 I like... Joseph, I'll also begin by saying that Saurabh, who's the organizer of this session, who couldn't be here, asked me to not focus on any one paper, but actually give an overview picture, which is what I'm going to do. And it's just serendipity that the topic that I've chosen actually blends in very well with the third paper of Joseph. So it's actually very nice. It's a nice continuity. It wasn't planned this way. Okay. So you've heard a lot about the affirmative action program. In India, I'll just give a very quick overview for those who don't know, which is that it's mostly a caste-based program, although you heard a lot about the gender component of it today, and it's directed against the scheduled castes, the scheduled tribes, and also increasingly now towards a group of intermediate communities, which are caste and non-Hindu communities, called the other backward classes, or the OBCs. And today, again, you heard a lot about the electoral component of the affirmative action program, but actually there's a very strong reservation quotas in education institutions and jobs, which are directed towards SCSTs and also for OBCs over the last 20 years. And in addition to that, the electoral reservation is only directed towards the scheduled castes and scheduled tribe communities, not towards OBCs. So that's the reservation for jobs in education. And women, as we heard, 33% of seats at the local bodies level only, which is what Victor talked about, is reserved for women. So that's a kind of overall summary of the Indian affirmative action program. There are other minor components to it, but I don't have time to go into that. So the belief that actually Joe also referred to, which is that affirmative action program is anti-merit. This is used for all affirmative action programs, but particularly in India, it's definitely raised more for the context of caste quotas in the context of gender quotas. In gender quotas also, there are stereotypes about women, but this argument that it's really unfair because it's anti-merit, this is an argument that's most compellingly made about caste quotas. And there are anti-reservation, anti-cota outfits that call themselves ironically youth for equality. So they believe that by opposing reservations of affirmative action, they're actually promoting equality. And they say things like, why should an essentially divisive element, which is caste, and carry over of the traditional society be a factor in determining something as modern as jobs? Why should people be divided on the basis of surnames? And why or why should merit not be the first, foremost, and the only factor that determines who gets the job? This is a verbatim quote from one of the protests that routinely gets organized. So if you see this quotation, what you realize is that the belief of the anti-reservation protesters is that the status quo in fact rewards merit. And the introduction of the affirmative action program disturbs the status quo. So that is why the state, that is why the affirmative action program is seen as an anti-merit program, but then it must follow that the status quo is, and this is the case, this is the case despite the fact that the status quo is obviously ridden by caste divisions. But that is never typically criticized as being anti-merit in some way or unfair in some other way. So that's really the point. Now, B.R. Ambedkar, who was the chairperson of the drafting committee of the Indian Constitution, he was India's first law minister and one of the most important writers and thinkers about caste discrimination, said many things, but one of the things that he said was that the caste system is often represented as a division of labor. It's a benign division of labor. But what he says is that it's not really a division of labor, but it's a division of laborers into social identities. So it's not efficient division of labor in the sense that economists understand it, and it's a hierarchy in which laborers are graded one on top of the other that has nothing to do with their intrinsic abilities. So the caste system itself is anti-merit, and so the proposed remedy has to be seen in this context. And it's not spontaneous, it's not based upon natural aptitudes, it's an attempt to assign tasks to individuals in advance, that is before they are born. Because as Victoria gave you the introduction to the caste system, caste is hereditary. So child even before he's born, or she's born, already is expected to follow a certain caste code. And that is why it is a stratification of occupations, which is the result of the caste system has nothing to do with aptitude or a naturalized kind of division of labor. Now, coming now to the modern system, and of course, many aspects of the caste system are weakening, it is not at all the case that modern occupations map neatly with the traditional caste divisions, and indeed in the modern economy there will be several individuals, millions of individuals who follow occupations that are not dictated by the traditional caste code. Now some part of it is to be expected because the modern occupational spectrum doesn't map easily on the traditional caste divisions. So there are a whole range of modern occupations for which there isn't a traditional caste counterpart. So you cannot follow, if you do a modern occupation, there isn't a caste counterpart to that. So in a trivial sense, it's obvious that the link between caste and occupation is breaking. But at a more fundamental sense, the question is, do traditionally lower caste individuals occupy lower runs of the occupational spectrum? And do traditionally high caste individuals occupy the higher run of the modern occupational spectrum? That's one question. And the second question is, when low caste and high caste individuals apply for same jobs with similar aptitude, are they treated differently? And there are a variety of studies that test for this, which is really the kind of overlap between caste and occupation in the present day context. And so there are correspondence studies which look at reward to merit. These are intention to hire. So these are resumes, identical resumes that are sent to employers. And employers are asked to choose in terms of callback rates. So who do they call back? Now, because it is not about hiring decisions per se, but it's only about intention to hire, it's what these studies are testing is how many people get called back. In that sense, it's kind of automatically controls for observables. Because all you're doing is you're looking at a paper resume and writing to the candidate, expressing an interest in the candidate. It's not a promise of a job. You can't say that it's because of unobservables. That's one set of studies. And then the other set of studies is the wage decomposition studies that of course Dita talked about. But that has the well-known problem of unobservables because these are wages of individuals who are already working and are earning a wage in the labor market. And the residual, which is not attributed to by characteristics is possibly discriminatory. But as we saw in the first presentation, this story is a little bit more complicated. Now, this brings to also a question of, if indeed it is the case that by the time individuals enter the labor market, their merit within codes is already different by broadcast groups. And in other words, if it is the case that in the labor market, upper caste individuals are entering the labor market with quote unquote greater merit and ability. And I'm using the word merit in codes at all times because if I have time, I'll talk about that later, about the creation of merit. And lower caste individuals are entering the labor market with lower merit. Then the question to be asked is, if it is not natural aptitude, then how do these merit differences get created by the time individuals reach a labor market? And I have a recent paper where I've tried to look at the early childhood roots of these merit differences. And indeed, it is the case that between 0 to 5 years, that's really the formative period where a lot of the differences that you see in adult life actually get reinforced and strengthened and created. And what we find in this paper is actually parental education is the key predictor of these early childhood differences. And when you look at parents that have more than 12 years of education, the caste differences in early childhood outcomes disappear. So really the key to break this intergenerational vicious cycle is really education. That's the conclusion that we draw from this particular paper. Then after children start going to school, so one is the early childhood which happens within the household and the community. Then after children start going to school, there is very clear evidence, both qualitative and quantitative, about teachers and peers discriminating against lower caste individuals. And the impact of that kind of discrimination can be very devastating both for self esteem, but also for again what is seen as merit by the time individuals reach the labor market. So these are all systematic and caste based. So for example, Hannah and Lenin's paper randomizes scripts, answer scripts and changes arbitrarily the gender and the caste of the student applications and shuffles them and gives them to 37 different evaluators. And then they take an average of the 37 indicators, of the 37 scores and then find that even controlling for heterogeneity in evaluators and then find that lower caste individuals, students and women are systematically given lower grades, even though the underlying script actually doesn't belong to them. So it's pretty compelling evidence in terms of discrimination inside schools. Then when you come to hiring practices, and all this is of course going towards why it's not really merit that's being rewarded in the labor market, but it's really discrimination and why affirmative action is therefore necessary. The role of networks in hiring practices is well researched, both in sociology and in economics well established and often who you know is more important than what you know in hiring decisions. I have a paper where I did a tracking study for over two years of university graduates. These are people with at least a master's degree. So these are highly educated individuals in India that have a very high chance of success in the labor market. And even in this highly educated group of individuals, we found that family background, employers always ask about family background. And this is a catchall phrase not only to look at caste, but also to look at caste. And it's really family background that determines in the end what kinds of jobs people get. So you find very clear differences in the jobs that people get depending on their family background. But there was a corresponding survey done at the time where employers were asked about what is it that they look for when they hire people. And everybody, including public sector, private sector and multinational corporations, gave a deep allegiance to merit. So everybody claimed that we want the best person for the job. So what we find is that employers pay a lip service to merit, but are completely blind to the structures on the ground that produce the so-called merit in the first place. So what they think of as merit are really stereotypes based upon caste and gender. And that's what they mistakenly believe to be merit. And that's what they reward in the labor market. In the private sector in India, personalized recruitment is very common. Jobs are not advertised. Employers actually justify that by saying that we find this convenient and efficient. It minimizes recruitment costs. It ensures commitment and loyalty, minimizes transaction costs of disciplining workers and handling disputes and grievances. So they prefer personalized systems of recruitment. But what that means is that coming back to the role of networks, that if you don't have the right kind of networks, you're unlikely to be even considered for the job. And this is all pre, you know, this is not even discrimination that happens at the stage where candidates are applying for jobs. This is selection that's taking place where it's highly likely that Dalit candidates will never even be considered for certain jobs because they never even apply. They don't even know that those jobs are getting advertised. I talked about the employer attitude survey. So basically, the opposition to affirmative action in India, but also everywhere actually, is very strongly based on commitment to the idea of merit. So people definitely believe that we want best people for all jobs. And therefore, affirmative action is often opposed, not often, always opposed with the idea that labor markets, if left untouched, always will reward merit. And it's really affirmative action that's interfering with that optimal working of the labor market. And of course, discussion and merit is conducted as if it's a neutral objective characteristic like height or weight. And there's a standard method to measure it. And so everybody can just tell what the merit implicit in a particular candidate is. And so what happens in these situations is that qualities of individuals are basically replaced by stereotypes that make it very difficult for highly qualified job applicants if they belong to the wrong social group to get jobs that they actually deserve. And so affirmative action should be seen in this context. I'll skip this part. This is about these entry scores, which are the determining factor for institutions. And what we have to understand about the Indian context is that many of these positions are highly competitive. And so even if the entry scores for the SCST students is lowered by 10 percentage points, it's still a very high, it's a very competitive entry. And so there's all this about death of merit by loading entry scores misses the point that even with that loading of 10 to 15 percent, you're getting highly meritorious applicants to universities. And there is a lot of opposition to affirmative action also because the belief, widespread belief is that affirmative action beneficiaries drop out. They are unable to compete. That's because they don't have the merit. And so they're unable to compete and they drop out. But the focus on dropouts in the popular imagination and the journalistic writings takes away from the millions of success stories. In other words, people who get admitted on affirmative action and actually complete their programs. And I have actually a paper in which I have a quantitative and a qualitative component. And the qualitative component is 60 in-depth interviews with successful graduates who came in on affirmative action. Because I think those stories need to be heard as well. Because we are so convinced, even those who are kind of well-meaning and support the affirmative action program are convinced about the dropout component of it. And so really that's something that we need to. And what I find in that paper actually resonates very well with findings in the context of the U.S. as well, which is that those who got into elite universities in the U.S. on the basis of affirmative action, even if they graduated with grades lower than their white counterparts, in terms of their long-term life outcomes, was substantively better than what anybody else in their families had been. And in fact, you find that the benefits that they obtained from the affirmative action admission goes just beyond the grades that they got at graduation. It is in terms of salaries. It is in terms of life satisfaction. It is also seen in terms of giving back more to the community. So this study finds actually that affirmative action beneficiaries do more pro bono work towards their communities than the whites, because the whites kind of take it for granted. Whereas these individuals are very acutely aware of the privilege that they've got. So it's a fantastic study. And in a minor way, my paper also mirrors that this view. Now, does affirmative action actually lower efficiency? I have a paper. This is the only one that has measured the efficiency impact of affirmative action in the context of the workplace. And we looked at a study of the Indian railways, and we actually don't find any negative effect of quotas on efficiency of Indian railways, which is the largest public sector employer in India. And it's amongst the largest employers anywhere in the world. Indian railways are a very big employer. And so it's a study with data over 23 years, and we don't actually find any adverse impact of affirmative action. On the contrary, when we look at people who are in top decision making positions who have come in on a quota, their presence actually, we find a mildly positive effect on efficiency. And I can talk about why that might be happening. And there's a very recent working paper which is looking at the performance of the Indian administrative officers who are the bureaucrats. And they echo exactly the same finding, which is that reserve category candidates do not have a negative efficiency effect at all. The mismatch hypothesis, which is that affirmative action in education is bad because you're placing candidates with a lower degree of preparedness in programs for which they're ill-suited to compete. And that's what leads to the dropout hypothesis. And there's a brilliant paper for India by Marianne Bertrand, Rima Hanna, and Sandeel Moolainathan. They have a unique data set, and they actually lay this myth to rest. I don't have time. I'm running out of time, so I can't. But they show that there is, in fact, no mismatch. People talked about unintended consequences in terms of backlash. I have a paper which was funded by you and your wider, which looks at another unintended consequence of affirmative action, which is in the form of increased stigmatization of beneficiaries. So it's possible that in addition to the stigma of their own cost status, they might further be stigmatized because of the fact that they came in through affirmative action, and therefore they are seen as less meritorious. So there's a further, and this is the paper that I've done using the you and your wider grant. And this stigmatization is important because it actually can increase the burden on beneficiaries to perform something that you talked about, which is that you have to go that extra mile. You have to work extra hard to prove that you're just as good. And so the stigmatization is serious. And what I did a survey, and I find actually that, yes, there is stigmatization. In other words, peers do stigmatize, do stigmatize affirmative action beneficiaries, but at least in my survey, not all the beneficiaries internalize that stigma. So when they were asked, do you think affirmative action beneficiaries are less meritorious, they didn't think so. I mean, I don't want to extend this argument. I don't want to push this argument. It's possible that in another context, there might be internalization, but it's not always the case that beneficiaries internalize the stigma. But the stigma is real. And we've seen consequences of that stigma in India recently through a high-profile suicide case of a research college, et cetera, et cetera. So stigma is something that definitely calls for public policy. But in my other paper where I talked about the dropouts, in that survey, I also asked about whether beneficiaries who are entitled to affirmative action would not use it for the fear of added stigmatization. And I find actually that that's not the case at all. They would much rather use affirmative action if they are eligible and if they can get a seat, rather than say, oh, no, no, no, we don't want to be further stigmatized because of, and the explanation is simple because they're already highly stigmatized. It's not as though not using affirmative action is going to lower stigma for them anymore. So, and stigmatization takes a variety of forms, including what are called microaggressions. But what I feel is that affirmative action has the potential to break the stigma because it also creates a Dalit, Dalit is a low caste, a low caste and a tribal middle class that can break the stigmatizing association between caste status and tribal status and incompetence. Because this middle class, which we already see in India, is small but is growing, has its presence in many different walks of life. And you can then say, well, this is not inherently lower merit, but it's really disadvantage through early life and a variety of factors. And so when you have successful success stories, those can really break the presumed association between lower caste status and merit. That's it. Thank you.