00:01 All right.
00:01 Thank you everyone for coming today.
00:03 It's a pleasure to welcome
00:05 our colleague,
00:05 Biju Rao,
00:06 uh,
00:07 uh,
00:07 to this policy research talk on,
00:10 um,
00:10 democratizing democracy.
00:12 Biju is a lead economist in,
00:14 uh,
00:15 IEC and he's well known for his wide ranging work that brings together
00:20 ideas from economics,
00:21 anthropology,
00:21 sociology,
00:22 and political
00:23 science to study social,
00:25 cultural,
00:25 and political context of poverty.
00:27 Um,
00:28 so it's gonna be very interesting to hear,
00:30 um,
00:31 uh,
00:31 Bijus,
00:32 uh,
00:33 describe,
00:33 uh,
00:33 a body of ongoing research that he's been working on for quite a few years,
00:37 falling under this,
00:38 um,
00:38 bucket of making democracy work better for development.
00:42 Uh,
00:42 we're also very fortunate to have with us today,
00:44 Vikram Raghaven.
00:46 Um,
00:46 as our discussant,
00:48 um,
00:48 kudos particularly and thanks for stepping in at short notice because,
00:52 uh,
00:53 our,
00:53 uh,
00:53 previously scheduled discussant had to travel at very short notice.
00:57 Um,
00:57 and so I'm grateful,
00:59 uh,
00:59 Vikram,
01:00 that you stepped in to,
01:01 to be our discussion.
01:02 Uh,
01:02 Vikram is a lead counsel in the bank's legal
01:04 department and it's probably the first time that we have
01:07 somebody from LEG as a discussant in,
01:09 uh,
01:09 in this series.
01:10 So you're,
01:11 uh,
01:11 uh,
01:12 you know,
01:12 we have high expectations for your comments as a result.
01:15 Uh,
01:16 anyhow,
01:16 I'll turn over to,
01:17 uh,
01:18 to,
01:18 to beat you,
01:18 and then we'll hear from Vikram.
01:20 Thanks,
01:21 A.
01:25 Uh,
01:25 it's a great pleasure to be here,
01:27 particularly in the time of COVID,
01:28 the room is half full,
01:29 which is good.
01:31 Uh,
01:32 let me just plunge right into it.
01:43 So
01:45 This is a fact.
01:47 And people are losing faith
01:49 in electoral democracy.
01:51 And there's a lot of evidence for this around the world.
01:53 Uh,
01:54 the economists,
01:55 uh,
01:55 collects something called the democracy
01:57 index,
01:58 and you can see a sharp decline in the index over the last 5 or 6 years,
02:03 yeah,
02:04 uh,
02:04 worldwide.
02:05 Uh,
02:05 the Pew Research,
02:06 uh,
02:07 uh,
02:07 Center also does,
02:08 uh,
02:09 worldwide surveys on this subject,
02:11 and you can see that the
02:12 satisfaction with electoral democracy in various dimensions
02:16 has also been,
02:17 uh,
02:18 pretty bad in the recent past.
02:21 Uh
02:28 So
02:28 the question is,
02:29 why is this happening?
02:31 And uh
02:33 What a set of important reasons can be classified as the three C's
02:38 clientelism,
02:39 corruption,
02:39 and capture.
02:41 Clientelism is when it's sort of the politics of patronage,
02:44 when,
02:44 uh,
02:45 politicians interact with a subgroup of the population for political support,
02:49 providing them with benefits,
02:50 and,
02:51 uh,
02:51 what that results in,
02:52 amongst other things,
02:53 is
02:54 a much greater focus on private goods rather than public goods generation,
02:58 uh,
02:58 but also makes,
02:59 uh,
02:59 the allocations related to democracy less efficient.
03:02 Corruption,
03:03 corruption is related to the funding of elections,
03:06 uh,
03:06 election campaigns are expensive things in
03:08 many parts we have electoral democracy,
03:10 uh,
03:11 people fund elections
03:12 by,
03:13 uh,
03:14 by taking money from the public pot,
03:16 uh,
03:16 and therefore run
03:18 elections in an endogenous way.
03:20 What's also interesting though is that
03:22 that then segues over time
03:24 into a process by which,
03:26 uh.
03:27 Elections are used as a business.
03:30 Politics is used as a business.
03:31 You enter the political game in order to make money.
03:34 Yeah,
03:34 so corruption
03:35 is a very,
03:36 very strong correlate of electoral democracy in many parts of the world,
03:39 and that's related obviously to capture
03:41 the control of the electoral system by
03:43 economic and social elites,
03:45 which you find again all over the world,
03:46 and there's overwhelming evidence for this.
03:49 Uh,
03:50 there are other reasons that have
03:51 come up more recently in the last 10 years or so,
03:54 uh,
03:54 the role of social media,
03:55 the role of Facebook,
03:56 uh,
03:56 in the manipulation of elections,
03:58 which has come up in many different elections around the world,
04:00 and importantly,
04:01 the use of big data,
04:03 uh,
04:03 both by private institutions,
04:05 what,
04:05 uh,
04:05 Zubov has called,
04:07 uh,
04:07 uh,
04:07 surveillance capitalism,
04:09 as well as the recent,
04:10 uh,
04:10 2017,
04:12 um,
04:12 uh,
04:12 op ed in the Scientific American by a bunch of people,
04:15 by over 20 to 30 political scientists,
04:17 uh,
04:17 sorry,
04:17 computer scientists.
04:19 Talking about
04:20 how the manipulation of citizens by by governments
04:24 is undermining the democratic system.
04:26 So data is being used for the purpose of manipulation
04:30 rather than empowerment.
04:31 And finally,
04:32 what John Keane,
04:32 the political scientist,
04:33 has called monitory democracy,
04:35 the idea.
04:36 That democracy is increasingly
04:38 sharply decentralizing,
04:40 not in the traditional sense,
04:41 but into
04:42 civic groups,
04:43 self-help groups,
04:46 organizations that monitor institutions like ours,
04:48 the Bretton Woods Institute,
04:49 for instance.
04:50 All of these are what are known as monitoring institutions,
04:53 but getting smaller and smaller and smaller.
04:57 And therefore in some ways making the
04:59 need for electoral accountability less salient.
05:02 Now what all of this is doing
05:04 is that it is devitimizing the electoral process,
05:07 electoral processes around the world,
05:08 not just in
05:09 our perception,
05:10 but in the perception
05:12 of,
05:12 of people from the top to the bottom,
05:14 yeah,
05:15 so,
05:16 uh.
05:17 There are huge implications for development from this,
05:19 OK,
05:20 practically every development outcome that we care about,
05:24 uh,
05:24 rests in many countries on the assumption of effective electoral accountability.
05:30 And when you don't have that,
05:31 you have compromised political institutions,
05:33 you have poor accountability,
05:34 political instability,
05:37 a poor public service delivery,
05:38 conflict,
05:38 everything we don't like
05:40 in,
05:40 in,
05:40 in the development process.
05:41 Yeah.
05:41 So
05:42 the question is,
05:43 what can
05:44 development policy,
05:45 what leverage does development policy have?
05:47 What,
05:47 what leverage do institutions like ours have,
05:51 uh,
05:51 to help with this,
05:52 and,
05:53 uh.
05:54 The work that I have been involved with has tried to work with that at the local level
05:58 in
05:59 institutions in countries where local democracy is prevalent
06:03 at the local level,
06:04 how do we
06:05 work with that to make them,
06:07 uh,
06:07 deliver better outcomes?
06:09 Yeah,
06:10 uh,
06:10 it's,
06:10 it's a process that I've been working on now
06:12 for several years running something called the Social Observatory.
06:15 Uh,
06:15 and I've written a recent article in Deadlist that
06:17 I'm calling reflective development that sort of says that
06:19 when you work on these things,
06:20 you've got to remember it's,
06:21 it's a Sisypient task.
06:23 It's not something that happens
06:25 with magic bullets at the end.
06:26 It's,
06:26 you climb up a bit,
06:28 you scroll down,
06:29 climb up,
06:29 scroll down.
06:30 It's you have to play this sort of counterbalancing
06:32 game to all the forces out there that.
06:34 To make it worse,
06:35 it's an endless game,
06:37 yeah,
06:37 but
06:38 it's a,
06:38 but it's an important
06:40 thing to do because you're not focusing necessarily on outcomes but on processes.
06:44 A lot of us work on improving outcomes.
06:47 The whole RCT revolution is about that,
06:49 but here
06:50 the emphasis is on improving processes.
06:52 I should also very quickly say.
06:54 That the work I'll be talking about,
06:55 a lot has been done by a person sitting in this room,
06:57 Netra Balai Swami,
06:58 who's sitting here who's now
06:59 the senior economist for the Poverty GP in Nepal.
07:02 She and I have been doing a lot of this work together,
07:03 so I'm so glad she's with us today,
07:05 and I will focus on two themes.
07:07 Deliberation,
07:09 incentivizing politicians and democratizing data and let me go
07:12 through each one of them one by one.
07:14 Let's start with deliberation.
07:15 Uh,
07:15 my apologies for that uh design flaw.
07:18 Uh,
07:19 the,
07:19 the,
07:20 what is deliberation now,
07:22 uh,
07:23 why,
07:23 why do we have electoral democracy to begin with?
07:25 I mean,
07:25 at the end it is
07:26 to solve the problem of collective
07:28 action,
07:29 yeah,
07:30 uh,
07:30 when we
07:31 collectively elect.
07:32 A representative represent our views in some council or a parliament.
07:36 They're representing the views of a large body of people and therefore
07:39 are
07:40 resolving the problem of collective action by aggregating votes.
07:43 In other words,
07:44 aggregating preferences.
07:45 Yeah,
07:46 deliberation is a much older process by which we come to collective decisions,
07:51 not by electing a representative,
07:52 but by persuading each other of a point of view
07:55 through a process of debate,
07:57 discussion,
07:57 we come to what is called a single peak preference.
08:00 We come to a collective understanding.
08:02 It's a,
08:02 it's a,
08:02 it's an old tradition that goes back not just
08:04 to the ancient Greeks but to the ancient Indians,
08:07 the,
08:07 the,
08:07 the Romans,
08:08 I mean,
08:08 all over the world,
08:09 Chinese,
08:10 every,
08:11 every system,
08:11 every religious tradition has
08:13 deliberation is very much part of its history,
08:15 yeah,
08:16 so it's about the transformation of preferences rather than
08:19 the aggregation of references to quote John Elstone.
08:22 Deliberation in developing countries,
08:24 on which,
08:24 by the way,
08:24 every participatory mechanism that that's out there in development
08:27 that the bank has spent over $100 billion on
08:30 involves some level of deliberations.
08:31 We've invested as an institution a lot in these processes,
08:34 but there are challenges.
08:35 The first challenge is low literacy.
08:38 When as a citizen you don't understand facts
08:40 and figures because you can't read and write,
08:43 then how does that affect the quality of deliberation?
08:45 Do citizens have what Michael Wilcox has called the capacity to engage?
08:49 How does inequality work?
08:51 How does caste and class work?
08:52 Can deliberation be made better with policy,
08:55 uh,
08:56 gender inequality?
08:57 We all know that there's an enormous of gender inequality in any group interaction,
09:00 both in terms of who speaks and who gets heard.
09:03 Uh,
09:04 can that be resolved?
09:05 So all of these questions are some things that we'll be looking at today,
09:08 and I'll be looking particularly at one important institution.
09:12 Which is the Indian village meeting.
09:13 It's,
09:13 it's called the Grams Saha.
09:15 I'm using that word a lot,
09:16 uh,
09:17 which is in effect the largest deliberative institution
09:20 in human history because it affects the lives
09:22 of over 800 million people living in 2 million villages.
09:25 It's huge,
09:26 yeah,
09:27 and the history behind it.
09:28 We just done a book Paramita Sanar and I on this
09:31 with Cambridge University Press last year,
09:32 uh,
09:32 what we call oral democracy.
09:34 We just go through the basic argument in that book,
09:36 and I'm going to read some excerpts which means that I have to put up my.
09:43 Oh,
09:43 which is right here,
09:44 great.
09:44 OK,
09:45 so,
09:45 uh,
09:46 a bit bit of history here,
09:47 and I'm gonna fast forward to the post 1947
09:50 history after India got independence from the British,
09:53 1992.
09:55 India ratified the 73rd Amendment to the Indian Constitution,
09:58 and it said various things.
09:59 This amendment,
10:00 it said there would be regular elections held every 5 or 6 years to
10:03 every village would have a village council.
10:05 They'd have regular elections.
10:07 There'd be regular village meetings.
10:08 Ram Sabas held at least 2 times a year,
10:10 but up to 4 times a year.
10:12 1/3 of seats at a minimum on all for all village councils and.
10:15 Village presidencies will be reserved for women.
10:17 A proportion of seats will be reserved for underprivileged groups,
10:20 scheduled castes,
10:21 and tribes,
10:21 and it was a hybrid institution.
10:23 You had elections,
10:24 elected representatives,
10:25 and village meetings.
10:26 So
10:26 the way the Indian constitution writes about it is that
10:29 it talks about the village council.
10:31 The village president is the executive of the village,
10:33 and the grounds of the village meeting is the parliament of the village.
10:35 Yeah,
10:36 and in previous what we found that.
10:39 The,
10:39 the prevalence of these meetings is highly correlated in a
10:43 good way with the reduction in corruption and clientelism.
10:45 We didn't manage to do a causal analysis,
10:47 but the correlation is very strong,
10:48 yeah.
10:49 Now what is the purpose of Rams of us?
10:51 A very important purpose is to ratify beneficiaries for anti-poverty programs.
10:55 There's a list that comes up through proxy means testing,
10:58 which
10:58 many states
10:59 are ratifying the grounds,
11:00 but they have to actually say that this list makes sense.
11:03 They debate expenditures on public goods.
11:04 They monitor and track the performance of councils,
11:07 uh,
11:07 so they have very,
11:08 very important functions.
11:09 But,
11:09 uh,
11:10 what is,
11:11 uh,
11:12 important to note,
11:13 uh,
11:13 is that,
11:14 uh,
11:15 this.
11:17 The enabling of these acts,
11:18 there's a constitutional amendment,
11:20 but the way the acts work in practice
11:22 is that
11:23 they have to be
11:24 enabled through
11:27 acts of the state government.
11:28 India is a federal structure,
11:30 so each state has
11:32 strong sort of jurisdiction over over village government,
11:36 and each state has to enable
11:38 a
11:39 state act that
11:40 interprets that constitution in terms of what the state sees as important
11:45 in its context and
11:46 The four states we'll be looking at are Kerala,
11:49 Tamil and Andhra Pradesh,
11:50 which is
11:51 those of you who know India,
11:52 this is
11:53 all data from 2004.
11:54 So Andhra Pradesh was still one state,
11:55 had been split into two,
11:57 and Kerala is by far the most famous of the state acts because
12:01 in 1996 they passed something called the People's Campaign,
12:04 and they.
12:04 They said
12:05 they devolved
12:06 40% of the state's budget
12:09 to village governments.
12:10 They also said
12:11 that uh
12:13 uh
12:14 all state plans would be done through a nested planning process,
12:18 going from streets
12:20 to neighborhoods,
12:21 to wards,
12:22 to villages.
12:23 And upward,
12:24 yeah,
12:24 and each system would collect ideas from people at that level
12:28 and slowly collate them as they went up.
12:30 It's a nested planning process.
12:31 And in order to enable this,
12:33 there's a huge thing called the People's Campaign.
12:34 They sort of said with all of the state training people in how
12:37 to participate and how to make decisions and all this kind of thing.
12:39 So it's a tremendous thing called the Kerala People's Campaign.
12:41 A lot has been written about it.
12:43 Kerala,
12:43 so we're calling high quality
12:45 uh uh decentralization.
12:47 Karnataka and Tamil Nadu,
12:48 but both neighboring states of Kerala,
12:49 we're calling medium.
12:50 Karnataka,
12:51 which is where I'm from,
12:52 is the state that pioneered this village Panchai system,
12:55 as it's called,
12:57 um,
12:57 and,
12:57 and,
12:58 uh,
12:58 also gives these village governments a lot of power,
13:00 but not as much as Kerala.
13:02 Tamil Nadu is similar to Karnataka in that sense,
13:04 it really enabled the space.
13:05 Uh,
13:05 the worst
13:06 of the lot is the state of Andra Pradesh,
13:08 which took a very technocratic approach to development
13:11 and emphasized the role of self-help groups and these
13:14 non-constitutional bodies gave them a lot of power
13:16 that.
13:17 Uh,
13:18 means that they,
13:19 they didn't even hold elections for a long time actually at,
13:21 at the village level.
13:22 They had to be called to the Supreme Court to hold those elections,
13:24 which means that it had a pretty weak,
13:26 uh,
13:26 franchise system we're calling it low quality.
13:28 Now,
13:29 uh,
13:30 the methodology here
13:31 is that we have recorded 3,
13:34 about 300 g of us in these 4 South Indian states,
13:37 transcribed and translated them from the South Indian language,
13:40 the 4 South Indian languages into English.
13:42 Uh,
13:43 and on the basis of the national experiment,
13:44 which I'll explain in a second,
13:46 and I'm focusing on two things.
13:47 I'm focusing on
13:48 to what extent can policy interventions affect the quality of discourse,
13:52 and secondly,
13:53 what's the role of literacy,
13:54 and I,
13:54 I should say this book took 10 years to write
13:56 part of it that I worked in for about 10 years
13:58 because it's all a handmade analysis.
14:00 This was way before machine learning was in our lives.
14:02 Yeah.
14:03 Now
14:03 this is an actual experiment.
14:06 And to explain it very quickly,
14:07 when you
14:08 start with a concept that linguists call,
14:10 uh,
14:10 linguistic communities.
14:12 What are linguistic communities?
14:13 The idea of linguistics is
14:14 that if
14:15 you share
14:16 a common language,
14:18 if you share
14:19 common social structures,
14:20 if you share common legislative histories,
14:23 if you share common sort of uh economic geographic sort of context,
14:27 then
14:28 you should speak more or less the same way,
14:30 which seems like a very reasonable thing.
14:31 There's a lot of evidence to show that that is in fact the case.
14:34 So the next experiment here is
14:36 prior to 1956,
14:38 India was organized.
14:40 Along the lines
14:42 of of of
14:43 of of of governmental structures of of of jurisdictions
14:46 that had survived for 300 years ever since the Mughal era
14:49 and in South India
14:50 there are several sort of political entities,
14:52 but I'm going to focus on 21 is Old Hyderabad State,
14:55 which is the south end of the Mughal Empire
14:57 that was run by the Nizam of Hyderabad,
14:59 uh,
14:59 who was a very feudal
15:01 king,
15:02 yeah,
15:02 and,
15:03 and,
15:03 and.
15:04 Presided royal feudal territory.
15:06 The rest of it
15:07 that we'll be talking about
15:08 is the Madras presidency,
15:10 which was ruled directly by the British.
15:12 OK,
15:12 these are very large political entities.
15:14 Hyderabad state had 5 languages spoken in it
15:17 Kannada,
15:18 Telugu,
15:18 Marathi,
15:19 Urdu,
15:20 Hindi,
15:21 and I think I forgot Marathi,
15:22 you know,
15:22 so there were,
15:23 there were several languages.
15:25 Midra's residency had 10 languages spoken in it.
15:27 I'm not going to recite all of them,
15:28 yeah,
15:28 so what happened is that in 1955 there's
15:31 something called the Panikar Committee that said,
15:32 OK,
15:33 this makes no sense.
15:34 We have to reorganize Indian states along linguistic lines,
15:37 yeah,
15:38 so that the states speak one unified language.
15:40 So they reorganized it into four states
15:43 Uttar Pradesh that spoke Telugu,
15:44 Tamla that spoke Tamil,
15:45 Kerala that spoke Malayalam and Karnataka that spoke Kannada,
15:48 yeah,
15:49 but
15:50 along the border when you try to do this crazy exercise,
15:53 you're going to make mistakes.
15:54 So in effect
15:56 what we have
15:57 are mistakes.
15:58 We have people who speak Kannada and Karnataka
16:00 10 kilometers away on the other side of the border
16:03 with
16:03 people who speak Kannada in the other state now,
16:06 yeah,
16:07 but prior to 1956 and 300 years,
16:09 they all belong to the same political entity.
16:11 They're 10 kilometers away from each other,
16:13 so they're the same linguistic community.
16:15 They have the same
16:16 economic base,
16:17 and importantly,
16:18 I've shown this in other work,
16:20 they shared the same caste structure.
16:22 So they're the same social and political structure.
16:25 In other words,
16:26 in everyday conversation,
16:28 they speak the same way.
16:29 I speak a lot waving my arms like this and loudly and,
16:32 you know,
16:32 passionately.
16:33 I'm not passionate.
16:33 I'm just South Indian,
16:34 OK,
16:35 so
16:36 I should meet other South Indians.
16:37 There are some in the room here.
16:39 Right,
16:40 so,
16:40 so they probably speak to you the same way,
16:42 yeah,
16:43 so,
16:43 so,
16:43 uh,
16:44 uh,
16:44 so,
16:44 so that's what we,
16:45 that's what we're working.
16:46 We're going to compare villages
16:48 on either side of these state borders that prior 19056 belong to the same
16:53 political entity and about 10 to 20 kilometers away from each other
16:56 and speak the same language,
16:58 OK.
16:59 And what I want to show you
17:00 is that if
17:02 the state borders matter today after 1956,
17:04 we can attribute that to something that happened after 1956.
17:07 So the main thing that happened in this case
17:09 was
17:09 the way the state governments implemented the 7th Amendment.
17:13 OK,
17:14 so here's some pictures of grams of us.
17:16 And just to give you a sense of the topics,
17:18 the main topic there is drinking water,
17:20 but you have roads,
17:20 you have electricity,
17:22 you have housing,
17:22 and so on that come up,
17:24 uh,
17:24 and I'm not going to read a few excerpts.
17:26 First we're going to compare
17:27 Chittoor district in Andhra Pradesh,
17:29 Southern Antar Pradesh,
17:30 with Dharmauri district in Tamil Na state,
17:32 and I'm going to read out the Rams Sabas to you,
17:35 yeah.
17:35 Uh,
17:36 I want to start with Chitto District.
17:37 I'm going to read out the entirety of the ground swa.
17:39 OK.
17:40 The president starts
17:41 today.
17:42 We're conducting this grounds server to discuss the problems in
17:44 our village and the various activities we've undertaken so far.
17:46 You can express your problems here.
17:48 A,
17:48 a,
17:48 a villager says there's no proper community hall
17:51 in this village for holding meetings or events.
17:52 We should construct a community hall.
17:53 The president says,
17:54 I will inform the government to construct a community hall.
17:57 Second villager says
17:58 there are no cement roads in this village.
18:00 Cement roads should be laid on all the village streets.
18:01 The president says.
18:03 Wherever we don't have cement roads,
18:04 I will instruct the government to construct cement roads.
18:06 Third villagers,
18:07 and this is a direct transcript.
18:09 The villager says
18:10 there are electric,
18:11 electricity poles in the streets,
18:12 but the lights are not there.
18:13 You should arrange for the lights.
18:14 The person says,
18:15 I will arrange for street lights soon.
18:17 Uh,
18:17 then another villager says,
18:18 We don't have a proper cemetery or graveyard of the village.
18:21 Uh,
18:21 some people throw dead bodies on the outskirts of the village,
18:24 where they're causing health problems for us,
18:25 for our children.
18:26 Uh,
18:27 president says,
18:27 I've given a complaint to the district collector about this.
18:30 I,
18:30 I hope he solves the problem.
18:32 That's it.
18:33 That's the groundswell.
18:34 Uh,
18:34 I want you to note that as I showed here,
18:37 where I don't have that table,
18:38 an average Gramsava is usually about 90 minutes long.
18:41 This thing lasts about 5 or 10 minutes,
18:43 yeah,
18:43 the entirety of the Gram server.
18:44 Now let's just go across the border 10
18:46 kilometers away to the Dharmaapuri district in Tamilan,
18:49 and I'm going to read out a brief excerpt from there.
18:51 Uh,
18:52 now I should say this is a Gramsam that lasts for 2.5 hours,
18:55 so I'm going to give you brief glimpses.
18:56 I will try to speak with a Tamil accent,
18:58 but forgive me if I don't.
18:59 So,
19:00 uh.
19:01 Say for your daily toileting needs,
19:02 you can use the unused land and fields,
19:04 referring to open defecation
19:06 in villages,
19:06 both men and women do that.
19:07 This is the president speaking.
19:09 But when these lands and fields are no longer there,
19:11 then what will you do with all your daily need?
19:13 Will it not become difficult?
19:14 Males can go anywhere,
19:15 but where the ladies go,
19:16 they will face a lot of problems.
19:18 So what I say is that all of you should have a toilet.
19:20 In cities there are thousands of houses.
19:22 Each house has a toilet.
19:23 It is,
19:24 it is clean.
19:24 Has the city been spoiled?
19:25 No.
19:26 So similarly,
19:26 if we have a toilet in each and every house,
19:28 then our village will also be clean.
19:30 I'm going to fast forward a bit.
19:31 In life there has to be some improvement.
19:33 You have to improve.
19:34 You do not have to spend more money for that.
19:36 For each dollar,
19:36 the government is funding ₹500.
19:38 In this village we'll build 20 public convenience latrines for ₹2000.
19:42 So please go ahead and do this.
19:44 And then the villagers start talking.
19:45 There's a lot of noise,
19:46 there are,
19:46 many voices,
19:47 and the villager says,
19:48 You're giving us only ₹500 to build a toilet.
19:50 How can we do that?
19:51 And then another authority says,
19:53 we're just aiding you of $500.
19:55 That should come out of your pocket.
19:56 I can't.
19:57 The village government doesn't have so much money.
19:58 There's a lot of shouting and screaming and debate on all of these things.
20:01 Is that enough money?
20:02 This is the kind of discourse you have going on in Tamil Na.
20:05 Just 10 kilometers away from Tamil speakers
20:08 Ira Pradesh.
20:10 So let's now go
20:11 again looking at Andhra Pradesh
20:13 to northern Andhradesh,
20:14 Mira,
20:15 the Merak district
20:16 and
20:17 And again I'm gonna read to you the entirety of the grounds.
20:21 The first villager says
20:22 there is no linking growth to our village.
20:24 The drainage and water problem persists.
20:26 I wish the government at least now will look into our problems and do something.
20:29 The villager says there is a problem.
20:31 Another,
20:32 another village has a problem of transportation.
20:34 Third villager says
20:35 Namaste or namaste now,
20:37 uh,
20:37 to all members of the Gramppanchai,
20:39 we have become independent more than 50 years ago,
20:41 yet we haven't developed.
20:42 Whenever a problem was raised,
20:43 it was only discussed.
20:45 Uh,
20:45 after Mr.
20:45 Ashok Desh was elected,
20:46 the press seem things seem to get better,
20:48 but nothing has happened since then.
20:50 Another villager says there has been some development,
20:52 but more needs to be done.
20:53 Another villager says something else.
20:55 The president ends by saying,
20:56 OK,
20:56 you expressed your problems.
20:58 We will note them down and try to do something.
21:00 But a villager at the end,
21:01 a wag at the end,
21:02 comes in and says,
21:03 What is the use of writing down?
21:04 Has the government done anything up to now?
21:06 You expressed your problems and said what needs to be done.
21:10 It is please give us an assurance there is a budget with this,
21:12 not listened to.
21:13 That's the end of the Grams Saba.
21:14 OK,
21:15 now we're gonna go
21:16 to my home state of Karnataka
21:18 and,
21:19 uh,
21:19 listen to the folks at Ber
21:21 who are basically the same people,
21:23 all formerly from the Hyderabad state,
21:25 and again,
21:26 a 1.5 hour Grams Saba,
21:27 a brief excerpt.
21:29 The secretary says.
21:30 Now I should explained it.
21:32 They're talking about a jet dam.
21:33 A check dam
21:34 is a is a small cement construction to collect rainwater.
21:37 OK.
21:38 The secretary says
21:39 one check dam is completed,
21:40 but there's provision for one vote in the budget.
21:42 The villager says we don't need another check dam.
21:44 The engineer who's the local engineer says you can't reject the check dam.
21:48 The secretary says
21:50 there is pressure on us to execute the check damn.
21:51 Villager says,
21:52 if it is so,
21:53 let us propose to remove silt instead of reconstructing a check dam.
21:57 The villagers,
21:57 the secretary says,
21:58 No,
21:58 no,
21:58 no,
21:59 please tell them we have to.
22:00 Why is this going on?
22:01 Because every time you do a construction of this kind,
22:04 there is money to be made.
22:06 Yeah,
22:06 and villagers know this game and they're trying to stop that from happening.
22:09 So literally they go on for about half an hour debating this issue.
22:13 They're getting checked.
22:14 The check time is getting checked,
22:15 yeah,
22:16 in within the ground.
22:17 OK,
22:18 so,
22:18 uh,
22:19 let me just end a bit,
22:20 uh,
22:20 with these excerpts since we run running short of time.
22:22 I don't wanna go on and on about this,
22:24 uh,
22:25 and let's go straight
22:26 to Palaga in,
22:27 in,
22:28 in Kerala.
22:28 So I want to show you what a Kerala Gramsaha is like,
22:30 and I can assure you
22:31 that the Tamil Na Gramsabas are similar across different parts of Tamil Na
22:34 and
22:35 very different from across the border in,
22:37 in,
22:37 in,
22:37 in Kerala.
22:38 Kerala,
22:38 remember.
22:39 This ground surba is the end of a long process.
22:41 They've had their nested planning at the village level.
22:44 They're at the street level.
22:45 They've come up with their decisions.
22:46 So by the time they come to the ground saba,
22:48 they're simply making announcements.
22:50 Yeah,
22:50 and you,
22:51 and you know,
22:51 World Bank folks and other bureaucrats will
22:53 be very familiar with these announcements.
22:54 I just want to read it out and tell you this is a like grounds.
22:56 It's very boring,
22:58 and this is an announcement,
22:59 and it goes on for an hour,
23:00 and the screener.
23:02 We're in the last phase of financial
23:03 year 2005-2006 planning according to government principles,
23:05 we're organizing this grants for planning the
23:07 development of the next financial year 2006-2007
23:09 as part of the financial year 2006-207,
23:11 we've organized working groups.
23:12 The new working group has suggested some projects.
23:14 The grants have also made suggestions.
23:16 Those suggested projects are submitted to you.
23:17 We will change those projects on the basis of the valid suggestions
23:20 and then formulate the working committee.
23:21 The working committee implements the project with
23:22 the help of the ruling committee.
23:23 We cannot implement so on and so forth.
23:25 He's basically making a set of announcements
23:27 that is telling you what the ground plan is for the next year.
23:32 It is a very efficient system,
23:33 and there's a lot of research showing
23:35 that actually does pretty well,
23:37 yeah.
23:37 So
23:38 the point of this
23:40 using this national experiment
23:41 is to show you
23:42 that state
23:44 policy matters for discourse.
23:46 It encourages the capacity of citizens to speak,
23:50 and it makes them speak in a way that is tailored
23:53 according to what the state government has empowered them to do.
23:56 Policy matters
23:57 in improving the quality of discourse.
23:59 We're not talking about outcomes yet.
24:01 Uh,
24:01 I'll talk about in a bit
24:02 now.
24:02 What is the effect of literacy?
24:03 There's a lot of stuff,
24:04 you know,
24:04 people are illiterate.
24:05 There's a range of,
24:07 there's a very illiterate part of the world in those days.
24:09 33% is a low level of literacy,
24:10 66% literate is about medium,
24:13 and then you have 100% literacy in Kerala.
24:15 And if you,
24:15 and what we did was basically compared,
24:18 uh,
24:18 low literacy and medium literacy and high literacy,
24:21 uh,
24:22 uh,
24:22 villages within the same state border within the same district.
24:25 So we're doing that across state comparison now.
24:27 We're looking within the same district to see how that varies.
24:29 And the basic point here
24:31 is that in Andhra Pradesh,
24:33 literacy makes no difference.
24:36 In Kerala,
24:37 literacy makes no difference.
24:38 Literacy makes a much bigger difference in the medium capacity states,
24:42 and the reason is obvious because there,
24:44 there is some room for literacy to make a difference
24:46 because the,
24:47 the,
24:47 it's sort of medium empowered in Kerala,
24:49 everybody's literate,
24:49 so everybody speaks the same way.
24:51 In Utar Pradesh,
24:52 it doesn't matter because the government doesn't do anything anyway.
24:55 Yeah,
24:55 so literacy matters,
24:57 but it is
24:58 superseded by state capacity.
25:00 Yeah,
25:01 so that's that book.
25:03 Let me switch a bit now to a second paper
25:05 that I worked on,
25:05 uh,
25:05 with Netra and with Ramya Pathasar who's now
25:07 unfortunately with McKinsey I can say that,
25:09 uh,
25:10 so this is basically an NLP analysis.
25:12 Now we're going fast forwarding to 2014 when we collected data from 100
25:15 g bars in the state of Tamil Na
25:17 and,
25:18 and,
25:18 uh,
25:18 we're gonna do some natural language.
25:20 Processing on it
25:21 as we're going to a different method altogether
25:23 and the two questions here given the context of
25:26 Grams about effective for citizen participation and is the gender
25:29 inequality in Grams of birth and can it be
25:30 improved and look we're going to look at three metrics
25:32 equality of participation agenda setting power,
25:34 and responsiveness of the state,
25:36 uh,
25:37 so.
25:37 Looking at the topic variation again,
25:39 the topics are similar to what we saw in 2004,
25:41 yeah,
25:42 water,
25:43 uh,
25:43 selection of beneficiaries,
25:44 employment and wages,
25:45 service failures,
25:46 and so on,
25:46 yeah,
25:47 uh,
25:47 but first looking at equality of participation,
25:49 you see
25:50 that there are tremendous differences in by gender and who speaks,
25:54 tremendous differences by gender.
25:56 So there's a lot of gender inequality in who speaks.
25:58 And
25:59 you can look at the length of pieces of speeches
26:01 as we just don't care whether they spoke or not,
26:03 but how long they spoke.
26:04 Again,
26:04 a lot of gender inequality,
26:06 but that gender inequality is not coming from citizens,
26:08 it's coming from politicians.
26:10 Male politicians tend to go on and on,
26:12 yeah,
26:13 and we have noticed that in the,
26:14 in our current presidential process here too.
26:16 OK,
26:17 so agenda setting power and,
26:18 and,
26:18 and,
26:18 so that we do more analysis with this.
26:20 I'm just gonna briefly show you what movement on that
26:22 agenda setting power is
26:24 when I speak as a speaker,
26:27 does that topic
26:28 get carried through over the next speech and over the next 5 speeches?
26:32 Am I as a citizen able to set the agenda and also how does it differ by gender?
26:36 And what we see there is basically this is interesting,
26:39 citizens have much more agenda setting power than officials or politicians.
26:44 Yeah,
26:45 But women
26:46 are slightly disadvantaged.
26:49 Right
26:50 now,
26:50 uh,
26:50 one of the things you should keep in mind in these,
26:52 in Tamilland in this,
26:53 at this point is that 50% of village Panchai presidencies are reserved for women,
26:57 and this is done on a quasi-random basis.
26:59 We're a natural experiment,
26:59 we actually have a real experiment going on here.
27:01 So we're going to see whether village presidencies
27:04 can counteract this gender inequality,
27:06 and we see that it does.
27:08 When you have a woman president,
27:10 women citizens have greater agenda setting power.
27:15 Yeah,
27:15 you can read the paper if you like.
27:16 I'm,
27:16 I'm just going to go through this very quickly.
27:18 And then the third thing you'll be looking at is responsiveness of the state.
27:21 If I speak on a certain topic,
27:23 what is the probability
27:24 that the official at the at the head of the table is
27:26 going to pick up on that topic and go on with it?
27:29 The
27:29 the
27:30 the dialogic responsiveness
27:31 of the state.
27:32 And what we see here
27:33 is that.
27:35 There's not much gender equality,
27:37 uh,
27:37 but,
27:37 but,
27:38 you know,
27:38 there is,
27:39 there is,
27:39 there is a bit,
27:39 and,
27:40 and we go to the regression to see that,
27:41 that,
27:41 that,
27:41 that women tend to be quite disadvantaged on,
27:44 on,
27:44 on-topic responses,
27:46 but look at that.
27:47 If there is a female president,
27:50 women tend to have more agenda and gender setting power.
27:53 So what does that mean?
27:55 Women are listened to more
27:57 when
27:58 there is a woman president.
28:01 Not just are they able to speak more,
28:02 they're listened to more.
28:04 It's what
28:05 uh
28:07 uh what they're calling a pessimic discrimination,
28:09 following the work of Miranda Finger,
28:11 yeah.
28:12 It's a discrimination is not just about whether you're allowed to speak.
28:14 Discrimination is also about whether you're being listened to.
28:17 We find that this
28:18 makes a big difference.
28:19 So,
28:20 The point of Grams Saba,
28:21 and I'll stop with the Graham Saba work here,
28:23 is to show you that,
28:25 that,
28:25 that citizens to participate.
28:27 Literacy matters,
28:28 but not in the way you think.
28:29 What is far more important is policy.
28:31 Gender equality exists,
28:33 inequality exists,
28:33 but it can be fixed
28:35 with
28:35 uh having women presidents.
28:37 Yeah.
28:38 Now,
28:38 I'm going to shift now to
28:40 away from the speech from deliberation to the idea that.
28:43 Of whether politicians can be incentivized to perform better
28:46 and this is a recent paper that we have not yet written up,
28:48 but I'm going to show you brown the brief results.
28:50 Now firstly,
28:50 I should say there's a large literature
28:52 on different ways of trying to improve a political performance.
28:55 Uh,
28:56 what most relevant here is on salary incentives were popular,
28:59 but for politicians in Brazil,
29:00 by Perraz and Finan,
29:01 who,
29:02 who find that higher wages for politicians to
29:04 make them perform better on the various metrics,
29:06 but generally
29:07 the literature is quite mixed now our.
29:09 Thing looks at two different kinds of incentives.
29:11 They're both public incentives.
29:12 They're not salary incentives.
29:14 We have two arms.
29:15 The first arm is a randomized trial.
29:16 The first arm
29:17 is,
29:18 uh,
29:18 uh,
29:19 giving
29:20 the president
29:21 a larger public budget.
29:22 The second arm
29:23 is simply giving them a certificate that they can use to run elections with.
29:26 They're getting no money,
29:27 yeah,
29:27 uh,
29:28 so.
29:30 Uh,
29:30 village panchays and Tamil Na perform a lot
29:32 of activities and public goods and private goods,
29:34 so that's just to keep in mind there's a lot,
29:35 they do a lot of work.
29:37 Uh,
29:37 so our experiment looks at 200,
29:39 uh,
29:40 villages,
29:40 village panchays across 10 blocks,
29:42 and we randomly assigned the village panchayats
29:45 into 3 categories a control group,
29:46 a treatment arm for the financial incentive,
29:48 a treatment arm for the kudos or the or the non-financial incentive.
29:52 Uh,
29:53 and,
29:53 and you know,
29:53 in 2015,
29:54 all the presidents of these villages received an official letter
29:57 in control of,
29:57 of,
29:58 of villages.
29:59 They were not told anything.
30:00 Basically you'll be graded into three categories A,
30:02 B,
30:02 or C,
30:03 but you'll get no incentives.
30:04 They didn't say anything about incentives in the treatment GPs,
30:06 uh,
30:07 that
30:07 letter was given that,
30:08 that you'll be graded into A,
30:09 B,
30:09 or C,
30:10 but along with that was said you'll either get
30:12 a financial incentive or a non-financial incentive.
30:14 Yeah,
30:15 uh,
30:16 OK,
30:17 that's sad.
30:22 Hm
30:29 What I do right?
30:31 Hm
30:36 OK.
30:37 Um,
30:37 but,
30:38 but
30:40 It's not moving.
30:45 OK,
30:45 so,
30:46 so let me,
30:46 let me try and fix that while I speak up.
30:48 So
30:49 while that's being fixed,
30:50 the,
30:50 the,
30:50 the,
30:50 the,
30:51 let,
30:51 let me basically give you the gist of,
30:52 of,
30:52 of the results.
30:54 Uh,
30:55 what we find
30:56 is that
30:58 I mean,
30:58 firstly,
30:59 I,
30:59 I got,
30:59 I got to qualify this.
31:01 It's very important how that monitoring is done,
31:03 how that ABC criteria are determined,
31:06 and what we do,
31:07 what we did very,
31:07 very carefully along with the Tamil Na government
31:09 was that the people assessing the performance of
31:11 these residents were not people from the village.
31:14 They were people from districts very far away.
31:16 A committee was formed of knowledgeable people,
31:17 ex-presidents and so on,
31:19 uh,
31:19 women's health group members who came along and set with a set of criteria to assess
31:23 villages and districts that were.
31:26 More than 100 kilometers away.
31:28 Therefore,
31:29 there was no
31:30 way that these guys could be,
31:31 could be,
31:31 could be bribed into performing uh differently into,
31:34 into,
31:34 into,
31:35 into,
31:35 into,
31:35 into,
31:35 into,
31:36 into giving them a better grade just because they were bribed.
31:38 Yeah.
31:39 As a result of that.
31:41 Uh.
31:44 The,
31:44 the monitoring process,
31:45 the scoring process was considered legitimate.
31:47 Nobody questioned it.
31:48 That's,
31:49 that's centrally important.
31:50 If you're going to have an incentive system of this kind,
31:52 how you score people is centrally important,
31:54 yeah.
31:55 Having said that,
31:56 what we found,
31:57 find is some remarkable results,
31:58 I hope I can show them to you,
31:59 uh.
32:02 OK,
32:02 so the first set of results is
32:04 that
32:06 The incentives improved both the provision of private goods
32:10 and the provision
32:11 of public goods,
32:12 yeah,
32:13 uh,
32:14 and,
32:14 and by the way,
32:14 the incentives are pretty big.
32:15 The great residents received about double their village budgets.
32:19 Uh,
32:20 the great residents received an equivalent amount,
32:22 and they could do a lot of work with that.
32:24 And I want to show you the certificate that they received,
32:26 which is very,
32:27 you know,
32:27 sexy certificate,
32:28 uh,
32:29 which they could use for,
32:30 you know,
32:30 I'm an a great president,
32:31 so please elect me kind of thing.
32:33 Uh,
32:34 we had a lot of data.
32:35 Uh,
32:35 you have sort of three criteria,
32:37 about 22% of the sample was A grade,
32:40 most of them were B grade,
32:41 and the rest were C grade,
32:42 uh,
32:42 and what you see that,
32:43 that,
32:44 you know,
32:44 whether you've had a financial incentive or a non-financial incentive,
32:47 they both equally improved,
32:49 uh,
32:50 access to private transfers.
32:52 Similarly,
32:53 they both equal,
32:54 they're not equally,
32:54 but you know,
32:55 the financial incentives did a little bit better on access to public investments,
32:59 but if you index uh all the public investments that were made,
33:02 they both did equally well,
33:03 more or less,
33:04 yeah,
33:05 and this is the
33:06 interesting result
33:08 that.
33:09 On average,
33:10 if you didn't have the incentives,
33:12 residents tend to favor their own neighborhoods and their own people.
33:16 But with the incentives,
33:17 they started favoring others.
33:19 It countered clientelism.
33:21 It went away from the group-based subgroup-based approach
33:24 that we're all worried about a democracy,
33:26 yeah.
33:27 So
33:28 what we find with this incentive system
33:30 is that it improves access to private transfers,
33:32 improves access to public goods,
33:33 uh,
33:34 and it reduces corruption,
33:35 nepotism,
33:36 and clientelism.
33:37 Yeah.
33:38 Now finally,
33:38 I'm going to go to democratizing data,
33:40 which is work
33:41 that we did several years ago,
33:42 but there is an update to it that I find quite interesting.
33:45 Here the point is,
33:46 with all this data manipulation taking place,
33:49 can't we give citizens access to their own data?
33:53 Can't we give them the ability to design,
33:56 collect,
33:57 and importantly,
33:58 analyze their own survey data?
33:59 This is what we pulled off
34:01 in a in a pilot with 200 villages,
34:03 uh,
34:03 across 32,000 women in Tamil Na.
34:06 and it's,
34:06 and you know,
34:07 and,
34:07 and the first,
34:08 the,
34:08 the.
34:09 itself was determined by women's health help groups
34:12 in these villages using a deliberative system.
34:14 Yeah,
34:14 and it's very interesting the kind of questions they came up with.
34:17 Let's take the first one.
34:18 We typically ask in an LSMS,
34:20 you know,
34:20 how much did you spend on rice in the last week or vegetables.
34:23 They had one question,
34:24 which is basically,
34:25 does the person who eats last in the family get enough to eat?
34:29 In other words,
34:30 does the mother who's serving everybody ever go hungry?
34:32 Yeah,
34:33 uh,
34:33 and other issues that were prioritized quite differently,
34:36 digital participation,
34:37 can you use a mobile phone you don't know,
34:38 can you read and,
34:39 uh,
34:39 uh,
34:39 and send text messages?
34:41 We'd also developed a decentralized system of data collection where
34:45 we selected one person,
34:46 one woman in every village,
34:48 uh,
34:48 to be the data collector,
34:49 gave her an,
34:50 a,
34:50 a,
34:50 a,
34:51 a,
34:51 a tablet with a copy-based app
34:53 that was designed extremely simply on a color coded method.
34:57 The data that were collected in the tablet
34:59 were dispatched directly to a cloud server,
35:01 so nobody got to manipulate it in the middle.
35:03 So,
35:04 and,
35:04 and,
35:04 and all the validation was done,
35:06 uh,
35:06 uh,
35:07 within,
35:07 within,
35:07 within,
35:08 within the system.
35:08 So there's no need for external validation.
35:10 So designed the users with low digital and then importantly
35:14 we visualized the data
35:16 for people in the village in a way that was co-produced in the sense that
35:19 we developed the visualizations.
35:20 We had a wonderful person called Cassandra Shia from MIT Media Lab working with us.
35:24 We went directly to the,
35:25 to,
35:25 to,
35:25 to folks in these communities and worked with them to design the visualizations.
35:28 We start with a village map
35:30 that has,
35:30 you know,
35:30 I,
35:31 I don't know
35:31 um if you.
35:33 Uh,
35:33 uh,
35:33 have ever been to South India,
35:34 but we tend to eat food in a banana leaf,
35:36 so that banana leaf is nutrition.
35:38 These are assets.
35:39 This is marriage and family.
35:41 That's mobility,
35:42 and so on.
35:42 Let me drill down a bit,
35:44 um.
35:45 This is on intra-household bargaining,
35:46 for instance.
35:47 Uh,
35:48 they looked at different questions in intra-household bargaining,
35:50 comparing two villages side by side.
35:52 Firstly,
35:52 there's a color coded mechanism.
35:54 The more green it is,
35:55 the better off you are.
35:56 But also it's very clear if the woman has a big face,
35:58 she has more say.
35:59 If the man has a bigger face,
36:00 he has more say.
36:01 The in-laws have a big face.
36:02 They are mostly,
36:02 yeah,
36:03 and you can very clearly see
36:05 that village
36:06 Aumal it does better than village
36:07 Malakode.
36:08 Yeah,
36:09 so there's a,
36:09 there's,
36:10 you know,
36:11 that's very easy for anybody even without
36:13 a PhD in economics to understand.
36:15 Yes,
36:16 this is an asset.
36:17 So this year,
36:18 you know,
36:18 if the asset doesn't have any box around it,
36:20 that means it hasn't changed the last year.
36:22 If it has a red box,
36:23 you've lost the asset in the right in the last year.
36:25 If it's a green box,
36:26 you've gained the asset in the last year.
36:28 Just compare the two villages,
36:29 you can immediately say which village is doing better.
36:31 Yeah,
36:32 it's sort of interesting thinking about why are we
36:34 visualizing it not as an average,
36:36 but uh as a,
36:37 as a sample.
36:38 Yeah,
36:38 we could basically each one represents a household,
36:40 and that's because visualization theory which I learned from Cassandra,
36:45 uh,
36:45 it,
36:45 it now says that using the central.
36:46 In a theorem,
36:47 if you just do a random sampling,
36:48 you get a much better
36:50 presentation with an idea of the distribution visually than if you just did it,
36:52 you know,
36:53 a,
36:53 a,
36:53 a,
36:53 a,
36:54 a measure of central tendency with a measure of deviance,
36:56 yeah,
36:56 so this is a better way of visualizing it,
36:58 uh,
36:58 particularly for folks who are not,
36:59 uh,
37:00 statistically literate,
37:01 yeah.
37:02 Marriage and family,
37:02 this is one of the most interesting visualizations that they developed with us,
37:06 and the idea really is that they all wanted
37:08 to sort of get away from traditional marriage practices.
37:09 So the height of the flower
37:11 is the age of marriage,
37:12 the number of leaves,
37:13 the number of kids you have.
37:14 If it's a red flower,
37:15 that means you married a relative,
37:16 a close relative,
37:17 a first cousin,
37:18 an uncle,
37:18 and so on.
37:19 If the,
37:19 it's a yellow flower,
37:20 you didn't marry a relative,
37:21 the flower is a bud.
37:23 That means your parents didn't ask your opinion
37:25 before they picked a husband for you.
37:26 If it's open,
37:27 they didn't,
37:28 and you can see immediately which village is doing better.
37:30 Again,
37:30 the color coded greenness tells you something about that.
37:33 So,
37:33 you know.
37:35 Diet and nutrition,
37:36 comparing the two villages,
37:37 if it's a blank banana.
37:39 That means somebody in that household said that they go hungry,
37:42 yeah,
37:42 then you can see an eggplant,
37:43 a papaya,
37:44 a chicken,
37:44 or rice that tells you how much they ate of that particular food,
37:47 uh,
37:47 in that village and in that,
37:49 in that year.
37:50 Now we,
37:51 we then sent this,
37:52 these visualization back
37:54 in the grounds that I talked about in the beginning.
37:55 It was shown there
37:57 and we found
37:58 that it affected village decision making.
38:01 It affected allocations.
38:02 Yeah,
38:03 unfortunately we had to cut this whole exercise short short because
38:06 the woman who was our great support,
38:08 the chief minister of the state.
38:09 Uh,
38:10 passed away
38:11 and,
38:11 uh,
38:12 the situation changed.
38:13 Uh,
38:14 but what is interesting,
38:15 and this is where I want to give you an update,
38:17 is that just last month
38:19 the Tamil Nadu government in its budget speech mentioned participatory tracking,
38:23 that's the budget speech right there,
38:25 allocated 100 crores,
38:26 which is what,
38:27 um.
38:28 A lot of money
38:30 to,
38:30 to be to implement the system across the state.
38:33 So that to me
38:34 is a tremendous
38:36 achievement
38:37 that
38:37 our little pilot has now been
38:39 implemented.
38:40 It's going to be implemented for rolled out for 40 million
38:42 households,
38:43 40 million
38:45 people,
38:45 yeah.
38:46 The state of Karnataka has come to me and asked me to help them do the same thing,
38:49 but bank systems make it rather hard for me to help them because there's no project,
38:51 you see,
38:52 so we have to think about how to work with that,
38:54 uh,
38:54 but,
38:55 but,
38:55 but,
38:55 but,
38:55 but,
38:56 but again.
38:57 As they're going to do great whether I help them or not,
38:59 right?
38:59 And so the question is to do it better.
39:01 So this is for folks in operations.
39:02 How do we solve this problem?
39:03 We're very interested here.
39:05 Interestingly,
39:06 the village strengthening project in Indonesia,
39:08 which is run by a wonderful DTL called Anna O'Donnell,
39:11 is going to be strengthening the village law in Indonesia and 75,000
39:14 villages across the country covering 120 million people in 5 years.
39:18 We are developing a system of participatory tracking that is based on public goods,
39:23 and this is a very crude prototype that we've developed.
39:25 It's basically a village.
39:26 Map
39:27 with village report cards that then get color coded,
39:30 yeah,
39:30 and we've got to go way,
39:31 way more and with this in the very beginning stages of it,
39:34 but it's going to emphasize public goods allocation and report cards
39:36 for public goods again visualized so that citizens can understand it
39:39 and make better local level decisions,
39:41 yeah,
39:42 so
39:43 I just wanted to run through three basic simple ideas.
39:48 Not that simple,
39:48 but
39:49 uh doable ideas is what I mean,
39:51 doable,
39:51 scalable ideas
39:52 to strengthen local democracy
39:54 in the context of World Bank projects,
39:56 that's important.
39:58 Yeah,
39:58 it's work that we can do.
40:00 One
40:01 is by
40:02 strengthening systems of democratic deliberation.
40:04 So we're not just relying on elections,
40:07 but we're supplementing elections with discourse,
40:10 and I've shown you that policy can make a diff to the quality of discourse.
40:14 Policy can make a difference to the,
40:16 to solving
40:16 gender inequality in that discourse,
40:18 and
40:19 we shouldn't be fearful of low literacy.
40:21 We shouldn't be fearful that it's going to be captured by elites.
40:23 There is no evidence we find that that has happened in the largest democratic
40:27 deliberation experiment the world has ever seen in India.
40:29 OK.
40:30 Secondly,
40:32 There is some hint
40:33 of evidence that we've given you
40:35 that we can incentivize politicians to perform better at the local level,
40:39 yeah,
40:40 by giving them electoral incentives that matter to
40:42 them for elections as well as more budget,
40:44 but it's very centrally important there
40:47 that
40:48 that system should be considered credible and legitimate.
40:51 There's a lovely UN report,
40:52 by the way,
40:52 on different incentive systems for local governments
40:54 that I would encourage you to read,
40:56 uh,
40:57 but what it does not talk about is how you want to assess performance.
41:01 If that performance assessment system is not
41:03 considered independent and fair and credible,
41:05 the whole system falls apart.
41:06 But if it is,
41:07 it can do magic.
41:09 Thirdly,
41:10 Democratizing data,
41:12 using digital tools to strengthen democracy.
41:14 There's a wonderful new report that we have.
41:16 The World Bank has produced by
41:18 Tiago Pescito and Steinberg
41:20 just last year
41:20 that works on this a great deal.
41:22 Tiago does wonderful work on democracy sitting in Madagascar,
41:24 I think now,
41:26 so I would encourage you to read that report,
41:28 see the work we've done,
41:29 and think about how you can do this in your own context,
41:31 in your own projects.
41:32 Yeah,
41:32 there's.
41:33 Tremendous potential here that we that remains untapped
41:36 and and democratizing data is important because it
41:39 counters it provides a counter narrative narrative to the
41:42 centralizing tendency of data,
41:44 big data that we're all so we're all getting so excited by big data so
41:46 we can manipulate people into how we think they're going to get better off.
41:49 How do we give them data so that they can get
41:51 better in a way that they think is better off?
41:53 That's the question,
41:54 yeah,
41:55 that's why
41:56 democratizing democracy
41:58 is such a centrally important goal for us.
42:00 Thank you.
42:07 Over to you,
42:08 Vikram.
42:09 If you'd like to stand or you can speak from here if you prefer.
42:13 Um,
42:14 It's an honor to be discussing for
42:17 Biju's talk today.
42:19 Um,
42:19 it's difficult for me to follow such a breathtaking presentation,
42:23 especially since I'm neither an economist nor a social scientist,
42:27 as Art gently pointed out,
42:29 um,
42:29 but I am emboldened to do so by an interesting line in the book's acknowledgment,
42:34 the first part of the presentation on the book,
42:37 uh,
42:37 of oral democracy.
42:39 Uh,
42:39 in it,
42:40 Biju thanks,
42:41 uh,
42:41 Doctor
42:42 Monica Birataburu for teaching him
42:45 how to listen.
42:46 I think that's a lesson that's important for many of us at the bank.
42:51 Yes.
42:52 Um,
42:53 and what follows is my takeaways from listening to this presentation
42:57 and to Bijo's work over the 19 years I've worked here.
43:01 Um,
43:02 the book is the most readable one.
43:04 I'm a slow reader and I reach for a hand sanitizer when I see a regression.
43:08 So
43:09 I yet managed to complete it,
43:11 uh,
43:11 in one treadmill session,
43:12 which is a good measure of its accessibility
43:15 for a non-economist like me.
43:17 Uh,
43:17 it makes several normative and technical claims,
43:20 many of which you've seen,
43:22 um,
43:22 and Bijo and his co-author Doctor Sanal offer a robust methodological defense
43:28 for these claims
43:29 while anchoring them in an extensive
43:32 multidisciplinary survey of the literature.
43:35 But what I found
43:36 especially unusual is their extensive use of these grams about transcripts,
43:41 uh,
43:42 which you've heard of,
43:42 uh,
43:43 uh,
43:43 uh,
43:44 a bit.
43:45 Because they give us a
43:46 rich,
43:47 raw sense of the drama and pathos involved
43:51 in these local colloquies,
43:53 right,
43:53 which are often dismissed as simplistic.
43:56 Um,
43:57 emotion isn't a quality that comes easily to economists,
44:00 even if they are behavioralists,
44:02 uh,
44:03 and or bank staff.
44:05 Um,
44:05 but then a polymath like Bijo is no ordinary economist,
44:08 and
44:09 the unvarnished transcripts which he and his co-author
44:13 present,
44:14 entertain,
44:15 amuse,
44:16 engage,
44:16 and enrage us.
44:18 Uh,
44:18 in many ways,
44:18 and I would commend them to you because
44:20 they are arranged like a carefully scripted morality play
44:23 who's,
44:24 uh,
44:25 which these,
44:26 the authors have been rehearsing for several years and they have been,
44:28 I think,
44:28 for 10 years.
44:29 This book has been in the making
44:31 and in fact Biju was a bit restrained today,
44:34 but,
44:34 uh,
44:34 you can see him with full gusto reading extensively from
44:37 these transcripts at his presentation at the Bangalore International Center.
44:41 Just Google
44:42 oral democracy in Bangalore and the link should pop up.
44:45 Now
44:46 Amartya Sen
44:47 teaches us that we inhabit multiple identities,
44:51 and this book
44:52 appeals to two of my own
44:54 my professional identity as an international lawyer for the World Bank
44:59 and my private identity as a lifelong student
45:02 of Indian constitutional
45:04 law and democracy and as a citizen.
45:07 Let me begin with the latter.
45:10 One of
45:10 Biju's mentors,
45:11 the historian Randall Austin,
45:13 this is his book,
45:14 um,
45:15 claimed that the Indian Constitution
45:18 was put to hourly use.
45:19 It's a long constitution.
45:20 Here's the pocket version.
45:22 Um,
45:23 Doctor,
45:24 uh,
45:24 Sonal and Bijo
45:26 established the salience and relevance of Austin's claim made many years ago
45:30 by showing us how the Constitution works at the community level.
45:34 Now many of you have seen the images or read
45:37 about the recent protests in India over the citizenship law.
45:40 These protests have featured mass readings of the Constitution's
45:44 rhapsodically worded preamble.
45:46 It heralds a republic
45:48 committed to social,
45:50 economic,
45:50 and political justice
45:52 and the equality of status and opportunity for all citizens,
45:55 and this book is a validation
45:57 of how that soaring rhetoric
45:59 can trickle down to local communities and empower ordinary citizens.
46:04 It shows us,
46:05 as the authors put it,
46:06 how democracy can be deepened
46:08 by giving citizens the chance to speak and to be heard
46:11 without getting into the
46:12 actual merits of this debate,
46:14 whatever you may think of it.
46:16 Now coming to the implications for this book for the bank and
46:19 my professional identity
46:21 as a
46:22 lawyer for the bank,
46:24 this year we celebrate the seventy-fifth
46:26 anniversary of the Articles of Agreement.
46:28 We don't have a pocket version unfortunately
46:31 of it
46:32 because it was not really thought of as,
46:34 you know,
46:34 a sacred charter.
46:36 Uh,
46:36 but it is at the core of the bank's DNA and it establishes,
46:40 establishes us as a multilateral institution that is at the core
46:44 of our very nature of an international financial institution,
46:49 and it was designed at the Bretton Woods conference which took place 76 years ago
46:53 to be an institution of universal membership.
46:56 To secure
46:57 this institution's institutional and financial viability and acceptance,
47:02 remember we had to go to the capital markets of the world
47:05 and uh sell our bonds.
47:07 The bank's founders at the Bretton
47:08 Woods conference insisted on depoliticization,
47:12 uh,
47:12 for the institution's successful functioning
47:16 and so into the articles they wrote three
47:19 strident prohibitions.
47:20 They first barred the bank
47:21 from meddling
47:22 in its members' political affairs.
47:25 They forbade bank staff from being influenced by the members' political character
47:30 or ideology.
47:32 And thirdly they insisted that the
47:34 bank must ensure that only economic considerations
47:37 are relevant to its decisions.
47:39 I get head rolls when I say this,
47:40 but that's what's in the articles.
47:42 Um,
47:43 now these restrictions have been strictly construed over the past 7 decades.
47:47 In fact,
47:47 general counsel Ibrahim Shehata,
47:48 after whom this room was recently.
47:51 Dedicated argued
47:52 that the article's bar on political
47:54 interference requires the bank to distance itself
47:58 from partisan politics
47:59 and the active participation in a member's political life.
48:03 That that was his ruling,
48:04 and he is
48:05 treated by the board as a demigod for his views on the article.
48:10 Now he also believed that the bank's political
48:13 neutrality required it to refrain from endorsing or mandating
48:17 a particular government form
48:20 or
48:21 or character in terms of template,
48:23 right,
48:23 even if the history of the bank,
48:24 especially deck
48:25 points entirely to the contrary,
48:27 uh in the uh especially through the Washington Consensus.
48:32 But Mr.
48:32 Shihada
48:33 was also careful to point out that the articles
48:36 do not bar the bank from either mentioning or considering
48:40 politics or the P word just as there was nothing to
48:42 prevent it from talking about corruption or the C word,
48:47 uh,
48:47 and so as we discussed at a board seminar in this very room 5 days ago,
48:51 uh,
48:51 the entire board met,
48:52 uh,
48:52 informally to discuss
48:54 the relevance of politics and the bank,
48:56 um.
48:58 We sort of uh
48:59 one of the takeaways was that the bank may legitimately conclude or consider
49:04 politics and demo democracy questions
49:07 uh in its operations and research provided there is a clear and direct link
49:12 to economics and development considerations
49:15 we can debate what that means later but what does this
49:18 mean for Bijo's thesis and its implications on the bank's operational work
49:23 first I don't think this book is likely to.
49:26 Nudge or move the bank in any
49:28 significant way from its long standing official position
49:32 of not getting involved in the writing of constitutions,
49:35 uh,
49:35 or the holding of national elections.
49:37 I know this is a stress point,
49:39 especially in our work in fidelity,
49:41 uh,
49:41 whether through aid or policy advice,
49:44 but
49:44 by chronicling how all of democracy affects
49:48 citizens' economic agency and development choices,
49:51 this book could help advance the bank's involvement.
49:54 In greater constitutional and legal devolution
49:57 beyond its judicial emphasis on fiscal federalism
50:00 and decentralized service delivery
50:02 and a variety of instruments for this purpose including policy based lending
50:06 whose 40th anniversary we observed this month
50:09 with a March 1980s sell to Turkey.
50:12 Um,
50:13 second,
50:13 as Biju suggests,
50:15 the Indian example shows how governments can themselves
50:18 promote participatory systems of governance
50:21 through legal and constitutional policy mandates
50:25 without being forced or nudged
50:27 by donors to do so.
50:28 There was nothing in the evidence to suggest that the bank
50:31 or anybody else,
50:32 uh,
50:33 benefited.
50:33 He suggested to the government of India
50:35 uh that uh they create an empower ground contracts.
50:38 In fact,
50:38 the history is to the contrary.
50:40 The founders of the uh of the republic at the constitutional debates rejected,
50:45 uh,
50:46 the notion of uh of uh Punjab Republic which wanted as
50:50 as we do,
50:51 uh,
50:52 uh,
50:52 beautifully raised to us Baker.
50:54 Uh,
50:55 Dali,
50:55 uh,
50:56 sort of founder of the Constitution was skeptical,
50:58 uh,
50:58 of the idea that you could have empowerment through these,
51:01 uh,
51:02 um,
51:02 uh,
51:02 through villages,
51:03 but obviously that changed and that change was really very much an internal,
51:07 uh,
51:07 change.
51:07 It was not prompted by any,
51:09 uh,
51:09 external agency,
51:11 uh,
51:11 and so it is very important,
51:13 uh,
51:14 for us that homegrown institutions like Rams Sabas
51:17 are used.
51:19 Credible laboratories which you've done uh to pilot and
51:22 implement development solutions and innovations in matters such as
51:25 beneficiary selection which the bank sort of tears its
51:27 head about how do we ensure it's correct?
51:29 how do we ensure it's robust and they've already been doing it as you tell us
51:33 uh targeted distribution and the new frontier of data collection
51:36 and credibility which is entirely new frontier for the bank
51:39 where we are still sort of you know uh
51:41 uh muddling our way through.
51:43 Um,
51:44 I certainly saw this firsthand in the West Bengal and
51:47 Bihar Panchat projects for which I was the project lawyer.
51:50 In both these projects.
51:51 Our focus was on the infrastructure
51:53 and finance of the Panchats,
51:55 uh,
51:55 but Bhu's book points us or shows us the
51:57 way to have more substantive engagements with these institutions.
52:00 In fact,
52:01 yesterday,
52:01 uh,
52:02 Bihar announced a new focus on Panchat-led growth and so.
52:05 This is very relevant to our work especially at
52:08 the India program but also more broadly this work
52:11 is an important validation of the bank's growing approach
52:15 to focus on strengthening and restoring the civic bonds between the state and the
52:20 citizen as we see in the manner strategy which Shanta gave us and is now
52:23 his successor sort of revalidated.
52:26 But it also has particularly useful lessons for
52:29 the bank in fragile conflict and violent situations,
52:32 which is my day job where national governments are distrusted,
52:36 ineffective,
52:36 or simply don't exist
52:38 and the bank searches for effective local institutions and counterparts
52:42 to do business with.
52:43 This is highlighted by the FCB strategy
52:46 that the board endorsed,
52:47 uh,
52:48 last week.
52:49 Finally,
52:50 the grounds of our experience
52:52 is very instructive.
52:53 To the bank's implementation of its new ESF,
52:57 environmental and social framework,
52:59 this framework
53:00 seeks to advance transparency,
53:02 non-discrimination,
53:04 public participation and accountability in the development process.
53:08 It sets out detailed requirements for
53:10 how the bank must identify relevant stakeholders
53:14 and
53:14 work with existing institutions as far as possible so your work is
53:17 tremendously helpful in pointing the way we're still in new ground there
53:22 there are lots of guidance notes but all of them exist in
53:25 the theoretical abstract sense and this book is a very important uh
53:29 uh validation of how,
53:30 uh,
53:31 at least ESSN which is the bank standard on participation.
53:35 Stakeholder consultation
53:36 can be uh implemented in a very practical way using Rams Sabas as a possible
53:41 interlocutor.
53:42 In fact,
53:43 the principal allegation in one of the ongoing cases before
53:46 the bank's inspection panel uh is that a bank's finance
53:49 project I actually was involved in it as a project
53:52 lawyer but not on this aspect which is before the panel
53:54 uh
53:55 it the the allegation is that,
53:57 uh,
53:58 in Jharathan state this is the low income states project for rural water supply.
54:02 Uh,
54:03 the,
54:03 um,
54:03 the project team failed to obtain the grounds of approval
54:07 for a rural water supply scheme under its jurisdiction.
54:11 The case is still pending,
54:12 and I don't want obviously to comment on it,
54:14 but one issue was whether the consultation at the local
54:17 at the level of the revenue village ground of us,
54:20 so which is at sort of the
54:21 uh at the.
54:22 Primary level was sufficient because
54:25 uh there were tribal hamlets uh within this revenue unit
54:29 that had their own grounds of us that may not have been consulted
54:32 so it also shows us we need to maybe drill down even
54:35 further and I think you're quite aware and sensitive to that issue
54:39 so those are my takeaways and thank you again for inviting me.
- add-style
- lp-body-content