00:00 Good morning everybody.
00:01 Really happy to be here with Kartik which uh
00:04 is not,
00:05 is not only
00:06 the author of one of the many,
00:09 uh,
00:09 11 of the most relevant books on,
00:11 on these issues uh recently,
00:13 but he has been uh
00:15 uh uh
00:17 walking with us the path of trying to rethink,
00:20 uh
00:21 Uh,
00:21 governance and the governance indicators
00:24 over the last few years and we hope that he will
00:27 still
00:27 be able to,
00:28 to collaborate with us on this.
00:30 Um
00:31 This is the main auditorium of the World Bank,
00:33 so,
00:34 uh,
00:35 similar gatherings have happened over,
00:37 over the last decades.
00:39 So it's not very difficult to mind people thinking about
00:42 issues like this 130 years ago.
00:45 Now,
00:46 the conversation may have been a little bit different.
00:49 It may have been about
00:51 Uh,
00:52 trade liberalization.
00:53 It may have been about
00:55 financial and labor market
00:56 deregulation.
00:58 It may have been
01:00 about
01:01 the need to open the economy.
01:03 And as a matter of fact,
01:05 in uh,
01:05 in many of these issues,
01:07 the common theme is
01:08 the state
01:10 given the step back
01:11 from the role
01:12 on,
01:12 on the economy and maybe it was needed,
01:14 right?
01:15 The only aspect in which the state was playing a,
01:18 a role
01:19 when when we were thinking about
01:21 growth and development and at that time,
01:24 the lines between growth and development
01:26 were
01:26 very thin.
01:27 People,
01:28 when we're,
01:28 people were talking about development,
01:30 they were actually talking about
01:32 growth.
01:33 Uh,
01:34 the,
01:35 the only role for the state was,
01:37 uh,
01:38 when the discussion about how to build human capital was brought,
01:41 and it's
01:42 the state has to have a role
01:43 in education
01:45 and health.
01:47 But even if we take it
01:49 uh at fair value
01:51 that the state was probably
01:53 not working,
01:54 what never occurred to,
01:56 to,
01:56 uh,
01:56 at the time was
01:57 that it could be improved
01:59 upon
02:00 the way in which the state and the government could work.
02:03 So since this was not placed,
02:05 this question was not placed there,
02:07 the obvious solution was to ask the state,
02:09 the state to move out of every,
02:11 of every single.
02:12 Uh,
02:13 of every single aspect of economic life,
02:15 including,
02:16 including privatizations,
02:17 right?
02:17 And that's what we were known to advocate for,
02:20 for many years and,
02:21 and also
02:22 we have abandoned them for a long time,
02:25 uh.
02:26 When you talk with people in the countries,
02:28 we are still
02:30 recognized for,
02:31 for that.
02:32 Now there's a,
02:33 a completely different conversation
02:35 uh today.
02:37 Now it's understood
02:38 that
02:39 not only service delivery but regulation
02:42 and many other functions are critical,
02:44 uh,
02:45 uh,
02:46 in a state so that the state,
02:47 so that the economy and a society could be,
02:49 uh,
02:50 work,
02:50 uh,
02:51 properly.
02:52 But also the ways and approaches to try to understand this have
02:56 changed.
02:57 Um,
02:58 uh,
02:58 we heard,
02:59 heard yesterday Rima,
03:01 uh,
03:02 basically,
03:03 uh,
03:03 describing her
03:04 richer search strategy from trying to get data from the governments
03:09 to try to work with the,
03:10 with the governments,
03:11 uh,
03:12 and,
03:12 uh,
03:13 Antarctica has taken it to
03:14 to an extreme now which,
03:15 which he directly works in the state capacity building
03:19 with the government,
03:20 uh,
03:21 of India.
03:22 The second distinction
03:23 is the
03:24 availability and the world of microdata,
03:27 which was not present
03:28 at the time,
03:30 and that created two problems at the time.
03:32 The first one was that
03:34 most of the empirical studies were
03:36 about
03:37 cross-country
03:38 regressions
03:40 and
03:41 because there was very little information about
03:43 a particular country,
03:45 then it was assumed
03:47 that the lessons
03:49 were equally applicable across country.
03:52 So,
03:53 so,
03:53 so the term best practice
03:55 uh was what we use
03:58 uh day in and day and uh
03:59 in respect
04:00 of the nature,
04:02 the of the country,
04:03 the culture,
04:04 the historical perspective,
04:05 and that's also something that Kartik,
04:07 uh,
04:07 the,
04:08 that,
04:08 uh,
04:08 Kartik is bringing to the uh
04:11 uh uh uh to the conversation.
04:13 The final point that I want,
04:15 I want to make is that,
04:16 uh,
04:17 Uh,
04:18 and it's a word of caution.
04:19 Uh,
04:19 India is now sur surpassed China
04:22 as the largest,
04:23 uh,
04:23 economy in the,
04:24 in the world,
04:25 and,
04:25 and I guess Kartik works on the assumption
04:28 that the size of the book is proportional to the size of the population
04:32 of,
04:32 of,
04:33 of a country.
04:34 So,
04:34 so it requires some commitment,
04:36 right?
04:37 Uh,
04:38 uh,
04:38 work with that.
04:39 There's a lot of things that you will learn from it.
04:41 Kartik,
04:43 thank you.
04:44 I
04:49 So the time can we reset that to 45,
04:51 I think that's what it was supposed to be,
04:53 because it's good,
04:53 um,
04:54 so.
04:56 Um,
04:56 thank you so much.
04:57 Really,
04:57 uh,
04:57 it's an incredible pleasure and honor to be here in the Preston Auditorium.
05:00 Um,
05:01 so there's a lot,
05:02 lot I'm gonna cover,
05:03 OK?
05:03 So,
05:04 uh,
05:04 so,
05:05 so be ready for the ride.
05:06 You can't really sleep,
05:07 OK?
05:07 Um,
05:08 so the book,
05:09 so this is based,
05:10 uh,
05:10 this,
05:10 this,
05:11 the talk is going to be based on a book written about India,
05:13 right,
05:13 but
05:14 the concepts and principles are going to be relevant globally,
05:16 and there's a global edition of the book that'll come out later this year,
05:19 and the,
05:20 the core point in some ways is to say
05:23 that most of our development.
05:24 Policy discourse focuses on what should governments do,
05:27 right?
05:28 I mean,
05:28 but we spend remarkably little time thinking about how will
05:31 they do it or can they even do it,
05:33 OK?
05:33 And so,
05:34 um,
05:34 a key goal here is to bring the topic of state capacity to the core of
05:38 the development discourse and then expand on that in ways I'm going to show you,
05:41 OK.
05:42 So
05:43 kind of when we just look at India itself,
05:45 right,
05:45 um,
05:45 and as kind of a big picture,
05:47 so the good news in India is India's doing reasonably well.
05:49 I mean,
05:50 on many measures,
05:51 right?
05:51 So it is one of the fastest growing large economies,
05:53 macro stability,
05:54 size,
05:55 demographics.
05:55 There's a lot going for India,
05:56 OK,
05:57 but it also faces profound challenges,
05:59 especially in basic service delivery,
06:01 in education,
06:02 in health,
06:03 in safety,
06:03 in justice with a backlog of 30 million court cases,
06:06 welfare,
06:07 and jobs,
06:07 which is kind of a global challenge,
06:09 right?
06:09 So.
06:10 Basically getting India to full potential,
06:12 um,
06:13 so it's doing reasonably well,
06:14 6,
06:14 6.5%,
06:15 though it's slowing down a bit,
06:16 but getting to full potential is critically constrained by kind of
06:20 weak state effectiveness in the delivery of essential public services.
06:23 Now,
06:23 the Indian state is also bipolar,
06:25 just like India itself,
06:26 right?
06:27 There's some things it does incredibly well,
06:28 so it does really well in what's called mission mode
06:32 when you need to execute at scale.
06:34 And get things like elections done or
06:35 vaccination campaigns done and stuff like that,
06:38 but when it comes to the day to day delivery of services,
06:40 that's kind of where
06:42 the challenges are.
06:42 OK.
06:43 Now what you have seen in the past decade is
06:45 significant improvements in state capacity in India using technology,
06:49 using kind of the direct benefit transfers as a way
06:51 of sharply reducing corruption and leakage in welfare programs.
06:54 So that's been a part where there has been meaningful improvement.
06:57 But there's a long way to go in human
07:00 intermediated frontline services when you've got the teacher,
07:02 the doctor,
07:03 the police officer,
07:04 the judge.
07:04 These are places where the technology can augment
07:07 productivity but cannot replace the frontline worker.
07:10 OK,
07:10 so,
07:11 so you hear the words of Rakesh Mohan,
07:12 who's one of the architects of the 90.
07:14 One economic reforms
07:15 and a former
07:16 chief economic advisor um and so what he says is
07:19 the way forward for accelerated growth in India and I
07:21 suspect this is true for many parts of the world
07:23 is being held back by major governance deficits in all areas
07:27 to do with public good delivery of public goods and services,
07:30 right?
07:30 Um,
07:31 and Lant Pritchett,
07:31 who's well known in this building,
07:33 like he's always more dramatic,
07:34 and Lant in 2008,
07:35 like,
07:35 you know,
07:36 famously said,
07:37 he said the weak performance of the Indian state
07:38 is one of the 10 greatest challenges facing humanity,
07:41 OK,
07:41 and part of this is motivated by the sheer scale of the country,
07:44 but also the depth of the problem,
07:46 and importantly,
07:47 the seeming intractability,
07:48 right?
07:48 So land almost put it on par with climate change because of
07:51 the importance of the intractability because
07:53 like everything is connected to everything,
07:55 OK,
07:55 so how do you make progress?
07:57 So my own assessment,
07:59 and part of why I wrote the book is that as we assess India at 75,
08:01 there is perhaps no greater barrier to achieving full potential
08:05 than improving the effectiveness of the Indian state.
08:07 And the other intellectual distinction of the book is that
08:09 almost all of our discourse focuses at the national level,
08:12 but here 90% of these issues are controlled by state governments.
08:15 And so this is a book that directly focuses on states,
08:18 and in many ways that's more relevant to the rest of the world
08:20 because Indian states are about the size of a typical medium sized country.
08:23 OK,
08:24 so,
08:25 so in terms of why I wrote this book,
08:27 OK,
08:27 is that in general,
08:29 and I'll come back like Artu was saying,
08:30 so at the very end I'm going to spend two slides situating the,
08:33 I'm going to dive into what I'm doing,
08:35 but I'll come back and
08:36 Situate this in the global discourse and development at the very end,
08:38 OK,
08:39 so,
08:40 um,
08:40 so what I'm doing here is,
08:41 so the default approach to improving public services
08:44 in kind of most global and national development conversations
08:47 is to say we need more money.
08:48 OK,
08:49 so
08:49 the education folks want 6%,
08:51 everybody,
08:51 health folks want more money,
08:53 the defense guys want more money.
08:54 Everybody wants more money.
08:54 OK,
08:55 so there's two problems with this challenge.
08:56 The first is we don't have the money.
08:58 OK,
08:58 so
09:00 a lot of these norms of 6% of GDP come from OECD norms of tax.
09:04 GDP ratios of 35%.
09:06 India is sitting at a tax to GDP ratio of 17.5%. OK,
09:09 so,
09:10 we just don't have the money.
09:11 B,
09:11 more importantly,
09:12 is that a key finding from empirical research
09:15 in the social sector over the past two decades
09:17 is that the translation of business as usual spending into outcomes is really weak.
09:22 OK,
09:23 so
09:24 the most important intellectual reason for writing this book is that
09:27 the common thread across 20 years of my work in education,
09:30 in health,
09:31 in welfare,
09:32 in the in child development.
09:34 Is that the return on investment of better governance and state capacity is often
09:39 10 times greater than simply spending more money in a government as usual way.
09:43 OK,
09:44 so the default focus of kind of sectoral development goals is
09:47 to say we need more money and we need more programs,
09:50 OK?
09:50 But you,
09:52 we spend remarkably little time then thinking about how does that
09:55 top line of expenditure translate into the bottom line of impact,
09:57 OK.
09:58 And what this 20 years of micro research has allowed us to do
10:01 is it's allowed us to much more sharply with a better microdata and randomized
10:06 controlled trials kind of have much sharper evidence on the impact of individual
10:10 interventions tracking all the way down to the ultimate outcome we care about.
10:13 OK,
10:14 so here's literally kind of,
10:15 you know,
10:15 my life's work in half a slide,
10:17 right?
10:17 But
10:18 this is,
10:19 but you know,
10:19 since this is an academic conference connecting the micro to macro,
10:22 I want to spend a little bit of time on each of these studies,
10:24 right?
10:24 So education.
10:25 Where does most of the money go?
10:27 OK,
10:27 so 85% of our budget in India goes to salaries,
10:30 right,
10:30 teacher salaries,
10:31 and most of the money goes to unconditional increases in pay.
10:34 And
10:35 we've got really rigorous studies in the QJE,
10:37 like literally looking at in Indonesia passed a
10:40 constitutional amendment to spend 20% on education.
10:43 That money mainly went to teacher salaries,
10:45 and we found zero impact of that.
10:46 OK,
10:47 so that paper was called Double for Nothing,
10:48 right?
10:49 And
10:50 but that cost $5 billion a year.
10:52 It had absolutely no impact on learning.
10:54 OK.
10:54 In contrast,
10:55 you can have even modest amounts of performance linked
10:58 pay that end up having a substantial impact,
11:01 right,
11:01 but we don't spend on that.
11:03 Coming to welfare,
11:05 a key goal,
11:06 you know,
11:06 so we have this AR paper on biometric smart
11:09 cards where we showed that rolling that out,
11:11 the reduction in leakage and corruption was over
11:13 10 times the cost of the intervention itself.
11:16 And then this recent econometrica paper in some ways is even more striking.
11:19 Because
11:20 a key goal of ERA,
11:21 which is this public employment program,
11:23 was to improve the market wage of the poor by
11:26 having the government be an employer of last resort,
11:28 right?
11:29 You can't implement a minimum wage if less than 10% of your employment is formal,
11:33 right?
11:33 It's all informal.
11:34 So how are you going to implement a minimum wage?
11:35 And the way you do that is by having
11:37 the government as an outside employer of last resort,
11:39 OK.
11:40 Now what we found here,
11:41 which is just mind boggling,
11:43 is that the impact of improving implementation on the market wage
11:46 was bigger than the impact of rolling out the program itself.
11:49 OK,
11:50 so
11:50 all of the attention again goes to creating the
11:53 program and getting the appropriation in the budget,
11:55 but then remarkably little attention goes to improving implementation,
11:59 and that was a bigger effect than rolling out the program itself.
12:02 OK,
12:04 again,
12:04 so education,
12:05 so this is a nice example of the micro evidence,
12:07 right,
12:08 how Micro evidence comes back to macro points.
12:10 So in a way,
12:10 the point here is exactly what you saw in Burnside and
12:13 Dollar AR 2000 of the complementarities between resources and governance,
12:16 right?
12:17 So
12:17 the aid was found to not be effective on average,
12:19 but in settings with good governance,
12:21 but that was not very robust.
12:22 It was cross country,
12:23 you know,
12:24 the Rudman et al critique just showed it wasn't very robust.
12:26 But what we're showing here is with micro evidence
12:28 randomized that simply providing extra resources has no impact.
12:32 But then when you combine that with a little bit of improved governance,
12:35 you get a massive.
12:36 with complementarity.
12:37 OK,
12:38 Education,
12:39 we have this AER paper on how
12:41 transformative personalized education technology can be,
12:44 but almost all the budget goes to procuring computers,
12:47 and we have study after study showing that the hardware by itself has no impact,
12:51 OK,
12:52 but we still haven't managed to see
12:53 a single government procurement that's managed to procure
12:56 this personalized software because they don't know
12:58 how to procure things based on outcomes.
13:00 OK,
13:00 so
13:01 they know how to write a contract and do L1 for a standardized.
13:04 Computer and that's what they kind of spend all their money on,
13:06 OK,
13:07 um,
13:07 but that has no impact and the returns are all in that personalized software.
13:11 Um,
13:11 recent paper in the JPE on child development where,
13:14 uh,
13:14 um,
13:15 what we see is India's ICDS is the world's largest early childhood
13:18 program catering to over 80 million kids aged 0 to 6,
13:22 and it is staffed by one frontline
13:23 worker who's responsible for everything from nutrition to
13:27 to home visits to mother's counseling
13:29 and
13:29 feeding and preschool education.
13:32 And what we find is that putting an extra half-time worker
13:35 dramatically improved learning outcomes and reduced malnutrition and stunting,
13:39 OK,
13:40 but
13:40 where does all the money go?
13:41 The money goes to unconditional pay increases of existing workers,
13:45 and that has zero impact on outcomes.
13:47 OK,
13:47 so that again reflects the politics of who has the bargaining power
13:50 and why the evidence is so important.
13:52 Uh,
13:53 farmer welfare,
13:53 we had this paper,
13:54 um,
13:55 where
13:55 the government,
13:56 as we know,
13:56 has been expanding cash.
13:57 Transfers to a whole set of people
14:00 and all we did was set an outbound call
14:02 center that was making 25,000 calls to farmers in two
14:05 weeks and reporting on whether they received their money
14:07 and we randomized at the level of the subdistrict official
14:11 and they were told that your performance is
14:12 being measured by talking to the last mile worker
14:15 and you saw 100x return relative to the spending about the money delivered on time.
14:20 OK,
14:20 so it's incredibly low cost and you can,
14:22 so the key.
14:23 is to combine the data,
14:24 the technology to get you performance measurement and management,
14:27 right?
14:27 So you've got all these pieces come together
14:29 and then
14:30 a recent paper on the courts by Manaswini Rao,
14:32 which I really like,
14:33 just shows that every extra district court judge
14:35 in India that clears another 200 cases,
14:38 and because 60% of these cases are land and credit,
14:41 that you're unlocking factors of production that leads to
14:44 a total increase in credit offtake and economic activity,
14:46 and again,
14:47 the total increase is.
14:48 30 times the cost of that judge.
14:50 OK,
14:50 so the key to all of this,
14:52 the value of the microevidence,
14:53 is not just saying it had an effect,
14:55 but quantifying and showing you how big the ROI is on kind
14:59 of getting these basic governance investments and state capacity things done.
15:02 OK,
15:03 so that's why I wrote the book,
15:04 that the reason you write the book like Reema was talking about yesterday is that
15:07 what goes into these individual papers,
15:09 there's so much you learn,
15:10 like I mean that you don't connect the dots,
15:12 and when you have enough to connect.
15:14 Dots is when you kind of take on a project like a book like this,
15:16 OK,
15:18 so what I'm going to do is now the core of the book
15:20 really is the Section 2 and 3 of building an effective state,
15:23 but before I do that,
15:24 I want to take a 5 minute historical detour,
15:26 OK,
15:27 and situating India in kind of
15:29 global context.
15:30 And before we get into the technocratic core of the book,
15:33 you have to think about the incentives and constraints of
15:35 the two key actors who make up the state,
15:37 right,
15:37 who are the politicians and the bureaucrats,
15:39 OK.
15:39 And when we think about state capacity,
15:41 it's a bit subtle,
15:42 right,
15:43 because I just told you we do some things really well and other things not so well.
15:46 And so how do you make sense of that?
15:48 And for that,
15:48 the Fukuyama framework of thinking about the scope and
15:51 the strength of the state is kind of,
15:53 I find the most useful.
15:54 So the point I want to make
15:56 is it's useful to kind of look at the modern state itself and
15:59 how the scope of the modern nation state has expanded over time,
16:02 OK.
16:03 And so for most of human history,
16:04 the nation state was essentially a security state.
16:07 So the only thing you did was defense and law and order,
16:10 and about 80% of your budgets went to those things.
16:12 It's not like you didn't build roads,
16:13 but the roads were built primarily for troop logistics,
16:16 and any economic benefits were ancillary.
16:18 OK.
16:18 It's only in the 18th and particularly 19th century
16:22 that you get the second phase of the state,
16:23 which I call the industrial state,
16:25 where governments now start having large scale investments in infrastructure,
16:29 roads,
16:29 railroads,
16:30 water,
16:31 public goods,
16:31 sanitation.
16:32 And it's not an accident that this happens at the time of the industrial revolution,
16:36 right,
16:36 because till then
16:38 production is decentralized in small scale,
16:40 right?
16:40 So it's only when you have scale production in a few
16:43 places that you then need the transport and the storage,
16:45 and there's complementarity then between the infrastructure
16:48 and the nature of production,
16:49 OK?
16:49 And
16:50 this is also the first time you get a significant increase in GDP per capita,
16:55 OK.
16:55 So,
16:56 but the problem with that is along with that growth comes massive inequality,
16:59 right?
16:59 So it's only in the last 100 years and
17:02 particularly in the last 75 years after World War II
17:05 that you get the third phase of the state,
17:07 right,
17:07 which is the modern welfare state
17:09 where you have progressive taxation and a lot
17:11 more public provision of key kind of benefits,
17:14 right,
17:14 food security,
17:15 healthcare,
17:15 pensions.
17:16 All of this is remarkably recent.
17:17 OK,
17:18 so the two key points I want you to keep in mind.
17:21 One is that welfare states are very expensive,
17:22 OK,
17:23 and so
17:23 if you just look at welfare spending as a fraction of GDP
17:27 till about 1930,
17:28 it's non-existent.
17:29 It picks up a little bit in the Depression,
17:31 and it really is a post-World War II phenomenon,
17:33 OK?
17:34 And the reason for this
17:35 is welfare states are expensive,
17:37 and having to do that,
17:38 you need the fiscal capacity and to build the fiscal capacity you need
17:42 the administrative ability to tax and the political willingness to be taxed,
17:46 OK?
17:46 And so,
17:47 and people are only willing to be taxed in existential crises like a war,
17:50 OK,
17:51 so it's the war,
17:52 it's the war effort that builds both the fiscal capacity
17:55 and the state capacity,
17:56 because remember the state is commandeering the best and brightest
17:59 in the economy to basically come run the war effort,
18:01 OK.
18:02 And so the peace dividend is then what gives you the modern welfare state.
18:06 OK,
18:06 so,
18:07 and so this is a remarkably recent phenomenon.
18:09 OK.
18:10 Now the second key point I want to highlight
18:12 is that the nature of the state,
18:13 the scope of the state,
18:14 is highly correlated with the extent of democracy.
18:17 OK,
18:17 so
18:18 the security state happens at a time when there's really few
18:21 constraints on the sovereign and there's no concept of individual rights,
18:24 OK.
18:25 Now what's a bit subtle about the industrial state
18:27 is the industrial state happens in the 19th century
18:30 at a time of limited democracy,
18:32 right?
18:33 So the US and the UK like to go
18:34 about lecturing the world about the values of democracy,
18:37 but in the 19th century when they were growing,
18:39 these were highly flawed democracies,
18:41 right,
18:41 with vote restricted to.
18:42 20% of the population.
18:43 OK,
18:44 so it's not an accident that that is the demographic that votes for the capex,
18:48 right,
18:48 because
18:49 in the long term the roads benefit the
18:51 poor through market integration and lower prices,
18:53 but in the short term it's capitalized in the property values,
18:56 and so it's the property owning class that typically wants the capex,
18:59 OK.
19:00 And so it's not an accident
19:02 that you only get the modern welfare state
19:04 after universal franchise.
19:06 OK,
19:06 so in any income distribution,
19:08 the left of the distribution wants higher taxes and more redistribution,
19:12 and the right wants lower taxes.
19:13 I wonder if that's where the term right and left come from,
19:15 right?
19:16 I think it came from the parliament.
19:17 Uh,
19:18 but you know,
19:18 but that is the existential conflict in every Western democracy today,
19:22 right?
19:22 That can mean,
19:23 and so there's a reason that the one bill the
19:25 Biden administration passed that passed bipartisan was a transportation bill,
19:28 right,
19:29 because that was,
19:29 it kind of sat with the old style of what,
19:32 you know,
19:33 people were willing to spend on.
19:34 OK,
19:34 so,
19:34 but the reason I'm highlighting all of this is then to kind of say,
19:38 What makes India so unique in human history,
19:40 right?
19:41 Why did I give you this background?
19:42 It's because before we to make sense of India,
19:45 you have to recognize that this is this insane
19:47 experiment in human history with no prior precedent of having
19:51 universal adult democracy on day one of independence,
19:54 OK?
19:55 And so this.
19:55 Is this really powerful picture from Arvind Subramania
19:58 where it just shows
20:00 that in the democracy index at comparable levels of income,
20:02 the US,
20:03 the UK,
20:04 South Korea are all kind of below 2.5,
20:06 which is roughly the fraction of the population with the vote,
20:08 and India is up there and stays up there.
20:10 OK,
20:11 now why is this so important?
20:13 Because at some level,
20:14 I mean,
20:15 so this is,
20:16 there are huge benefits,
20:17 OK,
20:17 there's huge benefits,
20:18 and the biggest benefit is it's an enormous moral triumph,
20:22 OK,
20:22 that's given
20:23 every Indian,
20:24 including the most marginalized,
20:25 the ability to participate in the political process
20:28 well before their counterparts in high income countries,
20:31 OK.
20:32 And it's also led to substantial policy wins for the poor.
20:35 So,
20:36 you know,
20:36 this point about the franchise being associated with
20:39 welfare is well documented in correlational political science studies
20:42 and in economics,
20:43 the best study is like the Tom Fujiwara Econometrica paper,
20:46 right?
20:46 Like I mean that when you have
20:47 this natural experiment that increases voting rights,
20:50 you end up with a lot more welfare spending.
20:52 OK,
20:52 so,
20:53 so in the,
20:53 in the Indian case,
20:55 my favorite example is think about the PDS for food security,
20:58 right?
20:58 So
20:59 the US does food stamps for food security.
21:01 For the poor in the 1930s at a GDP per capita of $18,000 normalized to 2011,
21:07 India does that in the 1960s at 1 15th of the GDP.
21:10 OK,
21:10 so
21:11 that's an example of how the democracy is a moral
21:13 triumph in terms of what it's giving the most marginalized.
21:17 OK.
21:17 India does school feeding programs at a much lower level of income,
21:20 whereas in these recent US elections,
21:22 Tim Walls was proudly talking about doing school feeding in Minnesota
21:25 at like $80,000.
21:27 OK,
21:27 like,
21:27 you know,
21:27 and India was doing that way,
21:29 way earlier,
21:30 right?
21:30 And so.
21:31 And this is again,
21:32 it's not just about the democracy,
21:33 it's also the fact that Indian elites are typically
21:36 educated abroad and get their ideas from global norms.
21:39 OK,
21:39 so this,
21:40 this early welfare is a combination of
21:42 top-down ideas and bottom up political support.
21:45 OK,
21:45 so this is a wonderful thing,
21:47 except that it also has two fundamental costs,
21:49 OK.
21:50 And because,
21:51 and the basic point I want to make
21:53 is,
21:53 remember what I said,
21:54 that all of these investments give you 10x ROI,
21:57 so to the economist,
21:58 that's like a $100 bill lying on the sidewalk,
22:00 right?
22:00 So why are we not making these investments?
22:02 And the reason
22:03 is that basically the challenge is that the scope of the
22:06 state has expanded before the strength to deliver against that,
22:09 OK?
22:10 And that gives you two key challenges.
22:12 The first is the public finance challenge,
22:14 which is
22:15 that voter demand and political supply of a welfare state means that
22:18 there just hasn't been enough money to invest in the state itself,
22:21 right?
22:21 So why have we chronically underinvested in the state,
22:25 and that's because the state.
22:26 Those investments take longer to pay off,
22:28 OK,
22:29 and it doesn't typically meet the electoral cycle,
22:31 right?
22:32 And again,
22:32 so Besley Parsan,
22:33 you know,
22:33 there's a theory literature that is all very consistent with how
22:37 you only get these investments in state capacity in times of existential crisis.
22:41 OK,
22:41 so that's in Chile.
22:42 So this is not a new argument,
22:43 right?
22:44 It's just kind of
22:45 helping make sense of India in light of that evidence.
22:48 OK.
22:48 Now the problem is not welfare,
22:50 OK.
22:51 And what I show in the welfare chapter
22:53 is well designed welfare programs can improve both equity and efficiency.
22:56 OK,
22:56 so the problem is not welfare.
22:58 The problem is doing welfare before building the pipes to do it effectively,
23:02 OK?
23:02 And then you get the second key problem,
23:04 which is then the political economy challenge,
23:06 right,
23:07 because what you've done
23:08 is you have expanded people's ability to make a claim
23:10 on the state before the state can serve everybody,
23:13 right?
23:14 And so then the nature of politics becomes
23:16 a politics of scarcity where every group is then
23:19 trying to divert the resources of the state towards
23:21 their group as opposed to broad-based public goods,
23:23 OK?
23:24 And this is magnified by kind of layering the idea of universal and equal,
23:28 equal franchise democracy
23:30 on one of the most fragmented hierarchical societies on the planet.
23:33 OK,
23:34 so the under,
23:34 and this is also true in Africa,
23:36 right?
23:36 So there's almost like an orientalistic sense.
23:39 In which Western political scientists who study democracy
23:41 in South Asia and Africa talk about,
23:43 oh,
23:43 clientelistic policies,
23:45 the truth is that there's a very simple economic explanation,
23:48 right,
23:48 which is you have expanded the claims of the state,
23:51 right before you have the capacity to meet the claims.
23:53 And so then by definition
23:55 there are more claims than there are resources,
23:57 and so you're going to have that politics of patronage and clientelism,
24:01 OK?
24:01 And the problem then is that gets you stuck in this low level.
24:04 Equilibrium,
24:05 particularly when you have
24:06 the mechanics of the first past the post election
24:09 where you only need 35% of votes to win and you have 4 or 5 parties.
24:12 So often you can win with like 20% of the registered voters,
24:16 right,
24:16 which means the incentives are to cultivate like a
24:18 dedicated base rather than cater to the median voter.
24:21 OK,
24:21 so the median voter theorem doesn't apply
24:23 in those cases of fragmented parties
24:26 and when you need to just turn out the base.
24:28 OK,
24:28 so.
24:29 And that gets you why the political incentives
24:31 then don't give you the state capacity,
24:33 despite having democracy,
24:34 OK?
24:35 Now the good news for India is that the nature of elections is changing,
24:39 OK,
24:39 where politicians are now increasingly under pressure to deliver.
24:42 OK,
24:42 so
24:43 if you look at the data in the elections till about 2000,
24:46 essentially there was this well known anti-incumbency effect.
24:48 But after 2000,
24:50 the performing chief ministers are more likely to get reelected.
24:53 OK,
24:53 so
24:53 the politician today wants to deliver,
24:56 OK,
24:56 and that's the key opportunity.
24:58 The problem is the politician is frustrated with the bureaucracy,
25:00 right,
25:01 and they keep blaming the bureaucracy for not being able to deliver.
25:04 But that's unfair
25:06 because
25:06 the bureaucracy's burden reflects the cumulative
25:09 result of 50 years of underinvestment.
25:11 OK,
25:12 so the way I tell the story is just like the
25:14 metaphor for the pre-liberalization Indian economy was the ambassador car.
25:18 OK,
25:19 that is still the metaphor for the Indian state.
25:21 OK,
25:21 so over time we've kept on adding burden to
25:24 the car but without strengthening the car itself,
25:26 right?
25:27 So the politicians are like the race car drivers deciding which
25:29 way to drive the car and which passengers to prior.
25:31 I,
25:32 but the car is barely moving,
25:33 OK,
25:34 and the bureaucrats are like the engine,
25:36 the body,
25:36 and the wheels who are trying to keep this moving and satisfy what the driver wants,
25:40 but you can't both be running and redesigning at the same time,
25:43 OK?
25:44 So
25:44 these guys,
25:45 they would rather be driving a Ferrari than an Ambassador,
25:48 but they're so busy keeping the Ambassador moving that
25:50 you haven't had the time to then rearchitect,
25:52 right,
25:52 the machinery of the state to then reflect
25:55 what is possible with newer data and technology,
25:57 OK,
25:58 so.
25:58 I'm going to skip all of this except to say
26:00 that there are deep structural challenges in the bureaucracy.
26:03 There's staff shortages.
26:04 There's limited accountability as we saw in our early teacher and Dr.
26:07 Absent's work,
26:08 but there's also misdirected accountability,
26:10 OK.
26:10 And so,
26:11 so it's interesting.
26:12 Indramit yesterday was saying the only thing the bank that
26:14 can get you fired is if there's a procurement problem,
26:16 right?
26:16 But that's true for the bureaucracy in India as well,
26:18 OK,
26:19 so it's not that there's no accountability.
26:20 It's just the accountability is all for compliance and not for outcome.
26:24 OK,
26:24 so that's why when Martin Reiser was reading the book,
26:26 he's like,
26:26 I felt like I was just reading about the World Bank,
26:29 OK,
26:30 but the basic point,
26:31 the example here,
26:33 my favorite example is edgTech.
26:34 OK,
26:35 so
26:35 even in cases where schools you manage to procure
26:38 the computers and give them to the schools,
26:40 I have gone to so many schools where you'll see the computer lab is locked.
26:43 And the key is with the principal,
26:44 OK,
26:45 and that's because nobody cares about whether the kids are using the computer,
26:48 right?
26:48 But if a computer is stolen,
26:50 there is hell to pay,
26:50 OK,
26:51 so which means that the best intention policy on the ground,
26:54 the incentives are just different,
26:56 it's an incentive for compliance,
26:57 OK?
26:58 And then
26:59 you know these are issues that are well known about bureaucracies,
27:01 but we have like a lot more research documenting,
27:04 and all of this is in that chapter,
27:05 and there's also a complex set of reasons,
27:07 right?
27:07 I mean,
27:07 it's a colonial legacy
27:09 which reflected the fact that
27:11 The entire Indian state,
27:13 and this is true in many post-colonial democracies,
27:15 that you've inherited a machinery that was designed to rule and not to serve,
27:18 OK,
27:19 and that gives you a lot of the pathologies
27:21 that you see even in the modern bureaucracy.
27:23 And then there's a combination of political,
27:25 economic,
27:25 judicial factors,
27:26 the risk aversion,
27:27 and then
27:28 the fact.
27:29 And,
27:30 and,
27:30 and here's something important that Land
27:32 has talked about,
27:33 right,
27:33 that premature load bearing,
27:34 that putting more load on the body than it can take actually weakens it further,
27:38 right?
27:39 And so as a weightlifter,
27:40 if you want to build muscles,
27:41 you lift 5% more than your capacity and slowly build it.
27:44 If you lift double,
27:45 you're basically going to badly injure yourself.
27:48 So
27:48 the example of that in India
27:50 is how
27:51 many education secretaries tell me they spend about 30 to 40% of their time.
27:55 In court cases,
27:56 OK,
27:56 now why are they in these court cases?
27:58 Many of them are litigation by staff themselves,
28:01 but often they are litigation by citizens who are saying we are
28:03 not getting what you have promised us like in this legislation.
28:06 OK.
28:07 So but the problem is that when you have committed to more than you have,
28:11 you're also spending time
28:13 deciding who gets the service.
28:14 Then you're spending time resolving conflict about who did not get it,
28:17 and that is all time that goes away from actually delivering the service.
28:20 OK,
28:21 so the way I tell the story in India,
28:23 the government as legislature will pass commitments beyond
28:26 the capacity of government as executive to deliver,
28:29 and then government as judiciary will hold government
28:31 as executive in contempt of government as legislature,
28:34 right?
28:34 And so every wing of the government is doing its job the way they think it is,
28:38 but they're tying the system in knots
28:40 because the fundamental problem is not one of intent,
28:42 but one of capacity.
28:43 OK.
28:44 So
28:44 and for the activists who try to To get the state to do stuff.
28:47 What I try to say is
28:48 we're not dealing with kind of an oppressive colonial state.
28:51 We're dealing with an overwhelmed democratic state,
28:53 OK,
28:54 that
28:54 has more expectations than it has capacity.
28:57 And but recognizing that the capacity constraint is the key binding constraint,
29:01 then becomes key.
29:02 So the summary of all of this
29:03 is that
29:04 basically India has not completed the transition
29:07 from being a colonial state designed to rule
29:09 to being a democratic state designed to serve.
29:11 And the converse of that
29:13 is that while much of the language of the book is on kind of technocratic efficiency,
29:17 cost effectiveness,
29:18 it really is a moral imperative,
29:20 right,
29:20 that
29:21 the task of building a more effective Indian state is
29:23 in fact the great unfinished task of Indian democracy itself.
29:26 OK,
29:26 so this is kind of the larger exhortation for why we need to do this,
29:29 right?
29:30 But this is still just chapter 2 and 3.
29:32 OK,
29:32 so,
29:33 so
29:34 this is not the core of the book,
29:35 right?
29:35 The core of the book is how do you actually do this.
29:37 Right,
29:38 it's one thing because state capacity is like motherhood and apple pie,
29:41 right?
29:41 Everybody loves it.
29:42 The question is how do you do it and how do you prioritize what do you do?
29:45 And that's really where the
29:47 one core set of contributions of the book is,
29:49 right,
29:49 which is saying
29:51 there's a ton of research,
29:52 but the book is not just a collection of papers,
29:54 it's kind of organizing the research in a framework
29:58 that is kind of original,
30:00 and it really comes from a theory,
30:01 right,
30:01 which is if you think about these six ideas.
30:04 They are relevant to the effectiveness of any org,
30:06 public,
30:07 private,
30:07 nonprofit,
30:07 right?
30:08 So
30:08 what the org literature says is what drives organizational culture,
30:12 the most important thing is what are your metrics,
30:14 what are you being measured on,
30:15 what are you being reviewed on,
30:16 what are your targets,
30:17 OK,
30:17 that sets kind of where the org moves.
30:20 The second key determinant of org culture is how do you recruit,
30:22 how do you pay,
30:23 how do you promote?
30:24 So
30:24 the way you promote is in fact the culture of the org
30:27 because that becomes the values that often get kind of subliminally,
30:31 you know,
30:32 disseminated through the organization of what values are you promoting for.
30:36 What are you spending your money?
30:37 What are the parts of the org where the budgets are expanding?
30:40 Where are the budgets shrinking?
30:41 And that pivots the direction of the org in a fundamental way.
30:44 What's your revenue model,
30:45 right?
30:46 I mean,
30:46 and that affects everything you do.
30:48 If you're a global multinational like the World Bank or like a large country,
30:51 what do you do in global head office?
30:53 What do you do in national head office?
30:54 What do you do in the field office,
30:55 right?
30:55 And that's that question of federalism and decentralization that.
30:59 Applies to every government in terms of
31:01 optimizing how you're organized.
31:03 And then the last part is the state in the market,
31:05 right,
31:06 post 1937 boundaries of the firm.
31:08 What do you do in-house and what do you outsource to other more competent providers?
31:12 OK,
31:13 so,
31:13 and
31:14 what I'll show you here is three things,
31:16 right?
31:16 One,
31:17 we do incredibly poorly on each of these things.
31:20 OK,
31:20 so that might seem very depressing.
31:22 But the more optimistic part is we also have a ton of high quality
31:26 research in each of these areas that can guide how to do better,
31:29 OK?
31:30 And then,
31:31 the more optimistic part,
31:32 which I haven't talked about,
31:33 is the structure of every chapter is what I call half science and half engineering,
31:38 right?
31:38 So that borrows from
31:39 Greg Mankiw when he came back to Harvard after spending time with his CEA chair,
31:43 wrote this nice.
31:44 JEP piece called The Macroeconomist,
31:46 a Scientist and Engineer,
31:47 right?
31:47 So,
31:48 and that's kind of what we're doing.
31:49 There's one half of each chapter that's social science.
31:52 We're doing research to understand the world,
31:54 and there's the second half that's a systems engineer saying
31:56 how do you use that understanding to build something better,
31:59 OK,
31:59 so
32:00 the structure of each chapter is that way,
32:01 and I'm only going to have time to give you a very high level
32:05 sense of these issues and how core they are,
32:06 right?
32:07 So data and outcome measurement.
32:09 So here's,
32:09 and all of this is based on research,
32:10 right?
32:11 So this is a lovely paper by Abhijit Singh,
32:13 and this looks at official data in India on learning outcomes in one state,
32:16 and that top bar is what's in the official data,
32:19 and this is what is in the independent audit data,
32:22 where Abhijit went back to the same schools and randomly sampled a
32:25 bunch of the same kids and retested them on the same questions,
32:27 OK.
32:28 And what you see is in the official data there's no child who scores below 65%.
32:32 In the independent audit data,
32:33 there's no child who scores above 65%,
32:35 OK?
32:36 And that's how much the data is inflated,
32:38 right?
32:38 And we see this in malnutrition.
32:40 And why is this so foundational,
32:42 right?
32:42 It's foundational because we've wasted 20 years as a country.
32:45 This is completely nonpartisan.
32:46 Every party,
32:47 every state,
32:48 OK.
32:48 And that's because Pratam has been doing independent household surveys
32:52 saying that we have a learning crisis since about 2005.
32:54 But even if a government official can admit that in private,
32:57 in public,
32:58 they have to work with the official data,
33:00 and the official data is all massively inflated.
33:02 OK,
33:02 so CGD,
33:03 I think,
33:03 had a recent paper showing
33:04 that public officials systematically overestimate the level of learning,
33:08 partly because that's in the official data.
33:10 OK,
33:10 so data is foundational.
33:12 And when we think about it's not just data collection,
33:14 but it's data,
33:15 it's analysis and data use,
33:17 OK,
33:17 so the story I'd like to tell to make this start
33:20 is Amazon,
33:21 one company,
33:22 one company has over 1000 PhDs,
33:25 OK,
33:25 1000,
33:26 and over 150 to 200 economics PhDs and over 800 to 1000 data scientists,
33:30 OK,
33:31 and they're just working.
33:32 On micro improvements right for one company,
33:35 right,
33:36 so think about how much more important it is that
33:38 we have that level of intellectual horsepower going into thinking about
33:42 public service delivery at scale given the welfare consequences of that.
33:46 OK,
33:46 so
33:46 the data chapter is foundational.
33:47 There's a whole bunch of ideas,
33:49 some of them we're already doing,
33:50 which we'll talk about in Q&A.
33:52 OK.
33:52 Personnel,
33:53 unlike what many people think,
33:54 the Indian state is actually remarkably understaffed.
33:57 OK,
33:57 so
33:57 we have the fewest employees per capita relative to every major country.
34:01 And the key reason for that
34:03 is government employees in India are the most overpaid in the world.
34:06 OK,
34:07 so this comes from multiple studies.
34:08 There's hedonic regressions,
34:10 and basically it reflects the political economy that the
34:13 insiders are so much more powerful than the outsiders,
34:15 OK,
34:15 that every time there's a budget,
34:17 it goes to increased pay,
34:18 and
34:19 the consequences of this are not just state capacity,
34:21 right,
34:22 because the simple econ 101 answer is you need to pay less and hire more,
34:25 OK.
34:26 But the problem is worse than that.
34:27 The problem is this structure of the public
34:29 sector pay being so much higher than the market
34:32 has massive general equilibrium costs in the
34:34 overall economy because it affects employment,
34:36 skilling,
34:37 job creation,
34:38 right?
34:38 So a recent study estimates that 80% of unemployment of educated youth,
34:42 even in the state of Tamil Nadu,
34:44 is because they're writing government exams
34:45 after government exams after government exams,
34:47 OK?
34:48 And that also explains why do we produce.
34:49 An education system with millions of qualified candidates with no skills,
34:53 and that's because the most lucrative employer in
34:56 the economy only wants exam taking skills,
34:58 right?
34:58 And so
34:59 the tail that wags the dog of the education
35:01 system is where are the returns to the education,
35:03 right?
35:04 And the returns come from cracking the government exam
35:06 and not from being productive in the economy.
35:08 OK,
35:09 so
35:09 the general equilibrium consequences of this are huge,
35:11 and I explore all of that in the book,
35:13 OK.
35:14 And public finance,
35:15 very,
35:15 very,
35:15 very simple picture,
35:17 but perhaps conceptually one of the most important ones that
35:19 applies perhaps to a ton of bank work as well.
35:22 And it's very simple,
35:22 right,
35:23 which is,
35:23 you know,
35:24 in the private sector you can't allocate 100 crores without doing a
35:27 discounted cash flow to say where is the IRR of this investment.
35:30 In government you can routinely spend,
35:32 you know,
35:32 thousands of crores without that discipline,
35:34 OK,
35:35 so you care about efficiency,
35:36 which is the return on investment of your spending,
35:38 but you also care about equity,
35:39 right,
35:40 because we're not a private company.
35:41 But so you can take every major category of government expenditure
35:45 and put it here.
35:46 Now reasonable people may disagree about Region 2 versus Region 4,
35:50 OK.
35:50 But what I want to highlight is there are so many things in region 1 that we don't fund,
35:55 and we spend hundreds of thousands of crores
35:58 on items in region 3 that are bad for both equity and efficiency.
36:01 OK,
36:02 so
36:02 part of the goal of good kind of economic analysis of budgets
36:05 is to shed light on kind of putting items in this quadrant
36:09 and then letting the political process.
36:10 I mean,
36:11 you know,
36:11 as economists,
36:11 we can't take the decisions,
36:13 but we can ask the right questions and present
36:14 the right data so that the political process can then
36:18 act.
36:18 So let me give you just two examples,
36:20 right?
36:20 So
36:20 in the case of rural development,
36:22 on one side you have something like ERA,
36:25 OK,
36:25 that we find is incredibly effective and reduces inequality,
36:29 right?
36:29 So it's up there in region 1.
36:31 OK,
36:32 but on the other hand,
36:32 you have farm subsidies and free electricity for farmers,
36:35 which is probably the single worst policy we have as a country,
36:37 right?
36:38 You can do the incidence analysis and see that the
36:40 top 5% of landowners get 50% of the subsidy.
36:43 It's a disaster for the groundwater table.
36:45 It distorts cropping patterns.
36:46 It kind of,
36:47 it's partly why the discoms are bankrupt,
36:49 and it also hurts industry because of the
36:51 cross subsidy from industry to agricultural subsidies.
36:53 OK,
36:53 so
36:54 it's one of these kind of
36:55 all pervasive
36:57 monsters.
36:58 OK,
36:58 now I'm not saying that we shouldn't support farmers,
37:01 but there's just much more efficient ways to do that,
37:02 OK?
37:03 So part of this is many of these things are well known,
37:05 but using that framework to think about every major sector of how do you do better,
37:09 OK?
37:10 So I'm going to skip revenue beyond just saying two important things,
37:14 OK?
37:14 One is when people are poor.
37:15 OK,
37:16 we normally think,
37:16 oh,
37:16 the government should do more,
37:17 OK,
37:18 but the key point is when the people are poor,
37:20 the government is also poor,
37:21 and the government in relative terms is poorer,
37:24 OK,
37:24 because tax to GDP ratio is lower in poor countries as you have high informality,
37:29 OK?
37:30 And so that increases the premium on value for
37:32 money on your spending because your revenue is much scarcer
37:36 and importantly,
37:38 Your tax instruments are much more inefficient,
37:40 right?
37:40 So the marginal cost of public funds in high income countries is 1.3% to 1.7.
37:44 In India that's estimated as 2.5 to 2.,
37:46 you know,
37:47 2.5 to 3%,
37:48 and it's public finance 101,
37:49 right,
37:49 because only 2% of the population pays income tax.
37:51 And so,
37:52 you know,
37:52 short-based high rates is much higher distortion in a quadratic loss function.
37:56 OK,
37:56 so,
37:57 so it's public finance 101,
37:58 but not enough development students learn public.
38:00 Finance,
38:01 OK,
38:01 and I think that's,
38:02 you know,
38:02 that's,
38:02 that's really important to fix.
38:03 OK,
38:04 uh,
38:05 here's federalism,
38:06 right,
38:06 which is again,
38:07 first principles of federalism would say service delivery
38:09 should happen at a local government level.
38:11 OK.
38:11 India is the most over centralized country in the world,
38:14 right?
38:15 So
38:15 you look at China and think of this as this big powerful communist country.
38:18 In practice,
38:19 over 50% of Chinese public spending happens at the local government level.
38:22 In India,
38:23 that number is 3%.
38:24 OK,
38:24 now the bigger culprit is not center to state,
38:27 but state to local,
38:28 and we can talk in Q&A about the political and
38:30 historical and social reasons for why this is the case,
38:32 but this is a cross-cutting issue,
38:34 right,
38:35 that you can't solve issues of teacher attendance partly because,
38:38 you know,
38:38 the ultimate stakeholder is the community,
38:40 and they have zero power in doing anything about this,
38:43 OK,
38:43 so we can talk about
38:45 and then finally state and market.
38:47 That one of the consequences of weak public service delivery is
38:51 mass exit to private services.
38:52 OK,
38:53 so in education,
38:54 the market share in urban India is about 50%,
38:56 in health,
38:56 it's 70%.
38:57 The private security industry hires 3.5 times as many
39:00 people as every government police person put together,
39:02 OK?
39:03 And
39:04 So the key point here
39:05 is that the government plays 3 very distinct roles in the ecosystem,
39:09 right?
39:09 You're a policymaker,
39:10 you're a regulator,
39:11 and you're a provider.
39:12 And the way you have to think about the
39:14 private sector is different in those three hats.
39:16 As a policymaker,
39:17 the private sector is your ally
39:19 because they help you deliver.
39:21 As a regulator,
39:21 you have to treat them as an equal.
39:23 As a provider,
39:24 they are your.
39:24 Competition.
39:25 OK.
39:26 The problem is 95% of government budgets
39:28 and personnel go to government as providers,
39:30 so their DNA is to be deeply suspicious of the private sector,
39:33 OK?
39:34 And there's many examples I can give you about how
39:36 the hostility to the private sector hurts the public good.
39:39 And the key point is that governments can greatly increase public welfare
39:43 if they pay attention to making markets work better.
39:46 And again,
39:46 there's a lot of fantastic micro research,
39:48 right,
39:49 showing how providing better information,
39:51 like I mean allows consumers to make better
39:53 choices and improves outcomes in the entire market,
39:55 right?
39:56 But that's a classic public good that's underprovided
39:59 and it's something that governments should do,
40:01 but they don't,
40:01 OK?
40:02 And similarly leveraging private providers to serve the poor
40:05 is also something that can massively improve welfare,
40:08 but both these functions also require state capacity,
40:11 right?
40:11 So you need state capacity for procurement and regulation and making markets work.
40:15 Better and we don't have that capacity either,
40:17 OK,
40:18 so the bottom line is I promise you I'll be optimistic.
40:20 OK,
40:21 so the reason I'm optimistic is the fact that we're at such a low base gives us so
40:25 much to work on and there is so much
40:27 free lunch that we can do to accelerate development.
40:29 OK,
40:30 so let me take my last 10 minutes
40:32 and think
40:33 we're not looking at a state for its own sake,
40:35 right?
40:35 We want an effective state because we
40:37 want to deliver essential attributes of development,
40:40 right,
40:41 education,
40:41 skills,
40:41 health,
40:42 safety,
40:42 justice,
40:43 welfare,
40:44 jobs,
40:44 OK.
40:45 And
40:46 so one way to think about situate what the book is doing
40:48 in the grander debates on development is over the past 50 years
40:52 one of the big debates has always been growth versus development.
40:55 OK,
40:55 so
40:56 every development outcome is positively correlated with wealth.
40:59 So the growth view is saying get 2% faster growth for 40 years,
41:03 everything else will take care of itself,
41:04 right?
41:05 And that in India is the Bhagwati.
41:06 Panharia view.
41:07 The development view,
41:08 which was famously Amartya Sen and Gres,
41:10 were to say,
41:10 hey,
41:11 the point of development is not GDP per capita.
41:13 The point of development is human capabilities,
41:16 and let's look at the Sri Lankas,
41:18 the Chinas,
41:18 the Vietnams,
41:19 the Cubass,
41:19 the Keralas of the world,
41:21 and what that tells you is you don't have to wait to get rich to get developed,
41:24 OK,
41:24 if you're willing to invest more in your social sector,
41:26 right.
41:27 And so that's been this big fight for like 50 years right
41:31 now
41:31 at some level they're both right because
41:33 clearly more growth helps human development and
41:35 more human development feeds back into growth.
41:37 So what are we fighting about?
41:38 In the end it becomes an ideological fight,
41:40 right,
41:40 that the center right likes capex and the left wants social sector,
41:44 right?
41:44 And my point of departure in this debate is to say,
41:46 hey guys,
41:47 the elephant in the room is that the quality of spending is so bad,
41:51 regardless of whether you do capex or
41:53 the social sector,
41:54 that the free lunch
41:55 is that if you can improve the effectiveness of public spending,
41:58 you can do more of everything,
41:59 OK?
42:00 So connecting my argument to macro development as a casse.
42:03 Paper that says 20% of the variation in GDP per capita is physical capital,
42:07 20% is human capital,
42:09 and 60% is TFP.
42:10 OK,
42:11 so you could think about my overall argument as
42:13 being about the TFP of the public sector,
42:15 and that's where the free lunch is
42:17 in terms of how you can do better without necessarily spending more.
42:21 OK,
42:21 so.
42:22 Now,
42:23 what is the reason for optimism,
42:24 and this comes from the shifting of the Preston curve,
42:27 which is one of,
42:27 and it's,
42:28 this is a different Preston,
42:29 but it's very nice to make this point in the Preston Auditorium.
42:32 OK,
42:33 this is Samuel Preston.
42:35 But
42:35 this picture is one of the great triumphs of
42:37 development in the second half of the 20th century,
42:39 right,
42:40 which is basically showing
42:41 that x axis is GDP per capita,
42:43 and in 1930,
42:44 that's what life expectancy was.
42:45 In 1960 that had shifted up.
42:47 In 2015,
42:48 that's gone up even further,
42:49 OK.
42:50 So what that means is at the same level of
42:52 income over time you're able to deliver better development outcomes,
42:56 OK,
42:56 and the key magic for all of this is new technologies,
43:00 right,
43:00 technologies that we didn't have,
43:01 right?
43:01 So
43:02 100 years ago
43:03 the president of the US could not save his son from a
43:06 bacterial infection caught playing tennis with a blister on his toe,
43:09 OK.
43:09 And today you have low cost antibiotics that that's
43:12 something that a daily wage laborer could have.
43:14 We got a COVID vaccine in a year,
43:15 which is a miracle,
43:16 right?
43:16 I mean,
43:17 compared to comparable flus in the past,
43:19 right?
43:20 So when you then think about what have been the key drivers of human progress,
43:23 the most important driver has been
43:25 the creation of new knowledge,
43:27 the transmission of that knowledge,
43:28 and then acting on that knowledge,
43:29 OK?
43:30 And health is the field that has had the Best established tradition
43:34 of evidence-based collection of knowledge,
43:37 institutional mechanisms for synthesizing that,
43:39 and then ways of pushing that into the policy regime like I mean to say we
43:43 are now going to build the public systems needed to deliver those benefits at scale.
43:46 OK,
43:47 so the question then is how can we accelerate development in every other sector?
43:52 How can you shift the precedent curve for governance and development more broadly?
43:55 And the way to ask the question
43:57 is what do today's poor countries have that
43:59 today's rich countries did not have 2030,
44:02 40 years ago?
44:03 And the key thing is we have much better
44:05 data and evidence and we have much better technology,
44:07 right?
44:08 So
44:08 the explosion of microdata
44:11 means that we now
44:12 understand so much more about how to improve quality of expenditure
44:16 that the free lunch
44:18 is that we spend so much on expensive things that have no impact.
44:21 Other very less expensive things.
44:23 And not being done.
44:25 And this is why you need the research and the evidence.
44:26 So part of my job is to motivate the community of
44:28 researchers as to why what we do matters so much.
44:31 OK.
44:31 So
44:32 as an economist,
44:33 I don't worry about how Toyota makes its cars
44:35 because you have market prices for inputs and outputs,
44:37 but you can spend money in the public sector badly.
44:40 You can spend taxpayer money badly for a very long time.
44:42 OK.
44:43 So here's a simple example of the power of this research,
44:46 right?
44:46 It comes from one of my studies on education and edTech in India,
44:49 and
44:50 the x axis here is the grade in which the kids are enrolled in.
44:53 The y axis is the learning level.
44:55 Now if you were making progress as per the curriculum,
44:58 you would be on that 45 degree line.
45:00 OK.
45:00 In practice,
45:01 the true level of learning is at the red line,
45:04 OK,
45:04 at about half of the curriculum.
45:06 So
45:06 the average 8th grade kid has about a 4th grade level of learning.
45:11 But more importantly,
45:12 with dynamic computer-based testing,
45:14 we can pinpoint where every child in the sample is.
45:17 OK,
45:17 so that means in the 8th grade classroom.
45:19 I have some kids at a 2nd grade level,
45:21 some at a 3rd,
45:21 some at a 4th,
45:22 all the way up to 8th grade.
45:23 And so then we,
45:24 it makes sense why simply spending more doesn't work,
45:27 because
45:28 one part is the absent teacher,
45:29 but even the very sincere teacher who's in
45:31 class and teaching is teaching to the textbook,
45:33 and that's way above the levels of most kids.
45:35 So jit qualified jit be motivatedjit be sincere.
46:04 So what have I just done to you,
46:06 right?
46:07 I have just given you 45 seconds in the shoes of a typical kid in a typical classroom.
46:13 In millions of schools around the world.
46:17 the one that didn't.
46:20 So,
46:20 no,
46:20 but,
46:21 but,
46:21 but you see,
46:21 the fact that 5% of the kids understood is exactly what would happen,
46:24 and that's who the teacher would be looking at,
46:26 right?
46:26 And then you make eye contact with the kid who's getting it,
46:29 and that's the kid up there.
46:31 Right?
46:32 And so the point is that the data then gives us,
46:35 there's nowhere to hide,
46:36 right,
46:37 about why is our spending not translating into outcomes,
46:39 and that's the power of the research,
46:41 OK.
46:41 So,
46:42 so if you look at the present curve right now,
46:44 we have 35% stunting in India,
46:46 and if you project the income elasticity of translation of
46:49 income growth and outcomes over the past 20 years,
46:52 even at 6% growth,
46:53 you'll still end up at 25% in 2047.
46:56 Now the interesting thing is,
46:57 even if you grow at 8%,
46:59 that only falls to 22%.
47:01 And that is also an upper bound of increasing the budget by 2% a year,
47:04 OK,
47:05 but if I can improve the TFP,
47:07 the effectiveness of public spending,
47:09 and double that going forward,
47:10 then even at 6% growth,
47:12 you do much,
47:12 much better.
47:14 And doubling seems ambitious,
47:15 but remember I've given you tons of ideas that have a 10x,
47:18 20x ROI,
47:19 OK,
47:20 so I'm only talking about doubling.
47:22 So that's possible
47:24 And then when you do that,
47:25 you get a tailwind back into faster growth,
47:27 and that makes it more likely that you'll get to 8%.
47:29 OK,
47:30 so the point about in a world of pandemics and kind of,
47:33 you know,
47:33 wars and
47:34 slowing global growth
47:36 where the fiscal space,
47:37 and we're seeing what's happening in Washington,
47:38 right,
47:39 you know,
47:39 where the fiscal space available for kind of just aid and development is shrinking,
47:44 like what are we going to do?
47:45 Like,
47:45 you know,
47:46 you need a new recipe,
47:47 right?
47:48 And that new recipe here is to really.
47:50 I think about the centrality of governance as the
47:53 force multiplier for everything we're doing in development,
47:55 and that's how you're going to accelerate development,
47:57 OK,
47:59 so yeah,
47:59 so I'm basically almost done,
48:00 like,
48:00 you know,
48:01 so what I have here
48:02 is just these 6 chapters,
48:03 and why are these 6 chapters what I'm focusing on?
48:06 Because in picking these chapters we don't have to choose between Bhagwati and Sen,
48:09 OK,
48:10 because
48:10 every one of these topics are intrinsically important for human welfare,
48:14 which is what Sen would want,
48:16 and One of them is instrumentally important for accelerating growth.
48:19 OK,
48:19 so
48:20 we should be able to build a broad consensus to say this is what we've got to do.
48:23 But my point is you're not going to get there just by raising budgets.
48:26 You're going to have to really
48:27 build an effective state,
48:28 OK?
48:29 And then there's the last part of making it happen,
48:31 which is in the beginning I'm arguing that
48:33 even with the current incentives of politicians and bureaucrats,
48:35 they should want to do that.
48:37 But eventually you'll need to institutionalize this practice with
48:40 institutions and activate citizens to kind of demand that.
48:42 And so what I'm going to do in my last 1.5 minutes,
48:45 OK,
48:45 is just
48:46 situate kind of what I'm doing in the global discourse and development,
48:49 OK.
48:50 And the first main piece is that
48:52 the discourse on institutions and development AJR has
48:54 been obviously the big elephant in the room,
48:56 OK,
48:57 but there's this very nice picture by Arvind
48:59 Subramanyam that highlights the limitations of that framework,
49:02 right,
49:02 which is it's not to say it doesn't work because
49:04 on average the democracies and inclusive institutions do better,
49:07 but you've got this big outlier of China and India that relative to democracy,
49:11 this is just like half the world,
49:12 right?
49:12 Like I mean that it doesn't explain as well,
49:14 not to mention the rest of East Asia.
49:16 So what I want to argue is that when we think about governance.
49:19 It's useful to actually think about three components of the quality of governance.
49:23 These are related but distinct,
49:25 OK?
49:25 And I'm calling this representation constraints and capacity,
49:29 right?
49:30 So representative representativeness matters to say who has a
49:33 voice in determining the objective function of the government.
49:35 The constraints matter to say is there a constraint in the executive,
49:38 whether it's property rights or freedom of speech or individual rights,
49:41 but those two are not enough if you also don't have the capacity to deliver.
49:45 OK,
49:45 so what I'm arguing is that too much.
49:47 Of our discourse on governance has focused on the
49:50 first two pieces and not enough on capacity,
49:53 and the advantage of focusing on capacity in some ways it can
49:56 also be slightly more value neutral for an organization like the bank
49:59 in terms of saying every country needs more kind of capacity to deliver better.
50:03 OK,
50:04 so my last slide
50:05 is this is kind of what was in the preface to the global edition,
50:08 right?
50:08 So
50:09 when I think about what is the,
50:10 so the book is written in a very
50:11 accessible way that hides the academic contribution.
50:14 So that's why it's sitting in the preface in the global edition.
50:16 But
50:17 what I'm doing here to the state capacity literature
50:19 is coming back and saying
50:21 there was a wave of kind of country case studies,
50:24 right,
50:24 with the developmental state Chalmers Johnson,
50:26 Alice Amsden,
50:27 Bob Wade,
50:28 and all of that work highlights the importance of the developmental state,
50:32 OK,
50:32 the centrality of an effective state for development.
50:34 The problem is that selecting on the dependent variable,
50:36 and you have more theories than data points,
50:38 OK,
50:38 so it doesn't let you actually do inference,
50:40 right?
50:41 And so it teaches us a lot,
50:42 but you can't do inference,
50:44 and that gets us to the next wave of cross-country regressions.
50:46 OK.
50:47 Now that was very valuable in terms of establishing certain broad patterns,
50:51 and the whole AJR Nobel is just because they found an
50:53 instrument right for what was otherwise a bunch of correlations.
50:55 OK,
50:56 so clearly that taught us a bunch of stuff,
50:58 but it was limited in two ways.
50:59 One,
50:59 you could never fully rule out endogeneity and sequencing issues.
51:03 And second,
51:03 It doesn't tell you how to get better,
51:04 OK?
51:05 It just says institutions matter,
51:06 OK.
51:07 So what I'm doing to that is now coming back to the state capacity literature,
51:10 building on 20 years of microfounded research,
51:14 and this is something many of you have contributed to and partly that's the,
51:17 I think
51:18 what I value this conference is I see this as bringing together an
51:21 epistemic community of people who are working on these topics from a micropro.
51:25 Perspective that then I'm providing some structure
51:28 into how all of that research fits together
51:30 to answer a bigger picture question of state capacity.
51:33 OK,
51:33 so that's the state capacity.
51:34 The second part
51:36 is
51:36 most of the work so far is focused on the state as an institution,
51:40 but not enough on the state as an organization.
51:43 OK,
51:43 so
51:43 we think about what the state does,
51:45 we don't think enough about how it does it,
51:47 OK.
51:48 The third is it's also providing a practical roadmap.
51:51 OK,
51:52 now
51:52 the way to think about the policy ideas in the book is it's not gospel truth saying,
51:56 oh,
51:56 you must do this,
51:57 but it's my starting point of saying if based on everything
52:01 I know you were to ask me where I would start,
52:03 that's where I would start
52:04 for iteration and further evaluation,
52:06 right?
52:06 So every one of the ideas in the book could use further iteration,
52:09 evaluation,
52:09 RCTs.
52:10 So there's tons of ongoing research to be done,
52:12 OK,
52:13 that can be on that agenda.
52:14 And finally,
52:15 I think this is important.
52:17 Because I do think state capacity is a topic where less is more,
52:20 OK,
52:21 where going deeper into a country teaches you more than cross country,
52:24 and that's because state capacity is built within a country,
52:27 OK?
52:27 It's not built across countries,
52:29 right?
52:29 So,
52:30 so what I'm doing here is combining the micro-founded research with the
52:33 area studies tradition of doing deeper
52:35 work that is historically contextually grounded,
52:38 but with principles that would then apply across.
52:41 And finally to the development literature.
52:43 I think like Arturo was saying,
52:45 the global development discourse itself has gone through three phases,
52:47 right?
52:47 There's 50s and 60s was all state led development that kind of hit a cul de sac,
52:52 wasn't working by the 70s.
52:53 Then we went to Washington Consensus and said get the state out.
52:56 That delivered growth,
52:57 but not enough development.
52:59 Then we went to the MDG SDG era of global development where
53:02 we said we want to focus much more on these developmental goals
53:05 and you see that in the patterns of bank
53:07 lending of how much more goes to social sector relative
53:09 To kind of the hard infrastructure
53:11 and now that's also hitting a bit of a wall that
53:14 the budgets are not translating into outcomes.
53:17 OK,
53:17 so what I'm arguing is for global development
53:19 thinking itself it's kind of important to recognize
53:22 that every development goal that the bank wants
53:24 has to be delivered through nation states,
53:26 OK?
53:27 And if the nation states don't have the capacity to deliver,
53:30 then that's your binding constraint for every development goal,
53:33 OK,
53:33 so that's 0.1.
53:34 The second
53:35 is.
53:36 Applying the lessons of shifting the Preston curve,
53:39 and I think that provides a really nice framework for us to
53:41 be optimistic about what is it that we're trying to do.
53:44 Why,
53:45 why does the research matter?
53:46 Why does the knowledge synthesis matter?
53:47 Why does the practice matter?
53:49 That Preston curve gives us that shifting of the Preston curve gives us the language
53:53 to kind of
53:54 activate our imagination,
53:55 OK,
53:56 as to what is possible.
53:57 So
53:58 the way I see this
53:59 is intellectually it's connecting the threads across 4 sets of no.
54:02 Is in development economics,
54:03 right,
54:04 because
54:04 Zen tells us that it's not about growth but about development,
54:07 which is what I'm doing by motivating this that way.
54:09 It's Deaton who highlights the shifting of the
54:12 Preston curve and the centrality of public action
54:15 and what the Banerjee Duflo Kramer Nobel
54:16 is doing is reflecting the 1000 micro studies
54:19 that gives us the body of evidence that can do that
54:22 in other sectors as well.
54:24 And then it's engaging with AJR by saying,
54:26 Hey guys,
54:26 this matters,
54:27 but there's this big missing piece.
54:29 There,
54:29 right,
54:29 which is the state,
54:30 and that's kind of how this is sitting in
54:32 the discourse on development.
54:34 And then finally it's also a template for building insights
54:37 from high quality micro RCTs to answering certain bigger questions.
54:41 OK,
54:42 so we can talk more about that.
54:43 And then finally
54:45 it is,
54:45 it recognizes the political determinants of state capacity and so
54:49 this is not like a naive technocrat's
54:51 manifesto that doesn't recognize the politics,
54:54 but what I'm trying to do on the.
54:55 Takes is just build a broad coalition by arguing that
54:58 the state capacity agenda should be able to unite the left and the right,
55:02 because if you're on the left and you care about equity and justice,
55:05 there's no better way to deliver at scale than have a more effective state.
55:09 And if you're on the right and you care about value for money,
55:11 cost effectiveness,
55:12 and laying the foundations for faster growth,
55:14 there is no better way.
55:15 You can't do this by getting the state out of the way,
55:18 and that's kind of,
55:18 you know,
55:19 the core point about
55:20 why you need the state.
55:21 And then there is an important point which maybe I'll take in Q&A,
55:24 also engaging with critiques of state capacity,
55:26 right?
55:27 You might worry about what happens if you enhance
55:29 capacity in the hands of a malevolent actor,
55:31 OK,
55:32 and we'll take that in Q&A.
55:33 And finally for orgs like the World Bank,
55:35 I think
55:36 my major kind of plea here
55:38 is what I've seen in the work of even
55:40 institutional banks and certainly for smaller ones over time
55:43 is we have gone from trying to
55:44 deliver development to basically delivering projects,
55:47 OK?
55:47 And when you deliver projects,
55:48 you focus on building PMUs,
55:50 and these PMUs are often staffed by consultants,
55:53 and you're trying to do first world fiduciary standards,
55:56 OK,
55:57 in settings that have much weaker governance,
55:59 right?
55:59 So like Iammeet was saying yesterday,
56:00 the one thing that gets Fired in the bank is if you're not meeting kind of
56:03 procurement standards of like a high income country,
56:06 right,
56:06 so you basically built these little fences
56:09 of operational excellence around your projects,
56:11 but that is at most 1 or 2% of national budgets.
56:14 OK,
56:14 so if you want to go from delivering projects to delivering development,
56:18 you have to bring the governance across the
56:21 sector to the core of the development agenda
56:23 and the core of the bank's agenda,
56:25 and I'm happy to talk more about that in Q&A.
56:27 So thank you so much and sorry for going over time.
56:38 Thank you.
56:38 Uh,
56:38 he did something unusual.
56:40 Am I allowed to do something unusual?
56:42 Can,
56:42 can we go back to your 27 slide?
56:45 Uh,
56:46 no,
56:46 because,
56:46 uh,
56:46 he,
56:46 he finished his,
56:47 his chat saying
56:49 that,
56:49 um,
56:50 he has reason to be,
56:52 uh,
56:53 to be up,
56:54 up probably to the index of the book,
56:56 I guess.
56:57 Uh,
56:57 yeah,
56:58 that one.
56:59 So,
56:59 to be optimistic,
57:00 and I,
57:00 I,
57:01 when I see this slide,
57:01 I,
57:02 I feel optimistic too
57:04 because if,
57:04 if you look to the,
57:05 to the index,
57:06 it has an 80% overlap with what we actually do in governance.
57:11 Uh,
57:11 there's even a,
57:11 there's even a team doing reimagining PFM so it's less likely
57:15 less uh
57:16 a constrained version of chapter.
57:19 of
57:20 of chapters 17,
57:21 so it seems that we are,
57:23 uh,
57:23 so that we are in the right uh uh track.
57:26 So,
57:26 uh,
57:27 we have a little time,
57:28 but I have a,
57:29 a proposal to how to do it.
57:31 Um,
57:32 usually my role is to insist that you go to the point.
57:35 If you don't want to the point,
57:36 that's fine,
57:37 but you have to speak as fast as him.
57:39 Uh,
57:40 uh,
57:40 other,
57:41 otherwise you have to make a very,
57:43 very concrete,
57:44 uh,
57:45 question.
57:45 OK.
57:45 And as,
57:46 and we have time for probably
57:48 33 questions.
57:53 You could also run to check to see who,
57:55 who could ask the question first.
57:58 Uh,
57:59 hi,
57:59 Dr.
57:59 Kartik.
58:00 Uh,
58:00 I think amazing presentation.
58:02 Let me treat this like an economic seminar.
58:03 Two questions on the internal validity and the external validity
58:06 on the internal,
58:08 that slide on teacher salaries.
58:10 I've done similar work in another place where we've seen
58:13 that about 65% of all civil servants
58:16 fall within the top two quintiles.
58:17 But then I also,
58:18 what I did was I plugged where the minimum,
58:21 uh,
58:21 minimum salary is
58:22 or the,
58:23 the poverty line is of that country.
58:25 About 23% of the civil servants earn below the poverty line.
58:28 So if you plug those things together,
58:29 if you plug the minimum
58:30 wage or the poverty line in your figure,
58:33 how does the implication of the figure then change,
58:36 especially in the context of when you're saying
58:37 we need 1000 PhDs
58:39 again and to be able to compete with Facebook.
58:41 Who could offer that market wage to these PhDs
58:44 to come into the government.
58:45 So how do you expect that bell curve to change
58:47 on the external validity?
58:49 I mean,
58:49 if you were a bit of a more of a provocateur,
58:51 you could have
58:52 said
58:53 government effectiveness as civil servant
58:55 effectiveness or the bureaucratic efficiency.
58:58 And so on that point,
59:00 I mean we're
59:01 in an institution where we've recently published a working paper which says,
59:04 and I quote in the abstract,
59:06 compensation of government employees
59:09 has no impact either positive or negative on growth in the long term.
59:14 So
59:15 how do we sort of engage with this discourse
59:17 in an institution where there is that sort of,
59:20 uh,
59:22 movement happening?
59:22 Thank you.
59:24 So I actually have a,
59:26 a question that it's not my own.
59:27 So Inder Mitjal,
59:28 our chief economist,
59:29 is actually watching it
59:31 online and he sent me a question.
59:33 So his question is,
59:34 why does the Indian state do fine in mission mode but not in normal mode?
59:38 So I'm assuming it's like
59:40 crisis versus normal time,
59:41 which applies to also other countries too that sometimes work better in
59:44 in
59:45 crisis mode than actual uh
59:47 normal mode.
59:51 Hi,
59:51 um,
59:52 thanks so much.
59:53 Uh,
59:53 this was a very fascinating,
59:54 uh,
59:55 presentation,
59:55 Kartik.
59:56 So you talked about the,
59:59 my name is Saeed Madadi.
1:00:00 I'm a public sector specialist in governance.
1:00:02 Um,
1:00:05 I did work with him.
1:00:06 Yes.
1:00:07 Um,
1:00:09 you talked about the Indian federalism,
1:00:11 and you said that compared even to China,
1:00:13 India is much more centralized if you think
1:00:15 of the service delivery at the local level.
1:00:17 How much of
1:00:19 that do you think is a result of the political
1:00:21 economic context of the Indian democracy in the aftermath of
1:00:25 the independence?
1:00:26 If you have a very large multi-ethnic society,
1:00:28 you might want to give voice to people without maybe as much power,
1:00:31 right?
1:00:32 And what is the implication of that for the state capacity today,
1:00:34 both in terms of,
1:00:35 is there any correlation or even affinity between state capacity and resources,
1:00:40 both vertically in the,
1:00:41 in the federal system,
1:00:43 but also across the country in different states.
1:00:45 Thanks.
1:00:46 Uh,
1:00:47 so
1:00:49 Uh,
1:00:49 so first,
1:00:49 I think,
1:00:50 see,
1:00:50 so I'll say two things,
1:00:51 right?
1:00:51 So one is
1:00:53 this,
1:00:53 this point about
1:00:55 are the salaries too high and this World Bank point about the level doesn't matter.
1:00:58 I think,
1:00:59 so where I kind of have a nuance in between view on that,
1:01:02 right?
1:01:02 I mean,
1:01:03 it is,
1:01:03 I agree with the view that the wage bill doesn't seem to matter,
1:01:06 mainly because the wage bill goes in level of pain,
1:01:09 OK?
1:01:09 But
1:01:10 if there is even a component of that that is linked to some metric of performance.
1:01:14 Now I have a,
1:01:14 but I have a very realistic proposal in the book,
1:01:16 right?
1:01:17 I know performance measurement in the public sector
1:01:19 is almost impossible and it's very difficult.
1:01:21 OK,
1:01:21 so I have never pushed for the scaling up of my
1:01:23 performance-based studies because I see those as a proof of concept,
1:01:26 right,
1:01:26 rather than,
1:01:27 so it illustrates how much slack there is rather than saying scale this up.
1:01:31 OK,
1:01:32 so and I talk about that
1:01:33 at some length in terms of how to learn from our cities,
1:01:36 right?
1:01:36 If you focus on the principle that is illustrated.
1:01:39 As opposed to the point estimate and see that whole scale of this thing,
1:01:42 because
1:01:43 then you run into all these troubles of external,
1:01:44 right?
1:01:45 So,
1:01:45 but what I've
1:01:46 suggested is saying,
1:01:47 can we at least have competence based,
1:01:49 OK,
1:01:50 like career planning,
1:01:51 because today when you join the government in India,
1:01:53 we've got the most perverse set of incentives that
1:01:55 you do all your work before getting in,
1:01:57 and then the moment you're in,
1:01:58 there's no incentive to kind of upgrade your skills,
1:02:00 right?
1:02:00 So the entire government is run by consultants,
1:02:02 OK,
1:02:03 like I mean the big four essentially is having a party in every department,
1:02:07 right?
1:02:07 And if you're a secretary,
1:02:08 the reason you use the consultants is precisely because they are more accountable,
1:02:12 they are flexible,
1:02:12 and they have the incentives to have the skills.
1:02:14 OK,
1:02:14 so,
1:02:15 so you're,
1:02:15 so the reason I'm not unsympathetic,
1:02:17 OK,
1:02:18 to that headline is that,
1:02:19 that I take it not as a way of saying,
1:02:22 hey,
1:02:22 pay doesn't matter,
1:02:23 shrink the state.
1:02:24 See that as
1:02:25 I'm not comfortable expanding the state without fixing its internal plumbing.
1:02:29 OK,
1:02:30 so that's,
1:02:30 I think,
1:02:30 that delicate balance.
1:02:32 So it's like this old Woody,
1:02:33 uh,
1:02:34 I think,
1:02:34 story from this movie,
1:02:35 right?
1:02:35 Like I mean and Annie Hall,
1:02:36 I think,
1:02:36 where
1:02:37 two guys in the restaurant where the first one says the food in this
1:02:39 restaurant is terrible and the second one says the portions are too small.
1:02:42 OK,
1:02:42 so you've got to kind of both,
1:02:44 you know,
1:02:44 expand the size and fix kind of the recipe,
1:02:47 right?
1:02:47 Um,
1:02:48 and then this point about poverty.
1:02:50 See,
1:02:50 I think the point is well taken,
1:02:51 but remember what I said,
1:02:52 right?
1:02:53 When the people are poor,
1:02:54 the government is also poor,
1:02:55 and the government is poorer,
1:02:57 OK.
1:02:57 So the important point to keep in mind,
1:02:59 even Drez and Sen,
1:03:00 OK,
1:03:01 like I mean,
1:03:01 who are not like,
1:03:02 you know.
1:03:03 Market friendly capitalist guys,
1:03:04 if you look at what they're saying in 2013,
1:03:06 they say that one of the biggest things China,
1:03:08 China and Vietnam,
1:03:09 teachers' salaries were 13 what they are relative to GDP as you have in India.
1:03:13 OK,
1:03:13 so
1:03:14 it was not about the level of pay.
1:03:15 It really was about,
1:03:16 so I think here is the other challenge.
1:03:18 The challenge is that
1:03:20 we have the wrong talent model in many ways for basic health and basic education.
1:03:24 OK?
1:03:24 We think we need highly qualified folks to go
1:03:27 in rural areas and teach that kind of kid.
1:03:29 And we pay high to kind of attract the best talent through a civil service exam.
1:03:33 The problem is that talent then has no connection to the community.
1:03:37 It doesn't want to live in the village,
1:03:38 right?
1:03:38 So
1:03:38 the problem of teacher absence partly comes because you've,
1:03:42 you've thought about civil services more from the perspective of careers of
1:03:44 the employees rather than the services they deliver for the citizens,
1:03:47 right?
1:03:47 So going back to Amartya Sen,
1:03:49 appropriate technology,
1:03:51 right?
1:03:51 Some of the most effective service delivery experiments in
1:03:53 India have been hiring locally and upskilling the local talent
1:03:58 because they are committed to that community.
1:03:59 And once you get to a higher level,
1:04:01 then you may need that,
1:04:02 right?
1:04:02 So I think that's why I don't buy this argument.
1:04:04 Like I mean that
1:04:05 you need higher pay because there's the spatial mismatch as
1:04:07 well as the right level of skills that you need.
1:04:09 Now the 1000 PhDs is kind of on the other end,
1:04:12 right?
1:04:12 Part of the problem in government is you have so much wage compression
1:04:15 that at the high end,
1:04:16 like You're massively underpaid relative to what
1:04:19 that talent would command in the market,
1:04:20 but at the low end for 95% of employees,
1:04:22 you're paid 5 times the market wage,
1:04:24 and because that's where the numbers are,
1:04:26 you don't have the fiscal space.
1:04:27 OK,
1:04:27 so,
1:04:27 and that's going to take,
1:04:28 it's not
1:04:29 politically easy,
1:04:30 but that's basically what Singapore did,
1:04:32 right?
1:04:32 Like I mean which is saying
1:04:33 at the high end we're going to have to pay to attract talent for certain key roles,
1:04:37 and that's a much,
1:04:38 I think,
1:04:39 complex,
1:04:39 but I think we need a nuanced conversation on
1:04:41 what those civil service reforms need to look like.
1:04:43 I'll give you another example,
1:04:44 right,
1:04:44 in the police.
1:04:46 So in India,
1:04:46 like there's a one the Director General told me,
1:04:49 he said,
1:04:49 I have a ton of positions for beat constables,
1:04:52 OK,
1:04:52 but that's not what I need.
1:04:53 I need people with data analytic skills who can do cybercrime and forensics,
1:04:57 and I'm not able to get those people,
1:04:58 right?
1:04:59 So it's just again
1:05:00 smart,
1:05:01 you know,
1:05:01 uh,
1:05:02 smart sizing the state,
1:05:03 OK,
1:05:03 um,
1:05:04 so I think,
1:05:04 uh,
1:05:04 Thita and Irameet really,
1:05:06 the,
1:05:07 I think the answer is,
1:05:08 so Devish Kapoor has talked about it,
1:05:09 right?
1:05:09 Like,
1:05:09 you know,
1:05:09 and I think
1:05:10 the basic point of functioning in mission mode is two or three things,
1:05:14 right?
1:05:14 One.
1:05:14 There's a well defined output and end date,
1:05:18 and you are marshaling the entire machinery of the state,
1:05:21 right?
1:05:21 So when you,
1:05:22 when India does elections,
1:05:23 we do the best elections in the world,
1:05:25 but that's because the entire machinery of the state is focused on that
1:05:28 and the election commission is able to
1:05:29 commandeer the entire civil service machinery,
1:05:32 right?
1:05:32 So
1:05:32 the point is I have capacity.
1:05:34 It's not like I don't have capacity.
1:05:35 I have this much capacity.
1:05:36 I have this much tasks,
1:05:37 OK?
1:05:37 But if I take all my capacity and put it on a few well defined tasks,
1:05:40 I will deliver.
1:05:41 OK,
1:05:41 so that's basically what that story is.
1:05:43 I think the federalism question is a really deep one.
1:05:46 And you know,
1:05:47 like I said,
1:05:48 so the chapter also talks about the political,
1:05:50 economic,
1:05:51 and social reasons for why India is so over centralized.
1:05:54 OK,
1:05:54 it wasn't like a power grab per se,
1:05:56 right?
1:05:57 So the political reason is because this is coming
1:05:59 in the aftermath of partition and the aftermath of
1:06:02 needing to integrate 500 princely states that had never
1:06:04 been part of one kind of unified union.
1:06:07 OK,
1:06:07 so
1:06:07 one of the reasons politically the central
1:06:09 government has so much power that they could
1:06:11 even dismiss democratically elected state governments was because
1:06:14 you were trying to build a nation,
1:06:15 OK.
1:06:16 The economic reason for this was the belief
1:06:19 that most development required high capital intensive investments,
1:06:23 and the central government was seen as the only entity that could
1:06:25 marshal those resources in a way that local governments could not.
1:06:28 But I think the most important and underappreciated reason.
1:06:32 For the over centralizing of service delivery
1:06:35 is because the framers of the Constitution were also trying to do social reform,
1:06:39 OK,
1:06:39 because
1:06:40 they were concerned that given both the
1:06:42 patriarchy and the feudalism of Indian villages,
1:06:44 that local elites would never allow the marginalized to get educated,
1:06:48 right?
1:06:48 So,
1:06:48 but Doctor Abedkar famously said,
1:06:50 what is the village but a den of ignorance and small-mindedness,
1:06:53 right?
1:06:53 So the framers of.
1:06:54 The Indian kind of state,
1:06:56 whether it's Nehru or Ambedkar,
1:06:57 believed in a modernizing bureaucracy,
1:07:00 right,
1:07:00 like a mean to be overlaid on a
1:07:02 patriarchal and feudal society and therefore bring modernity,
1:07:05 right,
1:07:05 like a mean to and accelerate that process,
1:07:08 and that's why
1:07:09 it over centralized and you see the importance of that even today,
1:07:11 OK,
1:07:12 because
1:07:13 ERA,
1:07:13 which is one of the biggest welfare programs in the world,
1:07:15 was staunchly resist.
1:07:17 Assisted by landlords,
1:07:18 right,
1:07:18 because
1:07:19 it is a government program that is threatening
1:07:21 the relative power of the local elites.
1:07:23 OK,
1:07:23 so
1:07:23 this belief that you need a modernizing bureaucracy to implement
1:07:27 the writ of a modernizing state over a traditional society
1:07:31 is what builds that,
1:07:32 and you can see the parallel in the US,
1:07:34 right?
1:07:34 So we think of the US as this very kind of,
1:07:36 you know,
1:07:37 local empowered federalism.
1:07:38 That's true,
1:07:39 right?
1:07:39 And because the US federalism is what we call
1:07:41 in kind of political science a coming together federalism,
1:07:44 right,
1:07:44 but any right that is not explicitly given to
1:07:46 the federal government stays with state and local,
1:07:48 right?
1:07:49 But the reason for the US Constitution being
1:07:51 that way is it's a very conservative constitution.
1:07:54 It's not trying to do any social reform,
1:07:56 OK,
1:07:56 so but the moment you have abolition and you
1:07:59 are trying to get universal franchise and civil rights,
1:08:02 you need the federal government to come in
1:08:03 with the Voting Rights Act that is enforced
1:08:05 by federal marshals because local elites in kind
1:08:09 of historic slave owning areas were not.
1:08:12 Allow the franchise,
1:08:14 right?
1:08:14 Like so
1:08:14 this is why you needed to centralize when you wanted to do social reform.
1:08:18 OK,
1:08:18 so then the question is where do we go from there now?
1:08:21 And part of my argument is 75 years after independence we've gone
1:08:24 from literacy rates of 20% to 80% and you've gone to much,
1:08:28 much better access to information.
1:08:29 OK,
1:08:30 so
1:08:30 today with smartphones,
1:08:32 I mean every person in the village,
1:08:33 so it becomes much harder,
1:08:34 OK,
1:08:35 so.
1:08:35 The point is therefore not that
1:08:37 you completely decentralize because that problem of
1:08:39 kind of local elite capture and power dynamics has not gone away,
1:08:42 but you can have a more nuanced decentralization,
1:08:45 right,
1:08:46 of that kind of puts more power with
1:08:48 local communities while having some checks and balances,
1:08:51 and that's why there's no silver bullet.
1:08:52 But the book again,
1:08:53 the chapter talks about certain specific ideas for how we can do that.
1:08:57 So excellent question.
1:08:59 Thank
1:09:00 you.
1:09:00 We,
1:09:01 we run out of time,
1:09:02 but I,
1:09:02 I want just to make a reference to your last point about the PMUs.
1:09:06 Actually,
1:09:07 uh around probably 1015 years ago,
1:09:10 we start moving away from that model.
1:09:12 Uh,
1:09:13 we start focusing on what we know now,
1:09:15 actually working on strengthening the country systems
1:09:17 and particularly in financial management and procurement.
1:09:21 That's,
1:09:21 we're in the right direction.
1:09:22 Excellent,
1:09:22 and I think you know,
1:09:23 but that's why the governance practice,
1:09:24 see,
1:09:24 it's not just that the governance practice needs to be stronger,
1:09:27 but it needs to imbue every aspect of the bank as a horizontal way of thinking,
1:09:31 right?
1:09:31 So more power to you.
1:09:33 Thank you.
1:09:33 OK,
1:09:34 so I guess we have a,
1:09:36 a short break,
1:09:36 right?
- add-style
- lp-body-content