What drives your success? - Evidence of the role of privilege

An earlier post titled, ‘The delusion of Merit-Hard Work-Ability’ argued that privilege plays a significant role in one’s success and that one shouldn’t forget it. It also argued that privilege comes in many forms.
This post presents evidence of privilege as a factor in success. In other words, it presents evidence of factors other than ‘ability’ that play a significant role in one’s success. It is to reiterate the fact that we should be less obsessed with the role of our hard work and ability and acknowledge the influence of other factors.
The detailed evidence on the impact of environment that one grows in, on one’s future earnings comes from the ‘Equality of Opportunity’ project. It is led by Prof. Raj Chetty of Stanford University (was earlier at Harvard) and his colleagues. They use huge administrative data of US tax records (4 crore people) and do a granular analysis of the ‘social mobility’ phenomenon.

Debunking Bhagwati and Panagariya’s ‘Debunking Kerala’s myth’

Arvind Panagariya and Jagdish Bhagwati (P&B from now on) have written extensively disputing post-independence governments’ role in Kerala’s higher metrics on education and health; at the same time celebrating the success of Gujarat. It forms a chapter in their book ‘India’s Tryst with Destiny’. The arguments regarding education are summarized in Panagariya’s 2012 column titled ‘Cracking Kerala’s myth’ in TOI.
P&B dispute Kerala’s success on two fronts
  1. While Kerala may be better on absolute metrics, the ‘change in literacy’ since independence is not impressive compared to India and especially Gujarat.
  2. Success in Kerala is not a state-led programme. It is due to initiatives led by private.
Prof. Ram Kumar of TISS has already written a response to P’s TOI column. I build on that adding more information.
Claim 1Kerala’s progress (change in % literacy) isn’t impressive compared to Gujarat and India’s.
P’s argument: “In 1951, it (Kerala) had a literacy rate of 47% compared with 18% for India as a whole and 28% for Maharashtra, the closest rival among the large states. By 2011, these rates had risen to 94, 74 and 83%, respectively. The gains made, thus, equal 47, 56 and 55 percentage points for Kerala, India and Maharashtra, respectively.
Debunking P:
The loophole in P’s argument is evident — comparing change from a lower base to that from a higher base. As per P’s argument, Kerala should have achieved 102% and 103% literacy to be called it as successful as Maharastra and India respectively.
A point to note here — Kerala hit 90% literacy rate by 1991 — before the India’s proverbial growth era.

*** 

Claim 2Kerala’s success is not state led.
P’s argument:

1. As per ASER 2010, Kerala has 53% of enrollment in private schools. This demonstrates that the success is not state led. " Based on the 53% figure, B & P write in their book The conventional and dominant story of Kerala as a state-led success doesn’t stand up to careful empirical investigation
2. Kerala’s success is attributable to historic factors and NOT to post-independence efforts.
Debunking P’s argument


1. B & P probably missed out being ‘careful’ in their 'careful empirical investigation’.
The 53% private enrollment in Kerala is misleading. The proportion of private schools in Kerala is just 7%!!
Why’s the difference? The word ‘private’ has two different meanings. One is that of a school entirely funded and run by the private individuals. This is the way private is usually understood. There’sanotherr set of schools that are funded by government and whose teachers also come from government’s cadre (depending on rules of the state) but the day-to-day operations are managed by trusts. These are called ‘aided schools’. Mistakenly, sometimes these aided schools are also mixed under conventional definition of ‘private’, which is what P&B do.
Since P&B quote 2010 data from ASER, we can consider 2010 school data to look at this. In 2010, Kerala had 36% government schools, 57% aided schools — schools funded by government but run by trusts and 7% schools run by private.

It implies that state had a huge role to play in Kerala and that 53% students are NOT out of state’s net, as misleadingly proclaimed by P&B.
2. Regarding Kerala’s post-independence role, as Prof. Ram Kumar points out, the case of Malabar is instructive of Kerala’s success. Malabar had 31% literacy in 1951 (Kerala — 47% in 1951) and has 91% in 2011.
The fact that even a backward region of the state, Malabar, has caught up to the levels of state aggregates, illustrate the post-independence efforts of governments in Kerala.
Malabar’s improvement is 60% which is more than India’s and Maharastra's, even going as per P’s earlier reasoning.

3. Robin Jeffrey in the book "Politics, Women and Well Being How Kerala became a Model” documents that Kerala was spending 40% of its annual budget on education by 1980s. If this isn't state-led effort, then not sure what is!

***

The actual outcomes

Literacy rate is an indicative measure but not a good measure of learning outcomes. It's the reason Pratham conducts ASER despite having literacy data from the census
P&B quote 53% private enrollment figure from ASER 2010 report but as per the same 2010 ASER report -
  1. % age of students of class 3–5 who can do subtraction or more46.6% in Gujarat and 79.2% in Kerala.
  2. % age of students of class 3–5 who can read class 1 level text63% in Gujarat and 87% in Kerala.
… and so on.
Kerala is way ahead of Gujarat in learning outcomes as per ASER.

It’s a mystery as to why the actual learning outcome data were not considered in P & B’s analysis.  It can’t be because they didn’t know about ASER. They do use private enrollment data (which is hidden inside the report) from ASER 2010 but carefully miss the learning outcomes data (which is the headline of the report). 


***

Bottom line:

Kerala’s success in education had significant role of post-independence governments.
Kerala is way ahead of Gujarat in learning outcomes.
Growth and learning outcomes: Kerala hit 90% literacy by 1991 - before the proverbial growth era. Gujarat is far behind Kerala in learning outcomes despite higher growth rate.

Promising effects of Mid-Day Meal program

The mid-day meal program, often termed as populist is one of the widely criticized government programs. The reasons range from ‘MDM is a populist policy’, ‘leakages in MDM’ to ‘MDM wasting teachers’ time’.

Earlier, some studies have suggested that MDM didn’t have effect on attendance. However, two recent studies suggest promising effects of MDM on both nutrition and learning outcomes.

On Nutrition: Using a clever research design of considering students whose enrollment in schools differs by an year but are born in adjacent months (children born in Dec-2001 are enrolled in 2006–07 while those born in Jan, 2002 are enrolled in 2007–08), this study finds that
School meals at age five compensated entirely for malnourishment from droughts in early childhood
Midday meals had an effect both on the weight and the height of children who received these meals. In particular, the effect was largest for children whose nutrition had suffered as a result of droughts in early childhood.
On Learning Outcomes: Using ASER data set and exploiting the staggered roll out of the program, a new study finds that
exposure to midday meals for the five-year duration of primary school increases test scores by 0.17 standard deviations (18% relative to children with the less than a year of exposure)
It is important to note that these effects are after taking the gaps in the program implementation into account. If only we could administer some of the policies well!

Transcending policy prescriptions emanating from parochial world views

In school education, when an expert points something as a constraint and offers a solution along those lines, ask yourself these questions
  • What is the background of the person — what is his/her area of expertise?
  • What is the constraint being pointed out and what’s the solution proposed?
In most cases, we find that, there is convergence between answers to first and second questions. It means that the answers could be in form of pairs such as
  • the person has expertise in pedagogy
  • the person identifies lack of appropriate pedagogy as the constraint and offers solution along those lines.
In other words, the problem identified by the person is same as the area of expertise of the person. This raises fundamental questions.
  • Did the person identify the problem as pedagogy because he/she is expert in that area and sees scope of improvement there? OR
  • Did the person consider all the existing constraints for analysis and then identify pedagogy as the constraint?
Without knowing answer to these questions, we can’t ascertain whether the person’s response is emanating from a systematic analysis or if it is just a parochial world view.

Do laws corrupt people?

What’s the fundamental reason for corruption?

One argument could be regarding incentives, institutions, lack of enforcement etc. If this is identified as the cause then the action would be to restructure the incentives and institutions. Most economists and policy analysts hold this view.

The other argument is that people are inherently corrupt. The strategy for reform should thus be to make people ‘non-corrupt’ — ‘change begins at home’. Common public tend to hold this view. This may seem naive but surprisingly this view is held by many people. During the peaks of Anna movement, many argued that — we are corrupt as a society — unless we change mentally- laws can do little.

The second argument regarding mental attitudes seems very convincing because of the simplicity of the idea and the solutions that follow it. It’s also easily picked up by those who are often worried about the declining standards of morality in humanity.

What’s the truth? Is it due to incentive structures or is it due to inherent nature of a society? or a mixture of both. If it’s a mixture of both, then what’s the one most contributing? We can consider an alternate hypothesis that satisfies the concerns of both the groups.

The narrative that marries both the above hypothesis could be — Majority of humans are all the same. They respond to incentives. Corruption was initiated due to bad incentive structures and since it continued for long, it got ingrained in people. So, when you just look at snapshot of current context, it would look as if the problem is with inherent nature of people. But, if you look temporally, you would realize that it’s not the case.

The strategy for reform to tackle corruption should hence approach it from the end of tweaking incentive structures, instead of approaching it from the other end, instilling moral education to people. Once you restructure incentives, the ingrained tendency would also get weeded out with time.

This argument clearly rests on the presumption that ‘incentives let corrupt nature seep through and it gets hardwired over time’. Any evidence which can suggest this could prove the point. Typically, case studies are used to demonstrate this phenomenon — examples of where corruption reduced after instituting appropriate incentives are used to argue that people aren’t inherently corrupt and that corruption can be addressed by changing incentive structures without resorting to moral lessons. Though, that would serve a purpose, it might still not be convincing. Evidence from neuro-science that actually demonstrates such human nature would be more helpful. We now have the evidence on these lines.

In a new study, researchers have found that
dishonesty gradually increases with repetition and gradually escalates and can get hard wired
Researchers mapped brains of participants using functional MRI (fMRI) and noticed that
amygdala, a brain area intricately involved in emotional responses, was most active when people first lied for personal gain. The amygdala’s response to lying declined with every lie while the magnitude of the lies escalated.
Critically, the extent of reduced amygdala sensitivity to dishonesty on a present decision relative to the previous one predicts the magnitude of escalation of self-serving dishonesty on the next decision
In other words, when you first do a dishonest act, there’s an emotional response which signals you not to act so. As you repeat the dishonest acts, the response, which urges you against it, stops declining in intensity. In long run, one gets just blunt to any such action.

In summary -

Inherent dishonesty of the majority is NOT the root-cause of corruption. Such traits of dishonesty that are visible now are cultivated over time due to the prevalent incentive structures, instituted by laws. Hence, the reform should be approached from the perspective of restructuring the incentives. Moral lessons to people would help little.

So, do laws corrupt people?

Yes. They can. If they are forcing people necessitating them to break the laws by initiating the  chain of reducing sensitivity to dishonesty.

It's not just laws, it applies to institutions too. A huge section of people necessitated to pay bribes to get things done get immune to `corruption` and take it for granted. The new normal for acceptable actions goes down.

The delusion of merit — hard work — talent

We are in a competitive world today. Success in education requires enormous efforts, hard work and discipline.
Growing in such environment, it is often natural for people to think — ‘I deserve this because I worked hard.’
The corollary of this is the argument of meritocracy — everything should be by merit.
In such arguments, people forget the role of privilege in their success.
John Rawls's Theory of Justice gave a powerful argument against such arguments. In short, Rawls argued that there is moral arbitrariness in the initial conditions
  1. The fact that the generation that you are living in values your talent is a matter of luck. For instance, in ancient world of Sparta, bravery and courage were extremely rewarded. In today’s world, it wouldn’t earn as much as those who do mental work.
  2. The ability to work hard and the expectations that drive you to work hard are also a function of the surroundings that you live in, which is again a matter of luck.
  3. Not to mention the role of economic status which gives the freedom to work hard instead of pulling you down by making you worry about subsistence.
A new research from Indian context on these lines has affirmed the second factor — the role of one’s identity in shaping one’s expectations and hence the drive to work hard.
I conduct a field experiment in India to investigate whether the norms associated with one’s social identity affect one’s aspirations, beliefs, and eventual outcomes.
I exploit cues within the subjects’ existing environment to “prime” either caste or gender identity, thereby introducing exogenous variation in the salience of one’s social identity.
Randomly assigning whether a subject’s caste, gender or neither is primed, I first elicit long-run aspirations of adolescents (parents) for their (their child’s) future economic outcomes.
I find that girls have aspirations and beliefs that are biased downwards when gender is primed, while parents of high caste adolescents have higher aspirations and beliefs about income and educational attainment when caste is primed.
To investigate the effects of identity on real (learning) outcomes, I set up a learning camp at each school for ten weeks and elicit their aspirations and beliefs for each of the tests conducted at the camp, also recording attendance and test scores.
Girls’ aspirations and beliefs are significantly lower when gender is primed, and learning outcomes are worse. Priming caste makes males from upper castes state higher aspirations and beliefs.
These findings suggest that even in the presence of equal opportunity, one’s social identity may have direct, psychological effects on individuals’ beliefs, aspirations, and eventual life.
This is in line with earlier research
  1. Beamen et al. find that presence of a female leader in a village (India) also led to a decrease in the gender gap in adolescent aspirations, changed parents’ perception, altered education attainment and useIt highlights the role of identity in shaping aspirations and expectations and the existent differences.
  2. Anirudh Krishna of Duke university conducted interviews of first generation IT workers in big MNCs of India to trace their trajectory — how did they end up in IT sector coming from a rural background. The common factor among majority of them was mentorship. At some point of time in each of their lives, they had the good fortune of getting access to information or guidance from somebody — often a teacher or parents’ employer — who shared information about a particular scholarship or opportunity. Such one off instances changed their trajectories. It highlights the importance of the role of guidance — which some have easy access to, which some are fortunate to get by luck and which some are unfortunate not to have.
This is not to suggest that merit shouldn’t be considered. The argument is that the role of privilege should also be acknowledged and that privilege comes in many forms.
We often tend to forget the role of privilege in our success and give excessive weightage to our efforts, hardwork.
Sometimes the only thing standing in between you and your success is the result of a God’s (if exists) coin’s toss — do you take birth in an affluent family of a privileged class (caste/race) in a liberal democratic society as a fully abled privileged gender OR do you take birth in a war wrecked country in an ultra poor family of an underprivileged class as differently abled underprivileged gender.

Ground Report: How does CCE function in government schools?

Disclaimer: Narrative below is description of teachers. Not mine.
Continuous and Comprehensive Evaluation in simpler terms mean that students are to be continuously tested over a period of time rather than at the end of year. The idea is to use the data from continuous evaluations as a feedback loop to remedy children’s difficulties.
As part of this, children are given regular home assignments, often involving small projects. The textbooks are also modified with the questions at the back of chapters being more dynamic in nature instead of asking to reiterate some lines mentioned in the chapter.
How does it work out in some government schools?
Initially this was the process.
  1. Teacher gives assignment
  1. Child finds it difficult to do the assignment because of mixture of reasons
  1. a) child can’t understand

  2. b) child doesn’t have resources like access to internet to gather material and money to buy materials to make projects etc.
3. Teachers worried that children are not doing assignments.
4. Children needn’t care because they aren’t going to fail anyways (no detention).
5. Teachers concerned about the workload due to increased paper work.
As it happens, the self correcting system, homeostasis, as Sheldon Cooper would call it, found a way it.
This is the way it works out now.

  1. Teacher gives assignment.

Modification: Teachers stopped giving assignments requiring use of internet and buying of materials.
2. Students submit the assignment.
Modification:


Entrepreneurial publishing companies have exploited the opportunity to prepare guidebooks — essentially a collection of model answers to all the assignments that students are supposed to do, as per the textbooks.

Students just copy text from guidebooks into notebooks and submit assignment. (Note that projects are no longer given to students).
3. Teacher corrects it. Obviously, it's a lot easier. Everyone has written same answer, from same guidebook.
Children happy. Teachers happy.
But teachers have some complaints
1. Net effect is negative. At least in earlier case, student could repeat memorized stuff. Now, even that’s not a possibility.

Earlier students used to at least memorize answers to the questions given at the end of the chapter. Now, neither they memorize nor do assignments properly.
Add no detention policy to that. There’s no reason for child to do anything.
2. Unnecessary extra burden which is of no use.


The elite class' condescension of middle class dreams

Ridiculing a school kid’s and their parents’ aspiration to get their child into IITs has become a common topic of ridicule these days.
The narrative goes as “Hey look! Crazy fellows. Getting anxious about IIT admission years before the admission process.”
Occasionally, this turns into a sermon on how the parents should leave kids alone to explore their interests.
Some of them also take pride in saying “We sent our kids to liberal arts school while this crazy world is running after engineering and medicine.”
This narrative displays nothing but the condescension of elite class and lack of appreciation of the reality of people who take such decisions.
Consider middle class families and those below. They live life on a brink with tight budgets saving every possible rupee. Their life is a life of uncertainty. Faced with these decisions daily, it is natural for them to aspire for a good life to their children where they don’t have to live the risky middle class life.
The most certain way of making this possible is by getting into a good engineering college (IIT) or a medical college (AIIMS). There might be other ways but these ways provide a certainty that once a student gets in to these colleges, some amount of security is assured.
Though people with networks may deny but the reality of the world is thatbrand matters. May be elite class can get internships to their children through connections, can make up for their child’s lack of good undergraduate education by spending on their foreign degrees, can get their children interview calls in companies, or can transfer their family businesses to their children in case everything falls apart. The middle class don’t have such options. The only resource they have is hard work and they would want to put that to appropriate use.
Hence, middle class parents strive to get their children into these colleges that grant them security and access to a better world. They would want to do everything possible to make it possible and not take any chance. In this hyper competitive world, if others’ kids are preparing from 6th standard, and if your kid isn’t preparing, then probably your kid is getting behind.
All of these make parents anxious about the process and make them scout for best possible schools which can help them make this possible.
Terming this anxiety as nonsense or ridiculing it is both insensitive and shows lack of appreciation of reality
The best thing middle class children can do for themselves is to get into a good undergraduate college first and then think of other things in life. The undergraduate college provides the necessary safety net which the elite class gets by birth. Probably the presence of the easily attained safety net is the reason for condescension of elite class. They can take off this invisible security net once and see how it feels.
Is this preferable for a society as a whole? May be not, but the economic compulsions of current age make this inevitable. May be, once a critical mass of people climb above the economic ladder and reach a stage where they don’t have to worry about ensuring basic necessities of life, then the default option can be to explore interests. Until then, people will act the way they are acting now.

I would in fact go a step ahead and argue (to be provocative) that everyone should opt for extra coaching in school provided in the name of IIT foundation. At least on this pretext, students get exposed to something new so that they aren’t spoiled by the routine curriculum that emphasizes rote learning and blunts children’s mind.

Intervention scale-ups as a torch to highlight state capacity constraints

Gulzar Natarajan is skeptical about intervention scale ups. He has a point especially when he says
interpretation of most positive results are complicated by very low baseline and overall marginal (yet statistically significant) effects.
The important point however is that —
to what extent can the interventions tested on a small scale be successful when they are scaled up and overlaid on structures of weak state capacity?
I have a contrarian take on this. I argue that
if done properly, scaling up interventions can enhance state capacity.
How?
Consider this argument: What if implementing interventions, which are sound in the ‘technical-know-how’, results in enhancing state capacity?
The state capacity constraints at lower level bureaucracy in education are
  • absence of support structures (CRCs and BRCs that act as academic support to teachers)
  • incoherence in monitoring norms 
  • lack of proper duty allocation norms and so on.

In the business as usual scenario, there is nothing at stake for the bureaucratic structure to disrupt the front line setup.
Now, the fact that one has to implement TaRL makes the bureaucracy feel the necessity to adjust some of the working norms and structures, thus enhancing the state capacity.
For instance, consider the case of Andhra Pradesh. Before TaRL was taken up for scale up last year, there was no proper structure of CRCs and most reporting activity of monitoring was regarding administrative aspects.
With the compulsion to implement TaRL, government felt the difficulties in the present set up. It setup a new cadre of CRCs, reorganized the roles and functions of some of the profiles in education bureaucracy etc. This to me, contributes to enhancing state capacity.
In a nutshell, the current state setup is like a foundation, on which the intervention is being laid out. If the process of overlaying the intervention on this weak foundation, highlights the weaknesses of the foundation and results towards correction, such scale-ups are a potential tool to enhance state capacity. They act as a torch light to highlight the constrictions.
Understandably, on the contrary, the danger is that if the government doesn’t recognize the weaknesses of the foundation and tries to gloss them over and force fits the intervention, it is bound to fail resulting in loss of money, energy, resources, time and contributes to the growing skepticism that government schools can’t be reformed!