Building a FAQ on Affirmative Action
#25
(12-Jan-2013, 10:45 PM)Kanad Kanhere Wrote: The case of "rich lower castes" is definitely not that acute but not something that should be discarded. Cultura capital is still an issue for the rich lower castes and hence they should clearly get some affirmative action as opposed to say rich upper castes.

Then there IS an affirmative action based on income for higher education. Below a certain income, education is subsidized. The only point then is what should be the distribution of resource between affirmative action based on income vs affirmative action based on caste. Certainly we can't discard the latter altogether.

Indeed. There needs to be a quantitative argument on the distribution of resources between economic and caste-based affirmative action. But the parameters for the actual tradeoff are difficult, if not impossible, to measure:

1. The incremental benefit to India as a whole (as opposed to particular families) of some X% reservations, or of Y% subsidized education.

2. The incremental cost to India as a whole of X% reserved seats.

The answers, as a function of X, will probably not be linear, otherwise a likely best strategy would be to allocate all seats to backward classes.

(When I say "benefit to India", I don't mean just in terms of money. If we strongly desire seeing the wealth gap between upper and lower castes close, then that should be accounted for in the "benefit" function.)

(12-Jan-2013, 10:45 PM)Kanad Kanhere Wrote: It might be cost effective to kill the poor to develop a country. But should that be ever done. (Apologies for the extreme example). The point is the argument about efficiency/cost is frequently invoked without due consideration to the fundamental principles of Humanism. Humans and social justice comes before growth.

There are many points I would like to make here:

1. It's my fault for not making it clear that I'm using "benefit" and "cost" from a humanist point of view rather than merely an economic point of view. So bringing self-respect and social power to dalits is indeed a benefit.

2. Economics cannot be separated from humanism. For example, income inequality is bad from a humanist point of view. However, I would happily trade off 2-3 GINI points for doubling per-capita income.

3. There are trade-offs in humanism as well. Consider your extreme example of killing poor people. Well, to some approximation, that is exactly what happens when a country goes to war. Yet war is sometimes necessary, and despite the nationalistic rhetoric used, it's ultimately about a humanist trade-off. (Though I agree that war could certainly also be better optimized to cause the least harm possible to participants.)

So I think you're setting up a bit of a false dichotomy with your statement. We need growth for humanism, because social justice is just one component of humanism. Let me flip that statement and say that the argument for social justice is often invoked without due consideration for the fundamental principles of economics. (Another example is the rule of law. Good police and judicial process is strongly correlated with per-capita GDP, and while correlation does not imply causation, it's not hard to imagine a causal link between more funding and better courts, something that India desperately needs.)

But of course I'm not suggesting some sort of universal growth/justice dichotomy here, though that is sometimes how capitalism plays out.

(12-Jan-2013, 10:45 PM)Kanad Kanhere Wrote: I would like to approach this question in another way. Is there any other system that could have achieved the betterment of the lower castes people (that have been benefited by reservation) in any other way? The point 3 should be considered only if we have competing systems. Otherwise, as you have said so yourself, something is most certainly better than nothing.

I would like to ask a question in all seriousness: if it so self-evident that reservations provide an irreplacable service for the lower caste, then why don't we have 100% reservations? This is not snark, if that's the best way forward then we should not hesitate to say so.


(12-Jan-2013, 06:57 PM)Captain Mandrake Wrote: What is this "bang for the buck" you talk about? Does the bang mean the same for people of various castes?

See my clarifications above on humanistic costs and benefits. Ultimately, the Indian govt made educational institutes not as a favor to the upper castes but because Indian needed educated people. To some extent, if the people passing out of these institutes are worse at their jobs, then many sectors of society are going to suffer. That's what I mean by "bang for the buck".

There are various complicating issues, to be sure -- this is only a probabilistic argument. I'm not trying to imply that every forward-caste student will go on to be professionally better than every backward-caste student. Many students, of all sections, will waste their education entirely, or will settle in other countries, or will go into finance. :/


(12-Jan-2013, 06:46 PM)Captain Mandrake Wrote: Sure a cursory look will reveal a few extreme examples of rich lower caste people and poor upper caste people. But we need more rigorous evidence. What percentage of people of various castes are below the poverty lines? What are the institutional barriers to upward mobility? Do these institutional barriers work the same way for people of different castes? I suspect that as you answer these questions you will start seeing the rationale behind affirmative action.

I meant a cursory glance at the statistics, not a cursory glance at people.
Reply
#26
(13-Jan-2013, 03:24 AM)anukool_j Wrote: Indeed. There needs to be a quantitative argument on the distribution of resources between economic and caste-based affirmative action. But the parameters for the actual tradeoff are difficult, if not impossible, to measure:

1. The incremental benefit to India as a whole (as opposed to particular families) of some X% reservations, or of Y% subsidized education.

2. The incremental cost to India as a whole of X% reserved seats.

The answers, as a function of X, will probably not be linear, otherwise a likely best strategy would be to allocate all seats to backward classes.

(When I say "benefit to India", I don't mean just in terms of money. If we strongly desire seeing the wealth gap between upper and lower castes close, then that should be accounted for in the "benefit" function.)

(13-Jan-2013, 03:24 AM)anukool_j Wrote: I would like to ask a question in all seriousness: if it so self-evident that reservations provide an irreplacable service for the lower caste, then why don't we have 100% reservations? This is not snark, if that's the best way forward then we should not hesitate to say so.

There are two important points to note
1. Presence of reservation system is better for lower castes than its absence. The benefit might not be quantifiable, but its there. (I am placing this argument from comparison of histories). It is in this sense that I said that unless until we have a better competitive system, reservation ought to stay.

2. The question about 100% allocation is what I always call the "dilemma of comet model". I will given an analogous example. Should we conduct olympics when there are so many hungry people? These are difficult queries to address because of conflicting values.

What most people will agree, in these difficult cases, is a reasonable middle approach, where the growth of non-reserved category is not stopped till reserved category catches up, but growth of reserved category is given a push, may be at the cost of slightly decreased growth of non-reserved category. And thats what I think we have in place right now in reservation system.

(13-Jan-2013, 03:24 AM)anukool_j Wrote: So I think you're setting up a bit of a false dichotomy with your statement. We need growth for humanism, because social justice is just one component of humanism. Let me flip that statement and say that the argument for social justice is often invoked without due consideration for the fundamental principles of economics. (Another example is the rule of law. Good police and judicial process is strongly correlated with per-capita GDP, and while correlation does not imply causation, it's not hard to imagine a causal link between more funding and better courts, something that India desperately needs.)

I didn't intend to assert a dichotomy. I wanted to put a perspective that some Humanistic principles cannot be compromised even at the cost of lost efficiency. Example we need to improve caste disparity even if it might cost us in terms of growth.
Reply
#27
(13-Jan-2013, 05:55 AM)Kanad Kanhere Wrote: 1. Presence of reservation system is better for lower castes than its absence. The benefit might not be quantifiable, but its there. (I am placing this argument from comparison of histories). It is in this sense that I said that unless until we have a better competitive system, reservation ought to stay.

I agree with the premise, but I think it only implies that we should stick to reservation till its benefits outweigh its costs. The presence of absence of other systems matters only insofar as it affects this tradeoff.

(13-Jan-2013, 05:55 AM)Kanad Kanhere Wrote: 2. The question about 100% allocation is what I always call the "dilemma of comet model". I will given an analogous example. Should we conduct olympics when there are so many hungry people? These are difficult queries to address because of conflicting values.

Well yes, that's a good analogy and another example that we should seriously consider. If put in stark terms like "India has to make a choice between lifting 5% of its population out of poverty and winning ~5 olympic medals each meet" the debate will be much more productive, and everyone will be able to clearly see the tradeoffs made.

Just because we have a middle-ground status quo does not mean that it is at all reasonable.

(13-Jan-2013, 05:55 AM)Kanad Kanhere Wrote: I didn't intend to assert a dichotomy. I wanted to put a perspective that some Humanistic principles cannot be compromised even at the cost of lost efficiency. Example we need to improve caste disparity even if it might cost us in terms of growth.

Personally, I would feel uncomfortable making that statement unless I knew approximately how much disparity we were reducing at the cost of how much growth. Like I said before, a 2-3 point drop in GINI seems acceptable if it results in doubling the per-capita GDP. Do you disagree?

Anyway, thank you for your time and thoughts. I'll update this thread if/when I have some quantitative data about the effect of reservations in India. I agree that without such data, India's current system seems not-too-bad, though it seems likely that the overall detrimental effect on the well-being of Indians is being underestimated.

If I may put some words in your mouth, we both agree that reservations are better than nothing, but we disagree on how much better. We also disagree on how much they cost the country as a whole. ("Better" and "cost" being used in a humanist sense.) I think we're both familiar with the general arguments on each side, so the only thing that should make us change our minds is data.

-Anukool.
[+] 1 user Likes anukool_j's post
Reply
#28
(13-Jan-2013, 07:13 AM)anukool_j Wrote: If I may put some words in your mouth, we both agree that reservations are better than nothing, but we disagree on how much better. We also disagree on how much they cost the country as a whole. ("Better" and "cost" being used in a humanist sense.) I think we're both familiar with the general arguments on each side, so the only thing that should make us change our minds is data.

-Anukool.

Agree on all counts.

Just one point that I would like to emphasize is that intuitively I can't think of any system that can maximize some humanistic principles without correcting the social disparity. So I think caste based affirmative action is a must.
Reply
#29
(13-Jan-2013, 03:24 AM)anukool_j Wrote: To some extent, if the people passing out of these institutes are worse at their jobs, then many sectors of society are going to suffer.

Sure. But that is only true if people graduating out of the institutes are worse at their jobs. That is why affirmative actions applies only at entry. Once you get into these institutions everyone has to clear the same bar to graduate. So your concern about not being good at the job is already being addressed. This is one of the biggest misconceptions among the opponents (not saying you are an opponent) of affirmative action. They somehow think that a doctor from a lower caste background who benefited from reservation is inferior to one from upper caste background. This is patently false.

But then as Kanad says in the post below, even if initiatives to reduce inequality impedes growth it might be worth it.

Quote:Example we need to improve caste disparity even if it might cost us in terms of growth.

But I am not convinced the reducing inequality through reservations actually impedes growth.
Reply
#30
(14-Jan-2013, 04:59 AM)Captain Mandrake Wrote: But that is only true if people graduating out of the institutes are worse at their jobs. That is why affirmative actions applies only at entry. Once you get into these institutions everyone has to clear the same bar to graduate.

Your conclusions do not follow from your premises. If people really didn't care about margins, no one would care about your GPA (as long as you passed) or your score in the boards, or the number of papers you published, or how many of them you were a first author on. In jobs/business things are a bit more vague, but relative grading within a batch still exists and is important.

And the evidence is that a disproportionately large number of reserved category students struggle through their courses and scrape through rather than excelling. So I think it's a reasonable conclusion that probabilistically, they would be worse at their jobs. (Or whatever related field they apply themselves to later.)

To put it in other words, it matter whether you clear that bar by an inch or a mile.
Reply
#31
(14-Jan-2013, 07:09 AM)anukool_j Wrote:
(14-Jan-2013, 04:59 AM)Captain Mandrake Wrote: But that is only true if people graduating out of the institutes are worse at their jobs. That is why affirmative actions applies only at entry. Once you get into these institutions everyone has to clear the same bar to graduate.

Your conclusions do not follow from your premises. If people really didn't care about margins, no one would care about your GPA (as long as you passed) or your score in the boards, or the number of papers you published, or how many of them you were a first author on. In jobs/business things are a bit more vague, but relative grading within a batch still exists and is important.

And the evidence is that a disproportionately large number of reserved category students struggle through their courses and scrape through rather than excelling. So I think it's a reasonable conclusion that probabilistically, they would be worse at their jobs. (Or whatever related field they apply themselves to later.)

To put it in other words, it matter whether you clear that bar by an inch or a mile.

I agree with Captain Mandrake, if I understood her/him correctly. It ofcourse matters whether one covers the bar by an inch or a mile, but what we are talking about are these three categories: Bad, Not so good, good, for the kind of a person's skill at his job is. Notice that its only category 1 that is harmful, the middle is harmful only in the sense of "growth" but not really in terms of degeneration of society (probably strong wording). So again as long as we are not creating the 1st category employees/employers it becomes a tradeoff between growth and social disparity.
Reply
#32
(14-Jan-2013, 10:53 PM)Kanad Kanhere Wrote: but what we are talking about are these three categories: Bad, Not so good, good, for the kind of a person's skill at his job is. Notice that its only category 1 that is harmful, the middle is harmful only in the sense of "growth" but not really in terms of degeneration of society (probably strong wording). So again as long as we are not creating the 1st category employees/employers it becomes a tradeoff between growth and social disparity.

Sure, but my point is that India needs to grow its GDP by about 15x just to be able to afford things like good nutrition and education, even assuming that inequality goes down somewhat in the process. While it is an unfortunate tradeoff to have to make, I don't consider the pursuit of growth in this kind of situation as peripheral as you seem to. It's quite urgent that India generate more wealth.

I already even quantified this. In the current scenario, I would be ok with a 3 point increase in GINI if it doubled GDP.
Reply
#33
(15-Jan-2013, 01:46 AM)anukool_j Wrote: Sure, but my point is that India needs to grow its GDP by about 15x just to be able to afford things like good nutrition and education, even assuming that inequality goes down somewhat in the process. While it is an unfortunate tradeoff to have to make, I don't consider the pursuit of growth in this kind of situation as peripheral as you seem to. It's quite urgent that India generate more wealth.

I already even quantified this. In the current scenario, I would be ok with a 3 point increase in GINI if it doubled GDP.

Several participants in this panel discussion on the now routine GDP vs GNH/Gini debate present arguments about why growth is far from peripheral and is an obvious pre-requisite to keep welfare spending sustainable.

Are we a nation of misplaced priorities? (NDTV Big Fight Jan 12 2013)

However, by what percentage GDP must grow to give a green signal to additional welfare spending, is a question on which no consensus is available. The cost of welfare-spending, says P Sainath, often lies well within the bounds of affordability and needn't wait for an explosion in GDP, as the compelling example at the very beginning of this talk demonstrates. It maybe worthwhile to seek figures on specific questions like "How many more IITs can the government institute at the current rate of GDP growth without growing the deficit?" The moot question, one of political will rather than statistics,may however be, "Are we willing to privilege welfare-spending over other forms of government expenditure to the extent that we will take spending cuts even if GDP stays static?"
[+] 2 users Like arvindiyer's post
Reply
#34
(14-Jan-2013, 10:53 PM)Kanad Kanhere Wrote: I agree with Captain Mandrake, if I understood her/him correctly. It ofcourse matters whether one covers the bar by an inch or a mile, but what we are talking about are these three categories: Bad, Not so good, good, for the kind of a person's skill at his job is. Notice that its only category 1 that is harmful, the middle is harmful only in the sense of "growth" but not really in terms of degeneration of society (probably strong wording). So again as long as we are not creating the 1st category employees/employers it becomes a tradeoff between growth and social disparity.

Actually I only model it with two categories; good and bad. The bad is weeded out by the educational institute with their standards for graduation. Only the good get out with a degree and contribute to society.

You are using a more detailed model by splitting the good into two more categories; exceptional and adequate.
Reply
#35
Recent posts in this thread (post 21 onwards) have had a topical focus on cost-benefit analyses of common implementations of affirmative action, with participants agreeing upon the need for affirmative action in principle and carefully distinguishing this from possible objections towards its manifestation in practice. However, such agreement on the fundamental premises of affirmative action is far from shared among Indians at large, among whom not just 'quota-skeptics' but 'affirmative-action-skeptics' can be found in abundance.

Underlying 'affirmative-action-skepticism' can be reasons as diverse as (i) bigoted adherence to the tenets of Varna Dharma or other notions of ethnic/racial superiority(which is called out for what it is in the protracted comment trail of this article) which regards equal opportunity itself as a natural impossibility, (ii) dogmatic belief in the 'invisible hand of the free market', which treats attempts towards equal opportunity as undue interference doomed to fail, and (iii) the approach of 'learned helplessness' where equal opportunity is recognized as desirable but too difficult or expensive. Given the absence of a society-wide consensus on the premises of affirmative action, it is worthwhile, even pressing, to forge one where equal opportunity is treated as enough of a priority to justify 'costs' that maybe incurred due to affirmative action. Written to contribute to this end, was this note from August 2011 archived on request in full in the post below.
[+] 1 user Likes arvindiyer's post
Reply
#36
On affirmative action

A positive sum game : There's a reason why constitutional measures to build a fairer more equal-opportunity society are referred to as affirmative action. The emphasis is on affirming the rights of the aggrieved than on curtailing or negating the rights of those identified as the supposed aggressors. Mistaking the affirming of the rights of the aggrieved as automatically equal to a negating of the rights of the rest, results from assuming a 'zero-sum game' view. This is not a zero-sum game as it is not without real value addition. There is a real payoff for affirmative action in terms of more enriching diversity in educational and democratic institutions and the emergence of a society less prone to the threats that structural inequality poses.

Speak of the 'payoff' of affirmative action, it did not mean that (i) we can get something for nothing, nor that (ii)the promised payoff is the only reason for promoting such policies.

The devil in the details : The payoff in terms of more diversity and harmony, is essentially a long-term payoff which requires substantial 'down payment', which opponents of such policies lament about as 'a price paid in terms of efficiency'. Granting for the sake of argument that there is indeed a price paid in terms of efficiency, we can say that it is wise for society to incur this cost than incur any further trust deficits, and that this is a worthy investment, and insurance, for society. The question of until when is such investment dictated by necessity, and when it can be retired, seems rather premature and inopportune to discuss at this juncture. Yes, discussions on how best this investment maybe designed are justified and even crucial. Whatever is said here is with the intent of making a case for affirmative action in principle, and it goes without case that there is no exemption of specific implementations of affirmative action in practice, from scrutiny on a case-to-case basis.

The price and the prized : Use of words like 'investment' and 'down-payment' may seem calculative and even condescending, but rest assured they are used only for purposes of illustration. However calculative this narrative of 'investment' may sound, one thing in its favour is that the resulting discourse will be less divisive than say, when a narrative of 'reparations' is resorted to. Critics lamenting 'the price paid in terms of efficiency' would do well to be reminded that according primacy to 'efficiency of an economy', even at the cost of individual sacrifices, itself is an unstated value premise and by no means the only one. According primacy to 'equal opportunity' is another candidate value premise no more deserving of dismissal at face value than the prizing of 'efficiency of the economy'. If an efficiency-seeking system chooses to bear some inequality risks, why can't an equal-opportunity-building system choose to bear some supposed inefficiency risks?

In sum, when equal opportunity is agreed upon by means of a social contract as a shared value premise, then it becomes an end in its own right and not just a means to an end of a safer society. When thus treated as an end in itself, 'incentives', 'disincentives' and any other consequentialist considerations greatly diminish in importance as objections.

Who's 'we' in 'we the people'? : While some may have an instinctive revulsion towards supposed compromises on efficiency in the name of empowerment, these arguments have enjoyed considerable support in the public sphere in other contexts. In debates on cheap outsourcing and cheap immigrant labor in the West, the claim that perpetuation of these policies is economically efficient seems to be a no-brainer. That such efficiency cannot be achieved at the cost of 'social capital' and other national priorities is however a commonly stated position. Such positions are often met with a plaint of 'protectionism', but societies are at liberty, even if not under the obligation, to 'put people first' rather than putting profit first. Some of the opposition to affirmative action stems from the fact that the people who are put first in these policies are not recognized by the critics as 'people like us' (PLUs). The unstated privilege that leads one to assume that 'people' means 'people-like-us' and that welfare means what is immediately beneficial to our group, is what needs to examined as a pre-requisite to any meaningful participation in an affirmative action or for that matter, any social justice debate.

- 26th August 2011

***
Footnotes and further work:
1) Need data on the 'cost' of affirmative action programs in terms of budgetary allocations, extra effort in terms of remedial classes etc. to justify the 'investment' in the same
2) Need a comprehensive survey of different 'designs' of affirmative action in terms of state-sponsored vs voluntary programs and also in terms of the criteria for receiving benefits
3) Need sociological data on different yardsticks of 'merit' in different societies and cultural differences about the same, to know what a society prizes and what price it is willing to pay to attain the same
[+] 4 users Like arvindiyer's post
Reply


Possibly Related Threads...
Thread Author Replies Views Last Post
  [split] moderation open thread(On affirmative action and not being sexist) Piyush Mishra 29 14,603 09-Sep-2012, 09:14 PM
Last Post: Kanad Kanhere



Users browsing this thread: 1 Guest(s)