Select text to annotate, Click play in YouTube to begin
00:00:00
[Music] foreign [Music]
00:00:18
[Music] I am delighted to to have you know this many people on the second day of a conference early in the
00:00:42
morning and a lot of young faces which makes me even happier um it's uh what my my brief here is to talk
00:00:54
to get into a bit the uh the innards of how jpal functions and the work we do and mostly what I want to tell you about is we have this um
00:01:07
we we we have this sometimes uh you know white might seem wild claim that we have
00:01:19
in some form or the other influence 600 million lives with programs that we've evaluated and more generally we I think uh try to be while we try to be
00:01:32
sober about it we want we want to sort of tell you that the so a little bit of about the scale of what we are doing now and why we are you know both proud and optimistic about where our
00:01:47
agenda is going with the goal especially in a room full of young people of getting you also excited about our work so let me let me uh this this graphic it
00:01:59
speaks for itself and it's sort of exactly you know you can see that in 2003 the dot was very very small and then it's it's become fatter and fatter and hopefully it's going to get even
00:02:12
fatter um the in the what feeds into that are randomized control trials that's sort of our bread and butter and and we have now
00:02:27
more than sixteen hundred of those across the world you can see this this graphic is uh lacks a bit the code in the sense that they're different colors
00:02:40
but I haven't told you what they are the colors are represent different areas of interventions there's Finance there is uh Health Care there is education there
00:02:55
is a direct social social support programs there is governance there is um [Music] study of you know interventions to
00:03:07
improve private markets business creation there's a enormous range of things that we do and that's those dots represent many of those and you can see that they're all
00:03:19
over the world they're clustered a little bit in South Asia East Africa and in the United States but nonetheless if you look and in France
00:03:32
um but if you look they are spread quite wide widely over all over all over the map and I I think to sort of but when we I
00:03:47
want to start by giving some content to the statement you know what does it mean to say 600 million people have been touched by programs evaluated by us what what that that sentence sounds great but
00:04:01
what does it mean actually so let me start with that um so this is this is probably one of our you know one of the things we keep coming back to many but this this is one
00:04:13
I think we are justifiably proud of uh Japan has uh been uh partnering an Indian NGO
00:04:26
come for nearly 20 years and the the idea of of the is of what this work with pratham constitutes is
00:04:39
sort of captured by the the title which is teaching at the right level and it's really a series of randomized control trials we say six on this slide but inside the six there were actually many
00:04:53
uh meaning at each time there were several interventions and in and if you take the the whole envelope of those they're more more like 12 or something and all of in all all of those the idea
00:05:08
was to test out a set of pedagogical interventions that pratham this wonderful Indian NGO that we've worked with uh
00:05:19
have have pioneered and the and at the base of that uh that is a experience that that a lot of people working in education had to start with which is
00:05:34
that while you know there were plausible things that you could do they didn't seem to have much effect on learning outcomes but on the other hand learning outcomes were rather disastrous uh just
00:05:47
in India uh about half of those in fifth grade cannot do a second grade mathematics so that's that's a fact that keeps coming back and then question is what do you do about it and what what's
00:06:01
what uh the the I think the series of of evaluations revealed was that uh you the curricular
00:06:14
often over ambitious and because they're over ambitious they kind of get to they they leave a lot of children behind very quickly and then there is no return
00:06:27
people follow the curriculum even if most children are not able to follow the curriculum the teachers keep following them and the net result is that children keep falling behind
00:06:39
and the Insight of teaching at the right level is exactly what it literally says which is that if you actually teach children what they need to learn and that requires changes in the pedagogy
00:06:52
you need to figure out what children need to learn and teach this to them but if you do it they learn very fast and that's that's the in a nutshell the idea of teaching at the right level it sounds banal
00:07:05
but alas it's actually been a long and I think somewhat uh daunting struggle to get that to be the uh something that's widely accepted and I think the process
00:07:19
of that uh in the next panel racial grannister who has been involved in much of that work will will speak about so I won't tell you much more than that but
00:07:31
it's it's uh one thing that it has meant is that we we have uh once we were persuaded that this was a was a a good
00:07:44
idea a good way to go we we started uh not we prathams started I mean we provided the evidence we've been armed with the reverence I think pratham persuaded a a series of
00:07:57
governments in Africa starting with Zambia uh to try to uh try to uh you know implement the same kinds of ideas that they had originally implemented in
00:08:11
India for us as part of our randomized control trial but then increasingly is a part of policies adopted by many of the Indian States and so we had after about 70 million
00:08:25
children had been reached by these programs based on the combination of pratham's push and some evidence that was backing them uh they they went
00:08:37
Global and there was um initial step which was to reach 1800 schools in Zambia which then uh I think impressed co-impact which is a donor
00:08:51
organization and with the help from co-impact and many others fcdo and many others uh there was a
00:09:03
an organization that was set up which is called tarl Africa which is committed to working with governments to scale these ideas of teaching at the right level inside government systems and that's I
00:09:18
think the word inside government systems is sounds again it rolls off my tongue but it's actually hides enormous amount of complexity and I'm going to let Rachel talk about that
00:09:30
when we come to the next slide the uh to take another exam so take a different example um what are other things that go go to
00:09:41
scale so here's a one that we worked on in Indonesia um there was a program and you might hear a little bit about our Indonesia work later in the day there's a program
00:09:53
which was the government was provides subsidies for rice to a large part of the population the the subsidies are the where the biggest part of the
00:10:06
government's social support program it was also true that when we measured it about 30 percent of the subsidies were reaching the beneficiaries what you do do about it turns out a bunch of obvious
00:10:19
self works surprisingly well one of them is giving people a card saying you are eligible for this program because people don't know that and then announcing the the list of people in public who
00:10:33
Etc so if you would do that package of giving people a card with what they're entitled to written on it it seems all of it matters it matters that you give your card at patterns that you've written on it but when you take that
00:10:47
package you can claw back almost half the missing subsidy so it was 30 percent you can get to close to 70 by having done that so it's it so when we showed that evidence the Indonesian government
00:10:59
actually decided to send the cards to all the beneficiaries and so that's a way as an example of the kind of cards that we had originally did a trial or
00:11:12
done a trial on was sent to uh the entire beneficiary population so that's that's an example of something that was uh another example of what goes into that
00:11:24
600 million number but that's not all we do part of what we try to do is more forward looking uh we identify where research is
00:11:36
lacking and try to try to move research in those areas we we're a network of researchers of
00:11:48
almost you know more than 700 researchers and you know in all over the world in many many universities LS not enough in the University of the South yet I'll talk about that a little bit
00:12:00
later but that that net being a network has many advantages uh we we collaborate we can we can try to help each other there's
00:12:13
not there's no hierarchy it has one downside which is people don't want to do research on something you can't get them to do it since it's voluntary but what we try to do to influence research is we try to raise funds to
00:12:26
these things called initiatives which we raise funds from various donors um many of them present today and then give that out uh through
00:12:38
competition so research competition to and that way try to influence so the the the point of having funds is that if you want to do research on this topic funds are available and so it it tries to
00:12:51
direct for research somewhat to to where we think the missing Gap gaps are so that's that's a forward-looking piece it's creating research in areas where we need research and uh sort of creating a
00:13:06
kind of a a a pipeline of oncoming research which then gets uh scaled to policy and then maybe can be uh can be scaled more globally so that's that's a
00:13:21
that's uh the we there's a a different model which is also part of the pipeline and that's that's the model of labs we've built labs in several countries there's a Egypt impact lab there is a a
00:13:36
mini do lab in Peru in the Ministry of Education there's a lab in Tamil Nadu in India which is a collaboration with the and in each case the idea is to have a
00:13:49
demand driven uh agenda so to have the to have the people interested in using the evidence to come to us and say this is the kind of evidence we would like
00:14:03
this is our so this is also a bit our work in Indonesia uh the government comes to us and says uh you know can you have something useful to say on this
00:14:15
particular point and then if you we try to uh try to design research based on that the the funding in these cases comes partly
00:14:27
from the government partly we fundraise but I and I I think the the great advantage of that is that you because you are sort of working closely with
00:14:39
with the government uh in a particular location you understand how the government functions better you understand their their mechanics their priorities their how how to get things
00:14:51
to work and what is easy and what is hard then that that helps the design of programs in ways that are hard to articulate but very important and and
00:15:03
therefore builds a pipeline uh one the the Egypt impact lab has been an extraordinary example of of this kind of research in Tamil Nadu where I have been
00:15:15
a little bit more involved there was a memorable random understand in 2014 which then connects to Tamil nadu's aim to be
00:15:30
poverty free by 2023 and part of that is we we have you know research that we have a steering committee that it constitutes mem of people members of
00:15:43
government but also some of us which sort of decides this is where the research is needed and moves the move the research in in tries to recruit researchers as I said
00:15:56
we don't tell people what to do when we try to encourage them to participate in the research and and create a body of knowledge that feeds directly into the policy because it has the advantage of
00:16:07
being demand driven and I think that through that I think we have both managed to create I think change their own understanding the government's own understanding of where
00:16:20
the problems are I give you an example um I think in India there's a presumption very strong presumption that the elderly are taken care of by the families and therefore they are fine and
00:16:35
they don't need other separate support one of the insights from the research with the government itself was surprised by was the fact that there were very large numbers of elderly more mainly
00:16:45
women living alone and they were both uh rather economically uh Under Pressure but also uh also
00:17:00
extremely depressed really extraordinarily depressed and I think that that's a fact that it was not known and but it once you know those kinds of facts you change policy because you start to see see things that you never
00:17:13
thought of and that's one of the an example of working closely through the Egyptian impact lab again the we we end up both and this is true in
00:17:27
Tamil Nadu as well we we create uh better pipelines for data inside government we one of the things that the government itself generates huge amounts of data but it it is not often the case
00:17:40
that the data is kept in a form that's necessarily you appropriable for research and one of the things we can we can do is create that and then work on various randomized control trials in
00:17:52
Egypt there's a lot of interest in labor market work and we've done a number of those again the idea is mostly to both both to both provide specific answers to
00:18:06
specific questions but also to provide uh build a culture of using evidence inside government so people get used to the fact that if you want to have a question there's a data a piece of data each is probably available in the
00:18:20
government in somewhere but you didn't look at it look at it let's try try to see the data and make what we can of it and that's that's one of the I think contributions of all these Labs uh
00:18:32
another place where we we feel that import it's a give very connected to what we do we want in the end a large number of a Workforce of
00:18:45
of researchers and people who work with JPL and JPL like organizations clients of of evidence and users of evidence and generators of evidence and those people
00:18:59
are uh there's not enough of them in the world and one of the things we have done is created this micro Masters program an online Masters program in data economics and development policy this has been
00:19:13
taken by you know courses have been taken by many many people you see the graphic uh 50 000 Learners have have participated in these online courses and from the online courses we actually
00:19:26
select those who are exceptional performers and try to try to attract them to come and finish their masters at MIT um
00:19:39
with a physical uh a one semester of physical work so a physical class attending classes so that it's a way to create a cohorts of uh really outstanding people with uh with a strong
00:19:54
background in in the kinds of work we do and related to that we're very interested in getting a widening the network in j-pal as I said we don't we
00:20:07
have less researchers outside the oecd and than we would like especially located working there and we've created a set of these programs for training
00:20:18
training people uh or giving them the opportunity to participate in our work of funding them trying to get create a cohort of people who are consumers but
00:20:32
also producers of evidence who will work in government but also perhaps do research and who will therefore create a broader culture of asking questions and answering them in in evidence-based ways
00:20:46
so this is uh and then we also do direct trainings on on on the kind of work we do they are tailored to various needs their and through that we have trained 20 000
00:20:59
people uh worldwide so this this is to give us a huge a sense of what we do um why we are excited about our our project maybe there's a little bit of vanity but nonetheless it's a it's for
00:21:14
us it's uh it's also useful to sometimes say these things to ourselves because most of the time we don't see it we we sort of live inside it and when you see these numbers I must say I feel starting from where we started 20 years
00:21:28
ago we've gone quite a way um thank you for listening to me and hopefully with your help we'll get there thank you [Music]
End of transcript