ABCDE 2024—Session 4: What should Developing Countries Do Differently in the Next Pandemic?

videoType

dynamic-media

videoDmUrl

keyFrameImage

timestamp

00:05 Um,

00:07 as we start the next session,

00:08 I'm gonna invite the panelists to come up

00:11 over here.

00:11 Um,

00:12 Professor C Kramer will join our last session of the day.

00:15 We will be discussing a topic of utmost importance.

00:18 What should developing countries do differently in the next pandemic?

00:21 I'm going at this stage,

00:22 invite

00:23 David Evans.

00:28 Our moderator,

00:30 uh,

00:30 who is the principal economic advisor of the social sector

00:33 at this other really good institution down the road,

00:35 the Inter-American Development Bank.

00:37 David.

00:42 Thanks so much.

00:42 I am going to invite our panelists to stay in their seats for the,

00:46 uh,

00:46 while we're hearing presentations so that they can actually

00:49 have a good view and,

00:50 uh,

00:51 not have to focus if they have to look down,

00:53 so,

00:54 um.

00:55 No,

00:55 no,

00:55 he wants to come here.

00:57 Yeah,

00:57 yeah,

00:58 right.

00:59 Well,

00:59 this is exciting.

01:00 Um,

01:01 so normally I would regret being the,

01:03 uh,

01:03 the last session standing between you and.

01:06 Dinner or cocktails or whatever,

01:08 what have you,

01:09 but

01:10 I don't in this case because we've got a fascinating panel ahead of us.

01:14 Um,

01:14 we,

01:15 our next and last session of the day is on

01:17 what low and middle income countries should be doing differently

01:20 in the next pandemic.

01:22 Uh,

01:23 I'm keeping a mask on in honor of that,

01:25 um.

01:26 Obviously COVID was a massive disruption from a human capital,

01:29 uh,

01:29 from a human perspective and also from an economic perspective with loss of life,

01:34 loss of human capital,

01:35 surges in poverty,

01:37 and much more.

01:38 So in this panel we'll hear from three experts first who have studied this in detail.

01:44 So first we'll hear from uh Mushfik Mubarak,

01:47 uh,

01:47 who's a professor of Management and Economics at Yale University.

01:51 And then we'll hear from Joanna Silva,

01:53 who's a senior economist here at the World Bank and also

01:56 a professor at the Catolica Lisbon School of Business and Economics.

02:00 So we'll hear from each of them and we'll have an

02:02 opportunity for just a couple of questions after each speaker.

02:06 After that,

02:07 we'll hear from Norbert Shatty,

02:08 Chief Economist for Human Development at the World Bank Group here,

02:11 and he'll give us a presentation.

02:13 Once Norman has spoken,

02:14 we're gonna invite Michael to come back up to the stage,

02:17 we'll invite Mamta to come up to the stage,

02:19 Mamta Murti,

02:19 the er vice president for human er Development at the World Bank,

02:23 to come up to the stage and we'll have uh a brief er

02:27 high level policy panel er talking about these issues from there.

02:30 So,

02:31 with that,

02:32 we will,

02:33 I will give the mic to er Mushvik.

02:42 Uh,

02:42 uh,

02:43 thank you.

02:43 I'll,

02:45 So while I wait for my slides to come up,

02:46 when um,

02:47 Norbert invited me to do this talk,

02:49 I thought,

02:50 you know,

02:50 15 minutes is too short for one paper,

02:53 and so I'm going to present about 5.

02:56 so

02:57 it's titled,

02:57 you know,

02:58 I,

02:58 I tried to reflect back on what we learned during the pandemic.

03:01 Uh,

03:01 so I'm going to

03:03 try and give you a few lessons that maybe we can take forward for the next pandemic,

03:08 especially keeping in mind,

03:09 um,

03:10 what's at stake for low and middle-income countries,

03:12 OK.

03:13 So,

03:14 the lesson number one,

03:15 and I'm going to try to be very,

03:16 very quick through through these messages,

03:18 is that it was an unprecedentedly large economic shock.

03:21 So initially,

03:22 March,

03:22 April,

03:23 May 2020,

03:24 we were really focused on the public health aspects,

03:26 how the virus was going to

03:28 behave,

03:28 but ultimately it became clear

03:30 that

03:31 Um,

03:32 you know,

03:32 food insecurity,

03:33 income losses,

03:34 employment losses

03:35 were the much bigger stories,

03:37 right?

03:38 So here,

03:38 given that we have eminent leadership of the World Bank here,

03:42 I'll say that that also suggests that maybe we should have

03:46 had institutions like the World Bank

03:48 rather than biomedically focused organizations like

03:52 Say the WHO

03:53 maybe play an even bigger role in coming forward and going forward,

03:56 I think the World Bank should keep that in mind,

03:59 right?

03:59 That ultimately,

04:00 when we,

04:01 when,

04:01 when crises happen,

04:02 even if the origin is conflict,

04:04 if the origin is a virus,

04:05 right,

04:06 um,

04:07 the,

04:08 the effects,

04:09 effects on human life are,

04:10 are,

04:10 are,

04:10 are large.

04:11 And here,

04:11 the World Bank actually played a very,

04:13 I mean,

04:13 so I'm showing you macro data here,

04:15 right?

04:16 Uh,

04:16 but then when we started thinking about in March,

04:18 April,

04:19 May,

04:19 how this was affecting the lives of people in developing countries,

04:23 right,

04:24 we quickly realized that we didn't have that information and

04:26 nobody was also able to collect data in person,

04:29 right?

04:29 So in the US you could solve that problem because people leave digital traces.

04:34 So we could indirectly infer what was going on.

04:37 In developing countries where a lot of people are in the informal sector,

04:40 right?

04:40 Even government survey methods were not sufficient.

04:43 So we quickly,

04:44 many researchers who work around the world,

04:46 we quickly put together this effort of aggregating data from across

04:51 many different countries,

04:52 right?

04:53 And we learned things like,

04:55 you know,

04:55 huge,

04:56 uh,

04:57 income losses in the median sample,

04:59 70% loss in income,

05:02 and that was also these income,

05:03 employment losses was actually also translating into

05:06 increases in food insecurity.

05:08 People were reporting missing meals,

05:10 right?

05:10 And I have to give credit to the World Bank here.

05:12 While this was sort of an

05:14 ad hoc enterprise by some researchers at various universities,

05:17 the World Bank really took this type of effort and

05:21 Put it on steroids,

05:22 right?

05:22 And there were huge,

05:23 uh,

05:24 large scale surveys,

05:25 like phone surveys being done in a

05:26 systematic way with nationally representative samples,

05:29 which I was not able to do,

05:30 right?

05:31 Um,

05:33 second lesson that became very clear,

05:35 uh,

05:35 in March,

05:35 April 2020,

05:37 is that

05:38 we can't really afford to blindly copy,

05:41 we,

05:41 I'm from Bangladesh,

05:42 low and middle-income countries can't,

05:44 couldn't really afford to just blindly copy.

05:47 The pandemic response policies that were being

05:49 promoted and proposed in the US and UK,

05:51 like things like if you

05:53 recall,

05:54 flatten the curve,

05:55 lockdowns,

05:56 right?

05:57 So things like that.

05:57 So what I did at that time was,

05:59 let's take that Imperial College model that that

06:02 governed our response in the US and UK,

06:04 right?

06:05 Let me just calibrate that to data from Pakistan,

06:08 Nigeria,

06:08 Bangladesh,

06:09 right?

06:09 Let's see if you get the same answer.

06:11 And quickly,

06:11 you figure out you don't actually get the same answer there,

06:13 right?

06:14 And the intuition is really easy for me to explain in 30 seconds.

06:18 One is the age distribution of the population is very different.

06:21 Developing countries are much younger,

06:22 and we already knew

06:24 from day one that the virus was a lot more fatal

06:27 to the elderly.

06:27 Number 2,

06:28 right,

06:28 you flatten the curve because

06:30 you don't want the curve of infections to go above

06:33 your

06:33 healthcare capacity line,

06:34 so there would be excess unnecessary deaths when people are not getting access to.

06:39 Uh,

06:40 hospital beds,

06:41 right?

06:41 But now imagine you're in rural Sierra Leone,

06:43 rural Bangladesh,

06:44 where the healthcare capacity line is down here,

06:47 right?

06:47 By,

06:48 by,

06:48 uh,

06:49 flattening the curve and delaying infections

06:52 for another 3 months,

06:53 you're not actually get that much out of it.

06:54 That's what the model tells you,

06:56 right?

06:57 And so then that made us focus a lot more on the,

07:00 OK,

07:00 the benefits are a lot smaller from lockdowns.

07:02 So what are the costs of lockdowns,

07:04 right?

07:05 And this is data from Nepal.

07:07 I'm on purpose showing you the early round of the data collected in March of 2020,

07:11 right?

07:11 And what people were mostly worried about wasn't health and safety.

07:16 Even in March of 2020,

07:17 rural Nepalese were mostly worried about their

07:19 food insecurity and the income shock.

07:21 Right?

07:22 So this is why we need to get that information out right away,

07:25 right?

07:25 To,

07:25 to be able to

07:27 target our responses more,

07:28 more accurately.

07:29 Now,

07:29 here's another place where I think

07:32 institutions like the World Bank could play a bigger role.

07:35 So,

07:35 something that became clear at that time is

07:37 regardless of your position on lockdowns or not,

07:40 it became clear that we needed to get money in the hands of poor people very,

07:43 very quickly,

07:44 right?

07:45 And that we also realized that we didn't know how to do that.

07:48 Right.

07:48 First of all,

07:48 we didn't even know who needed support,

07:51 right?

07:51 Because that requires a lot of data,

07:53 OK?

07:54 And then there were really interesting,

07:55 uh,

07:56 uh,

07:56 innovations like in Togo,

07:58 working with Josh Blumenstock,

08:00 who's at UC Berkeley,

08:01 you know,

08:01 they figured out how to,

08:03 uh,

08:03 use cell phone records to make indirect inferences about who might need support,

08:07 right?

08:07 So we need innovations like that.

08:08 And again,

08:09 those innovations have to be

08:11 economic,

08:12 technological,

08:13 those are not necessarily public health or medical innovations.

08:16 OK.

08:17 Uh,

08:18 right,

08:18 let me skip over.

08:19 Yeah,

08:19 and just,

08:20 just to,

08:20 just to say,

08:20 I'm,

08:21 I'm,

08:21 uh,

08:22 you know,

08:22 I also wrote at the same time that the

08:24 answer for the rich countries are actually quite different,

08:26 right?

08:27 In rich countries,

08:28 the right thing to do was to just like,

08:30 crush the virus like New Zealand did so that they could open up more quickly.

08:34 Whereas countries that

08:36 did not manage to have strict lockdowns,

08:38 they remained,

08:39 uh,

08:39 their economic activity remained depressed

08:42 for almost 2 years afterwards,

08:44 right?

08:45 Uh,

08:46 OK,

08:46 sorry,

08:47 these were,

08:48 um,

08:49 uh,

08:49 there were some hidden slides there,

08:50 but they're shown.

08:52 Um,

08:53 so,

08:53 uh,

08:54 researchers also had an important role to play,

08:57 you know,

08:57 uh,

08:58 people in this building and academic researchers.

09:00 Uh,

09:01 there was a lot of data and information and

09:03 insights that we could provide that would be of value

09:06 to government policymakers.

09:07 Everybody was flying blind,

09:08 we all were.

09:10 And government really needed that support.

09:12 So this is a simple insight,

09:13 right?

09:14 So,

09:14 when you look at what was happening in Nepal,

09:17 in terms of food insecurity,

09:18 this is what it looked like,

09:19 like the percentage of people,

09:21 this is just an index of food insecurity.

09:23 It actually looked not

09:25 too much.

09:26 It,

09:26 it looked better than what Nepal looked like 6 months before.

09:28 And the reason is 6 months before it was the period of lean,

09:32 it was like a lean season period,

09:33 right?

09:34 We,

09:34 we just got very lucky that when the pandemic hit with the post-harvest period.

09:39 Right?

09:40 But only because,

09:41 you know,

09:41 you had multiple years of data on what

09:45 seasonal fluctuations look like,

09:47 right?

09:47 That we knew that this was a lot worse

09:50 than what it really should be during this period,

09:52 even though the benchmark is 6 months ago,

09:55 it wasn't looking that bad,

09:56 right?

09:57 And in Sierra Leone,

09:58 um,

09:59 you know,

09:59 about a third of all

10:01 households are female-headed,

10:03 uh,

10:03 and we learned that

10:05 Um,

10:06 you know,

10:06 they experienced the pandemic very,

10:08 very differently.

10:09 They knew less,

10:10 they were engaging less in,

10:11 um,

10:12 defensive behaviors,

10:13 etc.

10:14 and much more food insecurity in female-headed households.

10:16 So there was information,

10:17 like more micro-level information on who the government should be

10:20 targeting more,

10:21 right?

10:23 Fourth lesson is that

10:25 it's not that low million countries

10:27 are just

10:29 uh recipients of information,

10:31 right?

10:32 There was a lot that we could teach each other as well,

10:34 right?

10:34 And I think a mistake we made in Bangladesh,

10:36 India,

10:37 Pakistan,

10:37 all over South Asia,

10:39 is that we

10:41 were looking towards US,

10:43 UK,

10:43 South Korea,

10:44 like getting our information from CNN and BBC was focused

10:47 on what was happening in the US and UK.

10:49 And we just missed the boat in terms of

10:50 just looking towards countries like Sierra Leone and Liberia,

10:54 which had pandemic experience,

10:56 right?

10:57 And they had these frugal innovations

10:59 that would have been much more

11:01 important for us to think about and copy rather than thinking about this,

11:05 uh,

11:06 sophisticated contact tracing system that South Korea could install,

11:10 right?

11:10 That wasn't all that useful for us,

11:11 but we were looking in the wrong places,

11:13 right?

11:14 And,

11:14 uh,

11:15 I,

11:15 I think the US also could learn a lot,

11:17 like our leadership,

11:18 uh,

11:18 you know,

11:19 was not,

11:19 uh,

11:21 very well

11:22 positioned to respond to the pandemic relative to many countries in Africa,

11:25 right?

11:26 And there's also other things that happen,

11:28 like through research in,

11:29 in developing countries,

11:30 right?

11:31 Like gen genome sequencing of the,

11:32 of the virus in South Africa,

11:34 uh,

11:35 which led to early identification of the Omicron variant,

11:37 vaccine development.

11:39 Um,

11:40 trials,

11:40 etc.

11:41 that was happening in,

11:42 in various developing countries.

11:43 OK OK.

11:44 And something that also became clear by 2021

11:48 is that it wasn't just developing country policymakers,

11:51 even international organizations who are supposed to lead on this,

11:54 CDC,

11:54 WHO.

11:56 I was surprised to learn they were also sort of flying blind,

11:58 right?

11:59 And this showed up

12:00 with very

12:02 strange uh and uh confused guidance around masking,

12:06 right?

12:07 And,

12:08 and so,

12:08 and talking to the WHO early on in the pandemic,

12:11 we had data that look,

12:12 when you look at non-experimental data on how

12:14 the virus is behaving in countries that had a

12:17 mask wearing norm versus not,

12:18 it's clear that masking seems to do something,

12:21 right?

12:21 And the WHO's reaction was like,

12:23 no,

12:23 that's not like,

12:25 without RCTs,

12:26 we can't

12:26 change our guidance,

12:27 right?

12:28 So then in the middle of the pandemic,

12:29 we ran an RCT and again,

12:31 it's a developing country,

12:32 Bangladesh.

12:32 Most of the authors of,

12:34 of this,

12:34 of this paper are from Bangladesh and we're residing in Bangladesh,

12:38 that we ran a trial of 350,000 people

12:41 to get people to wear masks and then track what happened.

12:44 And what we learned from that is that you can,

12:46 first of all,

12:46 you can get people to wear masks.

12:48 There's a set of 4 things that you have to do.

12:50 That includes free distribution,

12:52 but you also have to go back

12:53 and harass people a little bit

12:55 to keep their masks on,

12:56 right?

12:56 Like,

12:57 but politely harass people,

12:58 OK.

12:59 Um,

13:00 and we don't need it for Dave,

13:01 obviously.

13:03 Um,

13:03 and,

13:04 but then they keep their masks on,

13:06 right?

13:06 And once they keep their masks on,

13:08 there is a significant reduction in

13:11 COVID symptoms as well as symptomatic seropositivity,

13:14 right?

13:14 And in fact,

13:15 this in in rural Bangladesh,

13:17 this

13:18 eliminated about a third of all infections for the elderly,

13:21 right?

13:22 Um,

13:23 and,

13:24 you know,

13:24 in the developed world,

13:26 like the Washington Post also,

13:27 also learned from this,

13:28 right?

13:28 And,

13:29 uh,

13:29 and so did the WHO.

13:31 And another thing that became clear during the mass trial is that

13:35 I quickly moved from because the Delta wave had hit South Asia,

13:39 and people were dying on the street,

13:40 and we quickly stopped research and started distributing masks,

13:43 right?

13:44 And then it became clear that governments actually need support,

13:46 even,

13:47 even to figure out,

13:48 like on the fly,

13:49 figure out how to distribute mass,

13:51 right?

13:51 So we hired a person who used to work for the uh for the Gujarat state government,

13:57 right?

13:57 And he explained to us how to translate all of this information

14:00 in a format that makes it much more actionable for government,

14:03 right?

14:04 And,

14:05 um,

14:06 and that led to a lot of mass distribution,

14:08 uh,

14:09 prior to the Omicron wave,

14:10 OK?

14:11 And final lesson

14:13 is that the pandemic just didn't ever quite get over

14:17 in many poor countries after it was functionally over here,

14:21 right?

14:22 And that comes from this picture.

14:23 Let me just jump to this picture,

14:24 like 18 months after vaccines came to market.

14:28 Most of sub-Saharan Africa remain unvaccinated,

14:30 right?

14:31 Also,

14:31 the Caribbean poor countries in the Caribbean,

14:33 OK?

14:34 Why was this?

14:34 So when you ask this question,

14:36 why were vaccination rates lagging,

14:37 African scholars,

14:38 policymakers made the good point

14:40 that uh

14:41 vaccines were effectively being withheld from them,

14:44 right?

14:44 But

14:45 in response,

14:47 um,

14:47 the pharmaceutical companies claimed,

14:49 on this is on record that,

14:50 oh,

14:50 no,

14:50 no,

14:51 no,

14:51 that's not the issue.

14:51 If you were to send doses,

14:52 they don't get used.

14:53 Vaccination,

14:54 vaccine hesitancy is too high,

14:56 right?

14:56 So I was like,

14:57 OK,

14:57 that's an empirical question that can be answered,

14:59 right?

15:00 So let's run this survey in 15 countries.

15:02 But the trick was,

15:03 let's add it to add the US to it.

15:05 And what you learn from that is in every single low-income,

15:08 middle-income country,

15:09 vaccine hesitancy was lower than it is in the US.

15:12 So just on the face of it,

15:13 it's just like a BS excuse,

15:14 right?

15:15 So it wasn't vaccine hesitancy,

15:16 but we did learn is that

15:19 getting vaccines to people was really difficult,

15:21 right?

15:22 So then our final trial is the last thing I'll say

15:24 is let's

15:26 go to the remote places,

15:27 let's take vaccines and see what happens.

15:29 And what happens is when you run this trial for 48 hours,

15:32 you just take vaccines there.

15:34 The number vaccinated per community moves from 5

15:36 people to about 55 people within 48 hours,

15:39 right?

15:39 To get from 0 to 50,

15:41 you just need to get people's stuff,

15:43 right?

15:43 You don't have to do anything sophisticated with that,

15:45 right?

15:46 And we learned something from that is that,

15:47 OK,

15:47 like,

15:48 if you're going to have to pay this cost,

15:50 of

15:51 going out to remote areas,

15:53 it's insane to just take a single vaccine.

15:56 We should be taking

15:57 a bundle of useful health services,

15:59 right?

16:00 And so,

16:00 hopefully,

16:01 through the pandemic,

16:01 this was my bonus lesson 8,

16:03 is that we learned something useful for how

16:05 to deliver health services in remote rural areas.

16:08 Thanks.

16:16 Thank you so much.

16:18 Uh,

16:18 I don't know,

16:18 there we go.

16:19 Uh,

16:19 we are,

16:20 we're just gonna take 1 to 2 questions before we,

16:23 uh,

16:23 move on to the next panelists.

16:24 If anybody has a question from that,

16:26 uh,

16:26 exhilarating quick tutorial,

16:29 um,

16:30 of lessons.

16:31 Any questions?

16:33 Yes,

16:34 I have one there.

16:42 Very insightful.

16:43 Thank you very much.

16:46 Have we learned

16:47 these lessons

16:48 and

16:48 are we ready for the next pandemic?

16:50 Thank you.

16:51 Yeah,

16:52 um,

16:53 no,

16:53 we are definitely not,

16:54 sure.

16:55 Do we have,

16:55 do we have one more,

16:56 or is that,

16:56 yeah,

16:57 let's take this one more and then we'll,

16:59 uh,

16:59 and then you can answer both of those.

17:08 Hello again,

17:09 my name is Juliet Adams,

17:10 and I'm also an educator.

17:12 As you know that the,

17:14 the,

17:14 the COVID,

17:15 the lockdown,

17:16 it rocked the education,

17:18 um,

17:18 sector,

17:19 um,

17:20 with serious consequences.

17:22 I would like your take on what took place.

17:25 Thank you.

17:25 OK,

17:25 sure.

17:26 That second question,

17:27 I'm going to hand it over to Norbert,

17:29 who's going to talk about exactly this.

17:30 You'll hear 15 minutes of this.

17:32 So,

17:32 OK.

17:33 And have we learned these lessons?

17:35 I,

17:35 um,

17:36 I hope so,

17:36 but I'm not too optimistic,

17:38 right?

17:39 Um,

17:39 so I would say 11 reason is that,

17:42 look,

17:42 the next virus that comes,

17:43 you know,

17:44 like I said earlier in this talk that,

17:46 look,

17:46 this virus was really dangerous for the elderly,

17:48 right?

17:49 The next virus could affect the respiratory systems of children more,

17:52 right?

17:53 And we'll need a completely different set of responses and,

17:56 you know,

17:56 so we need to

17:58 stay on top of things and it's not just,

18:00 uh,

18:01 uh,

18:02 it's not just about like taking these

18:03 lessons and it'll immediately translate over.

18:05 That's one.

18:06 But on the other hand,

18:07 the reason what I'm optimistic is that I did

18:09 notice very early in the pandemic in March,

18:11 April,

18:12 May of 2020,

18:13 right?

18:13 Countries with pandemic experience like Sierra Leone,

18:16 Liberia,

18:16 Guinea,

18:17 etc.

18:17 they actually reacted much better than we did in Bangladesh or in India or Pakistan,

18:22 or even in,

18:23 in other parts of Africa,

18:24 right?

18:24 And that makes me think that experience does matter,

18:27 right?

18:27 Um,

18:28 so,

18:28 I hope I,

18:29 that's my optimism.

18:31 Let's

18:31 give another round of applause as Johanna,

18:33 thanks so much.

18:41 Thank you very much.

18:42 So this paper is about how you adjust to transitory shock,

18:45 so it's not specific about COVID,

18:47 but it,

18:48 it,

18:48 it has implications for that.

18:50 It's joint work with my,

18:51 with my colleague

18:52 Ana Marguerite Fernan who is over there.

18:55 Um,

18:56 so we have 3 main findings with this paper,

18:59 and let me start with that.

19:01 Normally when we think about permanent shocks like trade liberalization,

19:05 we do think that they have

19:07 permanent effects.

19:08 However,

19:09 when we think about transitory shocks,

19:11 we don't think they will have permanent effects of

19:13 the type of these other types of shocks.

19:16 So what we show in this paper is that indeed there are permanent effects of.

19:20 Repository shocks,

19:21 we are gonna prove it in this case

19:23 and this will happen in a setting of high informality where you could

19:26 think that informality could work as a full buffer of the shock,

19:30 but it doesn't.

19:31 We are also gonna show you that

19:33 um adjustment depends very much on the level

19:36 of informality but also on whether governments protect firms

19:39 uh through,

19:40 you know,

19:41 increasing in the level of concentration in the market or SOEs.

19:45 And quite surprisingly for us and maybe for you as well,

19:49 we are going to show you that most of the scarring actually occurs

19:52 among incumbent workers.

19:54 It's not displaced workers,

19:55 which is the type of workers that we normally

19:58 worry about,

19:59 but it's actually workers who continue in the same firm.

20:02 And now at this point you are wondering how come.

20:04 This happens because firms are also scarred.

20:08 And this scarring of firms,

20:09 what we show in this paper will come from firm exit,

20:12 but also a permanent downsize

20:15 in productivity

20:16 that will happen as a consequence of a transitory shock,

20:19 so quite surprising.

20:21 The last set of findings have to do with government responses,

20:25 and we normally think there are gaps in

20:27 these responses in the case of transitory shocks,

20:31 but we often don't know how large these gaps are.

20:34 So in this paper,

20:35 we look at data for Brazil and show that of all the losses that workers had,

20:40 unemployment.

20:42 and both the family welfare program of Brazil compensated 6% of their losses,

20:47 no more than that.

20:48 Training didn't even reply at all.

20:51 There is no response from training programs.

20:53 So lots of the losses

20:55 were borne by workers,

20:56 they were permanent,

20:57 they were borne by firms,

20:58 they were permanent and

21:00 policies didn't compensate for them.

21:02 In this paper,

21:03 we use

21:04 administrative data for,

21:05 for Brazil.

21:06 Uh,

21:07 and we link the matched employer employee records of Brazil

21:11 with the manufacturing census

21:14 and also the the the the the CEE,

21:16 so the trade data.

21:18 In addition,

21:19 which is rare,

21:20 more rare in this case is that we also Had individual level records for

21:26 the benefits of unemployment insurance at the level of each person in Brazil,

21:30 also the welfare program,

21:32 and also the individual records of the training program,

21:35 so we could see exactly which worker got what at which time.

21:40 With this rich data,

21:41 the way we identified the shock was that

21:45 we were trying to look at the foreign demand shock that

21:48 the global financial crisis created in each firm in Brazil,

21:52 and what we are going to compare is the exposure of the firms to this shock,

21:57 this exogenous variation in this exposure.

21:59 Uh,

22:00 is what's going to determine,

22:01 um,

22:02 you know,

22:02 then the impacts that we,

22:04 that we compute at the firm level.

22:06 Here is just a graph to illustrate the point that

22:09 this,

22:09 this shock that we are going to look at was really transitory,

22:13 is the deep dive at the moment of the global financial crisis.

22:17 The way we look at this shock is that,

22:19 We consider

22:21 the importance for each firm of each destination market in their export portfolio,

22:27 and then we multiply that by the change in GDP during the period of the crisis in that

22:33 destination.

22:34 So it's kind of the change

22:35 in the GDP of the firm,

22:37 if you will,

22:38 that only comes from external demand,

22:40 so it's in the partner country,

22:43 um,

22:43 and that's how we measure the,

22:44 the shock,

22:46 the temporary shock that each firm in Brazil has.

22:50 It's exhaustionous and it had been

22:52 used by others in other contexts.

22:54 So what we look at,

22:55 we look first at workers' outcomes.

22:57 So each line in our data set is a worker

22:59 and we follow him through time in his employment spell.

23:03 We have then the variable of interest,

23:04 the shock variable,

23:06 and then we control for,

23:07 uh,

23:07 you know,

23:07 lots of characteristics of workers,

23:10 of firms and have you,

23:11 you know,

23:11 the fixed fixed effects at the different levels

23:14 that will allow us to identify the effects.

23:16 What do we find?

23:17 So we find

23:18 first in terms of employment,

23:20 we see a permanent reduction

23:23 in the,

23:23 in the average month's work.

23:26 You see that they,

23:26 the,

23:27 the,

23:27 here are the plots of the coefficient of that regression,

23:30 the coefficient of the shock variable,

23:32 and you see that it felt

23:34 and it stayed significantly lower

23:37 until 2017,

23:39 so a decade after.

23:40 We look at wages and we see also a reduction

23:43 and

23:44 actually no improvement whatsoever,

23:46 so they went down and stayed down.

23:48 And in terms of hours worked,

23:49 which is surprising for labor economies,

23:51 they also went down,

23:52 and again this was a temporary shock.

23:55 In terms of magnitude,

23:57 what we computed to be the effect on the employment on

24:01 most work is about 3% and in terms of wages 6%.

24:05 So these are sizeable effects.

24:08 Now,

24:08 let's look at these mediating factors,

24:10 informality.

24:12 So labor markets,

24:13 local labor markets in Brazil,

24:14 they differ in the,

24:15 in the share of informal workers they have.

24:18 And what you see there is that

24:20 in places with lower informality levels,

24:23 workers

24:24 were better off.

24:25 Then in higher the response,

24:28 the response was smaller,

24:30 so the loss was smaller in higher informality places.

24:33 So it did work like a buffer,

24:36 but it didn't shelter workers completely.

24:39 They still had

24:40 employment losses and wage losses,

24:42 but they were smaller.

24:43 In terms of the level of concentration of the market and whether

24:47 you as a worker were working in an SOE at the time,

24:51 only those that were

24:53 in low concentration sectors

24:56 and in non-state owned enterprises bear the burden of the crisis.

25:01 The others,

25:02 the effects were not significant.

25:04 So it was a way to insulate workers,

25:06 but

25:07 that is,

25:08 as we show in the paper,

25:09 you know,

25:09 the cost then is spread to others.

25:13 Now,

25:14 who,

25:14 in terms of workers,

25:15 you could either stay at your firm or move to a different sector,

25:19 to a different firm in the same sector.

25:21 So what happened depending on what you did,

25:24 we could observe what they did.

25:26 So what you see is there in the first column are the initial results.

25:29 The second are the results for the workers who remain in their firm.

25:33 And you see that there are losses that are significant and negative and quite large.

25:39 For those that moved to other firms in the same sector,

25:42 uh,

25:43 there weren't losses or other firms in the tradable sector.

25:46 Um,

25:46 but the only other significant result is

25:49 for other firms in the non-tradable sector,

25:51 kind of the best way to escape the crisis losses was really uh for

25:55 you as a worker to move to another firm in the non-tradable sector.

26:00 And most of the losses as you see are really,

26:02 uh,

26:02 the coefficient is larger in the ones that remain in the initial firm.

26:07 This is quite surprising.

26:09 Now in terms of impact on firms,

26:11 we look at different outcomes,

26:12 effects on revenues,

26:14 profits,

26:15 on productivity calculated in different ways,

26:18 and here I find the results very interesting.

26:20 Here also we are plotting the coefficient of the shock variable.

26:24 But now our regressions are at the firm level.

26:27 What do you see there?

26:28 A transitory shock,

26:30 the global financial crisis,

26:31 its external dimension

26:33 led to a reduction in the net revenues that was permanent.

26:36 It was still there in the end of the period.

26:40 If you look at exit,

26:41 well,

26:41 it led to exit just a year after.

26:44 So the ones,

26:44 the firms that did exit,

26:45 they,

26:45 they,

26:46 they did exit because of the crisis,

26:47 they did it the year after,

26:48 not after that.

26:50 And in terms of size,

26:52 you see this contraction

26:54 in terms of size.

26:55 What did they do to the labor force?

26:57 So they adjusted the types of workers they had.

27:00 They stayed with less unskilled workers and less skilled workers,

27:04 but the reduction was much larger in the quantity.

27:07 That they were using of unskilled workers than skilled workers.

27:12 So as a consequence,

27:14 the,

27:14 the share of unskilled workers that each firm in Brazil was using

27:19 became lower,

27:20 as you can see in the last graph on the on the on the left.

27:25 So they kind of adjust the composition of their workforce.

27:29 What about the technology they were using,

27:31 the productivity.

27:32 As you can see here,

27:33 they adjusted down the materials,

27:36 so they use less materials.

27:37 The material per worker,

27:38 they did decrease.

27:40 Um,

27:41 the,

27:41 the effects on capital per worker in Brazil were not significant,

27:44 so that was not the main margin of adjustment for workers.

27:48 And then as you can see in the TFP effects,

27:50 and we use several measures,

27:52 uh,

27:53 you know,

27:53 really persistent effects

27:55 of,

27:56 uh,

27:56 on productivity.

27:57 So

27:58 firms that were hardest hit,

28:00 they had lower productivity not for one year but for a decade.

28:06 Um,

28:06 so you know,

28:07 11 thing that we know,

28:09 we say and,

28:09 and kind of now linking to the macroeconomist literature

28:13 is that there can be cleansing effects of crisis.

28:15 So a good news for productivity could be one

28:18 where the firms that exit the market are the weakest,

28:20 the least productive,

28:21 and therefore the market share of the

28:23 best firms actually increases with the crisis.

28:26 Here we are not in general equilibrium,

28:28 but we could at least ask which firms did exit in the case of Brazil.

28:32 Indeed,

28:33 they were the smaller,

28:35 they were the least productive,

28:36 so at least in that respect,

28:38 some,

28:38 some good news.

28:40 OK,

28:41 now looking at the asset,

28:43 the compensation mechanisms,

28:44 how did they

28:46 work in Brazil?

28:47 So we know for each worker in Brazil,

28:49 because of the the the type of data we have,

28:52 exactly how much he lost,

28:55 and we know what he received

28:57 in terms of unemployment insurance and

28:59 welfare programs and whether he got training

29:01 and what do we find?

29:02 So you there you see the response in terms of unemployment insurance.

29:06 So yes,

29:07 the ones,

29:08 the ones that were working at firms that were hardest hit were more likely to get,

29:12 they did get more unemployment insurance.

29:14 They did get

29:15 more uh both the familia program,

29:18 but what's really

29:20 the share that was compensated?

29:22 So among all the losses,

29:24 unemployment insurance compensated 4.3% of their losses,

29:28 and Bolsa familia,

29:30 so the welfare program,

29:31 1.7%. All in all,

29:34 6% of the losses were compensated.

29:36 All the rest stayed with the workers and the firms.

29:40 And training,

29:40 like I said,

29:41 didn't reply at all.

29:42 But you could be wondered,

29:43 this is so strange,

29:44 incumbent workers,

29:46 transitory shocks having permanent effects,

29:48 this permanent

29:49 loss in productivity of firms,

29:50 are they,

29:51 uh,

29:51 is this just Brazil?

29:53 So we did it for another country,

29:55 Ecuador,

29:55 that had the same type of very rich data.

29:58 But had a much less flexible labor market,

30:02 so there this margin that adjusted in Brazil,

30:04 the workers couldn't adjust as much.

30:06 So what did they do instead?

30:08 Unable to adjust labor as much,

30:10 they adjust capital.

30:11 And what happened?

30:12 The productivity losses were also persistent,

30:15 but much larger,

30:16 much larger.

30:18 So to conclude,

30:20 here an example um with data from Brazil and Ecuador of

30:23 transitory shocks having very persistent effects on workers and firms,

30:28 uh,

30:28 these effects,

30:29 uh,

30:30 seen,

30:30 is being seen even in context of high informality that,

30:34 you know,

30:34 works as a buffer but doesn't solve it all.

30:37 Governments protecting firms

30:38 and those at those firms being OK but not the rest.

30:42 Uh,

30:43 and then,

30:43 uh,

30:44 in terms of firms,

30:45 you see this really persistent fall in productivity with different

30:49 natures depending on labor market flexibility of the market.

30:53 And very limited response of the existing

30:57 social programs,

30:58 being their unemployment insurance or welfare programs in a country

31:02 where this exists,

31:03 and this just exists in about 1/3

31:05 of the countries in the world.

31:07 So

31:07 really worrying news for,

31:10 for,

31:11 for,

31:11 you know,

31:12 worker losses,

31:12 but also the extent to which policies are compensating them.

31:15 Thank you.

31:19 Thank you so much,

31:20 Joanna.

31:21 Um,

31:22 do we have,

31:22 uh,

31:23 do we have a question or so before,

31:24 uh,

31:25 we go on?

31:26 I'm gonna take the,

31:27 uh,

31:27 moderator's privilege.

31:28 I have one question while the mic goes up,

31:30 which is,

31:31 you know,

31:31 obviously we see here,

31:32 like the massive potential effects of shocks like the pandemic.

31:37 I'm curious if

31:38 sort of how this relates to the literature that we've seen

31:41 from the pandemic itself

31:42 and how you feel like this applies.

31:45 And we'll take the second question as well.

31:46 Yeah,

31:47 thanks,

31:47 Joanna for such a good presentation.

31:48 And very interesting work.

31:50 I had a,

31:51 um,

31:51 it was a lot of material,

31:52 so it may be obvious.

31:54 How were you able to rule out,

31:55 uh,

31:56 an explanation whereby it's the

31:58 weakest workers who stay in the firm,

32:01 and that's why you're seeing

32:02 what you're calling the scarring effect for those

32:04 who remain rather than those who leave.

32:09 So thank you very much for both your questions.

32:11 These are very good questions.

32:12 About the COVID is different because there was also a supply shock.

32:15 Here we are just looking at the demand shock

32:18 uh driven by the change in foreign demand.

32:20 So,

32:21 you know,

32:21 those other dimensions that come with the supply side that Norbert is gonna speak.

32:26 Uh,

32:26 will be additional to this,

32:28 so this is with a pure demand shock coming from abroad,

32:32 so totally exhaustionous to you.

32:34 This is

32:35 wall that is already happening.

32:37 Add to that the supply side shock and

32:39 you get COVID effects.

32:41 So that's when I,

32:42 uh,

32:42 I'll stay there,

32:44 um,

32:45 to,

32:45 to,

32:45 to Raquel's question,

32:47 so,

32:48 um,

32:49 Here,

32:49 here in the,

32:50 in the,

32:51 in this paper,

32:52 what we,

32:53 um,

32:53 what we,

32:54 what we are looking is the,

32:56 is,

32:56 is really

32:57 um

32:58 the,

32:58 the,

32:58 the,

32:59 the effects that,

33:00 you know,

33:00 just,

33:01 just the,

33:02 the,

33:02 the,

33:02 the,

33:03 the,

33:03 the,

33:03 the,

33:03 the,

33:03 the effects that

33:05 So we are not making

33:08 a statement about who stayed and who left.

33:10 So maybe

33:11 these workers who stayed were worse than the ones who left,

33:14 but what we can tell you is that very few

33:17 left,

33:18 so displacement was very limited in Brazil.

33:22 So in that sense,

33:23 um,

33:23 it's not that,

33:24 you know,

33:25 everybody left before the crisis,

33:27 and this case was really unexpected,

33:28 so they couldn't leave

33:30 before the crisis.

33:31 So we we we we do have some analysis comparing those who left and who didn't,

33:36 and they don't seem to be different.

33:37 So that's how we

33:39 kind of make a call about that part,

33:41 but it's like

33:42 being them how they are,

33:44 they did have these losses.

33:47 Thank you.

33:48 Thank you so much.

33:48 Let's give Joanne another hand.

33:51 And

33:53 now we'll have our last presentation of the session,

33:56 uh,

33:56 Norbert Shatty,

33:57 after which we'll transition into a brief round table.

34:03 OK.

34:04 Well,

34:05 uh thanks to everybody.

34:06 Uh,

34:06 I'm going to be speaking about the effects that the

34:08 pandemic had on uh early childhood and on education outcomes.

34:13 And I'm going to be speaking mainly on the basis

34:15 of a book we published about a year ago now,

34:18 which looked at these effects.

34:19 It was a book,

34:20 uh,

34:20 it was sort of the flagship,

34:21 uh,

34:22 flagship publication of the,

34:24 of the bank in terms of human development.

34:26 And I'm just going to summarize some of the evidence we have,

34:29 we have in there.

34:30 I want to talk a little bit about early childhood development,

34:32 a little bit about school-age children,

34:33 and a little bit about how much recovery has there been.

34:36 What can we say about that?

34:38 Let me start with young children.

34:40 So I have two slides on young children.

34:42 This is the first one.

34:43 What it shows is,

34:44 look at the graph on the left.

34:45 This is the evolution

34:47 of preschool enrollment rates in South Africa.

34:49 And then the vertical line corresponds to the beginning of the pandemic.

34:53 So what happens is preschool enrollment dropped substantially.

34:57 And on the right hand side,

34:58 we have this for different countries once

35:01 preschools reopened.

35:02 So this is no longer people are not enrolled in preschool,

35:05 because it makes no sense to be enrolled in preschool if preschools are closed.

35:08 These bars over there correspond after

35:10 preschools reopened,

35:12 and they're separate bars.

35:13 For children of mothers who have primary education,

35:17 at least some secondary and at least some postsecondary.

35:19 So two messages

35:20 everywhere we see large drops in preschool enrollment.

35:24 0.1,

35:24 large.

35:25 These are 15 18% point drops,

35:28 even after preschools reopened.

35:30 Kids did not come back

35:31 into preschool,

35:32 in particular kids of lower socioeconomic status households.

35:35 So that's message one.

35:38 Message too is that in addition to these preschool closures,

35:41 are not unrelated,

35:43 but in addition to that,

35:44 there were large drops in early childhood development measured in different ways.

35:47 So first on the left,

35:48 this is from data we collected ourselves

35:51 in Bangladesh,

35:52 where there was a cross section of children in villages,

35:55 these are 2 year olds,

35:56 cross section of children in villages in Bangladesh at age 2 in 2019,

36:01 and then we collected data in 2021.

36:04 So we went back and saw

36:05 Other children in those same villages.

36:07 And what you see is big,

36:09 big drops in early childhood development in

36:11 all dimensions of early childhood development,

36:13 and particularly large drops again

36:16 amongst children whose parents have low education levels.

36:19 So big drops and disequalizing declines as well.

36:24 And now look at the graph on the right.

36:25 This is from Sobral.

36:27 Sobral is

36:27 sort of the star municipality in Brazil in terms of learning outcomes.

36:32 They've

36:32 Better than anybody for,

36:34 for,

36:34 for a very long time.

36:36 So this is a high performing municipality,

36:37 if you want.

36:38 And these are the learning outcomes for children who are 4 years of age

36:42 in Sobral.

36:44 And what you see is the line for 2019 is this is how much a child normally learns in

36:48 in preschool.

36:50 And you see,

36:51 it's been scaled,

36:51 so it's 1 standard deviation is what they learn normally.

36:54 And what do you find in 2020,

36:56 they learn about 0.3,

36:57 0.4 standard deviations.

36:59 So they've learned about 30%,

37:01 40% of what they would have learned normally,

37:04 in a normal year,

37:05 during the pandemic.

37:06 So that's for,

37:07 for,

37:07 for preschool age children and for young children

37:10 even younger than preschool on the left.

37:13 So now let me go to school age children.

37:16 What happened during the pandemic is that schools

37:18 closed for an inordinately long period of time.

37:22 So that's 0.1.

37:23 There were big differences across regions.

37:25 So South Asia and Latin America and the Caribbean closed schools on average for

37:29 55 weeks.

37:30 It's crazy.

37:31 Schools completely closed for 55 weeks.

37:34 This didn't have to be so,

37:35 and I'll make the argument that it didn't have to be so.

37:37 First,

37:37 look what happened in ECCA.

37:38 Schools essentially closed for about 4 months,

37:41 as opposed to a year and a half years in South Asia and Latin America.

37:44 Second,

37:45 even within regions,

37:46 there was huge variation in the length of school closures.

37:49 So take,

37:50 I don't care,

37:51 whatever,

37:51 take East Asia and the Pacific,

37:52 which had about medium level school closures.

37:55 On average,

37:55 you have countries like the Philippines that closed schools

37:58 totally for 2 years.

37:59 Nobody went to school for 2 whole years,

38:01 and you have countries like Vietnam that look like kind of like ECCA.

38:04 The

38:05 countries that closed like Vietnam,

38:06 closed schools for 3 or 4 months,

38:07 and then fully reopened schools.

38:09 And if you look at Latin America,

38:11 that's exactly what you see in Latin America as well.

38:13 And interestingly enough,

38:14 the country that reopened schools first,

38:16 Uruguay,

38:17 quite quickly,

38:18 was the only country

38:19 that also had an effective distance learning uh uh

38:24 program.

38:25 So even though they were the only ones

38:26 who had actually competently administered distance learning,

38:29 they quickly realized distance learning was

38:31 no substitute for in-person learning,

38:33 and they were the first in Latin America to reopen schools compared to a

38:36 country like Honduras or Peru that also

38:39 essentially kept schools closed for two years.

38:41 So huge.

38:41 Variation,

38:42 really,

38:43 variation that's really hard to explain.

38:45 So here now we have GDP per capita in logs,

38:48 and you see what is the length of school closures.

38:50 On the left,

38:50 we have

38:51 countries in in in in South Asia.

38:54 And you see,

38:54 I mean,

38:55 Bangladesh closed schools for 2.5 times as long as India did.

38:59 Sri Lanka closed it for somewhere in between there.

39:01 Now go and look at la.

39:02 You have 3 countries that are essentially very

39:04 similar in terms of their GDP per capita,

39:06 in terms of their government effectiveness,

39:08 and so on.

39:08 And nevertheless,

39:10 some countries,

39:10 in this case,

39:11 Mexico,

39:11 closed schools for twice as long as did Argentina.

39:14 So in some sense,

39:14 what I'm saying is,

39:15 there was no real rhyme or reason

39:18 for why some countries closed schools for much longer than others.

39:23 So what happened now,

39:25 once schools reopened,

39:26 and we can actually see what happened.

39:27 So first,

39:28 most kids

39:29 in most countries,

39:31 I'm talking now school age kids,

39:32 as opposed to preschool aged kids,

39:34 most kids went back to school,

39:37 especially

39:38 in

39:39 middle,

39:39 upper middle income countries,

39:41 Essentially,

39:42 everybody went back to school,

39:44 once schools reopened.

39:45 There really wasn't much of a,

39:46 we have little evidence

39:48 of any sort of permanent effect on dropouts or anything like that.

39:52 If you look at lower income countries or lower middle income countries,

39:54 a little bit more of a mixed picture,

39:56 what you'll see is

39:57 substantial drops,

39:58 about 4% points in Ethiopia,

40:00 about 6% points in Pakistan.

40:01 So something to worry about.

40:03 In this case,

40:03 we don't know why it was so in Ethiopia and Pakistan,

40:05 and not so

40:06 in the middle income countries,

40:08 but this is a fact.

40:09 In the upper middle income countries,

40:11 essentially no effect on school enrollment and attendance.

40:15 But huge.

40:16 I don't know how to better

40:18 I'm afraid I sound like Donald Trump,

40:19 and it's a huge,

40:20 but in any event,

40:22 be that as it may very large effects

40:25 on learning outcomes.

40:28 So look first at the graph on the left.

40:30 Sao Paulo is the richest state in Brazil.

40:32 So this is they do

40:34 census-wide data collection,

40:35 census-wide testing,

40:36 all kids,

40:37 every kid.

40:38 They've been doing it for about

40:40 over a decade.

40:41 Over here,

40:42 we're just plotting the results for a decade.

40:44 So what you see is

40:46 comparing 2019 to 2021,

40:48 big drop,

40:49 but especially

40:51 large drop for the youngest children.

40:54 And in some sense,

40:56 that shouldn't come as a surprise.

40:57 All of us who had children during the pandemic,

41:00 sort of,

41:00 it was easier to get your child to

41:02 do something

41:04 educational

41:05 if that child was,

41:06 say,

41:06 1213,

41:07 1415 years of age.

41:09 Good luck trying to get your 8-year-old to pay

41:10 attention to some Zoom classes on a daily basis.

41:13 So

41:14 huge drops for the youngest children.

41:15 So in those children in grade 5,

41:17 Lord only knows what happened to the kids in grade 12,

41:19 and 3,

41:20 but

41:21 every reason to believe it,

41:22 if anything,

41:22 it was worse,

41:23 not so much in terms of what happened to some of the somewhat older children.

41:27 Now look at the right.

41:28 This is in Guanajuato,

41:29 a state in Mexico.

41:31 So this is interesting.

41:32 The yellow line is,

41:34 let's start with the green line,

41:35 that is

41:35 the amount,

41:37 the test score that grade 5 children had

41:40 in 2020.

41:42 Now you apply the same tests,

41:44 they apply the same test

41:46 in 2020 and in 2021,

41:48 and they apply it to grade five and grade six.

41:51 So there's one message you want to take from the graph on the right,

41:53 from the figure on the right.

41:55 Kids in grade six,

41:57 or of age to be in grade six

41:59 in 2021,

42:00 right after schools reopened,

42:02 knew less

42:04 in terms of their math knowledge than kids in grade five.

42:07 In the year before the pandemic.

42:09 So it was overall,

42:10 there were learning losses.

42:11 There,

42:11 it wasn't just that nobody learned anything,

42:14 nobody did learn very much on average,

42:16 but actually,

42:17 on average,

42:18 there was less knowledge for older children

42:20 in 2021 than there was for somewhat younger children in 2020.

42:24 And that's the point,

42:25 actually,

42:26 I have a,

42:26 I have a second slide on that,

42:29 but I'll I'll I'll I'll get to that in a minute.

42:31 So,

42:32 you know,

42:32 is this the case just for Guanajuato and Sao Paulo and a couple of other states,

42:36 it's hard

42:37 to make sense of all the data from different

42:40 uh uh different testing programs,

42:42 from different evaluations and so on.

42:43 We do our best to make sense of that.

42:45 And what we find here is on the horizontal axis,

42:49 well,

42:49 on the vertical axis,

42:50 sorry,

42:50 on the horizontal axis,

42:52 we have the length of school closures in months,

42:55 and on the vertical axis,

42:56 we have the learning loss in months,

42:58 as best we can calculate it.

43:00 And so if you're on the,

43:01 if you're on the 45 degree line,

43:03 it means you basically you learn nothing.

43:05 If your school was closed for 4 months,

43:06 you lost 4 months of learning.

43:07 If your school was closed for 1 year,

43:08 you lost a year worth of learning.

43:09 If you're below the line is on average,

43:11 you learned something,

43:12 and if you're above the line,

43:13 it means not only did you not learn anything,

43:15 you actually lost learning relative to where you would have been,

43:19 relative to what you had.

43:21 In the absence of the pandemic.

43:22 And that's the case,

43:23 as you see,

43:24 in particular,

43:25 for the lower income countries.

43:27 So we have

43:28 Ethiopia,

43:29 uh,

43:29 Malawi,

43:30 and Bangladesh.

43:31 In all of those,

43:32 the,

43:32 the amount of lost learning in months was larger

43:37 than the length of school closures also in months.

43:41 So a particularly bad effect for lower income countries,

43:44 both because they closed,

43:45 some of them closed schools for a very long time,

43:47 and

43:48 because for any given amount of school closures,

43:50 it translated into a larger learning loss in poorer countries.

43:54 And that's basically what this,

43:56 what this graph shows.

43:57 This is Bangladesh.

43:58 We have data for grade 6,

43:59 grade 8,

44:00 for two codes,

44:00 the 2020 and the 2022 code.

44:03 So let's look at the figure on the left,

44:04 it refers to math.

44:06 In grade 6,

44:07 a child can answer 67 questions normally correctly.

44:10 And in grade 8,

44:12 in the absence of the pandemic,

44:13 so in 2020,

44:15 on the same test,

44:16 that child can answer 85% of the questions correctly.

44:19 Now,

44:19 what happens to grade eight children in 2022?

44:23 They can answer only 59% of the questions correctly,

44:26 so they can answer the grade eight.

44:28 This is like the Guanajuato thing.

44:30 The grade 8 children

44:31 can answer

44:33 after being exposed to the pandemic,

44:35 can answer fewer questions correctly

44:38 than the grade 6 children who were unexposed.

44:41 So they're more than 2 years behind.

44:43 They're like 2.5 or 3 years behind relative to where they should have been,

44:47 had there not been a pandemic and had there not been school closures.

44:52 Last couple of slides is,

44:54 OK,

44:56 right,

44:56 there are these big learning losses.

44:57 Maybe kids just bounce back and we're set.

44:59 It's not that big of a deal.

45:01 You have this learning loss,

45:02 you know,

45:03 everybody,

45:03 you sort of converge back to some sort of

45:05 pre-pandemic mean,

45:06 everybody's OK,

45:07 not a big deal.

45:08 Not so.

45:10 Here's evidence from the US,

45:11 um,

45:12 which there's been a lot of work by by Tom Kane and others at Harvard and elsewhere,

45:17 which basically says,

45:18 no,

45:19 that isn't what happened.

45:20 In the US by 2023,

45:22 only about a third of the learning loss had been made up on average.

45:27 So 2/3 were still,

45:28 in 2023,

45:29 kids were still 2/3 of the way behind relative to the learning loss that they'd had.

45:34 That's 0.1.

45:35 0.2 that's the second point is,

45:37 it's not in the,

45:37 in the,

45:38 on the slide,

45:39 huge heterogeneity across states

45:42 and across

45:43 school districts within states.

45:45 Heterogeneity that is hard to explain just in terms of income or anything like that.

45:48 So some states and some school districts seem to have made up a lot of the difference,

45:54 and some school districts made up none of the difference,

45:57 or even continue to have further learning losses,

45:59 even into 2023.

46:01 So that's the evidence that we have

46:03 for for developed,

46:04 or one piece of evidence that we have for developed countries.

46:06 Now,

46:06 in developing countries,

46:08 again,

46:08 we don't have

46:10 For,

46:10 you know,

46:10 70 papers and 42 randomized evaluation or anything,

46:14 but we have some pieces of evidence.

46:16 We have evidence

46:17 from Guanajuato,

46:18 we have evidence from Tamil Nadu,

46:19 and we have evidence again from Sao Paulo.

46:21 And I'm just going to summarize that very,

46:22 very quickly.

46:24 There was some

46:25 some convergence back to where kids would have been,

46:28 but it is probably about half,

46:31 40,

46:31 50% of the learning losses have been made up.

46:34 It depends a little bit from country to country.

46:36 I wouldn't put a lot of emphasis on the country by country variation.

46:39 We just don't have enough,

46:40 you know,

46:40 all of these have confidence intervals,

46:42 so we don't just don't have enough to really be able to make a big deal out of that.

46:46 But my read of this is still

46:49 a lot of lost learning that has not been made up.

46:53 And some evidence,

46:55 and we can talk more about that in the policy panel,

46:57 that some things that governments did

46:59 to try to make up those learning losses helped.

47:02 And again,

47:02 I,

47:03 you know,

47:03 it's not enough here to be able to say,

47:05 well,

47:05 the most effective thing was to do this

47:07 afterschool remediation program,

47:08 rather than,

47:09 you know,

47:10 online tutoring or whatever.

47:11 But there's some evidence

47:12 that some of the recovery can be explained

47:14 by

47:15 the fact that some governments in some places,

47:18 try to do something to recover these learning losses.

47:21 So I'm going to stop.

47:22 Up there,

47:22 this is,

47:23 you know,

47:23 I leave it at that,

47:24 which is

47:25 what does all of this mean for policy moving forward,

47:28 which I see

47:28 as a as a brilliant segue to the policy panel,

47:31 which Dave is also going to be moderating,

47:34 and

47:35 I think I'm supposed to stay on the stage and

47:36 we get Mamta and Michael to come up as well.

47:38 Thank you very much.

47:39 That's right.

47:39 Let's give Norbert a hand.

47:45 All right.

47:45 Let's,

47:45 uh,

47:46 Monta Murti,

47:47 Vice President for Human,

47:48 uh,

47:49 uh,

47:49 Development here at the World Bank,

47:50 Michael Kramer,

47:51 University of Chicago professor.

47:53 Uh,

47:54 let's get you both on the stand.

47:55 Yeah,

47:55 Norbert,

47:55 why don't you come and join the party.

47:57 Um,

47:59 You are

47:59 that Michael Kramer.

48:02 All right,

48:02 well,

48:02 I wanna thank,

48:02 um,

48:10 again our presenters,

48:11 uh,

48:12 Mushfiq and Joanna and Norbert for setting the stage so powerfully.

48:15 I would love to have a full session

48:17 on any of those presentations,

48:20 um.

48:22 You know,

48:22 as we think about this,

48:23 um,

48:24 you know,

48:25 what does all this mean for policy?

48:26 We've seen a lot about what these,

48:28 you know,

48:28 what these shocks did,

48:30 um,

48:31 and as we think not just about COVID,

48:34 but as we think about the next pandemic,

48:36 right?

48:37 Um,

48:37 we don't know exactly when it'll be,

48:39 we don't know exactly what it'll be.

48:40 Like Muksviq said,

48:41 not all of the lessons will likely carry over.

48:44 Um,

48:45 so,

48:46 Norbert,

48:46 just coming on the stem,

48:47 and since we didn't get any questions for you,

48:49 what would you say was the biggest thing?

48:52 That when it comes to human capital,

48:54 that countries did wrong,

48:56 and er,

48:57 you know,

48:57 what you'd say they should do differently next time.

49:02 Suppose to use the

49:05 Uh,

49:07 Is this right?

49:08 OK.

49:09 I am

49:10 technology handicapped.

49:11 Um,

49:12 I asked my 10-year-old daughter to help me out with,

49:14 with technology,

49:15 so bear,

49:16 bear with me with that.

49:18 I think there are two,

49:19 from the,

49:19 from the previous pandemic.

49:21 I think there are two lessons that I would take at least on

49:24 with regards to education,

49:25 which is what I'm going to talk about.

49:27 The first is,

49:28 do not close schools

49:31 unless you absolutely have to close schools

49:34 and reopen them

49:35 as soon as you can.

49:37 It made no sense in the pandemic

49:39 for,

49:40 say,

49:40 the Philippines or Peru

49:42 to have schools closed for two years.

49:45 The consequences of this were devastating.

49:48 So that's the,

49:49 the,

49:49 the first lesson I would take

49:51 is that.

49:51 And I want to point out,

49:53 and I don't,

49:53 I took out some slides because I was told,

49:56 yeah,

49:56 you have too many slides,

49:57 you're gonna,

49:58 you're not going to get through this.

49:59 The American Academy of Pediatrics

50:02 was recommending full school reopening.

50:05 In June 2020.

50:08 In June 2020,

50:10 the American Academy of Pediatrics said

50:13 the costs of school,

50:14 we have very little evidence of infection happening through schools,

50:18 and the costs of school closures are immense

50:21 for the learning and well-being in general of these children.

50:25 We did not heed,

50:26 I'm not talking about the US in general,

50:27 I'm talking in general.

50:28 We did not heed this kind of advice.

50:31 Policymakers closed schools,

50:33 and in,

50:33 in some sense,

50:34 what I feel is they kind of

50:35 got locked into place in a particular way.

50:37 They kind of got locked themselves into,

50:39 we're not reopening schools until there's no evidence that

50:41 anybody's ever gonna get infected under any circumstances,

50:44 and once they,

50:44 once you sort of

50:45 backed yourself into that.

50:46 Corner,

50:47 it's hard to get out of that corner.

50:48 So that's,

50:49 I think,

50:50 one lesson.

50:51 Um,

50:51 and I would say also that in the case of developing countries,

50:54 this is a time

50:55 when public transportation was already open.

50:57 Everybody,

50:57 I mean,

50:58 all of you who have been in public

50:59 transportation in developing countries know what that's like.

51:01 So people,

51:02 you know,

51:03 cheek to jowl with each other at a time when schools were still closed,

51:07 restaurants were open,

51:09 workplaces were open,

51:10 everything was open except school.

51:11 School was the one thing that that wasn't open.

51:13 So the first thing I would say is,

51:16 Close schools only if you really absolutely have to,

51:19 and be aware of the costs of closing schools,

51:22 because they are very large,

51:23 and as I say,

51:24 they are lingering.

51:25 So that's 0.1.

51:26 0.2 I'm just going to flag it,

51:28 and perhaps there's a follow-up question,

51:30 which is,

51:30 you got to get ready for this stuff now.

51:33 The country,

51:33 some of it is just the countries that experienced pandemics before did better.

51:36 Absolutely,

51:37 I agree with Mushfik on this.

51:38 We have clear evidence of that.

51:40 But we also have evidence that countries that are prepared

51:43 for some kind of systemic shock like that beforehand,

51:46 and I can go into details of this,

51:47 did a lot better

51:48 than those that it had not.

51:50 And so in some sense,

51:51 what you want to do is you want to avoid the sort of

51:53 cycle of what,

51:55 what

51:55 uh people in health in particular talk about of,

51:58 what is it,

51:58 uh

52:00 panic and neglect.

52:01 At the time,

52:02 you panic,

52:03 and then,

52:03 OK,

52:04 it's over,

52:04 you don't do anything,

52:05 and then the next time,

52:06 you panic again,

52:07 as opposed to preparing.

52:08 So my advice is prepare now.

52:09 And I know I've gone on for a little bit too long,

52:11 but

52:12 But that's,

52:12 that's,

52:13 that's sort of my summary of,

52:14 of what we have to say about this.

52:16 Thank you so much,

52:16 Norbert,

52:17 and I,

52:17 I hope that we'll,

52:18 uh,

52:18 in a few minutes,

52:19 I hope we can circle back on a couple of those things.

52:22 Um,

52:22 so Mamta,

52:23 we've seen,

52:24 you know,

52:25 Norbert's talked a lot about the education effects.

52:27 We saw Jowana talking about employment shocks,

52:30 uh,

52:31 you know,

52:32 from the shocks,

52:33 uh,

52:33 Mushfaq,

52:33 a variety of lessons.

52:36 So across all of these,

52:37 I feel like there's a real theme of

52:39 kind of,

52:40 of resilience and.

52:42 What countries need,

52:44 you know,

52:44 what sort of,

52:45 what lessons countries can take away

52:48 from

52:49 sort of these studies and from the experience as you

52:52 worked with countries over the course of the pandemic.

52:55 So protecting human capital a little bit

52:56 more broadly than Norbert's talking about,

52:58 what would you say are kind of the big one or two takeaways?

53:05 Thank you.

53:06 Um,

53:06 thank you,

53:07 Dave,

53:07 and thank you for inviting me to this panel.

53:09 Um,

53:11 That's a big question.

53:12 I'm gonna give you a very short answer.

53:15 Um,

53:15 uh,

53:17 first of all,

53:17 I think protecting income.

53:20 Which is a big constraint,

53:22 uh,

53:22 for people to access services,

53:25 whether they're healthcare or whether it's healthcare or whether it's food,

53:29 uh,

53:30 um,

53:30 is quite important.

53:32 And,

53:32 um,

53:33 while,

53:33 uh,

53:34 low and middle income countries were quite constrained in terms of fiscal space,

53:38 the ones that were able to very quick either had a social registry

53:43 or were able to quickly set it up,

53:44 Mushfiq gave this great example of Togo,

53:47 um,

53:48 were able to provide some sort of floor under people's incomes,

53:52 and that was hugely important.

53:53 Many kids get their,

53:55 get food from,

53:56 uh,

53:56 early childhood centers,

53:58 and if your kid is not going to an early childhood center,

54:00 you need to be able to feed them.

54:02 And that means having an income at a point in time when you may,

54:06 I mean,

54:06 the economic activity has slowed down.

54:09 So,

54:09 I would say that,

54:10 um,

54:11 uh,

54:12 a very important thing is having a social registry,

54:15 uh,

54:16 and a means of making payments,

54:18 how,

54:18 however small,

54:19 in order to provide a floor under people's income

54:22 in the context of a supply and demand shock,

54:24 which is what a,

54:25 what a pandemic,

54:26 uh,

54:27 is.

54:28 Um,

54:28 the second thing I would say is that,

54:30 um,

54:30 uh,

54:31 being able,

54:33 being able to prepare,

54:35 being able to detect

54:36 that there's an outbreak

54:38 and put it out quickly,

54:40 uh,

54:40 outbreak of a virus is,

54:42 is very important,

54:43 and countries that had been,

54:45 had learned from Ebola were able to do it.

54:47 I'm hoping that there's some muscle memory now

54:50 in many countries and they're better able to do it.

54:53 It actually costs not that much

54:56 to have a reasonable surveillance system in place.

55:00 Now countries are different and the geographies are different

55:03 and the size is different and all of that,

55:05 but the basics of a surveillance system

55:08 is really having

55:09 community level people

55:12 who are reasonably well trained.

55:14 And can spot,

55:15 oh,

55:16 there's something going on here.

55:18 I see this outbreak of infections and there's a cluster,

55:21 right?

55:22 And

55:22 being able to report that up

55:24 and being able to respond quickly to that,

55:27 I think is very important.

55:28 And,

55:29 and,

55:29 and countries that are,

55:31 were able to do that,

55:32 were,

55:33 were better placed and,

55:34 and ready.

55:35 And That being able to contain that outbreak is,

55:39 is,

55:39 you know,

55:40 is a huge benefit.

55:41 It,

55:41 it means your economic activity doesn't slow down and,

55:44 and all of that.

55:45 I would say those are two really important things,

55:48 having a social registry and being able to put a floor under people's income,

55:51 and being able to detect

55:53 an outbreak very quickly and stamp it out.

55:56 Now.

55:58 We also know from this that it's sometimes very hard,

56:01 right?

56:01 We don't know what the next,

56:03 next virus is gonna be like.

56:05 Uh,

56:05 I don't know if you've watched,

56:06 uh,

56:07 the Planet of the Apes movies.

56:08 I'm a great fan of them.

56:10 Well,

56:10 in that,

56:11 the virus that,

56:11 that breaks out from a lab,

56:13 actually the survival rate is 1 in 500.

56:16 It's,

56:16 it's in a very incredibly infectious virus.

56:19 So if it's a virus like that,

56:21 then,

56:22 you know,

56:22 we're really,

56:22 we're really sunk.

56:24 And then being able to,

56:25 Access,

56:27 um,

56:27 you know,

56:28 masks,

56:29 uh,

56:29 whatever it is that you need,

56:30 hand sanitizer,

56:32 eventually vaccines,

56:33 maybe,

56:33 uh,

56:34 quickly and efficiently is very important,

56:36 and that is a huge governance challenge,

56:38 and I,

56:39 I just want to put that out there.

56:40 Uh,

56:41 it's a huge governance challenge because,

56:43 uh,

56:43 not every place manufactures these things,

56:46 uh,

56:46 not everybody's able to get access to it.

56:48 One of the things we learned this time around is

56:50 that countries were able to leverage their diplomatic connections.

56:54 And get access to things.

56:55 They were able to get access to masks,

56:57 they would get access to oxygen,

56:59 they were able to get access to vaccines based on their diplomatic connections.

57:04 So

57:04 if I was a national leader,

57:06 I would make sure that I am actually

57:09 maintaining good relations

57:12 with a variety of actors who can provide me with some of

57:15 the things that I might need for my country in the.

57:18 In the event of a,

57:20 of a,

57:20 of,

57:21 of a future pandemic.

57:22 That's not in the realm of economics,

57:24 that's in the realm of,

57:25 of politics and,

57:26 and leadership,

57:27 but

57:27 it's pretty important to,

57:29 to survival.

57:30 We can talk about,

57:31 um,

57:32 uh,

57:32 international governance of,

57:34 of,

57:34 of,

57:35 of,

57:35 uh,

57:36 of,

57:36 you know,

57:36 how do,

57:36 how do we share

57:38 things that are needed to save

57:40 populations from a pandemic,

57:41 but maybe in the next round.

57:43 No,

57:43 thank you so much,

57:44 and I think,

57:45 you know,

57:45 your point on protecting incomes,

57:46 your second point on surveillance,

57:48 I think also speaks to

57:50 Mushfiq's point that countries that had previous experience managing

57:54 a pandemic or an epidemic of some sort,

57:57 um,

57:57 we saw significant,

57:58 significantly better management and so

58:01 one can only hope that there will be some,

58:03 some learning from this time around broadly.

58:06 So,

58:06 uh,

58:07 another thing in one of the earlier presentations that we saw was,

58:10 uh,

58:11 the dramatic inequality in vaccine coverage,

58:15 well into,

58:17 uh,

58:17 the rollout of vaccines.

58:19 So,

58:19 uh,

58:20 Michael,

58:20 obviously you've worked on vaccines a little bit.

58:23 Um,

58:24 that was a joke.

58:27 Michael's worked on vaccines a lot,

58:29 so

58:29 if you,

58:30 you know,

58:30 how would you say,

58:31 how do we think about closing this massive equity gap that we saw in the figure that

58:37 that Musvik showed er next time around?

58:40 Right

58:44 So I think there are several,

58:46 uh,

58:46 several steps.

58:47 One

58:48 is

58:49 just increasing more capacity

58:52 at the global level.

58:54 You know,

58:54 that's something that I think

58:56 might initially think,

58:57 well,

58:57 that's just an efficiency argument

58:59 and that's nothing to do with equity.

59:01 Well,

59:01 first,

59:02 let me just reinforce how important that is from an

59:05 efficiency point of view and then let me tell you why I think that's also important,

59:08 uh,

59:09 uh,

59:09 from an equity point of view.

59:12 You know,

59:12 we just heard

59:14 the educational losses alone were $21 trillion.

59:19 OK.

59:20 The chance of uh,

59:22 my,

59:22 I might be slightly out of date of this,

59:24 but the chance of a pandemic are about 2% per year.

59:28 This is all worked out uh much more

59:30 thoroughly in a paper by uh Rachel Glenister and

59:33 Chris Snyder.

59:34 Yeah,

59:35 if you,

59:35 if you take a 2% chance,

59:38 that means the expected cost each year,

59:41 um,

59:42 call that

59:43 $420 billion.

59:44 I think they get $700 billion because

59:46 they're including things beyond just education losses.

59:50 OK.

59:50 What's the present value of that?

59:52 You know,

59:53 uh,

59:54 you know,

59:54 that's,

59:54 that's,

59:55 uh,

59:55 that's,

59:56 you're back up to $4.2 trillion again.

1:00:00 So how much,

1:00:01 you know,

1:00:01 how much should we be spending?

1:00:04 If,

1:00:04 if having some

1:00:06 extra vaccine capacity available,

1:00:10 not just vaccine capacity,

1:00:11 obviously mask capacity,

1:00:13 we could,

1:00:14 there,

1:00:14 there are masks that that work uh that are reusable.

1:00:17 Hospitals don't use them

1:00:19 because it's

1:00:20 a little bit easier to use,

1:00:21 uh,

1:00:22 use other masks.

1:00:23 Well,

1:00:23 you could get these reusable masks and require hospitals

1:00:26 to have those,

1:00:27 so if there's an emergency,

1:00:28 they would be needed.

1:00:29 So every category.

1:00:30 of

1:00:31 of of relevant equipment,

1:00:33 you know,

1:00:33 the cost of maintaining a stockpile of that

1:00:36 is really trivial.

1:00:37 Now,

1:00:37 obviously

1:00:38 there are certain adjuvants that are used in multiple vaccines.

1:00:41 We can do things like that.

1:00:42 We obviously don't know exactly what's going to hit us,

1:00:45 but there's a lot of preparation that could be done now.

1:00:47 You could put

1:00:48 vaccine capacity in place so you could switch it over

1:00:51 once a vaccine was distributed,

1:00:52 was developed.

1:00:53 It only took.

1:00:55 Uh,

1:00:56 the vaccines were developed

1:00:58 very,

1:00:58 very rapidly

1:00:59 after the,

1:01:00 after the,

1:01:01 uh,

1:01:01 after the,

1:01:02 the,

1:01:02 um,

1:01:03 after the pandemic hit.

1:01:05 This was a matter of,

1:01:06 I forget the exact time.

1:01:07 I've been trying to forget like everybody else,

1:01:10 but you know,

1:01:10 a month or something like that.

1:01:12 The,

1:01:12 um,

1:01:13 the,

1:01:13 what took time was testing them

1:01:15 and then building the manufacturing capacity.

1:01:18 You know,

1:01:19 the value of moving a vaccine ahead by a few days.

1:01:24 Greatly exceeds when you're talking about numbers like $21 trillion.

1:01:29 Any of these costs,

1:01:30 you know,

1:01:30 we estimated that Operation Warp Speed

1:01:33 in the US,

1:01:34 if that had advanced vaccine vaccine availability in the US by 12 hours,

1:01:39 that would have paid by itself.

1:01:41 OK,

1:01:41 so,

1:01:42 um,

1:01:43 so first putting in a bunch of capacity that

1:01:45 makes sense for individual nations on their own,

1:01:47 it makes sense globally.

1:01:49 And why is this,

1:01:51 uh,

1:01:51 I'll come,

1:01:52 you know,

1:01:52 um,

1:01:53 why is this an equity investment?

1:01:55 Well,

1:01:55 if we look at this at the global level and we take a fixed,

1:01:58 say,

1:01:59 that high income countries are going to get access first,

1:02:02 I don't think we should take that as fixed,

1:02:04 but let's say we take that as fixed.

1:02:06 If it's a 2 year lag till everybody gets vaccinated

1:02:09 and you cut that in half,

1:02:11 well,

1:02:11 you know,

1:02:12 if you're

1:02:13 2,

1:02:13 if you're,

1:02:14 you know,

1:02:14 a month into the queue,

1:02:16 that takes 2 weeks off your time.

1:02:18 If you're 2 years into the queue,

1:02:20 that takes 1 year off your time till you get vaccinated,

1:02:22 so just increasing global supply,

1:02:25 even without correcting any other equity problem

1:02:28 is uh is gonna help.

1:02:30 Second,

1:02:32 Oh,

1:02:32 what

1:02:33 it was

1:02:33 with uh,

1:02:34 uh,

1:02:34 I've done work on uh advanced market commitments for,

1:02:37 uh,

1:02:38 for pneumococcus.

1:02:39 Together I was asked by some,

1:02:41 some,

1:02:42 some

1:02:42 governments to think about this problem.

1:02:44 We didn't actually

1:02:45 recommend

1:02:47 an advanced market commitment.

1:02:48 We recommended something very much like

1:02:50 like Operation Warp Speed.

1:02:52 The US did that,

1:02:53 um,

1:02:54 not,

1:02:54 I don't want to claim because of our analysis.

1:02:56 The UK did something very similar.

1:02:58 The European Union was worried about,

1:03:01 you know,

1:03:01 saving a few pennies on the vaccine and didn't do it.

1:03:04 The worst case of,

1:03:06 uh,

1:03:06 of,

1:03:07 you know,

1:03:07 uh,

1:03:08 uh,

1:03:08 of worrying about,

1:03:10 um,

1:03:10 you know,

1:03:11 being penny wise and pound foolish.

1:03:13 Let me

1:03:13 talk about

1:03:14 a middle income country with,

1:03:16 you know,

1:03:16 I won't name the country.

1:03:18 Very capable policymakers,

1:03:20 they,

1:03:20 we made the same analysis,

1:03:23 uh,

1:03:23 we presented the same analysis to them.

1:03:25 They were convinced,

1:03:26 they wanted to get their order in early.

1:03:29 But they were afraid,

1:03:30 so the Operation Warp Speed and what the UK did was

1:03:33 they

1:03:34 paid to build the,

1:03:35 to get,

1:03:35 they got their orders in early

1:03:37 and they said build the factory.

1:03:38 We don't know whether this is going to work,

1:03:40 but you know,

1:03:41 fine,

1:03:41 we'll take the chance.

1:03:42 If it works,

1:03:43 it saves us

1:03:44 billions of dollars.

1:03:45 If it,

1:03:46 if it doesn't work,

1:03:47 we've spent a few million dollars or

1:03:48 you know,

1:03:49 maybe.

1:03:50 Tens or,

1:03:51 you know,

1:03:51 it's,

1:03:51 it's a very easy deal.

1:03:53 Um,

1:03:54 the,

1:03:55 these,

1:03:56 in this middle-income country,

1:03:57 they saw the logic,

1:03:58 they wanted to do it,

1:03:59 but they said to us,

1:04:01 hey,

1:04:02 if the vaccine doesn't succeed,

1:04:03 I'm gonna go to jail.

1:04:05 They're worried about being prosecuted for corruption.

1:04:08 So you know that's something

1:04:10 and you know Tristan's

1:04:13 research suggests that

1:04:15 a big part of the reason,

1:04:17 the majority of the reason,

1:04:18 you know,

1:04:19 there was obviously there were some cases of countries

1:04:22 placing embargoes and so on on on medical equipment,

1:04:25 but overwhelmingly countries,

1:04:27 the main reason countries didn't get their,

1:04:29 their uh lower income and middle income countries didn't get their

1:04:33 vaccines in as quickly is because they didn't get their orders in as quickly.

1:04:37 So

1:04:37 allowing

1:04:39 low and middle income countries that same flexibility,

1:04:42 and this is a matter of financing mechanisms which the bank could help with,

1:04:45 to say,

1:04:45 hey,

1:04:45 if you want to make an

1:04:47 investment

1:04:48 so you can get the vaccine early,

1:04:50 we're going to support you to do that,

1:04:51 that could be done for the next pandemic.

1:04:55 You know,

1:04:55 there's a bunch of other,

1:04:57 other things that we get done like human challenge trials,

1:05:00 coming up with procedures for those

1:05:02 to greatly accelerate the,

1:05:04 the vaccine development and testing process.

1:05:06 But uh,

1:05:07 but I think there's a lot,

1:05:08 a lot that can be done to get vaccines out quickly

1:05:11 and virtually any scale of investment,

1:05:14 um,

1:05:15 you know,

1:05:15 money that's a lot by vaccine standards is very low by

1:05:18 global GMP standards or uh

1:05:20 or the value of education standards.

1:05:24 Fabulous.

1:05:25 Um,

1:05:26 no,

1:05:26 thank you so much.

1:05:27 It's hard to beat that return.

1:05:29 So I know that we're um I know that we're over the scheduled time.

1:05:33 I wanna give our,

1:05:34 our panelists,

1:05:35 um,

1:05:36 I wanna ask one last question and er give them a chance for a last word,

1:05:40 so.

1:05:42 Yes.

1:05:44 And Herbert,

1:05:44 why don't you go ahead and then I'll plant mine and they can,

1:05:46 they can answer them together.

1:05:50 Um

1:05:53 So,

1:05:53 I was struck by Norbert's,

1:05:54 uh,

1:05:55 two lessons in terms of what,

1:05:57 uh,

1:05:58 how one should prepare,

1:05:59 right?

1:06:01 You,

1:06:01 you said that the,

1:06:01 uh,

1:06:02 worst damage was schools closing,

1:06:04 and the second one was

1:06:06 Uh,

1:06:06 that one has to prepare for the next one.

1:06:09 So I was looking at essential workers by industry in 2019 in the United States,

1:06:14 and there were 55 million essential workers,

1:06:17 you know,

1:06:17 uh,

1:06:18 30% were in healthcare and so on,

1:06:20 and 20%

1:06:21 were in food and agriculture.

1:06:23 Teachers are not on this list.

1:06:25 So I think one of the,

1:06:27 perhaps the best way to prepare for the next pandemic

1:06:30 is to declare school teachers

1:06:32 as essential workers.

1:06:34 Yeah.

1:06:37 Alright,

1:06:37 no,

1:06:37 that's a great point.

1:06:38 So

1:06:39 the,

1:06:39 the last question that I have,

1:06:41 and feel free to uh Norbert to respond to,

1:06:44 uh,

1:06:45 Inderrit's point,

1:06:45 chief economist privilege,

1:06:47 um.

1:06:49 So,

1:06:50 you know,

1:06:50 the question somebody asked after one of the er initial presentations was,

1:06:54 have we learned the lessons,

1:06:56 right?

1:06:56 And I think this is the question,

1:06:58 you know,

1:06:59 there will be another pandemic,

1:07:00 we don't know exactly when.

1:07:02 So,

1:07:03 do you think we're in the process of getting ready?

1:07:06 And if there were kind of one thing that you wanted to leave this panel with on sort of

1:07:12 one thing that needs to happen

1:07:14 to get there,

1:07:15 um,

1:07:16 what is it?

1:07:17 And er let's start with Mamta actually for this one.

1:07:21 Um,

1:07:21 I think we are in the process.

1:07:23 Well,

1:07:23 I,

1:07:23 I'm a,

1:07:24 I,

1:07:24 I'm an optimist,

1:07:25 so I have to say that we're in the process of getting ready.

1:07:27 Um,

1:07:29 uh,

1:07:29 but I want to link it

1:07:30 to the argue the discussion that we started at,

1:07:33 uh,

1:07:33 at the start of today,

1:07:34 the great incoherence,

1:07:36 right?

1:07:37 Um,

1:07:37 I think we are incoherent though,

1:07:39 in getting ready.

1:07:41 So,

1:07:41 we know from a paper that we wrote

1:07:44 with the WHO that the cost of getting prevention,

1:07:48 And preparedness investments by which is meant this

1:07:52 surveillance diagnostic,

1:07:54 being able to kill an infection within the first few days of its appearance.

1:07:58 The cost of that is relatively low.

1:08:00 It's about $15 billion US dollars a year for developing countries.

1:08:05 $15 billion.

1:08:06 That's nothing.

1:08:08 But we're in a world where we have created a financing facility

1:08:12 to

1:08:13 help with this,

1:08:13 and it's not raised

1:08:16 anywhere near $15 billion a year.

1:08:18 Sorry,

1:08:18 it's $15 billion a year for 5 years.

1:08:21 So it's $75 billion in total.

1:08:23 That facility has raised $2 billion in total

1:08:26 of the $75 that it needs over five years.

1:08:29 So

1:08:30 there's a great incoherence.

1:08:31 I mean,

1:08:31 Michael just talked about it.

1:08:33 We've walked through the valley of death.

1:08:36 OK,

1:08:36 Millions of people have died,

1:08:38 trillions.

1:08:38 Dollars have been wiped off

1:08:42 global GDP,

1:08:43 and the long-term consequences on,

1:08:45 on children are enormous.

1:08:46 So

1:08:47 if we were rational,

1:08:48 we would find the,

1:08:49 the $75 billion right,

1:08:51 over the next 5 years,

1:08:52 but we're nowhere near doing that.

1:08:53 So

1:08:54 the,

1:08:55 the single most important thing is to be able to detect and prevent an outbreak.

1:08:59 Forget.

1:09:00 And then we come to the medical countermeasures.

1:09:02 I won't say forget them.

1:09:03 I think we're going to need them,

1:09:04 right?

1:09:05 Um,

1:09:05 but we're nowhere near finding the resources to do that.

1:09:08 So I would say this is the single most important thing that we need to be able to find

1:09:12 the resources for and fund.

1:09:14 And we really have no excuses.

1:09:16 I mean,

1:09:16 I,

1:09:16 I,

1:09:17 I'd hate to say to my grandchildren,

1:09:19 oh,

1:09:19 I was around,

1:09:20 or my children,

1:09:21 maybe not my,

1:09:21 uh,

1:09:22 uh,

1:09:23 grandchildren.

1:09:23 I'm gonna be optimistic.

1:09:25 Um,

1:09:25 I,

1:09:25 I was around when COVID-19 happened and we had all

1:09:28 this talk and we knew how much was needed,

1:09:30 and we created a facility and,

1:09:32 and actually,

1:09:34 we didn't actually do much with it.

1:09:35 I,

1:09:36 I,

1:09:36 I would hate to be in a position where,

1:09:38 where we said that.

1:09:39 Um.

1:09:40 I,

1:09:40 I want to come back to this vaccine point.

1:09:42 I mean,

1:09:43 we don't know whether the next pandemic is,

1:09:44 is going to be vaccine amenable,

1:09:46 um,

1:09:47 but I do think thinking about

1:09:49 sharing the medical countermeasures is a very important issue.

1:09:53 And,

1:09:54 and,

1:09:55 um,

1:09:55 I,

1:09:55 I think we're not there as a world.

1:09:58 We have,

1:09:58 we are negotiating a pandemic treaty.

1:10:00 It's stuck.

1:10:01 That treaty is about sharing

1:10:04 medical countermeasures,

1:10:06 and,

1:10:06 and there's no agreement on that.

1:10:08 Um,

1:10:09 uh,

1:10:09 I think,

1:10:10 uh,

1:10:11 uh,

1:10:12 it's,

1:10:12 it's not just about putting in money,

1:10:14 putting in your orders.

1:10:15 Trade restrictions immediately come into play.

1:10:17 And that's why the pandemic treaty is so important.

1:10:20 We've got to agree that if we want to survive,

1:10:23 just imagine a Planet of the Apes

1:10:26 kind of virus,

1:10:27 right,

1:10:27 which is where only 1 in 500 is going to survive.

1:10:31 Don't we want to be in a situation where we're sharing medical countermeasures?

1:10:35 I know it,

1:10:36 it,

1:10:36 this is not a kumbaya thing.

1:10:37 This is really about,

1:10:38 don't we want to survive as humanity?

1:10:40 And so I think we need to be able to agree

1:10:44 that we should be able to share

1:10:46 at,

1:10:46 in a,

1:10:46 in a,

1:10:47 in a reasonable way,

1:10:48 whatever medical countermeasures needed.

1:10:51 I believe that the key to that is having deconcentrated manufacturing,

1:10:55 and it's an equity issue and not an efficiency issue,

1:10:58 and that's why we're trying,

1:11:00 like many other organizations to support

1:11:03 to support manufacturing and

1:11:05 of these medical countermeasures that could potentially be used.

1:11:08 We're trying to support manufacture and underserved geographies,

1:11:11 because one thing we know

1:11:13 is that the US was very good,

1:11:15 just to name names at providing.

1:11:19 vaccines to its allies.

1:11:21 So was China,

1:11:22 very good at doing that.

1:11:23 So was Russia,

1:11:24 very good at doing that.

1:11:25 And so you,

1:11:26 you want to be in a position where you have access

1:11:28 from your neighbor or from a country that values your existence,

1:11:32 right?

1:11:33 So,

1:11:33 so I think there's something about this

1:11:35 deconcentration which is going to be very important

1:11:39 to the,

1:11:39 to the effort of fight,

1:11:40 fighting a pandemic.

1:11:41 So

1:11:42 we're getting there,

1:11:43 is my,

1:11:44 my bottom line,

1:11:45 but we're going about.

1:11:46 It rather incoherently,

1:11:48 and I,

1:11:48 I would urge us to go about it more coherently.

1:11:51 I think

1:11:52 we can play a role as,

1:11:53 as an institution like the World Bank.

1:11:55 We can play a role,

1:11:56 um,

1:11:57 but I think we also have to play a role as citizens.

1:11:59 I mean,

1:11:59 after all,

1:12:00 we have a huge representation of countries here.

1:12:04 I think we also play a role as citizens in terms of advocating for the right things

1:12:08 to be done,

1:12:09 and,

1:12:09 and we should all play that role in our,

1:12:12 in our personal capacities as well.

1:12:14 Thank you.

1:12:16 Norbert,

1:12:16 are we on,

1:12:17 are we moving in the right direction?

1:12:19 And one thing.

1:12:21 So

1:12:22 I,

1:12:22 I think

1:12:23 in addition to the

1:12:24 don't close schools unless you have to,

1:12:26 and by the way,

1:12:27 Indermin,

1:12:27 I completely agree with you,

1:12:28 and some countries did give priority,

1:12:30 some developing countries did give priority in vaccine access to teachers

1:12:34 before other groups,

1:12:35 and,

1:12:35 and most did not.

1:12:36 So I'm,

1:12:36 I,

1:12:36 I completely agree with that.

1:12:42 Yes.

1:12:46 Good point.

1:12:48 Yes,

1:12:49 good point.

1:12:49 Good point.

1:12:50 Absolutely,

1:12:50 I agree with you.

1:12:52 So

1:12:53 I think it goes back to what I sort of hinted at before,

1:12:56 which is

1:12:57 You have to get ready now.

1:12:59 You can't be making this stuff up

1:13:01 when whatever shock hits.

1:13:03 It's not going to work.

1:13:04 If you're not prepared now,

1:13:06 you're not going to get prepared in the middle of a crisis where,

1:13:09 even if it's not a Planet of the Apes scenario,

1:13:12 even if it's just a sort of a run of the mill pandemic,

1:13:15 you're not going to be able to

1:13:17 get this ready while it is happening.

1:13:19 And so,

1:13:20 what does that actually mean?

1:13:22 I think what it means is

1:13:23 Figuring out

1:13:24 two things.

1:13:24 One,

1:13:25 if I have to close schools,

1:13:26 how am I going to ensure

1:13:28 that there is some way of getting these kids to learn something?

1:13:31 If some of that is distance learning,

1:13:33 you need to figure it out now.

1:13:34 If some of that is radio-based learning,

1:13:36 you need to figure it out now.

1:13:37 And you need to figure it out

1:13:39 and pilot it and evaluate and redo it.

1:13:42 And then I think you need to mainstream it.

1:13:44 It can't be just like,

1:13:44 OK,

1:13:45 so now we developed this,

1:13:46 and this is in case a pandemic hits,

1:13:47 because then you're You're going to get around to it.

1:13:49 You have to say,

1:13:50 OK,

1:13:50 we're going to develop a system of remote learning.

1:13:52 And we're going to,

1:13:54 some part of the curriculum,

1:13:55 we're going to do

1:13:56 through remote learning.

1:13:58 Because then it becomes sort of part of the fabric of the education system.

1:14:01 And then you can actually scale it up,

1:14:02 which is what the Uruguayans did.

1:14:04 And then you can also scale it down.

1:14:05 But it's sort of like,

1:14:06 yeah,

1:14:06 we developed it,

1:14:07 we tested it.

1:14:07 This kind of seems to work.

1:14:08 And,

1:14:09 oh,

1:14:09 OK,

1:14:09 now we put it on the shelf,

1:14:10 and now we wait for the pandemic hit,

1:14:12 nothing is gonna,

1:14:13 nothing is going to change.

1:14:14 And the same is with,

1:14:15 OK,

1:14:15 now try to figure out,

1:14:17 what do you do

1:14:18 on the day,

1:14:18 again,

1:14:18 this is sort of on the on the school closure.

1:14:20 What do you do on the day that the kids are back?

1:14:22 How do you get these kids to catch up?

1:14:24 Again,

1:14:24 if you're trying to figure that out on the day the school's reopened,

1:14:27 it's not gonna work.

1:14:28 You're gonna do some sort of ad hoc,

1:14:29 crazy thing,

1:14:30 and you're gonna draw a little bit from here and a little bit from there.

1:14:32 Figure it out now.

1:14:34 Figure out how do you get,

1:14:35 how do you make up learning losses?

1:14:37 And we actually know how to do that.

1:14:39 We know how to make up learning losses.

1:14:41 So technically speaking,

1:14:42 we know how to make up learning losses.

1:14:44 Look at what Pratam does in India.

1:14:45 There's a great program in Manizales in Colombia where they do exactly that,

1:14:48 that shows big impacts,

1:14:50 big ability to recover learning losses.

1:14:53 But again,

1:14:53 you need to figure this out now,

1:14:55 and in some sense you need to implement it now.

1:14:58 Sort of in the,

1:14:58 in equilibrium,

1:14:59 so that when you really need more of this,

1:15:03 you already,

1:15:03 it's sort of,

1:15:04 it's already kind of,

1:15:05 I don't want to say second nature,

1:15:06 but you've already figured it out how to do it.

1:15:08 People know what you mean.

1:15:09 The program is already there.

1:15:11 You just massively scale it up now.

1:15:13 So the big lesson that I draw is,

1:15:15 in that sense,

1:15:16 figure the stuff out now,

1:15:18 implement it now,

1:15:19 make sure the solutions work now,

1:15:21 and make it part of the sort of

1:15:23 standard package of how you deliver education services,

1:15:25 because Otherwise,

1:15:26 you're not gonna be able to respond.

1:15:28 So that's my

1:15:29 two cents' worth.

1:15:30 Oh,

1:15:30 thank you so much.

1:15:31 Before I go to Michael,

1:15:32 I just wanna announce to everybody that at midnight tonight

1:15:35 we'll be doing a showing of Planet of the Apes,

1:15:37 which apparently is a

1:15:40 huge theme here.

1:15:41 So,

1:15:41 back in this,

1:15:42 back in this room,

1:15:43 James Franco one,

1:15:44 which is the one that had the uh the virus.

1:15:47 All right,

1:15:47 Michael,

1:15:48 are we on track?

1:15:49 And uh one thing.

1:15:50 OK,

1:15:51 uh,

1:15:52 yeah,

1:15:52 we're not,

1:15:53 thank you,

1:15:53 uh.

1:15:55 Let's see if I've got this right.

1:15:57 Uh,

1:15:57 we're not,

1:15:57 we're not at all on track.

1:15:59 Um,

1:16:00 I think,

1:16:00 uh,

1:16:00 I,

1:16:01 I agree with Mamta about that.

1:16:02 I agree with,

1:16:03 uh,

1:16:03 with Norbert.

1:16:04 You know,

1:16:04 it's the theme of this conference.

1:16:07 We've

1:16:07 said we're gonna do a bunch of things and we're not making the resources available.

1:16:12 Um.

1:16:13 So what can we do,

1:16:15 you know,

1:16:15 I think the resources should be made available.

1:16:17 What can we do without the resources?

1:16:19 I think there's some institutional preparation

1:16:22 that's very low cost

1:16:24 and that we should put in place.

1:16:26 So I mentioned the middle income country policymaker

1:16:29 who felt that they would risk jail if they if they signed something.

1:16:32 Well,

1:16:33 we can set up procedures now.

1:16:34 To legitimize that

1:16:36 so that nobody would have to feel they would go to

1:16:37 jail and the World Bank could say if international institutions said

1:16:41 we're going to support these these purchases,

1:16:44 then

1:16:45 nobody would need to go to jail for that,

1:16:47 particularly if they say these are a list of reasonable candidates,

1:16:51 OK,

1:16:52 but you know,

1:16:52 more broadly.

1:16:56 Humans challenge trials.

1:16:58 This is where healthy volunteers,

1:17:00 in some cases,

1:17:01 there are plenty of people in high

1:17:02 income countries where there's good medical care,

1:17:04 young people who are willing to say,

1:17:06 yes,

1:17:06 I'll be infected with COVID and you can try out a vaccine on me.

1:17:10 Now,

1:17:11 you know,

1:17:12 complex ethical questions,

1:17:13 when are these allowed,

1:17:14 when are they not allowed?

1:17:15 But we should be setting up the panels now,

1:17:18 so if there are some groups that are at lower risk,

1:17:20 if this next disease is not fatal,

1:17:22 and if people volunteer,

1:17:24 then you can do the trials much more quickly.

1:17:26 We can set up the procedures for that now.

1:17:29 Um,

1:17:30 um,

1:17:30 you know,

1:17:32 other examples,

1:17:33 so

1:17:34 Mushfiq's presentation was

1:17:36 amazing.

1:17:37 And it was also very sad,

1:17:40 uh,

1:17:40 like the other presentations.

1:17:42 So,

1:17:43 let me focus in on the mask trial,

1:17:45 OK?

1:17:46 Now,

1:17:47 think about this,

1:17:48 we're in the,

1:17:48 we've,

1:17:49 we're in a pandemic.

1:17:52 We don't even know

1:17:53 whether masks work or not

1:17:56 and who works on it.

1:17:58 All respect to Mushfiq,

1:17:59 but why is an economist

1:18:01 having to do this,

1:18:02 you know,

1:18:03 I don't know how many years into the pandemic did he wind up doing to get this,

1:18:07 getting this trial going.

1:18:10 Yeah,

1:18:10 so why,

1:18:11 you know,

1:18:11 why was this not ready to go

1:18:14 immediately

1:18:15 after the,

1:18:16 once the pandemic started?

1:18:17 It should have been ready to go,

1:18:18 it should have been set up,

1:18:19 the procedure should have been approved,

1:18:21 the money should have.

1:18:22 Set up so it could be allocated as soon as,

1:18:24 as soon as this was available.

1:18:26 Um,

1:18:26 why,

1:18:27 um,

1:18:28 you know,

1:18:28 the same thing on school closures,

1:18:30 you know,

1:18:30 maybe,

1:18:31 maybe you could do RCTs on this,

1:18:32 maybe you couldn't,

1:18:33 you can think about that in advance,

1:18:35 but surely there was a lot of

1:18:37 non-experimental variation that could have been utilized

1:18:40 uh very well with the right data collection systems.

1:18:43 We should have thought about that because as you point out,

1:18:45 next time.

1:18:47 Maybe we will need to close schools,

1:18:49 um,

1:18:50 and there's,

1:18:50 and,

1:18:50 you know,

1:18:51 there,

1:18:51 there,

1:18:51 um,

1:18:52 uh,

1:18:53 you know,

1:18:53 something like,

1:18:54 um,

1:18:55 uh,

1:18:56 dose optimism,

1:18:57 you know,

1:18:57 filters in schools.

1:18:59 So I'm involved in a project early stage results,

1:19:02 we're seeing

1:19:02 filters,

1:19:03 this is post-COVID,

1:19:04 but.

1:19:05 Seem to be improving learning results

1:19:08 because of reduced pollution.

1:19:11 Now,

1:19:11 but

1:19:12 we should be testing those types of things,

1:19:14 you know,

1:19:15 we don't have a good,

1:19:16 we didn't have a good study ready to go at the time,

1:19:18 so all these

1:19:19 non-pharmaceutical interventions at the beginning

1:19:21 and then finally.

1:19:24 Dose optimization.

1:19:26 So for some vaccines,

1:19:28 so yellow fever,

1:19:29 there's a shortage of yellow fever,

1:19:30 I think it was Brazil,

1:19:32 um,

1:19:32 did 1/5 doses,

1:19:35 and that worked,

1:19:35 and the WHO endorsed it.

1:19:37 Well,

1:19:38 We don't know.

1:19:39 We didn't know what the right dosage was for COVID vaccines,

1:19:42 but,

1:19:43 you know,

1:19:43 based on some,

1:19:45 so I worked with,

1:19:46 started working with biostatisticians on this and people with expertise,

1:19:50 may well,

1:19:50 based on the antibody response,

1:19:52 which is now what's used for approving things,

1:19:55 one quarter doses of some of the vaccines would have worked.

1:19:58 Imagine if we had 4 times the supply,

1:20:01 Well,

1:20:01 we could have treated the world much,

1:20:02 much more quickly.

1:20:04 That research wasn't done.

1:20:06 The pharma companies don't particularly want to do that

1:20:09 and invest in that,

1:20:11 but you know,

1:20:12 we should have systems set up

1:20:14 for publicly funding dose optimization

1:20:17 from the beginning.

1:20:18 Um,

1:20:19 thanks.

1:20:21 Thanks so much.

1:20:21 OK,

1:20:22 so

1:20:23 just to er to sum up,

1:20:25 the one takeaway um from MoMA,

1:20:27 er detection systems,

1:20:30 uh,

1:20:30 from Norbert,

1:20:31 develop and mainstream,

1:20:33 institutionalize

1:20:34 the procedures now,

1:20:35 the technologies now,

1:20:36 um,

1:20:37 and from Michael,

1:20:38 do this institutional procedural preparation now,

1:20:41 so that things are ready to move.

1:20:43 So.

1:20:44 That's what we have to do,

1:20:45 uh,

1:20:46 please give a round of applause to our panelists

1:20:49 and to our presenters.

1:20:53 Thank you very much.

1:20:54 I know you'll be able to find uh these folks later in the day,

1:20:57 and we'll pass the mic back to Alan for a final word.

1:21:00 Thanks so much.

1:21:01 In the spirit of the Euros and the Copa America,

1:21:04 let me remind you that this is halftime.

1:21:07 There's another day tomorrow of,

1:21:08 uh,

1:21:09 of the conference that will be at the

1:21:11 Center for Global Development.

1:21:13 Uh,

1:21:13 today we tackle the 1st 4 areas of coherence,

1:21:16 and tomorrow we'll tackle the,

1:21:18 the,

1:21:18 the next 3.

1:21:20 The great incurrence for me is that we,

1:21:22 we struggle to stay on time,

1:21:25 but the conference was brilliant.

1:21:27 So

1:21:28 thank you to everyone.

1:21:29 A special shout out to Kathleen who

1:21:31 um helped us this morning and this afternoon,

1:21:33 and then also a special shout out to

1:21:35 Kelly who was running around and getting his steps in,

1:21:37 chasing people with a mic.

1:21:39 Thanks so much.

1:21:39 We'll see you guys tomorrow.

showAllTimestamps

no

transcript

Um, as we start the next session, I'm gonna invite the panelists to come up over here. Um, Professor C Kramer will join our last session of the day. We will be discussing a topic of utmost importance. What should developing countries do differently in the next pandemic? I'm going at this stage, invite David Evans. Our moderator, uh, who is the principal economic advisor of the social sector at this other really good institution down the road, the Inter-American Development Bank. David. Thanks so much. I am going to invite our panelists to stay in their seats for the, uh, while we're hearing presentations so that they can actually have a good view and, uh, not have to focus if they have to look down, so, um. No, no, he wants to come here. Yeah, yeah, right. Well, this is exciting. Um, so normally I would regret being the, uh, the last session standing between you and. Dinner or cocktails or whatever, what have you, but I don't in this case because we've got a fascinating panel ahead of us. Um, we, our next and last session of the day is on what low and middle income countries should be doing differently in the next pandemic. Uh, I'm keeping a mask on in honor of that, um. Obviously COVID was a massive disruption from a human capital, uh, from a human perspective and also from an economic perspective with loss of life, loss of human capital, surges in poverty, and much more. So in this panel we'll hear from three experts first who have studied this in detail. So first we'll hear from uh Mushfik Mubarak, uh, who's a professor of Management and Economics at Yale University. And then we'll hear from Joanna Silva, who's a senior economist here at the World Bank and also a professor at the Catolica Lisbon School of Business and Economics. So we'll hear from each of them and we'll have an opportunity for just a couple of questions after each speaker. After that, we'll hear from Norbert Shatty, Chief Economist for Human Development at the World Bank Group here, and he'll give us a presentation. Once Norman has spoken, we're gonna invite Michael to come back up to the stage, we'll invite Mamta to come up to the stage, Mamta Murti, the er vice president for human er Development at the World Bank, to come up to the stage and we'll have uh a brief er high level policy panel er talking about these issues from there. So, with that, we will, I will give the mic to er Mushvik. Uh, uh, thank you. I'll, So while I wait for my slides to come up, when um, Norbert invited me to do this talk, I thought, you know, 15 minutes is too short for one paper, and so I'm going to present about 5. so it's titled, you know, I, I tried to reflect back on what we learned during the pandemic. Uh, so I'm going to try and give you a few lessons that maybe we can take forward for the next pandemic, especially keeping in mind, um, what's at stake for low and middle-income countries, OK. So, the lesson number one, and I'm going to try to be very, very quick through through these messages, is that it was an unprecedentedly large economic shock. So initially, March, April, May 2020, we were really focused on the public health aspects, how the virus was going to behave, but ultimately it became clear that Um, you know, food insecurity, income losses, employment losses were the much bigger stories, right? So here, given that we have eminent leadership of the World Bank here, I'll say that that also suggests that maybe we should have had institutions like the World Bank rather than biomedically focused organizations like Say the WHO maybe play an even bigger role in coming forward and going forward, I think the World Bank should keep that in mind, right? That ultimately, when we, when, when crises happen, even if the origin is conflict, if the origin is a virus, right, um, the, the effects, effects on human life are, are, are, are large. And here, the World Bank actually played a very, I mean, so I'm showing you macro data here, right? Uh, but then when we started thinking about in March, April, May, how this was affecting the lives of people in developing countries, right, we quickly realized that we didn't have that information and nobody was also able to collect data in person, right? So in the US you could solve that problem because people leave digital traces. So we could indirectly infer what was going on. In developing countries where a lot of people are in the informal sector, right? Even government survey methods were not sufficient. So we quickly, many researchers who work around the world, we quickly put together this effort of aggregating data from across many different countries, right? And we learned things like, you know, huge, uh, income losses in the median sample, 70% loss in income, and that was also these income, employment losses was actually also translating into increases in food insecurity. People were reporting missing meals, right? And I have to give credit to the World Bank here. While this was sort of an ad hoc enterprise by some researchers at various universities, the World Bank really took this type of effort and Put it on steroids, right? And there were huge, uh, large scale surveys, like phone surveys being done in a systematic way with nationally representative samples, which I was not able to do, right? Um, second lesson that became very clear, uh, in March, April 2020, is that we can't really afford to blindly copy, we, I'm from Bangladesh, low and middle-income countries can't, couldn't really afford to just blindly copy. The pandemic response policies that were being promoted and proposed in the US and UK, like things like if you recall, flatten the curve, lockdowns, right? So things like that. So what I did at that time was, let's take that Imperial College model that that governed our response in the US and UK, right? Let me just calibrate that to data from Pakistan, Nigeria, Bangladesh, right? Let's see if you get the same answer. And quickly, you figure out you don't actually get the same answer there, right? And the intuition is really easy for me to explain in 30 seconds. One is the age distribution of the population is very different. Developing countries are much younger, and we already knew from day one that the virus was a lot more fatal to the elderly. Number 2, right, you flatten the curve because you don't want the curve of infections to go above your healthcare capacity line, so there would be excess unnecessary deaths when people are not getting access to. Uh, hospital beds, right? But now imagine you're in rural Sierra Leone, rural Bangladesh, where the healthcare capacity line is down here, right? By, by, uh, flattening the curve and delaying infections for another 3 months, you're not actually get that much out of it. That's what the model tells you, right? And so then that made us focus a lot more on the, OK, the benefits are a lot smaller from lockdowns. So what are the costs of lockdowns, right? And this is data from Nepal. I'm on purpose showing you the early round of the data collected in March of 2020, right? And what people were mostly worried about wasn't health and safety. Even in March of 2020, rural Nepalese were mostly worried about their food insecurity and the income shock. Right? So this is why we need to get that information out right away, right? To, to be able to target our responses more, more accurately. Now, here's another place where I think institutions like the World Bank could play a bigger role. So, something that became clear at that time is regardless of your position on lockdowns or not, it became clear that we needed to get money in the hands of poor people very, very quickly, right? And that we also realized that we didn't know how to do that. Right. First of all, we didn't even know who needed support, right? Because that requires a lot of data, OK? And then there were really interesting, uh, uh, innovations like in Togo, working with Josh Blumenstock, who's at UC Berkeley, you know, they figured out how to, uh, use cell phone records to make indirect inferences about who might need support, right? So we need innovations like that. And again, those innovations have to be economic, technological, those are not necessarily public health or medical innovations. OK. Uh, right, let me skip over. Yeah, and just, just to, just to say, I'm, I'm, uh, you know, I also wrote at the same time that the answer for the rich countries are actually quite different, right? In rich countries, the right thing to do was to just like, crush the virus like New Zealand did so that they could open up more quickly. Whereas countries that did not manage to have strict lockdowns, they remained, uh, their economic activity remained depressed for almost 2 years afterwards, right? Uh, OK, sorry, these were, um, uh, there were some hidden slides there, but they're shown. Um, so, uh, researchers also had an important role to play, you know, uh, people in this building and academic researchers. Uh, there was a lot of data and information and insights that we could provide that would be of value to government policymakers. Everybody was flying blind, we all were. And government really needed that support. So this is a simple insight, right? So, when you look at what was happening in Nepal, in terms of food insecurity, this is what it looked like, like the percentage of people, this is just an index of food insecurity. It actually looked not too much. It, it looked better than what Nepal looked like 6 months before. And the reason is 6 months before it was the period of lean, it was like a lean season period, right? We, we just got very lucky that when the pandemic hit with the post-harvest period. Right? But only because, you know, you had multiple years of data on what seasonal fluctuations look like, right? That we knew that this was a lot worse than what it really should be during this period, even though the benchmark is 6 months ago, it wasn't looking that bad, right? And in Sierra Leone, um, you know, about a third of all households are female-headed, uh, and we learned that Um, you know, they experienced the pandemic very, very differently. They knew less, they were engaging less in, um, defensive behaviors, etc. and much more food insecurity in female-headed households. So there was information, like more micro-level information on who the government should be targeting more, right? Fourth lesson is that it's not that low million countries are just uh recipients of information, right? There was a lot that we could teach each other as well, right? And I think a mistake we made in Bangladesh, India, Pakistan, all over South Asia, is that we were looking towards US, UK, South Korea, like getting our information from CNN and BBC was focused on what was happening in the US and UK. And we just missed the boat in terms of just looking towards countries like Sierra Leone and Liberia, which had pandemic experience, right? And they had these frugal innovations that would have been much more important for us to think about and copy rather than thinking about this, uh, sophisticated contact tracing system that South Korea could install, right? That wasn't all that useful for us, but we were looking in the wrong places, right? And, uh, I, I think the US also could learn a lot, like our leadership, uh, you know, was not, uh, very well positioned to respond to the pandemic relative to many countries in Africa, right? And there's also other things that happen, like through research in, in developing countries, right? Like gen genome sequencing of the, of the virus in South Africa, uh, which led to early identification of the Omicron variant, vaccine development. Um, trials, etc. that was happening in, in various developing countries. OK OK. And something that also became clear by 2021 is that it wasn't just developing country policymakers, even international organizations who are supposed to lead on this, CDC, WHO. I was surprised to learn they were also sort of flying blind, right? And this showed up with very strange uh and uh confused guidance around masking, right? And, and so, and talking to the WHO early on in the pandemic, we had data that look, when you look at non-experimental data on how the virus is behaving in countries that had a mask wearing norm versus not, it's clear that masking seems to do something, right? And the WHO's reaction was like, no, that's not like, without RCTs, we can't change our guidance, right? So then in the middle of the pandemic, we ran an RCT and again, it's a developing country, Bangladesh. Most of the authors of, of this, of this paper are from Bangladesh and we're residing in Bangladesh, that we ran a trial of 350,000 people to get people to wear masks and then track what happened. And what we learned from that is that you can, first of all, you can get people to wear masks. There's a set of 4 things that you have to do. That includes free distribution, but you also have to go back and harass people a little bit to keep their masks on, right? Like, but politely harass people, OK. Um, and we don't need it for Dave, obviously. Um, and, but then they keep their masks on, right? And once they keep their masks on, there is a significant reduction in COVID symptoms as well as symptomatic seropositivity, right? And in fact, this in in rural Bangladesh, this eliminated about a third of all infections for the elderly, right? Um, and, you know, in the developed world, like the Washington Post also, also learned from this, right? And, uh, and so did the WHO. And another thing that became clear during the mass trial is that I quickly moved from because the Delta wave had hit South Asia, and people were dying on the street, and we quickly stopped research and started distributing masks, right? And then it became clear that governments actually need support, even, even to figure out, like on the fly, figure out how to distribute mass, right? So we hired a person who used to work for the uh for the Gujarat state government, right? And he explained to us how to translate all of this information in a format that makes it much more actionable for government, right? And, um, and that led to a lot of mass distribution, uh, prior to the Omicron wave, OK? And final lesson is that the pandemic just didn't ever quite get over in many poor countries after it was functionally over here, right? And that comes from this picture. Let me just jump to this picture, like 18 months after vaccines came to market. Most of sub-Saharan Africa remain unvaccinated, right? Also, the Caribbean poor countries in the Caribbean, OK? Why was this? So when you ask this question, why were vaccination rates lagging, African scholars, policymakers made the good point that uh vaccines were effectively being withheld from them, right? But in response, um, the pharmaceutical companies claimed, on this is on record that, oh, no, no, no, that's not the issue. If you were to send doses, they don't get used. Vaccination, vaccine hesitancy is too high, right? So I was like, OK, that's an empirical question that can be answered, right? So let's run this survey in 15 countries. But the trick was, let's add it to add the US to it. And what you learn from that is in every single low-income, middle-income country, vaccine hesitancy was lower than it is in the US. So just on the face of it, it's just like a BS excuse, right? So it wasn't vaccine hesitancy, but we did learn is that getting vaccines to people was really difficult, right? So then our final trial is the last thing I'll say is let's go to the remote places, let's take vaccines and see what happens. And what happens is when you run this trial for 48 hours, you just take vaccines there. The number vaccinated per community moves from 5 people to about 55 people within 48 hours, right? To get from 0 to 50, you just need to get people's stuff, right? You don't have to do anything sophisticated with that, right? And we learned something from that is that, OK, like, if you're going to have to pay this cost, of going out to remote areas, it's insane to just take a single vaccine. We should be taking a bundle of useful health services, right? And so, hopefully, through the pandemic, this was my bonus lesson 8, is that we learned something useful for how to deliver health services in remote rural areas. Thanks. Thank you so much. Uh, I don't know, there we go. Uh, we are, we're just gonna take 1 to 2 questions before we, uh, move on to the next panelists. If anybody has a question from that, uh, exhilarating quick tutorial, um, of lessons. Any questions? Yes, I have one there. Very insightful. Thank you very much. Have we learned these lessons and are we ready for the next pandemic? Thank you. Yeah, um, no, we are definitely not, sure. Do we have, do we have one more, or is that, yeah, let's take this one more and then we'll, uh, and then you can answer both of those. Hello again, my name is Juliet Adams, and I'm also an educator. As you know that the, the, the COVID, the lockdown, it rocked the education, um, sector, um, with serious consequences. I would like your take on what took place. Thank you. OK, sure. That second question, I'm going to hand it over to Norbert, who's going to talk about exactly this. You'll hear 15 minutes of this. So, OK. And have we learned these lessons? I, um, I hope so, but I'm not too optimistic, right? Um, so I would say 11 reason is that, look, the next virus that comes, you know, like I said earlier in this talk that, look, this virus was really dangerous for the elderly, right? The next virus could affect the respiratory systems of children more, right? And we'll need a completely different set of responses and, you know, so we need to stay on top of things and it's not just, uh, uh, it's not just about like taking these lessons and it'll immediately translate over. That's one. But on the other hand, the reason what I'm optimistic is that I did notice very early in the pandemic in March, April, May of 2020, right? Countries with pandemic experience like Sierra Leone, Liberia, Guinea, etc. they actually reacted much better than we did in Bangladesh or in India or Pakistan, or even in, in other parts of Africa, right? And that makes me think that experience does matter, right? Um, so, I hope I, that's my optimism. Let's give another round of applause as Johanna, thanks so much. Thank you very much. So this paper is about how you adjust to transitory shock, so it's not specific about COVID, but it, it, it has implications for that. It's joint work with my, with my colleague Ana Marguerite Fernan who is over there. Um, so we have 3 main findings with this paper, and let me start with that. Normally when we think about permanent shocks like trade liberalization, we do think that they have permanent effects. However, when we think about transitory shocks, we don't think they will have permanent effects of the type of these other types of shocks. So what we show in this paper is that indeed there are permanent effects of. Repository shocks, we are gonna prove it in this case and this will happen in a setting of high informality where you could think that informality could work as a full buffer of the shock, but it doesn't. We are also gonna show you that um adjustment depends very much on the level of informality but also on whether governments protect firms uh through, you know, increasing in the level of concentration in the market or SOEs. And quite surprisingly for us and maybe for you as well, we are going to show you that most of the scarring actually occurs among incumbent workers. It's not displaced workers, which is the type of workers that we normally worry about, but it's actually workers who continue in the same firm. And now at this point you are wondering how come. This happens because firms are also scarred. And this scarring of firms, what we show in this paper will come from firm exit, but also a permanent downsize in productivity that will happen as a consequence of a transitory shock, so quite surprising. The last set of findings have to do with government responses, and we normally think there are gaps in these responses in the case of transitory shocks, but we often don't know how large these gaps are. So in this paper, we look at data for Brazil and show that of all the losses that workers had, unemployment. and both the family welfare program of Brazil compensated 6% of their losses, no more than that. Training didn't even reply at all. There is no response from training programs. So lots of the losses were borne by workers, they were permanent, they were borne by firms, they were permanent and policies didn't compensate for them. In this paper, we use administrative data for, for Brazil. Uh, and we link the matched employer employee records of Brazil with the manufacturing census and also the the the the the CEE, so the trade data. In addition, which is rare, more rare in this case is that we also Had individual level records for the benefits of unemployment insurance at the level of each person in Brazil, also the welfare program, and also the individual records of the training program, so we could see exactly which worker got what at which time. With this rich data, the way we identified the shock was that we were trying to look at the foreign demand shock that the global financial crisis created in each firm in Brazil, and what we are going to compare is the exposure of the firms to this shock, this exogenous variation in this exposure. Uh, is what's going to determine, um, you know, then the impacts that we, that we compute at the firm level. Here is just a graph to illustrate the point that this, this shock that we are going to look at was really transitory, is the deep dive at the moment of the global financial crisis. The way we look at this shock is that, We consider the importance for each firm of each destination market in their export portfolio, and then we multiply that by the change in GDP during the period of the crisis in that destination. So it's kind of the change in the GDP of the firm, if you will, that only comes from external demand, so it's in the partner country, um, and that's how we measure the, the shock, the temporary shock that each firm in Brazil has. It's exhaustionous and it had been used by others in other contexts. So what we look at, we look first at workers' outcomes. So each line in our data set is a worker and we follow him through time in his employment spell. We have then the variable of interest, the shock variable, and then we control for, uh, you know, lots of characteristics of workers, of firms and have you, you know, the fixed fixed effects at the different levels that will allow us to identify the effects. What do we find? So we find first in terms of employment, we see a permanent reduction in the, in the average month's work. You see that they, the, the, here are the plots of the coefficient of that regression, the coefficient of the shock variable, and you see that it felt and it stayed significantly lower until 2017, so a decade after. We look at wages and we see also a reduction and actually no improvement whatsoever, so they went down and stayed down. And in terms of hours worked, which is surprising for labor economies, they also went down, and again this was a temporary shock. In terms of magnitude, what we computed to be the effect on the employment on most work is about 3% and in terms of wages 6%. So these are sizeable effects. Now, let's look at these mediating factors, informality. So labor markets, local labor markets in Brazil, they differ in the, in the share of informal workers they have. And what you see there is that in places with lower informality levels, workers were better off. Then in higher the response, the response was smaller, so the loss was smaller in higher informality places. So it did work like a buffer, but it didn't shelter workers completely. They still had employment losses and wage losses, but they were smaller. In terms of the level of concentration of the market and whether you as a worker were working in an SOE at the time, only those that were in low concentration sectors and in non-state owned enterprises bear the burden of the crisis. The others, the effects were not significant. So it was a way to insulate workers, but that is, as we show in the paper, you know, the cost then is spread to others. Now, who, in terms of workers, you could either stay at your firm or move to a different sector, to a different firm in the same sector. So what happened depending on what you did, we could observe what they did. So what you see is there in the first column are the initial results. The second are the results for the workers who remain in their firm. And you see that there are losses that are significant and negative and quite large. For those that moved to other firms in the same sector, uh, there weren't losses or other firms in the tradable sector. Um, but the only other significant result is for other firms in the non-tradable sector, kind of the best way to escape the crisis losses was really uh for you as a worker to move to another firm in the non-tradable sector. And most of the losses as you see are really, uh, the coefficient is larger in the ones that remain in the initial firm. This is quite surprising. Now in terms of impact on firms, we look at different outcomes, effects on revenues, profits, on productivity calculated in different ways, and here I find the results very interesting. Here also we are plotting the coefficient of the shock variable. But now our regressions are at the firm level. What do you see there? A transitory shock, the global financial crisis, its external dimension led to a reduction in the net revenues that was permanent. It was still there in the end of the period. If you look at exit, well, it led to exit just a year after. So the ones, the firms that did exit, they, they, they did exit because of the crisis, they did it the year after, not after that. And in terms of size, you see this contraction in terms of size. What did they do to the labor force? So they adjusted the types of workers they had. They stayed with less unskilled workers and less skilled workers, but the reduction was much larger in the quantity. That they were using of unskilled workers than skilled workers. So as a consequence, the, the share of unskilled workers that each firm in Brazil was using became lower, as you can see in the last graph on the on the on the left. So they kind of adjust the composition of their workforce. What about the technology they were using, the productivity. As you can see here, they adjusted down the materials, so they use less materials. The material per worker, they did decrease. Um, the, the effects on capital per worker in Brazil were not significant, so that was not the main margin of adjustment for workers. And then as you can see in the TFP effects, and we use several measures, uh, you know, really persistent effects of, uh, on productivity. So firms that were hardest hit, they had lower productivity not for one year but for a decade. Um, so you know, 11 thing that we know, we say and, and kind of now linking to the macroeconomist literature is that there can be cleansing effects of crisis. So a good news for productivity could be one where the firms that exit the market are the weakest, the least productive, and therefore the market share of the best firms actually increases with the crisis. Here we are not in general equilibrium, but we could at least ask which firms did exit in the case of Brazil. Indeed, they were the smaller, they were the least productive, so at least in that respect, some, some good news. OK, now looking at the asset, the compensation mechanisms, how did they work in Brazil? So we know for each worker in Brazil, because of the the the type of data we have, exactly how much he lost, and we know what he received in terms of unemployment insurance and welfare programs and whether he got training and what do we find? So you there you see the response in terms of unemployment insurance. So yes, the ones, the ones that were working at firms that were hardest hit were more likely to get, they did get more unemployment insurance. They did get more uh both the familia program, but what's really the share that was compensated? So among all the losses, unemployment insurance compensated 4.3% of their losses, and Bolsa familia, so the welfare program, 1.7%. All in all, 6% of the losses were compensated. All the rest stayed with the workers and the firms. And training, like I said, didn't reply at all. But you could be wondered, this is so strange, incumbent workers, transitory shocks having permanent effects, this permanent loss in productivity of firms, are they, uh, is this just Brazil? So we did it for another country, Ecuador, that had the same type of very rich data. But had a much less flexible labor market, so there this margin that adjusted in Brazil, the workers couldn't adjust as much. So what did they do instead? Unable to adjust labor as much, they adjust capital. And what happened? The productivity losses were also persistent, but much larger, much larger. So to conclude, here an example um with data from Brazil and Ecuador of transitory shocks having very persistent effects on workers and firms, uh, these effects, uh, seen, is being seen even in context of high informality that, you know, works as a buffer but doesn't solve it all. Governments protecting firms and those at those firms being OK but not the rest. Uh, and then, uh, in terms of firms, you see this really persistent fall in productivity with different natures depending on labor market flexibility of the market. And very limited response of the existing social programs, being their unemployment insurance or welfare programs in a country where this exists, and this just exists in about 1/3 of the countries in the world. So really worrying news for, for, for, you know, worker losses, but also the extent to which policies are compensating them. Thank you. Thank you so much, Joanna. Um, do we have, uh, do we have a question or so before, uh, we go on? I'm gonna take the, uh, moderator's privilege. I have one question while the mic goes up, which is, you know, obviously we see here, like the massive potential effects of shocks like the pandemic. I'm curious if sort of how this relates to the literature that we've seen from the pandemic itself and how you feel like this applies. And we'll take the second question as well. Yeah, thanks, Joanna for such a good presentation. And very interesting work. I had a, um, it was a lot of material, so it may be obvious. How were you able to rule out, uh, an explanation whereby it's the weakest workers who stay in the firm, and that's why you're seeing what you're calling the scarring effect for those who remain rather than those who leave. So thank you very much for both your questions. These are very good questions. About the COVID is different because there was also a supply shock. Here we are just looking at the demand shock uh driven by the change in foreign demand. So, you know, those other dimensions that come with the supply side that Norbert is gonna speak. Uh, will be additional to this, so this is with a pure demand shock coming from abroad, so totally exhaustionous to you. This is wall that is already happening. Add to that the supply side shock and you get COVID effects. So that's when I, uh, I'll stay there, um, to, to, to Raquel's question, so, um, Here, here in the, in the, in this paper, what we, um, what we, what we are looking is the, is, is really um the, the, the, the effects that, you know, just, just the, the, the, the, the, the, the, the, the effects that So we are not making a statement about who stayed and who left. So maybe these workers who stayed were worse than the ones who left, but what we can tell you is that very few left, so displacement was very limited in Brazil. So in that sense, um, it's not that, you know, everybody left before the crisis, and this case was really unexpected, so they couldn't leave before the crisis. So we we we we do have some analysis comparing those who left and who didn't, and they don't seem to be different. So that's how we kind of make a call about that part, but it's like being them how they are, they did have these losses. Thank you. Thank you so much. Let's give Joanne another hand. And now we'll have our last presentation of the session, uh, Norbert Shatty, after which we'll transition into a brief round table. OK. Well, uh thanks to everybody. Uh, I'm going to be speaking about the effects that the pandemic had on uh early childhood and on education outcomes. And I'm going to be speaking mainly on the basis of a book we published about a year ago now, which looked at these effects. It was a book, uh, it was sort of the flagship, uh, flagship publication of the, of the bank in terms of human development. And I'm just going to summarize some of the evidence we have, we have in there. I want to talk a little bit about early childhood development, a little bit about school-age children, and a little bit about how much recovery has there been. What can we say about that? Let me start with young children. So I have two slides on young children. This is the first one. What it shows is, look at the graph on the left. This is the evolution of preschool enrollment rates in South Africa. And then the vertical line corresponds to the beginning of the pandemic. So what happens is preschool enrollment dropped substantially. And on the right hand side, we have this for different countries once preschools reopened. So this is no longer people are not enrolled in preschool, because it makes no sense to be enrolled in preschool if preschools are closed. These bars over there correspond after preschools reopened, and they're separate bars. For children of mothers who have primary education, at least some secondary and at least some postsecondary. So two messages everywhere we see large drops in preschool enrollment. 0.1, large. These are 15 18% point drops, even after preschools reopened. Kids did not come back into preschool, in particular kids of lower socioeconomic status households. So that's message one. Message too is that in addition to these preschool closures, are not unrelated, but in addition to that, there were large drops in early childhood development measured in different ways. So first on the left, this is from data we collected ourselves in Bangladesh, where there was a cross section of children in villages, these are 2 year olds, cross section of children in villages in Bangladesh at age 2 in 2019, and then we collected data in 2021. So we went back and saw Other children in those same villages. And what you see is big, big drops in early childhood development in all dimensions of early childhood development, and particularly large drops again amongst children whose parents have low education levels. So big drops and disequalizing declines as well. And now look at the graph on the right. This is from Sobral. Sobral is sort of the star municipality in Brazil in terms of learning outcomes. They've Better than anybody for, for, for a very long time. So this is a high performing municipality, if you want. And these are the learning outcomes for children who are 4 years of age in Sobral. And what you see is the line for 2019 is this is how much a child normally learns in in preschool. And you see, it's been scaled, so it's 1 standard deviation is what they learn normally. And what do you find in 2020, they learn about 0.3, 0.4 standard deviations. So they've learned about 30%, 40% of what they would have learned normally, in a normal year, during the pandemic. So that's for, for, for preschool age children and for young children even younger than preschool on the left. So now let me go to school age children. What happened during the pandemic is that schools closed for an inordinately long period of time. So that's 0.1. There were big differences across regions. So South Asia and Latin America and the Caribbean closed schools on average for 55 weeks. It's crazy. Schools completely closed for 55 weeks. This didn't have to be so, and I'll make the argument that it didn't have to be so. First, look what happened in ECCA. Schools essentially closed for about 4 months, as opposed to a year and a half years in South Asia and Latin America. Second, even within regions, there was huge variation in the length of school closures. So take, I don't care, whatever, take East Asia and the Pacific, which had about medium level school closures. On average, you have countries like the Philippines that closed schools totally for 2 years. Nobody went to school for 2 whole years, and you have countries like Vietnam that look like kind of like ECCA. The countries that closed like Vietnam, closed schools for 3 or 4 months, and then fully reopened schools. And if you look at Latin America, that's exactly what you see in Latin America as well. And interestingly enough, the country that reopened schools first, Uruguay, quite quickly, was the only country that also had an effective distance learning uh uh program. So even though they were the only ones who had actually competently administered distance learning, they quickly realized distance learning was no substitute for in-person learning, and they were the first in Latin America to reopen schools compared to a country like Honduras or Peru that also essentially kept schools closed for two years. So huge. Variation, really, variation that's really hard to explain. So here now we have GDP per capita in logs, and you see what is the length of school closures. On the left, we have countries in in in in South Asia. And you see, I mean, Bangladesh closed schools for 2.5 times as long as India did. Sri Lanka closed it for somewhere in between there. Now go and look at la. You have 3 countries that are essentially very similar in terms of their GDP per capita, in terms of their government effectiveness, and so on. And nevertheless, some countries, in this case, Mexico, closed schools for twice as long as did Argentina. So in some sense, what I'm saying is, there was no real rhyme or reason for why some countries closed schools for much longer than others. So what happened now, once schools reopened, and we can actually see what happened. So first, most kids in most countries, I'm talking now school age kids, as opposed to preschool aged kids, most kids went back to school, especially in middle, upper middle income countries, Essentially, everybody went back to school, once schools reopened. There really wasn't much of a, we have little evidence of any sort of permanent effect on dropouts or anything like that. If you look at lower income countries or lower middle income countries, a little bit more of a mixed picture, what you'll see is substantial drops, about 4% points in Ethiopia, about 6% points in Pakistan. So something to worry about. In this case, we don't know why it was so in Ethiopia and Pakistan, and not so in the middle income countries, but this is a fact. In the upper middle income countries, essentially no effect on school enrollment and attendance. But huge. I don't know how to better I'm afraid I sound like Donald Trump, and it's a huge, but in any event, be that as it may very large effects on learning outcomes. So look first at the graph on the left. Sao Paulo is the richest state in Brazil. So this is they do census-wide data collection, census-wide testing, all kids, every kid. They've been doing it for about over a decade. Over here, we're just plotting the results for a decade. So what you see is comparing 2019 to 2021, big drop, but especially large drop for the youngest children. And in some sense, that shouldn't come as a surprise. All of us who had children during the pandemic, sort of, it was easier to get your child to do something educational if that child was, say, 1213, 1415 years of age. Good luck trying to get your 8-year-old to pay attention to some Zoom classes on a daily basis. So huge drops for the youngest children. So in those children in grade 5, Lord only knows what happened to the kids in grade 12, and 3, but every reason to believe it, if anything, it was worse, not so much in terms of what happened to some of the somewhat older children. Now look at the right. This is in Guanajuato, a state in Mexico. So this is interesting. The yellow line is, let's start with the green line, that is the amount, the test score that grade 5 children had in 2020. Now you apply the same tests, they apply the same test in 2020 and in 2021, and they apply it to grade five and grade six. So there's one message you want to take from the graph on the right, from the figure on the right. Kids in grade six, or of age to be in grade six in 2021, right after schools reopened, knew less in terms of their math knowledge than kids in grade five. In the year before the pandemic. So it was overall, there were learning losses. There, it wasn't just that nobody learned anything, nobody did learn very much on average, but actually, on average, there was less knowledge for older children in 2021 than there was for somewhat younger children in 2020. And that's the point, actually, I have a, I have a second slide on that, but I'll I'll I'll I'll get to that in a minute. So, you know, is this the case just for Guanajuato and Sao Paulo and a couple of other states, it's hard to make sense of all the data from different uh uh different testing programs, from different evaluations and so on. We do our best to make sense of that. And what we find here is on the horizontal axis, well, on the vertical axis, sorry, on the horizontal axis, we have the length of school closures in months, and on the vertical axis, we have the learning loss in months, as best we can calculate it. And so if you're on the, if you're on the 45 degree line, it means you basically you learn nothing. If your school was closed for 4 months, you lost 4 months of learning. If your school was closed for 1 year, you lost a year worth of learning. If you're below the line is on average, you learned something, and if you're above the line, it means not only did you not learn anything, you actually lost learning relative to where you would have been, relative to what you had. In the absence of the pandemic. And that's the case, as you see, in particular, for the lower income countries. So we have Ethiopia, uh, Malawi, and Bangladesh. In all of those, the, the amount of lost learning in months was larger than the length of school closures also in months. So a particularly bad effect for lower income countries, both because they closed, some of them closed schools for a very long time, and because for any given amount of school closures, it translated into a larger learning loss in poorer countries. And that's basically what this, what this graph shows. This is Bangladesh. We have data for grade 6, grade 8, for two codes, the 2020 and the 2022 code. So let's look at the figure on the left, it refers to math. In grade 6, a child can answer 67 questions normally correctly. And in grade 8, in the absence of the pandemic, so in 2020, on the same test, that child can answer 85% of the questions correctly. Now, what happens to grade eight children in 2022? They can answer only 59% of the questions correctly, so they can answer the grade eight. This is like the Guanajuato thing. The grade 8 children can answer after being exposed to the pandemic, can answer fewer questions correctly than the grade 6 children who were unexposed. So they're more than 2 years behind. They're like 2.5 or 3 years behind relative to where they should have been, had there not been a pandemic and had there not been school closures. Last couple of slides is, OK, right, there are these big learning losses. Maybe kids just bounce back and we're set. It's not that big of a deal. You have this learning loss, you know, everybody, you sort of converge back to some sort of pre-pandemic mean, everybody's OK, not a big deal. Not so. Here's evidence from the US, um, which there's been a lot of work by by Tom Kane and others at Harvard and elsewhere, which basically says, no, that isn't what happened. In the US by 2023, only about a third of the learning loss had been made up on average. So 2/3 were still, in 2023, kids were still 2/3 of the way behind relative to the learning loss that they'd had. That's 0.1. 0.2 that's the second point is, it's not in the, in the, on the slide, huge heterogeneity across states and across school districts within states. Heterogeneity that is hard to explain just in terms of income or anything like that. So some states and some school districts seem to have made up a lot of the difference, and some school districts made up none of the difference, or even continue to have further learning losses, even into 2023. So that's the evidence that we have for for developed, or one piece of evidence that we have for developed countries. Now, in developing countries, again, we don't have For, you know, 70 papers and 42 randomized evaluation or anything, but we have some pieces of evidence. We have evidence from Guanajuato, we have evidence from Tamil Nadu, and we have evidence again from Sao Paulo. And I'm just going to summarize that very, very quickly. There was some some convergence back to where kids would have been, but it is probably about half, 40, 50% of the learning losses have been made up. It depends a little bit from country to country. I wouldn't put a lot of emphasis on the country by country variation. We just don't have enough, you know, all of these have confidence intervals, so we don't just don't have enough to really be able to make a big deal out of that. But my read of this is still a lot of lost learning that has not been made up. And some evidence, and we can talk more about that in the policy panel, that some things that governments did to try to make up those learning losses helped. And again, I, you know, it's not enough here to be able to say, well, the most effective thing was to do this afterschool remediation program, rather than, you know, online tutoring or whatever. But there's some evidence that some of the recovery can be explained by the fact that some governments in some places, try to do something to recover these learning losses. So I'm going to stop. Up there, this is, you know, I leave it at that, which is what does all of this mean for policy moving forward, which I see as a as a brilliant segue to the policy panel, which Dave is also going to be moderating, and I think I'm supposed to stay on the stage and we get Mamta and Michael to come up as well. Thank you very much. That's right. Let's give Norbert a hand. All right. Let's, uh, Monta Murti, Vice President for Human, uh, uh, Development here at the World Bank, Michael Kramer, University of Chicago professor. Uh, let's get you both on the stand. Yeah, Norbert, why don't you come and join the party. Um, You are that Michael Kramer. All right, well, I wanna thank, um, again our presenters, uh, Mushfiq and Joanna and Norbert for setting the stage so powerfully. I would love to have a full session on any of those presentations, um. You know, as we think about this, um, you know, what does all this mean for policy? We've seen a lot about what these, you know, what these shocks did, um, and as we think not just about COVID, but as we think about the next pandemic, right? Um, we don't know exactly when it'll be, we don't know exactly what it'll be. Like Muksviq said, not all of the lessons will likely carry over. Um, so, Norbert, just coming on the stem, and since we didn't get any questions for you, what would you say was the biggest thing? That when it comes to human capital, that countries did wrong, and er, you know, what you'd say they should do differently next time. Suppose to use the Uh, Is this right? OK. I am technology handicapped. Um, I asked my 10-year-old daughter to help me out with, with technology, so bear, bear with me with that. I think there are two, from the, from the previous pandemic. I think there are two lessons that I would take at least on with regards to education, which is what I'm going to talk about. The first is, do not close schools unless you absolutely have to close schools and reopen them as soon as you can. It made no sense in the pandemic for, say, the Philippines or Peru to have schools closed for two years. The consequences of this were devastating. So that's the, the, the first lesson I would take is that. And I want to point out, and I don't, I took out some slides because I was told, yeah, you have too many slides, you're gonna, you're not going to get through this. The American Academy of Pediatrics was recommending full school reopening. In June 2020. In June 2020, the American Academy of Pediatrics said the costs of school, we have very little evidence of infection happening through schools, and the costs of school closures are immense for the learning and well-being in general of these children. We did not heed, I'm not talking about the US in general, I'm talking in general. We did not heed this kind of advice. Policymakers closed schools, and in, in some sense, what I feel is they kind of got locked into place in a particular way. They kind of got locked themselves into, we're not reopening schools until there's no evidence that anybody's ever gonna get infected under any circumstances, and once they, once you sort of backed yourself into that. Corner, it's hard to get out of that corner. So that's, I think, one lesson. Um, and I would say also that in the case of developing countries, this is a time when public transportation was already open. Everybody, I mean, all of you who have been in public transportation in developing countries know what that's like. So people, you know, cheek to jowl with each other at a time when schools were still closed, restaurants were open, workplaces were open, everything was open except school. School was the one thing that that wasn't open. So the first thing I would say is, Close schools only if you really absolutely have to, and be aware of the costs of closing schools, because they are very large, and as I say, they are lingering. So that's 0.1. 0.2 I'm just going to flag it, and perhaps there's a follow-up question, which is, you got to get ready for this stuff now. The country, some of it is just the countries that experienced pandemics before did better. Absolutely, I agree with Mushfik on this. We have clear evidence of that. But we also have evidence that countries that are prepared for some kind of systemic shock like that beforehand, and I can go into details of this, did a lot better than those that it had not. And so in some sense, what you want to do is you want to avoid the sort of cycle of what, what uh people in health in particular talk about of, what is it, uh panic and neglect. At the time, you panic, and then, OK, it's over, you don't do anything, and then the next time, you panic again, as opposed to preparing. So my advice is prepare now. And I know I've gone on for a little bit too long, but But that's, that's, that's sort of my summary of, of what we have to say about this. Thank you so much, Norbert, and I, I hope that we'll, uh, in a few minutes, I hope we can circle back on a couple of those things. Um, so Mamta, we've seen, you know, Norbert's talked a lot about the education effects. We saw Jowana talking about employment shocks, uh, you know, from the shocks, uh, Mushfaq, a variety of lessons. So across all of these, I feel like there's a real theme of kind of, of resilience and. What countries need, you know, what sort of, what lessons countries can take away from sort of these studies and from the experience as you worked with countries over the course of the pandemic. So protecting human capital a little bit more broadly than Norbert's talking about, what would you say are kind of the big one or two takeaways? Thank you. Um, thank you, Dave, and thank you for inviting me to this panel. Um, That's a big question. I'm gonna give you a very short answer. Um, uh, first of all, I think protecting income. Which is a big constraint, uh, for people to access services, whether they're healthcare or whether it's healthcare or whether it's food, uh, um, is quite important. And, um, while, uh, low and middle income countries were quite constrained in terms of fiscal space, the ones that were able to very quick either had a social registry or were able to quickly set it up, Mushfiq gave this great example of Togo, um, were able to provide some sort of floor under people's incomes, and that was hugely important. Many kids get their, get food from, uh, early childhood centers, and if your kid is not going to an early childhood center, you need to be able to feed them. And that means having an income at a point in time when you may, I mean, the economic activity has slowed down. So, I would say that, um, uh, a very important thing is having a social registry, uh, and a means of making payments, how, however small, in order to provide a floor under people's income in the context of a supply and demand shock, which is what a, what a pandemic, uh, is. Um, the second thing I would say is that, um, uh, being able, being able to prepare, being able to detect that there's an outbreak and put it out quickly, uh, outbreak of a virus is, is very important, and countries that had been, had learned from Ebola were able to do it. I'm hoping that there's some muscle memory now in many countries and they're better able to do it. It actually costs not that much to have a reasonable surveillance system in place. Now countries are different and the geographies are different and the size is different and all of that, but the basics of a surveillance system is really having community level people who are reasonably well trained. And can spot, oh, there's something going on here. I see this outbreak of infections and there's a cluster, right? And being able to report that up and being able to respond quickly to that, I think is very important. And, and, and countries that are, were able to do that, were, were better placed and, and ready. And That being able to contain that outbreak is, is, you know, is a huge benefit. It, it means your economic activity doesn't slow down and, and all of that. I would say those are two really important things, having a social registry and being able to put a floor under people's income, and being able to detect an outbreak very quickly and stamp it out. Now. We also know from this that it's sometimes very hard, right? We don't know what the next, next virus is gonna be like. Uh, I don't know if you've watched, uh, the Planet of the Apes movies. I'm a great fan of them. Well, in that, the virus that, that breaks out from a lab, actually the survival rate is 1 in 500. It's, it's in a very incredibly infectious virus. So if it's a virus like that, then, you know, we're really, we're really sunk. And then being able to, Access, um, you know, masks, uh, whatever it is that you need, hand sanitizer, eventually vaccines, maybe, uh, quickly and efficiently is very important, and that is a huge governance challenge, and I, I just want to put that out there. Uh, it's a huge governance challenge because, uh, not every place manufactures these things, uh, not everybody's able to get access to it. One of the things we learned this time around is that countries were able to leverage their diplomatic connections. And get access to things. They were able to get access to masks, they would get access to oxygen, they were able to get access to vaccines based on their diplomatic connections. So if I was a national leader, I would make sure that I am actually maintaining good relations with a variety of actors who can provide me with some of the things that I might need for my country in the. In the event of a, of a, of, of a future pandemic. That's not in the realm of economics, that's in the realm of, of politics and, and leadership, but it's pretty important to, to survival. We can talk about, um, uh, international governance of, of, of, of, uh, of, you know, how do, how do we share things that are needed to save populations from a pandemic, but maybe in the next round. No, thank you so much, and I think, you know, your point on protecting incomes, your second point on surveillance, I think also speaks to Mushfiq's point that countries that had previous experience managing a pandemic or an epidemic of some sort, um, we saw significant, significantly better management and so one can only hope that there will be some, some learning from this time around broadly. So, uh, another thing in one of the earlier presentations that we saw was, uh, the dramatic inequality in vaccine coverage, well into, uh, the rollout of vaccines. So, uh, Michael, obviously you've worked on vaccines a little bit. Um, that was a joke. Michael's worked on vaccines a lot, so if you, you know, how would you say, how do we think about closing this massive equity gap that we saw in the figure that that Musvik showed er next time around? Right So I think there are several, uh, several steps. One is just increasing more capacity at the global level. You know, that's something that I think might initially think, well, that's just an efficiency argument and that's nothing to do with equity. Well, first, let me just reinforce how important that is from an efficiency point of view and then let me tell you why I think that's also important, uh, uh, from an equity point of view. You know, we just heard the educational losses alone were $21 trillion. OK. The chance of uh, my, I might be slightly out of date of this, but the chance of a pandemic are about 2% per year. This is all worked out uh much more thoroughly in a paper by uh Rachel Glenister and Chris Snyder. Yeah, if you, if you take a 2% chance, that means the expected cost each year, um, call that $420 billion. I think they get $700 billion because they're including things beyond just education losses. OK. What's the present value of that? You know, uh, you know, that's, that's, uh, that's, you're back up to $4.2 trillion again. So how much, you know, how much should we be spending? If, if having some extra vaccine capacity available, not just vaccine capacity, obviously mask capacity, we could, there, there are masks that that work uh that are reusable. Hospitals don't use them because it's a little bit easier to use, uh, use other masks. Well, you could get these reusable masks and require hospitals to have those, so if there's an emergency, they would be needed. So every category. of of of relevant equipment, you know, the cost of maintaining a stockpile of that is really trivial. Now, obviously there are certain adjuvants that are used in multiple vaccines. We can do things like that. We obviously don't know exactly what's going to hit us, but there's a lot of preparation that could be done now. You could put vaccine capacity in place so you could switch it over once a vaccine was distributed, was developed. It only took. Uh, the vaccines were developed very, very rapidly after the, after the, uh, after the, the, um, after the pandemic hit. This was a matter of, I forget the exact time. I've been trying to forget like everybody else, but you know, a month or something like that. The, um, the, what took time was testing them and then building the manufacturing capacity. You know, the value of moving a vaccine ahead by a few days. Greatly exceeds when you're talking about numbers like $21 trillion. Any of these costs, you know, we estimated that Operation Warp Speed in the US, if that had advanced vaccine vaccine availability in the US by 12 hours, that would have paid by itself. OK, so, um, so first putting in a bunch of capacity that makes sense for individual nations on their own, it makes sense globally. And why is this, uh, I'll come, you know, um, why is this an equity investment? Well, if we look at this at the global level and we take a fixed, say, that high income countries are going to get access first, I don't think we should take that as fixed, but let's say we take that as fixed. If it's a 2 year lag till everybody gets vaccinated and you cut that in half, well, you know, if you're 2, if you're, you know, a month into the queue, that takes 2 weeks off your time. If you're 2 years into the queue, that takes 1 year off your time till you get vaccinated, so just increasing global supply, even without correcting any other equity problem is uh is gonna help. Second, Oh, what it was with uh, uh, I've done work on uh advanced market commitments for, uh, for pneumococcus. Together I was asked by some, some, some governments to think about this problem. We didn't actually recommend an advanced market commitment. We recommended something very much like like Operation Warp Speed. The US did that, um, not, I don't want to claim because of our analysis. The UK did something very similar. The European Union was worried about, you know, saving a few pennies on the vaccine and didn't do it. The worst case of, uh, of, you know, uh, uh, of worrying about, um, you know, being penny wise and pound foolish. Let me talk about a middle income country with, you know, I won't name the country. Very capable policymakers, they, we made the same analysis, uh, we presented the same analysis to them. They were convinced, they wanted to get their order in early. But they were afraid, so the Operation Warp Speed and what the UK did was they paid to build the, to get, they got their orders in early and they said build the factory. We don't know whether this is going to work, but you know, fine, we'll take the chance. If it works, it saves us billions of dollars. If it, if it doesn't work, we've spent a few million dollars or you know, maybe. Tens or, you know, it's, it's a very easy deal. Um, the, these, in this middle-income country, they saw the logic, they wanted to do it, but they said to us, hey, if the vaccine doesn't succeed, I'm gonna go to jail. They're worried about being prosecuted for corruption. So you know that's something and you know Tristan's research suggests that a big part of the reason, the majority of the reason, you know, there was obviously there were some cases of countries placing embargoes and so on on on medical equipment, but overwhelmingly countries, the main reason countries didn't get their, their uh lower income and middle income countries didn't get their vaccines in as quickly is because they didn't get their orders in as quickly. So allowing low and middle income countries that same flexibility, and this is a matter of financing mechanisms which the bank could help with, to say, hey, if you want to make an investment so you can get the vaccine early, we're going to support you to do that, that could be done for the next pandemic. You know, there's a bunch of other, other things that we get done like human challenge trials, coming up with procedures for those to greatly accelerate the, the vaccine development and testing process. But uh, but I think there's a lot, a lot that can be done to get vaccines out quickly and virtually any scale of investment, um, you know, money that's a lot by vaccine standards is very low by global GMP standards or uh or the value of education standards. Fabulous. Um, no, thank you so much. It's hard to beat that return. So I know that we're um I know that we're over the scheduled time. I wanna give our, our panelists, um, I wanna ask one last question and er give them a chance for a last word, so. Yes. And Herbert, why don't you go ahead and then I'll plant mine and they can, they can answer them together. Um So, I was struck by Norbert's, uh, two lessons in terms of what, uh, how one should prepare, right? You, you said that the, uh, worst damage was schools closing, and the second one was Uh, that one has to prepare for the next one. So I was looking at essential workers by industry in 2019 in the United States, and there were 55 million essential workers, you know, uh, 30% were in healthcare and so on, and 20% were in food and agriculture. Teachers are not on this list. So I think one of the, perhaps the best way to prepare for the next pandemic is to declare school teachers as essential workers. Yeah. Alright, no, that's a great point. So the, the last question that I have, and feel free to uh Norbert to respond to, uh, Inderrit's point, chief economist privilege, um. So, you know, the question somebody asked after one of the er initial presentations was, have we learned the lessons, right? And I think this is the question, you know, there will be another pandemic, we don't know exactly when. So, do you think we're in the process of getting ready? And if there were kind of one thing that you wanted to leave this panel with on sort of one thing that needs to happen to get there, um, what is it? And er let's start with Mamta actually for this one. Um, I think we are in the process. Well, I, I'm a, I, I'm an optimist, so I have to say that we're in the process of getting ready. Um, uh, but I want to link it to the argue the discussion that we started at, uh, at the start of today, the great incoherence, right? Um, I think we are incoherent though, in getting ready. So, we know from a paper that we wrote with the WHO that the cost of getting prevention, And preparedness investments by which is meant this surveillance diagnostic, being able to kill an infection within the first few days of its appearance. The cost of that is relatively low. It's about $15 billion US dollars a year for developing countries. $15 billion. That's nothing. But we're in a world where we have created a financing facility to help with this, and it's not raised anywhere near $15 billion a year. Sorry, it's $15 billion a year for 5 years. So it's $75 billion in total. That facility has raised $2 billion in total of the $75 that it needs over five years. So there's a great incoherence. I mean, Michael just talked about it. We've walked through the valley of death. OK, Millions of people have died, trillions. Dollars have been wiped off global GDP, and the long-term consequences on, on children are enormous. So if we were rational, we would find the, the $75 billion right, over the next 5 years, but we're nowhere near doing that. So the, the single most important thing is to be able to detect and prevent an outbreak. Forget. And then we come to the medical countermeasures. I won't say forget them. I think we're going to need them, right? Um, but we're nowhere near finding the resources to do that. So I would say this is the single most important thing that we need to be able to find the resources for and fund. And we really have no excuses. I mean, I, I, I'd hate to say to my grandchildren, oh, I was around, or my children, maybe not my, uh, uh, grandchildren. I'm gonna be optimistic. Um, I, I was around when COVID-19 happened and we had all this talk and we knew how much was needed, and we created a facility and, and actually, we didn't actually do much with it. I, I, I would hate to be in a position where, where we said that. Um. I, I want to come back to this vaccine point. I mean, we don't know whether the next pandemic is, is going to be vaccine amenable, um, but I do think thinking about sharing the medical countermeasures is a very important issue. And, and, um, I, I think we're not there as a world. We have, we are negotiating a pandemic treaty. It's stuck. That treaty is about sharing medical countermeasures, and, and there's no agreement on that. Um, uh, I think, uh, uh, it's, it's not just about putting in money, putting in your orders. Trade restrictions immediately come into play. And that's why the pandemic treaty is so important. We've got to agree that if we want to survive, just imagine a Planet of the Apes kind of virus, right, which is where only 1 in 500 is going to survive. Don't we want to be in a situation where we're sharing medical countermeasures? I know it, it, this is not a kumbaya thing. This is really about, don't we want to survive as humanity? And so I think we need to be able to agree that we should be able to share at, in a, in a, in a reasonable way, whatever medical countermeasures needed. I believe that the key to that is having deconcentrated manufacturing, and it's an equity issue and not an efficiency issue, and that's why we're trying, like many other organizations to support to support manufacturing and of these medical countermeasures that could potentially be used. We're trying to support manufacture and underserved geographies, because one thing we know is that the US was very good, just to name names at providing. vaccines to its allies. So was China, very good at doing that. So was Russia, very good at doing that. And so you, you want to be in a position where you have access from your neighbor or from a country that values your existence, right? So, so I think there's something about this deconcentration which is going to be very important to the, to the effort of fight, fighting a pandemic. So we're getting there, is my, my bottom line, but we're going about. It rather incoherently, and I, I would urge us to go about it more coherently. I think we can play a role as, as an institution like the World Bank. We can play a role, um, but I think we also have to play a role as citizens. I mean, after all, we have a huge representation of countries here. I think we also play a role as citizens in terms of advocating for the right things to be done, and, and we should all play that role in our, in our personal capacities as well. Thank you. Norbert, are we on, are we moving in the right direction? And one thing. So I, I think in addition to the don't close schools unless you have to, and by the way, Indermin, I completely agree with you, and some countries did give priority, some developing countries did give priority in vaccine access to teachers before other groups, and, and most did not. So I'm, I, I completely agree with that. Yes. Good point. Yes, good point. Good point. Absolutely, I agree with you. So I think it goes back to what I sort of hinted at before, which is You have to get ready now. You can't be making this stuff up when whatever shock hits. It's not going to work. If you're not prepared now, you're not going to get prepared in the middle of a crisis where, even if it's not a Planet of the Apes scenario, even if it's just a sort of a run of the mill pandemic, you're not going to be able to get this ready while it is happening. And so, what does that actually mean? I think what it means is Figuring out two things. One, if I have to close schools, how am I going to ensure that there is some way of getting these kids to learn something? If some of that is distance learning, you need to figure it out now. If some of that is radio-based learning, you need to figure it out now. And you need to figure it out and pilot it and evaluate and redo it. And then I think you need to mainstream it. It can't be just like, OK, so now we developed this, and this is in case a pandemic hits, because then you're You're going to get around to it. You have to say, OK, we're going to develop a system of remote learning. And we're going to, some part of the curriculum, we're going to do through remote learning. Because then it becomes sort of part of the fabric of the education system. And then you can actually scale it up, which is what the Uruguayans did. And then you can also scale it down. But it's sort of like, yeah, we developed it, we tested it. This kind of seems to work. And, oh, OK, now we put it on the shelf, and now we wait for the pandemic hit, nothing is gonna, nothing is going to change. And the same is with, OK, now try to figure out, what do you do on the day, again, this is sort of on the on the school closure. What do you do on the day that the kids are back? How do you get these kids to catch up? Again, if you're trying to figure that out on the day the school's reopened, it's not gonna work. You're gonna do some sort of ad hoc, crazy thing, and you're gonna draw a little bit from here and a little bit from there. Figure it out now. Figure out how do you get, how do you make up learning losses? And we actually know how to do that. We know how to make up learning losses. So technically speaking, we know how to make up learning losses. Look at what Pratam does in India. There's a great program in Manizales in Colombia where they do exactly that, that shows big impacts, big ability to recover learning losses. But again, you need to figure this out now, and in some sense you need to implement it now. Sort of in the, in equilibrium, so that when you really need more of this, you already, it's sort of, it's already kind of, I don't want to say second nature, but you've already figured it out how to do it. People know what you mean. The program is already there. You just massively scale it up now. So the big lesson that I draw is, in that sense, figure the stuff out now, implement it now, make sure the solutions work now, and make it part of the sort of standard package of how you deliver education services, because Otherwise, you're not gonna be able to respond. So that's my two cents' worth. Oh, thank you so much. Before I go to Michael, I just wanna announce to everybody that at midnight tonight we'll be doing a showing of Planet of the Apes, which apparently is a huge theme here. So, back in this, back in this room, James Franco one, which is the one that had the uh the virus. All right, Michael, are we on track? And uh one thing. OK, uh, yeah, we're not, thank you, uh. Let's see if I've got this right. Uh, we're not, we're not at all on track. Um, I think, uh, I, I agree with Mamta about that. I agree with, uh, with Norbert. You know, it's the theme of this conference. We've said we're gonna do a bunch of things and we're not making the resources available. Um. So what can we do, you know, I think the resources should be made available. What can we do without the resources? I think there's some institutional preparation that's very low cost and that we should put in place. So I mentioned the middle income country policymaker who felt that they would risk jail if they if they signed something. Well, we can set up procedures now. To legitimize that so that nobody would have to feel they would go to jail and the World Bank could say if international institutions said we're going to support these these purchases, then nobody would need to go to jail for that, particularly if they say these are a list of reasonable candidates, OK, but you know, more broadly. Humans challenge trials. This is where healthy volunteers, in some cases, there are plenty of people in high income countries where there's good medical care, young people who are willing to say, yes, I'll be infected with COVID and you can try out a vaccine on me. Now, you know, complex ethical questions, when are these allowed, when are they not allowed? But we should be setting up the panels now, so if there are some groups that are at lower risk, if this next disease is not fatal, and if people volunteer, then you can do the trials much more quickly. We can set up the procedures for that now. Um, um, you know, other examples, so Mushfiq's presentation was amazing. And it was also very sad, uh, like the other presentations. So, let me focus in on the mask trial, OK? Now, think about this, we're in the, we've, we're in a pandemic. We don't even know whether masks work or not and who works on it. All respect to Mushfiq, but why is an economist having to do this, you know, I don't know how many years into the pandemic did he wind up doing to get this, getting this trial going. Yeah, so why, you know, why was this not ready to go immediately after the, once the pandemic started? It should have been ready to go, it should have been set up, the procedure should have been approved, the money should have. Set up so it could be allocated as soon as, as soon as this was available. Um, why, um, you know, the same thing on school closures, you know, maybe, maybe you could do RCTs on this, maybe you couldn't, you can think about that in advance, but surely there was a lot of non-experimental variation that could have been utilized uh very well with the right data collection systems. We should have thought about that because as you point out, next time. Maybe we will need to close schools, um, and there's, and, you know, there, there, um, uh, you know, something like, um, uh, dose optimism, you know, filters in schools. So I'm involved in a project early stage results, we're seeing filters, this is post-COVID, but. Seem to be improving learning results because of reduced pollution. Now, but we should be testing those types of things, you know, we don't have a good, we didn't have a good study ready to go at the time, so all these non-pharmaceutical interventions at the beginning and then finally. Dose optimization. So for some vaccines, so yellow fever, there's a shortage of yellow fever, I think it was Brazil, um, did 1/5 doses, and that worked, and the WHO endorsed it. Well, We don't know. We didn't know what the right dosage was for COVID vaccines, but, you know, based on some, so I worked with, started working with biostatisticians on this and people with expertise, may well, based on the antibody response, which is now what's used for approving things, one quarter doses of some of the vaccines would have worked. Imagine if we had 4 times the supply, Well, we could have treated the world much, much more quickly. That research wasn't done. The pharma companies don't particularly want to do that and invest in that, but you know, we should have systems set up for publicly funding dose optimization from the beginning. Um, thanks. Thanks so much. OK, so just to er to sum up, the one takeaway um from MoMA, er detection systems, uh, from Norbert, develop and mainstream, institutionalize the procedures now, the technologies now, um, and from Michael, do this institutional procedural preparation now, so that things are ready to move. So. That's what we have to do, uh, please give a round of applause to our panelists and to our presenters. Thank you very much. I know you'll be able to find uh these folks later in the day, and we'll pass the mic back to Alan for a final word. Thanks so much. In the spirit of the Euros and the Copa America, let me remind you that this is halftime. There's another day tomorrow of, uh, of the conference that will be at the Center for Global Development. Uh, today we tackle the 1st 4 areas of coherence, and tomorrow we'll tackle the, the, the next 3. The great incurrence for me is that we, we struggle to stay on time, but the conference was brilliant. So thank you to everyone. A special shout out to Kathleen who um helped us this morning and this afternoon, and then also a special shout out to Kelly who was running around and getting his steps in, chasing people with a mic. Thanks so much. We'll see you guys tomorrow.

showAllTranscripts

no

duration

PT1H21M56S

scene7File

worldbank/ABCDE2024-Session-4

scene7Domain

scene7FileAvs

worldbank/ABCDE2024-Session-4-AVS

title

ABCDE2024 Session 4

description

ABCDE2024 Session 4

showTimestampAndTranscript

yes