. such a system is preferred over other possible forms of identification that could suffice in any particular use-case). Implementing a centralised system of 'public infrastructure' may therefore not be necessary and may in fact reduce competition and civil liberties protections.
Instead, the focus of the government should be on enabling the private sector to develop relevant platforms and technologies that compete with one another on a level playing field, albeit with due consideration for regulatory, human rights and other problems that may arise in any given context. Such a system would also promote greater security. The use of federated databases, enabling the development of alternative technical solutions to be built on data, etc., would mean that problems associated with having a single source of truth or a single source of failure can be avoided.
Alignment with existing government policies
The CW proposes principles of open and interoperable delivery platforms. There are two concerns in the manner in which these are described in the CW.
The scope of the word 'open' as used in the CW is vague and appears to confuse concepts of "open access" and "open source". The CW suggests that each NODE will require a different degree of openness to adhere to specific objectives, context, or mitigate potential risks. This approach can dilute existing policies (mentioned above) that contain clear definitions and mandates on the use of open source solutions by the government.
It is imperative that the NODEs framework build on and strengthen existing government commitments towards the use of OSS solutions. This will unlock the benefits of OSS/Open APIs/open standards such as enhanced security and verifiability, no vendor lock-in, etc.
Preserving and promoting constitutional safeguards
The creation of NODEs platforms would significantly impact fundamental rights. We envisage three instances where the NODES environment needs to be careful about preserving constitutional safeguards.
Right to equality, right to life, and personal liberty: Digitisation at the scale contemplated by the CW may lead to concerns about access to services and possible exclusions therefrom. Ensuring rights protection may be particularly important in the context of the use of AI-based solutions and possible discrimination that may arise as a result. The understanding of what amounts to discrimination must be evolved by each NODE distinctly and will depend on the sector.
Right to privacy: Each of the NODEs will invariably result in the collection and processing of personal data and non-personal data by both government and private entities. The collection and use of personal data by different state entities must necessarily satisfy the tests laid down by the Supreme Court in the Puttaswamy decisions (2017 and 2018). Similarly, principles relating to the use of data by the private sector as laid down in the context of the Aadhaar judgment (Puttaswamy, 2018) must also be adhered to. Due regard must also be given to (the developing) regulatory frameworks concerning personal and non-personal data.
Federal structure: The NODEs framework must also consider the impact on the division of subject matter competencies under the Constitution. One could envisage benefits arising from NODEs in areas such as agriculture, judicial services, healthcare, etc. However, these sectors fall under the State List in the Seventh Schedule to the Constitution. Implementation of NODEs in these sectors should not result in de facto centralisation of federated competencies. Instead, mechanisms to ensure coordination and cooperation between different levels of government must be considered.
We, therefore, recommend that each NODE be backed by an appropriate statute, to the extent possible. This would ensure greater democratic deliberations, prevent excessive and arbitrary executive action, set out the rights of citizens and private entities, and clarify the scope/ limits of any particular project. Providing statutory backing would also limit mission creep, while delineating rights and obligations and governance processes. For instance, despite its various faults, the statutory mandate provided to the Unique Identification Authority of India and the restrictions on data sharing in the Aadhaar Act have proven invaluable in ensuring that biometric and other data is not made freely available for non-Aadhaar purposes by the public sector, including for instance, in criminal investigations. In contrast, projects such as FASTags (which aims to digitise highway toll systems) are being gradually expanded with plans to integrate the system with criminal tracking networks, amongst others.
Conclusion
The CW provides a basic overview of the concept of a NODE and identifies certain sectors in which such a system could lead to gains (such as the skills and health sectors). For various reasons outlined in our submission, our recommendation is to not proceed with implementing the NODEs framework in the manner currently outlined in the CW. We believe that the CW should be seen as an exploratory document. Greater clarity is required on the need for interventions on the scale envisaged in the document, particularly in view of the proposed centralised, stack-based approach. The NODEs framework should consider the need for openness at lower layers of the stack (infrastructural layers), adhere to existing government policies on the use of OSS, Open APIs and Open Standards, and consider policy developments concerning the regulation of personal and non-personal data. The CW should also ensure greater transparency and democratic accountability of governance frameworks and the processes for the creation of a NODE.
Michealson, 2017: Rosa Michealson, Is Agile the answer? The case of UK universal credit, in Grand Successes and Failures in IT - Public and Private Sector, IFIP Advances in Information and Communication Technology, Springer, 2017.
Ministry of Rural Development, 2013:Ajeevika skills guidelines, Ministry of Rural Development, Government of India, 2013.
