 The One World Project is initiative to examine a diverse set of inputs and perspectives regarding the introduction and evolution of technology as part of elections. We began in May 2021, but back then we used to only focus on blockchains. Now we've broadened out the project works with stakeholders in the electoral process to understand present and discuss knowledge from various domains. Media communications, disability rights, citizen rights, technology designs and constitutional powers. One vote is hosted at Hasgeek and Hasgeek is a platform for collaborations across practices surrounding technology design, law, policy, systems, data and related topics. So the collaborations take place through user generated content shared by practitioners. So far we have designed and hosted public deliberations looking at activities like voter verification technology, electoral role deduplication and the various internet based voting options. So a written an internet report and just finished up our first edition of an annual conference. What we aim to do here is to enable participants of the sessions to acquire the foundational knowledge and perspectives that are required to evaluate the intended and unintended consequences of technology intervention. To have a perspectives of identity, equity, privacy, security, right agency and look at the socio economic impact of such proposals. This particular series is being done in collaboration with the Jindal School of Journalism and Communication. About audience questions, if you wish to speak, please use the raise hand function and you'll be called upon. Or you can put your questions in the chat. If you're joining us from the YouTube live stream, we'd be monitoring that chat as well. We welcome Roshan Kishore and Govind Etiraj. Now I will hand this over to Professor Paricharty. Thanks Shantel and hello and welcome to another edition of Fight for Democracy elections 2022. An effort to raise and discuss some critical questions around one of the most important elections, many would say, but all elections are important actually. I mean, this is what we say each year before elections that this is the most important election. I was going through, you know, I worked in an organization where we dealt mostly with numbers. And that was NDTV, my boss for 20 years, Purnoy Roy, who was introduced to psychology in a way in his latest edition of his book. He starts the book by saying democracy lies at the very core of every Indian's DNA. It is intrinsic to our consciousness. It animates our conversations, energizes our minds and brings out the best and occasionally the worst in us. The more deprived we are, the poorer we are, the more alienated we are, the more we participate and are protective of our country's elections and democracy. And towards the end of the book, one of his last chapters actually he says, the source of all problems is India's poor data. Joining us today, Govind Rajyathiraj, the founder of India spent India's first data portal boom live. Again, one misinformation portal. In many ways, Govind really saw the criticality of data journalism and ushered this in our country way, way before other newsrooms really were exploring. And Roshan Kishore, an economist, who heads the data in department in Hindustan Times. So two journalists, Govind, with many, many years of television and print journalism experience. I want to start with you, Govind. If you can quickly tell us or give us a landscape of data journalism. I know data journalism has become the season's favorite, but what exactly is that? We had data earlier also. It's not like data has emerged just today or just yesterday. What has been the genesis of data journalism in India, growth, presence, and where do you see it moving? Right. So, Kishore, firstly, thank you very much for inviting me into this discussion. I think your question somehow implicitly presupposes the fact that data journalism is well-present, practiced, delivered. Actually, we are in a very early stage of the journey. So, in a way, this conversation itself is part of that early stage of the journey. As you have conversations, more people begin to understand, including ourselves. I mean, we may have set down this path, but I don't think we are any closer to being experts or any authorities on the subject. So, when we started India Spen, the whole objective was to firstly improve the quality of public discourse. And we started from there. We said, okay, how do we improve the quality of public discourse? We can do that by using data to tell stories. And the optimistic view is that if you use data and evidence to tell stories, then hopefully the quality of discourse will be better. And therefore, the electorate will be better informed and therefore will be in a position to make those who are in power accountable and move and eventually there will be better governance. Now, this is, depending on where you sit or stand, this is either utopia or at least the journey towards utopia of which data is a critical part. Now, the big challenge when it comes to data is obviously the communication of it. Now, there are two kinds of audiences. There are those who understand and appreciate exactly the role of data, mostly in public policy, academia and so on, this audience for example. And there are others who perhaps are not able to grasp it, not because they don't understand, but because it's outside the realm of daily life. And to that extent, it's really, let's say the challenges upon those of us who work on data sets or who write stories around data sets to convey the challenges or convey the proposition that we are trying to convey. So in that sense, if I were to be a little blunt, we failed or we've still not succeeded. Let me look at it the other way around. So I think the, now let me give you a couple of examples. And this is really from stories that we've done. And you talked about elections, we are all looking at Uttar Pradesh elections that are coming up. In 2017, we had done a story where we leaned on a pole done by a fourth line at a survey. And we said, because that was an issue, we were covering very closely, we were looking at air quality very closely. So we asked people, does air quality matter to you? And I'll come to the definition part in a moment. So some 46% of urban voters and 26% of rural voters in UP before elections said that the air they breathe is polluted. So they acknowledged that this was a problem. But on the other hand, this was not an issue that the political parties wanted to address, nor did it even come up anywhere in the scheme of things from the political parties point of view. So second, when asked to choose between reliable power, clean water and clean air. Okay, now clean air and I'm going to come back to that definition part in a moment. 40% of voters said reliable electricity was most important. 28% said clean water and 16% said air quality. Now what is the definition issue here, which makes it interesting, right? So elections are fought on broadly, let's say tangible issues and intangible issues. In this case, air quality is a very tangible issue because you're literally seeing it and breathing it. Now, maybe because you're seeing it and breathing it or not breathing all the time, it becomes invisible. But the fact is that you could not get more, let's say, tangible than this as opposed to many of the intangible issues on which people vote. So therefore, that itself gives you an indicator of how people vote. Now let me give you another example. Malnutrition is an issue in India. Now malnutrition as compared to air quality is not as tangible. It comes later. But nevertheless, we looked at a study, for example, in Dunn, I think, just let me give you, I'm trying to pull out the exact. So this study looked at 542 parliamentary constituencies but said that 72 of those constituencies were in the top two quintiles, that 20% of prevalence of child-matte nutrition. Of these 12 constituencies were in Jharkhand, 19 in Madhya Pradesh, 10 in Karnataka, 8 in UP and 6 in Rajasthan. Now this particular study, which was run by Harvard, did not get into what political leanings were or electoral preferences were. But I think it's safe to say that there was almost no connect between, let's say, health outcomes at this level, which is very, very fundamental, malnutrition and stunting and what eventually happened in the election. Let me give you one more quick example and then I'm going to pause. In 2018, Niti Ayog, which is under the current government, put out something called a Health Index Initiative. It provided disaggregated scores and rankings to Indian states and union territories according to their health sector performance. Now in this, Kerala, Punjab, Tamil Nadu ranked on top in terms of overall performance, health sector performance of Indian states. UP, Rajasthan and Bihar fared the worst. So UP is run by BJP, Rajasthan Congress, Bihar is BJP+. And then there were small states and then there were some other interesting patterns between, let's say, Kerala, Punjab, Tamil Nadu, Maharashtra, neonatal mortality and their overall sort of performance in this thing. Now, could there be, let's say, greater electoral or informed voter choice, possibly, but it's highly unlikely. And there are, unfortunately, also, which is the other problem. There are not enough studies to look at this issue so deeply as to understand why did voters vote in a certain way. And more importantly, did any of these aspects have any play in the way they voted? So we know all the obvious reasons or likely obvious reasons, but did any of this have a role in it? We don't have a clear idea. I mean, there are some studies, but it's not clear. So, and you have to remember that finally, as we speak today, there is a haze of misinformation which sits on all of this. So the haze also causes further distortion in the way people view what are, let's say, physical outcomes in terms of policies and so on. So even if something is not working, again, along the issues that I spoke of could be to do with water, could be to electricity, could be to clean air, health access. There could be a haze of misinformation which typically twists people's perceptions of what the real source of the problem is or takes them into a completely different thought zone and thus influences the outcome of elections. So I think as we look forward and I will pause after this is there are two or three issues. I think data is available. There are people now increasingly who look at data and tell stories. The challenge I think continues to be how do you ensure that these stories reach more and more people in a way that it affects their or influences their thinking in a positive way? I mean, and is able to influence the way they take their electoral decisions about who they elect and why they elect them. India's fundamental or foundational issues are health, education and environment. I could add gender but gender, let me maybe in some way subsumes all of this. So health, education, environment and gender, these are the four areas that we at India spend cover as well. If we want people to elect those they elect on the basis of performance in these four areas, data should be an important input. The challenge therefore now is to say how do we convey that data in a manner that people make the most useful or efficient electoral decisions. I'm going to pause there. Thank you so much. Thanks Govind for laying the land as it were. Roshan, I'm sure you would like to respond to number of issues that Govind has raised. But essentially what I understand is that in a democracy the electorate should be well informed. And today we are assuming that with data the electorate is going to be better informed. But there is a challenge because the data may not be correct. And you know how do you ensure that the data is good, the data is authentic. And the other thing is that reaching the voters because a large percentage of the voters will not be English language speaking people. How do you reach them because the journalism that we guys do mostly are going to be English language. Are there language journalists in the regions using data to do journalism in various other languages is something that I would like to know. And finally Roshan, you know you are in the political desk as well at the Hindustan time. I mean data driven let's say but what does data journalism bring to table when it comes to media coverage of elections in India. I'll start with the last question Krishna. Like you said when you pointed about the work that Roy has been doing at NDTV. So we have been talking numbers as far as elections are concerned for a very long time in India. There's nothing new about it. And if I were to slightly sort of push the envelope and include things such as economic and political weekly, which is informed journalism at a certain level. Then you know CSDS, Lokmiti numbers etc. So those numbers have been around. And you know journalistic work around elections I think can be divided into two parts. One is what you do when the campaign is on and the other part begins when you try and make sense of the results. So there I think you know the value which say dedicated data journalism desk can add is, you know, for example, you know, 2014 onwards, you know politically speaking we have entered a new epoch of sorts in our country where the BJP has been the dominant political party. And it's sort of sort of further strengthen that dominance. And with the BJP is a particular narrative in India, the BJP has been a party of the Hindu right and no, I mean, people largely say that the BJP victories actually mean that you know that sort of an ideology, a Hindu right excludes, exclusionary ideology has been getting ground in India. And I think that is not true. But one thing which data if you look at it carefully, not just during elections but over time in a continuum actually tells you is that the reality could be a bit more complicated than that. I'll give you one example, look at assembly election results and national election results after 2014. In all national elections at the state level the BJP seems to enjoy a vote, vote creamy. You know, Delhi is the best example the city where I work in one of the app actually finished third in the 19 Lok Sabha elections. It won the state with a very overwhelming majority before the 19 Lok Sabha elections and it did so even after the Lok Sabha elections. So it does tell at that you know voters actually make very informed choices. If you look at the voting pattern in say Odisha in 2019. The voter going to the same polling booth actually seems to have voted for two different political parties, and the BJD at the state level and the BJP at the national level. So, you know, I'm not for a second undermining the importance and I think the misinformation has always been an integral part of political tactics I think through centuries, not just in modern democracy. But the advent of internet and the WhatsApp ecosystem, you know, today I think, if we were to be honest enough we will all actually question as to how much control does journalism mainstream New Age journalism actually has a narrative setting because every political party, the best of its ability I think is trying very hard to set up its own parallel narratives the BJP WhatsApp ecosystem will have a different sort of a narrative, you know, the opposition ecosystem will have a different sort of a narrative. And this is not something which we're doing for the first time, you know, work by political scientists like Milan Vaishnava as an example, he works at the Carnegie Institute these days so his PhD looked at the issue of criminality in Bihar. And his research found out that, you know, a lot of good work in India political party watchdog such as ADR etc. They actually very painstakingly comp, no compile candidate level affidavits, and they bring out releases of what kind of criminal cases are pending against candidates what sort of financial background they have. But Milan's research actually found that it is not the fact that people did not know about the criminal antecedents of the candidates before elections. Their decision to vote or not to vote for them was driven by very different considerations. I mean they actually saw criminal candidates as no strong candidates and could potentially help them. So I think there if we look carefully at the data. First thing which it tells us or tells us to be very, you know, cautionary about it we should not jump to simple judgments in a country as diverse as India as big as India. Political results could actually be a result of lot more factors than the thing they are. That's my first thing, you know, we can talk about it later as we go on. If you have somebody looking carefully at the data, there's some nuance which you get in your coverage. Govind, you know, you mentioned you've given us some few examples. And I know that you all cover health, environment, gender, employment, I think for areas which you, for example, healthcare access, which is so important and gained so much of currency over the last two years. Now, if you can give us, you know, one or two examples of data driven exploration in social justice or rights issues, which highlights topics that are not discussed during campaigns. I mean, I know that most of the critical issues are not discussed during campaigns, but something that you feel that you all have done and can push then the conversation. Because you said that it improves the quality of public discourse. No, yeah, I mean, what I'm saying is that our mission is to improve the quality of public discourse. Now, to what extent we are succeeding, I don't know. I think, yes, it is a fact that there are many, many data inputs into the sort of melting pot of consciousness as it were. But is it doing anything or how much it is doing or do people vote intuitively in a different way? I think what people intuitively vote for we know, whether they vote for things that we believe are important from a fundamental point of view. And when I say fundamental, I mean health, education and environment, I'm not sure. And I think that the evidence usually kind of in a post facto way indicates that there is really no connection between many of these things. Or even if there is a change in political leadership, the political leadership itself does not do anything to really change matters on ground. And I think that's really the story of India in some ways over the decades. So let me give you an example of where we think we've had an impact. Maybe I don't think it's so much of so much social justice in the way you put it, Kishle. But at least I would say awareness about equity and it really comes back to the air we breathe. So about four years ago, we ran a project called Breathe. It was called hashtag breathe. And what we did was we basically, I mean, the first thing was to try and establish whether we had data on air quality in this country. And the answer was there wasn't. And the little data that there was was actually offline and it lay with people like Delhi Pollution Control Board or Central Pollution Control Board or whichever state it might be. But we were obviously looking at Delhi because Delhi was already at that point as it continues to be amongst the most polluted cities in the world. Now, and as we try to find more data, I mean, there wasn't enough data, nor was the data available real time. And so on a normal on a morning, if you were to get up and say, okay, what is the air around me like, you have no way of measuring. Now, this is before the advent of all those sort of more expensive air purifiers which come with built in air quality monitors and which I'm sure all of you who are from Delhi on this panel will have. But in Bombay so far, we've survived without. But so we what we did was we built our own monitoring stations using low cost sensors and built a complete built a device which was actually transmitting real time air quality data both PM 2.5 PM 10 and temperature. And we had a dashboard it was on breath.indiaspan.com where you could actually see it live. So you could actually go to Delhi and there were other cities as well and zoom in and you could see what the air quality at that point was or literally five minutes ago in Basant Kunj versus Dwarka versus Jhandevala extension versus Noida. And one reason it was Noida was because quite a few TV channel started putting our device on on the top of their rooftops and measuring it and then playing it back. So I think what the thing that happened after that you know till then a quality was an issue but it was an invisible issue. I mean it was literally enveloping everyone and yet people were not completely aware of the damage that it was doing except that you knew it because maybe children were coughing more or needed nebulizers and so on. I think what happened in that period and talking about 2016-17, 2016-17 towards 18 is that the awareness of the issue thanks to data went up dramatically for which we can claim some small credit. Obviously we put all our effort into Delhi because we knew that was where the pollution was most and people were also likely to notice the data I mean. And post which I think as you fast forward I mean the government had just started doing so but it then expanded its own suffer network for air quality monitoring. I mean one of our problems with that the government's network was that their monitoring stations were in very public places. For example traffic signals and junctions and we felt that you know you don't stay usually right next to a traffic junction or a signal or whatever you're normally a little inside. So readings can get distorted. I mean you can have much higher readings than what you would usually be experiencing let's say in a colony in Delhi or whatever. So I mean be that as it may I think we managed to bring this out. We also did a whole lot of other things on the social side. We worked with Twitter and you know you could just say hashtag breathe was uncunch and you would that the data would get pulled from the cloud and then it would flash on your Twitter timeline. So I think net net as we went towards to 2018 and like I said not only because of us the concept of data when it came to air quality became a page one or rather data became a page one data point. Right. So earlier on page one or whether it's in the times or whatever you know it would always be weather and maybe rainfall like in Bombay Mumbai where I am rainfall is a big obviously something you watch. Carefully in the monsoons but air quality has come there. Now this may or may not lead to the let's say equal effort on the policy execution side because the policy response I think I did has started coming around but has there been sufficient response in the policy execution perhaps not. But clearly everyone is aware of the issue. Everyone is now deep dive has deep driven into you know where is the pollution coming from the breakdown of pollution. Is it coming. Is it being imported from states outside Delhi. Is it because of construction dust. Is it because it could something else. So I think people are now aware and and I guess the role that people like us play is is to say that OK how do you make people aware then afterwards they can take their own decisions on what they want to do. So I think this is what perhaps the best use case I have in terms of sheer popularity and impact of where data that affected or had an impact on public life became. I mean we managed to cross the cross the boundaries and you know penetrate or the barriers rather and penetrate people's consciousness. I mean like I said it's happened along the way maybe with other people as well. Hindustan Times also did I don't know if Roshan you were there at that time but Hindustan Times was also trying to was also trying to do a similar effort in air quality. And I think they started almost at the same time but we were building our own devices so we could move faster. But to the credit of the Hindustan Times editor at that time I think it was Nick Dawes. They I think they also felt that this was something that they wanted to focus on. And Hindustan Times similarly has been a pioneer in attempting many such data projects which I will of course leave to Roshan to speak about not. But I think so the challenge therefore is how do you you know take issues that are important. Find the right combination of technology and visualization that will help people understand and grasp it the most in the most easiest way. This is I mean sometimes it's a it comes together. I mean like it did in this case because I met some environmentalist from Brazil who was doing similar work in the Amazon River in trying to measure and display. And some others in the United States who are doing some other work but not in air quality. But I think the challenge and the opportunity is to really find that happy confluence point where you can really show what is happening in a way that people understand it. And also in a way that it impacts them. You know you say AQI is yesterday was 300 and today is 900. Then you say wow what's going on here. You know and maybe you'll do something about it or maybe you will you know you will press or push your elected representatives to do something about it. Or like as we discovered in our own readings at that time or we would say something as simple as don't take a walk at 7am in the morning. Because that's the time the air in Central Delhi is the most polluted. If you do want to take a walk then do it at 12pm in the afternoon because that's when let's say the quality of air quality improves. Another factor is saying that air quality is linked to wind speeds. And we started showing wind speeds as well by the way because you could just by a simple API just pull it from I think from Google or somewhere. So if wind speeds are good then the air clears up. So you're also watching wind speeds. So you say okay nothing's going to happen on air quality but if the wind speeds pick up then I'm sure we'll have a nice day ahead of us. So of course this gets into a bit of a geek territory but it's something that you can experience and see for yourself. And need not be dependent on some data geeks sitting in India spend or wherever they are. That's why this platform is called has geek. They were the geeks who are monitoring this. Govind the question that I asked Roshan is that considering that majority of Indians and the Indian electorate are not watching us because this is not the language of the Indian electorate. Nor would they be reading India spend or Hindustan times. How do you convey what you are trying to convey to the rest of the people. So I mean there is a big challenge on that front but I mean I think all of us are making our own efforts. I mean you know India spend is also in Hindi and Tamil by the way. So it's not like we're only in English. A lot of people do read us in Hindi but will they read stuff like this are you know we have two kinds of audiences. One audience is obviously the public policy audience whether it's within government or outside academia and so on. The other audiences really are publishing. Those who read our articles via our publishing article publishing partners. So people read our articles because they may appear on first post or scroll or Hindustan. I mean not Hindustan times business standard Dow Jones news wires and so on. So we have a much larger audience than our portal. As you said I mean we prefer to call it a website but it's a much larger audience. But yes that audience is not the audience that if you're saying that UP is going to go into elections and there are 200 million or not 200 million people will not vote but let's say 150 million people will vote or I don't know whatever the exact number is but they're not going to be reading all of this. I think therefore it goes back to the original challenge of saying that how do we really take that data to those data points and convey them in a manner that that at least several million people will understand as opposed to the few hundred thousand or a couple of million that do today. So I'll have to I can only unfortunately answer till that point. All right thanks Roshan. You know we are talking about UP and I'm going to come to the my favorite question a little later which with reference to UP but you know as data you know pushes Indian political campaigns more and more the political parties are using data journalists in newsrooms are using data and you know with reference to UP I'm saying is there also the risk that it drags divisive cast politics with it because now we know how many say for example an entire booth if X percentage of people have voted the political parties actually can find out who have voted for them and who have voted not voted for them. So it's a danger right in some way and then they get disenfranchised. I mean in the next election we've seen we've had this example where that entire pocket of voters who did not vote for a particular party was were disenfranchised forever. You know I think there are two ways to look at it. You know the Indian society as we know it has always been divided on various lines in this class this religion and all that you know at some point in time class also used to be a big thing you know because there were conflicts between the haves and the haves. And these problems were there earlier also I mean if you remember you know when we used to do elections via paper ballots you know the election commission actually introduced the practice of mixing ballots for different booths at a very later stage. So earlier you actually know and that is when the results started getting delayed because they would actually mix ballots. So earlier also there was the possibility of political parties finding out with who did they get how many votes etc. What data and technology especially has done here is that it has allowed political parties to scale things on a completely unimaginable level and to come back to the question which was going. So the kind of data you know which we try and bring out in our work and the kind of journalism which we would like to see and believe you know people follow during the elections. I think political parties are actually running a completely parallel network to that. Almost all major political parties in India now hire political consultants. They do their own political surveys and I think the richer ones probably do rounds and rounds of surveys before elections right from candidate selection to what issue. So you know all of us heard about you know how you know iPads sort of help Mamta Menaji with the Dwara Sarkar thing and all that and it's just one example every political party does it. So and you know there is reason to believe you know of course we will not have substantive proof of it that political parties, especially the ones which have been in government have access to a lot of sensitive data, you know privacy concerns etc. We know how seriously they are followed or not followed in India. We have been hearing reports of welfare beneficiaries being approached specifically and targeted manner by political parties etc. So their data has sort of actually tilted the scales in a big way in favor of political parties were either in power or have a lot of monetary resources to hire those kind of people. That's one. Coming back to the question of data allowing I think you know divisiveness yes I think that threat exists. Technology has allowed that threat to be scaled at a different level. But I also think that it is ultimately the choice of the political party of the day. I mean if a political party wants it can do a survey on whether you know say improving the quality of government schools will let lead to more tail vegans attraction for it in say poor localities. I mean if it wants it can actually survey middle class localities where electricity bill debated etc can matter and of course I mean you can do it for all sorts of divisive purposes also. So I don't think data has actually sort of brought something new which did not exist in Indian political system or the electoral ecosystem earlier. What it has actually allowed political parties to do is scale things on a completely different level. So that way yes, you know the positive or the negative spin offs of it will be far greater than what they used to be separate 30 years ago. Go in what would be the drawbacks you know with the relentless news cycle that create for data driven investigative stories. You know the this 24 seven news creates a time crunch. So your stories or your topics remain alive only for a few hours or for a day. That's a challenge isn't it to hold readers attention because you spend a considerable amount of time you know doing these kind of investigations. Yeah I think we are not part of that race so and we've never we've not been part of that race for a while so we don't feel it. You know I mean if I maybe I as a consumer might be affected by you know 72 hour news cycles and so on. But as a producer or in our own little way are not affected by this at all. I mean the whole idea of starting this venture was to you know free ourselves from the cycle do one story a day spend anywhere between two weeks and two months doing stories. Works on projects like the equality one I just described and many other projects which take months sometimes a year to fruition in the way we wanted. And we've continued to do that I mean more I mean I guess the consumption of such material has only increased. And I think the and we do see that you know when we do something interesting or as others do something interesting the transmission does take place. You know the data is only building the argument on the back end if you want to use if I may use that term on the front end it's really the story that you see. You know if you know if I can convince you that I've done all this data crunching it back to present an argument to you that this is what the real picture is when it comes to let's say health care delivery or education or people's opinion about a certain issue I mean which is more survey based then you will believe me or the chances are that you will believe me because you trust the work that I've done on the back end. Now you don't have to understand what I have done at the back end necessarily you don't have to know that you know that here I mean know every line on what my of my Excel sheet and believe it and pass through it. So I think the consumption of what we say or what we've transmitted created and transmitted in the last many years has definitely increased I mean if indirectly if not directly. All I'm saying is that that the journey is still a long way to go because if you look at the population as a whole or the overall let's say electoral base of this country and because I mean your your the theme of your discussion is fight for democracy then we have a long way to go where people take decisions basis data and evidence and also I mean you may just like a vote you may just vote for someone because you like that person's face or what that person says and that's fine all we are saying is that to the extent that you can or if you could then also use data and evidence which people like us are working on so that your choices are informed whatever they may be. Roshan are there any recent examples of you know false narratives not false narratives very you know one can contest that term itself that has been created using data journalism. You're asking me to criticize one of our own. You know I mean it's a huge risk right when you say that oh these are the numbers as a layman I don't know where the numbers you're getting from. I think it's an important question. See I would not attribute motives to people because ultimately especially our peers in the profession I would rather not do that. It is eminently possible that two people would come to very different conclusions based on the same data and I'll give you a very simple example. For example you know in India you know I mean almost everybody agrees that there's a problem in agriculture. The real debate is what is the problem and therefore what is the solution to it. A lot of people including very senior journalists argue that India actually has a problem of oversupply in agriculture. You know that we had our share of problems you know supply constraint before the Green Revolution started and we would have to import food grains. Now we actually have like every year we do a surplus and every year we record breaking food grain production. Therefore a lot of people say that actually there's a problem of oversupply in agriculture. Now if you were to accept that diagnosis of the problem in Indian agriculture. Logically forget everything what I mean that person or that argument is saying is that you actually have to target negative growth rates in agriculture. Because if you have a problem of oversupply then why do you want to increase production further. Now that's quite bizarre and of course people who make these arguments stop short of saying so. But then you have to look at other parts of the problem and I personally have been arguing this for a long time that you know the problem is not oversupply. The problem is that a lot of an overwhelming majority of Indians actually do not have enough to eat in this country. And the fact that the government has been distributing free food grains to around 800 million people. Now the government has been saying this continuously that they've been giving 5 kilogram of food grains to around 80 crore people in this country after the pandemic. It's the biggest proof of that that people's incomes and purchasing power are so volatile that they can't even so low actually that they can't even buy 5 kg of wheat and rice after the pandemic. And you know anybody who goes on the field etc will be able to clearly say that a large number of the poor actually do not eat what a respectable person would call it that. So this is a good example of how on the basis of same data or I would say on the basis of looking at selective data you can actually reach a very long conclusions. So of course and this is something which you know the last time you had me in the journalism talk I made this that it is very important to have domain knowledge and knowledge of the context of the data which you're talking about. Otherwise you can actually I mean it is completely not surprising that you would reach wrong conclusions. And Roshan what is the data telling us about the coming elections. Well that you know it's very difficult because a lot of poll surveys happen. Unfortunately most journalists do not have access to that real time data. Not that they have a very good track record per se you know but but at least you know the only pollsters who have been making their data transparently available for research and academic purposes have been the CSDS local media people. We also have to understand that they ultimately are a quasi academic Institute, you know they do partnerships and all that with some media organizations to sort of fund the thing. But you actually have a lot more people in the game now who are actually who actually do this for purely commercial motives. So, for example somebody like Axis who has had a reasonable track record in predicting elections. This is not their bread and butter the bread and butter is actually proper polling. They do it for business houses, corporate etc. So if somebody has no access to significantly large resources, you know, I mean basic thing in statistics is the quality of the sample. I mean, if you give me one lakh rupees to do a survey and if you give me 10 crore rupees to do a survey, you know my sample is going to be I mean definitely much better than yours. So if these polling houses where to transparently share their numbers with journalism, I think the public discourse would be rich in India, as far as understanding what goes into elections is concerned. Unfortunately, that's not the case. That is not something which we as journalists can actually help because ultimately those people, you know it is between the people who do the surveys and the people who pay them. So there I think we would do a lot more better if we actually had access to that data. Otherwise, you know, it's, I mean my guess is as good as yours as to what is going to happen in elections. Can I come in on this? You know, I think that I mean we should distinguish between surveys and polls and the quantitative quantitative data that I'm assuming we're talking about otherwise. You know, when we talk about health outcomes or education outcomes or climate environment, you know, we're talking about quantitative data. This is not opinions about, you know, who will I vote for or who do I think is a great politician and so on. So I mean, I mean, when you look at large quantum of surveys, I mean they become data in their own right. And rightly so, I mean these are all, I mean I'm assuming scientifically run, not assuming, I know, I mean we all know that scientifically run the standards are global standards on how samples are selected and so on. So they're still surveys. It's not the same. I mean, I think we should make sure that we draw a small line of distinction between these two worlds. Yeah, absolutely. The only thing is like a big line of between the two things. Yeah, sorry. No, I completely agree with Govind there that I mean we have tangible statistics, you know, unfortunately, it's becoming the case that we have less and less of those statistics in India today. But when it comes to elections because we're talking elections, the two need not always sort of, you know, be congruent with each other. Now, for example, the BJP won Uttar Pradesh after demonetization that no matter which tangible data you look at it will tell you that demonetization led to economic pain. So there I think what makes elections complicated is that you might not have a correspondence between the tangible and the non tangible, which is political preferences in this that's my limited point. Yeah. Govind the other thing that I wanted to I mean because we don't have a lot of time left and I want to bring in the audience out here. Also people watching on the YouTube live stream, you can type your questions in so that we can get the question and ask the speakers but my personal question. Govind is that, you know, whenever we talk about say for example, if we are, if we've referred to elections in today's show, we've referred to Uttar Pradesh, at least twice or four times, and we have not mentioned Manipura and Punjab. Whereas Manipura and Punjab is much easier in terms of discussion through data because they're much smaller states. I mean Manipura's total population I mean look at wouldn't be would be a district's population in UP. And we also know that these smaller states and the borderlands of India are constantly falling off the map. Is there a way to where data journalism helps to actually bring these smaller states and falling off the map area, the shadow areas into mainland discourse? I don't know if data journalism specifically helps Kishilay, I mean I'm not able to sort of get my arms around this one. But I think, you know, when you use data to tell stories, I mean what you're trying to do is to make the argument more substantive. You're trying to build a case where in, and perhaps sometimes disprove a case, right, you may have a certain belief. Let me give you another example. You know one of the big statements that emerged after the 2012 and I think there was another case in 2014 rapes in Delhi. The first was the famous Nirbaya case, the second I think was it happened in an Uber car, was that Delhi was the rate capital of the country. And I think I remember we started looking, my colleague started looking at NCRB data, that's National Crime Records Bureau data in 2014 and found that at that point, Delhi was not nowhere near, but it was definitely not the top city when it came to a number of rapes in proportion to population. Actually it was Gwalior, that was the number one city that year. You could rightfully argue that, you know, maybe cases are under reported in Delhi and over reported in Gwalior and so on. But that's, I mean, but the fact is that the primary assertion that Delhi was the rate capital was not the correct one. But that was the popular, if you remember, and continues to be actually, it continues to be the popular perception. Now, which is, now how does this affect things, right? So when, if you remember again, go back to 2012 and if you remember the public outcry that followed. Everyone said Delhi is the rate capital and we need to have a greater policing or better quality policing. We need to have more women in the police force and so on and so forth. So now when something is driven by perception, your policy outcomes or responses can be a little different than what the right or what the, I mean, your prescription policy prescription can vary from what maybe it should be in an ideal situation. For example, you may say that, oh, Delhi is the rate capital and therefore we should have more women in police stations, right? Now that in itself is a good thing. But maybe you also need to have more women in police stations in Gwalior. And I think there was another state in Madhya Pradesh, Indore, I think, which was also ranking quite high up. Now those numbers have shifted subsequently as, you know, overall rape data or crime against women data has shifted as reporting has obviously increased and improved. But the fact is that you're tempted to take policy decisions on the basis of perceptions rather than data. So now if I were to go back to your Manipur question, you know, you've got it, whether we have a specific understanding for Manipur or not, if as long as we are applying these metrics to make policy decisions, whether it's Manipur or Delhi or Maharashtra or UP, then I think we are better off. And equally, if a certain perception is building up in public and you can't blame them, then we should counter that as public policy people and including in government, you know, maybe at that time no one had the guts in government to go and tell people that listen, you know, Delhi is not the rape capital because obviously, I mean, you will be booed out and so on. But the fact is that at some point, those in policy and in government should have tried to draw these distinctions and say, OK, I mean, let's let's attract this problem on the basis of existing data. And now let's build a policy framework to address crime against women. Right. The other and the last point on that, as you know, most crime against women more than 90% of cases. Now, this is commonly known. Okay. Even six, seven years ago, it was not commonly known is the perpetrator is known to the victim or the victim is known to the perpetrator. So now if you say that Delhi is a rape capital and you want to put more women in police stations or you have more night patrolling, but if more than 90, 95% of cases, the victim is known to the perpetrator, then aren't you trying to solve the wrong problem? Again, not to say that you should not have more patrolling or women in police stations. By the way, this statistic about 90% plus is consistent with many other parts of the world. India is not an exception. So I mean, just to highlight the point that you need to have data for any context, whether it's a small geography, like a small state, like you pointed out, a large or a horizontal issue like women's rights, like crime against women, and ensure that policy or public policy is informed by that and everyone is aware of that. And we as citizens also play some role in this in terms of asking the right questions or in demanding accountability and so on. So let me stop there. Alright, thanks. Roshan, you know, same question, but if I can put it more because you are on a, you're in the newsroom, where would you, I mean, how much of, how much of coverage do you, would you do on a state like Manipur, which is also going to election with UPA, because I haven't seen any of the newspapers. I don't read the Hindustan Times, sorry to confess that. Morning, I mean, I just reduced my newspaper and television of viewing, but I haven't seen much of Manipur or even Punjab actually on the newspapers. I think to be fair to news organizations, the repotage representation which states get is my large proportional to their importance in the parliamentary arithmetic of the country, which is why, you know, it's absolutely not surprising that UP gets more, which is why Bengal got more last time. But to come back to your question, no, I think that most people in this country, and I'm not even talking about, you know, the relatively underprivileged, even people who are otherwise well versed with say the popular discourse or the journalistic discourse in this country, actually know very little about the political diversity which sort of comes with the regional diversity in our country. They know very little. For example, and I'll give you a Manipur example, Irom Shamila contested the elections last time. You know, if this one person from Manipur, which a lot of people who follow the news in India know it was Irom Shamila, I mean, they probably know her more than they would know politicians in Manipur. And the electoral thing was, frankly speaking, quite a disaster. That tells you something about politics in that state. When it comes to supporting Irom on issues such as say opposing the half spas, he actually gets support which is across the political spectrum. It didn't translate into her electoral support. That tells you how politics in that state works. I'll give you another example. You know, Kerala is the only state where the companies actually run a government today. You know, if you go via the stereotype play again, you'd actually think that Kerala is full of revolutionaries. If you look at the 2019 results, the CPI and the left suffered huge reverses. And I would like to believe that was primarily on account of the Sabarimala issue, because the LDF government there took a stand supporting entry of women in Sabarimala, which again you would think that is a very progressive thing to do and therefore will have a lot of traction in Kerala. They actually made an intelligent retreat from the issue. Thanks to Supreme Court stay on that thing. I think election results are a very useful data set in themselves. So they actually tell us a lot about how society behaves otherwise and behaves politically. So this in itself is a very important insight for me. If you were to do nothing and just sort of compare the news cycle rest of the year and the election results, I think you can make some very useful deductive judgments. Thanks. You know, I'm going to open this for questions. I'll already see up a hand go up. But if I can, before I come to some use Samyokta, can I request my colleague, Dr. Richard Asen, who actually teaches data journalism to our students, because she's in chat typing something out there. I can't look at the screen and look at the chat. So I am handicapped in that. So if you can put your question. Well, my question is essentially about multitasking. So, you know, I mean, for data journalism, since we're teaching it, there's this book that's come out whole numbers and half truths, which I think really covers a lot of issues of data journalism in India. And Rukmani, who's the author, she basically compares household surveys to NCRB data. And she develops this, you know, nice two dimensional metric, where she shows states, which have high levels of household, I mean, they have reported it household surveys that there isn't, you know, high incidences of rape, but it isn't captured in NCRB data. And meanwhile, there are, you know, places that have high reporting. But, but you know, that's also captured. I mean, you know, the household surveys has it less but in the NCRB data it's more. So she basically shows that Delhi is overreporting, whereas there are many other places like Bihar, which have underreporting. So I was asking, do you think it's a good idea to develop a culture where we, you know, we shouldn't comment on data led or data driven ideas without looking at two or more data sources to make any point because one data source might be misleading. And also the other question I had was whether there's a role of open government activists and open data activists where journalists can partner with academics and with civil society. And who do you want? Who are you throwing this question to? Oh, to both, to both, sorry. Okay. Yeah, yeah, okay. So I think the answer to the second one is yes. And I think we are already embarked on that as we speak. And we've always been, I mean, from day one, we've always been keen to be part of or to host hackathons and the like. I think there's a separate discussion on is there enough free data now available in India as it used to be earlier in terms of the trend lines. I'm not sure it is now. I think there are a lot of areas where now there seems to be a slowing down of data sets. And it's clearly not available in the areas that you want. There's a lot of the data available, which you maybe don't want. But when you start really looking for employment or jobs and so on, I think there are some very valid questions. Also, you know, there are a lot of delays in, let's say, census data. Right now we are in 2022, we should be looking at another census. And the last Union budget talked about digitizing this process and, you know, using this digital kind of delivery to collect data and so on. Nothing has happened as yet, obviously, because of the pandemic and so on. But I mean, but what it does mean is that there are some very, very critical, large data kind of sets which are missing from our discourse, as they should be. So what do you substitute that with? You know, what are your second and third options? And it's a tough question. It's not easy. I mean, let me give you an example when we were trying to respond to this question of toilets that were being built by the government, right? So government was saying, government was saying that we've built 50, 60, 50 million, 60 million, 80 million. I think it went up 100 million toilets that were built. So now how do you show that? So we try to look at data. The data actually was funny because if I remember correctly, the data was showing a level of construction which was almost unimaginable, right? It was showing that we were constructing at X number of toilets a day, which by any stretch of imagination looked very large. But there was no way to counter that. So you had to believe it. I mean, you had to accept that maybe, you know, every state government, every whatever was pumping in money, effort, people, materials to build those toilets. So we went on the ground and we started taking pictures. And that's when we discovered, I think as many others did, that many of these toilets were not really toilets. I mean, they were being used as storage rooms for cement bags or something else, or basically not being used as toilets at all, or even the toilets were not built in the way that they should have been. There was no proper drainage system. Even the basic construction was flawed and so on. So when we say data journalists, I mean, maybe this goes back to the primary question. I mean, it's data and journalism. I mean, we have to combine the both, combine these two worlds, right? So what we use data to build the initial kind of set of arguments or data or positions, and then we go on ground and see what's going on, or we do that simultaneously. And that applies to every story we do, by the way. And that's how data journalism is really practiced. It's not about just sitting in front of a computer and, you know, that's one part of it. So I think where we can find multiple data sets or cross-referensible data sets, we should absolutely do that and we do that. But sometimes you will not find it. And oftentimes you have to back it up with classic journalism as we always do, but so should others. But I think to your larger question of can everyone do it, I doubt. I mean, I don't think people have even time for the primary set of data that I talked about. I mean, forget everything else. So I think the only option is to say, can we create a culture of asking questions? You know, if people ask questions, then at least hopefully you will get some answers and at least that means you're, you know, there is some critical thinking involved in before you form your decision, whether it's about voting for someone or forming a conclusion about what someone has done for you or not done. I mean an electorate representative since we are talking about elections and democracy. In fact, the toilet data story is my favorite for understanding and explaining data and showing how problematic it is because in a large number of women actually did not want toilets because that kind of restricted their mobility. The only mobility that a lot of women in North India had was going out of the home. And bringing toilet into their homes meant that they could not step out of their homes. And, you know, the toilet data story initially never obviously wouldn't have looked into this because they were just looking at numbers. So we need classic journalism to follow it up. Yes, I think the Ruchira's question was for Roshan also. So more than one data set always like I said, you know, you can, when I gave the agriculture example, you can reach very different conclusions depending on what data you are looking at or not looking at. As far as the question of a larger engagement and collaboration is concerned. Some of that has been happening in India. I mean, look at the energy is for example, you know, the government started like putting up online master roles and all that. The entire Jansunwai culture, which was built around examining master roles, I think was in a way an organized public effort to question official data, you know, and they were genuine problems which are found to be with during COVID a lot of people got together and we built that public database which was hosted on a public website. A lot of journalistic coverage actually was possible because of that database because it would have taken far too many people than what a typical journalism has a newsroom has. And this brings me to the larger question which Kishale asked Govind earlier also, you know, when there's so much news, this 24-7 news cycle is mad rush. I mean, can data journalism win this battle? You know, newsrooms are a daily battle. You come to the newsroom, you wage a battle every day and you go back to this no declaring victory, so to say, because newsrooms are institutions, they're not about people. But I think as, you know, this internet takes over the news cycle and, you know, the breaking news part has already been taken out of I'm talking business of journalism here. So if I am a newspaper or if I'm a website, you know, I have to make money to be able to pay my bills, pay my people. So when you're not in the breaking news cycle anymore, you know, and once in a while there are investigative stories, but if you are a newspaper or a website, you can't do investigative stories every day. I mean, it's very difficult. I think there if you have a dedicated data team which like Govind said, you know, the basic challenge is not to make graphs or beautiful visualizations. The basic challenge is to ask questions and then seek their empirical validity or lack of it. That I think in today's environment where every newsroom is actually trying to go behind a paywall, you know, try and compete with a Facebook or a Google in, you know, garnering as much of the digital revenue as possible. And that is only possible if you have some exclusive content. There I think data journalism and this entire approach of a data driven thing, especially in the polarized times we live in India today, where everything you say basically has an opposing view depending on your political persuasion is a useful thing to have for newsrooms. Thanks so much Samyukta. I'm guessing Samyukta is one of our students right Samyukta. Yes, Professor. Good evening, Professor. Good evening, sir. So my question is, do you think that data can, you know, change or influence an election. More in the western countries than in India, because you know, even if what I feel is even if you have a lot of data, you show right now how much unsafe UP is during Adityanath regime. I think something that plays more there is caste politics and religion politics. So I have this question that I mean, do you think it is more influential in the western countries where literacy rate is more Russian you want to go first. We are not accused the Russians of hacking our election so far so that way, you know, you know, I can say this with a lot of confidence, our election system is far better than American election system. You know, there they are talking stuff such as voter suppression. These things we have left I think in the 1980s. So literacy or lack of it is, I mean, ideally everybody should be literate and educated but literacy or lack of it, I do not think is a very important sort of Nobel weather of the health of your democracy. As far as the question of public discourse is concerned, see journalism, journalists and newsrooms can only go to a certain extent, you know, we can't dictate terms to the society. The society thinks that, you know, there's value or no premium in no making political decisions on no other basis than what we think are important. Then, you know, a political process has to engage with that. I mean, you can't, you know, there was a point of time in this country when you could have, you know, invited political backlash for sending women to school. I mean, that still happened in Afghanistan, which is why Malala got a Nobel Prize. So society has evolved over time. I think democratically we are a very mature society today. But on the larger question of whether it should be Roti Kapra and Makhan and nothing else, I think it will take a lot more politics to get there. I mean, we are still a very young democracy and we're just celebrating the 75th year of our democracy. I think, you know, when we kind of try to craft our mission, not that we spend much time doing it, but you know, our objective was to blend data with emotion, you know, so emotion drives everything. You know, and if one of the things I was in a way influenced by was the Anahazare agitations and what the Anahazare agitation showed us was the power of emotion. It could bring people out on the streets. It brought young people out fighting for, you know, change in change in the way we live our lives fighting against corruption and so on. And one of the things that struck me was as a flip side that if people actually, I mean, the Anahazare movement is, you know, I mean, in a way it fizzled out because part of it coalesced into Ahmadi Party and became the political force and the rest was gone, you know. But the interesting thing it showed up and on the flip side was that people could actually come out and fight for something. Till then, if you were an oldie like me, you were quite, you know, we all thought that this would never happen in India. That never would you have young people coming out whether connected with social media or otherwise and, you know, fighting for a cause in this case being corruption against corruption. So, and that's what, I mean, one of the things that got me thinking at that time was that if all of this could be blended with, let's say, 2% of data or 5% of data, could we have even more powerful discourse? Could the questions we ask be even more substantive? And that's, I think, really the question that we need to leave ourselves with. I mean, how much of data can influence or will influence? We don't know. I mean, people like us have taken it up as a cause and therefore we will continue to do what we're doing. But will it actually influence elections? Can it influence elections? I have no idea. I mean, I have no way of saying that what we've been doing all these years has had any more impact than not. But yes, at a certain level, at a certain section, we definitely, yes, we know that. But you look at the 2016 elections in the United States. I mean, I don't think data had anything to do with the election of Donald Trump or it had everything to do with disaffection of white voters, misinformation, fake news, technology and targeting by using analytic software like Cambridge Analytica. I mean, all of you said nothing to, you know, I mean, I don't know whether what role anything had to do with anything in that. I mean, it was completely another world. So, and people continue to believe so many things, including the fact that the election was stolen right now in the United States. Right. And so, so what is the role of all the work that people like us do? I don't know. And there are people like us in the United States as well. Some of whom we've looked at as models to emulate. But is that really affecting or changing public discourse in some ways you could argue it's only gone worse in this in this period. And you would be right. But we have to keep doing what we're doing and hope that, you know, like I said, if we can keep trying to improve the blend of data with emotion, because emotion will always be the stronger, the stronger ruling force, then hopefully we've done our job. Oh, yes, absolutely. I mean, in my long years of journalism, even if I can count five stories which has had impact, I would be very happy. But we got to carry on for that long period to get even if one life changes. That's that's that's useful. I think, Archal Poddar, do you want to come online and ask the question, Archal, because you typed it there. Are you there, Archal? I think there's a question in the chat box where she asks, how much has data journalism penetrated in Indian journalism today because Indian journalism. I think what what she's trying to say is that since Indian journalism is driven by popular perception and it's demand driven or clickbait is data journalism, does it follow a similar model? So Roshan probably would be a better place to answer this. Well, there's nothing wrong in journalism being demand driven. I mean, ultimately, you have to make revenues. Like I said, you know, the future of, you know, genres such as data journalism lies in the fact that eventually every newsroom will have to offer something exclusive. So think of data journalism as your exclusive feature writing team, you know, as your best sports writer and those kind of things. You know, it's not rocket science. You know what I mean, if you claim to be a journalist, it basically means that you do work which will be understood by the masses. It's not meant for a small minority of very sort of the kind of people. So there I think data journalism has a future and you know, increasingly more and more newsrooms, as I can see, are engaging. They either have a team or they at least try and engage with data there. So I don't think all's lost is that those clickbait things will always be there. But I, I also think that this, you know, I mean, internet has done bad things to the, I'm a journalistic profession but it has also done good things to the profession. I mean, I mean, Kislev when you started your journalistic career, and when I started my journalistic career, the amount of information I have at my disposal in real time, and the amount of convenience it gives me in story writing is immense. I mean, this is all because of the internet. So we have to live with the negative side effects of it and you know, and like any other profession, try to make the most of what are the positives which is given us. Fantastic. Thanks so much, Govind and Roshan and others. You know, I mean, we three are journalists, so I think we can only talk about stories and storytelling and storytelling which can be, which can be made better, more compelling using numbers, using visualizations, trying to remove the haze of misinformation, as, as, as Govind says, and I'm sure there are other people who will tell you, who will tell you better about data protection, which is, which is the other big concern, particularly during elections where your phone data gets transformed into electoral data where cast, religion, and even your income gets mapped through your electricity bill. I mean, you know, did we ever imagine that our electricity bill is giving away our entire data to political parties and political campaigners and how e-voting or blockchains, or whether that's going to be of any use or whether that's going to have more risks, but those are for others to really comment. We are wrapping this up. I think Shantel probably will have the, you want to. Right, just to thank everyone and say please come for our session on the 22nd. We will be talking about electoral bonds and we'll have Nitin Sethi and Jagdeep Chokar. Okay, and one last thing for the audience that when you go to India spend or when you read Hindustan Times or any newspaper, anything, please subscribe, please contribute because if you do not pay, you cannot expect quality information. So, and it costs you less than a cup of coffee in your nearest cafe. So, thank you. Go ahead and add that. Thank you so much. Thank you so much. Bye-bye. Bye-bye. Be safe everyone. Thank you.