Bhumi Purohit on Female Leaders and Bureaucratic Resistance

Shruti Rajagopalan and Bhumi Purohit discuss why female leaders are perceived as being less effective than male leaders in Indian bureaucracy

This episode is the fourth in a miniseries of weekly short episodes featuring young scholars entering the academic job market who discuss their latest research. In this episode, Shruti talks with Bhumi Purohit about her job market paper, “Bureaucratic Resistance Against Female Politicians: Evidence from Telangana, India.” They discuss female leaders’ access to networks, gender quotas, expanding women’s access to social and political capital, and much more. Purohit is a Ph.D. candidate in political science at the University of California, Berkeley. Her research focuses on comparative politics, political behavior and South Asia; in particular, she seeks to understand the behavioral and institutional barriers to women’s political representation.  

SHRUTI RAJAGOPALAN: Welcome to Ideas of India, a podcast where we examine academic ideas that can propel India forward. My name is Shruti Rajagopalan, and this is the 2021 job market series where I speak with young scholars entering the academic job market about their latest research on India. I spoke with Bhumi Purohit, a Ph.D. candidate in political science at the University of California, Berkley. She has a B.A. in public policy from Duke University and an MSc in contemporary India studies from University of Oxford.

We discussed her job market paper based on her dissertation-turned-book project titled “Laments of Getting Things Done: Bureaucratic Resistance Against Female Politicians in India.” We talked about the reasons for bureaucratic resistance against female sarpanches in Telangana, the importance of networks and social capital, gender quotas, state capacity and her next book project called “Bang for the Buck: Reforms to Maximize Public Funding Outcomes in India” (with Santhosh Mathew and Devesh Sharma).

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, Bhumi. Thank you so much for coming on the show. I’m thrilled to have you.

BHUMI PUROHIT: Thank you for having me, and thanks for creating this opportunity for job market candidates.

Bureaucratic Resistance Against Female Politicians

RAJAGOPALAN: You’re studying why there’s a perception that female politicians are less effective. One of the things that you find is that it’s really because of bureaucratic resistance. Bureaucrats are more likely to resist female politicians or just flat-out decline to help female politicians in policy implementations.

The mechanism, you explain, is that this is because bureaucrats believe that female politicians are going to be less important in their own career advancement in terms of promotions and so on. Of course, the perverse consequence of this is that this results in almost three times less help from bureaucrats for policy implementation when it comes to female politicians.

Can you just walk us through what is going on at the Panchayati-level female politicians who are elected? And what is the relationship between the bureaucrat and the female politician?

PUROHIT: Sure. The existing literature to date—the most seminal paper being the Chattopadhyay and Duflo paper that came out in 2004—shows that female politicians are better able to advance the needs of female citizens. But since then, there have been a variety of other papers that have come out that have mixed findings. They show that female politicians are not always as likely as men to be able to promote public-goods provision or get policies implemented in their constituency.

My dissertation-based book project is broadly looking at this puzzle, which is, when women achieve parity with men as elected officials, why do we see inequity in their ability to produce policy outcomes in some settings but not others? What I posit in my book is what I call the theory of bureaucratic resistance, as you said. What I mean by bureaucratic resistance is the outright refusal of bureaucrats to help women with policy implementation.

We can think of bureaucratic resistance as an extreme form of negative discretion. What I find is that, specifically, women are seen to be less instrumental for bureaucrats in helping them advance their own careers. The big assumption I’m making with my theory is that bureaucrats are seeking to maximize their ability to get desirable transfers. This assumption is based on a lot of existing literature that makes a similar assumption.

To study this theory, I look at the relationship between block development officers and sarpanches in rural governance in Telangana. I’m happy to go over later about why Telangana, but for those of you who are not familiar, BDOs are block development officers who are in charge of helping implement policy for a cluster of villages or Gram Panchayats in India. Sarpanches are elected to do the same, but at the village level.

They both are very codependent on each other. In that sense, bureaucrats are very reliant on sarpanches to implement policies. I confirm this through a survey of bureaucrats, where BDOs specifically say that the No. 1 person they go to when they’re thinking about how to implement a policy is the sarpanch. The sarpanch is similarly very reliant on the BDO because they consistently have to go to their office to get approval for projects, approval for funding, et cetera.

I thought this was a very useful dynamic to test the theory because both people are reliant on each other, but at the same time, they can’t really sanction the other person directly. That means that it reduces the risk of accidentally saying something negative to a researcher like me who they might fear will reveal things they might not want people to know.

I test the theory in Telangana through two surveys. One, I have an in-person survey of BDOs in southern and western Telangana. There, I partner with this fantastic research firm called Glocal Research that helped me implement the survey. I also partnered with the commissioner of Panchayati Raj and Rural Development Office to help me reach the BDOs. Through them, I have an almost complete census of 253 BDOs in the region of Telangana that I outlined earlier.

There, in the survey, I conduct various tests to show that bureaucrats have negative perceptions of women. Specifically, this has to do with three factors. One, they perceive women as less able to get work done than men as politicians. Two, they also perceive female sarpanches as less able to mobilize citizens. Three, they just have general negative implicit biases about women and leadership.

The next thing I wanted to see is if the bureaucrats’ biases lined up with women’s negative experiences with the bureaucracy. To do that, I conducted a phone survey of about 1,000 sarpanches in the same district as the BDO survey. I asked the sarpanches to reflect on the last few times they went to the BDO offices to receive help, and I said, “How often does the BDO refuse to help?”

There, after controlling for variables like dynastic status, proxy status of the sarpanch, socioeconomic status, et cetera, I find that women still report facing significantly more bureaucratic resistance than men, even though they go to the BDO with the same issues as the male sarpanches do. The explanation I find for this—one potential explanation is the size and quality of higher-level networks that women have with MLAs or district collectors, that perhaps men have access to much higher-quality networks than women do.

How Networks Affect Female Leaders

RAJAGOPALAN: This network part is what I want to dig into a little bit. There could be two mechanisms at play. Maybe you can highlight which one is stronger or if both are salient. One is, it seems like women politicians—in this case, the sarpanches—have lesser social capital or political capital that they can use to mobilize resources to get their work done. That seems to be one part of the story.

This is, of course, because women’s reservation—the constitutionally mandated women’s reservation at the sarpanch level—is relatively new within the Indian republic. This cohort of female politicians at that village level—it’s just about 25 years old in India, which is just a minute in terms of long-term democratic processes.

The second part is, it seems like there’s something going on, where the BDOs are . . . One would imagine that their transfers and promotions depend on how well they implement projects. But it seems that there’s something else at play, some kind of other mechanism within the bureaucracy, such that, even to their own detriment of not fulfilling their mandate and actually helping policy implementation, they are still able to get ahead without any problems.

There seems to be two different network effects going on. One is they seem to have some network which the men almost overcompensate or compensate for lack of implementation or help given to female politicians. For the women, there seems to be no network or social capital that they can rely on.

PUROHIT: I think both are at play. When we think about what BDOs do in their day-to-day job—anyone who’s familiar with the bureaucracy would know this. Devesh Kapur and Aditya Dasgupta had a project on this, a paper that came out recently that talks about bureaucratic overload, where bureaucrats have far more responsibility than they’re able to fulfill with regards to the time or resources or staff that they have available.

One thing going on here is that BDOs may not necessarily be working against themselves or working to their own detriment when they’re deciding to help some politicians over others. What I think is more likely is they’re actually using heuristics based on identity to make decisions on who to help and when, and when helping somebody is beneficial to them versus when it is costly.

When we think about it from this sense, that bureaucrats are working under resource constraints, then we can see why, based on their priors about women’s competence or women’s leadership ability, it might generally be beneficial or rational, rather, for them to focus on helping men.

To speak to the other aspect of it, I think it’s helpful to think about a shadow case that I study in my job market paper, which is to think about Dalit and Adivasi politicians who are also elected through quotas. What I find there is really interesting. I find that BDOs also think that Dalit and Adivasi sarpanches are not as competent as other caste or backward caste sarpanches. At the same time, Dalit or Adivasi sarpanches don’t report facing any different bureaucratic resistance than their counterparts.

The explanation that I have for women is that they face bureaucratic resistance because they don’t have access to higher-level politicians and district officials who can then sanction the BDO for resisting. But what I find, conversely, with Dalit and Adivasi politicians is that they have those higher-level networks. That, I think, is partly because there are these quotas in Telangana at the MLA level, based on caste.

There are officials elected, based on caste, at higher levels. There are also, therefore, strong incentives for sarpanches to mobilize based on caste. They’re seen as really good mobilizers by bureaucrats and BDOs, consequently. I think that’s a really good example to think about in terms of why women may be falling behind in terms of gaining the political and social capital that you mentioned.

Effects of Weak State Capacity

RAJAGOPALAN: A big part of this is that, given weak state capacity, it becomes easier to engage in these biases or discriminatory behavior. Because given weak state capacity, there are going to be projects that BDOs are not going to be able to implement. It is difficult to pull them up for it. It’s not laziness or lack of work. It’s just a question of, on the margin, who they are choosing to help and who they’re not choosing to help. In this case, it’s the women who they are least likely to help for that reason. If we improved overall state capacity, some of this should go away?

PUROHIT: Partly, but I don’t think that is likely without access to political capital because, if you think about the relationship, for instance, between an MLA and a BDO, even though the BDO might have negative perceptions of female MLAs, they cannot in theory act on these biases with an MLA because the MLA has influence over their transfers. They can influence and sanction the BDO for exercising discretion. In that sense, I think that part is also really key.

I do want to highlight that I think Telangana is actually, perhaps unfortunately, a better situation that I’m examining in the sense that in Telangana and Andhra, they have a mandal parishad system, which is very similar to blocks in the rest of the country. What it means is that, on average, each BDO is only directly working with about 24 sarpanches at a time. When you go to other states the average . . . The national average, for instance, is each BDO works with approximately 90 sarpanches.

Gender Quotas

RAJAGOPALAN: Wow. This is only likely to be worse in other states. Telangana is almost the best-case scenario, which is frightening given the kind of results that you find, right? As the follow-up, now, given that the big problem here is this access to networks and social capital and political capital at higher levels, do you think gender quotas at the state-assembly level and at union-parliament level are going to help mitigate this problem?

This has been on the agenda for a really long time. We’ve had various women’s reservation bills—33% quota for women pop in and pop out of our national imagination. But the reason I asked the question is, so far, we don’t have very good data on whether women vote as a group in the same way that Dalits and Adivasis and their caste subcategories vote as a group.

Are we likely to see women’s reservations at higher level mitigate this network effect, and some kind of girls’ club will form the way we see it? Or do you think that’s not the point? It’s never going to be as cohesive as the caste relationship that a male sarpanch from a particular caste will have with a male MLA of the same caste?

PUROHIT: I think partly this has to do with the relationship between power and structures. Specifically speaking, quotas existing, for instance, for Dalits at the MLA, at the state level, create an incentive structure for individuals to actually focus on the substantive needs of Dalit citizens. And it incentivizes parties to give tickets, for instance, at lower levels, to Dalits because they see an opportunity for these lower-level politicians to rise up in politics.

Even if they’re not rising up, they see that as an opportunity for them to locally mobilize on behalf of the party. Even if women are into voting as a block for female politicians, I think, just by virtue of providing women access to higher-level office through reservations—because we don’t see it naturally happening despite lower-level quotas—there needs to be a mechanism, like higher-level quotas, to get women elected at higher levels.

Just by the fact that women are then able to advance into higher-level politics through the help of quotas means there’s an incentive for parties to take women seriously at the local level, like the Gram Panchayat, because there’s a chance that the woman who’s running at the sarpanch level, and then doing well, can then help them mobilize others, whether it be men or women. Or there’s a chance that the woman can then rise up at the MLA level.

I think that would then enforce these structural incentives for parties to take women more seriously and thereby give women more political capital to also circumvent bureaucratic resistance.

At the same time, I think, to your other point, there is something going on here just about informal relationship at networks. I see that when I sit in a bureaucrat’s office. A male sarpanch can just come in and talk to a male BDO easily. They can shoot the breeze, but if a woman does that, there are cultural barriers that make her seem gossipy. There are these stereotypes associated with women doing the same thing. They can’t have the same informal networks that men can.

I think it’s both, but we can’t change stereotypes as easily. We can’t change biases as easily. Those are ingrained. What we can change are the structural incentives for parties and for these institutions to take women seriously.

RAJAGOPALAN: One way of going about it is the mechanism that you described, which is a higher-level MLA who can then punish the BDO for not helping out a female politician at the lower level. But on the other hand, do you think just gender quotas at the BDO level would fix the problem more easily? This problem of girl talk and female perception of being able to get work done—do you think that’ll ease up if we just directly attack the BDO level, and that’s where we should focus our energies?

PUROHIT: Yes, that’s a great question, and that’s something I thought about when I was designing the project, quite seriously, because I wondered if this was merely a problem of male-led patriarchy or if this was something more structural about the BDOs’ career incentives, and therefore who they think will help them advance those incentives.

I did dissect my analysis by the gender of the BDO, and surprisingly, I find that—or perhaps not surprisingly to some—I find that female BDOs are just as biased about female sarpanches as male BDOs. We don’t really see females sarpanches being treated any better by female BDOs.

I think that has to do with the way power mediates the way we see the world. There is a great theory, called system justification theory, that argues that the higher up you go in the power structures, the more you accept some of the underlying inequities, the more you’re likely to justify the power structures you see.

In this case, a lot of the BDOs are so different demographically from the sarpanches. The female BDOs have been studying for so long to pass these exams and get elected, and they’re likely from really well-off socioeconomic backgrounds. Whereas the sarpanches don’t always have that background, so I don’t think there’s the same level of collegiality and sameness they feel when they’re interacting with female sarpanches.

Quotas and Biases

RAJAGOPALAN: This is really interesting. I was speaking about this with Alice Evans. One of the things I’ve always worried about is that quota candidates are treated differently than regular candidates, even though we know that exams—whether it is medicine, engineering, IAS exams or the services—they are so competitive that no matter which category you belong to, you have to be really excellent to get through these exams.

There’s always this perception that if someone is a quota candidate, it must have been a handout, and they must not be deserving. Do you think, in fact, it is the gender and the caste quotas at the Panchayati Raj level which are promoting some of these biases or making it easier to engage in these biases? Or do you think it would have been, irrespective?

PUROHIT: I think some of the mechanisms I see don’t necessarily have to do with quotas. If it was just about the quota story, for instance, we wouldn’t see Dalit and Adivasis sarpanches not facing bureaucratic resistance because most of those sarpanches are also elected through quotas. If it was just a quota story, I could dissect women who were elected through quotas and the 5% who weren’t in my sample, and see that those 5% of women are treated just as well as the men, which I do not see.

I think there is some aspect to it that has to do with gender, but I also want to be mindful of claiming that it’s just negative biases in general because one thing I do is, I test for general negative biases, which is done through something called the Implicit Association Test. We test psychologically, at the subconscious level, how quickly do bureaucrats associate the idea of leadership with gender? I find that, on average, women and leadership are two concepts that they’re biased against.

They think of women as followers. They don’t think of women as leaders. This has to do regardless of quotas—that is the case. Just general negative biases don’t associate with bureaucrats thinking negatively of the actual female sarpanches they work with.

What we do see is specific biases that have to do with skills that help bureaucrats advance their careers—the ones I mentioned specifically being networks, mobilization of citizens and ability to get work done. Those are the factors that are driving bureaucratic resistance, not just general negative bias.

RAJAGOPALAN: It’s interesting. If there was some change in the incentive and transfer-and-promotion structure for BDOs, saying the number of females sarpanches you work with, you get extra points or something, that ought to immediately have some impact on this. Because what you’re basically saying is, on the margin, it’s cheaper to indulge their biases against women than it is of any other group or category, quotas or not. Some simple, literally human-resources incentive structure ought to be able to get at this a little bit more easily.

Expanding Women’s Political and Social Capital

PUROHIT: The other possibility is just expanding women’s ability to navigate the bureaucracy or expand their political capital. That may not necessarily have to do with implementing quotas at the higher level or helping them build relationships at the higher level, but one other possibility that is part of my larger dissertation project is something I’m working on with Rachel Brulé and Simon Chauchard, two political scientists, and Alyssa Heinze, who just joined as a Ph.D. candidate at Berkeley.

In particular, we’re interested in these solidarity networks that women can form amongst themselves. We want to pair experienced female sarpanches with inexperienced, first-time female sarpanches, and see if having those mentorship groups that meet regularly and help them navigate the challenges—not just with the bureaucracy, but with governance in general—can expand women’s political and social capital.

To give you a very concrete example of how this might happen, one big issue bureaucrat female sarpanches often told me about when I was in the field talking to them is, they are forced to go to the BDO’s office with their husband, often because they don’t have alternate means of transport.

On the other hand, if you imagine an alternate world where a group of women get together from neighboring villages and say, “Let’s just go together,” perhaps that changes the dynamics in the way the BDO interacts with these women and how seriously he takes a group of women versus a woman coming in with her husband. We’re optimistic about the potential of solidarity networks, and we’re eager to test what happens with mentorship groups once the pandemic subsides a little bit.

RAJAGOPALAN: I’m very excited to learn more about this project and this research. We see a lot of this within the academy, like women being able to form these mentorship relationships with other women, or upperclassmen in Ph.D. programs with those who are just entering the Ph.D. program, and so on. Anecdotally at least, it seems to have a lot of impact on the margin.

I would be very curious to see if it’s true at the village-politics level because there are so many other factors, like you said, such as family background and education levels, even for women politicians versus bureaucrats, caste, and so on and so forth.

You also have another book project that you’ve been working on, which is on the public-finance management system. Can you tell us a little bit about that?

Research on Public-Service Delivery

PUROHIT: Sure. My first research agenda focuses on women’s political participation, but my second research agenda is very much focused on public-service delivery. In a joint co-authored project with Devesh Sharma and Santhosh Mathew, who are both at the Gates Foundation, we’re working on finalizing a book manuscript that’s been accepted by Oxford University Press. The book is called “Bang for the Buck.” What we do is, we dive into India’s public-finance system to show how intricately the system is linked to public-service delivery outcomes.

We also outline the way its poor design has exacerbated issues of public-service delivery and how it’s long overdue for reform. And based on existing research, we provide some potential policy reform suggestions.

To summarize, we essentially find that the way the funds flow in government, combined with the lack of transparency and really poor data collection on where funds are located at any particular time, create these really perverse incentives for corruption, for fund leakages. And they cause long delays that citizens who rely on subsidies or businesses who signed contracts with government are likely very familiar with.

RAJAGOPALAN: There’s one really interesting overlap between the project that you’ve been working on for your dissertation and the public-finance management system, which is federalism at the local governance level. That seems to be the commonality.

Now, within politics, we seem to have devolved and decentralized rather well when it comes to electing through participatory democracy at that particular level. But on the other extreme, which is the fiscal side of things, we have not devolved at all, unless each state, like Kerala, for instance, has chosen to devolve a large proportion of the state finances to the local level, but most states are lagging. They’re quite far behind.

How much does that play with the dysfunction of the public-finance delivery system, and therefore the public-service delivery system?

PUROHIT: Fiscal federalism is definitely part of the problem. Part of the issue is, rather than central or state governments transferring funds directly to the implementing body, like the Gram Panchayat, funds go through various layers in government, which of course creates incentives for leakages.

Without going into the weeds a bit, if five people need to sign off before a Gram Panchayat receives funds for a program, then there’s a structural incentive for those five people to benefit from those funds and maybe take off a little bit before they’re passed on.

More importantly, I think it’s not just about fiscal federalism, but also about transparency in where funds are at a given time, and what is causing delays. For instance, there was this RCT that incentivized a policy reform with MNREGA, where funds are now sent directly to the implementing body. But there isn’t always full transparency, and there isn’t always full control at the local level in the information they get about why funds are delayed, for instance, or when they’re expected to arrive. When those issues happen, it’s not fully solving the system.

One of the arguments we make in the book is, it’s not just about ensuring that funds can be allocated at the implementing level, but it’s also about creating transparency systems that allow citizens to know what the status of the funds are, what they’re due to be owed, where they can go obtain them. Transparency also allows NGOs and social service actors to help citizens navigate these really complex bureaucracies that aren’t necessarily designed with citizens in mind.

That’s one of the things that we really focus on in the book—not just thinking about where funds go and who decides allocation of funds to begin with, or budget allocations to begin with, but also thinking more broadly at the bigger picture about last-mile delivery, who knows how much about public funding, and if they’re able to easily access funds when they actually reach the end implementing body.

RAJAGOPALAN: In terms of devolution, of course, one is the transfer of funds that you talked about, but the other aspect of fiscal federalism and devolution is entire categories of revenue-raising capacities that can be transferred to the Gram Panchayat level or the urban local body. This is developing their own capacity to raise their own revenues, in which case you don’t have to do this huge navigation of where the funds are.

The funds have been raised by the same people that they are effectively catering to in terms of public-service delivery. As we know from a big literature in public finance, that’s always the tightest link. If your taxpayers and your voters are the same group of people, that’s really when you have the tightest feedback mechanism in terms of governance and service delivery.

Pandemic Activities

RAJAGOPALAN: On a personal note, what have you been up to during the pandemic? I know you’ve been busy wrapping up your dissertation and your job market paper and so on, but what else?

PUROHIT: The pandemic’s been really fun to us in terms of having a lot more free time. One of the things I’ve been doing to get my mind off the dissertation and other work is diving into some of the recipes that I’ve always wanted to learn from my grandma or my mother that I never knew.

RAJAGOPALAN: I have done the same, so tell me more.

PUROHIT: I’ve been learning all sorts of Gujarati recipes and finding creative ways with some of the ingredients being missing out here in America. I’ve been learning how to make paatra and undhiyu and all these other delicious things that I’m always craving, but I can’t easily find in stores.

What about you, what have you been making?

RAJAGOPALAN: Well sort of everything. One of the things I started doing during the pandemic is making my own spices. Until now, my mom would visit or my mother-in-law, or I would visit them in India, and I had my steady stream of masalas—homemade masalas and pickles and things like that—that would come and go. Not seeing my parents for 18 months and not having visited India has depleted supplies. I have finally put on my big girl pants and said, “I’m going to learn how to do this myself.”

PUROHIT: There’s no greater joy than making your own spices.

RAJAGOPALAN: Yes. Now I’ve started making my own spices, my most recent high point of my cooking career, I must say, was making my own milagaipodi, which is what we eat with the idlis and dosas and things like that. It’s homemade. I figured that one out, so I felt particularly good. My mother likes to give fun recipes, things like “Roast the red chilies and take them off the pan just before they burn.” Until you burn them once, you don’t know what the point of “just before they burn” is.

What have you been binge-watching?

PUROHIT: I have actually been binge-reading these days. I love binge-watching television, but the last two weeks, I’ve been getting into some really good books. I started reading “Minor Feelings” by Cathy Park Hong, which was a great way to just learn the history of the Asian American experience that I had no idea about.

Then more recently, I picked up “Gold Diggers” by Sanjena Sathian, and that was a really fun, creative way to think about the striving that Asian Americans do and Indian Americans do growing up here, and the pressures they face, but it was a really fun read.

RAJAGOPALAN: Thank you for those recommendations, and thank you for doing this. This was such a pleasure.

PUROHIT: Thank you, Shruti. Thanks for having me.

RAJAGOPALAN: Thanks for listening to Ideas of India. If you enjoy this podcast, please help us grow by sharing with like-minded friends. You can connect with me on Twitter @srajagopalan. In the coming weeks we will feature weekly short episodes with young scholars entering the academic job market discussing their latest research.

Also check out our new initiative commemorating 30 years of India’s market reforms at The 1991 Project is an effort to revive the discourse on growth-centered economic reforms in India by focusing on the economic ideas that drove them. In the coming months, we will publish essays, data visualizations, oral histories, podcasts and policy papers demystifying the Indian economy and the 1991 reforms. You can see all the content and subscribe to our newsletter for updates at

About Ideas of India

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