In Service of The Republic: The Art and Science of Economic Policy
"Market failures come in four kinds: Externalities, Asymmetric information, Market power and Public goods."
"No private firm would produce public goods such as clean air as residents cannot be billed for the services rendered. It is not possible to exclude any person from the benefits. Excludability is essential for firms to obtain a revenue stream. Hence, we require state action to obtain tax revenues and provide the public good."
"When faced with a proposed state intervention, our first question should be: What is the market failure that this seeks to address?"
"from 1993 onward, in the US, federal officials were required to formally demonstrate the presence of market failure before planning interventions in the economy."
"we see three pillars of intervention: production (e.g., government running schools), regulating (e.g., government regulating private schools) and financing (e.g., government paying kids to attend private schools). These flow from the limited toolkit of intervention."
"One day, Chairman Mao organized millions of people to kill all the sparrows in China, on the grounds that sparrows eat grain."
"With the sparrows removed from the scene, the insect population surged, and the insects ate the grain. Chairman Mao may have had good intentions, and he may have had practical experience of sparrows eating grain, but the intellectual foundation of the intervention was faulty."
"The expenditure per student in government schools rose from Rs 2455 in 2010 to Rs 4385 in 2016. Learning outcomes declined in this period. In 2010, 50.7 per cent of children in class 5 in government schools could read a class 2 level text, but this declined to 42.2 per cent in 2014. Parents voted with their feet, and switched from free government schools to private schools charging a fee. Enrolment in government schools fell by 11.1 million while it rose by 16 million in private schools, from 2010 to 2016. 5"
"There is black humour in the phrase ‘the law of unintended consequences’. A government intervention that is intended to have a certain outcome will very often end up yielding a very different result. Such failures happen so often that these have been elevated to the level of a humorous ‘law’. It requires extreme clarity of thought to devise policy initiatives that hit the target; all too often the actual outcomes stray far from what was desired."
"We need to change course, from viewing government as a benign paternalistic agent that means well, to a more sceptical view of government. We need the state in order to address market failure, and we must give the state the monopoly of violence; but we have to be sceptical about the things that will be done by the state."
"The Indian state works poorly, experiences panic when faced with a crisis, comes out with an announcement of a policy package that seeks to address the crisis in a short-term way, and retreats back into neglect. Smooth, capable working on an everyday basis has not come about."
"There is a vast unfinished agenda, in India, of market failures where not enough work has been done. The justice system should be the first priority of civilization. This requires far-reaching work in rethinking laws, police, prisons, public prosecutors and courts. We have major new dimensions of health crises emanating from air quality, road safety, antibiotic resistance and substandard medicines. All these areas require subtle understanding of market failures, and commensurate state interventions."
"Numerous elements of state coercion present in India lack justification, insofar as there is no market failure that can justify the use of state power. Why does the state control the time at which a stock exchange opens and closes, or the objects that it trades in? Why does the state seek to imprison a farmer if her produce is not sold at a physical market specified by the state? Why does the state plan to put young people into jail if they trade in crypto-currencies? Why is the state forcing corporations to spend 2 per cent of their profit on good works?"
"What you measure is what you can manage."
"Taxation imposes a cost upon society owing to these distortions. The total cost of public expenditure is not just the Re 1 that is directly spent, but also the distortions or lost GDP associated with raising Re 1 of tax revenues. These issues lead up to a critical concept in public finance: the marginal cost of public funds (MCPF). The MCPF tells us the cost to society of Re 1 of public spending. This is typically much larger than Re 1."
"India relies heavily on the corporate tax, and has double taxation of the corporate form. In the last decade, corporate income tax and the dividend distribution tax added up to 35 per cent of total tax collection."
"Similarly, 80 per cent of the countries which introduced the Goods and Services Tax (GST) after 1995 have opted for a single rate GST. In India, various pressure groups lobby for higher or lower taxes on one industry or another, and this distorts the resource allocation of the economy. This drives up the MCPF."
"All these factors induce high inefficiencies upon the economy when the government chooses to spend Re 1. There is remarkably little work on measuring the MCPF in India. We believe that the value for India may be about 2.5 to 3.5. As a thumb rule, it is useful to reckon that the cost to society for every rupee of public spending is around Rs 3. The notion of Marginal Cost of Public Funds brings a whole new perspective upon revenues and expenses of government. It encourages us to be very frugal in spending. We should only spend Re 1 when we are sure that the gains to society exceed Rs 3."
"In the best advanced economies, the numerical value for the MCPF is from 1.5 to 2."
"Public administration, the management of government organizations, is harder than management in the private sector, for three reasons. The government wields coercive power (to compel private persons) or the power to spend public money (that is, in turn, obtained by coercing private persons to pay taxes). This is not found in private firms. Private firms keep score using profits or share prices. There is no comparable dashboard in government. Private firms have competitors, and face the threat of going out of business, while the government is a monopolist that will never go out of business."
"When the State trespasses beyond what is legitimately within its province, it just hands over the management from those who are interested in frugal and efficient management to bureaucracy which is untrained and uninterested except in its own survival."
"Without the toolkit of public choice theory, we tend to think that the solution to one failed bureaucracy is to set up another bureaucracy. We are willing to believe that the Lok Pal will be better than the CBI. We are willing to think that a new independent agency which reviews official statistics will be better than the Central Statistics Office (CSO). Public choice theory encourages us to think that all officials and all politicians are cut from the same cloth. We have to construct systems of checks and balances, that will work through rational incentives of all parties, and without assumptions that any one person is a saint."
"Politicians and officials are not benevolent; they are self-interested actors. What appears to be an entrenched mindset, or an entrenched organizational culture, is always endogenous to incentives. A government organization that is riven with corruption is not one which was unlucky to get a lot of corrupt people. It is one where the rules of the game facilitate corruption."
"The puzzle of policy design is that of finding the checks and balances, and the rules of the game, through which politicians and officials will generate good outcomes for society when they pursue their own self interest."
"In the 1860s, the US Congress paid railroad builders per kilometre of rail. This gave incentives to builders to take the longest route between two points. 5"
"Later, universities brought high-powered incentives into the management of researchers. Universities demanded publication orientation from researchers. This was done by counting publications in general, and particularly valuing publications in ‘high-prestige journals’. Once researchers were given career gains in return for publications, there was a strong incentive to increase the number of publications, even if this reduced the quality of research. In the extreme, ‘predatory journals’ have sprung up which exchange money for publications."
"Policymakers of a socialist vintage feel they should have a large number of levers of control through which prices can be controlled. Modern thinking in public economics guides us into focusing on the appropriate role for policy: addressing market failures. The state is not here to control prices based on rival political influences; the state is here to address market failures."
"By this logic, the presence of one large zombie airline harms the viability of all private airlines. When Air India is privatized, and the flow of public money into Air India is halted, all private airlines will benefit, as the prices charged by Air India for tickets are likely to go up."
"When a government faces a soft budget constraint, there is a greater temptation to support zombie firms to walk the earth for a few years more. This is one reason why public sector companies are a problem for the economy: there is a greater risk of them becoming zombie firms backed by the exchequer."
"The willingness of entrepreneurs to start a business requires an economic environment of limited liability, where the entrepreneur will be able to give up, put the firm into the bankruptcy process, and walk out of it with nothing more than a bruised ego, reputational damage, and valuable experience. A society that pillories entrepreneurs, and turns business failure into protracted disputes or entanglement in agencies, is one which will have less entrepreneurship."
"the powers of banking regulation have been utilized to block competition against incumbent banks (e.g., by preventing foreign banks from operating in India, and by preventing the entry of new Indian private banks) and competition against banking (e.g., by preventing adjacent industries from competing against banks for their business). In Indian banking, market power has been induced by RBI. Similarly, the market for agricultural products has been organized around monopoly power for agricultural produce market committees (APMCs). The state has promised to punish farmers if they sell to anyone other than the APMC. In Indian agriculture, market power has been induced by the APMC Acts."
"At the same time, we should recognize that no country got out of poverty through redistribution. All countries which managed to escape from mass deprivation did so through sustained GDP growth that played out over many decades. We should never lose sight of this prioritization of growth-oriented policies."
"Before Ronald Coase, economists viewed negative externalities as a story with a perpetrator and a victim. It is easy to slip into the assumption that the fishermen are the ones who have to be protected. However, the Coasean analysis yields an important result. Whether the fishermen have property rights, which limit the steel mill’s ability to pollute, or the steel mill has property rights, which limit the other side’s ability to fish, is not important. As long as property rights are clear, both sides will be brought to the table to negotiate. Coase showed us that both parties have a shared interest in finding the right solution, and minimizing total harm. The answer lies in their negotiation, and when this is feasible, we do not require state coercion."
"Consider a windmill company which places noise-making wind turbines near a residential community. The command-and-control approach is to view the noise as a negative externality. We would build a government mechanism which caps the noise that windmills can make. We would set up a bureaucratic procedure to identify violations and punish them. The Coasean approach consists of bringing the residents and the company into a negotiation. The company should pay something to the residents in return for the discomfort."
"Every legal instrument should state its objective at the outset. After three years have elapsed, an empirical examination of whether these objectives were met should be mandatory. In some situations, there is a role for ‘sunset clauses’ where laws are automatically repealed after a certain time period elapses."
"In the demonetization episode, such practices would have been quite useful. It would have helped to clearly state the objective up front. It would have helped, to undertake calculations about the costs being imposed upon society. It would have helped, to ask whether there were less intrusive interventions through which the same objectives could be achieved."
"In order to do better, and obtain the institutionalized application of mind, these practices need to be codified. Parliamentary law must encode requirements for cost–benefit analysis and ex post review of laws and regulations."
"As a thumb rule, each doubling of GDP (in real terms) calls for a fairly far-reaching change in the organization structure of government. It calls for a substantial rethinking of the boundaries and functions of departments and agencies, and of government processes."
"in the US, the federal Administrative Procedures Act (APA), 1946, prescribes such a machinery that must be used by all federal agencies."
"In 2017, SEBI decided to force these to be shifted from cash settlement to physical settlement. This is a major action, which has a substantial impact upon the life of market participants. The decision was taken without identifying a market failure, measurement or experimentation. There was no ex post review, so we do not know the magnitude of the adverse impact of this decision upon market quality."
"It would perhaps have made sense for the government to start by experimenting with these bans in small towns, and simultaneously establish a data-gathering and research process to measure the impacts. Perhaps the scaling up could be from a few towns to a district to a group of districts. At each step, evidence is required before making the next move. The packaging industry would also have time to develop and scale up plastic-free alternatives."
"Do things that you can undo."
"do public policy in India, we need to think like an engineer (what are all the moving parts, and the incentives that hold them together) and not a driver (a practitioner who operates in a given institutional apparatus)."
"Every reform hurts certain firms and certain persons. With more advance warning, they can plan their life better. This would reduce the costs for the economy as a whole."
"Healthy democracies are those where various interest groups are able to sit together, engage in discourse in good faith, and emerge with reasonable compromises. The essential foundation for the democratic process of negotiation is trust—a certain presumption of good faith."
"The trust building that is required for this is hampered when policy measures are hatched in secrecy and suddenly unveiled upon the populace."
"Reaching out to critics is tactically costly as this increases the say that critics have in stalling or subverting a reform. The authoritarian impulse, to favour action over talk, is fashionable. However, the best policy work gets done in the open. The participatory policy process, grounded in intellectual debate, generates a better reforms process. It helps avoid a policy process that is pure power play. The work is better rooted in the landscape of people and institutions."
"One of the most harmful things that is taking place in the Indian state is announcements that show up on a website in the evening, without any previous notice, and are effective next morning. These impose huge costs upon private persons, and drive up the ex ante fear of policy risk in the minds of firms."
"Authoritarian regimes, that neglected the value of feedback, have repeatedly failed in the human experience of thousands of years."
"In contrast, democracies look messy. They are riven with debate, dissension, and a tug of war. Power is dispersed across many individuals and many elements of the government. Neutral and intellectual voices weigh in on the conflicts that are played out in the public domain. This policy process is a world of ideas and rational thinking, and not merely an exercise in power play. The continuous debate in an environment of dispersed power is the reason why democracies work well. The continuous process of criticism and debate finds and solves mistakes."
"Public choice theory encourages us to see that everyone involved in government works for herself, and all policymaking is marred by conflicts. Conflict is the normal state, and we should not be uncomfortable about it. Differences between persons and agencies are normal and healthy, and should be played out in the public domain. Differences arise out of conflicting interests, differences in information sets, and legitimate differences in how information is analysed."
"If two people agree on everything, only one is doing the thinking."
"In the past, Indian policy reform consisted of low-risk projects like dismantling industrial licensing and dismantling barriers to globalization. The analytical clarity on those issues was strong and there was little that could go wrong. But as these first-order ideas are used up, future policy work will not be as unambiguous."
"In authoritarian regimes, all achievements belong to the leader, which makes it difficult to ever admit that a mistake was made. When credit is given more accurately, to the cast of thousands that work on any one policy initiative, it is easier for the leadership to accept that things have not worked out well and to introduce course corrections. The best management culture is one in which we get credit for not taking credit. This helps create a healthy environment of trying things, accepting that some did not work, and abandoning or fixing the troubled ones."
"The leadership of a private firm fears financial non-performance, which will ultimately lead to the loss of jobs and empire. Persistent weak performance can induce a sale of the firm to a new shareholder, who can impose painful changes upon the firm. Private firms face the threat of a bankruptcy process where the firm can be shut down or fundamentally reorganized. None of these possibilities influence employees in the government. There are rare events where a government agency is closed down. Politicians fear losing elections. Officials have no fear."
"Companies have feedback loops where the daily MIS shows how things are faring, where quarterly financial statements are put out, and the stock price is updated in real time. When mistakes are made, they kick off corrections. Governments have no comparable feedback loops."
"Firms wield no coercive power. They have to be nice to customers all the time. Governments wield coercive power. The danger of functionaries that mistreat individuals is ever-present. Process design for government organizations involves establishing checks and balances against this threat."
"A lot of decisions in private firms can be tactical. Governments become the most effective when they establish rules rather than discretion, and eschew day-to-day tactical responses. Governments are bound by equal treatment (Article 14 of the Constitution of India) while private firms have no such constraint."
"Firms tend to work on short time horizons. The best governments work on long time horizons."
"A country is not a company."
"During World War II, the British rulers established extreme restrictions upon the Indian economy. After Independence, these evolved into a central planning system. From the late 1980s onward, this has morphed into an ‘administrative state’, the rule by officials who have fine-grained control over the decisions of private persons, with functions that veer into legislative and judicial roles. This involves excessive interference in the lives of private persons. However, for the major part, this is the structure of the Indian state, and to do public policy in India is to operate within the contours of this administrative state."
"the SEBI Act, unlike the US Securities Act, prohibits private persons from suing for violations of securities law. Only SEBI has the right to initiate such actions. Victims of securities law violations do not have this right."
"India lacks a class action lawsuit mechanism, so there is no mechanism for (say) the shareholders of Satyam to sue. The design that was adopted in building SEBI was that only SEBI can initiate enforcement actions against a person. This is how we got to the administrative state. Now the difficulties of public management impact upon the enforcement process. Thus, persons who have experienced harm are supplicants before SEBI, requesting SEBI to enforce securities law."
"The agenda of liberalization and economic freedom requires going back to the foundations of common law: to a world of contracts, torts, private enforcement and class action lawsuits. This is what is required to scale back the administrative state."
"When the courts fail, we get the rule of officials."
"There is a bias, in India, for muscular responses to current newspaper stories. Letting the field unfold under laissez faire for four years is an unlikely event in the Indian policy process."
"Academic journals do not publish policy proposals, hence academic researchers are not keen to invent policy proposals."
"Investments in the early stages of the policy pipeline, and in human capacity, have to be in place. Otherwise, a crisis will be wasted. As an example, the rape in Delhi on 16 December 2012 created a political moment for reform of the criminal justice system. However, as the early stages of the pipeline had not been constructed ahead of time, and human capabilities for stages 5, 6 and 7 were weak, the response that was obtained in that terrible moment was weak."
"A lot more is possible from the slow process of improving things every day, than is commonly given credit for. The UK got to liberal democracy without anything like a French Revolution. We should not be entranced by dramatic wins; we should have the endurance to engage in ‘the slow boring of hard boards’ over long decades."
"In many fields, it is useful to think of state activity under a few pillars of intervention. A government can produce education services by running schools. It can regulate the working of private schools. It can finance private persons buying the services of private schools. These are three pillars of intervention—producing, regulating, financing."
"Each of the three pillars involves different kinds of management mechanisms. Positive externalities justify financing. Negative externalities, asymmetric information or market power justify regulating."
"The Indian state should not just be a mini-Swedish state. There would be many zeroes in the Indian optimization: We would try to go after a smaller set of market failures, where the welfare cost of the free market outcome is particularly large, and where the state capacity required in addressing the market failure is relatively small. Given our capacity constraints today, to quote Kaushik Basu, we have to engage in ‘libertarianism of necessity’."
"We would argue in favour of the largest possible state that is feasible while achieving competence, and this is likely to be a very small state."
"When state capacity is high, a government can embark on more complex plans. When state capacity is low, the only things that will work correctly are simple plans."
"In the international experience, waging war was an important pathway to developing state capacity. That pathway is not open to India, given the nuclear deterrent."
"Public choice theory and political science emphasize the dangers of concentration of power. Concentrated power will generally be abused for personal gains by the persons wielding the power. This encourages us to disperse power. Mature democracies work by limiting the power of every player in public life. As an example, the United States president has a narrow (‘enumerated’) list of powers under the US Constitution. Unlike the Indian arrangement, the US president does not even control the annual budget, nor does he control the draft legislation that is tabled for discussion by legislators."
"We in India have traditionally been over-optimistic about a benevolent state, and have skimped on the checks and balances that create accountability and performance. We need to take public choice theory to heart, and thereby bring sound procedures into law where agencies and personnel are mistrusted with power, so as to produce better performance by the state apparatus."
"consider the law that creates UIDAI. It has to contain two parts. The first part involves coercing private persons, to say that if they wish to be part of a government welfare programme, they have no choice but to submit to intrusive biometric requirements. The second part involves coercing the officials in UIDAI, and all public and private persons who utilize the UIDAI data, to engage in fair play towards the people. The need for this second part flows from public choice theory, the scepticism about the benevolence of the state."