col-xs-12
col-sm-12
col-md-12
col-lg-12
col-xs-12
col-sm-12
col-md-12
col-lg-12
videoType
dynamic-media
videoDmUrl
https://delivery-p136806-e1377785.adobeaemcloud.com/adobe/assets/urn:aaid:aem:291d5eab-5347-4eb9-9520-84b862ad5305/play?assetname=KCP-Part-2+Human-Development.mp4
keyFrameImage
videoDescription
The Knowledge for Change Program’s 20th-anniversary celebration (“KCP20+”), will be implemented through a series of events. This event focuses on "20 Years of Research, Data, and Analytics on Poverty, Inequality and Human Capital."
timestamp

00:00 So the second session is gonna focus on,

00:02 on human capital and what we've learned over the years and how the,

00:06 how the conceptualization of

00:09 What human capital is and

00:12 um how it's how it's created and how public

00:14 policy can help create it has evolved over time.

00:17 And,

00:18 uh,

00:18 we're gonna have two,

00:19 presenters again,

00:21 uh,

00:21 Kathleen Beagle,

00:22 uh,

00:23 who's the Research Manager and lead,

00:24 and lead economist in the human development

00:26 team of the Development Research Group,

00:29 who's here in person,

00:30 and then we're gonna have Patrick Premont online,

00:33 um,

00:33 a senior economist,

00:34 uh,

00:34 in the DM.

00:36 Uh,

00:37 group,

00:37 the development impact evaluation group,

00:39 uh,

00:39 uh,

00:40 here at the bank.

00:41 Um,

00:41 so you're gonna start,

00:42 how,

00:42 how long are you gonna speak for?

00:43 Uh,

00:44 about 1010 minutes.

00:45 OK,

00:45 OK.

00:46 And I'm gonna be stricter this time.

00:49 Thanks.

00:54 This is on,

00:54 right?

00:55 Now it's on.

00:56 OK.

00:57 Good morning.

00:57 Thank you for coming to everyone online.

00:59 Thank you for attending.

01:01 Um,

01:01 I'm pleased to give the first half of

01:03 this presentation on behalf of myself and Patrick,

01:06 uh,

01:06 and then I'll hand it over to him,

01:08 who is,

01:08 uh,

01:09 uh,

01:10 currently not in DC,

01:11 so he'll join us by video.

01:13 OK.

01:14 So,

01:14 you know,

01:14 taking a very quick

01:17 step back and asking,

01:18 like,

01:18 where are we today,

01:19 before jumping into

01:21 what we've learned from research,

01:22 um,

01:24 you know,

01:24 we want to make a few kind of very quick observations.

01:26 Um,

01:28 and let me premise this by saying,

01:29 the topic of this talk,

01:31 I think like with the last talk,

01:32 is quite overwhelming and all encompassing.

01:35 So,

01:35 struggling to figure out what to prioritize in 25

01:38 minutes was a challenge for me and for Patrick.

01:41 Uh,

01:41 so hopefully we,

01:42 we hit the,

01:43 the big spots for you,

01:44 but I think

01:45 there are a lot of areas that will be nice to hear discussion

01:48 during the panel from the panelists because we don't cover nearly everything.

01:52 Um,

01:52 but first we want to kind of remark on

01:54 the incredible progress in education as measured by,

01:58 uh,

01:58 enrollment numbers in these figures,

02:00 but also reflected in attainment and other

02:02 measures of education progress.

02:04 Um,

02:05 immense progress in the last couple of decades,

02:08 uh,

02:08 a lot of catching.

02:09 Up by lagging regions,

02:10 um,

02:11 what we don't show you here today is also on the gender side that girls have caught up

02:15 in enrollment

02:17 and attainment and in fact now in middle and high income countries,

02:21 and most of them girls are,

02:22 uh,

02:23 exceeding boys in terms of educational attainment.

02:26 And so where we have lagging er gender gaps are in um

02:30 low income countries,

02:31 and in fact in many countries the discussion now

02:33 is how to address the boy disadvantage in education.

02:36 So a lot has changed in the last few decades.

02:40 Uh,

02:40 we've also seen a lot of progress in health.

02:42 Here we have figures on maternal mortality rates,

02:45 but also a comparison of,

02:47 uh,

02:47 looking at the

02:49 universal health coverage index and improvements in that,

02:53 as well as,

02:53 uh,

02:54 improvements in the rate of out of

02:55 pocket expenditure that households have to take on

02:59 to attain healthcare services.

03:03 Of course,

03:04 there are a lot of challenges still at play and some of

03:07 these will be reflected in the future graphs in this presentation.

03:10 Uh,

03:11 we know that learning is very different than enrollment.

03:14 So here we have some statistics on the fraction of primary students

03:19 who are meeting minimum proficiency thresholds in math.

03:24 By income country grouping

03:26 and what you'll notice is uh

03:29 lower income countries,

03:30 far right,

03:31 light blue,

03:32 have the lowest shares of primary students who are proficient in math

03:37 for the grade that they are attending,

03:39 and we will also return to this challenge in a moment on the measurement side.

03:44 And in health,

03:45 er,

03:46 we've seen an increase in

03:48 coverage and the provision of healthcare services down to the community level,

03:52 even.

03:53 But we know that in low-income countries,

03:55 in particular,

03:56 health centers lack structural capacity.

03:59 This is equipment,

04:01 machinery,

04:01 medicines

04:03 to provide quality service delivery.

04:05 And here we have results from

04:07 5 countries on the percent of facilities

04:10 with an item

04:11 in stock

04:12 on the day of the survey,

04:14 and you'll notice that in some countries,

04:15 the coverage rates are very low.

04:18 Uh,

04:18 but even in the best performing countries,

04:20 it's,

04:20 you know,

04:21 it can be on the order of,

04:22 uh,

04:22 25% of facilities who do not have

04:25 particular types of equipment

04:27 that are standard for these,

04:28 that,

04:28 that are standards for these health centers.

04:33 Uh,

04:33 and

04:34 very recently we've seen

04:36 COVID be a huge disruptor

04:38 both for education and health.

04:41 And here we just have two figures to,

04:43 uh,

04:44 display this fact.

04:46 On the left you have,

04:47 uh,

04:47 simulations of the decline in learning

04:51 that's taking place

04:52 due to school closures.

04:54 Uh,

04:54 and on the right we see,

04:56 uh,

04:56 tracking the utilization of services during the epidemic.

05:00 We see

05:01 big declines in some countries on the use of

05:04 maternal and child healthcare services

05:06 during,

05:07 ah,

05:07 the COVID pandemic.

05:10 In this presentation,

05:12 we want to talk,

05:13 we're gonna bundle it into,

05:14 uh,

05:14 we're broke it up into three parts.

05:16 First,

05:17 I'm gonna walk you through

05:18 features some innovations in measurements of human capital.

05:22 We talked a lot in the last presentation about the importance

05:25 of data and concepts of measurement and methods of data collection,

05:29 uh,

05:29 for poverty and inequality,

05:30 and we will reflect on that

05:32 as it pertains to human capital.

05:35 And some of the things that we've seen,

05:36 the trends we've seen in the last

05:38 10 to 15 years that have improved the data,

05:41 uh,

05:42 the data landscape in terms of concepts of related to measurement.

05:46 Uh,

05:46 Patrick will walk you through a select set

05:49 of evidence areas on policy interventions in area,

05:53 in areas related to human capital.

05:55 And then we want to kickstart a discussion.

05:57 We threw out some areas that we think

05:59 are research priority areas moving forward.

06:03 Uh,

06:03 one big caveat is the focus of this presentation is on

06:07 children and young people,

06:08 but human capital,

06:09 and I,

06:10 and I've,

06:10 uh,

06:11 borrowed this figure from colleagues in human development,

06:15 a forthcoming

06:16 COVID,

06:17 uh,

06:17 human capital report by the

06:19 Human Development chief economist,

06:20 Norbitcheid and his

06:22 fantastic colleagues.

06:23 But to demonstrate that human capital is a lifelong process,

06:27 so we could have as well

06:29 had a presentation that focused on working age adults

06:32 or older people,

06:33 but

06:33 because of time,

06:34 we're going to focus on

06:36 young,

06:36 young adult,

06:37 young children,

06:38 children and young.

06:40 Adolescents,

06:40 I would say.

06:42 OK,

06:43 let's talk about measurement.

06:44 Uh,

06:44 people that know me know I'm a real data nerd,

06:46 so,

06:47 you were not gonna hear,

06:49 not gonna not hear about measurement in this presentation.

06:52 So a few highlights on the measurement side.

06:54 We know

06:56 that

06:56 schooling is not the same as learning,

06:59 and we've seen

07:00 a huge transition,

07:01 I would say,

07:02 uh,

07:02 which is featured in our world development report on education 2020 led by

07:07 DM Filmer and others,

07:09 um,

07:09 in terms of thinking seriously

07:12 about moving away from.

07:14 Enrollment and attainment

07:15 as

07:17 the only measures of educational outcomes happening.

07:20 Here we have a figure on the concept of learning adjusted

07:23 years of schooling,

07:24 lays.

07:25 You can read a lot more about this in the World Development Report.

07:28 But the point is there when you adjust

07:32 years of schooling

07:33 by the quality of that schooling as reflected in test score data.

07:38 Uh,

07:39 learning adjusted years are much lower in some countries,

07:43 and there's a huge amount of variation as to how much lower they are.

07:46 So,

07:46 you,

07:47 and there are a lot of details underlying this calculation.

07:50 You can use different,

07:52 different benchmarks for this calculation,

07:53 but the bottom line is,

07:55 in many countries,

07:56 the quality of education

07:58 is translating into lower levels of learning

08:01 that may be hidden by just looking at attainment or enrollment.

08:05 Other measures of this include,

08:06 or other achievements in the measurement of

08:08 this include the harmonized learning outcomes uh work

08:12 uh that KCP has also supported.

08:15 A second sphere of a real change in how we think about measuring human capital

08:19 is an increased use of the concept of

08:22 social-emotional skills and measuring that when we do,

08:25 uh,

08:26 studies on human capital.

08:27 And here I'm,

08:29 Highlighting the step skills measurement surveys

08:32 that the bank kickstarted about 15,

08:34 maybe more than 15 years ago,

08:36 um,

08:37 which looks at various dimensions of

08:39 social-emotional skills on the non-cognitive side

08:42 of human capital development.

08:44 Um,

08:45 among them include,

08:46 for example,

08:46 concepts of grit,

08:48 decision making,

08:49 um,

08:49 big five measures.

08:51 And what you see on the right-hand side is

08:53 what was interesting about the SEP skills measurement program

08:56 is that it went forward in a number

08:58 of countries to try to consistently measure this concept

09:01 and look at how these measures of

09:03 socio-emotional skills relate to performance in school,

09:07 ultimate earning outcomes in the labor market.

09:10 There are,

09:11 there are

09:12 huge challenges with measuring social-emotional skills.

09:15 Among them are.

09:17 Uh,

09:18 so we're seeing a lot of different ways of measuring them,

09:21 but there's a lot of

09:23 problems with the validation of these,

09:26 these scales because they are based on Western populations,

09:29 and I'll give you one example from work that was KCP supported

09:32 that shows when you look at the relationship between a measure of

09:36 consciousness and um

09:38 outcomes,

09:39 predictive outcomes on income.

09:41 What looks very important in the US does not look important in many other countries,

09:45 drawing into question what this measure is actually capturing.

09:49 And then finally,

09:50 I want to talk about reflect on quality of service delivery,

09:53 um,

09:53 starting with the qualitative service delivery surveys that were

09:57 rolled out and then moved into service delivery indicator surveys

10:01 to really kind of capture,

10:03 uh,

10:03 quality of services

10:04 and,

10:05 uh,

10:06 understand constraints to quality of service delivery.

10:09 Here I have,

10:10 uh,

10:10 we're looking at work that looks at

10:12 comparing different ways of measuring teacher effectiveness.

10:15 This is work that Dion and others have produced.

10:18 We also have seen a shift to

10:19 using standardized patients and mystery clients to understand

10:24 uh how providers actually treat patients inside health clinics.

10:27 This is work from India on TB patients,

10:30 looking at the variation of correct reporting depending on the time of day,

10:35 using mystery clients,

10:36 mystery patients.

10:38 And then,

10:39 uh,

10:39 I wanted to also mention in the policy research report

10:42 produced by the,

10:44 by DC,

10:44 by my wonderful colleagues in DC,

10:46 the concept of effective coverage,

10:48 which is a combination of both,

10:50 did a,

10:50 does a patient need

10:52 healthcare?

10:53 Do they go to get healthcare and what is

10:55 the quality of the healthcare service that they get?

10:58 And uh,

10:59 that combination,

11:00 um,

11:01 the idea of

11:02 effective coverage is really focusing on this.

11:05 Case cascade between need,

11:08 coverage,

11:08 and quality.

11:11 And finally,

11:12 we can bring these ideas together into the concept

11:14 of the no do gap in health services.

11:17 So,

11:17 uh,

11:18 we,

11:18 we know that quality of service delivery might be right,

11:21 might be quite low,

11:22 but we often don't know is that constraint on the

11:25 structural side with what equipment is at the center,

11:28 on the knowledge side of the providers,

11:30 or on the translation of that knowledge into the service that they deliver.

11:34 So,

11:34 for example,

11:35 here,

11:35 I have one illustration of.

11:37 The no do gap from work in China,

11:39 colleagues did with KCP support

11:41 that compares what happens when you look at TB treatment

11:45 between what providers would tell you with a vignette,

11:48 a hypothetical case,

11:49 and what happens when you bring in a mystery,

11:52 a mystery patient into that office,

11:54 and big variations.

11:55 So,

11:55 we see in the first two rows,

11:57 in fact,

11:57 they do very well in the vignettes,

11:59 but the actual action at the clinic level is,

12:01 is a much lower performance rate.

12:03 Um,

12:04 I also point you to the policy research report I mentioned a moment ago

12:08 has extensive,

12:09 uh,

12:10 discussion on the node gap also.

12:12 OK,

12:13 now I'm gonna turn it over to Patrick,

12:15 but I'm gonna

12:16 control the slides.

12:17 So,

12:17 hi,

12:17 Patrick.

12:20 Hey,

12:20 Patrick,

12:20 um,

12:21 I can't show you this,

12:22 the,

12:22 the,

12:23 the sign that is 2 minutes.

12:24 I'm really gonna ask you to,

12:25 to monitor your own time if you don't mind and keep it to 15 minutes.

12:29 Thank you so much.

12:30 I'll do my best.

12:31 Uh,

12:31 thank you,

12:32 Kathleen,

12:32 and,

12:32 and thanks everyone for,

12:34 for being here.

12:35 Um,

12:35 so Kathleen has highlighted,

12:37 you know,

12:37 huge improvements in the measurement of human capital.

12:41 Um,

12:41 but together with that in the last 15 years,

12:43 we have seen really a boom,

12:45 uh,

12:46 um,

12:46 in studies looking at the effectiveness of policies to improve human capital,

12:51 uh,

12:51 using many of those same measures.

12:53 So in the next 15 minutes,

12:54 uh,

12:55 what I'll do is I'll share some insight from That research.

12:58 Um,

12:59 but the research for the KCP is really rich

13:01 and so as Kathleen mentioned at the beginning,

13:04 uh,

13:04 it's really impossible to give a comprehensive overview.

13:07 So we have chosen a select set of studies

13:09 to highlight the broad evolution of research and,

13:12 and we hope you find that helpful.

13:14 Uh,

13:14 in doing this,

13:15 we want to highlight the topics we are not gonna discuss,

13:18 uh,

13:18 related to,

13:19 uh,

13:19 early childhood development,

13:21 teacher accountability,

13:22 uh,

13:22 pedagogy,

13:23 or adolescent or reproductive health,

13:26 not because they are not important,

13:27 but just because we,

13:28 we have limited time.

13:29 But we hope we can maybe hear from discussions on this or,

13:31 or engage in the discussion on,

13:33 on broader,

13:34 uh,

13:34 uh,

13:34 on these broader topics.

13:36 Uh,

13:36 we've organized the second part of this presentation,

13:39 uh,

13:40 to,

13:40 um,

13:41 um,

13:41 um,

13:42 start with specific interventions for households on the demand side,

13:45 um,

13:46 and then slowly look at,

13:47 uh,

13:48 interventions on the supply side before moving to broader analysis of,

13:52 of reforms on systems and Markets,

13:54 uh,

13:55 which itself I think highlights a little bit the direction,

13:57 the trajectory where,

13:58 uh,

13:59 the policies have taken over the last,

14:00 uh,

14:00 20 years.

14:02 So,

14:02 uh,

14:02 with this in mind,

14:03 the first,

14:04 um,

14:04 kind of specific,

14:06 uh,

14:06 household level demand site intervention is,

14:08 is cash transfer that has received a lot of attention.

14:11 Um,

14:11 and specifically the PRR reports,

14:14 um,

14:14 supported by KCP on CCTs about 15 years ago was really

14:18 influential in

14:20 laying out the theoretical underpinnings of CCTs

14:23 but also summarizing early impact evidence from,

14:26 uh,

14:26 uh,

14:27 Latin America and the first generation of CCTs.

14:30 Um,

14:30 I think we have learned a lot as CCTs are

14:32 expanding around the world on how to optimize their design

14:35 with a few key insights from the research.

14:38 Uh,

14:38 the first key insight is that

14:40 CCTs can really improve the outcomes on which,

14:43 uh,

14:43 they are conditioned.

14:45 Uh,

14:45 we have seen a lot of this,

14:46 uh,

14:46 on school enrollment and preventive health,

14:49 but also that has been used in other,

14:51 uh,

14:51 domains,

14:52 for instance,

14:53 trying to incentivize,

14:54 um,

14:55 uh,

14:55 safer sexual behaviors to reduce STIs,

14:58 um,

14:58 among other outcomes.

15:00 But the literature also has highlighted trade-offs,

15:03 uh,

15:03 and you can see here on the right,

15:04 a graph that shows that in Malawi,

15:07 uh,

15:07 a CCT was pretty effective at increasing enrollment,

15:11 but that an unconditional cash transfer was more effective

15:14 at delaying the age of marriage.

15:16 And so in a way,

15:18 um,

15:18 when we face this trade-off,

15:20 uh,

15:20 we may,

15:20 or to avoid those trade-offs,

15:22 we may need to combine

15:23 different types of cash transfer modalities in order to

15:26 achieve impact on multiple dimensions of human capital.

15:31 Now,

15:31 the literature uh has also consistently shown

15:34 that there are financial constraints to access,

15:37 um,

15:37 uh,

15:37 in multiple,

15:38 uh,

15:39 uh,

15:39 for schooling but also for,

15:40 uh,

15:41 access to health services.

15:43 Another example of that is in the WDR on learning which Kathleen already mentioned,

15:47 um.

15:48 Um,

15:48 showing that as,

15:49 uh,

15:49 user fees or school fees are removed,

15:52 uh,

15:52 enrollment increased quite substantially,

15:54 um,

15:55 also pointing in the same direction as the literature on scholarship scheme,

15:59 you know,

15:59 pointing to the fact that,

16:01 uh,

16:01 uh,

16:01 demand for services is,

16:02 is sensitive to,

16:03 to,

16:04 to cost.

16:05 But of course,

16:05 we know that addressing financial constraints is not sufficient,

16:08 it's not sufficient for the most vulnerable kids,

16:11 uh,

16:11 and households,

16:12 and also it's not sufficient because it doesn't

16:14 lead to improvements necessarily in learning.

16:18 One of the intervention that has been presented as

16:21 potentially improving both access and learning is school feeding.

16:25 Um,

16:26 there,

16:26 uh,

16:27 recent meta-analysis have shown that school meals can

16:30 increase health and nutrition status,

16:32 in particular,

16:33 if they are complemented by,

16:35 uh,

16:35 micronutrients or,

16:37 uh,

16:37 deworming interventions.

16:39 Um,

16:39 and interestingly,

16:41 also shown that,

16:42 um,

16:42 they can improve learning.

16:44 You see on the right here a graph showing,

16:46 um,

16:46 effect on,

16:47 on learning

16:48 of,

16:48 uh,

16:49 duration of exposure to school,

16:50 school meals in India showing a,

16:52 a substantial positive effects.

16:54 Um,

16:55 but behind this,

16:56 uh,

16:56 there is the research also highlighted the need for,

16:59 uh,

16:59 complementary school inputs in order to achieve these learning ga gains.

17:03 So in a way,

17:04 um,

17:05 in both those dimensions,

17:06 we have seen,

17:07 uh,

17:08 the potential role of complementarities

17:10 being important in human capital formation

17:13 and,

17:13 uh,

17:14 we have ongoing work at DAN in collaboration with the World Food Program,

17:17 uh,

17:17 to try to look at these,

17:19 uh,

17:19 synergies and complementarities in school feeding

17:21 a little bit more consistently.

17:24 In the spirit of thinking about complementarities,

17:27 another key result is that providing inputs alone

17:31 um is not sufficient to improve learning.

17:34 Um,

17:34 and here,

17:35 um,

17:36 results from studies

17:37 on providing textbook or on,

17:40 uh,

17:40 you know,

17:40 providing one laptop per child have been pretty clear in terms of the,

17:43 the limited effects on earnings,

17:45 on learning,

17:46 sorry.

17:47 Um,

17:47 and another example is a KCP study in,

17:49 in Lagos that shows that

17:51 Providing e-readers does not necessarily improve uh um uh learning

17:57 unless um it is,

17:59 um,

17:59 it includes,

18:00 those e-readers includes material from school curriculum

18:04 or compensate for lack of inputs in this case,

18:07 textbooks,

18:08 uh,

18:08 in school.

18:10 There is,

18:10 however,

18:11 uh,

18:11 recent evidence that's a little bit more encouraging on the use of,

18:14 um,

18:15 of technology and,

18:16 um,

18:16 uh,

18:17 uh,

18:17 to,

18:18 to improve learning,

18:19 in particular when software or edutainment,

18:22 uh,

18:22 is designed to uh really be engaging for children and teach,

18:27 uh,

18:27 uh,

18:27 at the level.

18:29 That,

18:29 that they need.

18:30 And so here we have some recent results from

18:32 northern Nigeria that are pretty powerful showing that combining

18:36 aspirational videos for parents

18:39 with literacy apps for um on smartphones for

18:44 children can really boost learning pretty substantially.

18:49 Now,

18:49 moving on to schools,

18:51 um,

18:51 there's been a,

18:52 a huge amount of progress,

18:53 uh,

18:54 over the last decades,

18:55 um,

18:56 due to many of you in the room here that have worked on this topic a lot,

18:59 uh,

18:59 to highlight what are the types of policies that are effective to boost learning

19:04 in school.

19:05 Um,

19:05 Kathleen,

19:06 highlighted the learning crisis among children,

19:09 um,

19:10 but another really striking,

19:11 uh,

19:11 finding is that

19:13 teachers themselves often do not master,

19:15 uh,

19:16 the subjects that they teach.

19:18 Um,

19:18 and you can see here on the,

19:20 the black line on the,

19:21 on the graph shows you.

19:23 Um,

19:23 the effects on,

19:24 uh,

19:25 learning,

19:25 uh,

19:26 among children

19:27 of,

19:27 um,

19:28 um,

19:28 ensuring that teachers that's,

19:30 uh,

19:31 at school

19:31 have themselves,

19:32 uh,

19:33 acquired the level of learning equivalent to the end of primary school,

19:36 and that in itself could,

19:38 uh,

19:38 substantially improve learning

19:40 even more when this is combined,

19:42 uh,

19:42 looking at the dashed line here with

19:44 an increase in time

19:46 spent by teachers teaching.

19:48 Um,

19:49 now,

19:49 that these are two levers,

19:50 uh,

19:51 to,

19:51 to improve,

19:52 uh,

19:52 learning.

19:53 Of course,

19:54 there are others.

19:55 Um,

19:55 we know that,

19:56 uh,

19:56 many teacher training program

19:58 or professional development programs are very far from being optimal

20:03 and that improving this along with,

20:05 uh.

20:06 Pedagogic uh sorry,

20:06 pedagogical approaches,

20:08 um,

20:09 that includes,

20:11 uh,

20:11 structured lesson plans or,

20:13 uh,

20:13 teaching at the child level,

20:15 um,

20:15 can be really,

20:16 uh,

20:17 great investments in boosting learning,

20:18 and this is something maybe,

20:20 uh,

20:20 Rukmini in the discussion may,

20:21 may touch on.

20:24 Um,

20:24 so as mentioned,

20:25 the,

20:25 the time spent by teachers teaching,

20:27 um,

20:27 is low in,

20:28 in many developing countries,

20:30 um,

20:31 but there are really multiple reasons for that.

20:33 Um,

20:33 and I think interestingly,

20:35 when

20:35 teachers are asked,

20:36 um,

20:37 if they think it's OK to be absent,

20:39 many of them here on the left say that they,

20:41 you know,

20:41 it's OK to be absent if the kids in school have

20:44 something to do in their absence or on the right,

20:46 if,

20:47 uh,

20:47 they are absent to do something helpful in their community.

20:50 And so that's kind of um insightful because it shows us that

20:54 um it's,

20:55 you know,

20:55 it's hard to shift

20:57 uh teachers' efforts,

20:58 um,

20:58 incentives,

20:59 financial incentives may help but

21:01 um it may only be part of the solution and other approaches to

21:05 improve accountability,

21:06 uh,

21:07 for instance,

21:07 to scorecard or other interventions may be needed and

21:10 uh again,

21:11 this is something that

21:12 uh Ritva may have a lot more to say,

21:14 uh,

21:14 in the discussion.

21:16 Um,

21:17 related to,

21:18 uh,

21:18 the question of incentives,

21:20 um,

21:21 many,

21:21 uh,

21:22 policies or,

21:22 or interventions have tried to provide grants

21:25 to schools to improve their performance,

21:28 um,

21:28 and we can illustrate some of the opportunities and challenges with

21:31 that from using the data from a study from Indonesia.

21:35 That shows here on the right that um

21:38 performance-based grants to school

21:41 improve student learning in secondary schools,

21:43 so this is column 4 here,

21:45 but uh actually the same grants do not improve

21:49 learning in primary schools and this is the column 2 here on the left.

21:54 Um,

21:55 so this is kind of interesting because it illustrates that,

21:57 um,

21:58 uh,

21:59 the same policy,

22:00 the same type of brand designed a similar way

22:02 in a similar setting actually lead to very different behavioral responses

22:06 in different types of school.

22:08 Um,

22:09 and so kind of illustrates the need to understand the,

22:12 the behavioral response from,

22:14 uh,

22:14 providers when those kind of,

22:16 uh,

22:16 grants are,

22:16 are,

22:17 are,

22:17 are offered

22:18 and kind of ties to a,

22:20 a broader literature with mixed results on the performance incentives.

22:26 Another set of intervention that's uh is trying to,

22:29 to shift behavior

22:30 um is uh the provision of information related to schooling.

22:34 Um,

22:35 and here we are showing you some results from a study in Mozambique,

22:38 uh,

22:39 that shows that providing information to parents about

22:42 your children's attendance in school,

22:44 uh,

22:45 not only increases attendance itself but also

22:48 uh increases learning in this case,

22:50 uh,

22:50 math scores.

22:51 And what's interesting is the mechanism is that it's through parents' behavior,

22:55 through parents basically getting more involved

22:57 in the monitoring of their children,

22:59 uh,

22:59 which in itself then leads to,

23:01 uh,

23:01 them learning,

23:03 uh,

23:03 more in school.

23:04 Uh,

23:04 and similar pathways have been highlighted in other studies,

23:07 for instance,

23:08 in,

23:08 in Angola.

23:10 Of course,

23:10 the provision of information can also

23:12 have broader effects

23:14 on education markets,

23:15 um,

23:16 and also induce,

23:17 uh,

23:18 behavioral responses from,

23:19 uh,

23:19 from providers and so for instance,

23:21 in Pakistan,

23:22 a very influential study showing that,

23:25 uh,

23:25 providing information on the,

23:26 uh,

23:27 on average test scores in school actually leads private providers to reduce fees,

23:32 uh,

23:32 because they don't have to signal

23:33 their quality to,

23:34 uh,

23:35 to fees anymore.

23:37 Um,

23:38 the,

23:39 of course,

23:39 the,

23:39 the analysis of,

23:40 of private school has,

23:42 um,

23:42 um,

23:43 taken a lot of attention in recent years trying to see

23:46 if they can be part of the solution to improve learning.

23:49 Um,

23:49 it has been the case in certain areas.

23:51 Here we show you in Uganda that's,

23:53 uh,

23:53 providing vouchers for students to,

23:56 for,

23:56 for private schools,

23:57 uh,

23:58 actually improve student score pretty consistently.

24:01 Um,

24:02 uh,

24:02 you see the distribution shifting,

24:04 um,

24:04 but behind this,

24:05 there are,

24:06 uh,

24:06 part of this is due to gains in enrollments partly driven from,

24:10 uh,

24:11 students with higher socio-economic backgrounds.

24:13 So,

24:13 of course,

24:14 um,

24:15 some distributional questions here on the,

24:17 uh,

24:17 the effects of those policies and who they might benefit to,

24:20 uh,

24:20 at the end of the day.

24:22 Um,

24:23 I will

24:24 wrap up with two more,

24:26 uh,

24:26 topics.

24:27 Uh,

24:27 the next one being on,

24:28 on health.

24:29 Um,

24:30 Kathlin already mentioned the,

24:31 uh,

24:32 uh,

24:32 very exciting PRR that was,

24:34 uh,

24:35 um,

24:35 issued,

24:36 uh,

24:36 a few months ago

24:37 that analyzes how changes in incentives,

24:40 uh,

24:40 introduced to system-wide reforms affect service quality and outcome.

24:45 And then specifically looking at pay for performance,

24:48 uh,

24:48 based on the quantity or the quality of services delivered.

24:51 Um,

24:52 what's interesting is that,

24:53 uh,

24:53 I remember many years ago we were all very exciting seeing,

24:56 seeing the first results from Rwanda on

24:58 performance incentives that were pretty positive.

25:01 Um,

25:02 but the PRR actually updates,

25:04 uh,

25:04 our overall thinking on this showing that,

25:06 um,

25:07 performance pay

25:08 actually has little,

25:09 little impact in relatively under-sourced health systems,

25:13 and this is because,

25:15 um,

25:15 there's a lot of,

25:16 uh,

25:17 uh,

25:17 issues in,

25:18 in health systems that are beyond the control of the frontline providers

25:21 that may not be able to control or,

25:23 or influence,

25:24 uh,

25:24 issues related to the supply chain,

25:26 to the quality of infrastructure or equipment,

25:29 uh.

25:29 Or even demand side constraints,

25:31 uh,

25:31 which will be needed for the incentives to be,

25:33 uh,

25:34 effective and so

25:35 what the PRR illustrates and the graph here on the right illustrates is that,

25:39 uh,

25:40 direct facility financing,

25:42 um,

25:42 with autonomy and,

25:43 and,

25:44 and accountability may actually be just as,

25:46 as effective,

25:47 uh,

25:47 uh,

25:48 in terms of,

25:48 um,

25:49 uh,

25:49 uh,

25:50 achieving gains in outcome,

25:52 uh,

25:52 but doing so at lower cost and,

25:54 uh,

25:55 with a relatively easier implementation.

25:59 So the,

25:59 I'll conclude my uh very brief overview on uh uh selected uh topics on human capital

26:05 by uh talking a little bit about skills for schools,

26:09 uh,

26:09 for youth uh that have exited schools.

26:11 Um,

26:12 we know that the evidence on training programs and in program,

26:14 employment programs is,

26:15 is very mixed,

26:17 um,

26:17 but I want to highlight some,

26:19 um,

26:19 innovation

26:20 that have been partly supported by the KCP that,

26:23 uh,

26:23 shows some,

26:24 some promise,

26:25 uh,

26:25 uh,

26:25 that have shown promise in recent years.

26:27 So one example is

26:29 interventions,

26:30 uh,

26:30 behavioral intervention.

26:31 trying to target

26:32 uh um behavioral skills,

26:34 in this case,

26:35 personal initiative

26:36 tying to Katlin's earlier discussion of,

26:38 of,

26:39 of,

26:39 uh,

26:40 of those behavioral skills,

26:41 um,

26:42 showing that this may be more effective

26:43 than interventions trying to teach traditional business training

26:47 in terms of improving earnings,

26:48 uh,

26:49 for microenterprises.

26:51 Similarly,

26:52 uh,

26:52 evidence,

26:53 recent evidence showing that interventions combining on the job training with

26:57 training in vocational training centers in Colombia or through

27:00 dual apprenticeship in Cote d'Ivoire

27:02 can be very effective in improving skills and then raising productivity

27:06 uh of young people when they enter the workforce.

27:09 Um,

27:10 but beyond all of this,

27:11 uh,

27:12 also in this topic,

27:13 I think we have learned that we need to think carefully about the trading markets,

27:17 uh,

27:17 about the incentives for training provision in firms or,

27:22 uh,

27:22 in public,

27:23 uh,

27:23 in the public sector,

27:24 uh,

27:25 the scope to improve demand,

27:26 and overall,

27:27 I think as in many of the cases we just quickly discussed,

27:30 um,

27:31 that these policies cannot just be assessed,

27:33 uh,

27:33 in

27:34 partial equilibrium.

27:36 So,

27:37 um,

27:38 uh,

27:38 I think we're almost at the end of the time.

27:40 I think,

27:41 I don't see you beyond waving yet,

27:42 but,

27:43 uh,

27:43 you're probably close.

27:44 Uh,

27:44 I just want to kickstart the discussion by highlighting a few,

27:48 uh,

27:49 uh,

27:49 topics that we think are important for future research and see,

27:52 uh,

27:52 hear a little bit from you,

27:54 uh,

27:54 whether those are the right priorities or there are others that you think are more,

27:57 uh,

27:58 uh,

27:58 higher order.

27:59 Um,

28:00 one of the big questions or the big trajectory we see on the policy side is

28:04 towards more integrated or multifaceted interventions,

28:08 um,

28:09 um,

28:10 bundling,

28:11 uh,

28:11 various types of,

28:12 of components including

28:14 packaging intervention.

28:15 of the demand on the supply side together to improve outcomes.

28:18 So this is very exciting but it also raises a lot of questions on how to optimize those

28:23 uh multi multifaceted intervention

28:25 and how to uh tailor them across different settings and,

28:28 and,

28:29 and target populations.

28:31 Related to that,

28:32 uh,

28:33 of course,

28:33 we need a better understanding of health and education markets.

28:36 We've made lots of progress over the last 10 years,

28:38 but,

28:39 uh,

28:39 as,

28:40 um,

28:40 uh,

28:40 the interaction between public interventions and private providers,

28:43 um,

28:44 is key to understanding overall how,

28:46 um,

28:47 behavioral shifts and,

28:48 and how,

28:49 uh,

28:49 changes in outcome,

28:50 um,

28:51 uh,

28:51 pan out.

28:52 Uh,

28:53 especially as we move more and more towards system-wide policies and,

28:56 uh,

28:57 as we are pushed,

28:58 uh,

28:58 rightly so at the bank to think about scale and how,

29:01 uh,

29:01 impacts,

29:02 uh,

29:02 uh,

29:02 kind of aggregates to

29:04 achieve higher outer policy goal at the,

29:06 at the economy level.

29:09 Of course,

29:09 um,

29:10 the last few years have,

29:11 um,

29:12 showed us that,

29:12 uh,

29:13 you know,

29:13 uh,

29:14 we've faced major challenges in terms of human capital,

29:16 in terms of COVID,

29:18 and are also facing new challenges related to

29:20 climate and understanding how those are affecting,

29:24 uh,

29:24 both human capital,

29:25 uh,

29:26 um,

29:26 accumulation but also how,

29:28 uh,

29:28 policies to address them also then

29:31 Uh,

29:31 improve human capital and intersect with the human capital formation

29:35 will be critical moving forward

29:37 as,

29:37 uh,

29:38 is the,

29:38 the focus on technology or the hope that with technology,

29:41 we can,

29:42 uh,

29:42 achieve better outcome

29:44 by facilitating the work,

29:45 facilitating the work of providers as well as,

29:47 uh,

29:48 facilitating behavioral change among,

29:50 uh,

29:51 among,

29:51 um,

29:51 um,

29:52 uh,

29:52 users.

29:53 The last thing I will,

29:54 I will say today is that um as you've seen,

29:57 there's a huge amount of work going on on both on the data and on the evidence side

30:02 and given this boom,

30:03 I think one of the challenge we face is really being

30:06 able to synthesize effectively this evidence for policymakers so that they can

30:10 use this for policy actions and I think

30:12 uh beyond the generation,

30:13 this

30:15 uh synthesis of,

30:15 of evidence moving forward will be critical to

30:18 uh have even more policy impact for,

30:20 with the work that we do.

30:21 Thank you so much.

30:25 Thanks Patrick,

30:26 and thanks Kathleen as well for those

30:29 really

30:30 Interesting presentations.

30:31 I mean,

30:31 I,

30:32 I,

30:32 I,

30:32 I do understand that,

30:33 the,

30:34 the,

30:34 the,

30:34 the

30:35 selection process

30:37 for,

30:37 for,

30:37 for what to highlight was pretty drastic,

30:40 but um I think we can all agree it's a really rich

30:42 agenda,

30:43 and I think this gave us a really good flavor of the kind of work.

30:45 Uh,

30:45 we have,

30:46 uh,

30:46 that kind of work that's going on.

30:48 Um,

30:49 we have a really interesting,

30:50 uh,

30:50 uh,

30:51 and impressive uh set of panelists uh with us.

30:55 Uh,

30:55 we have one person in the room who's Emily Gustafson Wright,

30:58 who's a senior fellow at the Center for Universal Education,

31:02 Global Economy,

31:03 uh,

31:03 and Development at the Brookings Institution.

31:06 Uh,

31:06 we have Pamela Jacquiela,

31:08 who's associate Professor of Economics at Williams College.

31:12 We have Rukmini Banerjee,

31:13 who's Chief Executive Officer of Pratham Education Foundation.

31:17 And then we have Rick Vereinke,

31:18 professor of Practice at Aalto University,

31:21 Helsinki.

31:22 Um,

31:22 I'm gonna go in the order that it was given to me,

31:25 uh,

31:26 just to,

31:27 it's somewhat random,

31:29 uh,

31:30 which means we're gonna start with Pam.

31:31 Uh,

31:32 Pam,

31:32 are you online and can you come in?

31:37 Yes,

31:37 I am online.

31:39 Great,

31:40 we can see you and we can hear you.

31:41 Uh,

31:42 thanks,

31:42 go ahead.

31:43 Fantastic.

31:44 Thank you very much.

31:45 Um,

31:46 so thank you for having me.

31:47 I'm delighted to see so many familiar faces,

31:50 uh,

31:50 online.

31:51 Um,

31:52 so I'm going to comment a bit on

31:55 the educational side of human capital because

31:57 I feel woefully underqualified to talk,

31:59 to speak to the health and healthcare side.

32:02 So the talk we just heard did a great job of emphasizing the

32:05 tremendous progress in education that we've seen over the last 25 and even 50

32:10 years.

32:11 When we set out the Millennium Development

32:13 Goal of achieving universal primary education,

32:16 um,

32:17 20 years ago,

32:18 uh,

32:19 we've now achieved that.

32:21 We've come near to achieving that goal.

32:22 It's not no longer,

32:23 it's not,

32:24 uh,

32:24 universal,

32:25 but

32:25 the international community has made,

32:27 made remarkable progress,

32:29 and,

32:29 uh.

32:31 We've now shifted toward thinking about,

32:33 uh,

32:34 as access has become less of a constraint,

32:35 we've shifted toward thinking about learning,

32:38 uh,

32:38 and in particular,

32:39 uh,

32:40 the

32:41 The presentation we just saw was emphasizing

32:43 policies to improve teacher performance in the classroom,

32:46 technological and pedagogical innovations,

32:49 uh,

32:49 to increase the likelihood,

32:51 uh,

32:51 that

32:52 access to schools and effort by teachers translates into learning,

32:55 and

32:56 there's also a broad set of policies

32:58 on early childhood education and interventions to

33:01 help,

33:01 uh,

33:02 prepare students to be ready when they enter the classroom.

33:05 These policies are all extremely good and constructive,

33:08 and I'm not here to criticize,

33:09 uh,

33:09 to criticize them.

33:12 And

33:12 there,

33:13 as Kathleen and Patrick rightly point out,

33:14 there are a lot of other policies that they didn't have time to,

33:17 to discuss in the,

33:18 in related domains that have been

33:20 really successful over the years.

33:21 But the task in these sorts of situations is to have an interesting take,

33:25 so I'm going to

33:26 take a stab at that.

33:28 Um,

33:28 and so I'm gonna start by sort of framing why

33:31 we've shifted from access to school readiness and learning,

33:35 um,

33:36 in recent years,

33:37 and that's because,

33:38 you know,

33:38 when we,

33:39 um,

33:41 Started to succeed with access,

33:44 uh,

33:44 and come near to achieving universal primary education and higher enrollment,

33:49 uh,

33:49 in secondary education,

33:51 uh,

33:51 we saw that access alone was not enough,

33:53 you know,

33:53 so when Nigeria and Indonesia were building schools in the 1970s,

33:58 when

33:58 many,

33:59 uh,

33:59 African countries We're eliminating primary school fees

34:02 in the 1990s and the early 2000s.

34:05 The idea was that access

34:07 would,

34:07 would be enough,

34:08 would translate into

34:09 the building of a skilled workforce,

34:12 uh,

34:12 the skilled workforce that countries need to fully participate,

34:15 uh,

34:15 in the modern economy and achieve inclusive growth,

34:18 but it wasn't enough.

34:19 And,

34:20 you know,

34:21 this shouldn't be surprising.

34:22 It's not much of an exaggeration to

34:24 say that the whole narrative of post-war development

34:27 is a series of

34:28 uh attempts to try things,

34:30 and then

34:31 learning that they're not enough.

34:33 Um,

34:34 and so,

34:34 you know,

34:34 it's a little bit like we're that Far Side cartoon where we have a

34:37 physicist is at the blackboard and you have some equations on their left,

34:41 equations on their right,

34:42 and in the middle,

34:43 they,

34:43 they've written,

34:43 and then a miracle occurs,

34:44 you know,

34:45 we try policies and then we hope a growth miracle occurs.

34:49 Um,

34:50 and so we've turned a lot of attention

34:51 to learning and early childhood and school readiness,

34:54 and we're hoping that now this focus on human

34:57 capital rather than access to education is going to achieve

35:01 the growth and prosperity miracle that we are all hoping for.

35:05 And

35:06 the suggestion I want to make today,

35:07 and it's clearly not

35:08 a new idea.

35:09 Particularly my idea

35:11 is that

35:12 everything we're doing focusing on the building of human capital is good,

35:17 but it's really unlikely to be enough,

35:19 um,

35:19 and it may lead to,

35:21 you know,

35:21 significant impacts that are even

35:23 economically and developmentally meaningful,

35:25 but it's not going to,

35:27 to get us as far as we need to go.

35:29 And a big reason for this,

35:31 and this is something that Patrick kind of brought up at the very end,

35:34 is that there's a tendency to focus on the supply side and not on the demand side.

35:39 Um,

35:40 and so I just want to talk a bit about,

35:42 um,

35:43 how we need to frame our entire discussion of human capital,

35:47 not only in terms of that supply side,

35:49 but in terms of political economy

35:51 constraints and social constraints that are going to limit and are currently

35:55 limiting both the incentives for individuals

35:58 and households to invest in human capital

36:00 and the relationship between

36:02 human capital that is built,

36:04 uh,

36:04 in the classroom and in youth,

36:06 and how that translates into

36:08 Outcomes in the economy,

36:09 participation in the economy,

36:11 and subsequently growth and shared prosperity.

36:14 And so I only have a couple of minutes.

36:16 I'm gonna try to do this quickly,

36:17 and I'm gonna do it through a gender lens because

36:19 I'm always looking at everything through a gender lens.

36:22 Um,

36:23 and so I want to make two points.

36:25 And so the first is that

36:26 over the last 25 years,

36:27 we've seen a major push

36:29 to increase girls' educational attainment,

36:31 and Kathleen talked about how now we're actually

36:33 thinking about how boys are lagging behind.

36:35 We've been remarkably successful.

36:37 Um,

36:38 girls' enrollment in primary and secondary schools have increased dramatically,

36:42 both in absolute terms and relative to boys.

36:44 And of course we know schooling isn't learning,

36:47 uh,

36:48 but it is correlated with learning.

36:50 So sometimes we can beat the schooling isn't learning drum a little bit too hard.

36:53 It is correlated with learning,

36:55 and we've also seen big increases in female literacy

36:58 and enrollment in tertiary education,

37:00 which requires some measure of human capital.

37:02 And so all the evidence suggests that both in absolute terms and in relative terms,

37:07 We've seen pretty substantial increases in women's

37:10 human capital over the last 25 years,

37:13 and that's in absolute terms and relative to men.

37:16 And so we can think about

37:17 what has happened with women as a little bit of a case study or

37:21 quasi-experiment and the broad impacts of increasing

37:25 human capital for a group of people.

37:28 And the bad news is that over the same time period,

37:31 we've seen remarkably little progress in

37:33 terms of women's labor force participation

37:36 or their ability to achieve fulfilling careers in many countries and contexts.

37:41 And that's a real problem for

37:43 growth and development.

37:44 So economists estimate that in the United States,

37:47 between 20 and 40%

37:49 of US growth over the last 50 years

37:51 has come about because women and minorities who were excluded

37:55 from high-skilled professions have been able to enter them.

37:58 And there's no reason to think that that type of misallocation

38:00 isn't just as big of a problem in developing countries.

38:03 Um,

38:04 but the problem is that as we've seen these big increases in women's human capital,

38:09 we've seen they are not predictive of women getting to enter the labor force,

38:13 increases in labor force participation.

38:16 Um,

38:16 and so that brings me to my second point,

38:18 which is that when we're thinking about human capital,

38:20 and

38:21 in particular thinking about investments in human capital that we hope

38:24 we're going to translate into impacts on the broader economy,

38:27 we really need to emphasize

38:28 not just the learning outcomes,

38:30 but also,

38:31 uh,

38:32 thinking about building

38:34 data,

38:34 building knowledge

38:36 about the demand side constraints,

38:38 and I'm gonna look at this through the gender.

38:40 as well,

38:41 because for women and girls,

38:42 we know

38:43 that at every stage of the process,

38:45 even though women and girls are in school now,

38:47 as much as boys in many places,

38:49 girls are facing,

38:51 uh,

38:51 these additional challenges that prevent them,

38:53 that reduce their incentives to invest in their human capital,

38:57 uh,

38:57 and it reduced the extent to which that

38:59 translates into their full participation in the economy.

39:02 So

39:02 girls face an

39:04 additional burden.

39:05 burden of domestic care work in the home.

39:07 They face discrimination from teachers

39:09 who nudge them into lower skill tracks and professions.

39:13 Uh,

39:14 they,

39:14 we know that they distort their educational choices to try

39:17 to avoid risks of street harassment and gender-based violence,

39:22 uh,

39:22 that they face a lower expected return to investing in their own education because

39:25 they might get kicked out of school in many contexts if they get pregnant.

39:29 And then of course,

39:30 they worry about the marriage market consequences,

39:32 and the incentives to invest in education

39:34 are reduced when there aren't complementary,

39:37 uh,

39:37 both,

39:38 you know,

39:38 access to child care,

39:39 things that allow them to

39:41 balance their family responsibilities with fulfilling careers in adulthood,

39:45 and also social constraints from their families and their spouses

39:48 that limit their ability to really have meaningful careers,

39:52 if those might come at some cost to their household work.

39:55 So none of these things is new to anybody in this room,

39:57 but I really wanted to sort of nudge us to when we talk about human capital,

40:02 not just think about what we're doing on the supply side,

40:06 but really think about the need to build knowledge on

40:10 the demand side and in particular how households and

40:13 individuals are perceiving the choices and the trade-offs they make

40:16 and the constraints that they face.

40:18 And of course,

40:19 the,

40:19 you know,

40:19 the World Bank has been a leader in

40:22 Not just,

40:23 you know,

40:23 service delivery indicators,

40:25 but also individual household surveys,

40:27 and there's a lot of great innovation going on in

40:30 measurement of beliefs and attitudes and perceived social constraints.

40:34 But I really would like to encourage us to think that,

40:36 uh,

40:36 to think about the need to,

40:38 to center those things when we're talking about human capital.

40:42 Because what we've seen is the good news is when households perceive opportunities

40:46 for women and girls to enter the labor force in a productive way,

40:50 they respond rapidly.

40:51 They invest in girls' education,

40:53 they change their attitudes about women's labor force participation.

40:56 We've seen that with business process outsourcing in India,

40:59 with uh

41:01 garment manufacturing in Bangladesh.

41:03 Um,

41:04 but the point is we need to think about

41:06 in a context specific.

41:07 way,

41:08 what are the binding constraints,

41:09 not just on access and not just within the classroom,

41:12 but that are preventing,

41:14 uh,

41:15 not just women and girls,

41:16 but lots of groups from being able to

41:18 have a pathway from building human capital

41:21 to contributing and fully participating

41:24 in the economy and think about building our measures and our interventions

41:27 to work on that margin as well.

41:29 So I'll stop there.

41:30 Thank you very much.

41:33 Thanks,

41:33 Pam,

41:33 for that was

41:34 really interesting and useful,

41:36 uh.

41:37 Contextualization and,

41:38 and slight challenge to the,

41:40 to the way uh

41:42 we're thinking about this

41:44 issue of building human capital and,

41:45 and,

41:45 and the benefits we might,

41:47 uh,

41:47 reap from that.

41:48 Um,

41:49 next up is Rukmini Banerjee.

41:50 Rukmini,

41:51 are you on?

41:57 Yep,

41:58 uh,

41:58 I'm here.

41:59 Can you hear me?

42:00 Hi,

42:01 Rukmini,

42:01 we can hear you and we can see you.

42:03 Thanks for being here.

42:05 Great.

42:05 Uh,

42:06 thank you for having me.

42:07 Um,

42:08 I really enjoyed the,

42:09 uh,

42:10 speakers before me,

42:12 and I was just thinking that,

42:14 uh,

42:14 you know,

42:15 we are celebrating,

42:16 I guess,

42:16 20 years of,

42:17 uh,

42:18 evolution of knowledge,

42:20 and,

42:21 uh,

42:21 especially as far as the

42:23 learning crisis and what is happening with children and schooling and learning.

42:28 Uh,

42:28 it almost seems like,

42:30 uh,

42:30 you know,

42:31 this,

42:32 uh,

42:32 difference between schooling and learning.

42:35 Schooling isn't learning,

42:37 overambitious curriculum leading to negative consequences.

42:41 All of these things now

42:43 seem to be,

42:44 you know,

42:45 part of the knowledge,

42:46 at least of a lot of people in this room,

42:49 as well as in kind of the highest policy circles.

42:52 Uh,

42:52 but

42:53 when we talk about knowledge and what is known changing life on the ground,

42:58 are there still things to be done?

43:01 Uh,

43:01 we just released in India the,

43:04 uh,

43:05 I've

43:05 lost track myself now,

43:07 I think the 17th annual status of Education report,

43:10 the UA report,

43:11 about 10 days ago,

43:13 uh,

43:13 where,

43:14 uh,

43:14 we are talking about,

43:16 um,

43:17 you know,

43:17 where are the basic learning levels in India,

43:20 and obviously because of the last two years of the pandemic,

43:24 there is a certain amount of decline

43:27 or lack of recovery,

43:28 you know,

43:29 in different places across the country.

43:31 Now,

43:32 whether you think about learning adjusted years of,

43:35 uh,

43:35 learning,

43:35 adjusted

43:37 years of schooling that the World Bank is coming up with,

43:40 or other kinds of measures,

43:43 uh,

43:43 my question often to myself as well as to our colleagues in India is,

43:48 to what extent

43:50 on the ground

43:51 are parents,

43:54 are teachers,

43:55 Accepting that this

43:58 big gap that they see between learning and schooling

44:01 is something that they need to work on

44:03 as,

44:04 uh,

44:05 you know,

44:05 as children go through the system.

44:07 Uh,

44:07 so for example,

44:08 there's been a lot of hand wringing on

44:11 the learning loss during COVID,

44:13 but I think there is,

44:14 uh,

44:15 other than in very enlightened circles.

44:18 There is less of a realization that

44:20 that dip has come on top of a much longer trend

44:24 of low learning levels in the world,

44:27 and therefore we need to get back to the structural reasons

44:30 for why this is the case,

44:32 uh,

44:32 and,

44:33 uh,

44:33 you know,

44:33 work on the solutions.

44:34 Some solutions exist.

44:36 Patrick and others have laid them out,

44:38 but work on the solutions which are actually longer term solutions,

44:42 not just as a response to COVID.

44:44 So if I think about the way that we have worked in India

44:49 in kind of understanding the problem,

44:51 it started with a very simple way of

44:54 assessing kids

44:56 using

44:57 a very basic tool

44:59 which didn't go at grade level.

45:01 And I think we also underestimate the,

45:04 the mindset

45:06 that almost

45:08 everybody has,

45:09 at least in a

45:10 low income country,

45:12 that schooling means a linear progression

45:15 year on year through a system.

45:17 And as you rise through the system,

45:20 that is spend more and more years in school,

45:23 you know,

45:23 somewhere you are different from those who have spent less years in school.

45:27 Despite the fact that the data for 10 years

45:30 is showing

45:31 that if you didn't get onto the treadmill at the right pace

45:34 early in your school life,

45:35 you actually you're not going to make it.

45:37 So we have enrollment,

45:39 for example,

45:39 if we look at enrollment in India in 8th grade,

45:42 uh,

45:43 and if you remember that

45:45 we have about 25 million

45:48 kids in each single age group.

45:51 The 8th grade enrollment

45:54 in say 2008

45:56 was about 11 or 12 million nationally,

45:58 and that number in 2018 or 201919

46:01 is like 23 million.

46:03 So while learning levels have not moved,

46:05 enrollments and completed years of schooling

46:09 have gone through the roof,

46:10 and the implications of that on what happens beyond,

46:13 you know,

46:13 I think is understood in a research sense

46:17 but not understood in a real and a practical sense on the ground.

46:20 So my one question that I would like to leave on this panel is

46:25 there is a lot of research results,

46:27 you know,

46:27 we have

46:28 uh panels and committees which are compiling the most cutting edge research,

46:33 but how long does it take

46:35 for the key elements of

46:38 our knowledge to actually get down onto the ground

46:41 and what methods or what,

46:43 uh,

46:44 what mechanisms

46:47 enable that to happen?

46:48 Now,

46:48 in some,

46:49 in a very perverted way,

46:50 I feel COVID could have been that instrument

46:54 that it shook everybody up in such a big way

46:56 that you are forced to think in different ways.

46:59 And,

47:00 you know,

47:00 just the fact,

47:01 for example,

47:02 if I quote from our latest AA report

47:05 that there are kids who are in 3rd grade

47:08 in India right now who actually had no schooling.

47:11 They came in.

47:12 Our schools were closed for 2 years.

47:14 They moved into 3rd grade in April of

47:16 2022.

47:18 And so by the time our asset measurement happened,

47:22 which was about 6 months later,

47:23 they actually had only 6 months of schooling.

47:25 And yet if you look at the basic level of learning.

47:29 There is 20 to 25% of kids in the country are at grade level,

47:34 and this is despite schools being closed.

47:36 So you would think that even these very basic facts

47:40 lead you to think about what is it that we are doing in school

47:44 or not doing in school

47:46 that is leading to this situation.

47:48 And if schools have been closed for two years

47:50 and we see a drop in learning for sure,

47:52 but you don't see zero,

47:54 what does that teach you?

47:55 What does that mean?

47:56 Uh,

47:57 and so,

47:57 you know,

47:58 I would really like,

47:59 uh,

47:59 you know,

48:00 I'm sure

48:01 there are others who are thinking about this,

48:02 about how does research get to the ground,

48:05 uh,

48:05 how about

48:06 the fact that we have solutions,

48:08 many of which have been laid out,

48:10 and,

48:10 you know,

48:10 the one that we have worked on a lot,

48:12 uh,

48:12 teaching at the right level,

48:14 evaluated

48:15 many times,

48:16 shown to be very effective,

48:18 but there isn't like an automatic

48:21 use of that data

48:23 despite the fact that the World Bank and

48:26 FCDO and whatnot have it as a good buy,

48:29 uh,

48:29 not a goodbye,

48:30 but a good buy,

48:31 uh,

48:32 uh,

48:33 out there.

48:33 We don't see like a massive sudden huge

48:36 demand from governments to use that solution,

48:39 which is well known.

48:40 So where are these gaps and what does the

48:44 knowledge producing community need to do

48:47 to a knowledge digesting community

48:50 who can then put some of these

48:52 uh

48:53 things that are now well known into practice.

48:55 So let me just stop there because

48:58 maybe others have

48:59 better solutions to offer.

49:03 Thanks,

49:03 Rukmini.

49:04 Another really insightful set of challenges to,

49:07 to how we think about um

49:09 the,

49:09 the,

49:09 the,

49:09 the production of this research and,

49:11 and where,

49:12 where it leads,

49:13 uh,

49:14 um,

49:14 and how it moves to practice.

49:16 Um,

49:16 next up,

49:17 we have,

49:17 uh,

49:17 Emily,

49:18 Emily Gustafson Wright,

49:20 um,

49:21 who's here in person.

49:24 Thanks,

49:24 Dion,

49:25 and uh thanks everybody,

49:27 great to be here.

49:28 Um,

49:29 so I'm at the Center for Universal Education at the Brookings Institution and um

49:33 we do,

49:34 uh,

49:35 I think a lot of digesting

49:37 of the research also and I think that that's a really

49:40 important

49:40 uh role is,

49:41 um,

49:41 you know,

49:42 translating a lot of these um really important findings from the bank and,

49:45 and elsewhere,

49:46 um,

49:46 as well as our own research,

49:48 um,

49:48 and including a number of areas such as scaling and gender equity.

49:53 Um,

49:53 what we call a breadth of skills.

49:55 Um,

49:56 there was a discussion on the slides around

49:58 non-cognitive skills,

49:59 we call them a breadth of skills,

50:00 systems transformation,

50:02 ah,

50:02 family engagement,

50:03 which was also mentioned,

50:05 um,

50:06 just to name a few.

50:07 I primarily focus

50:09 on costing and financing education and early childhood development,

50:13 and for the past decade or so,

50:14 I've been working on results-based financing.

50:17 So I will focus,

50:18 um,

50:19 my comments on results-based financing.

50:21 Um,

50:22 and,

50:23 um,

50:24 for human development or human capital,

50:26 um,

50:26 and I'll focus on four main,

50:28 uh,

50:28 4 main areas.

50:29 So first I'll comment on what was presented

50:32 with respect to the evidence and what wasn't.

50:35 Um,

50:35 and then I'll highlight some nuances that I think are

50:38 critical to consider when looking at the evidence or lack

50:41 or lack thereof.

50:43 Um,

50:43 third,

50:43 I'll describe the current landscape of results-based

50:46 financing in education and health and,

50:49 um,

50:49 what we might expect to see in the coming years.

50:52 Um,

50:52 and then finally,

50:53 I'll suggest some areas for,

50:54 um,

50:55 for further research.

50:57 Um,

50:57 so we heard some evidence on policies to

51:00 improve,

51:00 um,

51:00 human capital.

51:02 Um,

51:02 this included two forms of contingent financing,

51:05 uh,

51:05 conditional cash transfers and performance-based grants.

51:08 Um,

51:09 the results are mainly positive but,

51:10 but mixed,

51:11 um,

51:12 for both of the mechanisms,

51:14 um,

51:14 that were presented,

51:16 um,

51:16 in particular in terms of short versus long-term effects,

51:19 um,

51:19 but also for different,

51:20 um,

51:21 populations.

51:22 Uh,

51:22 fun fact,

51:23 I actually started my career,

51:24 um,

51:24 over two decades ago,

51:26 working on conditional cash transfers here at the bank.

51:29 Um,

51:29 the bank,

51:30 uh,

51:30 was skeptical,

51:31 um,

51:32 I would say,

51:33 um,

51:33 IDB was,

51:34 I think,

51:34 a bit more positive at that time.

51:36 Uh,

51:36 we were working in Brazil with the

51:38 then Bolsa Escola program.

51:41 Um,

51:42 so I guess I was,

51:43 um,

51:43 you know,

51:44 of course this is the,

51:44 this is the lens that I'm looking at this through,

51:46 and,

51:47 um,

51:47 I was surprised not to see a little bit more presented on the topic,

51:50 of course there wasn't a lot of time,

51:52 um,

51:53 maybe also this is just because,

51:55 um,

51:56 the,

51:56 um,

51:58 the research that's been done at the bank was funded by other,

52:00 other,

52:00 other trust funds or other,

52:02 um,

52:02 um,

52:03 other funders,

52:04 um,

52:04 but there is,

52:05 there has been quite a bit of work done at the bank,

52:07 um.

52:09 For example,

52:09 under the REACH Trust Fund,

52:10 um,

52:11 also an IEG review of P4R,

52:13 um,

52:14 and then,

52:14 um,

52:15 some work done by,

52:15 um,

52:16 GPRBA as well.

52:17 Um,

52:18 and of course lots of research that's been done,

52:20 uh,

52:20 outside of the bank on results-based financing.

52:23 So I'm not gonna go into the details,

52:24 um,

52:25 about the existing evidence.

52:26 Rather,

52:26 I'd like to make an observation

52:28 about the analysis of or the conversations

52:31 that are being had about results-based financing

52:33 or RBF literature,

52:34 and I have to admit I'm sort of using this platform

52:37 to address a bit of a pet peeve of my own.

52:40 Um,

52:41 so

52:42 there are quite a few different,

52:44 uh,

52:44 contingent payment or RBF mechanisms,

52:47 um,

52:48 that have been applied,

52:49 um,

52:49 to both education and and in the health sector,

52:52 uh,

52:52 over the years.

52:54 But what I've seen is often the findings uh tend to be bunched together,

52:58 uh,

52:58 kind of under one umbrella,

52:59 um,

53:00 and I,

53:00 and I feel that little attention is being paid to sort of the nuances

53:04 of the different types of mechanisms and or the projects themselves.

53:08 Um,

53:09 I feel like this has resulted sometimes in sort

53:11 of blanket decisions being made around our RBF,

53:14 um,

53:15 as a potential tool to address some of the challenges that we've talked about today.

53:19 Um,

53:19 and there are in fact considerable differences between

53:22 the different types of RBA,

53:24 um,

53:25 and,

53:25 uh,

53:25 and we've seen some of,

53:27 uh,

53:27 you know,

53:28 research done about this over the past couple of decades.

53:31 Um,

53:31 so the primary difference

53:32 between the different mechanisms is who bears the financial risk,

53:35 um,

53:36 or who,

53:37 who won't get paid if the results aren't achieved.

53:39 So that can be national government,

53:41 subnational,

53:42 um,

53:42 service providers,

53:43 or schools,

53:44 uh,

53:44 or health,

53:45 uh,

53:45 centers or teachers,

53:47 um,

53:47 individuals in the case of,

53:48 um,

53:49 CCTs,

53:50 um,

53:50 and in the case of the more recent,

53:51 um,

53:51 social development impact bonds,

53:53 which is where my research focuses,

53:54 um,

53:55 impact investors.

53:56 So,

53:57 who,

53:57 who isn't going to get,

53:58 who's going to get paid or not get paid,

54:00 um,

54:00 really matters.

54:01 Second,

54:03 Um,

54:03 I think the governance structure or the power dynamics,

54:06 um,

54:06 of these configurations matter.

54:08 So who's making the decisions about the

54:09 metrics and the threshold thresholds for payments,

54:12 for example.

54:13 Um,

54:14 3rd,

54:15 um,

54:15 the,

54:16 the,

54:16 the measurement and evaluation matter.

54:18 So are the results outputs or outcomes,

54:20 we've had a 10,

54:21 we've tended to see,

54:23 um,

54:23 a greater focus on outputs in traditional results-based financing.

54:27 We're seeing a shift towards,

54:28 uh,

54:28 farther to the right in the results chain

54:30 towards outcomes in social and development impact months,

54:33 for example.

54:34 Um,

54:34 also,

54:35 you know,

54:35 how and what are the thresholds for success,

54:37 um,

54:38 determined,

54:39 um,

54:39 what is the counterfactual?

54:40 Is there even a counterfactual?

54:42 Um,

54:43 and then finally,

54:44 obviously the context matters,

54:45 um,

54:46 and what happens around the project in terms of,

54:48 um,

54:48 technical support and,

54:50 and other elements.

54:52 Um,

54:53 so,

54:53 uh,

54:53 third,

54:54 what is the current landscape and what,

54:56 uh,

54:56 what are the trends that we are seeing?

54:58 Um,

54:59 so overall we are seeing,

55:00 um,

55:01 quite a bit of an upward trend,

55:03 um,

55:03 in the use of results-based-based financing.

55:05 I,

55:05 I do think that this,

55:06 you know,

55:07 kind of,

55:07 this comes from the,

55:09 the,

55:10 the access plus learning,

55:11 um,

55:12 a shift in focus to outcomes,

55:13 constrained budgets,

55:15 um,

55:16 you know,

55:16 I think that the

55:17 RCT movement,

55:18 I think,

55:19 was part of this,

55:20 um,

55:21 this push perhaps.

55:22 Um,

55:24 and,

55:24 um,

55:25 so I'm,

55:25 I'm not going to give sort of the,

55:27 it's,

55:27 it's very difficult to measure

55:29 sort of how many,

55:30 uh,

55:31 because there are so many different types of results-based financing,

55:33 kind of how many are we seeing in the world,

55:35 it is on the rise.

55:36 I'll focus,

55:37 um,

55:37 on,

55:38 um,

55:38 social development impact bonds,

55:39 which,

55:39 um,

55:40 is,

55:40 um,

55:41 an area that I,

55:41 again,

55:42 that I focus on and maintain a database of all of the projects in the world.

55:45 So this is just to give you sort of a sense of,

55:47 of what we're seeing,

55:47 um,

55:48 in the global landscape.

55:50 So,

55:51 Um,

55:51 there are about 240 impact bonds projects,

55:54 impact bond projects globally,

55:56 um,

55:56 that have been contracted.

55:57 Um,

55:58 the majority of those are in the employment and social welfare sector.

56:02 Um,

56:02 however,

56:03 the education sector was the one that,

56:05 that was the sector that grew the most,

56:06 um,

56:06 in the past year,

56:07 um,

56:08 an increase of 12 projects,

56:09 bringing the total number of education projects in the world to 40.

56:12 Um,

56:13 and this included the launch of 5 projects in Sierra Leone.

56:17 Uh,

56:17 which are meant to reach over 130,000 students

56:20 as part of the Education Outcomes Fund,

56:22 which is a pool of funds for,

56:24 um,

56:25 to pay for outcomes,

56:26 um,

56:27 which I can speak more about if you're curious.

56:29 Um,

56:30 and,

56:31 um,

56:32 It would,

56:32 yeah,

56:33 so that includes a multiple,

56:34 um,

56:34 results-based financing projects within that,

56:36 um,

56:37 and then there were,

56:38 um,

56:38 there were only two new health projects,

56:39 uh,

56:39 in the past year,

56:40 um,

56:41 bringing the total of,

56:42 um,

56:43 health projects to 36.

56:45 We,

56:46 we saw a little bit of a slow during the pandemic,

56:48 obviously,

56:48 but there seems to

56:49 be sort of,

56:50 they seem to be ramping up now.

56:52 Um,

56:52 what do we see in the pipeline?

56:54 Uh,

56:54 quite a bit,

56:55 um,

56:55 happening,

56:56 uh,

56:56 on the horizon in terms of education.

56:58 Um,

56:59 both impact bonds and outcomes funds like the Education Outcomes Fund,

57:03 um,

57:03 will be launching another project,

57:05 uh,

57:05 in Ghana,

57:07 um,

57:07 reaching over,

57:08 uh,

57:09 nearly 200,000 students,

57:10 um,

57:11 the Back to school Outcomes Fund in India,

57:13 a number of early childhood projects,

57:15 uh,

57:15 with Education Outcomes Fund as well as,

57:17 um,

57:18 independently,

57:19 and for example,

57:19 in Jordan,

57:20 um,

57:21 also in Chile,

57:22 we see a number,

57:23 um,

57:23 in,

57:23 in health,

57:24 also a few,

57:25 uh,

57:25 a number that are in both in early and latest design.

57:28 Um,

57:29 all right,

57:29 so in terms of the

57:30 future work in this area,

57:32 um,

57:33 So on the one hand,

57:34 we,

57:34 we know,

57:35 and I,

57:35 I didn't really go into this,

57:36 but we,

57:37 we know that there is some rigorous evidence,

57:39 uh,

57:40 a lack of rigorous evidence.

57:42 It's a tough experiment to do,

57:43 right,

57:44 um,

57:44 sort of,

57:46 uh,

57:47 project using this financing,

57:48 uh,

57:49 mechanism,

57:49 uh,

57:50 versus not,

57:51 um,

57:52 but on the other hand,

57:53 um,

57:53 some would argue

57:55 that it's really time that governments and donors

57:57 stop paying for things that aren't delivering outcomes.

58:01 Um,

58:02 so if that's the premise.

58:04 I posit the question should be,

58:05 you know,

58:06 when it's appropriate,

58:07 how can results-based financing be done better?

58:09 Um,

58:10 how can systems be set up to ensure more efficient design,

58:13 um,

58:13 and implementation of results-based financing for human development.

58:17 Um,

58:17 I'm currently working on a paper together with Thomas Poulson from the bank.

58:21 Um,

58:21 we're exploring

58:22 the four types of data needed for results-based financing.

58:26 Uh,

58:26 so in addition to cost data,

58:28 cost of inaction data,

58:29 and results data,

58:31 um,

58:31 we're,

58:32 um,

58:32 we believe it's really critical to invest in the monitoring data,

58:35 the real-time data,

58:36 um,

58:36 for adaptive management,

58:38 um,

58:38 in classroom classrooms and health,

58:40 uh,

58:40 settings,

58:41 um,

58:41 with a focus on equitable,

58:43 um,

58:43 outcome achievement.

58:44 Um,

58:45 and then finally back to my final point,

58:46 I,

58:46 I think it's really critical to understand

58:48 the nuances of these different projects,

58:49 so,

58:50 um,

58:50 rather than sort of

58:52 throwing the baby out

58:53 with the bathwater or going full steam ahead,

58:56 I think it's really important to,

58:57 to consider these different factors,

58:59 um,

58:59 that drive the results within them.

59:04 Thanks,

59:04 thanks,

59:05 Emily.

59:05 Another sort of reminder of,

59:07 of,

59:08 of how we should be careful in,

59:10 in how to interpret

59:12 the results and what to expect from these different interventions.

59:15 Um,

59:16 last but not least,

59:16 um,

59:17 Ritva Reineke,

59:18 uh,

59:19 Ritva,

59:19 are you online?

59:20 Ritva was,

59:20 was my former boss,

59:22 and,

59:23 uh,

59:23 under various configurations.

59:25 I'm very excited to see her today.

59:27 Uh,

59:28 over to you,

59:28 Ritva.

59:30 Thank you very much,

59:31 Dion,

59:31 and it's great to see you and many

59:34 other familiar faces,

59:36 um,

59:36 and colleagues.

59:38 And,

59:38 and anyway,

59:39 I have to say that uh it's great for me to join this celebration.

59:43 When I got the invitation,

59:45 I never knew this would still exist,

59:47 Knowledge for Change program,

59:50 because at the time when it was established,

59:53 I worked at the World Bank's research department and

59:57 One of my jobs was to establish this kind of a fund,

1:00:01 not just by myself,

1:00:02 but

1:00:03 as a group.

1:00:04 So it was uh

1:00:05 the longevity is amazing and I'm,

1:00:08 I'm,

1:00:08 I'm really thrilled.

1:00:09 So,

1:00:10 you,

1:00:10 you've done a lot of work over.

1:00:12 The,

1:00:13 yes.

1:00:13 I just will

1:00:15 uh make a few comments about the past,

1:00:18 um,

1:00:19 about research and policy,

1:00:21 and then

1:00:22 the future,

1:00:23 especially

1:00:24 based on,

1:00:25 on what Kathleen and um

1:00:28 Patrick presented.

1:00:30 So just about the past and it's,

1:00:32 it was fantastic to hear the various

1:00:35 comments because

1:00:37 at that time in the bank's research department,

1:00:41 the way research was organized,

1:00:44 we had

1:00:45 a group that studied demand for education and health using household surveys,

1:00:51 and then we had a public.

1:00:52 Economics

1:00:54 team

1:00:55 that used the budget data.

1:00:57 They did not interact.

1:00:59 They were separate endeavors.

1:01:02 And

1:01:03 at that time when KCB was also set up,

1:01:06 these two were integrated.

1:01:08 And what also happened was that

1:01:11 public economics in the bank really went micro.

1:01:15 Um,

1:01:16 and,

1:01:18 and there was a new research program on service delivery.

1:01:20 There was no analysis,

1:01:22 uh,

1:01:22 the research analysis

1:01:24 on service delivery,

1:01:25 and it was really great to hear,

1:01:27 uh,

1:01:28 the,

1:01:28 the fellow panelists to say that,

1:01:30 OK,

1:01:30 we need to go back to demand,

1:01:32 so something has been achieved,

1:01:35 uh,

1:01:35 for sure.

1:01:37 I think the key issue was that it was

1:01:40 two things.

1:01:41 One,

1:01:42 it was recognized that there is behavior

1:01:45 on the supply side as well.

1:01:47 There are incentives.

1:01:48 There are services that have guards of people

1:01:52 doing the service and,

1:01:53 and therefore there's behavior and incentives.

1:01:57 I think also the other thing was that this allowed

1:02:01 the bank's research also to get closer to the operational work and the country's

1:02:08 actual service delivery

1:02:10 because

1:02:11 we went beyond budget and I think the key issue was to realize also

1:02:15 budgets actually are very different from service delivery as such.

1:02:20 So at that time,

1:02:23 We were after impact,

1:02:25 we started the pioneering some of the

1:02:27 measurements like leakage of funds or absence rates

1:02:31 and competence of service providers,

1:02:34 and some of

1:02:35 you really presented very nicely.

1:02:38 We also put together making services work for poor people,

1:02:42 uh,

1:02:43 WDR which

1:02:44 Which,

1:02:45 which was hoped at the time that it would

1:02:48 spur further research in this area.

1:02:51 And when I looked at your presentation,

1:02:54 um,

1:02:54 uh,

1:02:55 Katherine and Patrick,

1:02:56 I,

1:02:56 I think there has been in research

1:02:59 huge progress

1:03:01 in the 20 years.

1:03:02 So,

1:03:03 uh,

1:03:03 that's really super good.

1:03:06 Uh,

1:03:06 then when I think about impact,

1:03:08 which was desired impact on actual service delivery,

1:03:13 particularly,

1:03:14 and I often think about poor countries,

1:03:17 I

1:03:18 I think definitely in education and it was,

1:03:21 by the way,

1:03:21 interesting that the whole panel

1:03:23 also is

1:03:24 uh people who have mostly studied education or worked in education.

1:03:30 My,

1:03:30 my view is that in healthcare,

1:03:32 this is much less so,

1:03:34 that kind of agenda,

1:03:36 uh,

1:03:36 much less in,

1:03:38 in healthcare.

1:03:40 OK,

1:03:41 so that's about the past.

1:03:43 Linking with policy,

1:03:45 uh,

1:03:45 research and policy,

1:03:47 that has been,

1:03:48 um,

1:03:48 a kind of dear topic for me.

1:03:51 I

1:03:51 used to work mostly in um

1:03:54 operations uh there in the bank

1:03:57 and,

1:03:57 and also a long time ago in UNICEF.

1:04:00 Um,

1:04:02 And in the bank,

1:04:03 um,

1:04:04 I had a chance

1:04:05 to put this research,

1:04:08 I mean,

1:04:10 for my,

1:04:10 for,

1:04:11 for a small part,

1:04:12 always,

1:04:12 one has to emphasize,

1:04:14 and into operations as Human Development Director

1:04:18 in the Africa region,

1:04:19 which was my

1:04:21 last um

1:04:22 assignment in the bank.

1:04:24 And it's easier said than done,

1:04:28 and I think Rukmini also

1:04:30 hinted

1:04:32 or

1:04:32 said that very clearly.

1:04:35 You have

1:04:36 certain

1:04:37 absolute winners and they don't go,

1:04:40 uh,

1:04:41 you know,

1:04:42 to the system.

1:04:43 Obviously.

1:04:45 Um,

1:04:45 it's also this thing,

1:04:46 uh,

1:04:46 that,

1:04:47 uh,

1:04:47 that I find that

1:04:49 research agendas that are born in academic circles don't necessarily

1:04:55 reach policymakers and sometimes they don't speak to them.

1:05:01 And RE a program,

1:05:03 the research on improving systems of education.

1:05:06 Um,

1:05:07 that many of you have been involved in which is a magnificent large program

1:05:12 is now struggling with that,

1:05:14 churning out

1:05:16 this

1:05:17 material to try and

1:05:19 get the message

1:05:20 through,

1:05:20 and it will be very interesting to see

1:05:23 how that works,

1:05:24 but it is not

1:05:25 very easy.

1:05:26 I think in,

1:05:27 we actually work together beyond in the Africa region 8D.

1:05:33 And you were leading the kind of the think tank part of it in,

1:05:37 in the region,

1:05:39 and I felt it was really,

1:05:40 really important to be

1:05:42 in there,

1:05:43 um,

1:05:43 and,

1:05:44 and we did work

1:05:45 on regional topics,

1:05:47 which for Africa,

1:05:48 we at the time felt,

1:05:49 and I think it,

1:05:50 I feel

1:05:51 Uh,

1:05:52 continued to be

1:05:53 population,

1:05:54 youth employment,

1:05:55 and service delivery,

1:05:57 and it was definitely much better when one is right there and the agenda is set

1:06:03 by the operational people that were in the bank

1:06:06 and then

1:06:06 the policymakers in the countries.

1:06:09 In my

1:06:10 own experience of the two decades in the bank,

1:06:14 I

1:06:15 was able

1:06:17 to combine research and operation

1:06:20 the best when I was a country economist working in Uganda.

1:06:24 Being,

1:06:25 being in their team entirely on the policymakers's terms,

1:06:30 uh,

1:06:31 it determined what was researched

1:06:34 and,

1:06:34 and then you were able to bring certain

1:06:38 Uh,

1:06:39 research that you knew in,

1:06:41 into the system.

1:06:42 So,

1:06:42 in this linking of research and policy,

1:06:45 it's a demanding task and my experience is

1:06:49 that once you are in the policymakers's team,

1:06:52 it somehow

1:06:53 is easier to pull it in there.

1:06:57 OK,

1:06:57 3rd thing,

1:06:58 looking

1:06:59 forward.

1:07:01 So

1:07:02 So interventions is,

1:07:04 is the term that is used

1:07:08 uh most.

1:07:09 And I think it's the and the randomista movement is the name of the

1:07:14 game today,

1:07:15 and,

1:07:15 and a lot has happened in that.

1:07:18 I remember we actually

1:07:20 in our team had Mike Kramer for 3 years,

1:07:24 part time,

1:07:24 not full time or part time.

1:07:26 I never would have believed

1:07:29 that that impact evaluation agenda

1:07:31 is so dominant.

1:07:34 What is much less dominant I feel is what

1:07:37 came out of the World Development Report 2004,

1:07:41 Making Services Work

1:07:42 for poor people,

1:07:43 which

1:07:45 Gave a way to think about systems

1:07:48 and,

1:07:48 and that features in my view,

1:07:50 much less today

1:07:53 to look at the politics,

1:07:55 the compact between providers and policymakers,

1:07:58 management and client power

1:08:00 and those elements that belong to it.

1:08:03 Um,

1:08:05 And I,

1:08:06 I think um

1:08:08 really two things in,

1:08:09 in that,

1:08:10 is that when you think of the system as a matrix,

1:08:13 as

1:08:14 the RICE program,

1:08:15 for instance,

1:08:16 effectively uses this very same

1:08:18 uh framework

1:08:20 things,

1:08:20 you have

1:08:22 These columns and rows

1:08:25 in the system

1:08:26 and there can be so many incoherent things

1:08:29 in that

1:08:30 um

1:08:31 system

1:08:32 and in those columns and in uh in those rows.

1:08:36 And then when you have an intervention,

1:08:39 when it lands in the system,

1:08:42 it,

1:08:42 it,

1:08:42 it is totally unknown

1:08:44 what kind of a cell is it.

1:08:47 Is it in the middle of incoherencies

1:08:49 or is it coherent in terms of that

1:08:52 what you demand is financed,

1:08:55 what you provide information on tells about performance,

1:08:59 things like that,

1:09:00 or motivation.

1:09:02 So,

1:09:03 like in my own research,

1:09:05 so,

1:09:05 so two things.

1:09:06 One,

1:09:07 in my own

1:09:09 little research activity,

1:09:10 uh,

1:09:11 which is in a very poor,

1:09:13 low-income country,

1:09:14 Mozambique,

1:09:16 when,

1:09:16 uh,

1:09:16 just recently,

1:09:17 we completed um

1:09:19 A kind of an

1:09:21 ethnography,

1:09:23 like a modest ethno ethnography,

1:09:25 let's put it this way.

1:09:26 It is astounding to see

1:09:29 how interventions are being piled up on the system

1:09:33 and the impact it has.

1:09:35 On motivational teachers,

1:09:38 behaviors of head teachers,

1:09:40 local officials.

1:09:41 It,

1:09:41 it focuses on,

1:09:42 on,

1:09:42 on local um

1:09:44 actors mostly

1:09:46 and then all the donor agencies being

1:09:48 in the,

1:09:49 in the mix.

1:09:51 So it,

1:09:52 It,

1:09:53 at minimum creates these incoherencies or enforces them

1:09:59 at least,

1:10:00 and sometimes the system looks.

1:10:02 Chaotic.

1:10:03 So,

1:10:04 in,

1:10:04 in

1:10:05 that

1:10:06 type of approach which,

1:10:07 which also

1:10:09 has been in the RICE program that I

1:10:11 have been involved

1:10:12 with a little bit,

1:10:14 um,

1:10:14 I,

1:10:14 I really like that

1:10:17 because you,

1:10:17 you need to know

1:10:19 how

1:10:20 this all

1:10:21 pushing these things actually affects the system.

1:10:25 OK.

1:10:25 The last point,

1:10:28 why I'm a systems-oriented,

1:10:30 country-oriented person.

1:10:32 So,

1:10:33 I,

1:10:33 I

1:10:34 really think

1:10:36 this is very important for Africa especially.

1:10:40 Why?

1:10:41 Because in Africa,

1:10:43 systems are massively expanding,

1:10:46 not in the north as we heard in the poverty session

1:10:50 and not in the south.

1:10:51 South,

1:10:51 but in the West,

1:10:53 in the center,

1:10:54 and East Africa.

1:10:56 In the next 2 years,

1:10:57 if KCP still exists,

1:11:00 in the next 20 years,

1:11:02 population in these areas will have pretty much doubled.

1:11:07 It will mean,

1:11:08 and these are young people,

1:11:09 it will mean that education systems

1:11:12 expand massively as they have and they will.

1:11:17 And then I would like to ask,

1:11:20 A question for the future,

1:11:23 does research have anything to say

1:11:26 to these massively expanding,

1:11:28 uh,

1:11:29 system,

1:11:30 systems that

1:11:31 like,

1:11:32 um,

1:11:32 I was impressed by JP uh

1:11:36 um

1:11:37 uh Sedo telling me this,

1:11:39 these numbers that if

1:11:41 Brazil,

1:11:42 Nigeria,

1:11:43 Bangladesh now have roughly 45 million

1:11:47 in the school system.

1:11:48 Brazil and Bangladesh will reduce slowly.

1:11:52 In Nigeria in 20 years,

1:11:55 it will be 90 million.

1:11:57 So who has

1:11:59 the experience of expanding.

1:12:02 Systems to,

1:12:03 to accommodate

1:12:05 that uh is uh and,

1:12:08 and does,

1:12:08 is research able to say something about it or is it just

1:12:12 an operational issue?

1:12:14 I mean,

1:12:15 another number,

1:12:16 these are just back of the envelope things,

1:12:18 but it was also some calculations that in Africa,

1:12:23 um,

1:12:23 16% of workforce should be teachers.

1:12:27 I mean,

1:12:28 Here in America,

1:12:28 healthcare is uh

1:12:30 like a mass sector and 7 every 7 person is working in healthcare,

1:12:36 is 16% is,

1:12:38 uh,

1:12:39 is something similar.

1:12:41 So,

1:12:41 so how,

1:12:43 how,

1:12:43 how can,

1:12:45 how can,

1:12:45 um,

1:12:46 research and,

1:12:47 and this work help,

1:12:50 uh,

1:12:50 you know,

1:12:52 make that expansion

1:12:54 possible?

1:12:55 So that was all I,

1:12:57 I

1:12:58 wanted to say and um

1:13:00 And I hope KCP conti continues to contribute

1:13:04 to this and,

1:13:05 and other issues.

1:13:06 So thank you very much,

1:13:07 Tion.

1:13:10 Thank you,

1:13:10 Rita,

1:13:10 um.

1:13:13 This has been fascinating because I think in both sessions,

1:13:15 we've seen sort of a,

1:13:16 a,

1:13:17 a strong core of work,

1:13:18 a sort of a general buy-in that work.

1:13:21 But then in the panel discussions,

1:13:23 I think it's been really interesting to see how

1:13:26 You know,

1:13:26 it's,

1:13:27 it's just never enough.

1:13:28 Sitting in a research department,

1:13:29 that's actually not bad news because there's a lot more to do.

1:13:33 But it's on a number of fronts.

1:13:34 It's around the,

1:13:34 you know,

1:13:35 Pam emphasized the complementarities with other factors,

1:13:38 uh,

1:13:38 Emily.

1:13:40 Uh,

1:13:41 uh,

1:13:41 among other comments,

1:13:42 you know,

1:13:43 the,

1:13:43 the details and the nuance and understanding differences and impacts

1:13:46 is really where we're gonna get a lot of information.

1:13:49 Um,

1:13:50 I thought,

1:13:50 I thought both Rukmini and,

1:13:52 and,

1:13:52 and,

1:13:52 and Ritva's points about the links between research and,

1:13:55 and,

1:13:55 and practice and operations and,

1:13:57 and programs.

1:13:58 I,

1:13:59 I think still need,

1:14:00 you know,

1:14:00 we're highlighting how that,

1:14:01 that,

1:14:02 that linkage,

1:14:02 and I appreciate Ritter,

1:14:03 that you,

1:14:03 you're pointing out how difficult it is.

1:14:05 It's just not easy.

1:14:06 Uh,

1:14:06 but that doesn't mean we have to stop trying to,

1:14:08 to,

1:14:09 to,

1:14:09 to,

1:14:09 to do that better.

1:14:10 And I think everybody's pointed to sort of

1:14:12 directions where they see,

1:14:13 um,

1:14:14 future avenues for research,

1:14:16 which,

1:14:16 you know,

1:14:16 obviously,

1:14:17 us in the research department,

1:14:18 we,

1:14:19 you know,

1:14:19 this is sort of what we now need to go back and,

1:14:21 and,

1:14:21 and,

1:14:21 and think through and internalize.

1:14:23 So,

1:14:24 with that,

1:14:24 um,

1:14:25 let me thank everybody.

1:14:26 Let me give a particular thanks to the

1:14:29 Uh,

1:14:29 the program and the KCP Program Management Unit,

1:14:32 Karina,

1:14:32 Marcello,

1:14:33 Bintao,

1:14:33 who's not here,

1:14:34 and,

1:14:34 and Gabby for putting together this event,

1:14:36 uh,

1:14:37 which was really insightful,

1:14:38 and,

1:14:39 and I,

1:14:39 I'm a lot of food for thought.

1:14:41 And thank you all for participating both in the room and,

1:14:43 and online,

1:14:45 and look forward to our next,

1:14:46 uh,

1:14:47 KCP,

1:14:48 uh,

1:14:48 event that'll be in,

1:14:49 uh,

1:14:50 a couple of months,

1:14:51 and,

1:14:52 uh,

1:14:52 announcements will go out soon.

1:14:53 So thanks again,

1:14:54 everybody.

showAllTimestamps
no
transcript
So the second session is gonna focus on, on human capital and what we've learned over the years and how the, how the conceptualization of What human capital is and um how it's how it's created and how public policy can help create it has evolved over time. And, uh, we're gonna have two, presenters again, uh, Kathleen Beagle, uh, who's the Research Manager and lead, and lead economist in the human development team of the Development Research Group, who's here in person, and then we're gonna have Patrick Premont online, um, a senior economist, uh, in the DM. Uh, group, the development impact evaluation group, uh, uh, here at the bank. Um, so you're gonna start, how, how long are you gonna speak for? Uh, about 1010 minutes. OK, OK. And I'm gonna be stricter this time. Thanks. This is on, right? Now it's on. OK. Good morning. Thank you for coming to everyone online. Thank you for attending. Um, I'm pleased to give the first half of this presentation on behalf of myself and Patrick, uh, and then I'll hand it over to him, who is, uh, uh, currently not in DC, so he'll join us by video. OK. So, you know, taking a very quick step back and asking, like, where are we today, before jumping into what we've learned from research, um, you know, we want to make a few kind of very quick observations. Um, and let me premise this by saying, the topic of this talk, I think like with the last talk, is quite overwhelming and all encompassing. So, struggling to figure out what to prioritize in 25 minutes was a challenge for me and for Patrick. Uh, so hopefully we, we hit the, the big spots for you, but I think there are a lot of areas that will be nice to hear discussion during the panel from the panelists because we don't cover nearly everything. Um, but first we want to kind of remark on the incredible progress in education as measured by, uh, enrollment numbers in these figures, but also reflected in attainment and other measures of education progress. Um, immense progress in the last couple of decades, uh, a lot of catching. Up by lagging regions, um, what we don't show you here today is also on the gender side that girls have caught up in enrollment and attainment and in fact now in middle and high income countries, and most of them girls are, uh, exceeding boys in terms of educational attainment. And so where we have lagging er gender gaps are in um low income countries, and in fact in many countries the discussion now is how to address the boy disadvantage in education. So a lot has changed in the last few decades. Uh, we've also seen a lot of progress in health. Here we have figures on maternal mortality rates, but also a comparison of, uh, looking at the universal health coverage index and improvements in that, as well as, uh, improvements in the rate of out of pocket expenditure that households have to take on to attain healthcare services. Of course, there are a lot of challenges still at play and some of these will be reflected in the future graphs in this presentation. Uh, we know that learning is very different than enrollment. So here we have some statistics on the fraction of primary students who are meeting minimum proficiency thresholds in math. By income country grouping and what you'll notice is uh lower income countries, far right, light blue, have the lowest shares of primary students who are proficient in math for the grade that they are attending, and we will also return to this challenge in a moment on the measurement side. And in health, er, we've seen an increase in coverage and the provision of healthcare services down to the community level, even. But we know that in low-income countries, in particular, health centers lack structural capacity. This is equipment, machinery, medicines to provide quality service delivery. And here we have results from 5 countries on the percent of facilities with an item in stock on the day of the survey, and you'll notice that in some countries, the coverage rates are very low. Uh, but even in the best performing countries, it's, you know, it can be on the order of, uh, 25% of facilities who do not have particular types of equipment that are standard for these, that, that are standards for these health centers. Uh, and very recently we've seen COVID be a huge disruptor both for education and health. And here we just have two figures to, uh, display this fact. On the left you have, uh, simulations of the decline in learning that's taking place due to school closures. Uh, and on the right we see, uh, tracking the utilization of services during the epidemic. We see big declines in some countries on the use of maternal and child healthcare services during, ah, the COVID pandemic. In this presentation, we want to talk, we're gonna bundle it into, uh, we're broke it up into three parts. First, I'm gonna walk you through features some innovations in measurements of human capital. We talked a lot in the last presentation about the importance of data and concepts of measurement and methods of data collection, uh, for poverty and inequality, and we will reflect on that as it pertains to human capital. And some of the things that we've seen, the trends we've seen in the last 10 to 15 years that have improved the data, uh, the data landscape in terms of concepts of related to measurement. Uh, Patrick will walk you through a select set of evidence areas on policy interventions in area, in areas related to human capital. And then we want to kickstart a discussion. We threw out some areas that we think are research priority areas moving forward. Uh, one big caveat is the focus of this presentation is on children and young people, but human capital, and I, and I've, uh, borrowed this figure from colleagues in human development, a forthcoming COVID, uh, human capital report by the Human Development chief economist, Norbitcheid and his fantastic colleagues. But to demonstrate that human capital is a lifelong process, so we could have as well had a presentation that focused on working age adults or older people, but because of time, we're going to focus on young, young adult, young children, children and young. Adolescents, I would say. OK, let's talk about measurement. Uh, people that know me know I'm a real data nerd, so, you were not gonna hear, not gonna not hear about measurement in this presentation. So a few highlights on the measurement side. We know that schooling is not the same as learning, and we've seen a huge transition, I would say, uh, which is featured in our world development report on education 2020 led by DM Filmer and others, um, in terms of thinking seriously about moving away from. Enrollment and attainment as the only measures of educational outcomes happening. Here we have a figure on the concept of learning adjusted years of schooling, lays. You can read a lot more about this in the World Development Report. But the point is there when you adjust years of schooling by the quality of that schooling as reflected in test score data. Uh, learning adjusted years are much lower in some countries, and there's a huge amount of variation as to how much lower they are. So, you, and there are a lot of details underlying this calculation. You can use different, different benchmarks for this calculation, but the bottom line is, in many countries, the quality of education is translating into lower levels of learning that may be hidden by just looking at attainment or enrollment. Other measures of this include, or other achievements in the measurement of this include the harmonized learning outcomes uh work uh that KCP has also supported. A second sphere of a real change in how we think about measuring human capital is an increased use of the concept of social-emotional skills and measuring that when we do, uh, studies on human capital. And here I'm, Highlighting the step skills measurement surveys that the bank kickstarted about 15, maybe more than 15 years ago, um, which looks at various dimensions of social-emotional skills on the non-cognitive side of human capital development. Um, among them include, for example, concepts of grit, decision making, um, big five measures. And what you see on the right-hand side is what was interesting about the SEP skills measurement program is that it went forward in a number of countries to try to consistently measure this concept and look at how these measures of socio-emotional skills relate to performance in school, ultimate earning outcomes in the labor market. There are, there are huge challenges with measuring social-emotional skills. Among them are. Uh, so we're seeing a lot of different ways of measuring them, but there's a lot of problems with the validation of these, these scales because they are based on Western populations, and I'll give you one example from work that was KCP supported that shows when you look at the relationship between a measure of consciousness and um outcomes, predictive outcomes on income. What looks very important in the US does not look important in many other countries, drawing into question what this measure is actually capturing. And then finally, I want to talk about reflect on quality of service delivery, um, starting with the qualitative service delivery surveys that were rolled out and then moved into service delivery indicator surveys to really kind of capture, uh, quality of services and, uh, understand constraints to quality of service delivery. Here I have, uh, we're looking at work that looks at comparing different ways of measuring teacher effectiveness. This is work that Dion and others have produced. We also have seen a shift to using standardized patients and mystery clients to understand uh how providers actually treat patients inside health clinics. This is work from India on TB patients, looking at the variation of correct reporting depending on the time of day, using mystery clients, mystery patients. And then, uh, I wanted to also mention in the policy research report produced by the, by DC, by my wonderful colleagues in DC, the concept of effective coverage, which is a combination of both, did a, does a patient need healthcare? Do they go to get healthcare and what is the quality of the healthcare service that they get? And uh, that combination, um, the idea of effective coverage is really focusing on this. Case cascade between need, coverage, and quality. And finally, we can bring these ideas together into the concept of the no do gap in health services. So, uh, we, we know that quality of service delivery might be right, might be quite low, but we often don't know is that constraint on the structural side with what equipment is at the center, on the knowledge side of the providers, or on the translation of that knowledge into the service that they deliver. So, for example, here, I have one illustration of. The no do gap from work in China, colleagues did with KCP support that compares what happens when you look at TB treatment between what providers would tell you with a vignette, a hypothetical case, and what happens when you bring in a mystery, a mystery patient into that office, and big variations. So, we see in the first two rows, in fact, they do very well in the vignettes, but the actual action at the clinic level is, is a much lower performance rate. Um, I also point you to the policy research report I mentioned a moment ago has extensive, uh, discussion on the node gap also. OK, now I'm gonna turn it over to Patrick, but I'm gonna control the slides. So, hi, Patrick. Hey, Patrick, um, I can't show you this, the, the, the sign that is 2 minutes. I'm really gonna ask you to, to monitor your own time if you don't mind and keep it to 15 minutes. Thank you so much. I'll do my best. Uh, thank you, Kathleen, and, and thanks everyone for, for being here. Um, so Kathleen has highlighted, you know, huge improvements in the measurement of human capital. Um, but together with that in the last 15 years, we have seen really a boom, uh, um, in studies looking at the effectiveness of policies to improve human capital, uh, using many of those same measures. So in the next 15 minutes, uh, what I'll do is I'll share some insight from That research. Um, but the research for the KCP is really rich and so as Kathleen mentioned at the beginning, uh, it's really impossible to give a comprehensive overview. So we have chosen a select set of studies to highlight the broad evolution of research and, and we hope you find that helpful. Uh, in doing this, we want to highlight the topics we are not gonna discuss, uh, related to, uh, early childhood development, teacher accountability, uh, pedagogy, or adolescent or reproductive health, not because they are not important, but just because we, we have limited time. But we hope we can maybe hear from discussions on this or, or engage in the discussion on, on broader, uh, uh, on these broader topics. Uh, we've organized the second part of this presentation, uh, to, um, um, um, start with specific interventions for households on the demand side, um, and then slowly look at, uh, interventions on the supply side before moving to broader analysis of, of reforms on systems and Markets, uh, which itself I think highlights a little bit the direction, the trajectory where, uh, the policies have taken over the last, uh, 20 years. So, uh, with this in mind, the first, um, kind of specific, uh, household level demand site intervention is, is cash transfer that has received a lot of attention. Um, and specifically the PRR reports, um, supported by KCP on CCTs about 15 years ago was really influential in laying out the theoretical underpinnings of CCTs but also summarizing early impact evidence from, uh, uh, Latin America and the first generation of CCTs. Um, I think we have learned a lot as CCTs are expanding around the world on how to optimize their design with a few key insights from the research. Uh, the first key insight is that CCTs can really improve the outcomes on which, uh, they are conditioned. Uh, we have seen a lot of this, uh, on school enrollment and preventive health, but also that has been used in other, uh, domains, for instance, trying to incentivize, um, uh, safer sexual behaviors to reduce STIs, um, among other outcomes. But the literature also has highlighted trade-offs, uh, and you can see here on the right, a graph that shows that in Malawi, uh, a CCT was pretty effective at increasing enrollment, but that an unconditional cash transfer was more effective at delaying the age of marriage. And so in a way, um, when we face this trade-off, uh, we may, or to avoid those trade-offs, we may need to combine different types of cash transfer modalities in order to achieve impact on multiple dimensions of human capital. Now, the literature uh has also consistently shown that there are financial constraints to access, um, uh, in multiple, uh, uh, for schooling but also for, uh, access to health services. Another example of that is in the WDR on learning which Kathleen already mentioned, um. Um, showing that as, uh, user fees or school fees are removed, uh, enrollment increased quite substantially, um, also pointing in the same direction as the literature on scholarship scheme, you know, pointing to the fact that, uh, uh, demand for services is, is sensitive to, to, to cost. But of course, we know that addressing financial constraints is not sufficient, it's not sufficient for the most vulnerable kids, uh, and households, and also it's not sufficient because it doesn't lead to improvements necessarily in learning. One of the intervention that has been presented as potentially improving both access and learning is school feeding. Um, there, uh, recent meta-analysis have shown that school meals can increase health and nutrition status, in particular, if they are complemented by, uh, micronutrients or, uh, deworming interventions. Um, and interestingly, also shown that, um, they can improve learning. You see on the right here a graph showing, um, effect on, on learning of, uh, duration of exposure to school, school meals in India showing a, a substantial positive effects. Um, but behind this, uh, there is the research also highlighted the need for, uh, complementary school inputs in order to achieve these learning ga gains. So in a way, um, in both those dimensions, we have seen, uh, the potential role of complementarities being important in human capital formation and, uh, we have ongoing work at DAN in collaboration with the World Food Program, uh, to try to look at these, uh, synergies and complementarities in school feeding a little bit more consistently. In the spirit of thinking about complementarities, another key result is that providing inputs alone um is not sufficient to improve learning. Um, and here, um, results from studies on providing textbook or on, uh, you know, providing one laptop per child have been pretty clear in terms of the, the limited effects on earnings, on learning, sorry. Um, and another example is a KCP study in, in Lagos that shows that Providing e-readers does not necessarily improve uh um uh learning unless um it is, um, it includes, those e-readers includes material from school curriculum or compensate for lack of inputs in this case, textbooks, uh, in school. There is, however, uh, recent evidence that's a little bit more encouraging on the use of, um, of technology and, um, uh, uh, to, to improve learning, in particular when software or edutainment, uh, is designed to uh really be engaging for children and teach, uh, uh, at the level. That, that they need. And so here we have some recent results from northern Nigeria that are pretty powerful showing that combining aspirational videos for parents with literacy apps for um on smartphones for children can really boost learning pretty substantially. Now, moving on to schools, um, there's been a, a huge amount of progress, uh, over the last decades, um, due to many of you in the room here that have worked on this topic a lot, uh, to highlight what are the types of policies that are effective to boost learning in school. Um, Kathleen, highlighted the learning crisis among children, um, but another really striking, uh, finding is that teachers themselves often do not master, uh, the subjects that they teach. Um, and you can see here on the, the black line on the, on the graph shows you. Um, the effects on, uh, learning, uh, among children of, um, um, ensuring that teachers that's, uh, at school have themselves, uh, acquired the level of learning equivalent to the end of primary school, and that in itself could, uh, substantially improve learning even more when this is combined, uh, looking at the dashed line here with an increase in time spent by teachers teaching. Um, now, that these are two levers, uh, to, to improve, uh, learning. Of course, there are others. Um, we know that, uh, many teacher training program or professional development programs are very far from being optimal and that improving this along with, uh. Pedagogic uh sorry, pedagogical approaches, um, that includes, uh, structured lesson plans or, uh, teaching at the child level, um, can be really, uh, great investments in boosting learning, and this is something maybe, uh, Rukmini in the discussion may, may touch on. Um, so as mentioned, the, the time spent by teachers teaching, um, is low in, in many developing countries, um, but there are really multiple reasons for that. Um, and I think interestingly, when teachers are asked, um, if they think it's OK to be absent, many of them here on the left say that they, you know, it's OK to be absent if the kids in school have something to do in their absence or on the right, if, uh, they are absent to do something helpful in their community. And so that's kind of um insightful because it shows us that um it's, you know, it's hard to shift uh teachers' efforts, um, incentives, financial incentives may help but um it may only be part of the solution and other approaches to improve accountability, uh, for instance, to scorecard or other interventions may be needed and uh again, this is something that uh Ritva may have a lot more to say, uh, in the discussion. Um, related to, uh, the question of incentives, um, many, uh, policies or, or interventions have tried to provide grants to schools to improve their performance, um, and we can illustrate some of the opportunities and challenges with that from using the data from a study from Indonesia. That shows here on the right that um performance-based grants to school improve student learning in secondary schools, so this is column 4 here, but uh actually the same grants do not improve learning in primary schools and this is the column 2 here on the left. Um, so this is kind of interesting because it illustrates that, um, uh, the same policy, the same type of brand designed a similar way in a similar setting actually lead to very different behavioral responses in different types of school. Um, and so kind of illustrates the need to understand the, the behavioral response from, uh, providers when those kind of, uh, grants are, are, are, are offered and kind of ties to a, a broader literature with mixed results on the performance incentives. Another set of intervention that's uh is trying to, to shift behavior um is uh the provision of information related to schooling. Um, and here we are showing you some results from a study in Mozambique, uh, that shows that providing information to parents about your children's attendance in school, uh, not only increases attendance itself but also uh increases learning in this case, uh, math scores. And what's interesting is the mechanism is that it's through parents' behavior, through parents basically getting more involved in the monitoring of their children, uh, which in itself then leads to, uh, them learning, uh, more in school. Uh, and similar pathways have been highlighted in other studies, for instance, in, in Angola. Of course, the provision of information can also have broader effects on education markets, um, and also induce, uh, behavioral responses from, uh, from providers and so for instance, in Pakistan, a very influential study showing that, uh, providing information on the, uh, on average test scores in school actually leads private providers to reduce fees, uh, because they don't have to signal their quality to, uh, to fees anymore. Um, the, of course, the, the analysis of, of private school has, um, um, taken a lot of attention in recent years trying to see if they can be part of the solution to improve learning. Um, it has been the case in certain areas. Here we show you in Uganda that's, uh, providing vouchers for students to, for, for private schools, uh, actually improve student score pretty consistently. Um, uh, you see the distribution shifting, um, but behind this, there are, uh, part of this is due to gains in enrollments partly driven from, uh, students with higher socio-economic backgrounds. So, of course, um, some distributional questions here on the, uh, the effects of those policies and who they might benefit to, uh, at the end of the day. Um, I will wrap up with two more, uh, topics. Uh, the next one being on, on health. Um, Kathlin already mentioned the, uh, uh, very exciting PRR that was, uh, um, issued, uh, a few months ago that analyzes how changes in incentives, uh, introduced to system-wide reforms affect service quality and outcome. And then specifically looking at pay for performance, uh, based on the quantity or the quality of services delivered. Um, what's interesting is that, uh, I remember many years ago we were all very exciting seeing, seeing the first results from Rwanda on performance incentives that were pretty positive. Um, but the PRR actually updates, uh, our overall thinking on this showing that, um, performance pay actually has little, little impact in relatively under-sourced health systems, and this is because, um, there's a lot of, uh, uh, issues in, in health systems that are beyond the control of the frontline providers that may not be able to control or, or influence, uh, issues related to the supply chain, to the quality of infrastructure or equipment, uh. Or even demand side constraints, uh, which will be needed for the incentives to be, uh, effective and so what the PRR illustrates and the graph here on the right illustrates is that, uh, direct facility financing, um, with autonomy and, and, and accountability may actually be just as, as effective, uh, uh, in terms of, um, uh, uh, achieving gains in outcome, uh, but doing so at lower cost and, uh, with a relatively easier implementation. So the, I'll conclude my uh very brief overview on uh uh selected uh topics on human capital by uh talking a little bit about skills for schools, uh, for youth uh that have exited schools. Um, we know that the evidence on training programs and in program, employment programs is, is very mixed, um, but I want to highlight some, um, innovation that have been partly supported by the KCP that, uh, shows some, some promise, uh, uh, that have shown promise in recent years. So one example is interventions, uh, behavioral intervention. trying to target uh um behavioral skills, in this case, personal initiative tying to Katlin's earlier discussion of, of, of, uh, of those behavioral skills, um, showing that this may be more effective than interventions trying to teach traditional business training in terms of improving earnings, uh, for microenterprises. Similarly, uh, evidence, recent evidence showing that interventions combining on the job training with training in vocational training centers in Colombia or through dual apprenticeship in Cote d'Ivoire can be very effective in improving skills and then raising productivity uh of young people when they enter the workforce. Um, but beyond all of this, uh, also in this topic, I think we have learned that we need to think carefully about the trading markets, uh, about the incentives for training provision in firms or, uh, in public, uh, in the public sector, uh, the scope to improve demand, and overall, I think as in many of the cases we just quickly discussed, um, that these policies cannot just be assessed, uh, in partial equilibrium. So, um, uh, I think we're almost at the end of the time. I think, I don't see you beyond waving yet, but, uh, you're probably close. Uh, I just want to kickstart the discussion by highlighting a few, uh, uh, topics that we think are important for future research and see, uh, hear a little bit from you, uh, whether those are the right priorities or there are others that you think are more, uh, uh, higher order. Um, one of the big questions or the big trajectory we see on the policy side is towards more integrated or multifaceted interventions, um, um, bundling, uh, various types of, of components including packaging intervention. of the demand on the supply side together to improve outcomes. So this is very exciting but it also raises a lot of questions on how to optimize those uh multi multifaceted intervention and how to uh tailor them across different settings and, and, and target populations. Related to that, uh, of course, we need a better understanding of health and education markets. We've made lots of progress over the last 10 years, but, uh, as, um, uh, the interaction between public interventions and private providers, um, is key to understanding overall how, um, behavioral shifts and, and how, uh, changes in outcome, um, uh, pan out. Uh, especially as we move more and more towards system-wide policies and, uh, as we are pushed, uh, rightly so at the bank to think about scale and how, uh, impacts, uh, uh, kind of aggregates to achieve higher outer policy goal at the, at the economy level. Of course, um, the last few years have, um, showed us that, uh, you know, uh, we've faced major challenges in terms of human capital, in terms of COVID, and are also facing new challenges related to climate and understanding how those are affecting, uh, both human capital, uh, um, accumulation but also how, uh, policies to address them also then Uh, improve human capital and intersect with the human capital formation will be critical moving forward as, uh, is the, the focus on technology or the hope that with technology, we can, uh, achieve better outcome by facilitating the work, facilitating the work of providers as well as, uh, facilitating behavioral change among, uh, among, um, um, uh, users. The last thing I will, I will say today is that um as you've seen, there's a huge amount of work going on on both on the data and on the evidence side and given this boom, I think one of the challenge we face is really being able to synthesize effectively this evidence for policymakers so that they can use this for policy actions and I think uh beyond the generation, this uh synthesis of, of evidence moving forward will be critical to uh have even more policy impact for, with the work that we do. Thank you so much. Thanks Patrick, and thanks Kathleen as well for those really Interesting presentations. I mean, I, I, I, I do understand that, the, the, the, the selection process for, for, for what to highlight was pretty drastic, but um I think we can all agree it's a really rich agenda, and I think this gave us a really good flavor of the kind of work. Uh, we have, uh, that kind of work that's going on. Um, we have a really interesting, uh, uh, and impressive uh set of panelists uh with us. Uh, we have one person in the room who's Emily Gustafson Wright, who's a senior fellow at the Center for Universal Education, Global Economy, uh, and Development at the Brookings Institution. Uh, we have Pamela Jacquiela, who's associate Professor of Economics at Williams College. We have Rukmini Banerjee, who's Chief Executive Officer of Pratham Education Foundation. And then we have Rick Vereinke, professor of Practice at Aalto University, Helsinki. Um, I'm gonna go in the order that it was given to me, uh, just to, it's somewhat random, uh, which means we're gonna start with Pam. Uh, Pam, are you online and can you come in? Yes, I am online. Great, we can see you and we can hear you. Uh, thanks, go ahead. Fantastic. Thank you very much. Um, so thank you for having me. I'm delighted to see so many familiar faces, uh, online. Um, so I'm going to comment a bit on the educational side of human capital because I feel woefully underqualified to talk, to speak to the health and healthcare side. So the talk we just heard did a great job of emphasizing the tremendous progress in education that we've seen over the last 25 and even 50 years. When we set out the Millennium Development Goal of achieving universal primary education, um, 20 years ago, uh, we've now achieved that. We've come near to achieving that goal. It's not no longer, it's not, uh, universal, but the international community has made, made remarkable progress, and, uh. We've now shifted toward thinking about, uh, as access has become less of a constraint, we've shifted toward thinking about learning, uh, and in particular, uh, the The presentation we just saw was emphasizing policies to improve teacher performance in the classroom, technological and pedagogical innovations, uh, to increase the likelihood, uh, that access to schools and effort by teachers translates into learning, and there's also a broad set of policies on early childhood education and interventions to help, uh, prepare students to be ready when they enter the classroom. These policies are all extremely good and constructive, and I'm not here to criticize, uh, to criticize them. And there, as Kathleen and Patrick rightly point out, there are a lot of other policies that they didn't have time to, to discuss in the, in related domains that have been really successful over the years. But the task in these sorts of situations is to have an interesting take, so I'm going to take a stab at that. Um, and so I'm gonna start by sort of framing why we've shifted from access to school readiness and learning, um, in recent years, and that's because, you know, when we, um, Started to succeed with access, uh, and come near to achieving universal primary education and higher enrollment, uh, in secondary education, uh, we saw that access alone was not enough, you know, so when Nigeria and Indonesia were building schools in the 1970s, when many, uh, African countries We're eliminating primary school fees in the 1990s and the early 2000s. The idea was that access would, would be enough, would translate into the building of a skilled workforce, uh, the skilled workforce that countries need to fully participate, uh, in the modern economy and achieve inclusive growth, but it wasn't enough. And, you know, this shouldn't be surprising. It's not much of an exaggeration to say that the whole narrative of post-war development is a series of uh attempts to try things, and then learning that they're not enough. Um, and so, you know, it's a little bit like we're that Far Side cartoon where we have a physicist is at the blackboard and you have some equations on their left, equations on their right, and in the middle, they, they've written, and then a miracle occurs, you know, we try policies and then we hope a growth miracle occurs. Um, and so we've turned a lot of attention to learning and early childhood and school readiness, and we're hoping that now this focus on human capital rather than access to education is going to achieve the growth and prosperity miracle that we are all hoping for. And the suggestion I want to make today, and it's clearly not a new idea. Particularly my idea is that everything we're doing focusing on the building of human capital is good, but it's really unlikely to be enough, um, and it may lead to, you know, significant impacts that are even economically and developmentally meaningful, but it's not going to, to get us as far as we need to go. And a big reason for this, and this is something that Patrick kind of brought up at the very end, is that there's a tendency to focus on the supply side and not on the demand side. Um, and so I just want to talk a bit about, um, how we need to frame our entire discussion of human capital, not only in terms of that supply side, but in terms of political economy constraints and social constraints that are going to limit and are currently limiting both the incentives for individuals and households to invest in human capital and the relationship between human capital that is built, uh, in the classroom and in youth, and how that translates into Outcomes in the economy, participation in the economy, and subsequently growth and shared prosperity. And so I only have a couple of minutes. I'm gonna try to do this quickly, and I'm gonna do it through a gender lens because I'm always looking at everything through a gender lens. Um, and so I want to make two points. And so the first is that over the last 25 years, we've seen a major push to increase girls' educational attainment, and Kathleen talked about how now we're actually thinking about how boys are lagging behind. We've been remarkably successful. Um, girls' enrollment in primary and secondary schools have increased dramatically, both in absolute terms and relative to boys. And of course we know schooling isn't learning, uh, but it is correlated with learning. So sometimes we can beat the schooling isn't learning drum a little bit too hard. It is correlated with learning, and we've also seen big increases in female literacy and enrollment in tertiary education, which requires some measure of human capital. And so all the evidence suggests that both in absolute terms and in relative terms, We've seen pretty substantial increases in women's human capital over the last 25 years, and that's in absolute terms and relative to men. And so we can think about what has happened with women as a little bit of a case study or quasi-experiment and the broad impacts of increasing human capital for a group of people. And the bad news is that over the same time period, we've seen remarkably little progress in terms of women's labor force participation or their ability to achieve fulfilling careers in many countries and contexts. And that's a real problem for growth and development. So economists estimate that in the United States, between 20 and 40% of US growth over the last 50 years has come about because women and minorities who were excluded from high-skilled professions have been able to enter them. And there's no reason to think that that type of misallocation isn't just as big of a problem in developing countries. Um, but the problem is that as we've seen these big increases in women's human capital, we've seen they are not predictive of women getting to enter the labor force, increases in labor force participation. Um, and so that brings me to my second point, which is that when we're thinking about human capital, and in particular thinking about investments in human capital that we hope we're going to translate into impacts on the broader economy, we really need to emphasize not just the learning outcomes, but also, uh, thinking about building data, building knowledge about the demand side constraints, and I'm gonna look at this through the gender. as well, because for women and girls, we know that at every stage of the process, even though women and girls are in school now, as much as boys in many places, girls are facing, uh, these additional challenges that prevent them, that reduce their incentives to invest in their human capital, uh, and it reduced the extent to which that translates into their full participation in the economy. So girls face an additional burden. burden of domestic care work in the home. They face discrimination from teachers who nudge them into lower skill tracks and professions. Uh, they, we know that they distort their educational choices to try to avoid risks of street harassment and gender-based violence, uh, that they face a lower expected return to investing in their own education because they might get kicked out of school in many contexts if they get pregnant. And then of course, they worry about the marriage market consequences, and the incentives to invest in education are reduced when there aren't complementary, uh, both, you know, access to child care, things that allow them to balance their family responsibilities with fulfilling careers in adulthood, and also social constraints from their families and their spouses that limit their ability to really have meaningful careers, if those might come at some cost to their household work. So none of these things is new to anybody in this room, but I really wanted to sort of nudge us to when we talk about human capital, not just think about what we're doing on the supply side, but really think about the need to build knowledge on the demand side and in particular how households and individuals are perceiving the choices and the trade-offs they make and the constraints that they face. And of course, the, you know, the World Bank has been a leader in Not just, you know, service delivery indicators, but also individual household surveys, and there's a lot of great innovation going on in measurement of beliefs and attitudes and perceived social constraints. But I really would like to encourage us to think that, uh, to think about the need to, to center those things when we're talking about human capital. Because what we've seen is the good news is when households perceive opportunities for women and girls to enter the labor force in a productive way, they respond rapidly. They invest in girls' education, they change their attitudes about women's labor force participation. We've seen that with business process outsourcing in India, with uh garment manufacturing in Bangladesh. Um, but the point is we need to think about in a context specific. way, what are the binding constraints, not just on access and not just within the classroom, but that are preventing, uh, not just women and girls, but lots of groups from being able to have a pathway from building human capital to contributing and fully participating in the economy and think about building our measures and our interventions to work on that margin as well. So I'll stop there. Thank you very much. Thanks, Pam, for that was really interesting and useful, uh. Contextualization and, and slight challenge to the, to the way uh we're thinking about this issue of building human capital and, and, and the benefits we might, uh, reap from that. Um, next up is Rukmini Banerjee. Rukmini, are you on? Yep, uh, I'm here. Can you hear me? Hi, Rukmini, we can hear you and we can see you. Thanks for being here. Great. Uh, thank you for having me. Um, I really enjoyed the, uh, speakers before me, and I was just thinking that, uh, you know, we are celebrating, I guess, 20 years of, uh, evolution of knowledge, and, uh, especially as far as the learning crisis and what is happening with children and schooling and learning. Uh, it almost seems like, uh, you know, this, uh, difference between schooling and learning. Schooling isn't learning, overambitious curriculum leading to negative consequences. All of these things now seem to be, you know, part of the knowledge, at least of a lot of people in this room, as well as in kind of the highest policy circles. Uh, but when we talk about knowledge and what is known changing life on the ground, are there still things to be done? Uh, we just released in India the, uh, I've lost track myself now, I think the 17th annual status of Education report, the UA report, about 10 days ago, uh, where, uh, we are talking about, um, you know, where are the basic learning levels in India, and obviously because of the last two years of the pandemic, there is a certain amount of decline or lack of recovery, you know, in different places across the country. Now, whether you think about learning adjusted years of, uh, learning, adjusted years of schooling that the World Bank is coming up with, or other kinds of measures, uh, my question often to myself as well as to our colleagues in India is, to what extent on the ground are parents, are teachers, Accepting that this big gap that they see between learning and schooling is something that they need to work on as, uh, you know, as children go through the system. Uh, so for example, there's been a lot of hand wringing on the learning loss during COVID, but I think there is, uh, other than in very enlightened circles. There is less of a realization that that dip has come on top of a much longer trend of low learning levels in the world, and therefore we need to get back to the structural reasons for why this is the case, uh, and, uh, you know, work on the solutions. Some solutions exist. Patrick and others have laid them out, but work on the solutions which are actually longer term solutions, not just as a response to COVID. So if I think about the way that we have worked in India in kind of understanding the problem, it started with a very simple way of assessing kids using a very basic tool which didn't go at grade level. And I think we also underestimate the, the mindset that almost everybody has, at least in a low income country, that schooling means a linear progression year on year through a system. And as you rise through the system, that is spend more and more years in school, you know, somewhere you are different from those who have spent less years in school. Despite the fact that the data for 10 years is showing that if you didn't get onto the treadmill at the right pace early in your school life, you actually you're not going to make it. So we have enrollment, for example, if we look at enrollment in India in 8th grade, uh, and if you remember that we have about 25 million kids in each single age group. The 8th grade enrollment in say 2008 was about 11 or 12 million nationally, and that number in 2018 or 201919 is like 23 million. So while learning levels have not moved, enrollments and completed years of schooling have gone through the roof, and the implications of that on what happens beyond, you know, I think is understood in a research sense but not understood in a real and a practical sense on the ground. So my one question that I would like to leave on this panel is there is a lot of research results, you know, we have uh panels and committees which are compiling the most cutting edge research, but how long does it take for the key elements of our knowledge to actually get down onto the ground and what methods or what, uh, what mechanisms enable that to happen? Now, in some, in a very perverted way, I feel COVID could have been that instrument that it shook everybody up in such a big way that you are forced to think in different ways. And, you know, just the fact, for example, if I quote from our latest AA report that there are kids who are in 3rd grade in India right now who actually had no schooling. They came in. Our schools were closed for 2 years. They moved into 3rd grade in April of 2022. And so by the time our asset measurement happened, which was about 6 months later, they actually had only 6 months of schooling. And yet if you look at the basic level of learning. There is 20 to 25% of kids in the country are at grade level, and this is despite schools being closed. So you would think that even these very basic facts lead you to think about what is it that we are doing in school or not doing in school that is leading to this situation. And if schools have been closed for two years and we see a drop in learning for sure, but you don't see zero, what does that teach you? What does that mean? Uh, and so, you know, I would really like, uh, you know, I'm sure there are others who are thinking about this, about how does research get to the ground, uh, how about the fact that we have solutions, many of which have been laid out, and, you know, the one that we have worked on a lot, uh, teaching at the right level, evaluated many times, shown to be very effective, but there isn't like an automatic use of that data despite the fact that the World Bank and FCDO and whatnot have it as a good buy, uh, not a goodbye, but a good buy, uh, uh, out there. We don't see like a massive sudden huge demand from governments to use that solution, which is well known. So where are these gaps and what does the knowledge producing community need to do to a knowledge digesting community who can then put some of these uh things that are now well known into practice. So let me just stop there because maybe others have better solutions to offer. Thanks, Rukmini. Another really insightful set of challenges to, to how we think about um the, the, the, the production of this research and, and where, where it leads, uh, um, and how it moves to practice. Um, next up, we have, uh, Emily, Emily Gustafson Wright, um, who's here in person. Thanks, Dion, and uh thanks everybody, great to be here. Um, so I'm at the Center for Universal Education at the Brookings Institution and um we do, uh, I think a lot of digesting of the research also and I think that that's a really important uh role is, um, you know, translating a lot of these um really important findings from the bank and, and elsewhere, um, as well as our own research, um, and including a number of areas such as scaling and gender equity. Um, what we call a breadth of skills. Um, there was a discussion on the slides around non-cognitive skills, we call them a breadth of skills, systems transformation, ah, family engagement, which was also mentioned, um, just to name a few. I primarily focus on costing and financing education and early childhood development, and for the past decade or so, I've been working on results-based financing. So I will focus, um, my comments on results-based financing. Um, and, um, for human development or human capital, um, and I'll focus on four main, uh, 4 main areas. So first I'll comment on what was presented with respect to the evidence and what wasn't. Um, and then I'll highlight some nuances that I think are critical to consider when looking at the evidence or lack or lack thereof. Um, third, I'll describe the current landscape of results-based financing in education and health and, um, what we might expect to see in the coming years. Um, and then finally, I'll suggest some areas for, um, for further research. Um, so we heard some evidence on policies to improve, um, human capital. Um, this included two forms of contingent financing, uh, conditional cash transfers and performance-based grants. Um, the results are mainly positive but, but mixed, um, for both of the mechanisms, um, that were presented, um, in particular in terms of short versus long-term effects, um, but also for different, um, populations. Uh, fun fact, I actually started my career, um, over two decades ago, working on conditional cash transfers here at the bank. Um, the bank, uh, was skeptical, um, I would say, um, IDB was, I think, a bit more positive at that time. Uh, we were working in Brazil with the then Bolsa Escola program. Um, so I guess I was, um, you know, of course this is the, this is the lens that I'm looking at this through, and, um, I was surprised not to see a little bit more presented on the topic, of course there wasn't a lot of time, um, maybe also this is just because, um, the, um, the research that's been done at the bank was funded by other, other, other trust funds or other, um, um, other funders, um, but there is, there has been quite a bit of work done at the bank, um. For example, under the REACH Trust Fund, um, also an IEG review of P4R, um, and then, um, some work done by, um, GPRBA as well. Um, and of course lots of research that's been done, uh, outside of the bank on results-based financing. So I'm not gonna go into the details, um, about the existing evidence. Rather, I'd like to make an observation about the analysis of or the conversations that are being had about results-based financing or RBF literature, and I have to admit I'm sort of using this platform to address a bit of a pet peeve of my own. Um, so there are quite a few different, uh, contingent payment or RBF mechanisms, um, that have been applied, um, to both education and and in the health sector, uh, over the years. But what I've seen is often the findings uh tend to be bunched together, uh, kind of under one umbrella, um, and I, and I feel that little attention is being paid to sort of the nuances of the different types of mechanisms and or the projects themselves. Um, I feel like this has resulted sometimes in sort of blanket decisions being made around our RBF, um, as a potential tool to address some of the challenges that we've talked about today. Um, and there are in fact considerable differences between the different types of RBA, um, and, uh, and we've seen some of, uh, you know, research done about this over the past couple of decades. Um, so the primary difference between the different mechanisms is who bears the financial risk, um, or who, who won't get paid if the results aren't achieved. So that can be national government, subnational, um, service providers, or schools, uh, or health, uh, centers or teachers, um, individuals in the case of, um, CCTs, um, and in the case of the more recent, um, social development impact bonds, which is where my research focuses, um, impact investors. So, who, who isn't going to get, who's going to get paid or not get paid, um, really matters. Second, Um, I think the governance structure or the power dynamics, um, of these configurations matter. So who's making the decisions about the metrics and the threshold thresholds for payments, for example. Um, 3rd, um, the, the, the measurement and evaluation matter. So are the results outputs or outcomes, we've had a 10, we've tended to see, um, a greater focus on outputs in traditional results-based financing. We're seeing a shift towards, uh, farther to the right in the results chain towards outcomes in social and development impact months, for example. Um, also, you know, how and what are the thresholds for success, um, determined, um, what is the counterfactual? Is there even a counterfactual? Um, and then finally, obviously the context matters, um, and what happens around the project in terms of, um, technical support and, and other elements. Um, so, uh, third, what is the current landscape and what, uh, what are the trends that we are seeing? Um, so overall we are seeing, um, quite a bit of an upward trend, um, in the use of results-based-based financing. I, I do think that this, you know, kind of, this comes from the, the, the access plus learning, um, a shift in focus to outcomes, constrained budgets, um, you know, I think that the RCT movement, I think, was part of this, um, this push perhaps. Um, and, um, so I'm, I'm not going to give sort of the, it's, it's very difficult to measure sort of how many, uh, because there are so many different types of results-based financing, kind of how many are we seeing in the world, it is on the rise. I'll focus, um, on, um, social development impact bonds, which, um, is, um, an area that I, again, that I focus on and maintain a database of all of the projects in the world. So this is just to give you sort of a sense of, of what we're seeing, um, in the global landscape. So, Um, there are about 240 impact bonds projects, impact bond projects globally, um, that have been contracted. Um, the majority of those are in the employment and social welfare sector. Um, however, the education sector was the one that, that was the sector that grew the most, um, in the past year, um, an increase of 12 projects, bringing the total number of education projects in the world to 40. Um, and this included the launch of 5 projects in Sierra Leone. Uh, which are meant to reach over 130,000 students as part of the Education Outcomes Fund, which is a pool of funds for, um, to pay for outcomes, um, which I can speak more about if you're curious. Um, and, um, It would, yeah, so that includes a multiple, um, results-based financing projects within that, um, and then there were, um, there were only two new health projects, uh, in the past year, um, bringing the total of, um, health projects to 36. We, we saw a little bit of a slow during the pandemic, obviously, but there seems to be sort of, they seem to be ramping up now. Um, what do we see in the pipeline? Uh, quite a bit, um, happening, uh, on the horizon in terms of education. Um, both impact bonds and outcomes funds like the Education Outcomes Fund, um, will be launching another project, uh, in Ghana, um, reaching over, uh, nearly 200,000 students, um, the Back to school Outcomes Fund in India, a number of early childhood projects, uh, with Education Outcomes Fund as well as, um, independently, and for example, in Jordan, um, also in Chile, we see a number, um, in, in health, also a few, uh, a number that are in both in early and latest design. Um, all right, so in terms of the future work in this area, um, So on the one hand, we, we know, and I, I didn't really go into this, but we, we know that there is some rigorous evidence, uh, a lack of rigorous evidence. It's a tough experiment to do, right, um, sort of, uh, project using this financing, uh, mechanism, uh, versus not, um, but on the other hand, um, some would argue that it's really time that governments and donors stop paying for things that aren't delivering outcomes. Um, so if that's the premise. I posit the question should be, you know, when it's appropriate, how can results-based financing be done better? Um, how can systems be set up to ensure more efficient design, um, and implementation of results-based financing for human development. Um, I'm currently working on a paper together with Thomas Poulson from the bank. Um, we're exploring the four types of data needed for results-based financing. Uh, so in addition to cost data, cost of inaction data, and results data, um, we're, um, we believe it's really critical to invest in the monitoring data, the real-time data, um, for adaptive management, um, in classroom classrooms and health, uh, settings, um, with a focus on equitable, um, outcome achievement. Um, and then finally back to my final point, I, I think it's really critical to understand the nuances of these different projects, so, um, rather than sort of throwing the baby out with the bathwater or going full steam ahead, I think it's really important to, to consider these different factors, um, that drive the results within them. Thanks, thanks, Emily. Another sort of reminder of, of, of how we should be careful in, in how to interpret the results and what to expect from these different interventions. Um, last but not least, um, Ritva Reineke, uh, Ritva, are you online? Ritva was, was my former boss, and, uh, under various configurations. I'm very excited to see her today. Uh, over to you, Ritva. Thank you very much, Dion, and it's great to see you and many other familiar faces, um, and colleagues. And, and anyway, I have to say that uh it's great for me to join this celebration. When I got the invitation, I never knew this would still exist, Knowledge for Change program, because at the time when it was established, I worked at the World Bank's research department and One of my jobs was to establish this kind of a fund, not just by myself, but as a group. So it was uh the longevity is amazing and I'm, I'm, I'm really thrilled. So, you, you've done a lot of work over. The, yes. I just will uh make a few comments about the past, um, about research and policy, and then the future, especially based on, on what Kathleen and um Patrick presented. So just about the past and it's, it was fantastic to hear the various comments because at that time in the bank's research department, the way research was organized, we had a group that studied demand for education and health using household surveys, and then we had a public. Economics team that used the budget data. They did not interact. They were separate endeavors. And at that time when KCB was also set up, these two were integrated. And what also happened was that public economics in the bank really went micro. Um, and, and there was a new research program on service delivery. There was no analysis, uh, the research analysis on service delivery, and it was really great to hear, uh, the, the fellow panelists to say that, OK, we need to go back to demand, so something has been achieved, uh, for sure. I think the key issue was that it was two things. One, it was recognized that there is behavior on the supply side as well. There are incentives. There are services that have guards of people doing the service and, and therefore there's behavior and incentives. I think also the other thing was that this allowed the bank's research also to get closer to the operational work and the country's actual service delivery because we went beyond budget and I think the key issue was to realize also budgets actually are very different from service delivery as such. So at that time, We were after impact, we started the pioneering some of the measurements like leakage of funds or absence rates and competence of service providers, and some of you really presented very nicely. We also put together making services work for poor people, uh, WDR which Which, which was hoped at the time that it would spur further research in this area. And when I looked at your presentation, um, uh, Katherine and Patrick, I, I think there has been in research huge progress in the 20 years. So, uh, that's really super good. Uh, then when I think about impact, which was desired impact on actual service delivery, particularly, and I often think about poor countries, I I think definitely in education and it was, by the way, interesting that the whole panel also is uh people who have mostly studied education or worked in education. My, my view is that in healthcare, this is much less so, that kind of agenda, uh, much less in, in healthcare. OK, so that's about the past. Linking with policy, uh, research and policy, that has been, um, a kind of dear topic for me. I used to work mostly in um operations uh there in the bank and, and also a long time ago in UNICEF. Um, And in the bank, um, I had a chance to put this research, I mean, for my, for, for a small part, always, one has to emphasize, and into operations as Human Development Director in the Africa region, which was my last um assignment in the bank. And it's easier said than done, and I think Rukmini also hinted or said that very clearly. You have certain absolute winners and they don't go, uh, you know, to the system. Obviously. Um, it's also this thing, uh, that, uh, that I find that research agendas that are born in academic circles don't necessarily reach policymakers and sometimes they don't speak to them. And RE a program, the research on improving systems of education. Um, that many of you have been involved in which is a magnificent large program is now struggling with that, churning out this material to try and get the message through, and it will be very interesting to see how that works, but it is not very easy. I think in, we actually work together beyond in the Africa region 8D. And you were leading the kind of the think tank part of it in, in the region, and I felt it was really, really important to be in there, um, and, and we did work on regional topics, which for Africa, we at the time felt, and I think it, I feel Uh, continued to be population, youth employment, and service delivery, and it was definitely much better when one is right there and the agenda is set by the operational people that were in the bank and then the policymakers in the countries. In my own experience of the two decades in the bank, I was able to combine research and operation the best when I was a country economist working in Uganda. Being, being in their team entirely on the policymakers's terms, uh, it determined what was researched and, and then you were able to bring certain Uh, research that you knew in, into the system. So, in this linking of research and policy, it's a demanding task and my experience is that once you are in the policymakers's team, it somehow is easier to pull it in there. OK, 3rd thing, looking forward. So So interventions is, is the term that is used uh most. And I think it's the and the randomista movement is the name of the game today, and, and a lot has happened in that. I remember we actually in our team had Mike Kramer for 3 years, part time, not full time or part time. I never would have believed that that impact evaluation agenda is so dominant. What is much less dominant I feel is what came out of the World Development Report 2004, Making Services Work for poor people, which Gave a way to think about systems and, and that features in my view, much less today to look at the politics, the compact between providers and policymakers, management and client power and those elements that belong to it. Um, And I, I think um really two things in, in that, is that when you think of the system as a matrix, as the RICE program, for instance, effectively uses this very same uh framework things, you have These columns and rows in the system and there can be so many incoherent things in that um system and in those columns and in uh in those rows. And then when you have an intervention, when it lands in the system, it, it, it is totally unknown what kind of a cell is it. Is it in the middle of incoherencies or is it coherent in terms of that what you demand is financed, what you provide information on tells about performance, things like that, or motivation. So, like in my own research, so, so two things. One, in my own little research activity, uh, which is in a very poor, low-income country, Mozambique, when, uh, just recently, we completed um A kind of an ethnography, like a modest ethno ethnography, let's put it this way. It is astounding to see how interventions are being piled up on the system and the impact it has. On motivational teachers, behaviors of head teachers, local officials. It, it focuses on, on, on local um actors mostly and then all the donor agencies being in the, in the mix. So it, It, at minimum creates these incoherencies or enforces them at least, and sometimes the system looks. Chaotic. So, in, in that type of approach which, which also has been in the RICE program that I have been involved with a little bit, um, I, I really like that because you, you need to know how this all pushing these things actually affects the system. OK. The last point, why I'm a systems-oriented, country-oriented person. So, I, I really think this is very important for Africa especially. Why? Because in Africa, systems are massively expanding, not in the north as we heard in the poverty session and not in the south. South, but in the West, in the center, and East Africa. In the next 2 years, if KCP still exists, in the next 20 years, population in these areas will have pretty much doubled. It will mean, and these are young people, it will mean that education systems expand massively as they have and they will. And then I would like to ask, A question for the future, does research have anything to say to these massively expanding, uh, system, systems that like, um, I was impressed by JP uh um uh Sedo telling me this, these numbers that if Brazil, Nigeria, Bangladesh now have roughly 45 million in the school system. Brazil and Bangladesh will reduce slowly. In Nigeria in 20 years, it will be 90 million. So who has the experience of expanding. Systems to, to accommodate that uh is uh and, and does, is research able to say something about it or is it just an operational issue? I mean, another number, these are just back of the envelope things, but it was also some calculations that in Africa, um, 16% of workforce should be teachers. I mean, Here in America, healthcare is uh like a mass sector and 7 every 7 person is working in healthcare, is 16% is, uh, is something similar. So, so how, how, how can, how can, um, research and, and this work help, uh, you know, make that expansion possible? So that was all I, I wanted to say and um And I hope KCP conti continues to contribute to this and, and other issues. So thank you very much, Tion. Thank you, Rita, um. This has been fascinating because I think in both sessions, we've seen sort of a, a, a strong core of work, a sort of a general buy-in that work. But then in the panel discussions, I think it's been really interesting to see how You know, it's, it's just never enough. Sitting in a research department, that's actually not bad news because there's a lot more to do. But it's on a number of fronts. It's around the, you know, Pam emphasized the complementarities with other factors, uh, Emily. Uh, uh, among other comments, you know, the, the details and the nuance and understanding differences and impacts is really where we're gonna get a lot of information. Um, I thought, I thought both Rukmini and, and, and, and Ritva's points about the links between research and, and, and practice and operations and, and programs. I, I think still need, you know, we're highlighting how that, that, that linkage, and I appreciate Ritter, that you, you're pointing out how difficult it is. It's just not easy. Uh, but that doesn't mean we have to stop trying to, to, to, to, to do that better. And I think everybody's pointed to sort of directions where they see, um, future avenues for research, which, you know, obviously, us in the research department, we, you know, this is sort of what we now need to go back and, and, and, and think through and internalize. So, with that, um, let me thank everybody. Let me give a particular thanks to the Uh, the program and the KCP Program Management Unit, Karina, Marcello, Bintao, who's not here, and, and Gabby for putting together this event, uh, which was really insightful, and, and I, I'm a lot of food for thought. And thank you all for participating both in the room and, and online, and look forward to our next, uh, KCP, uh, event that'll be in, uh, a couple of months, and, uh, announcements will go out soon. So thanks again, everybody.
showAllTranscripts
no
duration
PT1H14M55S
scene7File
worldbank/KCP-Part-2 Human-Development
scene7Domain
https://worldbank.scene7.com/
scene7FileAvs
worldbank/KCP-Part-2 Human-Development-AVS
title
KCP Part 2HumanDevelopment
description
KCP Part 2HumanDevelopment
showTimestampAndTranscript
yes
col-xs-12
col-sm-12
col-md-3
col-lg-3
col-xs-12
col-sm-12
col-md-7
col-lg-7
  • add-style
  • lp-body-content
lp-heading-top-medium
lp-heading-bottom-medium
col-xs-12
col-sm-12
col-md-2
col-lg-2