What do top US economists think of school vouchers?

I came across this interesting survey of economists, conducted 5 years back. Thanks to Prof. Dynarski's NYT column. Prof. Dynarski works on Charter Schools. Her research shows that Charter Schools in urban Massachusets have been beneficial and hence she is a passionate supporter of Charter Schools.

Coming to the survey, it's an initiative of University of Chicago. It asked some of the top US economists to vote on a set of public policies. All of this is in American context. The idea is to get a sense of economists' thinking on important public policy issues. The list of economists includes the likes of Daren Acemogulu (Why Nations Fail), Raj Chetty, Richard Thaler (Nudge), Angus Deaton (Nobel Laureate), Michael Greenstone, and so on. 

One must note that this survey is not a survey of "education economists". It is a survey of economists on a wide range of issues. It means that the economists didn't necessarily vote only on their subject of expertise. So, you have environment experts sharing their opinion on school choice, health economists sharing their opinion on school choice and so on.

One of the questions in the survey was to vote on the following statement on school vouchers.
Public school students would receive a higher quality education if they all had the option of taking the government money (local, state, federal) currently being spent on their own education and turning that money into vouchers that they could use towards covering the costs of any private school or public school of their choice (e.g. charter schools).
The respondents (economists) have to select one of the options (Strongly Agree - Agree - Uncertain - Disagree - Strongly Disagree - No opinion). It also asked to rate the confidence of their voting. It means that if I select "strongly agree" and I choose the confidence of my decision as 10, it means that I am absolutely sure of my opinion. It also asked to put in comments if any.

The result is that 36% of the top economists choose either Agree/Strongly Agree, supporting vouchers. If it is adjusted for the confidence of their choice, it goes up to 41%. Check this for complete results.

It looks like there is a considerable degree of skepticism about markets in economics (vouchers) among top economists. In contrast, 90%+ of them agree on markets for cab-sharing.

The good thing about the survey is that it also gave the option to explain the reasoning behind their voting. My reading of these comments is that majority of them (who chose both agree and disagree) seem to think that vouchers are good "on an average" but are concerned that it leaves behind some students, who might then find it difficult.

Some of the responses below. 

Economists who voted "Strongly Agree" or "Agree".

Judith Chevalier: The kids whose parents don't pay attention or are poorly informed can be worse off. Otherwise agree.

Aaron Edlie:  On the plus side, incentives are better under vouchers. On the negative side, decision making might be in the hands of those with less info.

Claudia Goldin: Many public school students would benefit but some with little choice might not. On net it would be beneficial and increase competition.

Nancy Stokey: It's the only way to break unions.

Richard Thaler: Hurray for charter schools but to go to full vouchers it is necessary to deal with possible unraveling if no schools want the bottom kids.

Anil Kashyap: Hard to know what the equilibrium will be, but so many kids are trapped now eventually most of them would have better choices.

Economists who voted either Disagree/Strongly Disagree or Uncertain

Pinelopi Goldberg: Treating students as clients does not improve education. Evidence on voucher programs very mixed - no robust evidence of positive effects.

Janet Curie: More motivated and able students would take advantage of the vouchers, but the students left behind would likely be worse off.

Daren Acemogulu: Vouchers likely to improve things in short run given the awful state of US public schools. But we know little about their long run effects.

Angus Deaton: And what about the kids that don't take up the vouchers?

Richard Schmalensee: Many would of course benefit, but those in rural areas or with irresponsible parents wouldn't. Charters aren't magic.

David Autor: Maddeningly sweeping! Some students would benefit and the average effect might indeed be positive. But some students would surely be harmed.

Katherine Baicker: Those using vouchers would likely be better off, but others might be worse - need to consider system-level and distributional effects.

These are only some of the responses. Please check the results page for the complete list of comments. There are also other questions in the survey. You may want to check responses to those questions too.

Moral bankruptcy of demonetization defenders

We easily forget history. Once in a while, it's useful to rehash lessons from history.

Once upon a time, slavery existed on earth and for a good amount of time. It did so because it was carefully sustained by building a narrative supporting it. Michael Munger of Duke University has a good podcast on Slavery and Racism. He summarizes the narrative justifying slavery as follows
 since I (master) owned them (slaves) I have a much better reason to take care of them because they are still going to be valuable to me 5 years from now. Whereas if I rent labor, I don’t care: that guy can die; I’m just going to pay him just enough to induce him to come work for me. I don’t have to provide housing; no health care. Whereas with a slave, if he hurts his leg, I’m going to take care of him.
In other words, the narrative was that slaves can't take care of themselves, hence it's better for them to be slaves. They would at least be taken care by their master.

Any attempt to argue against slavery was countered with such proclaimed economic benefits of the institution of slavery.

The thing about arguing only in terms of economic benefits is that one can defend anything using it - dowry, disenfranchising illiterates and women and so on. The proclaimed benefits can be easily listed down for anything that has sustained for long. It's easy to fall into the fallacy of "something that's going on without resulting in huge resistance (like riots) has to be good because if it wasn't good, it wouldn't have sustained". One only needs a handful of sycophant economists to come up with a narrative handcrafting the benefits. The benefits will be amplified and costs (suffering) are ignored.

By doing so, the defenders carefully substitute the normative moral reasoning for proclaimed economic benefits. These economic benefits never factor in the non-quantifiable moral issues in the calculation.

The difficult part is that it's easy to build a narrative that persuades people into turning off their normative lens. As Michael Munger confesses in the podcast
If you read letters — and this is something that I found so difficult in a way — and in fact, I have nightmares and when I was working on the paper and looking closely at letters would wake up at night and feel bad about this, because I was developing some sense of sympathy for people who objectively I think are terrible. They are slave owners.
If one is not the victim and is removed from the situation, it's not easy to perceive the human suffering. So, it's easy to buy into the narrative of benefits, disregarding the suffering. For instance, consider the narrative: "illiterates and women don't know have a knowledge of the world. Giving them the power to vote will be detrimental to democracy". It appeals to our superficial sense of reasoning. Looked at it superficially, such argument based on benefits looks very convincing. It's the power of narratives that turns off the normative lens.

Now, the same thing is happening with demonetization. The enormous human suffering is being ignored because it is apparently worth the proclaimed benefits. While sycophant economists are busy manufacturing narrative of proclaimed benefits, it's becoming easy to fall for it, especially those who don't have to suffer because of the luxury of being in places where they can sustain without cash. There are several issues with this.

One, normatively speaking, the whole exercise violates one of the fundamental principles of our society/law that it's not ok to punish innocents, even if it's at the expense of not punishing the wrong-doers. A large number of people who have no role in the black money making are being put to enormous trouble and death in some cases, for the wrong-doings of a few. It's literally a carpet bombing.

Two, demonetization is as beneficial as slavery. Sure, there may be benefits somewhere, but at what cost and suffering? If a drug to a disease results in hair growth on a bald head, it's not the reason to pray for that disease and deliberately spread it.

Three, much of the proclaimed benefits needn't even be true. They are just post-facto justifications. If demonetization is such a transformative and groundbreaking reform, one's to wonder, why no one who's vehemently supporting it now, have argued for it in the past?  Some economists, who are part of the current sycophant group, even have written books in the past on Indian economy. Demonetization proposal doesn't even get a cursory mention in those books. All this post-facto justification is nothing but sycophancy, deliberately ignoring the large-scale human suffering. Some even term the human suffering as noise, which is reflective of the nature of justifications.

How different are the defenders of demonetization from those who defended slavery? By trying to come up with such dubious justifications of economic benefits, the demonetization defenders are acting like defenders of slavery, by substituting normative moral reasoning with dubious, manufactured 'economic benefits'.

It's not just the case of demonetization. The logic of over-emphasizing proclaimed benefits over normative issues is leading to serious perils in other aspects also. It was recently seen in US elections. Some prominent figures advocated for Trump because he can get things done, ignoring the normative issues. India isn't an exception, where the practice of electoral victory vindicating all the past wrong doings of the leaders is deeply instituted.

Such disregard to normative reasoning should be seriously resisted. If only proclaimed economic benefits (that don't factor in unquantifiable aspects like loss of life) is the metric, let's draw a line in the population to demarcate net generators and net consumers, and kill all the net consumers. It may certainly bring economic benefits.

This can't be the way forward. It's time to re-up our normative lenses, else we will slide down to the days of justifying acts like slavery, with our lens of dubious economic benefits.

Does development increase/decrease emigration?

Our intuitive understanding is that people emigrate from poor and developing countries to developed countries, in search of better opportunities. So, once these countries develop, emigration slows down. It's a reasonable hypothesis, except that there's a small twist to it.

Emigration actually increases in short term, as the country develops, till a threshold income level. It decreases only after that.

Michael Clemens constructed these graphs from emigration data. Emigration follows an inverted-U pattern. Emigration increases till the GDP per capita is about $6000-$8000 (PPP), after which it decreases.

There can be several reasons behind the inverted U pattern of emigration, where emigration initially increases with development. 
  1. Developed countries issue more visas as countries grow
  2. Financial constraints to emigrate ease, as countries grow
  3. People can access more information, as countries grow..
.... and so on.

A new paper explores the micro behaviour of emigration. It finds that
Between one-third and one-half of the slope of the increasing segment is due to the changing skill composition of working-age populations, and another third is due to changing network size. The microeconomic channel (including financial incentives and constraints) only accounts for one fourth of the total effect in low-income countries, and for less than one fifth in lower-middle-income countries.
In other words, skill acquisition, along with network effects are the major reasons for the initial increase of emigration, as countries grow.

Countries might have to accept the inevitability of this emigration pattern. One must also note that this emigration of high-skilled workers need not necessarily harm the country. It might infact be beneficial as the evidence suggests. Instead, obstructing emigration might turn out to be harmful.

PS: Regarding India: India's GDP per capita is $6,266 (PPP) in 2016. If India were to follow the same pattern, we should see a decrease in emigration.

Class vs. Caste - Does equalizing class equalize life outcomes?

It is a well-known fact that there's an inequality of life outcomes in India. However, there's considerable debate on the mechanisms of inequalities. Four factors prominently feature in the discourse - class (economic status), caste, gender, and rural vs. urban. Disaggregating impacts of these individual factors is problematic because there is an overlap - the median income of higher castes tends to be higher, maybe they also tend to be urban, and so on.

In an important paper, Ashwini Deshpande of Delhi School of Economics disentangles the impacts of class and caste. In other words, it answers the question - if class conditions are equalized (economic status is equalized), does caste still play a role in life outcomes? It's only for rural areas.

Excerpts from her interview below
even when a lot of the class conditions are equalised, caste seems to have an independent effect on future life outcomes. The point at which the differences are really stark is performance in the school leaving examination (Class 12). From then on the differences just keep multiplying.
Disaggregating the results further by the individual caste (Brahmins vs. Dalits), she says
even when I compare poor Brahmins with poor Dalits, and middle class Brahmins with middle class Dalits etc. there are differences that are pretty stark. I have information on a number of caste categories, but the most significant and persistent difference is between the Brahmins at the top end of the upper castes, and the Dalits at the bottommost end of the caste hierarchy. The other differences are sometimes significant but it’s these two ends of the caste hierarchy where the differences seem to persist even when class conditions are equalised. Most national or macro data sets do not allow us to break up the upper castes to Brahmins and other upper castes, so this nuance is not highlighted.
The other important insight from the paper is that many don't avail reservation due to difficulty in procuring certificates, contrary to a section of people who argue that it's because people feel stigmatized to avail reservation.
I found that of those who do not use quotas ever, only 17% fear stigmatisation. The bulk of non-use comes from procedural difficulties that are created by the administrators and institutions who are reluctant to give them these places in the first place.
It may not be a conclusive evidence but it hints in the direction that caste handicaps are too strong to be dislocated by just bridging class differences.

We find similar evidence in competitive exams too. UPSC Civil Service examination is a good proxy to test the role of privileges because the cost of appearing for the exam (coaching, opportunity costs) and the role of other factors (support structures, family's financial health etc.) is high in this exam.

In UPSC Civil Service examination, 85%-90% of top 100 rankers are from General Category, 8-10% from OBC, 2-3% from SC and occasionally 1% ST. It holds true even if the rank range is extended to 500. It's astonishing because 85%-90% of the rankers are from a caste group that constitutes only 30% of the population. Probably, it also reflects the role of caste privileges, as found in Prof. Deshpande's paper. 

The birth factors seem to be playing too much role than it's presumed. One must also observe that if 'pure meritocracy' was followed, meaning that if there was no reservation, the caste composition of the civil servant recruits would have been highly skewed as above, reflecting more of the inequalities at birth and not necessarily pure merit. Equalizing class, as is argued in proposals to replace caste reservation with class reservation, doesn't seem to be enough to bridge birth inequalities, as the above evidence suggests. 

Caste handicap is too hard to ignore in the light of this evidence. All is not well, as the urban elite seems to assume.

More on China and poverty reduction

One impressive statistic of China is that its poverty declined from 53% in 1981 to a mere 8% in 2001; a 45% decline in just two decades. In comparison, India's poverty declined from 45% in 1991 to 27% in 2011, a 28% decline in two decades. By any standards, Chinese rate of poverty reduction is astonishing.

What's the Chinese story then?

Interestingly, much of the Chinese poverty reduction came within first half decade of 80s. By 1987, poverty fell down to one-third of its level at 1981, i.e. 17%. Two-thirds of the gains were achieved within initial 5 years.

Martin Ravallion has a paper on Chinese growth story. It's a 10-year-old paper. I am not sure of recent developments. This paper has four key insights.

On characteristics of Chinese poverty reduction

1. Agriculture has contributed to poverty reduction more than other sectors.
While migration to urban areas has helped reduce poverty nationally, the bulk of the reduction in poverty came from rural areas. Growth in the primary sector (primarily agriculture) did more to reduce poverty and inequality than either the secondary or tertiary sectors. Starting in 1981, if the same aggregate growth rate had been balanced across sectors, it would have taken 10 years to bring the poverty rate down to 8%, rather than 20 years.
2. Poverty reduction is slower in places with high inequality.
historical evidence suggests that more unequal provinces will face a double handicap in future poverty reduction; they will have lower growth and poverty will respond less to that growth.
On mechanisms of growth

3. De-collectivisation of agriculture is a major factor behind significant poverty reduction during 1981-1987. It is complemented by government's agricultural produce sourcing policy, setting income floors, macroeconomic stability and public investments.
The early 1980s saw high growth in primary sector output and rapid rural poverty reduction in the wake of de-collectivization and the privatization of land-use rights under the household responsibility systemQ. (Agricultural land had previously been farmed by organized brigades, in which all members shared the output more-or-less equally.) Since this was a one-off event across the whole country, we cannot test its explanatory power. However, the literature has pointed to the importance of these reforms in stimulating rural economic growth at the early stages of China’s transition (Fan, 1991; Lin, 1992; Chow, 2002). And (as we have seen) rural economic growth was key to falling poverty in that period. It would not be unreasonable to presume that the agrarian reforms around 1980 accounted for the bulk of rural poverty reduction in the first half of the 1980s, which (as we have also seen) accounted for roughly three-quarters of the total decline in the national poverty rate over 1981–2001.
4. Trade reform had little impact on poverty and inequality.

The final point is counterintuitive since much of the discourse around China has been that it reduced poverty due to its strong export-driven policies, driven by strong manufacturing.

Pranab Bardhan draws lessons for India from Chinese experience in a paper around the same time (2007).

1. Growth elasticity of poverty is lower in India, compared to China.
The Indian pace of poverty reduction has been less than China’s, not just because growth has been faster in China but also because the same 1 per cent growth rate reduces (or is associated with reduction in) poverty in India by much less. The so-called growth elasticity of poverty reduction is much higher in China than in India; this may have something to do with the differential inequalities in wealth in the two countries (particularly, land and education).
2. Majority gains in China are from de-collectivisation of agriculture. It's not a policy translatable to India. One key insight from China that's useful to India is that poverty's response to growth depends on initial inequalities. India fares poorly on that front compared to China. Bardhan points out

The Gini coefficient of land distribution in rural India was 0.74 in 2003; the corresponding figure in China was 0.49 in 2002.
India’s educational inequality is one of the worst in the worldaccording to a table in the World Development Report 2006, published by the World Bank, the Gini coefficient of the distribution of adult schooling years in the population, a crude measure of educational inequality, was 0.56 in India in 1998-2000, which is not just higher than 0.37 in China in 2000 but even higher than almost all Latin American countries (Brazil: 0.39)

Interestingly, the high growth states in India have lower growth elasticity of poverty (lower decrease in poverty for 1% increase in mean incomes). Bardhan points out
growth elasticity of poverty reduction depends on initial distribution of land and human capital. Purfield (2006) indicates that in the period 1977-2001, this elasticity was quite low in high growth states like Maharashtra and Karnataka, and high in states like Kerala and West Bengal.
3. Bardhan also points out globalization slowed down poverty reduction rate in certain areas.

One key lesson for India from this experience seems to be that we need to factor in initial levels of inequality, for a faster poverty reduction. In this context, maybe it's the time to act on a long pending government's promise to provide land to landless Dalits and STs, to start with, to accelerate the poverty reduction.

Importance of mechanisms of growth - Poverty reduction potential of growth

Esther Duflo makes this insightful observation in an interview
The biggest problem with just focusing on growth is that we have no idea what causes growth. And this does not hold just for me, but a whole lot of macroeconomists who have studied numerous determinants of economic growth. So focusing on just growth can be a bit useless. Another issue which has become important vis-à-vis growth is its failure to guarantee upward social mobility for future generations. 
It's a subtle point but often missed in the discourse. Growth-Poverty relation is also relevant in this context.  Economists call it the 'growth elasticity of poverty', the percentage decrease in poverty for 1 percent increase in mean per capita incomes.

An old column of Maitreesh Ghatak summarizes this argument clearly.
While it is true that in the post-liberalization era growth has indeed lifted millions out of poverty, it is also true that the extent to which growth has made a dent on poverty (growth elasticity of poverty being the technical term) has been lower in India than in China and other comparable countries
In other words, the number of people that were pulled out of poverty by 1% growth, is lower in India, as compared to China. Ghatak argues that it's mainly because of lower human capital in India.
This is mainly because of the shockingly low levels of human capital for a large chunk of the population. 
Consider this fact: the wage rate more than doubles if you move from low-skilled to medium-skilled jobs, or if you move from medium-skilled to high skilled jobs. If the child of an unskilled worker becomes highly skilled, then individual income will increase four-times within one generation. Yet, as a recent report by the Pew Research Center points out, between 2001 and 2011 the share of Indians who would be considered middle income increased from 1% to 3%, whereas in China it went up from 3% to 18%. While poverty has fallen, instead of an emerging middle class, what India has experienced is a larger class of low income earners.  
In this context, it is saddening that there are no serious efforts to revamp education and health care in the country, barring exceptions of Delhi and partially Bihar. As someone remarked "the struggle to make governments focus on education and healthcare has come from 'pushing them to do something productive' to damage control - to stop them from destroying what's remaining.

Regressive step by CBSE: Meddling with appointment of principals in private schools

CBSE has brought new rules regarding the appointment of principals in CBSE affiliated private unaided schools. As per this new rule, i) principals have to pass a test (even the existing principals have to write the test); ii) committee to recruit principal to a school should include a CBSE nominee and a state government nominee, who will have a veto power; iii) government schools are exempted from this.

This is a regressive step on many counts in an already over-regulated private school sector.

One, tests in India mean nothing. They are poorly designed. In fact, it might not be even possible to measure the qualities required for a principal, through a written test. It's not a recruitment for a match referee where one is testing candidates on the knowledge of intricate details of the game's rules.

It's useful to ask the following questions. There are 10,000 CBSE unaided private schools in India as of today. There are many good and outstanding principals among them. How did all of them become good principals without passing a test? How were they running the schools till now?

Instead, requirements of passing a test, reduces the available pool of candidates for recruitment, making it difficult for private schools. 

Two, the involvement of CBSE and state governments on panel interferes with the autonomy of the schools. It's not clear as to, on what grounds does CBSE think that it (CBSE and state governments) can choose better principals, than the school management? 

Worse, these nominees have a veto power. The new rule might not do anything apart from creating new avenues for corruption.

Three, continuing the flawed principle of exemption to government schools from RTE infrastructure requirements, this rule also exempts government schools from the above requirements. 

Even if one assumes that above rules make sense and are beneficial to schools, does that mean that government schools shouldn't be beneficiaries of such rule?

Governments don't make any significant efforts to improve the quality of public schools. Instead, they leave no opportunity to drag down the well-functioning private schools.

2016's Economic Survey talks of regulatory cholesterol. In private school sector, it's not even cholesterol, it's a deliberate insertion of a leg piece in the throat, obstructing the windpipe, choking the person to death.

Update: Few days after the above decision, government retracted the rule of mandating exams for principals.

Seasonal Hunger: Cash Transfers to assist migration

Mushfiq Mubarak has a good study (RCT) on seasonal hunger. Villagers face a tough time between sowing of seeds and harvesting the crop. They don't have source of income during this time. They are forced to cut down food consumption and other expenditures.

Mubarak et al. provided a small cash transfer ($8.5) to assist temporary migration to cities. This is a small amount to cover bus ticket and food for a couple of days. Such small amount had huge effects.
22% households sent a seasonal migrant;
A 30 to 35 percent increase in food and nonfood expenditures for the families of the individuals who accepted the incentive and migrated;
Consumption of 550 to 700 more calories a day per person, equivalent to an extra meal per person daily, for those who took advantage of the travel subsidy;
As much as a 19 percent increase in household income during the lean season for those who took the incentive
Recurring migration as households that received incentives one time became more likely to send someone to work during the lean season in subsequent years, even though incentives were not offered in those years.
 This had spill over effects too.
  1. When large number of people migrate, it reduces population in village and hence increased their wages.
  2. It increased migration even among the non participants. It's because people migrate in groups to share costs.
 If the cost of migration is small ($8.5) with huge benefits, why didn't households migrate before? The authors say that
migration is risky, mitigating risk requires individual-specific learning, and some migrants are sufficiently close to subsistence that failed migration is very costly.

Stop saying "correlation is not causation" unnecessarily

Lately, the term "correlation is not causation" is being thrown around loosely. This is in the tradition of casting aspersions on evidence that one disagrees with. Throwing this phrase only puts the burden on the opponent while one can just sit and do nit picking. We have reached a stage where "correlation can never mean causation."

Correlation gives a good hypothesis to explore. It should no means be discarded because it is just a correlation. 

Correlation can be causation. If one doesn't want to believe a causation because of the presence of confounding factors, one needs to present those confounding factors.

Often, there's a misconception that causation can only be proven by RCTs. It's again incorrect. Causation can be proven if one can demonstrate an appropriate mechanism through which the correlation occurs.

For instance, if you quote the correlation between ice cream sales and the number of books published in a year, it may not be attention worthy. It's because there is no possible mechanism which explains this mechanism. On the other hand, if there's a correlation between ice cream sales and the number of ENT cases, it's attention worthy. It's because one can clearly see a mechanism through which this can possibly happen. If the mechanism through which ice cream causes throat disease, which results in ENT cases can be proven, it’s as good as an RCT.

For a good understanding of this technique, check out Raj Chetty's work in social mobility in US. He starts with the data that some areas have high social mobility and some have less mobility. There is a possible confounding factor here - may be people of certain kind live in a place which results in higher mobility, and it has nothing to do with the place. The way to test this is to check those people, who have changed cities and see how their mobility varies. If a person shifting from low mobility area to high mobility area ends up having higher mobility, it can be inferred that mobility is due to place, not people. Further, if the extent of mobility is proportional to the number of years of exposure to high mobility areas, it's a further evidence. All of this proves causality. No RCT here.

It also illustrates the importance of theory. Theory gives us the mechanisms through which the effect occurs.

Ashish Jha does a good job reminding these arguments in the context of his recent study.

On needing RCT to prove causation
Remember the RCT that assigned people to smoking (versus not) to see if it really caused lung cancer?  Me neither…because it never happened.  So, if you are a strict “correlation is not causation” person who thinks observational data only create hypotheses that need to be tested using RCTs, you should only feel comfortable stating that smoking is associated with lung cancer but it’s only a hypothesis for which we await an RCT.  That’s silly.  Smoking causes lung cancer.
On alternative explanations - confounding factors
There must be an alternative explanation! There must be confounding!  But the critics have mostly failed to come up with what a plausible confounder could be.  Remember, a variable, in order to be a confounder, must be correlated both with the predictor (gender) and outcome (mortality).  We spent over a year working on this paper, trying to think of confounders that might explain our findings.  Every time we came up with something, we tried to account for it in our models.  No, our models aren’t perfect. Of course, there could still be confounders that we missed. We are imperfect researchers. But that confounder would have to be big enough to explain about a half a percentage point mortality difference, and that’s not trivial.  So I ask the critics to help us identify this missing confounder that explains better outcomes for women physicians
The other important aspect here is the "possible threat of effects". If there's a correlation between smoking and lung cancer, it is better to err on the opposite side and take precautions assuming that it causes cancer. The stakes are high to wait till the causation through RCT is proved.

The same applies to climate change too. By the time Montreal Summit was going on in 1987, there was no indisputable evidence that CFCs are damaging the ozone layer. One had only preliminary evidence. The indisputable evidence came only years later. Countries exercised wisdom in identifying the damage and taking precautions well before hand.

In the context of growing relevance to evidence, unfortunately, there is a trend to misuse the data interpretations. The common form is to just point fingers at opponent's evidence without taking any responsibility for the burden of proof. Ashish's article is a gentle reminder to catch such arguments.

Correlation is a double edged sword. It can affect either ways. One needs to exercise caution and diligence in interpreting it.

Hayek and Demonetization

A twitter user made an insightful observation on demonetization and Hayek's famous essay titled, 'The use of knowledge in societies'. 

The point is that demonetization is being called "good idea but bad implementation". The argument is that demonetization is not just bad implementation, it is a bad idea too. Hayek's analogy helps here.

In his essay, The use of knowledge in societies, Hayek made an insightful observation on the role of prices and the infeasibility of planned economy. The argument is simple. 

Many think that economics is about the appropriate allocation of resources. Hayek says that allocation of resources is a mathematical problem and is easy to solve. The problem, however, is about our "knowledge of resources". The data of resources isn't stored in a central database from which data can be downloaded and be slapped with algorithms to spit out appropriate allocation. Knowledge is dispersed throughout the economy. If one may use Foucault's concept of power, knowledge is spread the way blood flows in the body through capillaries. There's no precise way to capture by a single person.

Prices come into play here. Price is a system that aggregates the micro knowledge and gives it a concrete shape. Hence, they should be relied up on.  Centrally administered planning. fails because of its inability, rather infeasibility to gather such knowledge.

The second complexity arises due to the dynamic nature of knowledge. One can theoretically argue that given enough time, one can gather knowledge. However, the trouble is that knowledge is not static. It's changing every second. So, even the best possible person can't detect it real time.

Now, apply it to demonetization. Cash in the economy is also like knowledge, spread throughout the economy. The argument that demonetization is a good idea but bad implementation misses the same point that many economists miss, as Hayek argued above. It assumes that government had all the information but failed to execute. The point is that it's not possible to have all such information. No matter how well it was planned, it was bound to fail because the information is just not enough.

Demonetization is thus as bad an idea as a planned economy that determines prices centrally.

One may make theoretical arguments against this line of reasoning but it's a fresh way to look at things.

Importance of diversity in profiles of social science researchers

At this time of the year (December), World Bank's Development blog posts summaries of job market papers of fresh Ph.D graduates in economics. It gives a sneak peek into the latest in the field of development economics. One of the striking features of the research papers seems to be that all women related papers are written by female Ph.D students. Maybe male Ph.D students researching on women issues exist there exist but if the WB series is a proxy for top papers in empirical development economics, it doesn't seem to be so.

Why is that only (majority) female graduate students are researching on gender issues? It highlights the role of 'choice of questions for research'.

One's choice of questions is partly rooted in one's own experiences, beliefs and so on. Amartya Sen and Albert Hirschman are good examples. Sen's experience of Bengal's famine made him research on famines and guided much of his work. In recent times, work of Prof. Roland Fryer of Harvard University, on race and education illustrates this. From his experience, Prof. Fryer had an intuition that students of colour feel it uncool to excel academically. That intuition was behind his major work on race and education.

All of these illustrate the importance of diversity and the value it brings to academia. The fallout of lack of diversity in profiles of social science researchers is that certain themes go out of radar.

In the aftermath of Trump's victory, one of the after-shock reflections was that, why couldn't sociologists predict this? Responding to that question, a sociology professor pointed out that 'sociology discipline' is losing its diversity in the profile of candidates in its doctoral candidates. In the absence of such diversity, there's a narrow focus only on certain themes and some important themes are getting missed out.

On a broader note, can readily identify four reasons behind the phenomenon of under-researched themes or certain themes going out of radar.

1. Selection and Competition: In recent times, the top Ph.D programmes in social sciences have become extremely competitive. Getting into these universities needs recommendations from top notch professors and undergraduate degrees from top universities. It's well known that privilege plays a huge role in securing undergraduate admission in the top universities. It just gets carried into selection of Ph.D programmes too.

Essentially, you end up having a pool of candidates researching on issues of development who might have never experienced those issues of poverty, gender, hunger, oppression etc. It doesn't mean that one needs to experience these issues to be able to research them. It's just that it's also good to have people who have experienced them because they can bring new insights.

2. Obsession with mathematical rigour: Research has two elements, rigor and choice of the questions. Everyone in academia can be rigorous but if they are missing out asking certain questions due to lack of perspective, it reduces our understanding of the world. Ph.D programmes are actively assisting the process of narrowing down of research by placing disproportionate emphasis on rigour, neglecting the choice of questions.

Apparently, the famous behavioural economist Sendhil Mullainadhan was advised by his Ph.D adviser to include some mathematical equations in his Ph.D thesis. It's because Sendhil's adviser thought that people don't take it seriously, if papers don't have mathematical equations. Those were the days when behavioural economics was less known and was struggling to gain prominence.

3. Compartmentalisation of  knowledge: There's a tendency to look at 'economics', 'sociology', 'political science', 'anthropology', 'psychology separately. Such compartmentalisation narrows the perspective.

In an interview, Tyler Cowen asks Dani Rodrik (appreciating Dani Rodrik's ability to bring new perspectives),  "what should be done to Ph.D programmes to make more Dani Rodriks?". Rodrik says that economics schools these days are recruiting mathematicians and not economists and that a training and knowledge of other disciplines helps in understanding the world. In fact, Dani Rodrik argues that most of his ideas came from his interaction with non-economics friends and his early training in political science.

4. Method over question: There's is a recent trend of arguments over methods - quantitative vs. qualitative, RCT vs. Non RCT. People are first deciding to do RCTs and then they choose a question to do the RCT. As Angus Deaton says, the method is taking prominence over the question. When the method is pre-fixed, the range of questions that one can explore with it gets narrowed.

In summary, research is not just about rigour, it's also about the questions that are researched. Questions asked by researchers depend on their personal experiences and biases. It means that if the pool of researchers aren't diverse, the questions asked may also be narrow, leaving out crucial information about the world. It's time to think of making research diversified.

What's common on to Electricity consumption - Voting behaviour - Car Offers - Policy Advocacy?

Long before, I worked on an RCT evaluating an intervention aiming to reduce electricity consumption. The intervention was simple. Every week, households would receive letters with data about their consumption previous week and the average consumption of their households in the previous week. The hypothesis was that everyone wants to save electricity but households may not know if their consumption is appropriate or they are overusing. If people are given a benchmark (neighbour's consumption) to check if they are using appropriately, they would adjust their consumption.

Since letters had to be sent every week, we needed a location where we can collect data digitally. It happened to be a large apartment complex. It also happened that the apartments had a generator for backup. There were two prices for electricity. One price for the "usual electricity". The price of generator electricity was higher.

The results showed that people significantly reduced their consumption. Telling people that they are over-consuming as compared to their neighbours, worked. The interesting aspect is that the electricity consumption reduced only for "usual electricity". It didn't change for "diesel generated electricity". It happened because diesel electricity was high. So, people already made a conscious effort of saving it. Letters didn't help as they were already saving diesel electricity. Letters helped saving usual electricity, as they weren't sure of "whether they were using appropriate amount".

In discussion with my boss, I had then mentioned that this pattern can be linked to voting behaviour. The model is same as above. In the above case, there is a product (generator), about which people are sure. So, external influence didn't matter. It mattered only when people were unsure, usual electricity. Similarly, there are two types of voters: strong supporters and others. Strong supporters would vote for a party irrespective of the situation. So, they wouldn't be necessarily affected by campaigning. The other supporters are malleable. 

Two situations arise. When a party is extremely unpopular, then everyone has decided NOT to vote for it. In a two party system, opposition gains. In a not so clear context, a party can increase their votes by focusing on the malleable people.

Convincing the malleable people is a tough task. Local leaders come into play here. Since parties can't reach out to everyone, they target these local leaders, who act as a lever to convince others. I later developed a more comprehensive framework including other variables but this was the crux: convinced voters and neutral voters (fence sitters).

After few years, I was excited to know that Abhijit Banerjee and others are working on a similar area, called social networks. They were getting into the depth of this idea on the modalities of influencing people. Essentially, the spread of information through local leaders (nodes) depends on the "connectivity of the person" and the person's "credibility". Further, one doesn't need to focus on everyone in the community. There may exist an optimum of nodes for a given population size, through which information can be effectively disseminated to the rest. They are figuring out these things.

A bit on the influence of "credibility of person". During the "intolerance debate", some argued that there is actually no data supporting raise in intolerance in the country. They had used crime data for this analysis. I had then written in News Laundry that crime data is not a good proxy for measuring perception of intolerance for two reasons. One, crime is an extreme measure of intolerance. Crime data hence doesn't capture verbal and other forms. Two, the perception of intolerance is magnified when certain key leaders of the society engage in it and get away with it. It's similar to the social networks argument: The messenger of information matters.

Coming back to our central argument of malleable people and non-malleable people, some time later, I learned that Malcolm Gladwell discussed a similar idea in his book Tipping Point. 

Famous economist textbook author Hal Varian had long back worked on a similar model of consumer and price behaviour. His model was on informed consumers and uninformed consumers. His idea was based on offer prices of cars. His argument was that only uninformed consumers would be affected by such advertisements and so on.

Central to all these ideas is the presence of two sections of people: "malleable" and "non-malleable". Malleable people are swayed by others (neighbours etc). Again, not all neighbours are same. Credibility of messenger and connectivity etc. also matter.

Cass Sunstein has a new paper on people's support to policy. His finding is that if people are not well informed of a policy, being told that majority supports it, will increase their odds of supporting the policy. This is similar to our model. Informed people don't get influenced but uninformed people get swayed by information of neighbours.

Simple idea but wide application!

One big lesson from PISA for low performing countries

One consistent lesson from PISA results is that low-performance is not a destiny. Countries can improve their performance. Contrary to popular discourse to design and come up with fancy policies, the process of transformation requires boring grunt work. For instance, this is the story of Argentina's improvement
For Esteban Bullrich, the minister in charge of education from 2010 to 2015, the initial aim was to make sure that pupils were being taught. Teachers were spending 12-15 days per year on strike, or about 7% of  the time they should be in class, according to his calculations. To try to reduce those absences he first made his mobile-phone number public and began fielding calls directly from angry teachers. He extended the school day.
Then he offered teachers something of a deal: higher salaries in exchange for taking their job more seriously. The grip of unions in deciding promotions was loosened. And he made teacher training more rigorous and practical.
Argentina's efforts, as with most other countries that have improved, didn't include fancy grand schemes. It required them to do boring, grunt work of correcting the fundamentals of governance like addressing teachers' concerns, navigating through unions, bargaining for performance. It's a lesson that's to be ingrained. On the contrary, all our efforts to reform bypass these aspects. We instead try to parachute interventions to classrooms or launch grand schemes without addressing the first order issues. We have to strongly resist such approach.

This is for low performing countries suffering from capacity constraints. For countries that don't have  capacity constraints, the policies that are implemented using the capacity may be a constraint. They require a different approach to reform.

Traps to be avoided while interpreting PISA results

PISA 2015 results are out. Singapore has topped the list. Naturally, there is a discussion on top performing countries and lessons that can be learnt from them. While interpreting the PISA results and inferring lessons, one needs to avoid the following traps.

One, pressure cooker model vs. fermentation model. Education systems of countries can be categorised into two types - pressure cooker model and fermentation model.

In pressure cooker model, an artificial pressure is created on students to excel. This is typically through high stake school leaving exams. For instance, if a student who doesn't score above a certain level in 10th grade is ineligible to pursue university education, it creates an artificial pressure on students to excel. Students spend enormous time on tuition classes getting trained to solve 'questions'. Students in that environment have an edge in standardised tests that test them on 'questions'. Most top performing Asian countries fall in this category - Singapore, China, South Korea etc.

In fermentation model, the process of education is organic. Unlike the pressure cooker model, there is no artificial pressure to excel. The process is more through low-stake testing, motivating children to perform, letting students learn naturally without pressure. Scandinavian countries fall in this category - Finland, Estonia etc.

If you want to emulate, what approach would be suitable for your country or is it desirable? Certainly, creating artificial pressure is a brute force method but can some countries take that? 

A critic of pressure cooker systems said "Saying that Singapore has best model of education is equivalent to saying that Kota has best model of +2 education." 

Two, selection effects. In places like Singapore, Shanghai, students are put into different tracks of education based on their performance in mid-school, around 5th-6th grade. So, students who study 9th grade and take part in international assessments are those who are screened at earlier grades and hence are from the sample of good students. PISA says that it corrects for this sampling bias but there's a big debate around Shanghai.

Just a warning. Indians doesn't need to use this as alibi to justify their poor results. Even students of the best performing schools of India are below the PISA's international average.

Three, interpreting systems in form of a diagnostic model. One of the common interpretations of successful systems is that - good systems invest in teacher education, have a culture of improvement etc. It's important to not to fall into this trap of reasoning.

Everyone knows that imparting education requires good teachers, good school leadership etc. Professing it as a lesson is similar to saying that one needs to daily go to gym and eat healthy food to build a good body. Everyone knows the ingredients. The problem, as with the gym is the ability to daily do that.  In context of education, it translates to the ability to impart teacher training, which is the state capacity. 

If the lesson is interpreted as "teacher training", then the usual policy that flows from it is to start a new teacher training system. The problem however is that we have realized the importance of teacher training long back. We imparted numerous trainings till now. The problem is not with either lack of realization or non-existence of training. The problem is with doing training better, implementing it in a better manner. We lack the capacity to do that. We should instead focus on building that capacity, resisting the urge to launch new programmes.

The perils of weak state capacity

Inefficient implementation of policies is often thought of as the major consequence of weak state capacity. It’s only partially true. I argued in my book that there are other serious consequences of weak state capacity.

First, the straight forward consequence is that weak state capacity decreases the returns to governments’ investment. Governments make huge investments in resources. When they don’t function properly, we end up wasting a great deal of money. For instance, government may recruit 1 lakh teachers but if they do not attend schools or don’t deliver value, the money is essentially going down the drain. The additional 1 lakh teachers have not resulted in any value addition but increased fiscal burden. Lant Pritchett and Yamini Aiyar estimate that with the existing inefficiencies of public school system, it would need an extra 2.78% of our GDP, or Rs. 2,32,000 crore to be able to reach the learning levels of private schools in India.

Two, repeated failure due to state capacity deficiencies causes experimentation fatigue. It resists new innovations. Any new innovation will be viewed with suspicion and may not be taken seriously while implementing because we already know that it’s going to fail. Such prolonged situation of weak state capacity can kill the spirit of the system and the people.

Three, weak state capacity kills good ideas. If a good idea fails due to implementation bottlenecks, the failure is attributed to the idea. Continuous and Comprehensive Evaluation (CCE) is a good example. Continuous low stake diagnostic testing is a good idea. The implementation challenges are now posing aspersions on the idea of low stake diagnostic testing itself.

Four, weak state capacity shrinks the imagination of ‘what can be done’. If government repeatedly fails to implement policy ideas, overwhelmed by implementation deficiencies, many end up thinking that innovative models can never be implemented in our systems. Over time, this perception gets ingrained in our belief systems and can limit our imagination to dream big. We start thinking of solutions with the presumption of ‘we can’t afford complex ideas’, thus leaving many good ideas out of our radar.

For instance, Finland revamped its curriculum shifting from individual subject based learning to theme based learning. Instead of learning geography, history, economics separately, you would study Europe as a whole (theme based). In our context, where we can’t even teach letters to children, our cognition restricts our ability to think of such ambitious innovations.

Five, weak state capacity reduces policy options. Let’s take a problem, say, skill building. There are many policy options to skill building. Germany uses apprenticeship model. Implementing such model requires certain prerequisites. The World Bank report on skill building says “The strongest constraint in the transferability of such programmes is that they require a strong institutional framework, in particular a clear legal framework.

Now, if a country doesn’t have strong institutional capacity to set up legal frameworks, German model is out of that country’s pool of options. Thus, when faced with a policy problem, weak state capacity limits the pool of solutions from which we can adopt. It naturally affects the quality of our policy. Ideally, we should be able to learn from a big pool of ideas and make the best of all, without having to think of capacity challenges.

Six, weak state capacity leads to policy traps and forces us to settle for second-best solutions. This is at the core of many debates in development.

Regulation and certification is one of the core functions of modern state. Let’s take the example of haircut. There is no great danger if governments don’t certify barbers and allow non professionals to do the hair cuts.

Now, move on to school education. The debate gets stronger here. Some argue that teaching by non-professionals is going to have significant costs and hence teachers should be strongly regulated. By that reasoning, low cost private schools shouldn’t be allowed. Some others argue that low cost private schools should be allowed because they are preferred by the people. The absolutists however argue that crucial aspects like education shouldn’t be left to people alone. There should be strict standards.

Its a policy trap. We don’t have capacity to build strong public systems in short run. We thus have to settle for second best solution of allowing education service delivery through untrained professionals.

A person once rightly asked a professor - "You have well functioning public school system in your country. Why are you are coming to us and asking us to settle for low budget private schools?" Weak state capacity is at the root of this question.

Health care is an example on the extreme end. There may be some case to argue that not regulating teachers may not be harmful. It poses a difficult conundrum in case of doctors. The recent debate on Jishnu Das’s papers on health care in Madhya Pradesh documenting the effectiveness of quacks and public doctors, and the effect of training quacks in West Bengal is a good example.

Our intuitive reasoning suggests that only trained professionals should be allowed to practice medicine because the stakes are high. Such reasoning challenges the reality, where a large number of untrained doctors, called quacks, deliver health care service.

Some argue that quacks are preferred by patients and hence should be allowed. They argue that government doesn’t have the capacity to train large number of doctors who are available to people all the time and provide equivalent service. Hence, quacks should be made central to service delivery by training them. This is an example of settling for second best solutions in the absence of capacity to embrace first best solutions.

It evokes strong response from some who argue quacks shouldn’t be allowed to practice. They argue that health care delivery shouldn’t be subjected to such work around (jugaad) — second best solutions and that we should focus on building strong public systems.

Weak state capacity is at the root of these debates. In some cases we can afford to ignore some good solutions because we don’t have the capacity. In case of critical sectors like health care, it is difficult to argue to reject first best solutions due to lack of capacity and instead embrace second best solutions. We should instead build capacity and do whatever is required because some sectors are too important to be left for second-best solutions.

We must also note that till now, there haven’t been any strong efforts by governments to reform our public health care. It’s a case of lack of political will. So, if second best solutions are advocated, without even attempting to reform, it may not go down well. It may not be even desirable. The recent experience of Delhi government’s efforts to revamp public education and successful results rekindles hope that given sufficient attention, public systems can be reformed.

All the above discussion says that weak state capacity affects our cognition as much as it affects the policies. It strongly illustrates the need to enhance our state capacity. It’s time we stop thinking of solutions in terms of schemes or designing new policies and start focusing on first order fundamentals, the state capacity.

The complexity of 'minimum wage' effect on employment

Debates on minimum wage are one of the oldest. Proponents argue that every worker should get certain minimum wage. The rationale is rooted both in ethics and correcting the monopolistic markets. Opponents argue that setting minimum wage is an external disruption to the market equilibrium. The firms would shut down because they won't be able to pay the wages, decreasing employment. It would then hurt the very poor.

David Card, showed for the first time, in 1990s that minimum wage laws in US (for that level) didn't reduce employment. This was a revolutionary finding at that time. Subsequently, literature on minimum wage in developed countries supported this finding.

However, the effects of minimum wages in developing countries is not well researched. A recent paper on the effects of minimum wage in Honduras suggests that it did have negative effect on employment.

In Indian context, MNREGS is a kind of minimum wage law. It sets wage floor, providing bargaining power to workers. Employers have to essentially give more wage than MNREGS wages in order to recruit workers and retain them. The General Equilibrium effect of MNREGS show that it didn't distort employment.

The story in developing countries is thus complex. It gets more complex if one considers sub sectors. Consider the case of budget private schools in India. Teachers in these schools are paid around Rs.4,000 to Rs.5,000. It's the affordable wage for the employers, for the amount of fee they collect. Let's say that the minimum wage is fixed at Rs.15,000 and if it is strictly enforced, the schools have to either shut down or increase their fee by three times. Either way it would hurt the students.

It would be great if minimum wage research should also looks into sector wide effects (may be it's already done but I am not aware). The solutions aren't obvious to me though. Differential minimum wages as per sectors is an obvious idea but it's not practical. May be we should just leave it there thinking that lack of implementation of policies is some times a boon!

Influences on intelligence: Cultural and Biological

Tyler Cowen interviewed Joseph Henrich in his Talks with Cowen talk show.

Henrich says the following on the cultural influence on intelligence.
we have this tendency to think that our intelligence is about raw brain processing power. But I make the case that a lot of our intelligence comes from our culturally downloaded tools. 
A simple example is all of you have a numbering system that you can count without bound. But the smallest-scale human societies will count one, two, three, many. They can’t differentiate 36 from 37. You can see the full variation of numbering systems by looking at these body-part counting systems.
Some groups in New Guinea will have a body-part counting system that goes up to 27. Another one will have one that goes up to 17. Somewhere in cultural evolution, we developed this ability to count without bound. Once we do that, we actually get new cognitive ability. When you grow up with this new kind of system, you get abilities you didn’t have before.
Then there’s also a relative one. Between me and the door — I could say she’s to the left of the door. That’s by drawing a line between myself and the doorway, and then using that as a reference point.The same thing is true of our spatial cognitive abilities. In English, there’s three different spatial reference systems. There’s absolute: north, south, east, and west. There’s body-centered — left, right, front, back.
But in some languages, they just have north, south, east, and west. They can’t tell you to drive on the left or drive on the right because it’s not one of the systems that’s built in there. Once you have those, you can redeploy them in all kinds of fancy ways to do new stuff. 
Cultural evolution is making us smarter by giving us all these new cognitive tricks.
In summary, Henrich seems to argue that if intelligence is the ability to analyse, then culture provides us some cognitive tools, that helps us in deciphering the problems.

I am inclined to support this argument although I am not sure if the following interpretation is correct. One of the ways to get good at Maths, sport or any other activity is to get a database of enough examples ingrained in your brain. If faced with a new problem, these can be tweaked to solve the problem. Such practice helps us identify patterns. Vishwanathan Anand argues the same about Chess, where I think it's more applicable.

Henrich makes another important point, the distinction between biological and genetic differences.
People often conflate biological differences with genetic differences. Culture changes our biology even when it doesn’t change our genetics.
culture’s been shaping our genetic evolution, but it also shapes our biology. A simple example is everyone in this room — I would say with . . . I can’t be 100 percent sure — but you have a specialization in your left hemisphere, and you have a thicker corpus callosum than you would otherwise.
You’ve acquired a particular cultural skill, literacy, that changes your brain and makes you biologically different and actually thickens that information highway between your two hemispheres. When you hear spoken speech, you get greater full-brain activation patterns than you would if you’d still been illiterate. Culture changes our biology and causes us to think differently. 
A simple example is everyone in this room…[has] a specialization in your left hemisphere, and you have a thicker corpus callosum than you would otherwise.
You’ve acquired a particular cultural skill, literacy, that changes your brain and makes you biologically different and actually thickens that information highway between your two hemispheres. When you hear spoken speech, you get greater full-brain activation patterns than you would if you’d still been illiterate. Culture changes our biology and causes us to think differently.
Read or listen to the interview. Henrich shares good insights on relevance of psychology research based on western subjects, religion, society and so on.

India's largest scam was in its most liberalised sector. Why?

During the debates on Lokpal many argued that the antidote to corruption is competition. They cited telecom sector as an example. With competition, customers got quality services and corruption reduced. It was hence argued that liberalisation is the way to address corruption and instituting lokpal is an incorrect way to address corruption.

Ironically, India's largest scam, the 2G scam, happened in most liberalised sector, the telecom. Why is that so?

It calls for a more nuanced understanding of types of corruption. Corruption can be categorised into two types - collusive corruption and extortive corruption. Extortive corruption is the petty corruption at local level where officers demand bribe to issue certificates. Such corruption thrives under scarcity and monopoly. In collusive corruption, government officials collude with others to do favours.

The liberalisation of telecom sector, allowing private companies into the sector, reduced scarcity of phones and abolished monopoly of BSNL. It thus reduced corruption.

With time, the scope and nature of corruption has gotten complex. Earlier it was about issuing phones. Now, it is at higher policy making levels.

2G scam is an example of shift in nature of corruption from local level to policy making levels. While corruption at local levels could be addressed by abolishing monopolies, the monopoly of government's policy making role can't be abolished. One needs a different mechanism to address such problem.

Unlike petty corruption that's clearly visible, corruption at policy making levels is hard to recognise. It requires strong domain knowledge to detect any such wrong doings, making it difficult to expose. We need more preventive measures in such situations. Deterrence through strong, swift, certain punishment is a strong preventive measure to address corruption at higher levels. Lokpal is well suited to do this.

Without recognising this distinction in nature of corruption and strategies needed to address corruption of different kinds, Lokpal was unfairly blamed for not being suitable to address all types of corruption. Such narrative helped governments to discredit Lokpal and stall it.