Ashwini Deshpande on Gender and Caste Discrimination and Affirmative Action in India

Shruti Rajagopalan and Ashwini Deshpande discuss the economics of discrimination, affirmative action based on caste, female labor force participation and more

In this episode, Shruti speaks with Ashwini Deshpande about caste discrimination in labor markets, the reservation system and its critiques, education of women, how endogamy perpetuates caste, Bollywood films and much more. Deshpande is a professor of economics and the founding director of the Centre for Economic Data and Analysis at Ashoka University. Her Ph.D. and early publications have been on the international debt crisis of the 1980s. Subsequently, she has been working on the economics of discrimination and affirmative action, with a focus on caste and gender in India. She is the author of “The Grammar of Caste: Economic Discrimination in Contemporary India” and “Affirmative Action in India.”

SHRUTI RAJAGOPALAN: Welcome to Ideas of India, where we examine the academic ideas that can propel India forward. My name is Shruti Rajagopalan, and I am a senior research fellow at the Mercatus Center at George Mason University. Today my guest is Ashwini Deshpande who is a professor of economics and the founding director of the Centre for Economic Data and Analysis at Ashoka University. She is also  the author of the books The Grammar of Caste: Economic Discrimination in Contemporary India” and “Affirmative Action in India.” We talked about caste and gender discrimination in India, the relationship between merit and affirmative action (or reservations as it is known in India), the low and declining female labor force participation, Indian cinema, Lata Mangeshkar, and more.

For a full transcript of this conversation, including helpful links of all the references mentioned, click the link in the show notes or visit

Hi, Ashwini, welcome to the show. I’m really excited to have you here.

ASHWINI DESHPANDE: Hi, Shruti. It’s a real delight and an honor to be on the show because I’ve been following it. You do such excellent, fantastic interviews. Really delighted to be here. Thank you so much for inviting me.

Taste-based Discrimination versus Statistical Discrimination

RAJAGOPALAN: I want to start with your book, the main work on discrimination relating to caste and gender. You are an economist. In the standard new classical economics, the canon is the Gary Becker model of economics of discrimination, from the ’57 book. The idea is that firms and employers that do discriminate in the market on grounds that have nothing to do with productivity will bear a cost of discriminating in the market.

This doesn’t mean that there is no discrimination in the equilibrium outcome. It just means that people who choose to discriminate will pay a cost for doing so, and that cost may or may not be large depending on cultural and other circumstances. Furthermore, that market competition will therefore make it more costly to discriminate and hopefully reduce these kinds of discriminations which are not related to productivity. How does this play out in India, especially with caste discrimination and gender discrimination, which have some overlap but can still take quite different forms?

DESHPANDE: Gary Becker’s theory of taste for discrimination—so it’s taste based. Even though there’s a cost associated with discriminating in a market setting, the reason that economic agents discriminate is because it’s a part of their taste and preferences. Even though it’s couched in neoclassical terms—and obviously it’s a neoclassical theory; it’s very much monetary equivalent of the taste for discrimination—I think that it does capture what, in other words, we call prejudice. In other words, if you really have a strong taste for discrimination, you don’t mind paying the cost for it.

In India, we see examples of that kind all the time, where you would say, I’m not going to go to a particular area to eat, for example, or to get my car serviced, or I won’t look for a house in that particular neighborhood even though it might be actually cheaper for me to buy something in that neighborhood, because I don’t want to live next to these neighbors or anything like that. The car service example is a very real one. People might choose a car service that’s more expensive rather than go to a cheaper one, which makes economic sense, so they are bearing a cost.

I think individuals do that all the time. Having said that, in the neoclassical canon, there’s another theory of discrimination which is not taste based, it’s statistical. The real challenge in the literature on discrimination is, how do you distinguish between taste based and statistical? I don’t think that there is a 100% clear answer to that. I think both things happen.

We see examples of stereotypes about women, or people from certain parts of the country, or certain caste groups which are articulated very effortlessly by people in India, as it is fairly untouched by political correctness in large parts of the country. Is that taste-based discrimination? Is that statistical? Some of it is obviously taste based. Some of it is statistical as well. Empirically or theoretically, to distinguish between the two channels—theoretically it is easier to distinguish; empirically to distinguish between the two channels is not that straightforward.

RAJAGOPALAN: Now in India, given how fractionalized our interactions are based on caste, do people really face a very competitive market when it comes to indulging in their prejudice?

DESHPANDE: By that, do you mean, are markets competitive enough that they will drive out people, let’s say, employers who are discriminating? Is that the question?

RAJAGOPALAN: Employers, but also something like, say, the housing market. Technically, you are paying a price when it comes to, let’s say, discriminating against someone because they eat meat, which is again, a very caste-based and religion-based preference or prejudice in certain circles. The housing market is really not that competitive. It’s not easy to increase the supply.

It is highly ghettoized for other reasons. There are various other reasons other than eating meat that certain groups will not be allowed in a particular building. Are most of the landlords really paying a price for their discrimination, or does the price they are paying capture the full amount of discrimination? Maybe that’s the other way to ask the question.

DESHPANDE: No. Because it’s so ingrained in their tastes, in principle they could find a tenant that would pay higher. If you just think of price as in just the monetary equivalent, they might be paying a price. We don’t know that because we don’t know the counterfactual. It could well be that if there’s a reshuffling, higher-paying tenants might be matched with landlords who are looking for higher rents, and so on.

What if the landlord’s aversion to meat-eating is extremely strong? Then the question is not really the price that they’re going to get for their apartment. What matters really for them is to get a tenant who doesn’t eat meat in their apartment. That is a very important factor in their calculations. I think what Becker’s theory made us aware of is the fact that it’s not always the monetary amount of that transaction that you see as the price. The monetary price of the transaction, all of these factors go into—it’s defining what the price feels like.

I always give my students this example of the weather app on our phones. It says “16 degrees, feels like 12,” what it really feels like. That’s what it feels like. This is what Becker pointed out as the monetary equivalent of the taste for discrimination. If a person ate meat in my apartment, and if I had a very strong aversion to meat-eating, even though that person might be giving me, let’s say, ₹50,000 a month as rent, to me, the act of my apartment being defiled by meat makes my rent feel like ₹30,000 or whatever it might be. Then it’s really the ₹30,000 that’s entering my calculation.

In Becker’s model, that’s exactly what happens, which is you literally adjust the price by the discrimination coefficient. Every individual is rationally aware of each of our millions of discrimination coefficients that we possess. Obviously, that’s not the way the real world functions, but the aversion for a particular feature or a practice or a group could be strong enough that it factors into your monetary calculations. I think we focus on the pecuniary, but what we need to realize is that discrimination shapes our feelings, that kind of a factor.

Caste Discrimination in the Labor Market

RAJAGOPALAN: Here now, I want to move a little bit away from the housing market and, let’s say, move to the labor market. When it comes to caste discrimination, there are multiple problems. One of course is just plain prejudice. The other is the consequences of historical oppression on building of human capital. The Becker model is basically talking about the kind of discrimination which has nothing to do with productivity.

In caste discrimination, there’s an element where the prejudice or the discrimination affects productivity, and there is additional discrimination other than affecting productivity. What does the Indian landscape look like when it comes to these two aspects of discrimination?

DESHPANDE: Here I’d urge you to move a little bit away from Becker and toward Arrow and statistical discrimination, because Arrow most elegantly showed how statistical discrimination can become a self-fulfilling prophecy. If there is a group that believes that they are not going to be eligible for higher-quality jobs, and if it is costly to invest in human capital to be eligible for those, it works backward, and you don’t actually invest in the human capital because you are not going to recoup those costs. I’m talking now strictly as an economist. The employer’s beliefs about the lower productivity of a group becomes a self-fulfilling prophecy. Effort goes down because you know that you’re not going to be rewarded for that effort.

Now, listeners who are listening at this moment might say, “You’re talking too much like an economist. Nobody thinks like this.” When you think about girls, for example, if the cues that girls are given from childhood is that, “Don’t go in for certain kinds of professions,” or “Play with dolls”—or girls are nudged toward the more housekeeping kinds or caring, nurturing kinds of games and options—they’re less likely to go into occupations or streams of education that are not geared for them to fulfill those roles.

The landscape or values or social norms or cultural ethos of a particular society builds in unconsciously into our own expectations and into what we see as our roles. It may be directly parents telling children, don’t do this or that. But it also could be just internalization at a subconscious level by a whole variety of actors, such that it results in a situation where girls are likely to choose pink and play with dolls and less likely to do STEM, or something like that. You know what I mean?


DESHPANDE: It becomes a self-fulfilling prophecy. Employers can turn around and say, “See, we told you we don’t have enough girls who are scientifically trained, don’t have enough women applicants who are scientifically trained.” The state of the labor market that gives out certain signals that go all the way down, and those signals then become a self-fulfilling prophecy. I find Arrow’s statistical discrimination theory really powerful here in terms of explaining how a cultural ethos affects choices and outcomes early on in the pipeline. Yes, there is a pipeline issue for sure.

Affirmative Action and Caste Discrimination

RAJAGOPALAN: This is where I find some of your work on affirmative action very interesting, because what you find is that despite the fact that there is so much stigma and prejudice a lot of the beneficiaries of the reservation programs in India face, especially the Scheduled Caste/Scheduled Tribe candidates, they don’t choose to drop out of those benefits. They actually still engage in everything that needs to be done to pass those exams, to get into those educational institutions regardless of the consequences. There are consequences on self-esteem and mental health, and so many other consequences. They don’t apply the same way in the private-sector market as their non-affirmative-action counterparts and so on.

Is it that affirmative action can bridge this gap or break the link with the self-fulfilling prophecy? Is that how I should read this work? Or is the way to read the work that for certain groups—in particular Dalit/Adivasis in India—the disadvantages due to historical oppression are just so great that stigma is not going to be the relevant hurdle that keeps them behind? The benefits of getting out of their traditional occupations is just too big to be deterred by some kind of stigma or prejudice.

DESHPANDE: Both of these. You’ve hit the nail on the head. This is what I always say when people talk about the stigma of reservations or stigma of affirmative action. Of course, there’s been excellent work in the United States on this, most famously the book “The Shape of the River” by Bowen and Bok. They show how, despite facing microaggressions in Ivy League universities when African American students who got admitted on affirmative action, 10 years down the line in their lives, their lives were on a much higher trajectory, socioeconomically speaking, compared to their peers who didn’t get into these Ivy Leagues through affirmative action, and compared to their previous generations.

If you ask them, “Would you rather not have been a student at Yale or Harvard because you face those microaggressions or stigma and had a very different life with poorer socioeconomic outcome?”—life is about tradeoffs. In fact, the interview, the record that any day the beneficiaries of affirmative action are happy with where they were at that stage in their lives, relative to not having been given this chance at all.

In my paper with Katherine Newman, “Where the Path Leads,” we find exactly the same thing. We interview students who got into three universities, which at that point when we did the study were the top universities in India: Delhi University, JNU, et cetera. Students openly told us about how they would not dream of any other life, even as they told us that in the university, they did face discrimination, microaggressions, not being invited to people’s homes or a separate seat in the hostel where food was being served, et cetera. All of these stigmatizing instances did occur.

As I keep telling my students, these individuals come from groups or communities that are already highly stigmatized. To think that by not getting reservation, their stigma is going to be any less—they are the most stigmatized groups in India today. What are they losing? If anything, they’re going to gain. There is now ample study that shows us the immense benefits of getting reservation. That’s one. The other is, of course, it breaks the stigmatizing association between being Dalit and being educationally backward or being considered incompetent. Just today, I had lunch with a couple of colleagues, and they were asking me about B.R. Ambedkar.

Having somebody like Ambedkar be a student at LSE shows you what’s possible. It holds an individual role model, but it also shows the possibilities that can take place. Of course, Ambedkar focused a lot on educate, agitate, organize, et cetera. Things subsequently have not been so straightforward. It is not straightforward that if you educate yourself, automatically hurdles in later life are going to disappear. Unfortunately, we are not there yet, but certainly, that’s the first step.

Having reservation, I think it’s a very important policy that allows us to break the stigmatizing association between belonging to a particular group that is stigmatized for its supposed lack of competence, which has only occurred because the caste system kept education away from certain groups consciously. That was a part of the caste code. It wasn’t as though Dalits were incapable of getting an education. They were denied education. Having educated members of your community then allows for those stigmatizing associations to weaken. Having said that, of course, it’s not a cakewalk, and it’s not straightforward, the transformation.

Critiques of Affirmative Action

RAJAGOPALAN: Here, I want to also get into the question of merit. There are two very typical—I want to call this the upper-middle-class and upper-caste critique of reservation. One is the stigmatization, which is very almost paternalist. The second is one of merit: somehow, this sense that those who are getting these jobs or these educational institution admissions because of affirmative action, they are not as meritorious as the candidates who are in the general category. I don’t know if this was ever true. Maybe it was in the very early years of the republic.

We’ve had litigation to prevent SC/ST candidates from coming into the general category because their exam scores and civil services scores are often good enough to qualify them in the general category. They may be outliers. What is the relationship between merit and affirmative action? Is it anything other than sour grapes, and we couldn’t get into these schools—upper-middle-class critique? Or is there something here that there is genuinely a big gap in merit, and that actually gets bridged through the education process or through the job skilling process?

DESHPANDE: This is a long question that has answers in several parts. Let’s just first understand the role of cutoff. The way the admissions take place to many universities, there is a cutoff, a score. If 80% is the cutoff, if you have 80.1 you get in, and if you have 79.9 you don’t get in. That’s how it works.

While the cutoff is seen as an indicator of merit, it’s really a market-clearing mechanism. A cutoff is a number that arises such that the number of entrants to a college is equal to the number of seats that are vacant. If there are 100 people applying to 100 positions, everybody gets in. If there are 1,000 people applying to 100 positions, 900 will not get in no matter what. As you can see, as the number of positions is fixed and the number of applicants rises, the cutoff is going to keep rising. First of all, that is just a literally a market-clearing mechanism. It’s a price that equates to demand and supply. That’s one thing.

The second thing is that when you look at the results of the competitive exams, whether it is the class 12 exams, whether it’s the MBBS, engineering college entrance exams, the numerical marks that students have been getting have been widely increasing over the decades. I got 80% in my class 12 exam, which for my time was considered a really good score. Today 80% is nothing. There are people who will say you barely passed. Today people are getting 99%, 100%—grade inflation.

What we really want to do is a relative ranking. That’s what is happening when admissions are going on. It is true that the cutoffs for reserve category students are lower. For caste categories, the difference in the cutoffs is announced publicly. We know the gap between the last general categories student who got in, and the SC cutoff. These are in the public domain. I’m not saying it’s insignificant, but it’s well known and it’s transparent.

A slight digression to my answer is that India also has a lot of other quotas. Most notably, something like a management quota, or something where in private engineering colleges, parents can give a donation, and somebody gets in. Those cutoffs, I don’t even know if there are any cutoffs there. In this merit argument, we always forget to talk about graduates who got into a college on all of these very opaque and nontransparent quotas.

The first point is that, yes, there is a gap in the entry scores between the general category and the Dalit students. However, after you take admission, there is a process of two to four to five years, depending on what course it is, after which you graduate. To think that somebody who was at, say, a particular position at the time of entry will always remain at that position at the time of exit is not true at all.

Just because somebody entered with a lower cutoff, which is well known, to assume that they will always remain lower at the exit point is incorrect. It may happen, because catch-up. But to believe that a complete catch-up can happen in two years where there’s been historical disadvantage of so many prior years is a little bit over-optimistic. Maybe they will graduate with lower scores as well.

We don’t have good data on the entry scores and the exit scores correlation, and that is not very strong. Because when I was at the Delhi School [of Economics], I had done some back-of-the-envelope correlations between entry and exit scores. Those were not strong.

Finally, the real question is one that has actually been examined in a paper by Marianne Bertrand, Sendhil Mullainathan and Rema Hanna, where they looked at the “Shape of the River” thing, where they’ve looked at the position of affirmative action beneficiaries 7 to 10 years down the line. In that paper, they got data on general-category students who were displaced by reservation. Take 100 seats and 1,000 applicants, 900 are not going to get in anyway, even if there was no reservation. Supposing 22 seats are reserved, then only 22 general category are displaced. In this paper, they are able to identify the actually displaced students.

They look at the outcomes of those displaced by reservations 10 years down the line, and they look at the outcomes of displacing students 10 years down the line. They find that those who got displaced, their occupation and income levels were not significantly lower than their parents or their peers who got in. In other words, they landed on their feet. Those who displaced, in other words the affirmative action beneficiaries, their occupation and income outcomes 10 years down the line was significantly higher than their parents and their peers.

Moral of the story is that reservations help the beneficiaries much more than they cost the losers from reservations. Any policy has winners and losers, and reservation is no different in that regard. I think the benefits to those who benefit are far greater than the presumed losses to those who lose out. Because the presumed loss is literally, “I didn’t get into this college, that’s it, but I land on my feet. I’ll do something else.”

Increasing the Education Supply

RAJAGOPALAN: I want to get into a couple of things. One is, the “I didn’t get into this college,” that is given a certain amount of supply. And to me, I feel like a lot of this grouse against the reservation policy from the general-category students who are mostly upper caste, I think it is rooted in the fact that our education system has not scaled with population. The size of the pie has simply not grown. What ends up happening is the reservation becomes a very easy target for people who presumably worked pretty hard, most of them, and then there is this huge sense of unfairness.

One very simple solution is just increase the size of the pie. We could have 1,000 more engineering colleges. We could triple the size of each IIT and still produce very high-quality students because the intake of the students would still be excellent, and of course you need to invest in the infrastructure. To me, it seems to be a supply-side problem, which has squeezed human capital formation in a particular way, and which may be leading to adding to the prejudice and discrimination that people face in society. Is that a good way to think about what’s happening in India?

DESHPANDE: To a large extent, yes. Shruti, I’ll just add one qualification to what you said, which is the supply of good-quality educational institutions. Across the length and breadth of India, you’ll find these teaching shops that have mushroomed without any regulation, without any attention to quality, curriculum, et cetera. Now you see large numbers of students, but they’re not regulated, no idea what’s being taught there. That’s why the good-quality institutions become so coveted, the IITs or the IIMs or Delhi University in its hayday and so on and so forth.

It really boils down to the priority that as a nation we need to give to education. Over there, economists are very fond of putting primary education in contradistinction with higher education —whenever anybody talks in favor of expanding higher education, they say, “No, but we need to focus on primary.” I always wondered why we need to do either/or. I think there’s a big demand for higher education. Why is it that the expansion of primary and the strengthening of primary needs to go on simultaneously? We need to focus on high-quality higher education.

You’re right. A lot of this grudge is because there are few spots and competition is intense. For that reason, I just want to say one more thing, which I wanted to say earlier in response to the merit question, is that actually competition is intense also for the reserve categories.

RAJAGOPALAN: Yes, exactly. That’s what I find so incredible about this merit conversation.

DESHPANDE: When you hear the merit conversation, or especially in popular debates, it seems like literally somebody walking on the street can just walk into an IIT. That doesn’t happen. Even for reserved category, the rejection rate is very high. While the SC/ST cutoffs are lower than the general category, there is intense competition even within those categories. If you think of cutoffs as an indicator of merit, then we have the best of each category.

Affirmative Action and Private-Sector Success

RAJAGOPALAN: You’re absolutely right. There’s fierce competition in every category, and we are getting the best of each group. Even if there are some gaps between the best of each group, hopefully that gap should narrow right over four years of education or three years of education.One of the interesting things about people from the reserved categories is they tend to have worse options or a worse chance in the private sector than the public sector. Of course, some of this is because public-sector jobs are specifically reserved for them. But some of this is also because private-sector jobs are so network-based in India, and you need to come from particular neighborhoods, speak English in a particular way, come from particular families, and have certain references and so on.

How much of the affirmative action candidate’s success in the private-sector market is because of this historical disadvantage, and how much of it is because the gap between their levels and the general-category levels didn’t quite completely bridge over the three or four years of education? This seems to be a thing that keeps popping up, that if they’re so good, why aren’t they doing so well in the private-sector market?

DESHPANDE: I think it’s both. I think there is one paper that finds that they continue to play catch-up. I think that’s based on one of the engineering colleges. While that paper gets cited as some sort of a proof that reservations are not so good after all, to my mind, even if you find that the gap remains or—it’s very hard to measure, first of all. How do you even measure that gap? Anyway, especially when you have streams and you have different options that students take, it’s really hard to measure.

For argument’s sake, let’s assume that the gap remains the same, which is hard to measure. And it’s hard to believe—also, the way the question should be framed is not so much as did the gap narrow, because I don’t think that’s the right way to assess the success of reservations. Because I don’t think the success of reservations is that in a college, in a course, the gap has to narrow. That may or may not happen.

I know that wasn’t your question, but that’s a very common thing that’s asked. “What’s the point? They’re going to remain at the bottom of the class,” et cetera. I know that you didn’t ask that question, but I’m answering it anyway because it’s a question that comes up a lot. What happens in the job market after the degree is a combination both of the possibility that the gap never closed, as well as the cultural deficits that mark many students who come in on reservation. These cultural deficits have their history in the early life, what economists called pre-market.

Unless you have a very strong bridge program, a conscious attempt to equalize a certain kind of cultural capital that you need to succeed in the labor market, these gaps are not going to bridge just by studying a particular course. For example, I tell you from my experience at the Delhi School [of Economics], is that when the private-sector employers used to come to recruit, they would have a group discussion, which was basically students sitting in a room where this person from the Citibank or whatever is there. The person is asking a question, and you have to quickly respond very smartly in English. It’s hard.

I thought of myself in those settings, and I probably will take a long time to answer the question. That is what they’re expecting, these compact, one-minute answers to these questions. I’m not saying that it’s only hard for reserve-category students. I’m sure that it’s hard also for many general-category students who don’t have the cultural privilege of having studied in the super fancy English medium school that gave them all these references, cultural expressions, smartness, debate, somebody who’s been a debater, et cetera. All of that confidence that comes with being any of these.

I found that a lot of reserve-category students opted out of the placement cell at the Delhi School. They didn’t even become members because they didn’t want to face the group discussion. How did I find all this out? When I was working on my paper “Where the Path Leads,” I used to ask them, “Did you join the placement cell?” which is a voluntary activity. They said, “No, we didn’t, because we didn’t want to be in this group discussion scenario.”

That’s an elimination round. If you don’t perform well in the group discussion, you are eliminated. They just didn’t want to have to face the possible humiliation of not performing well. The private sector, as you very rightly say, relies a lot on networks, relies a lot on markers of smartness, where not just students from stigmatized caste students, but also rural students suffer a lot.

RAJAGOPALAN: People from small towns.

DESHPANDE: Small towns.

RAJAGOPALAN: There are a whole bunch of markers. Women—

DESHPANDE: Markers. Not all women, I would say, but some women. It’s a lot about what you’re wearing, how you conduct yourself, and casual references to something or the other that reveals your class status and so on. As you know, India has several axes of disadvantage, and reserved-category students definitely have a lot of those disadvantages from early in life. That’s hard to overcome in a two-, four-, five-year course. It requires a lot of sensitization of the employers also. I don’t think college is going to be sufficient to bridge these kinds of gaps. Certainly not.

RAJAGOPALAN: I also feel like, more than the broader discussion of affirmative action and how we think about different groups and catch-up, there is a particular problem with the absolute top levels, whether it’s the private sector or whether it’s the state. It’s been captured by one group that thinks of success in one particular way. Successful people speak English like this. They watch these shows. They listen to these bands. They went to these schools, and they live in these neighborhoods.

Caste as Social Safety Net

RAJAGOPALAN: Pretty much all of the caste narrative in India is at some level played out through networks. Whether we’re talking about how they can enter a private-sector, high-end-of-the-private-sector job, whether it is people coming out of rural employment and getting into urban employment in the informal sector, whether it’s construction, whether it’s people moving to cities and they’re finding a place to live in the slum, everything is based on caste networks in India.

Is the reason for persistence of caste networks the fact that India doesn’t have any other safety net? Is it the substitute for that? Is it because India is so strongly endogamous?

DESHPANDE: I think it’s both marriage and employment, because as you say, we have of course been talking in terms of administrative categories of scheduled caste, scheduled tribes, et cetera. Caste is really jati, so all the way in which rural migrants to cities, or getting a construction job, what we call ticketers or the contractors, they are recruiting at the jati level. Because the administrative categories don’t mean anything in that space, they apply only to the public sector. It’s really a defining feature of most Indians’ lives.

Even when students such as the ones I teach say, “We don’t really think about caste because it doesn’t really matter,” but I say, “What about your marriage? How will you marry?” It doesn’t even occur to them that there’s a problem that you advertise or look for a match using your same jati, that it’s perpetuating the jati network.

Because at the end of the day, it can help you find a job as well, just as it can keep you away from certain jobs. You might get stuck at a low level, but what are the options then? I won’t say comforting, but it basically is a kind of social insurance which is not necessarily a good equilibrium because it can keep you trapped in a very low-level, low-productivity, exploitative, non-desirable situation.

RAJAGOPALAN: It helps them deal with volatility, and their lives are very precarious. Here, the question is what would be the policy intervention that could solve something like this? Is it an insurance program? Is it an unemployment insurance program? Is it a universal basic income? What is the social safety net?

DESHPANDE: I think a combination of things. I can’t think of a single policy that singlehandedly will substitute because at the end of the day, networks matter everywhere. There might be different kinds of networks in other countries. Today, Boris Johnson just resigned as the prime minister of the U.K. There’s been so much discussion on this elite boys’ club that he and his ministers ran. In the U.S., everywhere—I can’t even think about it, but it’s certainly utopian to think that India would be that exceptional society that you transition out of networks, because networks matter everywhere.

The question is, should jati be the defining feature of those networks? I think that is not a desirable outcome for a very large number of jatis. It also promotes mediocrity if you belong to the right kind of jati, because just by virtue of being somebody’s son or daughter, you will get certain benefits, whereas you may be completely unsuited to getting those benefits. It doesn’t just affect quality by not recognizing the merit of those who are at the bottom or are excluded, but it also rewards mediocrity. There’s also an overpayment to certain members of the upper caste, just as there’s an underpayment, talking about it as a labor economist.

I think that doesn’t make for a very dynamic and efficient society. I would say a bunch of different things where the labor market needs to be made genuinely meritocratic. People oppose reservations because they think it’s anti-meritocratic, and the assumption is that in the absence of reservations, our labor markets are meritocratic. They’re anything but. I think that working on what would make for genuinely meritocratic labor markets—which are compatible with reservation, I’ll argue—universal basic income could be one component, but that’s just a small part of what people need for sustenance.

Also, you think of the old-style welfare state. When you fell ill, when you needed food, libraries to go to, or parks to play in or any of these things, all our social needs, where are they? How are we fulfilling them? Do we have a society in which there is a neutral provision of services and facilities that fulfill our social needs as human beings? If the only people I visit are my kith and kin, my caste members, then that’s my network. Then I need that. I need that for survival, for feeling like a member of society. I think when you think of all of these different dimensions, then a set of policies—all of that has to be preceded by recognizing that reliance on caste networks is not making India efficient.



Endogamy and Gender Discrimination

RAJAGOPALAN: For me, I think the worst thing about jati-based labor markets as a social insurance or something like that is the consequences on women. Everywhere else in the world, which didn’t exactly have India’s caste system, but every society was stratified and had classes in some way—as the structural transformation was taking place in society, those break down.

In India, whether Indians move from colonial government to an Indian government, despite the fact that there are constitutional rights, despite liberalization, exogamous marriages are only hovering at about 5% to 6%. They’ve always been stably hovering at that level. There’s no movement. To me, that is the most shocking thing about what’s happening. And endogamy is the control over women.

I see all this conversation about we must smash the patriarchy and all of that, but in India, annihilation of caste and annihilation of patriarchy in India is pretty much the same road. How does this network-based society impact gender discrimination other than through marriage? What is going on there?

DESHPANDE: Marriage is obviously the instrument that keeps it alive, that keeps it perpetuating. It becomes perpetuating. Control over women’s bodies, their options are very strongly correlated to where the women are in terms of their caste hierarchy. People need to understand that when we talk about Indian women—and of course there are many, many common barriers that all Indian women or all women, not just Indian women, all women face everywhere—but the options are very strongly differentiated by where one lies in the caste hierarchy.

Having said that, of course, I must emphasize that there are women like the two of us here on the podcast, where there is a small group (that is small in proportional terms but large in numerical terms, because any small proportion of 1.6 billion ends up being a large number of people) that have—it might be through education, because of anti-caste reformist practices that have happened in several parts of the country (not recently, but in the previous century)—have had a role to play in creating a slightly progressive outlook that has led to endorsement of girls’ education, letting girls do what they want. I think it’s important for listeners to also realize that. Because I don’t want to give the impression that it’s just a complete state of darkness with no light at all.

I come from Maharashtra. That has had a very significant social reform movement, of course, in the 19th century. But also when you go back to the Bhakti movement of the eighth and the ninth centuries, which were very strong anti-Brahmanical, very progressive egalitarian movements with a huge very rich literature, poetry, and which many, many Maharashtrans genuinely believe in. The Bhakti poetry informs a lot of the way Maharashtrans would think about women or about the role of women, or about caste even.

That’s important. Maharashtra is not far from the only state in India that has had these kinds of anti-caste—Tamil Nadu is one, West Bengal. Several parts of India have historically had very strong anti-caste, egalitarian movements that have targeted patriarchy and caste oppression together. Just putting that on the table. Having said that, that’s still not the majority. For the majority of women, the higher up they are in the caste hierarchy, there are greater constraints on the choices they make, primarily because it’s a question of honor and it’s a question of property. It’s plain and simple.

That is an issue even though if you think of legislation—for example, changes in the Hindu Succession Act that now allow women to inherit property—there has been a lot of legislation in the right direction. But I think it’ll be a while before you start feeling the impact of those changes on the ground in terms of greater equality, in terms of property ownership. I would say asset ownership or property ownership remains one of the biggest areas where you have male-female disparity. Even in communities where girls study well, et cetera, but property ownership, privilege. Property ownership still remains a big—as long as that’s the case, there will be restrictions on who the girls are going to marry because at the end of the day, it boils down to sharing wealth and sharing property.

Female Labor Force Participation

RAJAGOPALAN: We’re talking about caste being the embarrassing thing about India and the dark picture. The other is female labor force participation. The numbers are just horrifying. They are now in the mid-teens; some people say the low teens because there’s a problem in the way the data are capturing the information. There are two, three things going on with female labor force participation.

There was this big feeling when liberalization happened that as the structural transformation happens and India grows richer, there are going to be more women in the workforce. That didn’t quite exactly pan out. The economist response to that is there is a U-shaped curve. As communities get richer, first women drop out of the labor force because for most people, the household income has gone up. Now they feel like they can make that tradeoff. Children should be raised by the woman in the house if it’s a cultural preference. In India, of course, it’s this honor-income tradeoff. The woman can stay at home and we can preserve the honor.

Then the U-shaped curve is, as you get further down the development path and as you get really rich, then you’re going to see more and more women like you and me, for instance, those who got educated and have the ability to work. And work is not just because you have to support your family; it’s also a means to other ends, and it’s a form of agency, and so on. Is this U-shaped curve playing out in India the way we think it plays out in other parts of the world? Is there something else going on in India? Why is there this war against women in the labor market?

DESHPANDE: Again, a long question that has several parts to the answer. First, I don’t see the evidence of a U shape, in the way that is being argued for other countries in the world, in India. Where you see the U shape is with education. The highest female labor force participation rates are either illiterate women or highly educated women. That also puts the question mark on A, the income-effect story, and B, this honor-income tradeoff that gets talked about a lot.

I have a lot of skepticism about that argument because if you look at the honor-income tradeoff story, then is the honor when you talk about the poorest women or when you talk about the middle-income women?

Now, one of the issues that I have with a lot of commentaries on India female labor force participation rate is that people confuse the low levels with the decline. I think that they’re related, but they are separate phenomena. Even at its very peak, the recorded female labor force participation rate in India did not exceed 40%. It remained at 36%, 38%.

Now, when you look at ethnographic accounts of village studies done by sociologists, anthropologists, if you travel by train, if you travel by bus, you see fields, you travel through villages—that 30%, 40%, whatever percent figure, the low level I’m talking about now, that doesn’t square with the visuals of women working in fields, doing small things, bustling about.

Pune is a city where a lot of my family lives. If you came to Pune, you would not believe that there’s such a low female labor force participation rate because you see women bustling about everywhere. What’s going on there? I’m not talking about Pune, but I’m just talking about the Indian story. On the low-level question, a big component of that low level is that family enterprises—and this includes farming—run on the back of free labor provided by women. These are economic activities. I’m not talking about care and child care and cooking and cleaning and all the rest of it. These are activities that when a man does them, they are counted in the GDP, and the man is counted as a worker. When the woman does them—

RAJAGOPALAN: They are not counted.

DESHPANDE: —she’s not counted as a worker. A large part of the low level is explained by the fact that unpaid economic activities of women are not counted, and these women are not counted as workers. My submission to you is that many more women are workers. If you count working in activities for pay and profit, if that is counted as a definition of work, which is what the ILO [International Labour Organization] uses, I think that’s an issue that needs to be fixed. This is not to count their care work hours and all that.

RAJAGOPALAN: I agree. On this, I have a question. How much is the level affected by this?

DESHPANDE: My work with Naila Kabeer, where we did a primary survey in West Bengal, we actually calculated the difference. Our survey was in seven districts of West Bengal which, using the NSS [National Sample Survey] data for that time, these seven districts had a labor force participation rate of 18 or 20. I forget exactly what. Just counting these economic activities, our labor force participation rate estimate for these seven districts is 56%. That’s the order of magnitude difference. This is not expanding the definition of work. We take the strict economic GDP definition of work. That’s one thing.

The other question is the decline now. I could go on talking about this, but let me stop here. Now, let’s come to the decline. I think we need to understand when the decline started, where it is happening, before we put forward theories about them.

If you believe NSS data, as I do believe NSS, the decline started in 2004. According to NSS data, the decline is a rural phenomenon. Within rural, it is a phenomenon that affects Adivasi women. It’s really a decline that shows up for rural ST women, if you believe NSS. If you don’t believe NSS, you believe CMIE [Centre for Monitoring Indian Economy], there’s a decline in both urban and rural. Anyway, they don’t go back that much; they only started in 2016. From CMIE, we will never know what happened earlier, because CMIE did not collect data on employment. We have to use NSS, whether we like it or not.

Now, when we think of the decline starting in 2004, then we have to ask ourselves whether the cultural explanations that are very popular in the ecosystem, particularly among economists who discovered culture—after having neglected culture, now everything is culture. In 2004 in India, obviously, there are cultural factors that particularly disadvantaged rural Adivasi women in a way that led them to withdraw from the labor force? Does that sound like a plausible explanation? Or does the fact that the job—


DESHPANDE: Well, MGNREGA, they get counted as workers. MGNREGA is not it. The famous jobless growth period coincides, really, with 2004, 2005. I have a graph, which unfortunately, this is not a video; otherwise, I would have shown you. I’ve tweeted that graph earlier, where you see the rise in the population that’s eligible for jobs, that is the working-age population, and the curve that is employed people that starts diverging from 2004. What’s really happening is that as the job scarcity is increasing, women are getting disproportionately impacted by the job scarcity.

I have two separate papers in which I show the effect of demonetization and the COVID lockdown. Both of those disproportionately impacted female labor force participation. The real story is that when jobs are scarce—

RAJAGOPALAN: Women get thrown out.

DESHPANDE: —women get disproportionately impacted.

COVID and Demonetization

RAJAGOPALAN: Yes. I want to get into the stuff you did on COVID and demonetization. On this, what you find is actually quite remarkable, because this entire time, whether it is culture or whether it is income-honor tradeoff or so many other things like public safety, motherhood, care work—all these explanations are about the supply side of women’s labor supply not being forthcoming. Even when I said MGNREGA for Adivasi women, for me, that was a supply-side explanation, that if now one member of the family can qualify for MGNREGA, then no need for the other person in the family to work, presumably, or they might choose to drop out of the labor force and so on, right?

DESHPANDE: I see. Okay. Understood.

RAJAGOPALAN: All these are in some sense, supply-side explanations. What you find is quite remarkable, that it’s actually a demand-side story. Now, in this demand-side story, it sounds like very classic discrimination. Now, is this productivity-based discrimination, skill-based discrimination or is this just gender prejudice?

DESHPANDE: When you think of the drop in rural areas, other than agriculture, for women, avenues of employment are practically nonexistent. You can be a schoolteacher, you can be a nurse in the primary healthcare center, but that about sums it up. ASHA workers, Anganwadi workers are fighting for years to be classified as full-fledged government workers, and that’s not happening yet.


DESHPANDE: And paid, which is a real tragedy. If you just increase proper employment, even through these three channels, that would open up. I think that women particularly at the bottom of the caste hierarchy, Dalit women and Adivasi women, they get particularly hit because their families don’t own as much farmland as upper-caste families would. For them, other than fisheries if you live near the water or whatever, the options for rural Dalit or Adivasi women to engage in economically productive work, if their family enterprises are going through a crisis, that the men have migrated, are very limited.

In rural areas, I would say gender discrimination? Yes, maybe, but I think it’s more about just no avenues for these women. My colleague Kanika Mahajan’s work shows that mechanization that happened in the period exactly that I’m talking about, 2004 et cetera, the kind of mechanization that happened, substituted more for agricultural tasks that were dominated by women—seeding, sowing, implanting and things like that, or plucking at the smaller harvest in-season, et cetera. Technological change disproportionately disadvantaged women in rural India, exactly at the time when you see a decline in rural female labor force participation rate.

In urban India, we talked about the absence of a social network, welfare services. And that’s the reason why in urban India, women have always had a lower labor force participation rate than rural India, because where do you leave the children? What do you do with the elderly family members, et cetera? For urban Indian labor force participation rates, COVID is an exception, but pre-COVID, they hadn’t really dropped; they just remained low.

They’ve always been low, and they didn’t change. All this thing of rising urban crime, et cetera—Jitendra and I, not in this paper but elsewhere, we’ve actually done fairly robust correlations across countries, and we don’t find a correlation between sexual violence against women and their labor force participation rates. Not in India, nowhere else. Yes, sexual violence against women is a huge problem.

RAJAGOPALAN: There’ll be some substitution, right? Where you choose among the various opportunities, you’ll substitute, but the levels won’t decline, is what you’re saying.

DESHPANDE: Yes. In Delhi, as the resettlement colonies get pushed further and further away from the centers—urban India lives on the backs of paid labor, so all of our domestic services are substituted by paid labor—these women travel huge distances because they need to work. They do need the work, and these are not necessarily illiterate women; they could be middle-level educated women.

The point is that, take for example a mobile repair shop in India—or a courier service, which is more or less a substitute for the postal service in urban India, certainly—why can’t women be courier delivery people? Now, DHL in India has started employing women. There is that element of a mindset of gender discrimination, where we don’t see women as possible or capable.

RAJAGOPALAN: Uber drivers or auto drivers or courier or whatever.

DESHPANDE: Uber drivers, deliveries or couriers, or all these middle class—in urban areas. In rural areas it’s a different problem. I think that is a lot of merit to the old-fashioned economic arguments of just a paucity of jobs, absence of support services that will allow women to access the jobs that exist, and so on and so forth, before we jump into the cultural norms.

Also this honor-income tradeoff—I mean, think about rural Adivasi women. What is the honor-income tradeoff for them? In my paper with Jitendra, when we document the number of switches that women do, enter and exit the labor—these are the same women; we are following the same women because it’s panel data—the number of switches by Dalit and Adivasi women is the highest. Every four months, you see them entering and exiting. Not every woman, but our data is every four months, which means that when they do get a job, they’re working.

Are Cultural Factors Affecting Women’s Labor Force Participation?

RAJAGOPALAN: Is the honor-income tradeoff more about how certain castes are trying to be more and more Brahminized, as they call it? They’re mimicking certain features or restrictions on women, because that’s what the uppermost castes have historically done? Is there any merit to that?

DESHPANDE: There could be. I’ve seen that. On the field trips that I’ve been, you see the spread of, for example, ghoonghat, veiling, covering your head, in castes that previously never practiced ghoonghat. Dalits never historically had ghoonghat barriers. You might see women, obviously, women even more aggressively supporting ghoonghat than earlier, and so on.

I’m not an anthropologist, so I would not be the best person to comment on how widespread these changes are. And indeed, if they are in all parts of the country or not, I don’t know. I only know a few parts of the country where I’ve done fieldwork. There is discussion about sankritization, including constraints on women and so on. Whether that translates into preventing a woman from taking up a job when it exists, and when she can access it, I think that’s a leap of faith. We need evidence to show that. We don’t know.

RAJAGOPALAN: Does it come into your very first point, when we were talking about a self-fulfilling prophecy of how certain cultural codes and norms just make you believe that there is no possibility of anything beyond getting married when you’re 18 or 20? What is even the point of investing in high levels of human capital? Is there a reason that doesn’t quite work the same way for gender, because of caste norms, in a way that it works for caste stratification?

DESHPANDE: It doesn’t work even for caste. I was explaining to you Arrow’s model. On the ground, as we saw, students don’t internalize these biases. They want to study.

RAJAGOPALAN: Do women, is my question.

DESHPANDE: Neither do women. If you look at the expansion of higher education in India, precisely during the period when you see a fall in women’s labor force participation, it is increasing.

Especially women in higher education. Which means leaving home, traveling long distances, battling all those prejudices and cultural factors and stares and groping, all the uncomfortable experiences that are supposed to prevent women from accessing jobs, or that are supposed to be leaving their families to tell them, “No, you can’t work because you’ll meet strange men; you will be groped on the bus,” et cetera. But this very same woman went to college, which was far away from her home. She took public transport; she went there. Whether she fought inside the home or her parents allowed her, we don’t know; we will not know. The fact is, the numbers are showing that women are going to college.

RAJAGOPALAN: In absolute terms, women are going to college, but for instance, Girija Borker’s work shows that women change; they have to incur a higher cost. They have to either spend more money or spend more time to take the safest route to get to college, or to get to the place of work, relative to a man. There is a certain cost that women bear because of their gender, but it may not turn into a tradeoff that actually affects levels at the workplace or something like that.

DESHPANDE: Yes. I don’t doubt the cost at all. The fact is, is it leading Indian women to say, “Okay, I’m not going to go to college because the streets are not well lit, or I don’t have a bus connection”? They’ll wait for hours because that’s their chance. That’s the other part of it, yes. The other thing is that we now have in India a phenomenon of, what is it? Educational hypogamy, which is women, on average, are more educated than their husbands, which was not the case 20 years ago.

Second, the total fertility rate in India has declined. It’s now at replacement. Urban is 1.8, total is 2.1, rural is 2.2 or 2.3 or something.

RAJAGOPALAN: Which means some of these motherhood supply-side factors can’t be the explanation, right?

DESHPANDE: That’s one. The interstate or community variation—and all of this is converging very fast. Educational gaps, male-female, are converging very fast. Again, the community differences, et cetera, are narrowing, and so on. Maternal mortality rate significantly dropped in India. All of these factors, Shruti, are factors where elsewhere in the world were the base factors for female labor force participation rate to rise.

Now, unless you argue that Indian people are wired strangely, and Indian people are not like the rest of the world—which I don’t buy at all; we are all people, we are all human beings—we have to look for alternative explanations other than cultural explanations for why we are not seeing absorption of this increasingly educated female human capital power that lies in our country. I think we’re losing out. India is losing huge amounts of potential addition to GDP by not absorbing the productive potential of Indian women, which is abundant.

Increasing Female Labor Force Participation

RAJAGOPALAN: I think there are some things that could be done on the margin, even on the demand side, which have to do a little bit with the supply side. For instance, I’ve seen recent literature—this is work by Ashish Sedai. He’s talking about access to indoor piped water or access to electricity, how that improves the status of women relative to men, but through employment opportunities for two reasons. One, there is a reduction of time allocated in the household tasks.

But more importantly, they’re more reliable—if there isn’t piped indoor water, that means you’re using some strange source of water, your kids fall sick more, which means you might live in a community where women take more sick days, so employees just don’t want to employ women. If you’re a community that gets indoor piped water or gets electrification, then the norm changes. Because now, women take the same number of sick days as men, which means employers are not going to not hire women at all, or are in fact more likely to hire women if the jobs open up and so on.

There is something to do with public goods provisioning and things like that. But you’re right in that they can’t be this simple, straightforward, “Oh, women’s supply of labor has dried up.” I think you’re right, that we need to walk away from that kind of simple theory of what’s happening in India.

DESHPANDE: As long as you continue to believe that, then there’s nothing to be done. Obviously, some cultural factors have taken over, and Indian women have just decided, “Okay, you know what? We can’t work anymore.” Okay, then end of story then, right? It isn’t the end of the story. Of course, having a talk about economic factors, I must mention one cultural norm, by the way, that definitely needs changing in India, South Asia, is the norm that places almost exclusive burden of domestic chores on women. Whether the women do it themselves, they get it done.

RAJAGOPALAN: Yes. It’s all women.

DESHPANDE: It’s all women. Of course, it’s a huge commitment of time, but it’s also a huge commitment of headspace and mental energy.

RAJAGOPALAN: You showed it in the COVID work, right?


RAJAGOPALAN: That was a clean break. You can do a very nice experiment where both men and women are at home, and somehow, magically, still it’s women doing—the first couple of weeks, the men chip in, and then it just disappears.

DESHPANDE: By the time you reach December 2020, men, according to CMIE, are doing even less than they did before COVID. It’s like done it for one month, and that’s it, our quota for life is over. I think there, messaging, incentives, role models—again, to think that any one policy is going to change any of this is obviously not. So, for example, this very important provision of electricity, water, and piped gas or LPG. All of these are very good examples of policies that should be promoted to reduce time on domestic chores, and it is also gender neutral. Men can operate these things too.

RAJAGOPALAN: Exactly, yes.

DESHPANDE: Certain companies are doing some fabulous advertising campaigns, like Ariel, the washing detergent. They have started this hashtag, #sharetheload. They are beautiful ads, but these reach a very small set, like the urban—

RAJAGOPALAN: Yes, the top 1%. These are for us; we are mimicking the West. And even there, they’re not sharing the load exactly, but they pretty much only work for people like us, right?

DESHPANDE: Yes. I do feel that the conversation on sharing domestic chores, and that is everybody’s job, I think it’s an important one and must be aggressively pushed.

RAJAGOPALAN: Yes. That norm-setting has to happen at every level. It can’t just happen in the labor force conversation. It has to happen in school. It has happened in kindergarten. It has to happen as older sisters are looking after and doing the care work when there are younger brothers. This is a more complicated thing than it looks like at first blush; it’s not just about loading the washing machine.

DESHPANDE: No, in fact, there’s a slogan called “Beti Bachao, Beti Padhao” which is an excellent slogan, which is, “Educate the daughter, save the girl child.” Obviously, we all believe in this idea. But on my way to work, I pass a holding for this particular slogan, where the girl child is shown in the kitchen making rotis, making bread. Obviously, I would like to believe that it was unintentional. However, what is the message that the holding is giving us? Unconsciously, it’s saying we need to save the girl child because we want her to work in the kitchen. I don’t think that’s the reason—

RAJAGOPALAN: The narrative is, if you educate the girl child, your next generation will be better educated, because the mother will teach the kids or help them with their homework or something. It’s not really to absorb that human capital in the labor market. You’re right. There’s something very funky going on.

DESHPANDE: I think there, the change in the conversation is needed. The current government has such a massive majority, and it has the ear of the people. If they could just give the right kind of messages about gender equality inside the home, it would have a tremendous reach.

RAJAGOPALAN: I agree. I do agree. Some norm-setting is always helpful.

The Reservation System and Historical Oppression

RAJAGOPALAN: I want to go back to our reservation discussion. My main critique of the reservation policy in India actually has nothing to do with merit and stigma, which is the usual grouse. My argument is actually twofold. This is not against reservation per se, but it’s more about how we’ve done it. One is, the Ambedkar vision started with reservations for Dalit and Adivasis because of historical oppression and discrimination. Then of course, we’ve moved quite far from that.

There is the Mandal movement, which is the first step away. It becomes a Dalit- Bahujan movement. I don’t want to get into the tension of that movement, but at that time, there were many Dalits who felt that people who were caste Hindus and part of systematically oppressing them just 50 years before are now being included in the same group as them. Leaving that aside, then you have states like Tamil Nadu which break the Mandal rule, which exceed the above 50% through the Ninth Schedule and stuff like that.

Now we have this 103rd constitutional amendment, which is actually further breaking the link between historical oppression, discrimination and reservations. Now it is social and economic backwardness. Your work shows that as we keep adding more and more groups to be eligible for reservation policies, the newer groups are—the problem is not just that they’re not historically oppressed. It’s that even within those groups, the beneficiaries are not the poorer ones, right?

It’s always the rich, elite Jats or the Kapus or the Patidars who are getting the benefits of these. That is one problem that I have with not reservation policy per se, but how we have conducted this. Related to that is, the entire goal was to annihilate caste, right? It seems like what we’ve done is constitutionalized caste. The process by which we are tagging on more and more people to get these benefits is not helping us make them lose their caste identity. It’s the opposite. The benefits are linked to their caste identity. From the moment the child is born, the first thing you do is get them the SC/ST or OBC certificate. Then that’s the perpetuation.

If you were the one who was designing this from scratch today, what is a good way to think about this problem in a way that we are not strengthening caste identities?

DESHPANDE: One of the reasons why this is a complicated issue is we think we have now designed the reservation policy as the only “magic wand” that is supposed to solve all manners of caste disparity. As long as we think of reservations as a solution to caste inequalities, it is going to reach a dead end as it does often. The demand from additional groups on the reservation train is linked to what we were talking about earlier, which is just the demand for good jobs. I think a lot of problems are getting articulated when people demand reservations or to be included in the reservations group.

Somehow, we think that, okay, we just add one more group to this shrinking pie, that problem is going to be magically solved. I think we have to disentangle the problems faced by various communities and then address them. I definitely believe that the original Ambedkar vision, which was a limited time to create a level playing field so that you have equality of opportunity, which could lead to inequality of outcomes. That’s a separate matter, but the focus should be on equality of opportunity.

While Ambedkar might have phased it for college-level education, if you think of equality of opportunity as the main focus of the policy, then you must make sure that there’s equality for opportunity at every level. It doesn’t mean a quota, but it means that the kind of schools that children go to in rural India, the kind of health facilities that they have any—everything that makes zero to five, the early childhood years, other—

RAJAGOPALAN: Also, the source of water, and your chance of survival. Sometimes, it’s so basic.

DESHPANDE: The cradle of adult life disparities is actually zero to five, early childhood. If you think of equality of opportunity as the guiding principle of policy, then you will make sure that in addition to giving reservation at the college level, you will equalize opportunities at all of these steps. If that had been done, then we could have talked reservations being redundant after, say, 15 to 20 years. I think that’s how Ambedkar envisioned it. Unfortunately, reservation is seen as the beginning and the end of it. Now we’ve given reservations. Whether they come in, they sink, they swim, it’s just a mechanical policy that everybody has to follow.

Coming to the other communities, there is a genuine discontent and a genuine anxiety about stable livelihoods, particularly from agrarian communities. We’ve looked at Jats in Haryana, we’ve looked at Marathas in Maharashtra, Patels in Gujarat. We’ve looked at their demands. We’ve looked at their socioeconomic outcomes in a variety of dimensions—my work with Rajesh Ramachandran. These three communities, in their respective states, even do better than the legal OBCs right now. Based on any indication of backwardness, there is no justification for these communities to be included in the reservation bandwagon.

Because of the crisis in Indian agriculture, because of the anxieties that our children may not get a good job—and this, in the village, is going to get a reserved category job. Even if it’s just one person, it’s seen as something that makes the myth of reservations giving jobs. Reservations don’t create employment.

RAJAGOPALAN: Yes. It’s a tiny fraction of the total workforce. It’s irrelevant.

DESHPANDE: It’s irrelevant. Actually, when you think of the total number of jobs, and if you think of what number of proportional jobs is impacted by reservations and the reserved category, it’s a tiny fraction. But look at the fight over that little tiny fraction.

RAJAGOPALAN: We know that from liberalization, everyone benefited, but the upper-caste, English-speaking, service-sector groups, especially urban, benefited the most. Is this a feeling that their rural counterparts, the high-status, landowning rural counterparts never got those benefits from liberalization? In some sense, they got left behind, right? Because we never liberalized agriculture, that rural structural transformation that was supposed to happen never really happened.

Our land sale laws are so terrible that they couldn’t quite exactly monetize their biggest asset. Their children are not well educated. Is this where that anxiety is really stemming from, that they didn’t benefit from the structural transformation?

DESHPANDE: That is one. The other is, with the entry of big corporations into agriculture—while on the one hand, there isn’t a land market, and we haven’t monetized agriculture, but on the other hand, there are these big fears of corporate takeover.

RAJAGOPALAN: Your bargaining power has now completely changed as opposed to your high status before.

DESHPANDE: Look at the farm laws, look at the anxiety that got exhibited through the agitation. Whether it is justified, not justified, is that the solution or not, I’m not going into that. I’m not an expert on those topics. But the point is that the fear of a corporate takeover, along with missing out relative to your urban counterparts, that status doesn’t count for much, because a sarkari naukri [government job]  is a sarkari naukri and you see some little boy from a Dalit family who’s in the city, having a sarkari naukri, brings in stable income every month. You see the protest on that. All of this Agniveer Scheme all of this is a jobs problem.

RAJAGOPALAN: Absolutely. I saw the graph you tweeted after the agniveer announcement and my tweet. We’ll try and pull that up when we link the transcript. It’s extraordinary. There’s no way this scheme is going to plug the hole. There is no government jobs program that is going to plug this hole.


RAJAGOPALAN: We just have to create jobs in the Indian economy. The crux of whether it is women, whether it is caste discrimination, whether it is upper-caste anxiety, everything finally seems to be a labor problem in India. That’s what it’s boiled down to.

DESHPANDE: We talk about our youth dividend. We should, by the way, talk also about our gender dividend, because I think women embody in themselves huge amounts of productive potential, especially after the expansion of education. I think we are losing out. The Indian economy is losing out, because we are not utilizing our educated workforce, young people, women who could be in the workforce and contribute.

RAJAGOPALAN: Absolutely. Also, we think of jobs as the end, but actually, the jobs are also the means of building human capital further. It’s a missed opportunity both in harnessing the human capital and also building it further. It’s like two missed opportunities. It’s really frightening how bad things are.

On Bollywood

RAJAGOPALAN: But on the depressing note, I want to move to some happier topics. One of the happier topics is our common love for Bollywood movies, right?


RAJAGOPALAN: It seems that yours is very much rooted in music, which seems to be the same for me. I know more Bollywood songs, and I know songs from movies, everything about those songs, and I haven’t even seen those films. For me, it’s very much rooted in a music of a particular kind. Where is this coming for you? Is it also the music, or is it both the movies and the films?

DESHPANDE: Music and the films. When I was a child, watching Hindi movies was not considered things that cultured people do. It was very infra dig. But my father, who was a very unusual person—from my father’s side, the whole family was crazy about the movies. My father had absolutely no qualms about taking us to movies. My classmates used to be shocked that we went with our parents to watch trashy movies.

My father had this classification, which I really like, which is that there are four kinds of films. There are the good good films, there are the good bad films, there are the bad good films and there are the bad bad films. His point is that the good good—sorry, the good bad. The good are your Satyajit Ray, and the good bad is Manmohan Desai blockbuster. It’s a bad film, but it’s a really good bad film.

RAJAGOPALAN: Good bad film.

DESHPANDE: His point is, you guys must be exposed to both the good good, which also we watched a lot of, but also the good bad. The category that he hated the most were the bad good films, the films that are giving a message, or some art film that is so boring. His point is, I’d much rather you watch a Govinda film than watch some arty-farty film with a message.

RAJAGOPALAN: NFDC [National Film Development Corporation] depressing thing.

DESHPANDE: Boring, whatever. As a result, I grew up watching movies at a time when my classmates were absolutely shocked that I was watching all those movies, and the music. The music of the golden era of the Hindi films, which is 1950s to 1970s, is my most favorite music other than Hindu classical.

RAJAGOPALAN: Which is reflected in that time, actually; that’s not surprising.

DESHPANDE: It’s not. It’s difficult to separate the two. That’s my go-to music for all times. As I say—not just me, I didn’t invent this—but basically, from that era, there’s a Hindi film song to describe every possible human emotion. Whatever micro, minute feeling that you’re feeling, there will be a song that will describe exactly that. That’s why it’s so enriching and so fulfilling.

Lata Mangeshkar as Icon

RAJAGOPALAN: We recently lost Lata Mangeshkar, and she was the last person from the Nehruvian era, the immediate—all the hope that came with the post-independence movement. She lasted till post-liberalization. What is it about Lata Mangeshkar that we think about or we care about, or what does that icon really mean?

DESHPANDE: I gave a talk recently in Delhi on the music of Lata Mangeshkar and the idea of India. Basically, the best way to understand what Lata stood for is to read Nasreen Munni Kabir’s book, “Lata Mangeshkar . . . in Her Own Voice,” because as with all superstars, there’s a lot of mythology surrounding the individual. Lata Mangeshkar was all too human, with all the human strengths and weaknesses that we see in any person. What I want to argue is that she was an extremely independent woman with her own thoughts and ideas. She led her own, she shaped her own life, not caring so much for social approval.

She comes across as an intelligent, smart, thinking woman who is capable and who makes decisions for herself. She was a reigning superstar in the film industry. Now, I know that there’s been a lot of discussion of her friendship with Bal Thackeray. My answer to that is, tell me any famous person from the Bombay film business or cricketing fraternity that wasn’t friendly with Bal Thackeray. In that case, they then brought to their families—like the father and the Thackeray family were known to each other.

Other than that, in her personal life, she made friends with all kinds of people. Several Muslim artists, whether they were songwriters or music directors or actors, were extremely close friends of hers. She worked hard on learning Urdu, because she wanted to be perfect in Urdu. Even though when she spoke in her interviews, you could hear a Marathi person speaking, but then she sang, you would not be able to know. Her pronounciation was perfect.

It shows a respect for other languages other than your own. It shows a respect for other religions other than your own. It’s a kind of person that I think all Indians should be. In terms of her virtuosity, or in terms of her musical skills, you really need to know a little bit about music to appreciate how phenomenally talented she—and Asha Bhosle by the way, also—how phenomenally talented they are.

You take any song of theirs, try to break it down. Now you get these apps where you can slow down a song. I would challenge people to download those apps and play the song in. Then you will be blown away by what they do, what they introduced to those songs. This is not what the director has told them to do. A genius of that kind needs to be applauded.

RAJAGOPALAN: For me, I agree with your second point much more than your first, in the sense that the virtuosity is just unquestionable. Also just the riyaas [rehearsing], the practice, the commitment that has gone into honing this skill. That’s clear, you can’t last very long without it. Also, you can’t perform that well in that kind of technology. You have to be pitch perfect when you’re singing live with 70, 80, 100 musicians playing with you, and things like that.

For me, I feel like sometimes we make too much of Lata Mangeshkar’s—that syncretic, plural nature in the sense that she stood for—not because she was friends with Bal Thackeray. I feel like that comes more from the lyrics writers, but she’s the voice. For me, the incredible person there is Sahir Ludhianvi. He’s the syncretic face for me, or Gulzar or Kaifi Azmi.

I have somehow always been able to separate the two in my mind, but I feel like a little bit too much is made of confusing the identity of the people who wrote the words with her own identity. That’s almost a disservice to her because she’s her own person, whereas we’ve turned her into some kind of Indian melting pot.

The thing—all good things must be embodied in this one person, but is it actually embodied across very large number of people? You know, the harmony, the unity across caste, class, whether you’re singing about Hindus or Muslims or things about whatever is happening with the poor. That’s where I lie on that.

DESHPANDE: I agree with you. But I think in the current atmosphere that India is in, we’re seeing two kinds of things. One is particularly after she passed away or when she was alive, there’s this huge attack on her from the left, as it were, or from the woke crowd, calling her a Hindu nationalist and just dismissing everything about her. I don’t buy the arguments that are made.

RAJAGOPALAN: It’s also a lack of knowledge, frankly.

DESHPANDE: It’s lack of knowledge. Calling her the epitome of Brahminical privilege. Give me a break. You know, you need to know.

RAJAGOPALAN: She had to start earning for her family at 13. Come on. She’s not Brahmin. That’s the other thing.

DESHPANDE: Her father was. She’s the daughter of a second marriage. Her grandmother was a devadasi. She has a very strong devadasi lineage. Even when you’re showing awareness of these categories, you’re being completely blind about her.

Decline in Hindi Cinema

RAJAGOPALAN: That’s like, we’re imposing a current thought of a very strange kind on a past figure. I don’t care much about that.

I wanted to ask you, what’s happening with women in Hindi cinema? Is it a better lot for women today in the sense of roles? I’m not talking about women working in film. Are the female parts better today? Are they worse? Is it a mixed bag, in the sense there have been always some great female parts written in the past, there have been always some terrible female parts, and it’s the same? We always have had good good and good bad films.

I ask specifically because the music, I know, has declined. Both the music has more generally declined, but the number of female songs, the amount women speak or sing, the number of duets, it’s all gone downhill. When you listen to female playback singers these days, they’re lucky if they get four lines to sing in a duet. How do you see this picture evolving in front of your eyes?

DESHPANDE: Let me answer the second part first, which is, I think the big tragedy of Hindi cinema is the almost close to extinction of the Hindi film song. I think there are two reasons responsible for it. One is the globalization, unfortunately, which is because Western audiences, including Indian diaspora in the West, is used to shorter films, under two hours—the more realistic Hollywood genre of making film, non-musicals, unfortunately, has pervaded Hindi film sensibility, which didn’t use to be the case. One casualty has been the Hindi, as Paromita Vohra says now, the song appears at the end credits, almost as if you’re embarrassed to bring it out, like a flashy buaji [aunt] at a wedding you don’t want the wedding guests to see. It’s so true.

RAJAGOPALAN: I’m going to weddings just for the flashy buaji so I’m particularly disappointed.

DESHPANDE: Exactly. I’m hugely disappointed by this trend, because I think that the fake equation of “if a film has songs, it is not real” is completely misplaced, because every film is a genre. We see Superman movies. Do we believe Superman is real? No. But we believe in the genre, right? That’s what we’re going for. James Bond, for that matter, or any of these. There should have been a willingness to accept Hindi film songs through the genre that those films represent. Anyway, that’s a rant for another time.

I think the demise of the Hindi film song as a genre, it’s—relative decline of women versus men? There could be, but I doubt if that’s even a significant question anymore, because I think the song is dying. To me, that’s the biggest tragedy. If it’s an item number song, it could be the male voice, or it could be a female voice. If it’s rolling out at the end credits, it really doesn’t matter to me.

RAJAGOPALAN: I don’t buy the death of the musical as the explanation for the decline in quality. I’ll tell you why: because there are background scores. If you wanted quality music, you could build really great background scores, but people are not lip-syncing and it’s not a standard musical. But the background scores are also terrible. Maybe this is a very elitist opinion, but correct me if I’m going a weird way. I think it might be because the reach of Hindi films has increased.

Earlier, still a very small elite were watching Hindi films before the advent of television and the films being screened on television, which meant that the people who had access to other forms of entertainment, classical music and things like that, were the main patronage to also go watch Hindi film. As that group has expanded further and further, the music has now got more influences, which is a lovely thing, but it’s lower on the scale of virtuosity. There are fewer classical musicians who are part of the big orchestras, or there are few classical musicians who turn composers in Indian films, and stuff like that. Is that a good way to think about how the audience has changed?

DESHPANDE: No, I would say the opposite. I think the real problem is that the elite discovered Hindi films, but things were so much better when the elite was not watching Hindi films. We had songs. We had dances. We had fairytale romances. We had so much fun. Now everything has to be realistic, everything has to be short, everything has to be crisp. First, most actors are always abroad—characters in the movies, I mean, not actors. Why? Why do they always have to be abroad? Just because now we have a more elite, globalized audience that needs to see certain things that they’re familiar with. I would say the exact opposite.

RAJAGOPALAN: You’d say the opposite.

DESHPANDE: It’s exact opposite. The films of the ’50s, ’60s and ’70s were mass—the music was created for the masses even when it was heavily classically based. That was the beauty of the film industry of that time, which is that take a medium that is considered elite, but convert it into a complete mass form.

RAJAGOPALAN: A mass medium.

DESHPANDE: I credit Hindi film, the whole, the trinity of the music director, the lyricist and the singers for creating this absolutely magical genre of art that was completely mass based. I feel like if the elites had stuck to watching their Ray or whatever reason, whatever that would have been, we would have had at least our regular movies with songs. Now, when I see something like “Gangubai Kathiawadi” with actual songs, I say, “Oh, thank goodness. There are songs in this film.”

RAJAGOPALAN: The songs are not bad. Sometimes, when the songs are bad, you’re like, “Oh, good God, I wish they had made a serious film without any music altogether.”

DESHPANDE: That’s the part of the music. I’m actually thinking a lot about this. Some of my thoughts may not be very articulate, but I’m thinking a lot about—

RAJAGOPALAN: No. This is good.

DESHPANDE: —what has contributed to the demise of the Hindi film song as a genre. Maybe I’ll write something about it.

RAJAGOPALAN: That would be lovely. Actually, I’ve been thinking about something that maybe I’ll do a thread on or write something. I was thinking about philosophy in Indian lyrics. There’s a subgenre of cabaret songs and club songs and dance numbers, which used to be really philosophical. I’m talking about “Mud Mud Ke Na Dekh” or “Aage Bhi Jaane Na Tu” from “Waqt.” Shashikala is singing the song on stage, so, it’s literally that person, “Vamp” is singing these songs, or Nadira. That seems to have completely disappeared, where now the club is not a space for any philosophical discussion or living in the present or an existential question.

It’s just all objectification or love or pining or something. Once upon a time, even the Nautanki songs asked the existential question —like in “Seeta Aur Geeta,” the Nautonki song is “Arey Zindagi Hai Khel”; that’s the song which is asking the great question. I’ve been thinking about this a lot. Maybe I’ll write something on this.

DESHPANDE: That will be great. Even something like “Dil Ka Haal Sune Dilwala, Seedhi Si Baat Na Mirch Masala,” what happened in Maharashtra recently, when “varna pakadh lega policewala” [or the police will catch you] sort of fun song like that. Anyway, to come back very quickly to your first question about changing portrayal of women, I know that ostensibly, on the surface, it seems like portrayal of women is women are more independent, they’re smoking, drinking, wearing revealing clothes, having sex before marriage. All that is great. I think that reflects the mores of our time, but just because characters played by Waheeda Rehman and Meena Kumari and Madhubala, et cetera, wore saris or dressed in a more traditional way, it doesn’t mean that the characters were devoid of substance.

RAJAGOPALAN: Rosie [from “Guide”] is actually more independent than anyone else in modern Hindi film, right?

DESHPANDE: Yes. I’d written a short piece called “New Voices in Hindi Films Songs.” I really argue that, again, this is influenced by a Hollywood way of looking at Hindi films which is, if a woman is wearing trousers and skimpy tops, she’s liberated, whereas if she’s in a sari, she’s not. I’ve given many examples in some other pieces I’ve written on that.

RAJAGOPALAN: I don’t buy that at all. I’m totally on your side on this one. That’s a good note to end on, that all these things are complex. It’s not easy to paint anything with a single brush.

Ashwini, thank you so much for doing this. I have very long been an admirer of all your work. I’ve been reading your stuff for so long and waited to have this conversation with you. Thank you for doing this. This was such a pleasure.

DESHPANDE: Thank you, Shruti. This was wonderful, absolutely wonderful. Thank you.

About Ideas of India

Host Shruti Rajagopalan examines the academic ideas that can propel India forward. Subscribe in your favorite podcast app