1
00:00:00,780 --> 00:00:10,610
Thanks for coming to what is now the fourth or fifth one of our esteemed development seminars.

2
00:00:10,620 --> 00:00:20,700
This time we have 1 or 2 more left on the program, so have a look, um, online to see what you might be missing if you don't come to the next year.

3
00:00:21,690 --> 00:00:29,310
Uh, but today I'm very happy to be able to introduce, um, Vivek, uh, Srinivasan.

4
00:00:30,240 --> 00:00:35,670
Um, Vivek's, uh, program manager of the program on Liberation Technology at Stanford.

5
00:00:36,450 --> 00:00:40,620
And he used to work as an activist, the India's Right to Food campaign,

6
00:00:40,860 --> 00:00:46,200
and what he was doing that was trying to make sure that everyone in India has access to basic public services.

7
00:00:47,220 --> 00:00:56,280
And that then led him to his PhD at the Maxwell School on understanding Tamil Nadu success in providing basic services.

8
00:00:57,240 --> 00:01:02,650
And then that work is in his recent book that's actually published by Oxford University Press.

9
00:01:02,670 --> 00:01:06,810
It's called Delivering Public Services Effectively, Tamil Nadu and Beyond.

10
00:01:06,960 --> 00:01:10,050
I think it just came out, uh, a few months ago. He has a copy of that.

11
00:01:11,340 --> 00:01:17,390
Um, I was going to say a bit about the topic that Vivek's talking about, um,

12
00:01:17,460 --> 00:01:22,680
combating corruption with mobile phones, but I think I'll just leave that for you to talk about.

13
00:01:23,220 --> 00:01:28,290
Um, we will we'll talk for about half an hour now.

14
00:01:28,320 --> 00:01:33,960
We'll have about 20, 30 minutes for questions and conversations afterwards.

15
00:01:34,590 --> 00:01:45,659
We have some drinks outside for everyone to be able to partake in and mingle and ask any questions that didn't get asked here and then.

16
00:01:45,660 --> 00:01:51,959
I think some people might be going to the pub afterwards to continue some of those sessions, but for now, let me just hand over to Vivek.

17
00:01:51,960 --> 00:01:56,250
So thank thank you for coming. Thank you. Thanks, Mark, and thank you, Ben for organising this.

18
00:01:56,250 --> 00:02:02,010
It's fantastic to be here at Oxford and with especially for a who is podcast, among other things.

19
00:02:02,010 --> 00:02:08,640
I've been listening and so I've been consuming many of your work, um, in my very long commute in California.

20
00:02:09,180 --> 00:02:10,559
So great to be here.

21
00:02:10,560 --> 00:02:19,320
So what I do is to talk about the ongoing project, which is two years old, but it's fairly young, um, which I'm doing in different parts of India.

22
00:02:19,800 --> 00:02:25,740
So talk about the genesis of the project and also some of the tweaks that I had to do along the way in order to make it,

23
00:02:26,250 --> 00:02:30,630
um, a little bit more effective, or at least to give, uh, give it a greater semblance of promise.

24
00:02:31,740 --> 00:02:39,720
This would be, um, Genesis is in my activist past as, um, someone who was working with India's Right to Vote campaign.

25
00:02:40,170 --> 00:02:44,700
The campaign was very effective in getting a series of different legal entitlements on the right to food.

26
00:02:45,210 --> 00:02:45,630
Um,

27
00:02:45,870 --> 00:02:52,439
the first legal entitlement of the campaign was able to get was to make school feeding a legal right for every child going to public school in India,

28
00:02:52,440 --> 00:02:58,829
which now covers about 130 million children. But the biggest of them came with what's called as the Employment Guarantee Act.

29
00:02:58,830 --> 00:03:02,730
Uh, has anybody heard about India's entrega, the National Rural Employment Guarantee Act?

30
00:03:03,600 --> 00:03:07,499
So let me give a very quick background. It's it's a kind of a workfare program.

31
00:03:07,500 --> 00:03:13,950
What it does is that any person living in a rural area can apply for work with the government, and they should be provided work within 15 days.

32
00:03:14,520 --> 00:03:19,440
The nature of work is casual manual labour. So you do have to work and you're given minimum wages.

33
00:03:19,860 --> 00:03:28,020
Um, which is about uh, I mean ₹150 today, which is probably around £2, uh, for a very hard day of work.

34
00:03:28,410 --> 00:03:33,719
So typically the only people who come for this kind of work is anyone who doesn't have other things to fall back upon.

35
00:03:33,720 --> 00:03:36,390
And the family can get up to 100 days of work in a year.

36
00:03:36,780 --> 00:03:42,810
So it's really seen as a fallback option so that people don't have to migrate in distress or they don't have to fall hungry.

37
00:03:43,830 --> 00:03:51,240
But it's also become one of the largest welfare programs anywhere in the world, uh, with a fairly substantial allocation in India's national budget.

38
00:03:51,600 --> 00:03:53,549
So why are you even getting these entitlements?

39
00:03:53,550 --> 00:03:58,890
We were also very conscious that if you get these big budgets, but if it doesn't reach people, it doesn't mean anything at all.

40
00:03:59,400 --> 00:04:04,950
So the method that we started following was from the success of what is called as the right to information movement in India,

41
00:04:05,520 --> 00:04:12,509
and they followed a particular, um, process, which is that they would go to a village and according to India's information law,

42
00:04:12,510 --> 00:04:15,300
they would demand all the records for that particular project.

43
00:04:15,630 --> 00:04:19,200
So if it's, let's say welfare project, they'll say, give me a list of all the works that have happened.

44
00:04:19,590 --> 00:04:27,120
Give me a list of all the people who worked in it, how much money was each person paid, what assets we created, etc., etc. so you obtain the records.

45
00:04:27,120 --> 00:04:31,020
Now all the records are written chronologically, right? One of the accounts chronologically.

46
00:04:31,530 --> 00:04:37,530
But when you want to verify the records, which is the object of this particular exercise, it's actually a very long process.

47
00:04:38,250 --> 00:04:43,170
Uh, but let's say that even though we do have a strong right information law in the country.

48
00:04:43,170 --> 00:04:51,330
So if you look at just the legal, uh, uh, aspect of what kind of records citizens can get and what records are exempted,

49
00:04:51,690 --> 00:04:54,269
India has one of the strongest right information laws in the world.

50
00:04:54,270 --> 00:04:59,250
But despite that, we often have to go on a protest like this one in order to get the records.

51
00:04:59,310 --> 00:05:03,240
Um, it's. Especially if you're really challenging, uh, the people in the administration.

52
00:05:03,750 --> 00:05:07,110
Once you get the record, this is what it looks like in the welfare program.

53
00:05:07,110 --> 00:05:11,310
For example, this is called the muster roll out, a list of attendance, someone who attended a project.

54
00:05:12,240 --> 00:05:16,950
You have to, um, you get reams and reams of muster rolls for any given project.

55
00:05:18,090 --> 00:05:22,470
Now you have to do a simple exercise, which is that we typically take the name of the first person,

56
00:05:23,070 --> 00:05:26,370
flip through all the pages of records, find out how many days this person worked,

57
00:05:26,370 --> 00:05:31,110
take the second name you flip through all the page pages again, and then, um,

58
00:05:31,320 --> 00:05:35,930
ask them to be prepared for a simple question that you can ask an individual.

59
00:05:36,210 --> 00:05:42,750
According to the records, you worked in this project for so many days, and you have made so much money, and it's a very laborious process.

60
00:05:42,750 --> 00:05:48,810
And this was one social audit that I organised for, uh, that preparatory work that you did for one audit of 25 villages.

61
00:05:49,230 --> 00:05:57,210
So I had to house and feed about 100 people for a whole week just to prepare the records before we could go to a household and ask people question.

62
00:05:57,930 --> 00:06:03,389
So thanks to this process, and it requires a lot of knowledge in terms of asking for the right papers.

63
00:06:03,390 --> 00:06:04,710
So to give you an example,

64
00:06:05,160 --> 00:06:11,520
it's possible for an official to give you all the correct records and give and just not give you a photocopy of the wrong papers.

65
00:06:11,970 --> 00:06:19,170
So you need to know, um, just to compute the whole amount, um, number of days that people have worked and see if it matches with the total.

66
00:06:19,620 --> 00:06:25,679
Uh, you need to ask. You just can't go and ask an official to give me all the records you need to know, uh, give me the muster roll management book,

67
00:06:25,680 --> 00:06:31,319
etc., etc., which means you need to know the program so it requires a lot of knowledge and skill to do a social audit.

68
00:06:31,320 --> 00:06:33,660
And it also requires a lot of preparatory work to do.

69
00:06:33,660 --> 00:06:40,080
One, thanks to this, I felt that even though the social audit process seemed to be very powerful,

70
00:06:40,470 --> 00:06:44,730
powerful to the point that a lot of social audits and the public hearings that follow them,

71
00:06:45,330 --> 00:06:50,700
um, you find people who have been making money for a long time coming and returning money in a public meeting.

72
00:06:51,180 --> 00:06:57,960
Um, typically you find that officials were indulging in corruption, um, get very, very threatened during the process of a social audit.

73
00:06:58,290 --> 00:07:00,600
This is an experience most auditors have.

74
00:07:01,170 --> 00:07:07,950
But the process is so cumbersome, and you need skills that you often have external people then coming and doing audits that you rarely have,

75
00:07:07,950 --> 00:07:12,480
audits repeated in any given village, um, multiple times over many years.

76
00:07:13,110 --> 00:07:20,460
So a lot of um officials have told me that I know, okay, you're coming and you're shaking me up today, but you won't be here forever.

77
00:07:20,910 --> 00:07:23,520
So which means that I go back to my ways. So why even do this?

78
00:07:24,750 --> 00:07:32,250
So it is in this context that I started feeling that the, um, even though social audit is a powerful, um,

79
00:07:32,310 --> 00:07:36,690
there are there's a lot of road work that happens which could be done easily using

80
00:07:36,690 --> 00:07:40,440
technology in a context where e-governance is taking off in a large way in India.

81
00:07:40,770 --> 00:07:47,249
So all these accounts that are getting available digitally in many programs, many welfare programs in particular, simultaneously,

82
00:07:47,250 --> 00:07:54,059
there's a mobile phone revolution with, uh, a very large proportion of households in rural India getting access to mobile phones.

83
00:07:54,060 --> 00:07:58,650
So even in one of the poorest states in the country, about half the households have mobile phones.

84
00:07:59,190 --> 00:08:03,450
Um, as the Census Commission put it, there are many more households that have mobile phones today than toilets.

85
00:08:04,200 --> 00:08:04,739
Um,

86
00:08:04,740 --> 00:08:13,680
so that presents to me an opportunity where I can access public records online and disseminated to people on a continuous basis using mobile phones.

87
00:08:13,680 --> 00:08:16,170
So which is the attempt that I'm doing in this particular project?

88
00:08:16,620 --> 00:08:24,420
Um, so what I will do is to go through some of the challenges that I've had, um, and also explain, um, how we are trying to tweak the process.

89
00:08:24,750 --> 00:08:28,020
So it's happening now in four different states in each place.

90
00:08:28,020 --> 00:08:34,620
We work on a different program, uh, Employment Guarantee Act being one that, um, because it's implemented across the country.

91
00:08:34,620 --> 00:08:35,610
That's something we work on.

92
00:08:36,090 --> 00:08:44,549
Um, the second program that we focus on a lot is the public distribution system, which is a way of giving subsidies, rice and other grains, um,

93
00:08:44,550 --> 00:08:51,780
to the population living below the poverty line, which is also a national program, but we work on it in certain states due to data availability.

94
00:08:52,770 --> 00:08:58,559
Let me pause here for a moment and take a short detour, which is that in the context where I'm trying to do this kind of work,

95
00:08:58,560 --> 00:09:07,170
that had a bunch of people who started asking the question of whether transparency works in the first place, um, to name three recent naysayers.

96
00:09:07,260 --> 00:09:12,090
One of them, um, um, a colleague about whom we were talking about a moment back, Jemmy Weinstein.

97
00:09:12,750 --> 00:09:19,980
Um, so these are scholars. All of them are political scientists have done randomised experiments, um, using a transparency measure.

98
00:09:20,430 --> 00:09:24,270
Jeremy Weinstein and Humphreys, what they did was to create a scorecard of,

99
00:09:24,270 --> 00:09:29,280
uh, performance of parliamentarians, um, in I'm not sure if it's in Nigeria.

100
00:09:29,520 --> 00:09:32,190
Um, let me go to country for the moment.

101
00:09:32,490 --> 00:09:39,090
And then they distributed, uh, in a very, very, very large, um, um, scale in the country in which they are working in,

102
00:09:39,090 --> 00:09:43,860
uh, and they wanted to see whether this form of transparency changes how people vote, uh,

103
00:09:43,860 --> 00:09:47,850
whether people who are MPs who are better performers tend to get voted in better,

104
00:09:48,570 --> 00:09:52,170
um, similarly, Dan Posner and Guy Grothman did transparency interventions,

105
00:09:52,170 --> 00:09:57,420
and all of them came back with, uh, practically the same response that transparency interventions don't work.

106
00:09:58,320 --> 00:10:01,450
So it is in this context that I'm kind. Of embarking on my book.

107
00:10:01,460 --> 00:10:07,690
So let me talk about the model that they have and how, um, we proposed to go ahead in my book.

108
00:10:08,650 --> 00:10:15,430
So all of them, uh, use the quantum mystic models in a sense that you have a principal who has a contract with the agent.

109
00:10:15,430 --> 00:10:23,080
So in this case, uh, the citizens have, in some sense, uh, an informal contract, a social contract, if you might want to call it, with the government.

110
00:10:23,830 --> 00:10:29,110
But you cannot observe all the actions of the officials or, let's say, the politicians and so on,

111
00:10:29,530 --> 00:10:33,700
thanks to which, uh, the principals in question cannot enforce the contract.

112
00:10:34,210 --> 00:10:40,510
So which is very few, uh, bring in a measure of transparency, which is that you, uh, bring in these monitoring mechanisms.

113
00:10:40,900 --> 00:10:45,309
Your ability to enforce the contract kind of becomes better. So that's the model with which they go.

114
00:10:45,310 --> 00:10:53,170
So they want to create monitoring mechanisms for citizens to be able to, uh, view the, uh, actions of politicians or bureaucrats.

115
00:10:54,130 --> 00:10:59,200
Um, so here you basically have a 1 to 1 relationship between the principal and the agent,

116
00:10:59,710 --> 00:11:03,490
um, which maps in this case, let's say, uh, between citizen and government.

117
00:11:04,720 --> 00:11:13,420
So the model, um, that you have here of transparency is that you get public record and send it to citizens, but citizens are seen as atomised people.

118
00:11:13,900 --> 00:11:17,260
So you send information to people on a very, very independent manner.

119
00:11:17,260 --> 00:11:24,969
And it's a really a choice theoretic, uh, issue here for the citizens that based on the information you get, you, uh, make your choice.

120
00:11:24,970 --> 00:11:32,560
Let's, uh, let it be a voting choice or something. And it also assumes that citizens will be able to act on the information that you have.

121
00:11:32,980 --> 00:11:38,049
So, uh, Jeremy Weinstein, um, in fairly naturally would assume that if you have,

122
00:11:38,050 --> 00:11:41,080
let's say, better information about how your politicians perform in office,

123
00:11:41,470 --> 00:11:47,200
you will be able to vote, um, according to the information that you have, which is all great.

124
00:11:47,200 --> 00:11:52,480
But for most citizens, a government is a massive, massive, massive entity.

125
00:11:52,750 --> 00:11:57,610
And a lot of us actually feel fairly powerless, uh, in dealing with the bureaucracy of the government in general.

126
00:11:57,970 --> 00:12:06,460
So the assumption that citizens can act in a very seamless manner using the information that you get, uh, is flawed.

127
00:12:06,490 --> 00:12:10,930
Um, and that's something that needs to be taken into account in building a transparency intervention.

128
00:12:11,560 --> 00:12:16,300
So one of the things that you're trying to do is that the model of, uh, government.

129
00:12:16,450 --> 00:12:20,919
But I mean, I'm sorry, sending public records to atomised individuals, um,

130
00:12:20,920 --> 00:12:26,890
is something that I'm trying to challenge a little bit or at least work upon by bringing an element of activism into it.

131
00:12:27,430 --> 00:12:36,700
So, um, in all the areas that I work in, instead of getting public records and sending it to citizens by, uh, practically drafting the information,

132
00:12:37,210 --> 00:12:44,530
we work with activists in each of these areas who may be able to mobilise and do the things that are required in order to confront,

133
00:12:44,530 --> 00:12:55,809
uh, convert the information into action. Um, along with that, I also tend to see, um, um, citizens not just as activist individuals,

134
00:12:55,810 --> 00:13:01,900
but as a community in places where citizens are better organised, in places where they are able to act collectively.

135
00:13:01,900 --> 00:13:05,950
The ability for information to translate into action is a little bit higher.

136
00:13:06,490 --> 00:13:09,310
It's a fairly commonsensical it's a fairly simple thing to say,

137
00:13:09,340 --> 00:13:14,200
but if you think about information interventions that are often created, it's really targeting towards individuals.

138
00:13:14,620 --> 00:13:19,839
So is there a difference between creating, um, an information intervention for a community versus an individual?

139
00:13:19,840 --> 00:13:24,070
I'll come into some of those, um, challenges, um, as I go along.

140
00:13:25,360 --> 00:13:33,099
So what we have done is to change the model where we say that, uh, we want to get a public records and pass it on to,

141
00:13:33,100 --> 00:13:41,620
um, a community, but doing it, um, in a way that we can help the activists that we work with to build a community.

142
00:13:41,620 --> 00:13:48,610
I'll give you a quick example. Oh, one of the things that we've done is to create, uh, information broadcasting mechanism where any time,

143
00:13:48,790 --> 00:13:51,700
let's say, a person gets benched, another person is given foodgrains.

144
00:13:52,060 --> 00:13:56,530
Or if a truck reaches a village with food grains, we call people and let them know,

145
00:13:56,540 --> 00:14:00,370
okay, according to our records, a truck getting so much grains has reached your village.

146
00:14:00,970 --> 00:14:07,330
This should be adequate to give you, uh, give all the beneficiaries of the public distribution system your one month quota.

147
00:14:07,930 --> 00:14:12,940
So. Which means that when you go to your ration shop, at the ration shop, dealers say, I don't have enough grains to give everybody,

148
00:14:12,940 --> 00:14:20,499
you know, that it's a, it's a like, um, but along with that, uh, initially that was my only focus.

149
00:14:20,500 --> 00:14:27,250
So just to keep the activists informed. But what we also started doing is to tell the activists, suppose if you want to organise a meeting,

150
00:14:27,940 --> 00:14:32,409
um, or if there is other kind of information that you want to pass along, use our system.

151
00:14:32,410 --> 00:14:36,940
We have a lot of phone numbers that we collected from you use our system to make announcements.

152
00:14:37,540 --> 00:14:41,469
This to my surprise, I mean, this wasn't something that I anticipated in the first place.

153
00:14:41,470 --> 00:14:45,310
This is has been the most favourite feature that the activists have been using.

154
00:14:45,600 --> 00:14:49,239
And of course, everybody likes to listen to your own voice being broadcast to hundreds of people.

155
00:14:49,240 --> 00:14:54,430
But in addition, what we realised is that it's such a powerful tool for mobilisation.

156
00:14:54,730 --> 00:14:59,140
To give you an example, one of my activist, uh, partners, um, has a lot of.

157
00:14:59,480 --> 00:15:02,690
Bananas, or these gatherings of protest gatherings.

158
00:15:03,470 --> 00:15:06,790
The normal way in which they work is that they have a bunch of volunteers.

159
00:15:06,800 --> 00:15:11,090
They will contact each of the volunteers who will in turn go and, um, talk to people.

160
00:15:11,120 --> 00:15:15,380
Um, tell them that there's going to be one such gathering and then ask the people to come.

161
00:15:16,460 --> 00:15:21,050
Now the volunteers tend to be from the same community, so they don't have the same kind of authority as,

162
00:15:21,050 --> 00:15:26,390
let's say, the activist, uh, himself or herself might. And it also takes a lot of time and effort to do this.

163
00:15:26,390 --> 00:15:31,370
So on one occasion when they had to organise that, that relatively quickly, he just used, uh,

164
00:15:31,370 --> 00:15:37,909
used the system that we have to blast the call to, uh, a few thousand of the, uh, community that they are in touch with.

165
00:15:37,910 --> 00:15:43,070
And they ended up getting twice the number of people turning up for the turnout than they would using the normal process.

166
00:15:43,760 --> 00:15:47,989
Um, some of my, uh, partners have been using it very regularly to inform people.

167
00:15:47,990 --> 00:15:51,709
So every time they would meet a government official, they would make a call and say, we met an official.

168
00:15:51,710 --> 00:15:56,780
This is what exactly happened. Or if there's a protest in a different village, they would say, this is what happened in this village.

169
00:15:56,780 --> 00:16:01,790
Or if there's a nice thing happening in a particular village, they'll say, this program was implemented beautifully in a different village,

170
00:16:02,330 --> 00:16:10,330
and it has kind of created a mechanism for community building in a much, much more, uh, sustained and a relatively costless manner.

171
00:16:10,340 --> 00:16:14,900
And a couple of my partners are, uh, people who started the work very, very recently.

172
00:16:15,230 --> 00:16:19,850
So it's easier for me to see the impact of this mechanism with them than with an established organisation.

173
00:16:20,360 --> 00:16:21,770
So in that sense, um,

174
00:16:23,030 --> 00:16:31,729
disseminating public records and community building happened together in this particular process and enabling that has made a world of difference,

175
00:16:31,730 --> 00:16:37,460
both for the uptake of the system by the activist partners, but also in, um,

176
00:16:37,790 --> 00:16:41,780
creating a semblance of possibility that the issues that we are talking about will get addressed.

177
00:16:43,130 --> 00:16:50,900
Now, along with transparency, we also realised that, um, um, I mean, all the partners told us that don't stop sending information.

178
00:16:50,900 --> 00:16:57,110
We need to do something about the grievances. So we started doing very small things like collecting, um, feedback.

179
00:16:57,440 --> 00:17:05,479
Um, based on the calls. Um, this is kind of, um, has an interesting implication because suppose if you, let's say,

180
00:17:05,480 --> 00:17:10,309
create an idea of mechanism, um, let me pause by saying that I work in rural areas.

181
00:17:10,310 --> 00:17:13,940
I work only with feature phones. I don't work with smartphones, I don't work with the internet.

182
00:17:14,540 --> 00:17:20,599
So, um, for any complaint that you might want to submit to the government, you need to put a lot of details.

183
00:17:20,600 --> 00:17:25,729
And so you need to say who is the complaint about? What is the program that you're talking about within the program?

184
00:17:25,730 --> 00:17:29,960
Which month does your complaint pertain to? Uh, what is the specific project?

185
00:17:30,350 --> 00:17:35,210
Which village are you talking about? And then you have to address it to the right official and whatnot.

186
00:17:35,510 --> 00:17:39,800
So creating a cogent complaint that actually people can act upon is not a simple thing.

187
00:17:40,370 --> 00:17:45,259
So if you were to create an IVR mechanism in which you ask people to feed all this data in,

188
00:17:45,260 --> 00:17:50,120
it'll be practically impossible as a technical, uh, technically mediated mechanism to do that.

189
00:17:50,660 --> 00:17:53,510
But suppose if I send a piece of information where I say,

190
00:17:53,600 --> 00:17:59,960
according to the records in this particular project, you left two months back for which you had paid ₹300.

191
00:18:00,260 --> 00:18:01,790
Can you confirm if you got that money?

192
00:18:02,960 --> 00:18:09,650
And I already know who I am calling, uh, because I have their phone number and I have that identity number, the particular job identity number.

193
00:18:09,980 --> 00:18:15,680
So I know who I'm calling, which village they belong to, which market is it pertains to, and I have all of that information.

194
00:18:16,070 --> 00:18:22,100
So all they have to do is to press 1 or 2 at the end of it in order to give me a very cogent, consistent reply.

195
00:18:22,700 --> 00:18:26,719
So grievance forwarding is something that we've added along with the transparency intervention.

196
00:18:26,720 --> 00:18:30,680
And just adding that makes a huge difference to the whole project.

197
00:18:31,670 --> 00:18:36,170
Now this as well, we had to tweak in many, of many ways in each area that we work in,

198
00:18:36,710 --> 00:18:40,400
in certain places, we know that the bureaucracy is very, very proactive.

199
00:18:40,400 --> 00:18:47,290
And they are um, um, uh, and for example, in Andhra Pradesh, which is an area that you work in, in the Employment Guarantee Act,

200
00:18:47,300 --> 00:18:53,720
we knew that the the senior bureaucrats are extremely proactive, and they have also created a grievance redressal mechanism online.

201
00:18:54,710 --> 00:18:59,060
So taking this these complaints and passing it on to the online mechanism made sense for me,

202
00:18:59,060 --> 00:19:04,190
because if the lower level officials did not address these grievances within two weeks,

203
00:19:04,340 --> 00:19:10,040
uh, that information will go to a top level official as an underdressed grievance and the ticketing mechanisms that they have.

204
00:19:10,820 --> 00:19:17,750
And that is in a different place. Um, the activist told us that if you do have, let's say, um, submit individual complaints,

205
00:19:18,290 --> 00:19:22,520
there will be so much pressure on the complainant by local officials that they will be made to withdraw the complaint.

206
00:19:23,210 --> 00:19:27,710
And you'll see in lots of cases, complaints, getting close, saying that this complaint was false,

207
00:19:28,280 --> 00:19:33,890
or that suppose if you make a complaint saying that the grain that I got from the public distribution system was of a poor quality,

208
00:19:34,490 --> 00:19:39,799
the inspector who goes to verify the complaint will say, show me the grain, and if you don't have the grain anymore,

209
00:19:39,800 --> 00:19:43,970
and because this person comes two months later, then they'll say this complaint has no basis.

210
00:19:44,360 --> 00:19:48,500
It's annoying how to document the complaint and withstanding local pressures are also critical.

211
00:19:48,920 --> 00:19:53,450
So the partner there told us that, uh, instead of submitting individual complaints,

212
00:19:53,450 --> 00:19:59,270
let's wait till we get at least ten complaints from a particular village, because then people will be able to resist pressure better.

213
00:19:59,410 --> 00:20:06,010
An individual will be able to. And instead of giving each complaint separately to the administration, we will collect them.

214
00:20:06,460 --> 00:20:10,900
The activist will work with the community to prepare them, documented in ways that are,

215
00:20:10,930 --> 00:20:15,400
um, compatible, and then give a collective complaint and follow up along with them.

216
00:20:15,760 --> 00:20:21,730
So in that sense, um, in each place we realise that we have to strategize depending on the local context,

217
00:20:22,180 --> 00:20:26,500
in order to follow up on the grievances that arise through the transparency mechanism.

218
00:20:26,890 --> 00:20:33,460
And putting the two together, or at least conceiving the two together, can make a huge difference to the impact that a project might have.

219
00:20:35,410 --> 00:20:42,220
And finally, we also realise that, um, along with information dissemination via mobile phone,

220
00:20:42,730 --> 00:20:46,750
it makes a lot of difference, often to have print outs in certain cases.

221
00:20:47,200 --> 00:20:53,800
Now one of my partners, for example, has what I mean, uh, social workers who cover ten villages.

222
00:20:53,830 --> 00:20:58,989
Um, each know it's easy for me to send, um, uh,

223
00:20:58,990 --> 00:21:03,940
each person's entitlement or the arrival of a truck to a village to all the people in that particular village.

224
00:21:04,360 --> 00:21:10,420
But sending all that information via voice calls to this to the social worker will be completely meaningless.

225
00:21:10,840 --> 00:21:14,860
So the moment we give them a printout saying this is what is happening in these villages,

226
00:21:14,860 --> 00:21:20,290
or when we highlight certain problems that are happening in each place, it becomes a lot easier for them to navigate with.

227
00:21:20,620 --> 00:21:27,939
So in that sense, um, having online plus offline communication, um, partly to bridge the gap with people that we are not able to reach,

228
00:21:27,940 --> 00:21:34,629
but partly to present information in a different way, um, to people who may need it also started making a difference,

229
00:21:34,630 --> 00:21:38,230
um, in how our partners have been perceiving the project so far.

230
00:21:38,560 --> 00:21:45,460
So we started moving away from this very individual centred information dissemination mechanism to one that is focussed on the community,

231
00:21:46,150 --> 00:21:50,010
um, and enabling activists to communicate with the community,

232
00:21:50,020 --> 00:21:57,370
um, in order to build the community in some sense also became fairly important for the process, uh, creating grievances and offline mechanisms.

233
00:21:57,370 --> 00:22:02,650
So those were some of the initial tweaks that we did, but then we ran into other problems as we went along.

234
00:22:03,220 --> 00:22:08,440
Now it's easy to on in public records, um, in the Employment Guarantee Act alone,

235
00:22:08,440 --> 00:22:11,890
if you go to the website, you will find that there are 220 plus reports.

236
00:22:12,550 --> 00:22:16,600
Now each of these reports presents a different information, all of which is meaningful.

237
00:22:17,230 --> 00:22:21,190
So one of the early exercises we did was to go to villages where we would ask people, okay,

238
00:22:21,220 --> 00:22:25,450
would you would it be useful for you to know when the wages have been sent to your bank account?

239
00:22:25,510 --> 00:22:28,270
Would it be useful if we sent this particular information and whatnot?

240
00:22:28,810 --> 00:22:34,390
So what we realised in the process is that people had different problems in different areas.

241
00:22:34,870 --> 00:22:39,159
And so it raised the question for us as to what is it that we need to make transparent.

242
00:22:39,160 --> 00:22:45,729
It's a question that surprisingly, nobody really asks in the, um, in the technology for transparency community,

243
00:22:45,730 --> 00:22:52,090
because a lot of this effort is really to get existing datasets and to visualise them in easy to consume ways.

244
00:22:52,540 --> 00:23:00,320
So for us, it became really critical to ask what it is that you want to make transparent and the different strategies that we are following.

245
00:23:00,370 --> 00:23:06,400
I can kind of communicate that. By I mean we can summarise it by the acronym Shuo, which I'll get to in a second.

246
00:23:07,240 --> 00:23:14,590
So the first that we realised is that, um, we have to focus on the ease of understanding messages, but there are a lot of cases.

247
00:23:14,590 --> 00:23:22,149
But even though information is there, it's so incredibly complicated or it's so incredibly nuanced that sending that,

248
00:23:22,150 --> 00:23:25,960
especially via a voice call or SMS, is practically impossible.

249
00:23:26,410 --> 00:23:28,389
That does act as a filtering mechanism.

250
00:23:28,390 --> 00:23:34,299
Maybe we will come up with the easier way of delivering these messages in the future, but ease of understanding for us is critical.

251
00:23:34,300 --> 00:23:39,750
So the information has to be short, self-contained, so that we don't say, uh, we don't, uh,

252
00:23:39,760 --> 00:23:43,870
refer to something in the message and say, refer to something else because it's a voice call or an SMS.

253
00:23:44,680 --> 00:23:50,230
So we don't have to qualify or give explanations. And then we also want to repeat what we said in every message.

254
00:23:50,620 --> 00:23:53,799
Now, if you're thinking about a voice call that people can actually listen to and

255
00:23:53,800 --> 00:23:56,650
make sense of doing it for more than a minute and a half doesn't make sense.

256
00:23:57,100 --> 00:24:01,960
So it has to be short, self-contained, relatively simple, uh, elemental messages that you can actually send.

257
00:24:02,950 --> 00:24:08,560
So along with ease. The second, um, thing that became really critical for us is how valuable is the message?

258
00:24:09,040 --> 00:24:10,180
To give you an example.

259
00:24:10,280 --> 00:24:18,339
Uh, in the works program, uh, after you work in a project, you're supposed to be paid within 15 days in some parts of, uh, in some areas,

260
00:24:18,340 --> 00:24:24,940
we we work a new form of corruption has come up where the payment agencies will take will get the money that is sent to them.

261
00:24:24,940 --> 00:24:28,510
These are intermediary agent private agencies that the government makes payments through,

262
00:24:29,380 --> 00:24:32,920
and they send it out on interest because interest rates in rural India. Atrocious.

263
00:24:32,930 --> 00:24:36,160
I mean, you can get, uh, 20 to 30% in a month.

264
00:24:37,600 --> 00:24:42,639
So they would send it out on interest for a month or two. But for a poor labourer and a pensioner, this is pure torture.

265
00:24:42,640 --> 00:24:46,360
If you don't get your money on time and then you get you pay the money later.

266
00:24:46,840 --> 00:24:51,370
So here have people found it very valuable to know that the money has arrived in their bank account,

267
00:24:51,370 --> 00:24:54,700
because the contract with the agency is that you have to pay it within four days,

268
00:24:55,330 --> 00:24:59,240
so they can immediately track the, I mean, and they can actually go and talk to the agency.

269
00:24:59,290 --> 00:25:00,390
Didn't say, now give me the money.

270
00:25:01,110 --> 00:25:07,520
Whereas in some other places they said, look, I know I work and within 15 days I get my money, which is the duration.

271
00:25:07,530 --> 00:25:11,460
So if you send me this information and when the money reach the bank account, it doesn't make any difference for me.

272
00:25:12,210 --> 00:25:16,620
So, um, sending the right message is fairly critical.

273
00:25:16,620 --> 00:25:22,739
Now, if you think about a lot of transference interventions, they're often based on some centralised person sitting in value.

274
00:25:22,740 --> 00:25:26,970
I mean, creating values for messages saying these are important messages, let me disseminate them.

275
00:25:27,450 --> 00:25:30,809
But how do you identify which message is useful in which context?

276
00:25:30,810 --> 00:25:36,990
Because the local problems are so varied. So that has important design consequence for transparency interventions.

277
00:25:38,610 --> 00:25:40,650
So which I'll talk to in a bit.

278
00:25:40,770 --> 00:25:49,110
But even that needs to be kind of qualified because there are places where people would say that, uh, yes, this information is useful for us,

279
00:25:49,590 --> 00:25:55,770
but I can't do anything about it because if I take this issue of I'd be beaten up in that kind of a scenario,

280
00:25:55,770 --> 00:25:59,130
even if it's a valuable information, it's not something that you can really act upon.

281
00:25:59,640 --> 00:26:08,820
So what we've kind of decided is that, um, one the information that we send needs to be relevant to a particular problem that people have.

282
00:26:11,720 --> 00:26:17,120
And so the word that we use is kind of Actionability is the information that you send actionable.

283
00:26:17,540 --> 00:26:20,990
A heuristic for this is that, um,

284
00:26:21,140 --> 00:26:25,969
suppose if you send information saying that so many million dollars has been allocated in the national budget for, let's say,

285
00:26:25,970 --> 00:26:30,080
building roads, you, the most that you can do with this information is to say,

286
00:26:30,080 --> 00:26:33,890
okay, more money needs to go into roads and less money needs to go into roads. You can't verify it.

287
00:26:34,460 --> 00:26:39,350
Whereas if we send an information saying so much money has been allocated for building a road in your village,

288
00:26:39,350 --> 00:26:44,270
you can monitor and say, this road was built in my village or not, or even going down.

289
00:26:44,270 --> 00:26:49,910
If you say that if a particular person was paid, um, so much money, then it is something they can immediately verify.

290
00:26:50,390 --> 00:26:59,330
So information becomes more and more actionable when you go from macro issues like policies and budgets and so on to relatively micro issues,

291
00:27:00,380 --> 00:27:07,880
um, including payments that are made to individuals, project implemented in the villages, resources allocated in villages, or how to information,

292
00:27:07,880 --> 00:27:11,870
how to access this particular program, how to file a complaint, who to talk to for this, etc.

293
00:27:12,260 --> 00:27:18,500
So a lot of that focus has shifted in terms of the micro level, partly because it is actionable.

294
00:27:19,430 --> 00:27:24,409
And we um, as I said, information should be it should be relevant to a current problem that people face,

295
00:27:24,410 --> 00:27:28,040
but it should also be something that people feel that they can act upon and resolve.

296
00:27:28,640 --> 00:27:32,420
And the only way for us to know that is by having some kind of a local feedback.

297
00:27:32,870 --> 00:27:38,419
Without that, there's nowhere on earth will be able to find out if it is a problem that people know that they can tackle,

298
00:27:38,420 --> 00:27:42,530
given the current level of organisation, etc. So very often you might have, let's say,

299
00:27:42,530 --> 00:27:47,510
a very enterprising bureaucrat in the locality who's so responsive that you feel that you can act on this particular information.

300
00:27:48,200 --> 00:27:55,400
Um, or you may be in a place where, um, the senior bureaucrat rather than the local bureaucrat is much more capable of acting.

301
00:27:55,910 --> 00:28:04,700
So in that sense, having a sense of organisation, having a sense of, um, the particular social context, um, but most importantly, um, the,

302
00:28:04,880 --> 00:28:07,480
the level of unity that, uh, the community has,

303
00:28:07,490 --> 00:28:12,890
all of these things have implications for your transparency, intervention to, um, translate into action.

304
00:28:13,220 --> 00:28:17,209
So, like I said, build a building. Unity has been fairly important for us.

305
00:28:17,210 --> 00:28:23,720
But equally, building on a community building on an existing community has been a strategy that we've been using.

306
00:28:24,470 --> 00:28:27,980
And communities that are organised typically do a lot of different things.

307
00:28:27,980 --> 00:28:32,450
They build alliances with, uh, I mean, different people in the administration or otherwise.

308
00:28:32,840 --> 00:28:39,680
They have different kinds of skillsets. Um, and they bring a different kinds of people who all act as resources for action.

309
00:28:40,190 --> 00:28:46,370
So this view of, um, transparency working in a kind of a collective action framework, if you think of it,

310
00:28:46,700 --> 00:28:53,420
think of it as a social movement theorist rather than an economist who thinks about, um, information changing individual choices.

311
00:28:54,020 --> 00:29:00,050
It seems to be a lot more powerful way of building, um, uh, uh, an intervention that has a chance to work.

312
00:29:01,550 --> 00:29:06,050
So in essence, what we are saying is that if you have to build, um, uh,

313
00:29:06,590 --> 00:29:12,260
an intervention that has a chance of working, it should be based on local strategies, strategies of the community,

314
00:29:12,260 --> 00:29:16,580
not just what problems they face, not just what problems they feel they can act upon,

315
00:29:17,060 --> 00:29:21,200
but also on what alliances they have and what are some of the current priorities.

316
00:29:22,940 --> 00:29:29,870
So that's kind of, um, a set of, um, heuristics that we've been kind of developing has been going on in this particular project.

317
00:29:29,870 --> 00:29:36,079
So there's a very specific way in which we work is that in each of the areas that I work and the activist partner,

318
00:29:36,080 --> 00:29:41,750
we work with them on a continuous basis to. So they let us know that these are the current problems that we have.

319
00:29:42,590 --> 00:29:47,300
We then bring all the information that we can find on the website to say, Will any of this be of use to you?

320
00:29:48,020 --> 00:29:53,989
They will identify the kind of information. Then we try and put it in the most useful or the most easy to understand manner.

321
00:29:53,990 --> 00:30:01,760
Test it out with the people that we work with, and start disseminating that information consistently for at least a period of one year so that,

322
00:30:01,760 --> 00:30:04,999
um, information that you send doesn't like the social audit.

323
00:30:05,000 --> 00:30:07,910
It's not a one shot thing. It's something that happens on a continuous manner.

324
00:30:08,420 --> 00:30:13,760
So we also collect feedback in the process and pass it on to the activists as well as the government,

325
00:30:13,760 --> 00:30:17,720
depending on the local priorities and the local context and the allies that we have and whatnot.

326
00:30:18,290 --> 00:30:22,129
So the, the schwa um acronym, in some sense, it's fairly simple.

327
00:30:22,130 --> 00:30:25,730
It's very commonsensical, but it's yet very, very consequential.

328
00:30:26,420 --> 00:30:33,500
So localising information, um, based on uh, local problems and organisation is very, very critical in the way I think about it.

329
00:30:34,220 --> 00:30:38,660
And it's not something that a lot of transparency interventions can easily account for, at least as they exist today.

330
00:30:38,660 --> 00:30:45,380
To give you an example, the Government of India today has created a committee to examine how, uh, instead of just putting information online,

331
00:30:45,770 --> 00:30:49,880
whether they can collect people's phone numbers, the beneficiaries phone numbers, and send information to them.

332
00:30:50,480 --> 00:30:55,430
So the project that we have is that just this system of blasting information to

333
00:30:55,430 --> 00:30:58,460
all the beneficiaries of welfare programs is not really going to take us too far.

334
00:30:58,910 --> 00:31:03,110
So how do we localise? How do we contextualise, how do we build it upon a community, etc.?

335
00:31:03,110 --> 00:31:09,620
It requires a lot of, um, thinking, a lot of design thinking, so that such policies that have great potential.

336
00:31:10,180 --> 00:31:14,710
Um, do not lose out in a way that Weinstein, Grossmann and others found out.

337
00:31:15,760 --> 00:31:20,560
So the strategies of all of his partners over the last two years keep changing constantly.

338
00:31:21,130 --> 00:31:24,490
That is, if you look at a lot of our technology for transparency interventions,

339
00:31:24,490 --> 00:31:30,790
it takes such a lot of time for us to take a data set and clean it up and to visualise it that we want it to be a long standing thing.

340
00:31:31,180 --> 00:31:34,360
So the information interventions have to be really dynamic.

341
00:31:34,480 --> 00:31:40,210
The messages should keep changing. So how do you build a dynamic system is another thing that we've been kind of contending with.

342
00:31:41,800 --> 00:31:48,650
Um, the least that I want to finish with is that there's a lot of focus today on data sets.

343
00:31:48,670 --> 00:31:51,670
You know, we want to open up data sets, and we want to get whatever data sets.

344
00:31:51,670 --> 00:31:55,710
We can connect them to each other and to visualise them. Yes, that will be great.

345
00:31:55,720 --> 00:32:01,270
But I feel like without doing all the attendant work, which requires localisation and prioritisation,

346
00:32:01,780 --> 00:32:06,190
any information system that we create is unlikely to be successful,

347
00:32:06,610 --> 00:32:10,300
especially if we want to focus on rural areas and on uh,

348
00:32:10,360 --> 00:32:17,320
users would be having basic feature phones and like asking a person to go to a website and select from 220 different reports.

349
00:32:17,920 --> 00:32:24,010
When you want an information dissemination mechanism through voice or items, you are very limited in the kind of information that you're dealing with.

350
00:32:24,520 --> 00:32:29,170
So a lot more needs to happen than just putting information online.

351
00:32:29,530 --> 00:32:32,700
We need to think about organisation. We need to think about building communities.

352
00:32:32,710 --> 00:32:38,290
We need to think about building alliances and a lot else for transparency to do Park.

353
00:32:38,560 --> 00:32:43,450
So I'll stop here and then hope to hear your questions and suggestions around it.

354
00:32:43,760 --> 00:32:55,750
Thank you. Thank you. Let me turn this monitor off so I can take notes without you seeing what I'm thinking about you.

355
00:32:57,280 --> 00:33:03,760
So it will take up to about 20 minutes or so for, um, questions like, would you like to take a few at a time?

356
00:33:04,030 --> 00:33:08,170
Um, any help you want? So. So maybe we could take 2 or 3 at a time.

357
00:33:08,500 --> 00:33:11,710
Um, there are any comments or questions? I see Benson for the first one.

358
00:33:13,120 --> 00:33:23,140
Yeah. Um, so when you have the list of naysayers saying that, you know, opening up information, being transparent doesn't actually have real effect.

359
00:33:23,950 --> 00:33:28,490
Um, I didn't read them, but it strikes me as that's quite obvious.

360
00:33:28,510 --> 00:33:32,050
You know, of course, if the government opens up for data, I'm not going to look at all the data.

361
00:33:32,680 --> 00:33:41,740
But, um, the what it may do is allow people like you and lots of other organisations to take that information and make it,

362
00:33:42,220 --> 00:33:47,410
uh, valuable, and certainly has that check checked. What's the effect of the tiny amount of people,

363
00:33:48,130 --> 00:33:54,730
the tiny fraction of the population that actually uses that information and then make something valuable out of it as they're being tested,

364
00:33:54,730 --> 00:33:56,580
or if they, the naysayers, talks about that.

365
00:33:56,590 --> 00:34:03,100
So what these three authors did was to take, um, some information that a lot of groups believe would be useful.

366
00:34:04,060 --> 00:34:08,110
Um, they did the job of the info military curated, the information,

367
00:34:08,110 --> 00:34:14,440
created visually appealing um cards, even organise meetings, um, all over the country.

368
00:34:14,590 --> 00:34:18,069
Uh, explaining the performance, let's say of MP's or another case.

369
00:34:18,070 --> 00:34:21,100
It was explaining the educational achievement of the children,

370
00:34:21,100 --> 00:34:27,280
etc. etc. these were all very careful interventions, very careful interventions, very thoughtful interventions.

371
00:34:27,760 --> 00:34:33,640
Uh, but they had taken into account, uh, and they just didn't, uh, look at what data sets are getting used.

372
00:34:34,250 --> 00:34:37,450
They worked on very concrete information after a lot of homework.

373
00:34:38,140 --> 00:34:45,550
Um, so the problem that I see with their work is that they thought about information acting in a very atomistic sense.

374
00:34:46,390 --> 00:34:50,709
Plus, the information was prioritised centrally. In some sense, it was not dynamic.

375
00:34:50,710 --> 00:34:54,310
The scorecards were prepared with a lot of work, but the score was static.

376
00:34:54,670 --> 00:34:58,930
They didn't, um, particularly respond to the demands that local communities were making.

377
00:34:58,930 --> 00:35:03,610
So, um, did communities want to know what foreign policy issues these guys commented on?

378
00:35:03,610 --> 00:35:05,280
Probably not. Right.

379
00:35:05,290 --> 00:35:16,209
And that's this is where I feel like even though all of these, uh, projects were based on very carefully curated, um, um, um, transparency mechanisms,

380
00:35:16,210 --> 00:35:22,080
uh, they did not act on a lot of other things that, uh, I'm trying to do in this project, partly in response to their concerns.

381
00:35:22,200 --> 00:35:25,810
Quick question. You do you think that you've proven them wrong with what you do?

382
00:35:26,350 --> 00:35:28,720
Um, I don't know. I don't know the impact of my project.

383
00:35:28,810 --> 00:35:38,140
Um, so we are still in early stages, but I know that if I were to just blast information to citizens, it will not succeed that much I know for sure.

384
00:35:38,380 --> 00:35:42,520
Um, so. So, like I said, this is a work in progress. These are things that I'm struggling with.

385
00:35:44,090 --> 00:35:48,909
So, yeah, I have a couple of questions. One long and one short that, um,

386
00:35:48,910 --> 00:35:52,899
I'm really interested in the approach that the right information campaign in

387
00:35:52,900 --> 00:35:59,139
India to to work with Aruna Roy was here and here ago and she made the point.

388
00:35:59,140 --> 00:36:03,360
It's very similar to yours when you say, well, it's not enough. Digital information is not enough.

389
00:36:03,370 --> 00:36:09,930
We need other kind of information, she said. We almost won our campaign and people.

390
00:36:09,980 --> 00:36:13,280
We're coming in saying we can put everything online. As Betty was saying.

391
00:36:13,640 --> 00:36:18,860
And we said, that's not enough. We have to go to the village level and put on a board written,

392
00:36:18,860 --> 00:36:24,169
handwritten language that can be spoken and understood by the people, the relevant information.

393
00:36:24,170 --> 00:36:25,730
At that point, people are in power.

394
00:36:26,180 --> 00:36:32,570
And even if you're part of deliberation, technology, agenda and so forth, you have a very nuanced approach saying this is not enough.

395
00:36:33,190 --> 00:36:38,600
But in developed and in developing country, there is kind of a digital complacency approach.

396
00:36:39,050 --> 00:36:42,230
A lot of governments are saying information is out there.

397
00:36:42,410 --> 00:36:43,729
So we have done our job.

398
00:36:43,730 --> 00:36:54,590
But what is specific of India is they have to find in other places that sort of forces both government and activists to make the next step,

399
00:36:54,890 --> 00:37:01,490
because you're never satisfied that independently from what is the stand that you're taking and I haven't find elsewhere,

400
00:37:01,490 --> 00:37:05,620
maybe in the broader president, find it in in other countries I haven't personally.

401
00:37:05,630 --> 00:37:12,130
So this is, this is this is the first question and uh, and I can I will just focus on this one and maybe that's not it.

402
00:37:12,770 --> 00:37:17,600
So should I be for one more question, I forgot to make sense before I answered Benson.

403
00:37:18,440 --> 00:37:25,730
I have two questions. The first is methodological. How do you choose the number of agents that you have in each and each village?

404
00:37:25,910 --> 00:37:29,330
Do you take the people that you have their phone numbers and text all of them,

405
00:37:29,330 --> 00:37:32,510
or do you choose people that are strategically located within the community?

406
00:37:33,020 --> 00:37:40,150
And the second question is that I saw what you did with simplifying the information, saying that we don't care about the budget, the overall budget.

407
00:37:40,160 --> 00:37:43,160
This is the amount of money allocated to your road.

408
00:37:43,940 --> 00:37:47,659
But my question is kind of critical in the sense that if you simplify that information,

409
00:37:47,660 --> 00:37:54,500
don't you perpetuate the kind of information poverty in the sense that now they know how much has been allocated to the specific road,

410
00:37:54,500 --> 00:37:59,479
but they have no idea how much from the budget it is, and they can't even compare in the sense that,

411
00:37:59,480 --> 00:38:03,170
okay, this is what we have, this is what we're going to make do. So.

412
00:38:05,450 --> 00:38:06,390
Just talk with your question.

413
00:38:06,410 --> 00:38:13,969
I mean, I don't want to trivialise the work that has happened and right information and relate allied movements around the world.

414
00:38:13,970 --> 00:38:19,570
So the very fact that 100 countries today have a right to information law is because people have been acting on it.

415
00:38:19,580 --> 00:38:28,340
And, uh, over a period of time, I think the disconnect is between people who focus on right information online in terms of creating data sets.

416
00:38:28,910 --> 00:38:32,470
Um, but there's a much wider right information community everywhere.

417
00:38:32,480 --> 00:38:35,150
I mean, investigative journalists, for example, uh,

418
00:38:35,150 --> 00:38:41,840
in different parts of the world often rely very heavily on access to information through right information, FOIA or whatever that is called.

419
00:38:42,320 --> 00:38:46,130
And I do see that there's a lot of sophisticated use of the law, per se.

420
00:38:46,790 --> 00:38:55,099
Uh, but the, uh, does the sophistication extend to using the law, uh, creatively when it comes to putting information online?

421
00:38:55,100 --> 00:39:01,940
I think there's a long way to go for that. Um, data driven journalism is one thing that seeks to do this in a clever way.

422
00:39:02,450 --> 00:39:06,170
Um, but when it comes to interfaces that are created for citizens,

423
00:39:06,620 --> 00:39:10,549
I think we have a long way to go in, um, doing things that are much more meaningful.

424
00:39:10,550 --> 00:39:14,900
But that's, I think, a reflection of the fact that all of us are contending with this new beast.

425
00:39:15,560 --> 00:39:23,330
Um, but I do think that, uh, India is definitely not in isolation in terms of, um, thinking about clever ways of using information.

426
00:39:23,330 --> 00:39:28,549
The difference, if any, um, is that among the people that I have interacted with in this community,

427
00:39:28,550 --> 00:39:37,340
there's a lot more focus on rural areas and working with communities that often don't have access to information by themselves.

428
00:39:37,790 --> 00:39:44,839
And so in that, I think maybe it's an accident that, uh, but I, I have not I was based in a village when she started this book,

429
00:39:44,840 --> 00:39:48,650
and then we discovered that it was a very powerful mechanism for doing things.

430
00:39:49,250 --> 00:39:53,750
Uh, but yeah, I wouldn't want to make a blanket statement that that isn't so fascinating use of

431
00:39:54,050 --> 00:39:59,660
that information in other places and on how many people we send information to,

432
00:39:59,660 --> 00:40:09,170
etc. So as of now, um, we initially I started with, uh, um, heuristic that we'll try and collect at least 100 phone numbers from each village.

433
00:40:09,740 --> 00:40:13,340
But it turns out that in certain programs it works reasonably well.

434
00:40:13,340 --> 00:40:16,640
Um, like the public distribution system where, um,

435
00:40:17,150 --> 00:40:22,610
as long as I can tell people that you have enough grains reaching your village to satisfy everybody's entitlement, it works well.

436
00:40:23,090 --> 00:40:30,160
But when it comes to the Employment Guarantee Act, um, I need to send that information to each person who worked in each project there.

437
00:40:30,170 --> 00:40:35,690
What ended up happening is that the amount of people that I could inform out of 100 in each village became very, very low.

438
00:40:36,080 --> 00:40:40,400
So because people tend to work at different times, uh, not everybody does 100 days of work.

439
00:40:40,730 --> 00:40:46,400
Very often I would have only 6 or 7 people to whom I can disseminate information among many, many who may have worked.

440
00:40:46,730 --> 00:40:52,580
So now I'm trying to increase the base to at least one third of the, uh, beneficiaries of each program.

441
00:40:53,240 --> 00:40:57,889
Um, but the process of collecting phone numbers is very, very difficult, and numbers keep changing.

442
00:40:57,890 --> 00:41:03,320
So it's a lot of homework that you need to do. There is a different reason as to why I do not want to blast information to everyone,

443
00:41:03,320 --> 00:41:07,700
which is that I want this to be a tool that is used by activist organisations who use right information.

444
00:41:08,120 --> 00:41:12,260
So which means that my system will be effective only if it is low cost.

445
00:41:12,710 --> 00:41:20,540
So any system that blasts information is not going to be effective. Now, uh, a hopeful stage, but I want to get to is that, um,

446
00:41:20,540 --> 00:41:27,020
you create an information dissemination mechanism where we randomly send information to a small number of people, say, in each habitation.

447
00:41:27,410 --> 00:41:32,000
So people have some basic information and of course, people talk to each other, especially in rural context.

448
00:41:32,780 --> 00:41:37,700
And then for the rest, we have an incoming mechanism with all they have to do is to dial a particular number,

449
00:41:37,700 --> 00:41:41,540
and they can listen to that number pertaining to them, information pertaining to them if they want to.

450
00:41:41,900 --> 00:41:43,879
So that way the cost of information is shared.

451
00:41:43,880 --> 00:41:52,250
Now this is actually really interesting because the cost of hosting data, um, and the cost of incoming phone calls is very negligible.

452
00:41:52,580 --> 00:41:56,210
There are a lot of activist organisations may be able to use this, um,

453
00:41:56,420 --> 00:42:00,110
because the predominant cost in this kind of a mechanism is the cost of the call.

454
00:42:00,110 --> 00:42:05,630
If we can distribute the information, or rather, that cost sharing by asking people to call into the system,

455
00:42:06,290 --> 00:42:09,489
um, um, it creates a great avenue of consistent transparency.

456
00:42:09,490 --> 00:42:11,960
And people will only call if they feel like it's a problem. Right?

457
00:42:12,290 --> 00:42:17,060
So it's a way of automatically also detecting who has problem instead of blasting information to everyone.

458
00:42:17,390 --> 00:42:21,469
So but we are not quite there because I want to spend enough time in broadcasting

459
00:42:21,470 --> 00:42:25,400
information without people having to call us until we find out what it is that,

460
00:42:25,850 --> 00:42:28,310
uh, people find useful and how to deliver that, and so on.

461
00:42:28,610 --> 00:42:35,150
Once we get there, I'm really hopeful of creating these mechanisms where different activists will be able to host their own information,

462
00:42:35,810 --> 00:42:41,400
um, that others can call into. Right.

463
00:42:41,550 --> 00:42:48,300
Two questions. One was getting off from net, which was, uh, maybe gone from leading up methodology.

464
00:42:48,300 --> 00:42:55,170
Uh, I was just wondering about the scope of it. You have done it in four states, but then how do you, uh, plan to scale it up?

465
00:42:55,170 --> 00:42:59,280
If we have 1.2 billion people in 70% of them live in more conditions?

466
00:42:59,820 --> 00:43:04,560
Do we have enough activists in all the regions to get the information that's relevant to all these people?

467
00:43:05,010 --> 00:43:12,090
And the second question I had, which was more of interest to me, is, uh, you're talking about an effective transparency intervention system.

468
00:43:12,810 --> 00:43:17,340
So you were basing it off of, uh, context and localisation of the context.

469
00:43:17,790 --> 00:43:18,449
But, uh,

470
00:43:18,450 --> 00:43:25,140
what is your impression of the government's initiative into this when they bring out the UAD project and there's no budgetary constraints of that?

471
00:43:25,500 --> 00:43:29,610
So what are your impressions of that system, which is trying to do very much the same thing.

472
00:43:29,610 --> 00:43:33,299
That's what you're trying to do. Yeah. You it doesn't do that.

473
00:43:33,300 --> 00:43:38,730
By the way, you said only to uh, record identify the correct individual, which is a unique identity project,

474
00:43:38,730 --> 00:43:43,980
which is the biometric, uh, uh, identification of every citizen that the government of India is trying to do.

475
00:43:43,980 --> 00:43:51,420
That's what he's referring to. My UAD and a transparency intervention like this will not work until the government actually puts data online.

476
00:43:52,140 --> 00:43:59,580
Right. To that extent, um, it is critical for, um, e-governance mechanisms to be there so that, um,

477
00:43:59,580 --> 00:44:05,309
every transaction gets recorded in the first place digitally and that it is made available proactively online,

478
00:44:05,310 --> 00:44:07,290
preferably in open data formats and so on.

479
00:44:07,740 --> 00:44:13,050
So there has been a lot of effort around the world, especially through the open data movement, to make governments do that.

480
00:44:13,710 --> 00:44:17,190
Um, what I'm saying is that it doesn't we shouldn't stop there.

481
00:44:17,190 --> 00:44:25,440
We should go one step extra to make sure that the mass of information that gets put online gets channelled by Info Ministries.

482
00:44:25,530 --> 00:44:31,110
Um, based on all the criteria that we discussed, as far as the scale of it is concerned,

483
00:44:31,290 --> 00:44:38,760
one thing that I am fairly certain of is that at least in a context like India, it does have a very vibrant civil society that is very widely spread.

484
00:44:39,240 --> 00:44:47,819
So if we create mechanisms where activists can very easily extend it to their local areas, um, we'll be able to scale it up fairly substantially,

485
00:44:47,820 --> 00:44:51,719
especially if we predominant keep on incoming phone calls, uh,

486
00:44:51,720 --> 00:44:55,560
because then they don't have to bear enormous cost in, you know, hosting this kind of a system.

487
00:44:56,100 --> 00:44:59,100
So the question that I would ask is, yes, even this will be limited, right?

488
00:44:59,100 --> 00:45:02,400
Obviously even this is not going to reach, let's say, even the majority of the population.

489
00:45:02,820 --> 00:45:08,370
So the question that I would want to ask is, yes, we won't have an information intervention that reaches the majority,

490
00:45:08,880 --> 00:45:14,040
but would it make sense for us to acknowledge that just by sending information to those who are,

491
00:45:14,550 --> 00:45:18,900
um, in some sense unorganised, um, or not organised,

492
00:45:18,900 --> 00:45:23,610
or who don't have an ability to respond to this information, should we just tell blast information to them?

493
00:45:24,120 --> 00:45:27,150
So would that be of value? It's a it's a question that we need to ask.

494
00:45:28,170 --> 00:45:35,219
And I think there are interesting ways of, um, getting the government to disseminate information, um,

495
00:45:35,220 --> 00:45:41,070
along with activists, even without necessarily setting of activists, um, even without actively collaborating with them.

496
00:45:41,070 --> 00:45:44,490
So one, uh, design that I'm thinking of now, I'm not sure if I could work on it,

497
00:45:44,880 --> 00:45:48,480
would be to consider 220 reports that I talked about an Employment Guarantee Act.

498
00:45:49,080 --> 00:45:53,879
Let's say that there are 6 or 7 reports that are very critical at the grassroots, um,

499
00:45:53,880 --> 00:45:58,740
in a sense that it's micro information, it's easily verifiable, etc., etc. instead of putting,

500
00:45:58,740 --> 00:46:02,010
let's say, the six or even 20 in 1, um,

501
00:46:02,670 --> 00:46:07,950
have phone list so that people will have to call and go through an IVR mechanism to identify the right information.

502
00:46:08,280 --> 00:46:13,079
Let's say you create 20 different phone numbers, or let's say you create even five phone numbers.

503
00:46:13,080 --> 00:46:14,490
Each contains four reports.

504
00:46:15,150 --> 00:46:20,969
Now one activist in a particular area can just distribute a pamphlet to people if that particular problem is very prevalent.

505
00:46:20,970 --> 00:46:26,160
Saying that, call this particular number to get that information so that it I think clever ways

506
00:46:26,160 --> 00:46:29,640
that we may be able to navigate where you create an information ecosystem that,

507
00:46:29,650 --> 00:46:38,160
uh, typically non-tech savvy activist doesn't have to work on, but they will still be able to customise information according to local priorities.

508
00:46:38,640 --> 00:46:42,570
Um, it's something that I think needs a lot of thinking and experimenting with.

509
00:46:43,080 --> 00:46:48,780
But in my mind, without doing that, uh, creating, um, universal transparency mechanisms,

510
00:46:48,780 --> 00:46:51,000
especially using mobile phones is not going to take us too far.

511
00:46:56,830 --> 00:47:03,610
Um, so I just wanted to ask you about the kind of politics, because these are an intensely kind of political activity, you know?

512
00:47:03,610 --> 00:47:08,589
So can you tell me about the kind of communities that and, I mean, when you talk about the community,

513
00:47:08,590 --> 00:47:15,460
who are you acting in kind of these local villages and what implication to that have around kind of different politics?

514
00:47:15,680 --> 00:47:22,659
So I, I don't, um, choose the community that we work with, um, sitting at Stanford and trying to do this work.

515
00:47:22,660 --> 00:47:31,150
I'm way too far remote to do anything like that. What I do instead is, um, go back to my activist roots and identify the set of, uh,

516
00:47:31,210 --> 00:47:36,790
people that I think have had, uh, substantial impact over the last ten, 15 years that I've known them.

517
00:47:36,790 --> 00:47:40,210
And a couple of young activists who seem to be really, really promising.

518
00:47:40,930 --> 00:47:50,319
Um, so, um, the dynamics within, um, I don't really get into that question at all in terms of their priorities, their method of acting, etc.

519
00:47:50,320 --> 00:47:56,860
I don't really get into that. Um, which is, um, so I should acknowledge that there's that a potential limitations there.

520
00:47:56,870 --> 00:47:57,069
Right.

521
00:47:57,070 --> 00:48:05,200
So depending on who they choose to work with, maybe it's cast as organisation, like maybe they have, um, uh, they have of exclusions and whatnot.

522
00:48:05,620 --> 00:48:09,100
It is something that I acknowledge. Um, and it's something that I'm happy to look at.

523
00:48:10,000 --> 00:48:17,260
Um, and I don't um, there have been times when one of the so in, in each place that I work,

524
00:48:17,260 --> 00:48:24,040
I have one, um, team member today who actually, uh, is embedded with the NGO that I work with.

525
00:48:24,670 --> 00:48:31,180
Um, in some cases, they have come back to me with feedback saying, look, uh, the way in which this particular group operates, um, needs change.

526
00:48:31,180 --> 00:48:34,719
Can you go back to the leader and tell them that, you know, you need to do these things differently?

527
00:48:34,720 --> 00:48:39,820
And I, uh, put my foot down on it because it's not my station to go and tell activists who

528
00:48:39,820 --> 00:48:43,870
live and work in that area for decades as to how they should change their work.

529
00:48:44,530 --> 00:48:48,070
Um, but it it has its inherent limitations, which I'm happy to acknowledge.

530
00:48:48,100 --> 00:48:55,059
I mean, just to follow up on that, I mean, there's, um, like a lot of using Indian moments around these kind of anti-corruption party and things.

531
00:48:55,060 --> 00:49:02,740
So do you find yourself being connected with a sense of politics, which then kind of could cause issues, or is that not so as of now?

532
00:49:02,740 --> 00:49:07,000
It just happened so that, uh, in different states that I work are ruled by different parties.

533
00:49:07,600 --> 00:49:12,950
So I have been in conversation with, um, I don't I'm not in conversation with a political party directly.

534
00:49:12,970 --> 00:49:18,730
I'm in conversation with senior bureaucrats, uh, who obviously work very closely with chief ministers.

535
00:49:19,240 --> 00:49:23,890
Um, the, the, um, criteria that I have is that in general, um,

536
00:49:24,370 --> 00:49:32,500
this much data is never put online unless the chief minister is makes this program a political priority or the bureaucrats there are really proactive.

537
00:49:32,890 --> 00:49:37,000
So that is a selection bias, um, in how I operate at this point.

538
00:49:37,660 --> 00:49:43,209
Um, and I think that it's a productive way of getting, um, on this kind of project at this point because without,

539
00:49:43,210 --> 00:49:48,670
um, responsive bureaucracy, uh, at least a project like this is not likely to succeed.

540
00:49:49,180 --> 00:49:52,749
Um, so which is again, to say that, uh, what transparency mechanism?

541
00:49:52,750 --> 00:49:57,400
I mean, there are lots of different ways of doing things that are lots of different ways of making a government to be responsive.

542
00:49:57,970 --> 00:50:04,450
And I don't want to, uh, name the digital transparency, um, uh, thing with the same hammer that I have in every single place.

543
00:50:04,900 --> 00:50:11,080
So I want to acknowledge that it won't work. And probably most of the pieces encountered in local opposition.

544
00:50:11,920 --> 00:50:19,300
Um, so far, um, I mean, again, the project is fairly young because of the fact that I work through,

545
00:50:19,330 --> 00:50:23,920
um, responsive bureaucrats and because I work with people who are very, very organised.

546
00:50:24,520 --> 00:50:30,910
Um, there hasn't been any local resistance that we have, uh, confronted, um, yet.

547
00:50:31,570 --> 00:50:37,840
But then if the project does become, um, viable and if it really creates a challenge, that's something we might see.

548
00:50:37,840 --> 00:50:40,930
But in fact, the paradoxical thing that we saw is that a lot of, um,

549
00:50:40,930 --> 00:50:46,120
field level workers have started approaching us to disseminate certain types of information.

550
00:50:46,480 --> 00:50:52,450
What often ends up happening is that a field level worker who lives in the same village would often go to people and tell them, this is the rule.

551
00:50:52,450 --> 00:50:55,780
That is, they'll say, you're telling me this is a role because you don't want to give me this benefit?

552
00:50:56,050 --> 00:51:02,950
Um, so they have started approaching us and saying, look, the system that information goes through, your mechanism has more credibility with people.

553
00:51:02,950 --> 00:51:07,840
So can you please still inform people of this rule so that it takes burden off us?

554
00:51:08,230 --> 00:51:15,250
So that is that was completely unexpected. Um, but so there are areas, I think, where it will work well with local officials,

555
00:51:15,250 --> 00:51:18,670
but I'm sure there are areas where we have conflicts as we go along.

556
00:51:18,670 --> 00:51:24,490
But it is a conflict that I expect the partner to manage because the, uh,

557
00:51:24,520 --> 00:51:30,400
strategy and issues and everything are driven by them and they do it in their natural course of operating.

558
00:51:31,570 --> 00:51:37,250
Um, I'd like to ask you about the root cause of the problem or the root problem, which is, uh, corruption itself.

559
00:51:37,290 --> 00:51:43,390
Um, I was recently told by, um, an academic at a New Delhi who was describing to me the corruption is something that

560
00:51:43,570 --> 00:51:48,280
really trickled down in the sense that it was pervasive in some aspects of the culture,

561
00:51:48,280 --> 00:51:55,840
in the sense that even people who were meant to deliver services at the most basic local level just have it as part of just ingrained.

562
00:51:56,360 --> 00:52:00,950
Just because it's happening high on up and they might not be getting their due.

563
00:52:01,000 --> 00:52:04,969
Um, and they just sort of put a little bit money aside for themselves.

564
00:52:04,970 --> 00:52:11,240
And so I'm just wondering if there's something there as far as informing or having sort of an awareness raising

565
00:52:11,240 --> 00:52:18,020
or even informing workers that at a sort of most basic local level of what their wages are supposed to be,

566
00:52:18,020 --> 00:52:20,540
of what their project is meant to look like.

567
00:52:20,540 --> 00:52:26,629
So I don't know if you've targeted them as opposed to just also the citizens in the community who are meant to benefit from the services.

568
00:52:26,630 --> 00:52:35,300
And so the um, because, again, I the kind of programs that I have selected to work on are programs that are very,

569
00:52:35,300 --> 00:52:40,160
very pervasive in a sense that they are implemented in practically every single village across the country.

570
00:52:40,160 --> 00:52:47,629
And, uh, they reach a very large proportion of people, uh, especially the poor in the country,

571
00:52:47,630 --> 00:52:50,510
or at least, uh, they're supposed to reach a very large proportion of them.

572
00:52:50,990 --> 00:52:58,960
So to that extent, as long as the information reaches this particular, um, very substantial group, I think we are reasonably fine.

573
00:52:58,970 --> 00:53:07,550
What has happened? Um, this is not by design that I mean, but for some of my activist partners have started adding the phone numbers of officials.

574
00:53:08,390 --> 00:53:14,960
Um, so that have been they want to, um, they want to inform them as well as local notables and whatnot, as a part of the system.

575
00:53:15,590 --> 00:53:19,010
Um, so in that sense, it's kind of happening organically to a certain extent.

576
00:53:19,010 --> 00:53:22,999
But, um, that's not my main target group in some sense.

577
00:53:23,000 --> 00:53:29,690
But there's a different, um, answer to, um, your earlier, the earlier part of your question as well, which is that,

578
00:53:29,780 --> 00:53:38,390
um, while corruption is pervasive, um, it is not uniform everywhere, and it is a changing phenomenon.

579
00:53:38,810 --> 00:53:42,230
There are states where public services tend to work remarkably well.

580
00:53:42,470 --> 00:53:48,500
Um, if that's a particular issue that you want to discuss more deeply on the practical.

581
00:53:48,800 --> 00:53:52,790
Precisely what my current book is on, which is to look at, uh, one particular state.

582
00:53:52,790 --> 00:53:55,399
The public services are delivered remarkably well.

583
00:53:55,400 --> 00:54:00,220
And I look at what is the, uh, what are the dynamics that led to this particular political commitment?

584
00:54:00,230 --> 00:54:07,070
And if you're interested in giving a talk, um, within Oxford again, day after tomorrow on the same subject, I'd be happy to send you the location.

585
00:54:07,490 --> 00:54:15,680
So the the key lesson in some sense is that where people are organised and when people are able to demand their rights effectively,

586
00:54:16,220 --> 00:54:17,940
the level of corruption does change.

587
00:54:18,020 --> 00:54:25,220
So what I'm hoping to do is to find strategies to enable that, in part by by supplementing current efforts with digital technology.

588
00:54:27,440 --> 00:54:30,780
I think about you. You can't be honest with what I thought I was going to go on.

589
00:54:30,800 --> 00:54:35,629
But nonetheless, I wondered, how far did your suits design lead you?

590
00:54:35,630 --> 00:54:42,320
Space to actually assess the impact on the quality of citizenship or practical citizenship in these environments?

591
00:54:42,830 --> 00:54:51,559
Um, so I'm, uh. So what I'm not trying to do is to look at the quality of citizenship, because there is a big selection bias and, uh, how, I mean,

592
00:54:51,560 --> 00:55:02,360
who I work with in itself, what I'm trying to do is to see whether my own intervention, um, can have an impact on the effectiveness of organisations.

593
00:55:02,360 --> 00:55:10,610
So one of the things that I'm trying to do is, uh, in areas where I work in a fairly substantial number of villages and trying to, um,

594
00:55:11,690 --> 00:55:16,399
randomly select villages within which the activists already have a presence, um,

595
00:55:16,400 --> 00:55:21,889
and then have this intervention for a year or two and see whether there is a difference in how far these programs work.

596
00:55:21,890 --> 00:55:32,000
Well, in these villages. Uh, it has interesting methodological, um, problems, partly because we send information based on local problems,

597
00:55:32,000 --> 00:55:35,030
which means that there is an endogeneity in the intervention directly.

598
00:55:35,840 --> 00:55:41,540
Um, the second element that I do not want to compromise on is that if you look at a lot of, um, I mean,

599
00:55:41,720 --> 00:55:46,580
our cities in general tend to be effective if your intervention is very, very, very standardised.

600
00:55:47,420 --> 00:55:55,459
Now, as far as I'm concerned, you cannot standardise and, uh, transparency intervention because it has to respond to local concerns.

601
00:55:55,460 --> 00:55:59,090
It has to respond to opportunities. It has to respond to context.

602
00:55:59,420 --> 00:56:05,330
So I want my intervention to be able to vary over a period of time, depending on what the priorities are and what comes up.

603
00:56:05,930 --> 00:56:11,390
Um, so I don't want to just blindly send exactly the same information to each place, which would have made it standardised.

604
00:56:11,870 --> 00:56:15,889
So but still, despite that, what I can do will be the following.

605
00:56:15,890 --> 00:56:20,240
Which is to say that, um, let's say there are 100 villages that I choose of them.

606
00:56:20,930 --> 00:56:23,540
These 50 are controlled by. This is a level of corruption.

607
00:56:23,540 --> 00:56:29,599
This is the timeliness with which payments are being made, etc. etc. in the other 50, I did have an intervention.

608
00:56:29,600 --> 00:56:33,409
I can document the intervention accurately because I know every message that goes out.

609
00:56:33,410 --> 00:56:37,610
I know how many people pick up the phone, I know how long they listen to the message and whatnot.

610
00:56:38,210 --> 00:56:44,750
And then I qualify with that by saying that in places where I have an intervention, this is the difference in the result.

611
00:56:44,750 --> 00:56:50,630
But I don't want to associate the metric by saying that seven phone calls leads to a point 5% reduction in corruption.

612
00:56:52,100 --> 00:56:57,500
It'll still, I think, help us to make, uh, qualified, measured yet a tentative assessment.

613
00:56:58,520 --> 00:57:01,669
I think we have time for about one more question, and I have seen Ben with his hand up,

614
00:57:01,670 --> 00:57:08,420
but is there anyone who hasn't asked the question who has a burning question before we give Ben's final opportunity?

615
00:57:09,680 --> 00:57:12,980
And, yeah, this is, um, what we talked about earlier, as you're looking at, um,

616
00:57:13,490 --> 00:57:18,150
maybe could you give a few minutes presentation of deliberation to criticism?

617
00:57:18,230 --> 00:57:21,920
Because I'm seeing some people here, um, for instance, and what that entails.

618
00:57:22,670 --> 00:57:27,979
So, I mean, it's, um, little sister to Oxford Internet Institute, uh, in some sense.

619
00:57:27,980 --> 00:57:33,650
Uh, so basically it was started by, um, a couple of political scientists and a computer scientist,

620
00:57:33,650 --> 00:57:35,870
partly because the political scientist who or one of them,

621
00:57:35,870 --> 00:57:43,069
Larry Diamond, who's teaches democracy, um, increasingly started having students who kept asking him, what about Facebook?

622
00:57:43,070 --> 00:57:46,520
What about Vietnam, what about this, etc., that he didn't have answers to?

623
00:57:47,060 --> 00:57:54,650
So he paired up with, uh, computer scientist called Terry Winograd, who is also the founder of the design school at Stanford University.

624
00:57:55,220 --> 00:57:59,810
Um, so initially, um, it was really a platform for us to learn.

625
00:57:59,810 --> 00:58:04,120
So we have a seminar series where, um, I mean, in two quarters, um,

626
00:58:04,130 --> 00:58:07,640
each year they've been inviting people to first learn our sense as to what's going on.

627
00:58:08,330 --> 00:58:11,629
And given the fact that, uh, we have two political scientists in the group,

628
00:58:11,630 --> 00:58:20,150
one of the early focus of the program has been on, uh, the impact of, um, digital technologies in international relations.

629
00:58:20,150 --> 00:58:24,790
So particularly on how authoritarian regimes have been using digital technology.

630
00:58:24,800 --> 00:58:29,750
So we've done some conferences around it, and there's a book that has come out of process, um,

631
00:58:29,750 --> 00:58:35,850
since I joined the this is about four years back because my focus has been on transparency as an activist,

632
00:58:35,850 --> 00:58:41,420
and that's probably the one thing that I kind of at least understand, but I don't really understand international relations at all.

633
00:58:41,930 --> 00:58:44,180
So we've been doing a lot more work on, um,

634
00:58:44,930 --> 00:58:51,020
in conversing with people who have been having the transparency initiatives to see what kind of initiatives people are taking up,

635
00:58:51,020 --> 00:58:54,550
what are the limitations, um, what are the interesting innovations that are happening?

636
00:58:54,620 --> 00:58:58,460
So we've done several conferences in India as well as in the US.

637
00:58:59,630 --> 00:59:02,840
Um, the most interesting initiative that has come out of that is, uh,

638
00:59:02,940 --> 00:59:10,190
out of this process is a massive open online course that will start in May on the use of technology for transparency,

639
00:59:10,190 --> 00:59:19,190
which is a consortium of organisations and scholars, uh, who will all talk about different ways in which technology shaping transparency practices.

640
00:59:19,190 --> 00:59:26,480
I've been trying to recruit Mark to this today. Um, so David will talk about how to transmit technology.

641
00:59:26,510 --> 00:59:33,350
Is being used to track political finances, elections, fighting corruption at the grassroots, etc. etc. um,

642
00:59:33,350 --> 00:59:39,920
apart from having short lectures which will be a part of the course, uh, I'm also trying to solicit people who have projects,

643
00:59:40,400 --> 00:59:45,110
either those who have ideas and want to build a tool which, um, Mark, among others,

644
00:59:45,110 --> 00:59:51,710
I'm hoping will be able to tip in, or people who have tools and who want that to be adopted internationally.

645
00:59:52,160 --> 00:59:59,750
So I'm trying to bring such live projects into the course, and we'll also be having, um, hackathons and other activities in my estimate,

646
00:59:59,750 --> 01:00:05,630
I think at least 10 to 15 countries, because I'm partnering this with National Democratic Institute, which has a presence in about 80 countries.

647
01:00:06,170 --> 01:00:12,470
Um, and we're also working with code for India, code for Pakistan, and a number of other hacker groups that have thousands of volunteers.

648
01:00:12,890 --> 01:00:17,690
So I'm it's it's something that I'm really looking forward to because it will be very action oriented,

649
01:00:17,720 --> 01:00:22,100
um, plus a very different model of teaching in some sense.

650
01:00:22,100 --> 01:00:23,750
And I'm not even sure if I can call it teaching.

651
01:00:23,760 --> 01:00:29,950
I have not understood if it's a conference or a course or a community building exercise, but it's it's it's own beast.

652
01:00:29,960 --> 01:00:36,020
Um, so in case any of you are interested about that, please do talk to me and I can tell you more about what we have in on cards.

653
01:00:36,560 --> 01:00:40,780
Um, so, like I said, uh, Patty, being one of the founders of design school,

654
01:00:40,790 --> 01:00:44,989
we've also been organising and been offering a course called Designing Liberation Technologies,

655
01:00:44,990 --> 01:00:50,150
where students go to Kenya, work with a set of NGOs that identify problems that they're working with.

656
01:00:50,750 --> 01:00:53,840
They come back to Stanford, learn design thinking as a part of the course,

657
01:00:53,840 --> 01:00:58,130
and then try to come up with, uh, phone based tools that could address some of these issues.

658
01:00:58,460 --> 01:01:05,360
Uh, it has led to a few promising projects, one example being that a lot of people in the slums that they've worked in lose

659
01:01:05,360 --> 01:01:09,770
documents constantly due to fire or flood or other kinds of natural disasters.

660
01:01:10,220 --> 01:01:15,799
So she's creating a simple intervention of, um, scanning the documents and storing them in the cloud,

661
01:01:15,800 --> 01:01:20,450
because when people have photocopies, they're able to get new copies of the documents relatively easily.

662
01:01:20,720 --> 01:01:23,090
So in that sense, when we find a promising project,

663
01:01:23,090 --> 01:01:26,930
we've been trying to offer seed funding to the students to go and actually implement it in reality as well.

664
01:01:27,260 --> 01:01:31,280
So those have been some of the initiatives, but we are really hoping to grow as we go along,

665
01:01:31,280 --> 01:01:37,610
because there's a lot of such work that happens across Stanford in practically every department imaginable.

666
01:01:38,150 --> 01:01:44,780
So we're trying to form a community of all of us so we can actually collaborate with each other, but also showcase to the world a little bit better.

667
01:01:44,780 --> 01:01:49,609
What action happens within Stanford? Great.

668
01:01:49,610 --> 01:01:56,690
Thanks. Thanks for this great work that you're doing. And, uh, thank you guys for for coming to listen.

669
01:01:56,960 --> 01:01:59,120
And please join me in thanking Vivek for.