The Story Rules Podcast E20: Rukmini S – Understanding India through data (Transcript)

Rukmini S Cover art v1
5. General

The Story Rules Podcast E20: Rukmini S – Understanding India through data (Transcript)

This transcript has been created using a combination of AI transcription tools and (some painstaking) human effort. Please excuse any typos, grammatical mistakes, inaccurate time stamps, or other errors. Specifically, the time stamps would not account for the intro portion of the podcast.

You may share portions of the transcript with due credit. Enjoy!

And, if you find this content useful, it would be great if you could leave a review on your preferred podcast platform.

Intro Hook:

“I think if there is this response, and if there is outrage to data by people who feel that it doesn’t accurately reflect reality then instead of reflexively blaming those people for being oblivious or choosing to not accept what reality is, I do think it reflects on a failure of those communicating the data. If the process and the methodology for collecting the data isn’t clearly explained, it becomes that much easier to bat that data away. I find that the best estimates we have on consumption expenditure which show that in 2017-18, if you spent more than 8,500 Rupees a month as an individual it puts you in the top 5% of urban India. That’s something that people often push back against, find unbelievable, or want to say that “There are so many people I know who are spending this, that; you can’t tell me I’m in the top 1% of the country”, (I think that it) comes from us not being able to properly explain how these numbers were calculated and what goes behind it. Explaining to people “Is your objection that this is being missed? Let me show you how it is actually captured in the data,” “There’s all this black money – let me tell you how consumption doesn’t mean that black money isn’t captured.” I completely believe that there’s a failure of communication rather than a refusal to accept reality on people’s parts.

Welcome to the Story Rules podcast with me, Ravishankar Iyer, where we learn from some of the best storytellers in the world, find their story, and unearth the secrets of their craft.

Today we speak with Rukmini S, eminent data journalist and author of ‘Whole Numbers and Half Truths’ a seminal book on data in India.

Quick – answer whether these questions are True or False:

1. Delhi has the highest rate of crimes against women in India

2. Most of India’s migration is rural-urban 

3. UP is safer for women than many big states as per an NCRB report

4. India has a large middle class

5. You are a part of that middle class

If you answered ‘True’ for any of the questions above, you need to read Rukmini’s book.

I teach how to craft narratives with data and one of the things I used to take for granted was the ‘data’ part. Earlier on the podcast, when I had interviewed Brent Dykes’ (author of ‘Effective Data Storytelling’), we had discussed about the importance of ensuring that the data part of the equation is tied up and not taken for granted.

And if you aren’t rigorous about getting the right data, you end up with narratives which may be divorced from the truth. Just like those True/False statements above.

But if those statements are not true, then what is the truth?

As per Rukmini’s book, the answer is, um, complicated. In a series of ten illuminating chapters she covers a wide range of topics about India – from crime to education to income, to eating habits to how we vote and how we fall ill – and deftly unveils a truer picture of India.

In this conversation, Rukmini shares her approach to researching and writing the chapters. She shares her productivity approach of getting writing done during the pandemic despite being a parent to two young kids! She gives her nuanced take on India’s unique data architecture and why it’s in all our interests to safeguard and nurture it. Finally she also shares the need for better communication and storytelling of data findings, especially if they seem to contradict the audience’s perception based on lived reality.

Let’s dive in.

Ravi (0:21)

Hi, Rukmini. Welcome to the Story Rules podcast!

Rukmini (0:24)

Thank you for having me!



Normally, when I start talking about a book that has especially captured my attention so much, I would ask the author questions about the book’s contents first and then go to the writing approach, style, techniques, etc. In this case though, I plan to reverse the order. For two reasons: one, I feel that if we go into the content, two hours would not be enough. There’s so much to chew in. But more importantly, I want to pick your brains on the writing style and approach. To me, that’s quite valuable.

Coming to the choice of writing the book, I remember you talking to Amit Varma on his podcast, where he goaded you and you had encouragement from Pratap Bhanu Mehta also. Which is great, I appreciate that the push came. But I’m surprised as to why the push was needed because if you look at the breadth and depth that this book has – and for listeners who have not read the book, I urge you to pick it up – it’s as relevant now as it was when it came out one and a half years ago, if you just read the chapter titles; it’s a primer on how India is based on data; How India Tangles with Cops and Courts; What India Thinks, Feels, and Believes; How India (Really) Votes; How India Spends Its Money; How India Works. Just looking at the chapter titles, it’s so ambitious and each chapter’s content does full justice to its title. This is turning out to be a long context, but I just wanted to put it out there. I’ve worked with Indian statistical data sources in my consulting years and even later. It’s not that I was completely lacking knowledge about India’s stats and yet, every study that you were giving in every section of every chapter was giving me very surprising TIL moments, “Oh my God, I didn’t know this!”

Now, coming to my question…why is it that, when you look at a country like India which is 1/6th of humanity, why don’t we have more books like this? I’m not even talking about the fact that this book is written so well…that’s a secondary, bonus benefit; but why don’t we have this covered at all? This, to me, was the biggest surprise. Why is it that this is the first time I’m reading such a book? That’s my first question to you, and related to that is: if this is what I found most surprising, when you have been talking about this book over the last one and a half years, what about the reaction of various people – the audience, reviewers, etc. – has surprised you the most?

Rukmini (3:18)

Thank you for your kind words about the book, and I’m glad that you felt the content in the chapters lived up to the ambitious chapter titles. I did feel like I was packing a lot in, and I’m glad if it comes off as substantial in that way.

My answer to the 2 or 3 elements to your first question tie in together, which is, how is it that writing the book hadn’t occurred to me and how is it that we haven’t had this sort of book so far? I think it’s a combination of something that comes from a place of relative modesty, and immodesty. Let me give you the more modest one first: As a career, full-time daily news journalist, I never had the time to pull back, stop, and think. I never took a break. The only breaks I took were maternity leave for my children, which doesn’t really lend itself well to dreamy thinking while looking out of the window. I never stopped to pull back and see how all of this fit into a larger theme. If, at a dinner party, you asked me whether I had a grand narrative about India, I would have said no. I possibly would still say no, but I do think that I was not able to see the wood for the trees all this while because I had never stopped to think about it. When I mentioned that both Pratap Bhanu Mehta’s and Amit Varma’s taking of my views with such seriousness pushed me towards the book, it’s not flattery – it really is that I hadn’t stopped to think about it as, “Oh, I am pointing out multiple facets that add up to some sort of coherent India story.” I hadn’t seen it that way. That is the relatively modest part.

The relatively immodest part is something that I think many journalists, if they haven’t already had this reckoning, should have this reckoning now. Which is that we, as journalists, feel that we have said everything. When I started putting the book together and I was talking to my agent, Anish Chandy, of Labyrinth agency, and then to Westland’s publisher Karthika – who was the editor for my book – I so often felt that it wasn’t worth mentioning this or that because it was already known. It isn’t just that it was unknown and I was revealing these great truths, but it was also a moment of realization – not a moment, it’s been a continuing journey since then, of realizing – that a lot of what we do as journalists disappears into the aether, partly out of the format that we do it in, and partly out of the fact that we’re bombarding people with information, and also because it doesn’t end up aggregating very much. While we feel that there are moments in time where we’ve made all of these points, I believe that we, as data journalists, haven’t truly advanced public understanding commensurate to the work we’ve been doing. In a sense, the book ended up being that: it ended up being a moment of stopping to think of the points that I had, in part, made over time; coming to see new directions that I wanted to take this in; and aggregating knowledge that – as you’ll see by the number of citations and references in the book, have been made by other people – had not been pulled together in this fashion. If this makes it sound like a compendium of sorts, I don’t think that is the case. These sorts of connections between two different studies or news articles had not been made. I do think that that is something I’m offering anew, but I think the reason they were not made before is because we are often single-source journalists as news reporters, because the scope of what we do is so limited in size and ambition. In literal ways, it is 450 words, and what that does is it curtails your thinking and ambition around it. You do not make connections between two disparate sources of data, let alone two disparate schools of thought, because that’s all the mandate you have to say in one go. I think this is why I hadn’t thought of this book until now, and perhaps partly why we haven’t had this book so far, because some people feel that it’s been said already, and the ones who don’t know about it have perhaps just missed all of the information that most journalists feel that they have already said beyond anything that needs saying.

Ravi (8:40)

Great points, Rukmini. I think Amit had alluded to this concept, ‘the curse of knowledge’, which is something that I face every time and I’ve seen others struggle with it, too. This happens all the time; senior people in a meeting might say something with the assumption that everybody knows it, but you don’t know what others don’t know. It’s important to keep that in mind.

Rukmini (9:02)

Just to add, when we feel that we’ve already said something and people don’t know it, or “how do they not know it?” Or, “shouldn’t they know it already?” I think what that is, is a commentary on the way we have communicated it in some way. One of the big reasons that people don’t understand a lot of Indian statistics is because it’s explained in a way that’s almost designed to not really get through to people. Part of this belief comes from my own experience, and one of the reasons that writing around Indian data has been so poor, historically, is because many of the people writing don’t fully understand what they’re talking about. I do think that perhaps the majority of people who write about the GDP would not be able to explain to a layperson exactly how the GDP is calculated. This doesn’t mean that they are faking it, or swinging by without actually knowing what they’re talking about, but they wouldn’t be able to explain in absolute, clear detail, how some of these things are calculated. I know from experience that the things that I was least certain about, in terms of data, are the ones I did the poorest job of explaining when I was doing journalistic writing as well. I remember at one point stopping to write a piece about how opinion polls were conducted…I enjoyed writing it, and from then on, everything made a lot of sense to me. But I should, in all humility, think back about whether, until then, the writing I had done around opinion polls was perhaps without the benefit of knowing how opinion polls were conducted. It’s not just that people assume that others have knowledge, it’s also that if that knowledge exists or has been shared, it’s often shared in ways that don’t really get through to people.

Ravi (11:13)


There are a couple of insights that are coming to me. There might be others like you, Rukmini, who, by the strength of the work they’ve done over the years – if not decades, have in them the ability (for want of a better word) to write a book about it. I think there is value in just consolidating and creating a compendium – as you were saying, not just a compendium…something where you step back and find connections or distil patterns, figure out overarching stories and narratives that might be hidden. I think there’s a lot of value in that. Yet, to your point, maybe some people may not know where they are good or bad, and having external validation might help. Especially if somebody tells you that you’ve got it in you to share it with the world. I hope that what you’ve written will trigger more people to think about that, look out for validation, and Pramit (Bhattacharya) – if you’re listening, I think you should be writing a book! I’m sure there are many others you can think of who probably have books in them, but are, for whatever reason, not (writing them).

(Was there) anything that surprised you in your journey (as you were) talking about the book, (such as) the reactions, etc.?

Rukmini (12:37)

Yes, a couple of things surprised me. One is that I was taken aback by some of the responses to the chapter on people’s thoughts and beliefs. Again, because I thought it was quite widely known. Broadly, data has shown us that the vast majority of Indians do not hold what would be called progressive views on things like intermingling/inter-group mixing. This would include things like inter-caste and inter-religious marriage and similar issues. I remember feeling even while writing the book that it was surprising that I was still making this point anew. Yet again, I was surprised by how surprised people were by those parts of the book.

In terms of broader issues that the book brought up, there were two key ones that surprised me. One, I was surprised by the extent of disbelief that a lot of young people had in official numbers; the widespread sense among a lot of young people that numbers in Indian official statistics are so poor or so badly manipulated that they are unusable, which is something that I disagree with for very specific reasons and I enjoy talking about this to young people because I do want more people to look at data. The more people look at data, the more (likely it will be to spot) when there is a specific problem as it comes out, and then it will be fixable. That’s one (point).

The other thing I’ve been very surprised by is just how much misinformation worries people. People worry about older people being taken in by a standard WhatsApp forward, that sort of thing…and for themselves, the younger people worry about not knowing what is credible information and how to distinguish between what’s credible and what isn’t. As well as how they will ever fight or win the battle for credibility when there always seem to be multiple sources of “truth” available; “How can I make my point with data convincing, when someone else will just go and pick another source of data?” is something that comes up a lot and which essentially drove me to write my second book. I had intended for my second book to be something else, but I was so taken aback by how much misinformation bothers people that I decided to pivot to that for the second book.

Ravi (15:17)

I love the title of the book, which is ‘How to Talk to Your WhatsApp Uncle and Other Essential Tools for Survival’, coming in late 2023?

Rukmini (15:25)

Yes, that’s right. Which reminds me, I’ve got a draft I need to work on.

Ravi (15:31)

Haha! Looking forward to that, Rukmini.

I definitely want to tap into some of the thoughts you (put forth) regarding data and its sanctity, trust, credibility. But I want to come back to the writing process itself.

Let’s start with the research process which precedes the writing one. When you are going out to try and answer a few questions in your mind like how much crime is really there? Or how much do Indians really earn? What do they spend? etc., there is this dichotomy or this tussle that I’ve always faced between having a hypothesis about the data that you want to research – which is useful, (because) if you don’t have any hypothesis then it’s just overwhelming (to figure out) what questions you want to answer and even if you get some insight, is it surprising or not? If you have a hypothesis, you can compare against it. That’s one; the other side is that if you go in with a hypothesis then confirmation bias kicks in and you only look for data (pertaining to that hypothesis). How do you balance this challenge out?

Rukmini (16:46)

I think I go in with questions rather than hypotheses. I know that that is the foundation of a hypothesis, and I don’t have complete knowledge on most things, so I go in with questions because I do not know the answer to those things. I was interested in what research showed on the likelihood of crime being unreported. The universe of official statistics means that the only numbers that are there are crimes that were reported, while all of us know anecdotally that a lot of crime goes unreported. Of course, you could go in with a hypothesis that the crimes against women which we would imagine in India tend to carry with them social stigma, the notions of shame and honour, would be far more likely to be under-reported. Instead, I tried to go in with the question of what crimes are more likely to be unreported without bringing in priors. I truly think this comes out of a sense of “I don’t know the universe of what crime in India looks like”; I have an interest in understanding sexual crime because that is an area I’m interested in, but I wouldn’t have any pretentions of understanding what the universe of crime in India looks like. It’s not something I’ve studied, it’s not something there’s good research on. So, it really was a question rather than a hypothesis there and what I ended up finding by looking at some studies that are referenced in the book, is that the crimes most likely to be under-reported are those involving theft and property, and not violent crimes to the body. This might not always be replicable across contexts, and might not universally be the case, but this direction comes from going in with an open question. Otherwise, what I would have done is to try and purely look at studies that look at the extent of under-reporting of crimes against women in different contexts. This helps you zoom further and further out. I think that helps people to really understand what a universe looks like rather than a subset of what you’re looking at. I do think that is what I typically go in with.

This is similarly the case with COVID as well, which is something that I was flying blind with when I was reporting, not having a background in either health or science reporting. What I did in that particular case, which preceded the book, was that I started a nightly podcast in the early days of the pandemic, literally just for me. I recorded on my phone and put it out on SoundCloud without any publishing aims at all, and every day (there) was an open-ended question to which I truly did not have an answer. The more we are able to be upfront with ourselves and with other people about not knowing answers, and going in with a genuine spirit of questioning – the more we are able to come back with a broad set of answers rather than things that either confirm or do not confirm our priors, but are still very much based around our priors. I think that just makes for a much more interesting and wider scope of enquiry.

Ravi (20:22)

That makes sense. Especially when you go into a topic where you may have some sense (of the subject) but you are being quite modest and saying, “I know very little, so let me go in with as open a mind as possible.”

Just to push on that, Rukmini…let’s say you have found something in a geography and let’s take the stupendous research you’ve done on violent crimes against women by studying several hundred court cases, and coming to the finding that in a majority of the cases there was actually no crime committed as per the (relevant) section; it was the parents putting out that story to separate their children who had gotten married out of the family. Let’s say you’ve found that in one geography – let’s say Delhi or Mumbai. And for whatever reason, say you’re going to do a similar exercise in Chennai or Telangana…you would go with a prior, then. Would you agree? Or would you (say) you have a different approach?

Rukmini (21:37)

Let me be honest that if I had gone in, in that case – because I did the same exercise for Mumbai and Madhya Pradesh, in this particular example – if I had gone in and found that it wasn’t the same pattern, I think it’s a good thought exercise to think about what I would have done. Would I have reported null findings? As people in the sciences are encouraged to do? is a very valuable thing in the sciences! I can’t promise that I would have gone to a newspaper and said, “I want to write a story and this is what I found in Delhi, and this is what I didn’t find in Mumbai.” I should be honest about that. But I did go in uncertain that I would find the same thing. For one, you do know that there are cultural differences between different parts of the country so it seemed entirely possible to me that this violent and heavy-handed criminalization of the sexual agency of young people could be something that parents in Mumbai might not want to pursue. It was a possibility, so I wasn’t certain I would find the same thing. Similarly, I thought that perhaps the expression of free will and free sexual agency might not be something that I would see in rural Madhya Pradesh, so I was prepared to not find these cases. But as it happened, it ended up being a fairly universal finding and then much more rigorous academic research by others after me has shown that particularly in the case of teenagers…when it comes to cases in the POCSO Act…that’s really something people continue to go on finding. I do take the point that it would be interesting to think of what I’d do if I was to find an uninteresting, as such, result.

Ravi (23:35)

But it’s still remarkable for someone of your experience and knowledge, to go in with as clean a mind as possible. To just give you some contrast, I worked in consulting for 7 years…Often, young consultants who’d just joined were told from day 1 that “You need to have a prior”, or “You need to have a hypothesis.” They might be consulting on an industry they’ve never worked on before, but they would be told to read up as much as they can, look at prior studies, and let’s say you’ve read up about Market X and now you’re going to study Market Y – you’re supposed to go in with priors, and then look at data that confirms or will dissuade you.

Rukmini (24:20)

But you know, this is so much the case with Indian journalism as well. I don’t think it’s often as explicit as this, but it’s very much the case that a lot of journalists go in with priors. Additionally, perhaps one of the reasons I’ve tried so strongly to turn away from this is because this was the advice that was given to me, too,  in the early days of (my career in) journalism, from people who I believe were well-intentioned and came from a position of wanting to make sure that the journalism we were doing spoke true to power, championed the cause of poorer people and I think their thinking was that journalism should not be seen as supportive of policies designed to keep the rich richer. Even if that was the intention, I don’t think the ends justified the means at all, and the means very much was to tell us young journalists that if the data showed a positive story – something good happening in the numbers; things getting better – then it was not of interest; it was not worth writing about. I think, comprehensively turning away from that has been something that I’m quite keen to do and that I definitely tell journalism students to work on so that it’s not the case that they feel obliged to report only bad news. Nobody is asking for rose-tinted storytelling, but there is no need to force-fit a narrative either.

Ravi (25:56)

Great point, Rukmini.

Coming now from the research…let’s you’ve got a bunch of findings, and you’ve broadly thought of the chapter theme, can you walk me through the steps or the process of actually structuring a chapter? Broadly, there could be two ways – and you could take any particular chapter you want from the book – one way could be to dump all your thoughts or ideas on a piece of paper or Word document and then make sense of it like a jigsaw puzzle, or do you end up directly writing in a stream of consciousness kind of way? What’s your approach?

Rukmini (26:32)

I usually work from a skeleton out. I try to make sure that I have a sense of the rough ideas I want to cover, and that brings coherence to how the information is organized. As you can imagine, and as would be the case for most people writing books, the information is vast. Deriving meaning out of it seems like a hard thing for me to do. I would much rather dwell on the research and what’s in there, decide on what the structure should be and then pull the information from there. What that also ends up doing, is that it doesn’t became a case of the tail wagging the dog because sometimes the way you organize the information takes you in a narrative direction that’s really more an artefact of how you’re coming and reading all of it, rather than how you’ve taken yourself out and thought independently about it. I try to ensure that getting out of the information, thinking independently about the structure, then going back to the information I’ve amassed is something I try to do.

I also sometimes have a few key questions I want to answer, and I try to make sure I get to some of the points of “But what does this not answer?” “What does it not tackle?” “What does it leave out?”

I don’t know how much this comes through in the book, but one thing I try to do in general and talk to people about is to try and move away from a need to project certainty about things we aren’t certain about. A lot of journalists feel obliged to come up with final statements of fact about things that aren’t always fully known, so I do try to make sure that there’s space to question all of that, to ask whether this is the full, final truth.

Ravi (28:41)

Can I paraphrase by saying that there’s one part which is getting lost in the weeds, so to speak, and what you’re alluding to is the part where you leave the data as it is and you mentally step back to try and look at the bigger picture of the map and say, “Okay, there are missing parts here; this is where that fits in,” and so on. Is that what you’re alluding to?

And what is your actual process for doing it? Maybe you can take a specific chapter if you remember it, as an example. Do you take a pen and paper, do you use chits…what do you do?

Rukmini (29:18)

It’s very literally getting away from the data to come up with a storyline, which would involve not looking at my screen at that time when I’m trying got think about it. If I’m only looking at the information, that cues me up to think in a particular way. I do try to look away from the page with all of the information, to either a fresh page on my laptop, or to pen and paper which I do use. I do a lot of note-taking to draw a structure out.

Let me suggest to you, for example, the portions on the economy in the book. Here, I really had to walk a tightrope a little bit because what I was trying to do was to see what the numbers can tell us, as well as to bring in debates about the numbers themselves – about the methodology and the concepts the numbers were meant to cover. When I was drawing up the structure, I would try to do a few things and mix the 2 together. One is to make a hierarchy of what the numbers do tell us. With consumption, for example, I wanted to very clearly say this is what data on consumption can tell us. Then I additionally drew up a set of questions and answers around what are the contemporary debates and key foundational debates around using consumption to try and understand the economy. Then some things emerge as tangents from those numbers. For example, if you’re talking about food which occupies a large portion of the consumption basket in India, it’s hard not to get side-tracked into the conversation around whether calorific consumption is a good way to understand whether people are spending more or less on food. The key debate in India that happened about 20 years ago around why – when the country is getting richer – the calorific consumption is going down. Some of those things would not have emerged in the first putting down of the structure, they would have come a little later. Largely, what I did was to build these skeletons of each chapter and then go chapter-by-chapter in terms of filling them out. That’s the way I work – from the spine out.

Ravi (31:47)

If I look at it on a piece of paper, would the spine look like a list, or a tree, or something? What kind of structure would it have?

Rukmini (31:55)

A list with sub-headings under it.

Ravi (32:01)

So would you have 3 or 4 broad themes for each chapter, and within those, further (classifications)? Or there might have roughly been more or less?

Rukmini (32:12)

I think there were 6-8 within each chapter, and then things under them. Again, this was important to talk about because it was a very data-first way of looking at things. Just as there were key debates to address, there were also key data sources to include. I had to sometimes think about how it is important while talking about consumption to talk about these two or three data sources, because they, themselves, bring in specific debates. Whether you use consumption or income to understand how people live their lives is quite a central discussion and each of them has their own data sources. Under each subsection would have been the questions I wanted to address, as well as the data sources. Sometimes some of them don’t say the same thing, so I’d try to answer what this site says and what the other site says.

Ravi (33:10)

Let’s say that with this, you have broadly identified the 4 or 5 key points that you want to make. How would you decide the flow that “I’ll start with this, then lead on to this”, and so on?

Rukmini (33:26)

I think I’ve evolved quite a bit as the writing went on. I would have started from a “here’s the key thing I want to say” part, but when the writing of it finished, that doesn’t really come in separate drafts; that usually comes right then, which is: I would first write the key point I want to make and then do a bit of writing around it. In some cases, that was how it came (about). In other cases, there were conversations that I had with other people that adhered themselves to my way of thinking about that topic so much that that came first, before anything else I knew I had to talk about that person.

Ravi (34:12)

I must point out that in every chapter, you have managed to write a very pithy, one-sentence summary of what that chapter (looks at.) It’s almost like you ask the question in the chapter name and answer it immediately. I just want to give a few examples. Chapter 1 is How India Tangles with Cops and Courts. Your first line is: “Relying on police statistics and media reporting of police reports has created a distorted picture about the reality of crime in India, and perverse incentives for the police force.” It’s such a pithy line that captures everything. When I’d gone through the chapter and came back and read it, I was like “Okay, now I get it in one line.”

Let me give another example, (from) ‘What India Thinks, Feels, and Believes’. I think this is even tighter: “At its core, India is conservative – even fundamentalist. If there is going to be change, it will take work.”

These are powerful. It’s not something that comes immediately. How is your process of reflection, stepping away from this data, to come up with this one line that summarizes everything?

Rukmini (35:25)

I think it came out of something more prosaic. We had decided to put a line explaining each chapter either in the table of contents or somewhere…there was this decision that we need to have a line explaining it somewhere. I have to say that as a journalist who had to suggest her own headlines as well as what is called as ‘strap’ – the lines that come under this, this is second nature to me. I cannot ever give someone a piece of writing without a headline and strap. It’s just absolute second nature to me. In a sense, it’s good practice, right? Even if you were just writing up something as information to send someone else, it’s a great way of condensing what you write. Sometimes I find that another reason why it’s important to do this is that journalists often get very angry about the headlines and straps that are given to their articles or the way that they’re shared, because that’s the line that gets shared on social media. If it’s controversial, it gets them in a lot of trouble. They get upset and they say, “I’ve written 800 words; you’ve not read all of this. It’s just the headline and strap you’re getting upset about.” Sometimes the problem is not that the strap is incorrect, but that your whole argument – when boiled down into one line – is saying something quite sharp, that you have tried to soften by waffling over 800 words. It’s good practice, because it also forces you to accept what it is that you have argued in 5,000 words in that one line. But I have to say, it’s just second nature to me.

Ravi (37:15)

I’m glad you say that it’s good practice and that it’s something that everybody must do, because in my experience – and I’ve been teaching this for about 6 years – it’s incredibly rare in the workplace. It’s a skill that I think everybody can, and should, develop. (When) they are creating slides, it’s very rare (for them) to put a headline on top. I frequently give examples from journalism. While writing an email, the subject will be “Project Overview”, or “Project Report” – like, what does that even mean?! And so, I think that’s a great skill for everybody to develop.

I’m just curious – what’s the difference between a headline and a strapline?

Rukmini (37:51)

The headline is just usually 5 or so words in bold, right on top. And in the newspaper, you’ll usually see 2 lines in italic underneath – that’s called a strapline.

When I started work, we were never even consulted on headlines and straplines. It would just go as it is. And then we would always moan and complain about it, so at some point some editor started suggesting to us, “Just put a suggested headline at least, because if the person on the desk is short pressed for time, they might just take it or at least they’ll know the key point of what you’re trying to argue.”

Now, I’m actually expected to do it as part of (the article). Not because I am good at it, I don’t think I am very good. But journalism has moved to expecting the journalists themselves to put it in. Plus, the organizations I write for often have different house styles in how they do the headlines. Some out of the space or format; some out of their internal house style. I have to now know how to do this in 4 or 5 different ways. It’s good practice for everyone. Some places I write for have a character count; and I know what their house style is. They don’t take questions; they don’t like things that are (scandalous) – I mean even I don’t want a particularly controversial headline – but it has to be interesting, it has to grip people, so you’re forced to do all of this now.

Ravi (39:23)

What do you think about this whole clickbait-y thing that, unfortunately, some have taken too far but (do you think) there is any merit in that? How do you manage the balance between a headline that tells you what’s there but also makes you want to actually read the article?

Rukmini (39:38)

If the article can be summed up in one line, which the headline is teasing you to get to, then that doesn’t sound like a very deep article. Perhaps a clickbait-y headline is forced to do work because you can get all of it in just one line and the rest of it is just padding or waffling around it. In an article that considers multiple questions, and goes in with a more questioning approach and isn’t claiming to answer everything in a one-line yes or no thing, a clickbait-y headline couldn’t work. It isn’t just a question of being able to answer it. You often see people on Twitter, for example, who revel in posting the headline and then answering it in one word. I think most good journalism can’t be answered in one word. Even if the headline was written in that way, you can’t sum the article up (in a single line). I think we have had a history of extremely boring headlines, and we still often do. It could use a dose of improvement. For the most part, I think Indian journalism’s headline problem is not clickbait – they’re just uninteresting. Too dull. Could have been written any time in the last 50 years. “Wheat Exports Go Up”, “Government Worried About Rising Prices” – how can this possibly be good enough when people are often reading a full 12, 14, 16, 18 hours after it’s happened?

Ravi (41:17)

What might be a good publication that you look up to, for the quality of their headlines?

Rukmini (41:23)

I like quite direct headlines; I don’t enjoy the New York Times wordplay much. I prefer pretty direct headlines. I have often enjoyed headlines from Vox, which are written well but get to the point quite simply. An Indian publication is…there’s an interesting exercise that the publication does, because their headline is one word. It’s often a lovely, little, mild teaser or tangent to what they’re doing. You never read it and feel “Why was that the headline?” It’s very charming.

Ravi (42:24)

Mad curiosity-inducing, I’m sure.

Rukmini (42:25)


Ravi (42:28)

In your journalistic training, is there any interesting book/resource you’ve come across for writing better headlines and straplines?

Rukmini (42:36)

I haven’t, because this isn’t an area I have paid much attention to. No, I would come up short there.

Ravi (42:47)

This was great in terms of structuring and figuring out the one-line theme.

I want to move to the beginning of the chapter. I found that in many of the chapters, you used the story of an individual, whose case was a part of the data that made up (the chapter). I thought that was great! I know you have your reservations about going in with the trope of a human story, but I personally really enjoyed them. I think it gave me a great ‘in’ to the story. Did you specifically think about the beginning separately, once you’d had your structure worked out? Some people call them ‘hooks’ – what do you think of opening hooks? What are different options that you think of and experiment with? And maybe you could talk a little bit about the human stories also.

Rukmini (43:39)

That’s a great point. As someone who writes about data for a living and also tries to bring storytelling into it, it’s really something I should have figured out by now but I still wrestle with this. I still don’t know what the right answer is, but perhaps for me it is to not have a formula.

As you mentioned, many of the chapters begin with a person’s story. Everyone will tell you that this is the best way to open data journalism because numbers intimidate people and are often seen as boring; stories are what’s interesting; stories will pull you in. I recall the ‘openings with a person(al) story’ of so many stories I’ve read over the years and they stay with me. I get it…I would say in most of the cases in the book, this came organically because there was a person who I had thought of right through, and whenever I thought of the topic, that person’s story came to mind and so it came through organically. I would say there are a couple of examples – or at least one, in the book – where it came later. Where I had to think about how to open and decided to open with this person’s story. I wonder if people can tell the difference?

I still don’t know where I lie on it. If my thesis is that data is interesting, it tells important information about how the country operates – I can make data interesting; I’m going to understand the number and make it clear to you; I’m not infantilizing you; we can do this together; we’re going to talk about numbers together. If that is my mission then I shouldn’t always have to open with a number. Why don’t I try not opening with a number? I do believe that in my journalism, I often open with the number itself, when it’s something important. Most of my journalism doesn’t have to do with a number anyway. It’s asking questions in which numbers are part of the answer. I don’t necessarily open with a number because it isn’t one single number. I don’t know; I suppose the smart and easy thing to do would be to accept that this really is how people prefer to consume information, and to continue doing it this way. The downside of it is that it immediately becomes a cliché and over time, that starts putting people off. So, I encourage young people and journalists to experiment with introductions in whatever way they would like.

There is a newspaper that, when others and I worked in Delhi, as young journalists we’d often compare the treatment that our respective desks gave our stories. There was one newspaper in particular that rewrote everything that everyone wrote. And the newspaper read that way – it read in one voice. Perhaps, that gives it an amount of coherence, and maybe that’s good. But I suppose in things that aren’t as cut and dry as newspaper journalism, I think there’s a lot of scope to experiment with how people write. I would really prefer if people didn’t get too bogged down by rules around how to communicate, even though I know I often start with a story.

One of my big problems with how to use stories in data storytelling is the fact that the choice of whose story you tell is very subjective and complicated. By telling one person’s story, I’m necessarily not telling another person’s story. I’ve decided to tell that median person’s story only; I’ve decided not to tell the outlier’s story. Is that fair? Is that right? Should I be choosing this one story? What I’ve done through some of the stories is to push back against that. I have a whole chapter in the book that has a lot of data around love and marriage. I have the story of a person in it who is in an inter-caste relationship, and at the end of the chapter I go back and ask whether the numbers make him feel like an outlier because he’s in an inter-caste relationship while it’s so rare in India. He comes back to me and says “they don’t, because your numbers are about marriage – they’re not about love;” he says, “many of us are in inter-caste relationships but they don’t always translate into marriage because people are not always able to take these leaps of faith.”

In a sense, that story has undone my chapter because my chapter has argued, with all of this data, towards something. If his story is true, then it means that there is a lot of inter-caste love in the country that numbers don’t capture, but I’m okay with that. I’m okay with using stories to complicate certainties that statistics sometimes project.

Ravi (49:13)

I love that example, Rukmini, As Amit says, people contain multitudes. So, you can never have one person as a perfect representation of anything. To your credit, you did mention that in the book, so it’s not that you’re hiding that fact. In a way, you’re undermining (for want of a better word) your own argument…which is fine. I think readers will be able to see that nuance and be okay with it. Once you’ve reached a place where you’re so comfortable with writing, you might find these as tropes and maybe formulaic, but coming back to a large percentage of mid to senior level leaders who are dealing with or presenting data all the time – I find this could be used a little bit more. The idea is not that in every quarterly review you start with a personal story, but maybe insert a few of them in there because if it’s only numbers throughout then I’m not able to see who are the customers, employees or the people behind these numbers. I think that’s an interesting lesson for folks to do.

You mentioned some other options to open…is there anything else that strikes you as a good opening tactic?

Rukmini (50:36)

Just to add to the point of bringing in stories of people –  I very much believe that there is a place for this, and that it is important, and perhaps other ways of doing it could be by not simply using that story to embody the numbers that have gone before. You could use numbers, and then the person’s story could talk in some way to the implications of those numbers. If you use numbers to talk about, say, learning loss during the pandemic, then you talk about somebody who has experienced it, perhaps the story could carry things forward if it’s the story of someone who was able to recoup this learning loss, or if it’s about someone who, had something else been available, would not have undergone this. It can take the issue forward beyond what the numbers are able to show. That’s definitely something I encourage. I will say that it’s not just that I had to choose stories for this book; what I gravitate to most is really people’s stories. I enjoy talking to people, and as you and Amit say, people contain multitudes and I enjoy that mix of things. In fact, one of the problems with the way we relay people’s stories is that we flatten them too much. If we’re going to instrumentalize them for the purpose of what these numbers are able to convey then we take out all of the humanity behind them; we take out the nuance that is within them. If we were able to tell people’s stories in a more rounded way, where it’s not that somebody neatly embodies this particular number –

Ravi (52:29)

I think you should share the example of your colleague refusing to take chai. That’s a great example of this.

Rukmini (52:36)

Oh, yeah!

In the portions in the book where I talk about intergroup mixing, I talk about untouchability and inter-caste hostilities. Sometime around 2012 or so, I was in Madhya Pradesh to report on the wheat harvest. There was a buffer crop and there were all these issues around the wheat harvest. I was going to a remote Adivasi hamlet to meet the son of a farmer who had lost his life while waiting 7 days in line to sell his wheat at the mandi there. There was a glut; Trucks just lined up in the thousands on the highway this farmer had lost his life on. I was on my way there to talk to his son. Since I’d come from Delhi, I was able to have the resources to have a car to take me to this person’s house. A local Hindi reporter joined me for the trip. When we went to his house – he was recently bereaved, yet he made sure to offer me a cup of tea. While we were leaving, he uprooted sweet potatoes from his kitchen garden and gave them to me with the mud still on them, because he couldn’t send me off empty handed. This is something I experience everywhere I go in everyone’s houses! Hospitality and generosity isn’t even the word for it.

When we reached the main road, the journalist immediately told the taxi driver, “I’m desperate for a cup of tea. Let’s go to the nearest tea shop.” I said, “But you said no to tea in this guy’s house. Why did you say no?” He said, “Well, I’m Brahmin. I can’t accept food and water from his house.”

After that, he wanted me to come to his house because he had a college going daughter. He was so impressed by me going alone, and I took my own photos on that trip – I didn’t even have a photographer with me. (He was) so impressed that there was this young woman making her way in the profession travelling alone, he wanted me to come and meet his daughter and inspire her to stand on her own feet. I told his daughter the story of what had happened with him. She was so embarrassed, she rounded on him and said, “I hope you said it’s because you’re fasting! I hope you didn’t tell them that this is the reason you didn’t take it!”

This was a man who held strongly onto notions of untouchability, but still wanted freedoms for his daughter. So, people do contain multitudes.

Ravi (55:29)

And his daughter – she didn’t actually condone his behavior, she just said I hope you gave a good, valid excuse. It’s a fascinating story (and perhaps one where) many other storytellers might have just taken that one slice of him refusing tea and left it at that; that rounded story was a great thing.

Stories are powerful if told with their full, lovely detail.

As I mentioned, (are there) any other ways to open that you can think of, that impress you?

Rukmini (56:05)

Sometimes, opening directly with a number is useful and powerful, and gets you straight to the point. Sometimes, setting the context of the past is also useful. If you are talking about something like life expectancy or maternal mortality – things in which we have made enormous strides within our lifetimes, starting with what things were like some time ago is always interesting. I like slightly meandering stories that have a lot of detail which aren’t necessarily immediately getting to the point as well. Some years ago, I was reporting from Jharkhand during the state elections, I think this was in 2014. I was able to meet the descendant of the great Adivasi leader Birsa Munda, in his village. It was a complex story to tell…perhaps one that I didn’t even have the full ability to tell at the time – or maybe even now – because I was trying to look at the legacy of the Jharkhand Mukti Morcha and other offshoots; the politics of the state, as well as some amount of development made by Adivasis on development indicators; as well as land, some questions like that. I decided to begin with a long-ish description of Birsa Munda’s descendant and him taking out the key that hung around his neck for Birsa Munda’s memorial, that was now in a state of great disrepair, but was lit up every year on the Jayanti celebration. Sometimes, a story that isn’t sharply to the point is also a nice, lyrical detour.

Ravi (58:19)

I think it’s great for the audience to be there with you, right? They can picture themselves in the scene, and see that (set-up) directly. I think it’s great.

These are some cool ways to start; let me now come to the other part, which is the ending. In a lot of your chapters, just like you had a one-line summary at the beginning to give a sense of the entire chapter, the ending was a great way in how it tied up all the loose threads in our minds by saying “this is where we’re at now.”

I just want to give an example of the ending from a chapter on crime (How India Tangles with Cops and Courts). This is what you say: “Registration of crime is a culmination of multiple realities: the existence of a grievance, the empowerment of an individual to report it, the decisions behind the police choosing to register it and mechanisms for accountability. With the last three components so sorely lacking in most of India, it might at the very least be time to stop treating crime statistics as a spectator sport.”

To me, when I read this…in a couple of lines, it does a very powerful job of making me see the whole chapter once, in just these 2 lines. Again, from a writing technique point of view: do these come to you in the flow of writing, or do you have to sometimes step back from the screen – take a walk; get a cup of coffee, and then it will come to you?

Rukmini (59:56)

I had to work on endings much more strongly than introductions. Sometimes it’s not always possible to neatly tie it up; sometimes a one line tie-up feels a bit instrumental and you don’t feel like doing that; sometimes you do need to return to a story you’ve mentioned earlier; other times you feel that you’ve dealt with it then and don’t need to come back to it.

One of the interesting things with journalism used to be that I used to really enjoy crafting nice endings, but frequently the last three paragraphs would get lopped off if the word count fell short. Then I would fight with the desk person the next morning and he would always say, “If it mattered so much you should have put it up front.”

I had to, for a long time, abandon attempts to end neatly because that was the line which would always go. I used to add at the bottom – which, I’m sure, was extremely annoying for the desk of a newspaper working under their own constraints – I’d always add in caps: PLEASE DON’T CUT LAST LINE. I mean, what is the poor guy supposed to do? He couldn’t hack the three paragraphs before to just bump this line up. I had to fall out of love with a neat ending line.

I was able to work on it a lot more when I started writing op-eds for newspapers. There, ending strongly does matter. The point of an op-ed is in relatively more words – though still not a lot – between 800 and 1,000 words, you are making an argument. In most news reporting, especially in the early part of my career, I was not making an argument; I was purely reporting what I saw. More recently, what I do is closer to analytical reporting so there is something of an argument or at least questions being asked over there. Op-eds forced me to think of an ending. Typically, when an op-ed is going in a newspaper, it’s taken seriously. When the people on the desk are there, they’re not going to just hack off lines without running it past you. I would have to begin in a way that was up to me, coherently build an argument through the op-ed, and end in a way that ties it all together. I tried to carry that on here with the chapters as well. There were definitely a few cases where the feedback from my editorial staff at Westland was to end better or do a stronger ending, and so consciously working on that thanks to their feedback is something I had to do.

I do enjoy a cleverly ended piece, and I think it’s quite important to work on. Sometimes you begin with a bang, you work on the argument, and you find that the piece just tails off. So, working hard on an ending particularly in an era where we’re not going to run out of physical space is something I’d encourage people to work on.

Ravi (1:03:08)

It reminds me of something from psychology, I think, known as the “peak-end rule.” When humans go through any experience, they don’t remember the entire experience. It could be something as random as a dentist visit to a holiday; we remember peaks from the visit and we remember the ending. If you had a very bad flight back from the holiday, you might remember that. That kind of ties in here.

So what you’re saying is that you may not necessarily write it in the flow, you may actually go back and think about it if that helps you?

Rukmini (1:03:40)

Yes. I think I do have a tendency to just peter out off what I was writing. I either remember that I need to work on that and come back to it later, or I would see that it has often come back to me as feedback, maybe from the publication I’m writing for, or in this case it was Karthika and other people at Westland.

Ravi (1:04:03)

Moving on to some of the other elements of writing. One thing I really like is the use of surprise in writing. If you’ve got some data point that surprises even you – of course, it might surprise the audience even more – then I’ve seen people tend to make it land with a bang or flourish. They would be like “This was something that made me realize something very different.” Then there will be a full stop and a paragraph, and the line will follow in the next thing. You might find that most writing on Twitter and LinkedIn follows this format to a T; A lot of new age journalists also do this. I didn’t see a lot of that in your writing. If something was surprising, it was there plain as night and day. What do you think about the use of surprise in writing?

Rukmini (1:05:00)

I can’t say I don’t enjoy it while reading other people’s writing, but I can’t really imagine constructing something I write around an attempt to build surprise. I do care about the information being surprising. While explaining the concept of ‘news-yness’ to young journalists, who often want to know if it’s newly happened then is it news, is that it and is nothing else news-y? I do try to remind people of the role of surprise as being important to a lot of news journalism. I do think I enjoy reading it; I enjoy it very much in fiction. Possibly less so in journalism and non-fiction. But I would have wanted the data points and content itself to be surprising to people, but I can’t imagine constructing a narrative around the desire to surprise. I’m not sure. Maybe it’ something that, with more work, has a good payoff. I’m not entirely sure about that.

In general, I perhaps pay less attention to the actual craft of writing than I should. I am always overcommitted and short on time, so I tend to steamroll through most writing from my journalism days to the book. So slowly, well-considered work is something I’ve not had the chance to do.

Ravi (1:06:49)

Your productivity is incredible. The fact that Anish Chandy gave you a 5-month deadline for a book, while you have 2 kids at home, and you’ve got other assignments, and it’s the pandemic – that you took it on and managed to do it is crazy!

Rukmini (1:07:10)

I think ignorance was bliss there. I had no idea what timelines you were actually expected to write a book in, so I thought this was Anish –

Ravi (1:07:16)

You had much shorter timelines for your articles.

Rukmini (1:07:19)

That is exactly it. Anything that’s beyond the same calendar day is news to me.

Ravi (1:07:28)

On that, are there any lessons you can share with other parents? It’s one thing to write as a non-parent, and it’s very difficult to find time for longer term goals and there’s always the urgent time where – as you say in other conversations – that one Lego piece can be found only by you. What would you do to manage distractions both from family but more importantly, your own (mind wandering?) Like when you get stuck on something, wouldn’t you be like “Let me just go and look at the news” or in my case, the cricket score for some time? How do you manage those?

Rukmini (1:08:06)

No, I don’t claim being immune to procrastination, but I will say that by being constantly overcommitted as I always am – both personally and professionally, and this was particularly the case for the book – the period of time that I had was it. There was no elbow room. If I missed working at that time on it, then I was not going to have time for large chunks. For the whole of 2020, right through the year, most of my work was happening early in the morning or late at night. My kids were even smaller then.

Ravi (1:08:52)

By “early” – how early are we talking about, Rukmini?

Rukmini (1:08:53)

6 to 8. I’d give myself 6 to 8 in the morning. I don’t work 6-8 in the morning anymore…the kids are back in school so life is sunnier. Perhaps, I could do journalistic work while also watching my kids because it required much less deep thinking but putting things together, making phone calls, sending emails, that sort of thing which I could do. As could the rest of the world – work with their kids around them. But I couldn’t work on the book while also watching my kids who were home with me all day. From January 2021 onwards, until May or so, I had babysitting off and on. Through the Indian summer of May 2021, my parents were here full-time, watching the children. If there are people out there who could get good quality work done while simultaneously watching their children then good from them! I couldn’t. The way for me to get it done was for me to be in a completely separate room with the certainty that a kind and competent person like my parents or my husband was watching the children. Better still, being in an entirely different house. I did a lot of the writing from a friend’s house who was also a journalist; she would also work from home, so we would just sit at our respective tables and work and only break for lunch. I do think having a lot of external deadlines over my head, perhaps because I’ve always been in a deadline driven environment (helped me). In this case, the deadline that my parents were going to leave by the end of April or so, made me think I had to get it done while they were there. That’s what really helped me pack in the productivity.

Ravi (1:10:49)

Did you break deadlines down into smaller ones, like I need to send this chapter by this week, so on?

Rukmini (1:10:55)

Yes. There, the book writing process helped. I had to hand in two chapters first, and I had to send those in by a particular date. In fact, the book getting picked was contingent on those 2 chapters going in so they had to be done well.

Getting started is such a big part of it. Once you’ve had to do 2 chapters in a way that makes it good enough to be able to pitch to publishers, then you’re up and rolling. There are so many processes in publishing that me handing it in was just step 1 and then there would be all these processes after that. Knowing that helped me stick to it.

That said, I think I’m going to be slightly behind on the deadline for my second book so let me not crow too much about this.

Ravi (1:11:48)

Fair enough. Coming to the content now, I want to talk about the culture of data in India and all that. One thing that struck me while reading this, and also while listening to you talk about it, a lot of the content we are reading in this – especially on crime and people’s attitudes about caste or religion – they are emotionally quite taxing or disturbing. How do you manage your emotional steadiness while you’re getting all this data? I can’t even imagine the work you’ve put in to read so many court cases on violent crime in India. It’s tough. Of course, journalistic training might have helped, but I want to know what is your approach to manage your emotional steadiness when you deal with that information, re-think about it and write about it, so on.

Rukmini (1:12:53)

Yes, in some ways my journalistic training helped because I have done field reporting in very difficult circumstances, as all journalists have. I’ve reported on floods and terrorist attacks, gone to people’s houses when they’ve experienced violent crime and had to talk to them about it when they might not want to talk to you about it – really, the deep end that you get thrown into as a young journalist. That has given me the ability to separate information that I’m dealing with for work with my own life and what I take home with me. That’s one. The other part is that I do try to switch that stuff off when I’m at home, particularly when I’m with my children. For me, it’s not so much the information I’m reading but what I consciously switch off is social media. For maybe the last several years, I don’t read my mentions on social media and I think that’s an extremely important way for me to ensure I maintain my emotional stability and not feel anxious and to not have it overrun my interactions at home and with my family. I couldn’t recommend it more, especially to people with something of a public profile who may encounter this stuff.

But I would say, all of this preparedness and experience didn’t prepare me for one part of it, which was the 2nd Covid wave. It seems trivial to even talk about your emotions as someone consuming information, given the loss that people personally experienced. Forget what I felt as a regular citizen; I would say that while writing about it journalistically and for the book, I found it difficult because of what people had experienced. I found dealing with the dismissal of people’s experiences and the data-driven debates around it (to be) particularly difficult to engage with, just because it felt like such a catastrophic event to have befallen so many people that to be getting into petty politicking around it felt like a huge disservice and something that I was not going to engage in at all. I still find that talking about those stories, some of the people I know well and speak to fairly often, thinking about those times – what people went through, and the debate around all of it…that’s the most recent example I can think of about struggling to make that balance. Otherwise, I’m quite well prepared for most other things. I would also say in the last few years I’ve done much less field reporting than the past. What young journalists in particular, many of whom are women, experience when they’re reporting violent crime or hate crime, riots on the field – those are far more significant challenges than what I’m talking about.

Ravi (1:16:46)

It’s remarkable, I think.


I want to move on to the concept of data culture. It’s a bit of a long, philosophical question; Here’s how I think about it: One way to look at data is, like an economics student, from a demand and supply point of view: for any other product, there is a demand for good data and there should be a supply of good data. Let’s talk about demand. One way to introduce that argument is that data matters. There are two quotes that come to me when I think about data, one is almost a frivolous one, from Sherlock Holmes – “Data! Data! Data! I can’t make bricks without clay.” Without good data, you can’t make good decisions. Most organizations understand this; there’s this famous Edwards Deming quote: “In God we trust. All others must bring data.” A lot of companies would very proudly call themselves as data-driven in decision making. While they may have some flaws, I think that broadly, many of them do deal with a lot of data. If you move the lens form companies and organizations to countries as a whole, how does a country – especially one like India – deal on this scale? How much do we demand good data for decisions? It doesn’t look like we are very good with it…there’s a lot of false narratives that are peddled as your book so eloquently lays out. The whole narrative about ‘most violent crimes take place through strangers’, which is not the case; the one that ‘voters care mainly about development and economics’, which doesn’t seem to be the case; that ‘India is mostly a vegetarian nation’, which is definitely not the case. There are tons of false narratives floating around, so one might feel that India is particularly disadvantaged when it comes to data. But to take a wider lens, I don’t think it’s necessarily an India problem but a human problem – humans struggle with processing data; we prefer simple narratives and like to see the world in simple cause-effect relationships. Politicians across the world peddle simple narratives. Media does it too. It’s a common thing across the world.

Coming back to my question…if you step back and look at both the global and the over-time view of India, how data sensitive or how data-demanding is the Indian political administration, general public, media, etc.? How is it today, and how has it changed over time? Do you see it improving, or becoming worse, as compared to the world?

Rukmini (1:19:44)

I think I have a long answer for this.

I think data has, and remains central to the work of the government in India. The centrality of data to the decolonizing moment and to the process of planning is really, in some ways, the story of how Indian data progressed. The plan or the thinking that Nehru and (PC) Mahalanobis had at the time in the early 50s, around the role that statistics were to play in developing a planned economy with the stated role of pulling people out of poverty resulted in the putting in place of a statistical architecture that stands in good stead even today. It has always been an important part of policy making and I think that this has improved over the last several years. The centrality of data to decision making is quite impressive when you talk to people within the government and understand the data they’re looking at and who it’s feeding into, and I do believe it’s getting better and better data is being harnessed for the decision-making process. However, that’s not the full story. The decisions around what is done with those numbers and policy decisions necessarily have to do with political incentives that may not always be aligned purely with what the numbers are showing. I think it is important to differentiate, in some ways, between political narratives and the actual hard work of governance and policy. I think numbers and statistics play a much greater role in policy than perhaps in political narratives. But there are undoubtedly cases in which there is a dangerous divergence between numbers and policy and politics as they come together. Crimes against women is one of those examples where most people were either looking at wrong data or the wrong interpretation of it in the activism that happened in 2012, to then determine that iron-fisted policing of public spaces was going to be the main problem to solve, as well as the notion that things like raising the age of consent for women and lowering discretion in the hands of judges were important policy decisions to make as against looking at the data in more detail to understand what the numbers on sexual crime were actually capturing, which is something I get into in the book. I do think that was one instance in which politics and policy galvanized around a poor reading of the numbers to take us in a dangerous direction and one that I think has done tremendous harm in the last ten years to the overall freedoms of people and of Indian women.

There is certainly a lot of data-free politicking that happens. Additionally, there is sometimes cherry-picked data-driven politicking that happens. It’s not that numbers are not used, but that they are used in a dishonest way to make dangerous points sometimes. We often see this around the weaponization of numbers around fertility, for example, which politicians repeatedly try and use to argue that Muslim fertility is a big problem in India while a closer look at the numbers shows that fertility is falling across all groups, Muslims included, and is converging around a very low number. In fact, a number that is so low as to now worry some demographers. Having numbers and using them for progressive purposes are definitely two different things.

Additionally, I think we’ve had some instances in the last few years of suppression of inconvenient numbers. In my view, that makes it incumbent on people who work with numbers to be more vigilant and to be more outspoken in their criticism when this does happen. I’m sure there are attempts by governments across the world to try and suppress inconvenient numbers, I don’t think this places us in Banana Republic territory. It is incumbent on us to pay more attention when this happens.

Ravi (1:24:43)

That’s such a nuanced answer. I think, for me, the distinction is now clearer that there is a public narrative and a political narrative by politicians, and there is policy which is different, which I had not differentiated in my mind.

For example, the suppression of the Labor Force, Unemployment Survey, the Consumption Survey, and now the postponement of the Census – some of these, as you say, might just be for political gains, but do these also have an impact on policy being delayed or in case of the Census, actually deny it in some cases?

Rukmini (1:25:29)

A couple of dangerous things happen when this sort of thing occurs. One is, when you draw the bar that you make it acceptable to suppress inconvenient numbers by giving the excuse of changes in methodology or problems with methodology and it isn’t met by widespread opposition – which is the case with the Consumption Expenditure Survey, which was withdrawn after it showed the first decline in real income in decades – an ostensible reason was that there were problems with methodology; of course, problems with methodology could have always been discussed post-facto. The whole survey didn’t need to be junked. When you draw that line, you do set a precedent. You do serious damage to public faith, I think, because when the survey is conducted there will be diminished faith in the results of it. It does have that impact. I think it additionally demoralizes people working within the statistical system, where they can feel that work that they do which is within the directions of the central government and very much within international norms can be, at will, stopped. I think that makes them feel much smaller.

I think another thing that happens is when you demonize and criticize the methodology of surveys you disagree with, and if it is bought wholesale by people – which is something that happens; the success of a lot of propaganda means that a large proportion of people who read any international survey that is seen as critical of the government reflexively feel that it’s a survey designed to make the government feel bad and cannot possibly be true. When all of that begins to happen, I think it does damage to the overall integrity of the data space. That is something that’s happened, I think. We’ve seen it with disagreements around consumption…I’ve said it in the book as well, these fault lines have existed for decades. In a way we are having new iterations of arguments that we have had for decades. But what we’re saying now is such a demonization of the other side that perhaps some of these processes might wither away if you create enough public opinion and momentum around the fact that the survey itself is bad.

I have to say I have no idea why the census is so delayed. I don’t think it necessarily can be explained by politics because I would need to see a much more convincing case as to why the census is so explosive that it needs to be suppressed in this form. I haven’t seen a single convincing answer for what is going on with the census, but it is certainly worrying. It’s worrying that we haven’t seen a convincing answer – nobody has felt obliged to give a clear explanation for why it remains delayed.

Ravi (1:28:57)

What I was reading about was the supposed huge gap that might come between administrative data across a lot of ministries versus the ground situation in terms of many assets in the household elements.

Rukmini (1:29:12)

The thing is that the census captures so little of all of that. It doesn’t capture a lot of assets in great detail. I think there are other ways to test against administrative data. We have a pretty robust sample survey process that goes on which does test against administrative data and does find gaps, but not ones large enough to justify this. I still remain quite confused.

Ravi (1:29:43)

To build on this and take a bit of a devil’s advocate position…let’s say something like the amount of data suppression that happened around the 2nd wave of covid where nobody could believe what the government was putting out because it was (unbelievable). But in some of these other situations where the data that’s coming out on the fact that real incomes are coming down, or wages have come down – what happens when, for not just the government but the public at large – the data that comes out contradicts what they perceive as their lived experience (which could be a very bubble like existence) that, “How can you say wages are coming down? I have been paying more salary to XYZ,” so when that happens I think there is a gap that opens between lived experience and data, and someone has to fill that gap. Ideally, the people with the lived experience should grapple with the data and figure it out but given the polarized times we live in, is that sometimes a role for the data provider to become a bit of a storyteller and try and bridge that gap? To say that, “I know this what you feel; I know you feel that Uber and Ola drivers are coming and there’s a gig economy, but here is how this survey actually includes that also.”

So to reconcile with some of these possible arguments, do you think that the data providers themselves should take on the responsibility of bridging that gap?

Rukmini (1:31:16)

Absolutely. I think if there is this response, and if there is outrage to data by people who feel that it doesn’t accurately reflect reality then instead of reflexively blaming those people for being oblivious or choosing to not accept what reality is, I do think it reflects on a failure of those communicating the data. If the process and the methodology for collecting the data isn’t clearly explained, it becomes that much easier to bat that data away. I find that the best estimates we have on consumption expenditure which show that in 2017-18, if you spent more than 8,500 Rupees a month as an individual it puts you in the top 5% of urban India. That’s something that people often push back against, find unbelievable, or want to say that “There are so many people I know who are spending this, that; you can’t tell me I’m in the top 1% of the country”, (I think that it) comes from us not being able to properly explain how these numbers were calculated and what goes behind it. Explaining to people “Is your objection that this is being missed? Let me show you how it is actually captured in the data,” “There’s all this black money – let me tell you how consumption doesn’t mean that black money isn’t captured.” I completely believe that there’s a failure of communication rather than a refusal to accept reality on people’s parts.

Ravi (1:32:51)

And the leaders of statistical organizations must probably build that skill and do that communication better.

Rukmini (1:33:00)

That’s certainly one way of thinking, but given that we now have an entire field of work that is dedicated to explaining data whether it’s analysts, data journalists, researchers – it is incumbent on all of us in this community to share it. I am okay with saying that the government’s job should extend to good methodology, collecting the data well, making it public, and explaining the methodology well to the extent that at least researchers or experts understand it. The communicator’s job is to then demystify that and explain it to people. For years, I wrote about consumption expenditure without spending much ink on explaining how these numbers were collected. At that time, it was because of word count restrictions; now, I think there are no excuses to write in that way. I think this also extends to other things like opinion polling, which people often feel is “What is this? You’ve surveyed 1,000 people and you claim that you know how the country works?” it means that I’ve done a poor job of explaining how opinion polling works.

If data is clearly presented, and still fails to convince, then perhaps there is a motivated reasoning issue there. In the absence of that, transferring blame onto people for refusing to accept reality and feeling that their lived reality makes more sense than the numbers they see says more about the communication than the consumption of that data.

Ravi (1:34:46)

Well said, Rukmini.

We talked about the demand for data in the culture; let’s talk about the supply part where you mention that “India’s statistical architecture is vast and impressive”, you also (suggested that) it’s like a 7 out of 10, which surprised me because as a consumer of data during my consulting years, I always felt that we lack so much data – we don’t have good data on income, wealth and so on. This was interesting. If you could step back and (talk about) what aspects of India’s statistical infrastructure would you say we are really doing well in, as compared to other comparable countries and where do you think we could be doing much better?

Rukmini (1:35:27)

I think we did well until now on a lot of health administrative data. On health infrastructure, the National Family Health Survey, which is part of the global demographic health survey system, has been conducted very well. It’s conducted every 5 years so there is that issue, but it’s a rich and good data source. In addition, we have data from things like the National Health Mission which is supposed to give much more recent and up-to-date data. I think the National Sample Survey framework also is impressive in the scale of what it collects which tells you things from consumption to employment to other key understanding around debt investment, agriculture, many of those sections. What has changed is that we need much more high frequency data in India. The era of being able to rely on surveys that were conducted every 3 to 5 years and the findings of which came out in PDF form 2 years after that is now gone. Maybe that was okay right up till the 80s, but we should have moved to much more high frequency data soon after that. I think given the depth of work on the economy and the fact that tech-driven data work on the Indian economy has run far ahead of what the government is able to do, has meant that that gap looks so much more glaring. We lack high frequency data on the economy, on employment, all of that.

Again, I’d say that the government has attempted to make changes so we move to the Periodic Labour Force Survey which is now conducted every quarter, which is much more frequent than before. But given that the Centre for Monitoring Indian Economy – a private data source – is able to give weekly data on employment…the government seems that much further behind.

Historically, some of the improvements that have happened in the last few years have been on administrative data. We have really gotten much better at collecting administrative data. I think NREGA has been a great example of just the scope and breadth of what administrative data can do and that’s similarly stretched across many other government schemes, and there are people in the government that are keen on open data and putting more data out in the public. On all of those fronts, there is a lot we’ve done that is of impressive quality.

Ravi (1:39:57)

Coming to the close of the conversation, Rukmini, (I) just want to talk about your influences for your writing and specifically, I’d love to know (if there is) any book, fiction, non-fiction, or both, that you have gifted the most to others.

Rukmini (1:40:11)

That’s a great question. The book that I gift the most is not necessarily (related to) the way that I write, but one that I gift for other reasons. The book I gift the most is P. Sainath’s first book, ‘Everybody Loves a Good Drought’. The reason I gift it is because I’m typically giving it to young journalists and I want them to have a sense of what the country can be like, again to not think that everything is neat because many of the stories in this are not neat. It isn’t how I write but it’s definitely a book I share a lot.

In recent years, I’ve been sharing books by Tim Harford with people. I enjoy his podcast that I was lucky to be on as well, it’s called More or Less on the BBC and it’s really a wonderful way of sharing the full context with numbers. There are all the classic books, which is How to Lie with Numbers and the book by a person who used to work at OkCupid which was the book I was reading around the time I was writing my book as well. All of the book is trying to interpret data from OkCupid and other early internet sites to try and understand human behaviour. Of course, when I read that book, I was extremely envious that I don’t have access to data from Tinder or Google or Amazon that I would love to use for my book. In terms of influences, I think some of them are more on the fiction side than on the non-fiction side. I do read a fair amount of non-fiction but the writing that typically stays with me tends to be fiction. My favorite books (include) ‘A Suitable Boy’; I was recently reading Nikhil Menon’s ‘Planning Democracy’ on Mahalanobis and his role in the early days of planning. It’s a great book and the other book I recommend to people who want to understand the 40s and 50s is ‘A Suitable Boy’. That’s the book I go through to even think about that moment and all the feuding of the Princely States, that sort of thing.

I remember reading Upton Sinclair’s ‘The Jungle’ a long time ago, and being struck by how non-fiction can be moving as well. Another book that I enjoy very much is a Hindi book called ‘Raag Darbari’ which is a satire on a very specific moment in time and I think it is in some ways an embodiment of what non-fiction reporting probably showed at the time.

I think there is some really fantastic journalistic longform reporting going on all across the country and I enjoy the range of styles that journalists bring to it. I think (from) overseas I used to like a lot of Vox in its early days, and I used to think of it perhaps as a good model about what data journalism or explanatory journalism could be like. I’ve settled on preferring writing that is a bit more empathetic and a bit less neat. I think Vox, too, was like that. It accepted that there could be uncertainty in the world.

That’s some of the writing that I’ve enjoyed the most.

Ravi (1:44:24)

Such lovely recommendations, Rukmini. Thank you.

I also remembered Cautionary Tales by Tim Harford which is a great narrative journalism podcast.

Where can people get to know more about you? Maybe you could end with a short line about your upcoming book.

Rukmini (1:44:42)

I still do a bit of independent journalism for a few publications including Mint and IndiaSpend. I’m also hoping to do new things this year which I hope to be able to talk about in a little while.

My new book is also with Westland and it deals with misinformation and disinformation; me taking cues from what the readers said they were so concerned about. The book tries to look at 10 key simple, statistical concepts that could help readers potentially distinguish between what’s real and fake and try and pierce through misinformation and disinformation. The book is meant to come out at the end of this year and will also be published by Westland.

Ravi (1:45:33)

Wonderful. Really look forward to reading that and maybe we can have another chat once that comes out.

Thank you so much for coming on this podcast! Of course, I have learned a lot about better writing (from you) even though you don’t call yourself a “writer”, but I also think (I have learned) a lot more detail and nuance about India’s data architecture and culture. Thank you for coming.

Rukmini (1:45:58)

Thank you for having me! It was a real pleasure.

And that was Rukmini S, one of India’s leading data-journalists who is illuminating and shaping the way we understand our own country through data.

A few things which stood out for me in the conversation:

  • Data matters. Let it not be suppressed or unfairly questioned, because the findings are not palatable
  • Communication matters – it is the responsibility of the data presenters and wider data journalist community to explain data findings and place them in the right context
  • Writing matters: I’m glad Rukmini took time from her schedule to write this seminal book and hope to read many more from her pen

If you find this content valuable, please rate and review this podcast on iTunes, Spotify, Google Podcasts, or wherever you listen to them. It’ll help others like you discover these insights!

This podcast was hosted by me, Ravishankar Iyer. Audio editing by Kartik Rajan. Transcript editing by Amisha Jha and all-round support by Sanket Aalegaonkar.

Until next time, may the force of good stories be with you

Get Storytelling tips in your Inbox

Subscribe to the 'Story Rules on Saturday' newsletter

Get a free e-book that decodes the hidden storytelling structure used by leaders like Jeff Bezos, Bill Gates and Warren Buffett.
Your infomation will never be shared with any third party