col-xs-12
col-sm-12
col-md-12
col-lg-12
col-xs-12
col-sm-12
col-md-12
col-lg-12
videoType
dynamic-media
videoDmUrl
https://delivery-p136806-e1377785.adobeaemcloud.com/adobe/assets/urn:aaid:aem:ef1265c8-de70-44ae-ae11-7d5112f7bcf5/play?assetname=ABCDE23_Session_1.mp4
keyFrameImage
timestamp

00:02 And after this very inspiring presentation by Intermit,

00:06 uh,

00:06 we will start the second

00:08 session part of the session.

00:11 Um,

00:11 my name is Manuella Francisco,

00:13 and I'm the global director of the Global Practice

00:17 for macro Trade and Investment.

00:19 I have with me a very distinguished panel of academics.

00:23 Let me start by presenting them,

00:25 uh,

00:26 from my left to the right.

00:28 Um,

00:28 we have Professor Ufuk

00:31 Ax sorry.

00:32 Professor Jung Sin Shin,

00:35 a professor

00:36 at the Washington University of Saint Louis,

00:40 we have Professor Ufuk

00:42 Akish,

00:43 apologies for my bad pronunciation,

00:46 Professor of Economics at University of Chicago.

00:50 We have Professor Michael Grubb,

00:53 professor of Energy and Climate change at UCL University College of London.

00:58 And we have Professor

01:00 right

01:00 next to me,

01:01 Professor John Friedman,

01:03 Chair of Economics department at Brown University.

01:07 The four papers you hear about today

01:10 present some of the theoretical and policy underpinnings

01:14 underpinnings that will help us better understand

01:17 the growth challenge in middle income economies.

01:20 We,

01:21 we will hear a presentation that is structured around

01:24 the 3 themes of the upcoming world development report

01:28 that

01:29 is structured around enterprise,

01:31 social mobility,

01:32 and energy transition.

01:36 We will start with Professor Akishit that

01:39 will focus on the role of creative destruction

01:43 by

01:44 generating economic growth through innovation and entrepreneurship,

01:48 which will form the general framework of the Schumpeterian growth of the

01:54 of the World Bank development report.

01:56 Over to you,

01:57 Professor Akus.

02:03 So good morning everyone.

02:05 Uh thanks to the organizers for this fantastic conference.

02:10 Uh,

02:11 it's always difficult to talk after Inarmid,

02:13 who is a great presenter,

02:14 and,

02:15 uh,

02:15 but the nice thing is that he also made a

02:18 nice transition to what I'm going to talk about.

02:21 So,

02:21 uh,

02:23 today I will talk about economic growth,

02:24 but from an angle of creative destruction.

02:28 The literature on creative destruction has made

02:31 enormous advances over the past 30 years,

02:34 I would say,

02:36 and

02:37 this,

02:37 you know,

02:37 once you start thinking about economic growth

02:39 through the lens of creative destruction,

02:41 you suddenly start uncovering many,

02:43 many facts because now you're able to

02:46 map the economic growth idea to micro-level data and.

02:51 The microdata analysis is guided by the Schumpeterian theory,

02:56 as I'm going to illustrate.

02:57 It's a,

02:57 it's a very nice and very informative angle that the literature has developed,

03:03 and this is also the main thrust of the of the next WDR.

03:07 And so led by Sean Mikal and Indermit,

03:11 we have a very large team working on the WDR.

03:14 A lot of them are here,

03:16 and I would like to give you a little bit of an idea about what we are working on.

03:20 Uh,

03:21 at this new WDR,

03:23 so first of all,

03:24 we are taking a comprehensive approach.

03:26 So we,

03:27 we try to question everything in an economy

03:30 and we try to say

03:32 how can we save the middle income countries

03:34 that are trapped in the middle income trap?

03:36 How can we save them out of that trap through technology and as Indermet also showed,

03:41 as you go

03:43 through the stages of development,

03:45 as you get closer to the frontier.

03:48 Rather than capital accumulation,

03:50 what starts to matter is technology development and productivity.

03:54 And of course when it comes to productivity and technology,

03:57 the question is how can we introduce new technologies in an economy.

04:02 Just focusing on firms would be a mistake because

04:05 of course firms are some abstract entities,

04:07 but in those entities there are individuals working and this.

04:11 The whole approach,

04:12 the comprehensive approach is quite crucial.

04:15 So the whole reasoning starts with a

04:18 pool of individuals in any society,

04:20 which

04:21 we are illustrating here at the bottom right.

04:24 I think the pointer is not working,

04:25 but so at the bottom right,

04:27 and of course individuals in any society are ranked according to their talents

04:33 at the time of the birth.

04:34 Of course children are born with heterogeneous talents.

04:38 And societies with their own systems

04:41 somehow let some of those kids rise up in the society.

04:44 It could be based on talent

04:47 or it could be based on

04:48 some

04:49 ethnic background or family characteristics.

04:52 It depends on the

04:53 system that exists in the society.

04:56 So then,

04:56 of course,

04:57 in an ideal scenario.

04:59 We would have the most talented kids sent

05:03 to schools

05:04 when it comes to innovation-led growth or technology-led growth here.

05:09 The education starts to become more technical

05:12 and so then whom are we educating in

05:16 our technical schools becomes an important question.

05:18 Who are being sent in the society into those technical schools?

05:22 So then

05:23 education

05:24 becomes a different notion rather than looking at the average years of schooling,

05:29 now we are trying to look at

05:30 these elites

05:32 selected in the society that are sent into those technical schools.

05:36 Who are they?

05:37 Are they going there based on their ability or based on some other characteristics?

05:42 So then once we send the kids into schools and train them,

05:45 some of them stay in the country and some of them leave.

05:48 So suddenly brain drain becomes also part of the equation.

05:51 How do we treat brain drain?

05:53 Should we turn our back

05:54 and just complain and whine about the brain drain,

05:57 or

05:58 should we somehow embrace it?

05:59 Indeed today Marta is going to talk about a

06:04 fantastic paper of hers in more detail about this.

06:07 But

06:07 brain drain is also part of the equation.

06:11 So once we have people trained in the society,

06:15 some sent abroad and some coming back,

06:17 now we have the workforce in the society and

06:20 firms are hiring those trained workforce

06:23 and start competing among themselves.

06:25 So what do they invest in?

06:26 Are they investing?

06:27 In new technologies or if so through

06:30 what incentives or through what market mechanisms

06:34 and then the firms will create some value added

06:37 and as a result we will have this cycle.

06:39 Now

06:40 what is the observation about middle income countries?

06:42 The observation is that

06:44 through this pipe

06:45 the water is coming very,

06:46 very weak at the end.

06:48 So clearly something is blocked in the system,

06:51 and the question is where.

06:52 It can be anywhere,

06:54 right?

06:54 That's why rather than just focusing on one side of the economy,

06:58 we are just taking this comprehensive approach

07:00 and try to

07:02 identify the problems and potential solutions and policy recommendations

07:06 through this approach.

07:08 So why the Schumpeterian

07:11 model?

07:12 Why the Schumpeterian approach?

07:13 Of course when it comes to economic growth,

07:15 there are many,

07:16 many theories.

07:17 Indeed we are not denying any

07:19 other theories.

07:20 Of course the institutions are important.

07:22 Capital accumulation through the solar model is important,

07:25 but now we are also trying to bring in the technology

07:28 into the equation.

07:30 When it comes to technology and innovation,

07:32 there are competing theories.

07:34 One of them,

07:34 of course,

07:34 is the Romer model,

07:36 which is the Nobel winning model by Paul Romer.

07:40 But when you think about the Roman model,

07:41 of course it was a brilliant model,

07:43 but the way that model operated is that

07:45 there is this continuum of varieties,

07:47 and through each innovation we are adding new varieties to the economy,

07:51 and

07:51 we just add more and more varieties,

07:54 and whenever somebody introduces a new variety,

07:56 that person becomes

07:57 the

07:58 Permanent monopolies in that market and produces forever.

08:02 The same person keeps producing

08:04 forever

08:05 and the Roman model.

08:06 Of course it's a brilliant model endogenis technological change.

08:11 It doesn't talk about exit and turnover,

08:13 which is exactly the problem,

08:15 as I'm going to illustrate in a second,

08:18 especially in middle income countries,

08:19 because once you start thinking about turnover

08:22 and exit,

08:23 there you suddenly realize there are many,

08:25 many frictions that are embedded into the economy and the data

08:29 and that you can only read through the Schumpeterian theory.

08:32 So what does the Schumpeterian theory do?

08:34 So there is a firm.

08:36 Rather than introducing a variety,

08:38 it introduces a better version of an existing

08:42 technology.

08:43 So

08:43 for instance,

08:44 firm 1 is the

08:45 current producer,

08:46 but then Firm 2 comes in with a better technology.

08:49 Firm 1 is out and Firm 2 is in.

08:51 And this way through this turnover in the economy.

08:54 Firms are just replacing each other and

08:56 they are introducing new versions of technologies.

09:00 Of course this was the first version of the Schumpeterian model,

09:04 but then

09:05 Since 1992,

09:07 after Agnonhovi,

09:08 the literature has evolved immensely.

09:10 Now creative destruction is not only

09:13 taking place in our existing theories between entrants and incumbents,

09:16 but also among incumbents.

09:19 Indeed,

09:19 when we look at the data,

09:21 50 to 75% of the productivity growth

09:24 is coming from successful factor reallocation among incumbents,

09:29 and we cannot

09:31 avoid this,

09:32 obviously,

09:32 and that's why the type of framework that we are.

09:35 Considering is way

09:37 beyond Ajonhoitz initial framework,

09:40 I don't have time unfortunately,

09:41 so that's why I'm not able to go into those details.

09:43 But basically

09:44 here we have

09:46 a creative destruction model where not only

09:48 entrants and incumbents compete among themselves,

09:51 but also incumbents also compete

09:54 among themselves.

09:55 So once you start with this logic,

09:57 suddenly you realize that we might be,

10:00 we might be living in many different scenarios in different industries,

10:04 for instance.

10:05 Entrants through their new technologies are trying to push the frontier forward.

10:10 What can happen is that of course incumbents will somehow join join the right here.

10:16 So either incumbents can also react positively

10:19 to incumbent pressure.

10:21 In this case there will be the pro-competitive effect.

10:23 Indeed

10:24 there were some interesting papers

10:26 that argued that China's pressure

10:28 on the US market had some pro-competitive effect

10:32 or some paper.

10:34 Also argued that it had a negative effect and then some papers

10:38 argued that indeed in the continental Europe it was a positive,

10:40 etc.

10:41 So as you can see,

10:43 it's a very empirical question,

10:45 but we need to be flexible for these types of

10:48 different hypotheses so that we can read the data better.

10:52 Alternatively,

10:53 of course,

10:54 incumbents can

10:55 can resist the change,

10:57 and they are not always going to introduce a

10:59 better quality when an entrant comes in with a

11:02 better technology.

11:03 I might just say,

11:04 Well,

11:04 I lost the game.

11:05 Let me just leave the market.

11:06 I might try to rely on some

11:08 other strategies.

11:09 And indeed,

11:09 in many,

11:10 many middle income countries this is a major problem as we see in the data.

11:15 Incumbents can also collude.

11:17 It's not only one incumbent that the entrants might be fighting against.

11:20 It could be that the incumbents might be colluding among themselves.

11:23 And once you start reading or approaching the data

11:27 or the whole growth story through this logic,

11:30 suddenly it opens up the avenue into many interesting questions.

11:35 Of course we are aware of the fact that

11:38 innovation might be too early or too soon for some countries or for some sectors.

11:44 And

11:46 technology can be upgraded not only through innovation,

11:49 but sometimes countries can also utilize their advantage of backwardness,

11:53 and we are not the first ones to say this.

11:54 Indeed,

11:55 Gar Shankron in his

11:56 1960s book had already made this point very strongly,

12:00 forcefully.

12:02 If a country is far from the frontier,

12:05 Rather than forming these really expensive R&D labs and

12:09 try to innovate and compete with the frontier countries,

12:12 they might as well open up to the world

12:15 and get connected to the outside world and try to

12:19 bring those technologies that have been invented

12:22 anywhere around the world to the country first

12:24 and start learning from it and building from it.

12:27 Do we see an example of this?

12:28 Yes,

12:29 I guess

12:31 Jungs will

12:32 talk more about this,

12:33 but

12:34 The way we are thinking about the economic growth process is in different stages.

12:40 So on the x axis you can see the proximity to the frontier.

12:43 As you go to the right,

12:44 a country gets closer to the world technology frontier.

12:47 As we go to the left,

12:48 it gets far away from the frontier.

12:50 And if a country is very far from the frontier.

12:53 You know all they might need at that point is having some more capital and some more

13:00 resources to build more

13:02 pipes,

13:03 more roads,

13:04 or more bridges.

13:05 But

13:06 as we are getting closer to the frontier now,

13:08 rather than capital accumulation,

13:10 now technology has to be evolving,

13:12 and

13:13 again,

13:14 it's the basic intuition of the solar model.

13:16 The Solar model says

13:17 you can grow just so much with the catch up growth,

13:20 but once you reach your steady state from that point on,

13:23 you need to upgrade your technology.

13:24 How are you going to do it?

13:26 What we argue here is that

13:28 if a country is a middle income country,

13:30 there's an opportunity to bring technologies from outside.

13:34 And how are you going to do it?

13:36 Well,

13:36 now the trade suddenly has a different meaning.

13:39 Rather than

13:40 being open means that rather than just trying to sell your goods,

13:43 it might mean

13:45 opportunities to learn from the world.

13:47 And also

13:49 through diaspora abroad,

13:50 a country can make use of the knowledge at the world technology frontier

13:55 and learn from it.

13:56 And of course as the country starts building on what it learned,

14:00 then

14:00 it can start innovating

14:02 on itself once it's ready to do the innovation.

14:06 Do we have an example?

14:07 Yes,

14:08 Korea's story,

14:08 of course Korea's story is

14:11 very,

14:12 very involved,

14:13 but one of the things that was shown by

14:16 Yong Honghim and his co-author in his job market paper this year,

14:19 they showed they digitized the Korean archives

14:22 and looked at the government contracts,

14:24 and what they showed is that

14:26 The Korean government in the beginning,

14:28 in the early 70s,

14:29 when they were making the huge leapfrog,

14:31 the government was very heavily subsidizing technology adoption from Japan.

14:36 So the country was of course was not fully ready to do its own innovation,

14:40 but there were a lot of technologies that were licensed from Japan,

14:44 and then they learned from those experiences and build on it.

14:47 So this is a very important observation,

14:48 I think.

14:49 Rather than

14:50 putting an unreasonable target to middle income countries,

14:53 we need to put reasonable targets first so that

14:55 they can use it as a stepping stone and then

14:58 jump.

14:59 So

15:00 I highlighted the Schumpeterian dynamics so much,

15:03 but

15:03 what do we see once we start thinking about the Schumpeterian dynamics?

15:07 What do we learn from the microdata

15:10 as something new?

15:11 The basic premise of the Schumpeterian theory says that

15:14 entrants will be more productive and replace the less productive incumbents.

15:19 Let's look at some

15:20 middle income countries that's totally randomly selected.

15:24 So

15:25 here

15:25 we are comparing

15:26 exiters and entrance productivity

15:30 in Turkey,

15:31 and

15:32 the blue line is showing you the productivity of the exiters and

15:36 the red line is showing you the productivity of the entrants.

15:40 So clearly the basic premise of the Schumpeterian theory

15:43 is not working in the case of the Turkish manufacturing sector.

15:47 So

15:47 this is something that we have to fix

15:50 as part of the story.

15:52 So rather than creating destruction,

15:54 maybe this is a destructive creation in that sense,

15:56 because it's lowering the productivity.

15:58 Indeed,

15:59 when you look at the total factor productivity evolution in Turkey since the 1970s,

16:03 there's

16:05 a steady decline.

16:06 So let's switch gears and let's look at another country that did well

16:11 for a while and for quite some time now it's doing quite poorly

16:15 and here in this study

16:18 what we did is we looked at the growth strategies of the market leaders.

16:24 As the firm is evolving and becoming the dominant market leader,

16:29 does it become more innovative,

16:30 or does it rely on other alternative strategies?

16:33 Number 1 here means market leader.

16:36 Number 20 means that's the 20th largest firm.

16:39 So as you go to the right,

16:40 the firm size gets smaller.

16:41 As you go to the left,

16:42 firm size increases.

16:44 What we see is very interesting as firms are becoming the dominant market leader.

16:49 Innovativeness goes down.

16:50 The red dots go down,

16:52 but the number of politicians that they hire

16:55 relative to the firm size increases,

16:57 so they are getting more tangled

16:59 to the political system

17:01 rather than trying to push the frontier.

17:02 Again,

17:03 it's the simple Schumpeterian logic.

17:06 Once I'm in the market,

17:08 I need to resist,

17:09 and that's exactly what we observe here.

17:12 When we look at the Indian case again,

17:14 we see a lack of creative destruction,

17:17 and here on the top

17:19 left figure you see the life cycle of firms in the US and India.

17:22 In the US firms are growing very rapidly.

17:24 There's an up or out type of

17:26 competitive dynamics

17:27 when it comes to India.

17:29 The average firm size conditional on survival is just flat.

17:33 It means that the small firms,

17:35 they are not only growing,

17:36 they are not getting kicked out either.

17:37 And as a result,

17:39 the bottom right figure here is showing.

17:41 The fraction of small firms in the economy

17:44 as the cohortages.

17:46 As the cohortages,

17:47 the dashed line shows that in the US,

17:49 small firms are eliminated very,

17:51 very rapidly.

17:52 In India,

17:53 small firms are not being eliminated,

17:55 and as a result,

17:56 the fraction of small firms remains.

17:58 So this means that

18:00 the fact that we are observing so many small firms

18:04 in the middle.

18:04 Countries

18:06 could be a symptom of

18:07 not only lack of financing for them,

18:10 but

18:11 it could be lack of competition by larger firms.

18:14 Maybe the problem is in the mid-sized or large sized firms,

18:17 and maybe they are not growing sufficiently

18:20 to drive out the small firms so that they can create high quality,

18:23 high paying jobs,

18:24 so that all these trying to be entrepreneurs.

18:27 Can get reallocated and earn decent wages in these growing firms,

18:32 and this is what we need to consider.

18:35 Indeed,

18:36 of course,

18:36 when it comes to middle income countries,

18:38 informality is a major concern

18:41 and here once you start thinking through the Schumpeterian lens,

18:45 informality is not a cause but a result,

18:48 and informality could be a result of

18:51 Lack of competition in the market.

18:52 And here,

18:53 for instance,

18:54 again when we look at some regulations and what we see

18:58 is that

18:59 regulations that exist in middle income countries

19:02 can even encourage more

19:04 informality.

19:05 But if we could normally push the firms to larger sizes,

19:10 we would automatically fight

19:12 against informality.

19:14 So I'm running out of time.

19:16 Just a few

19:18 issues on the human capital side.

19:20 The report

19:21 will also focus on human capital.

19:23 And here we look at the talent allocation.

19:25 Innovation and technological upgrade cannot happen

19:28 just by itself or firms' investment.

19:30 Firms should also have good

19:32 engineers and inventors.

19:34 Without having good players,

19:35 a team cannot win the championship.

19:38 With that logic,

19:39 we have to also have a strong workforce.

19:41 And of course if

19:42 in a society there is discrimination or financial

19:45 frictions and not having equal opportunities for people.

19:49 We might be leaving out important talent out,

19:51 so this is,

19:51 this is an important

19:54 consideration in the report.

19:55 And then we look at the data again,

19:57 it's not the average years of schooling that matters for technological progress.

20:02 It's technical training,

20:03 technical education,

20:06 even in the historical US,

20:07 where we should draw a lot of lessons from,

20:10 in the historical US we see that

20:12 college education was extremely important for innovativeness.

20:16 And

20:17 not only having college education,

20:19 also having liquid secondary market for technologies,

20:22 but today in many of the middle income countries

20:25 there's barely college graduates.

20:27 There's barely any regulation for how to protect

20:30 intellectual property.

20:31 So

20:32 we need to talk about these aspects.

20:34 And

20:35 in successful countries like Denmark today we see that PhD

20:39 is the strongest predictor of innovation.

20:42 It's that technical.

20:43 It's at that level,

20:45 basically.

20:46 And of course when it comes to human capital we

20:48 don't necessarily need to grow them within the country only.

20:51 We can also tap on the global talents.

20:54 China,

20:54 for instance,

20:55 80% of Chinese students go back to their country

20:59 after studying,

21:00 and can you imagine what Big spillover there is in the case of China.

21:04 The number of returnees to China exceeds 1 million a year.

21:09 And that way,

21:10 of course,

21:10 we are talking about a Chinese miracle,

21:12 but we have to also think about these spillovers,

21:15 global spillovers that are taking place

21:17 and what happens in middle income countries

21:20 right now.

21:20 Middle income countries,

21:22 when you look at for instance,

21:23 who is leaving a country and who is coming back,

21:26 it's typically the most productive

21:28 individuals or scientists who are leaving the country.

21:31 And the least productive

21:33 ones are coming back because of the bad policies,

21:36 but rather than trying to force

21:38 those

21:40 immigrants to come back,

21:42 one thing one can do is one can try to connect with them

21:45 and try to learn from them.

21:47 Indeed,

21:47 Martha Prato will talk in great detail about

21:49 this rather than turning our back to migrants,

21:54 to people abroad,

21:55 we can try to build a bridge from them and learn from them

21:58 to utilize the advantage of backwardness.

22:01 And finally,

22:01 there will be also the energy part in the report,

22:05 but just to give you an idea about how we are thinking about this,

22:09 you know,

22:10 we can produce dirty technology in

22:12 the Schumpeterian sense or non-green technology,

22:15 let me put it that way.

22:16 But there can be also green alternatives,

22:18 green technologies to use.

22:20 What's the comparison here?

22:22 The key critical thing here is that

22:25 today we are living in a world where there is a race between green technologies

22:29 and non-green technologies,

22:30 and green technologies by construction

22:33 are less advanced.

22:34 We need to reconsider the government policies because

22:37 some government.

22:38 Policies that might have well intentions might be slowing down this transition.

22:43 Indeed we do see some evidence on this.

22:46 For instance,

22:46 the subsidies that are being provided to fossil fuels

22:50 in many countries

22:53 are just going against the transition,

22:54 and at best it will just slow down the transition.

22:57 So that's why

22:58 thinking about this creative destruction and clearly.

23:00 Energy transition is the poster child of creative destruction.

23:03 We are trying to replace

23:05 an old

23:06 energy or technology with a green one.

23:08 That's very much in the realm of Schumpeterian theory,

23:11 and these are some of the elements that we are considering in the new.

23:16 Sorry for the limited time,

23:18 but hopefully it gave you a good idea about what we are working on.

23:22 Thank you.

23:24 Many thanks Professor Akhijit.

23:26 So we will now hear from Professor Friedman.

23:30 Uh,

23:30 his presentation focused on intergenerational mobility around the world.

23:35 His work focused on social mobility and the importance of selecting,

23:39 developing,

23:40 and enabling

23:41 talents,

23:42 talented individuals.

23:45 Over to you.

23:45 Well,

23:46 thank you very much for the uh invitation to speak here.

23:48 Um,

23:49 I should note at the start that I'm not a development economist.

23:51 Uh,

23:52 I do not specialize in economic growth and so,

23:55 uh,

23:55 I think everyone in this room is gonna have a better

23:58 sense of the exact types of policies,

24:01 the exact types of analysis like what Ufuk

24:03 was talking about that are really going to,

24:06 uh,

24:06 help,

24:07 uh,

24:07 move development forward across the world.

24:11 What I wanna talk about is intergenerational mobility.

24:14 Now

24:15 intergenerational mobility,

24:17 uh,

24:18 featured in that,

24:19 uh,

24:19 extremely nice schematic that Ufuk started with

24:22 as part of economic growth directly in terms of

24:25 figuring out how to get,

24:28 uh,

24:28 the most diverse broad set of talent

24:31 into the position of being innovators or high productivity employees

24:36 in a given country through education

24:39 and I think that's a very important part of what intergenerational mobility does.

24:43 I also think that intergenerational mobility is uh

24:46 a broader part of the institutions of a country

24:50 which if not maintained appropriately can lead to real problems

24:55 um and I think there are many countries across the world today where

24:58 the

24:59 lack of intergenerational mobility or

25:00 perceived lack of intergenerational mobility,

25:03 the sense that too many people are born into situations from

25:06 which they just don't have a shot at moving up in.

25:09 Uh,

25:09 the income ladder or,

25:10 or in the social ladder

25:13 where parents feel like their kids do not have the

25:15 opportunity to have a better life than they do,

25:17 that is not only an economic problem that becomes a

25:20 social problem that becomes a political problem that can really

25:23 uh

25:24 cut the legs out from under what might otherwise be

25:28 a high growth,

25:29 high productivity regime.

25:31 So let me start with just

25:33 a couple of facts from

25:35 uh some great data that the World Bank has produced,

25:38 uh,

25:39 right,

25:39 people always think of the American dream as,

25:41 you know,

25:42 the possibility that,

25:43 you know,

25:43 uh,

25:43 people,

25:44 especially immigrants could come to this country,

25:46 um,

25:47 in the early 20th century and kind of no matter

25:49 what your background there was the opportunity for success,

25:52 and I think.

25:53 You know,

25:53 most people know at this point that,

25:55 you know,

25:55 that is not something that's really borne out in the day today.

25:58 Uh,

25:59 here,

25:59 uh,

26:00 the particular statistic being calculated is the probability that a

26:03 child born into parents who are in the bottom half

26:07 of the income distribution move into

26:09 a position by the time that they themselves

26:11 reach the labor market that's in the top quartile

26:15 and so,

26:15 you know,

26:15 obviously each of those are defined,

26:17 um,

26:17 locally.

26:18 And what you see is that um you know the

26:21 US is very much in the middle of the distribution of

26:25 um

26:25 middle income countries.

26:27 There are some countries like Brazil and Pakistan

26:30 who have lower rates of upper mobility,

26:32 but there are also many countries here like Jordan and Thailand which have

26:36 uh much higher rates of upper mobility than the United States.

26:40 Now,

26:40 these statistics though are still,

26:43 I think really only glossing the surface of what we can learn

26:47 about how to increase intergenerational mobility.

26:50 And so what I wanna talk about today is,

26:53 uh,

26:53 what I think are a very exciting set of new tools

26:57 that have come out of work that,

26:59 um,

26:59 my co-authors and I and and many others have done,

27:02 um,

27:02 across the world.

27:04 Um,

27:04 these are techniques I think that,

27:06 uh,

27:06 mostly were pioneered in developed countries,

27:09 but what's been very exciting is to see a lot of new work applying them to many,

27:14 especially,

27:15 uh,

27:15 middle income,

27:15 but,

27:16 but even,

27:16 um,

27:17 countries at the,

27:18 um.

27:19 Very lowest levels of the development ladder.

27:21 Um,

27:22 and I think that it's extremely important,

27:25 right,

27:25 especially in this middle income range,

27:27 to think about,

27:28 uh,

27:29 alongside growth and kind of elimination of,

27:32 of abject poverty,

27:33 to think about maintaining or increasing or not losing ground on social mobility,

27:38 um,

27:39 as we go along.

27:39 And so,

27:40 um,

27:40 over the next 10 minutes I'm just gonna

27:42 kind of give you a little bit of a,

27:43 a tour.

27:44 I'm gonna hit kind of 3 highlights of techniques that have been,

27:47 um,

27:47 pioneered that,

27:48 that I,

27:48 I think might be helpful,

27:50 uh,

27:50 to this end.

27:52 So first,

27:52 let me just start with this map of the United States.

27:54 This is a heat map

27:56 where in each

27:58 city area,

27:59 essentially,

28:00 which are each of the colored polygons,

28:02 we're measuring the average incomes in adulthood for children who grew up to

28:08 families at the 25th percentile,

28:09 kind of the.

28:10 Average below average family,

28:12 uh,

28:12 and the heat map colors in blue children who

28:15 are doing very well in red children who are,

28:18 are not doing as well

28:19 and what you see across the United States

28:21 is just there's an enormous amount of variation.

28:24 Children born in Charlotte,

28:26 uh,

28:26 down in the southeast of this country.

28:28 Um,

28:28 are only making $26,000 on average as adults.

28:31 Children born in Dubuque,

28:33 um,

28:33 which is part of that very blue region in the

28:36 upper,

28:37 uh,

28:37 Midwest of this country,

28:38 are earning nearly twice that amount.

28:41 Now,

28:41 of course you could look at this map and say fine.

28:43 The United States is the size of a continent,

28:45 you know,

28:46 of course,

28:46 intergenerational mobility is going to vary at this scale.

28:49 What I think might be surprising is that even when you zero into

28:55 a particular city to make it really local,

28:58 you still see this range of variation in mobility.

29:02 So this is the same map that I showed you at the national level,

29:06 but now focusing only on the New York metropolitan area.

29:09 We've now colored things at the census tract level.

29:12 So these are essentially neighborhoods.

29:14 There are about 4000 households per

29:18 census tract.

29:19 The colors are essentially the same.

29:22 Right,

29:22 so that same variation that you saw between the

29:24 southeast of the United States and the Upper Midwest,

29:26 so those are areas that are,

29:28 you know,

29:28 still within the same country,

29:29 but,

29:29 you know,

29:29 thousands of miles apart.

29:31 Here you see that same variation between neighborhoods that are in many cases,

29:35 literally across the street from one another.

29:37 Um,

29:38 and in particular I'm,

29:39 I'm gonna come back to this example later on,

29:41 um,

29:42 you know,

29:42 you see down in,

29:43 um,

29:44 in Brooklyn,

29:45 right,

29:45 there's some areas where there's,

29:46 I'm gonna kind of point it out here,

29:48 the pointer is not working,

29:49 but kind of right in,

29:51 in this region here,

29:52 um,

29:53 there's these little blue islands in what

29:56 is otherwise a very low mobility neighborhood.

29:58 Uh,

29:59 the blue areas are,

30:00 uh,

30:00 what's called the Nehemiah houses,

30:02 and they were developed in the 1980s on literally a plot of vacant land.

30:06 Uh,

30:06 by a local organization as an attempt to create kind of a new society moving away from

30:12 some very low mobility areas,

30:14 uh,

30:14 that were called the Van Dyke houses.

30:16 So I think the first lesson that I want people to take away

30:19 is that mobility is going to vary not only at a national level

30:23 but in an incredibly local level.

30:25 And so kind of that's the level of granularity

30:27 that one has to take when thinking about this problem.

30:30 Now of course the examples that I'm showing you

30:32 here from my work are in the United States,

30:34 but

30:34 um over the past 10 years we've seen very

30:36 similar analyses of many countries around the world.

30:39 So here are 4 examples from developed countries in Europe,

30:44 um,

30:44 which show kind of very,

30:45 very similar patterns.

30:47 More relevant I think to this audience here are examples from.

30:51 Um,

30:51 Asher et al.

30:52 on the right and from Allesina et al.

30:54 on the left,

30:55 um,

30:55 of course one has to use data creatively in

30:58 order to generate these types of very local maps.

31:01 Uh,

31:01 on the right they're using education as a,

31:03 as a measure of socioeconomic status.

31:05 On the left they're using literacy,

31:07 but,

31:07 uh,

31:08 I think the big picture here is that not

31:10 only is intergenerational mobility varying at a national level,

31:14 it's varying

31:15 at uh a subnational and even very local level,

31:18 uh,

31:19 within these countries.

31:20 So second,

31:22 um,

31:22 let me just come back,

31:23 um,

31:23 to this map here.

31:25 Uh,

31:25 this looks at intergenerational mobility overall pooling all different types of,

31:31 uh,

31:31 individuals

31:32 and what analysis in,

31:34 in this country has shown,

31:35 this is analysis by my co-authors,

31:37 um,

31:37 right.

31:38 Chetty and Nathan Hendren

31:39 along with uh Maggie Jones and Sonia Porter,

31:42 is that this is sharply divided the

31:43 United States between individuals of different races

31:47 and so here I'm showing a similar map

31:50 with black men on the left and white men on the right.

31:55 Um,

31:55 and looking at these maps,

31:57 right,

31:57 it might look like we're using two different

31:59 color schemes for the two different maps.

32:02 Uh,

32:02 that's actually not the case.

32:03 It's the same color scheme on the bottom.

32:05 It's just that upward mobility for,

32:07 um,

32:07 black individuals,

32:09 black men born into poverty on the left

32:12 is,

32:12 is basically everywhere less than,

32:15 uh,

32:15 the,

32:16 the same outcomes for,

32:18 uh,

32:18 white men born into families at the same level of poverty on the right.

32:22 And so I,

32:23 I think it's not just that overall mobility,

32:25 it's thinking about

32:26 um how do opportunities

32:28 exist for children from these uh other dimensions in society,

32:32 um.

32:33 Similarly,

32:34 uh,

32:34 these,

32:35 uh,

32:35 types of findings have come out in work

32:37 in developing countries.

32:39 So this is,

32:40 um,

32:40 uh,

32:40 a scatter plot that I made based on results from that same Allesina et al.

32:45 um,

32:45 analysis where

32:47 they've calculated mobility separately

32:49 for,

32:50 uh,

32:51 Christian individuals and Muslim individuals in

32:53 many different countries in Africa.

32:55 And so we have uh the upward mobility in

32:58 each country for Christians on the horizontal axis,

33:01 for Muslims on the vertical axis,

33:04 um,

33:04 and what you see,

33:05 uh,

33:05 relative to the 45 degree line is that almost all countries are

33:10 below and to the right.

33:11 And so what that's saying is that to a varying degree,

33:13 uh,

33:14 you see more upward mobility for Christian individuals than Muslim individuals.

33:18 So again,

33:19 you know,

33:19 that's the type of

33:20 uh equality of opportunity that I think matters quite a bit,

33:24 not only from an economic perspective,

33:26 but from a broader perspective as well.

33:29 Third,

33:30 um,

33:31 we've seen quite a bit of evidence that these types of local variation in

33:36 upward mobility is not just about sorting

33:39 of particular people to particular places,

33:41 but rather it's a causal effect in large part of the.

33:44 Itself.

33:45 So how do we know that?

33:47 Um,

33:47 we've learned that from an exposure design.

33:51 And so the idea here is that one's,

33:53 uh,

33:54 the,

33:54 the impact of one's neighborhood

33:56 varies by how long one's lived in it,

33:58 right?

33:58 It's really an exposure-based model.

34:01 And so let me come back to those two

34:03 neighborhoods in,

34:04 in Brooklyn that I started with the Nehemiah houses,

34:06 which are a very high upper mobility neighborhood.

34:08 The Van Dyke houses,

34:09 which

34:10 are a very low mobility neighborhood,

34:12 and essentially just across the street

34:13 from the Nehemiah houses.

34:15 The dash line at the top is our outcomes for children who grew up

34:20 their entire lives in the Nehemiah houses,

34:22 which is much higher than the outcome,

34:25 uh,

34:25 the dashed line at the bottom for children who

34:27 spent their entire lives in the Van Dyke houses.

34:29 So the idea of this causal design is to separate out,

34:33 you know,

34:33 maybe those kids living in those two different,

34:35 um,

34:35 uh,

34:36 sets of houses are very different for other reasons.

34:39 And so instead of just comparing kids who grew up entirely in one versus another,

34:42 we're going to look at those who moved from one to another at

34:45 a particular point in childhood.

34:47 And so first,

34:48 let me think about children who were born in the Van

34:51 Dyke houses and then moved at a very early age,

34:54 let's say age 2,

34:55 to the Nehemiah houses.

34:56 Those children have outcomes that look very similar,

34:59 uh,

34:59 slightly below,

35:00 but very similar to those children who grew up,

35:03 their entire lives in the Nehemiah houses.

35:05 And then we're just gonna repeat this,

35:07 uh,

35:07 children who moved from the Van Dyke houses to the MEA houses at,

35:10 at age 3,

35:11 age 4,

35:11 and so on.

35:12 And what you see is that the more time you spend

35:15 in the Van Dyke houses before moving to the Nehemiah houses,

35:19 that is the less opportunity that you have to be exposed to that,

35:23 uh,

35:23 higher mobility environment,

35:24 whatever that is in the Nehemiah houses,

35:27 the worse your outcomes get by.

35:29 The time you've not moved until you're in your early twenties,

35:33 moving at that age to the my houses seems to really have no effect.

35:37 Um,

35:37 and then in particular,

35:38 one can then look after age 24.

35:40 We're measuring incomes at age 24,

35:43 so this is essentially a placebo check to see,

35:45 well,

35:45 maybe the people who move at these different ages are different,

35:48 and here you basically find no effect.

35:50 Um,

35:50 and now this,

35:51 this mover's design,

35:52 which has this kind of.

35:53 Characteristic hockey stick,

35:55 uh,

35:55 for those of you who are from North America,

35:58 um,

35:58 type of pattern,

35:59 uh,

36:00 this too has shown up in a remarkably broad set of circumstances across the world,

36:05 you know,

36:05 here's something similar not only in Denmark and Australia on the top,

36:09 but you see very similar exposure patterns,

36:12 uh,

36:12 in Africa from the Allesina work and in uh Brazil,

36:16 um,

36:16 on the bottom.

36:17 And again you see very similar things where it's not

36:19 just that exposure to these local places is incredibly important,

36:23 it seems to be exposure during

36:25 childhood

36:26 is the most important thing.

36:28 And of course the exact definitions of what that means

36:30 are going to differ between places and and how one is

36:33 um measuring upward mobility,

36:36 um,

36:36 but this really does seem to be a very robust fact in many places across the world.

36:42 Now,

36:43 the fact that these local exposure to to place are causal,

36:47 then of course

36:48 just raises the question of,

36:49 well,

36:50 what is it about these local places that that is causal,

36:53 and here I think,

36:55 um,

36:56 you know,

36:56 again,

36:56 the answers are going to differ.

36:58 What do we get from looking in the US at these types of things?

37:03 It's not just raw growth that seems to matter actually.

37:06 This slide might be helpful.

37:07 This is a scatter plot of average growth across US cities on upward mobility

37:12 in the 30 largest cities.

37:14 There's just no correlation at all.

37:16 So just broad-based growth does not necessarily help.

37:20 What does seem to matter are these features that are much more about

37:24 not only human capital development and exposure,

37:27 you know,

37:28 especially exposure to abject poverty is not a good thing.

37:31 But you see that the social circumstances and the

37:34 networks seem to matter quite a bit as well.

37:37 And so in particular you see the third one here,

37:40 greater social capital.

37:41 What seems to be going on here is exposure not only to adults who

37:46 can

37:47 not only inspire.

37:48 you to give you information about the pathways that

37:50 are available that can help make you make the choices

37:53 that might maximize your potential as a child,

37:56 but also exposure to other peers of yours

38:00 that matters tremendously as well.

38:01 And so here's a map of the United States

38:04 from um.

38:05 Some of Raj Shetty's recent work on social capital,

38:08 this looks enormously like the picture of upward mobility that I showed you before,

38:13 um,

38:13 and so my sense is that these types of uh social exposures are going to matter,

38:17 uh,

38:17 in addition to all these more directly human capital focused activities,

38:22 you know,

38:22 here's a similar thing looking across

38:24 regions.

38:25 From the,

38:25 the Allesina work,

38:27 um,

38:27 again,

38:27 what pops out here is not only some of

38:30 those raw economic factors like distance to a railroad,

38:34 um,

38:34 but it's things like distance to a Protestant mission which is gonna,

38:37 gonna be a stand-in for a lot of the cultural differences that,

38:39 you know,

38:40 exist,

38:40 uh,

38:40 across these institutional differences that exist across these,

38:43 these places.

38:45 So,

38:45 um,

38:46 kind of three takeaways as,

38:47 as my time is ending.

38:49 Um,

38:49 this is the first talk I've ever given where like the luge,

38:52 my time is being measured down to 10,000 of a second.

38:55 It's extremely precise.

38:57 Um,

38:57 luge is the only Olympic event where time is measured in the 10,000 of a second.

39:01 Um,

39:02 so 3 takeaways.

39:03 First,

39:04 it's the local childhood environment,

39:06 I think that really plays the central role

39:08 for developing prospects for upward mobility,

39:11 and it's environment again,

39:13 local,

39:13 I think is the key word here.

39:14 It's really

39:15 about kind of who you're interacting with on a daily basis.

39:18 You shouldn't even think about this is at the level of a city.

39:20 It's more like the,

39:20 the neighborhood that you're living in.

39:22 Second.

39:23 What I think this shows is that one,

39:26 much like Ufuk was talking about,

39:27 you can use these large scale observational data

39:30 to make,

39:31 uh,

39:31 to,

39:31 to learn something about what types of

39:33 interventions might be most valuable and where,

39:36 uh,

39:36 right,

39:36 this is,

39:37 you know,

39:37 not going to be as precise or as direct as an RCT for a particular intervention,

39:43 but this is going to give you perhaps a better sense of like what bucket to look in.

39:47 And third,

39:48 I think it's again we have to move beyond these,

39:51 uh,

39:51 you know,

39:51 very traditional models of human capital development,

39:54 uh,

39:55 right?

39:55 I think it's really about a lot of these broader sociological forces,

39:59 um,

39:59 that have a lot to do with economic mobility and inequality as well.

40:02 Thank you very much.

40:04 Many thanks Professor Friedman for the

40:07 very interesting and sobering presentation.

40:09 We'll hear next from Professor Shin.

40:12 Uh,

40:12 Professor Shin will present,

40:14 uh,

40:14 a presentation focused on growth through creative destruction

40:18 from the point of view of firms,

40:21 uh,

40:21 focusing on the South Korean growth miracle.

40:24 Over to you.

40:25 Thank you.

40:26 So South Korea is

40:28 one of the well-known economic success stories.

40:31 It went from poor to rich in one person's lifetime,

40:35 actually more like 30 years.

40:37 And a lot of research has been done,

40:39 a lot of theories have been expounded on how they

40:42 made this miracle.

40:44 What has not been possible until now

40:46 was an analysis of what is happening at the micro level,

40:50 at the plant level.

40:52 Is there a systematic pattern that accompany

40:54 this macroeconomic transformation at the micro level?

40:57 So we're able to do that.

40:59 So this is the motivating figure.

41:02 The bottom line is the GDP per capita

41:04 at domestic constant prices.

41:07 So if you look at it,

41:08 two things,

41:09 it's a low scale.

41:11 From say 1970 to 2000 in 30 years,

41:14 the GDP per capita

41:16 uh

41:16 goes up by a factor of 10,

41:19 from 700 to 7000.

41:20 This is remarkableable growth rate.

41:22 The other thing you notice is that the growth

41:24 rate kind of slows down over each decade.

41:27 In the 70s and eighties,

41:28 it's about 9% to 8% per year growth

41:31 and 202,010 is 4% and 3% per year and

41:35 Koreans call that a recession.

41:38 And then,

41:39 the blue line is manufacturing value-added per worker,

41:42 so that's your measure of productivity of the manufacturing sector.

41:45 Of course,

41:45 it's

41:46 higher than the GDP because manufacturing tend

41:48 to be more productive than service sector.

41:51 Uh

41:52 broad patterns are consistent.

41:54 The subtle differences that manufacturing

41:57 grew really productive in the 90s,

41:59 in the value added per worker sense,

42:01 despite the '97

42:03 East Asian financial crisis,

42:05 and then it really slowed down after 2010.

42:09 So,

42:10 we're using

42:11 newly digitized data,

42:13 manufacturing

42:14 census data.

42:16 We already had data from 1982 onwards,

42:19 but what is new is the previous period,

42:22 196,

42:23 1017,

42:23 1981.

42:24 So,

42:24 one limitation is that the panel dimension of the data

42:28 is available only from 1982,

42:30 and then we're gonna actually spend some time in the second half of my

42:34 presentation about that part.

42:35 So first,

42:36 we start with the first half,

42:38 the cross-section data only.

42:40 It starts from 1967.

42:43 Since

42:43 the cutoff of employment is 5,

42:47 we're only looking at plants with at least 5 employees.

42:51 Until 2007,

42:51 after that,

42:52 it becomes at least 10 employees.

42:54 So,

42:55 you look at the right figure,

42:56 the average size of the manufacturing plants rises dramatically

43:00 in the 1970s and collapses dramatically in the 1980s.

43:03 There's some difference.

43:04 1970s is the period of

43:07 heavy and chemical industry promotion,

43:09 so a lot of this rise is

43:11 compositional.

43:12 So

43:12 you're moving away from light manufacturing to heavy

43:15 and chemical industry and chemical industry obviously has much

43:18 more employees

43:19 per uh plant.

43:22 And then the

43:24 collapse in the 1980s,

43:25 actually,

43:26 that's not compositional.

43:27 So it is happening in basically all industries within having chemical industry,

43:31 within manufacturing and other industries.

43:33 So there's some differences in how it goes up and down.

43:36 There are different ways to look at size.

43:38 This is what fraction of employment

43:40 in total employment is in plants with at least 2 250 or 500 employees.

43:46 Same pattern goes up and down.

43:50 And then this is log log plot.

43:52 So some of you are very familiar with this plot.

43:54 It basically shows the entire distribution,

43:56 so it's sort of a reverse of the cumulative distribution function.

44:00 So

44:01 at any point on the x axis,

44:03 pick two,

44:03 that's the logo where it stands,

44:04 so that's 100,

44:05 and you ask of all the plants,

44:07 what fraction of plants are actually larger than 100 employees.

44:11 And it's,

44:12 it's,

44:12 it's moderately declining.

44:14 So solid line in the middle you see is 1967,

44:17 which is showing every decade.

44:19 And then you see that in 1977 it shifts out

44:23 quite a bit.

44:23 It's almost like a for short of stochastic dominance.

44:26 So it's basically what it is is that the entire distribution shifts to the right,

44:30 so all the plants are basically much larger in 1977 and 196967,

44:34 but that was the peak of the plant size distribution,

44:37 and then you see

44:38 1987 and 1997,

44:40 2007,

44:41 actually you see these things coming back,

44:43 so things are getting smaller.

44:44 So it's very

44:45 consistent throughout um

44:47 The entire distribution,

44:48 that's just something at the top.

44:50 So

44:51 what,

44:51 at this point you take a stock and

44:53 compare the picture of

44:55 manufacturing value added per worker and the size distribution changes.

44:59 What you see is that there's no clear pattern

45:01 either in levels because

45:03 value per worker continues to grow,

45:04 but the size ra goes up and down.

45:07 Or in terms of growth rate.

45:09 So we had a fast growth in the 70s and 80s and 90s,

45:12 but then

45:13 sometimes science division goes up

45:15 and comes down,

45:16 so there's no simpler relationship.

45:18 So it prompted us to

45:20 look at

45:21 it that's over time,

45:22 we don't see any clear pattern between size and

45:24 value per worker or the growth rate of that.

45:27 So we

45:27 wanted to see whether actually there's any pattern

45:30 across countries.

45:31 So more research is definitely needed.

45:33 uh we just went to OECD database and grabbed whatever we could.

45:37 On the left panel,

45:38 these are countries including South Korea that report in

45:41 terms of plants rather than farms or enterprises.

45:44 So what you see is that

45:46 Um,

45:47 you're attempting to make,

45:48 uh,

45:48 oh,

45:48 there's a downward sloping

45:50 pattern here.

45:51 First,

45:51 only 6 countries here,

45:53 and also you know that there are many other countries,

45:55 for example,

45:56 India,

45:57 who talked about

45:58 many other countries with a really small employment

46:01 and very low value added per worker,

46:03 so there's no simple relationship.

46:04 The right hand side is from same OECD data

46:07 uh looking at countries that report in terms of enterprises and firms,

46:11 you see that there's no clear pattern here either,

46:14 at least among these countries in terms of the

46:15 size of a farm measured by average employment per farm

46:19 and the value added per worker of the sector.

46:21 This is a manufacturing.

46:23 So there doesn't seem to be a very simple relationship between

46:26 size and productivity or growth.

46:29 OK.

46:29 The other thing we can do

46:31 with this

46:32 static data,

46:33 cross section data is

46:34 do simple

46:35 calculation of misallocation,

46:37 sharing clinotype.

46:39 Um

46:39 you can see I plot,

46:41 but let's talk about the right one.

46:43 So if you look at it really carefully,

46:44 oh,

46:45 the data from 69 to 77 is missing because those

46:48 years we don't have a capital at the plant level,

46:50 so we cannot do the calculation.

46:52 So what you see is actually

46:53 there seems to be a downward slope in degree of missile location.

46:57 So location efficiency is actually improving

47:00 between

47:01 1968 and up to mid 1980s,

47:03 but then it gets worse in the

47:06 1980s and later part of the 1980s and 191990s,

47:09 and it gets really bad in the 2000s.

47:11 I think the last phase is more about the issue of missile specification.

47:14 I will not read too much into that.

47:16 But to the extent that we think this is a pattern misallocation,

47:19 then you again don't see a really

47:21 clear pattern between the degree of misallocation and

47:23 either level or growth of the value added per worker,

47:26 because again,

47:27 value per worker in manufacturing is growing constantly

47:30 and then it was growing fast in the 70s

47:33 when misallocation is coming down,

47:35 but then it's also growing

47:38 fast when misallocation is getting worse.

47:39 So there doesn't seem to be a very clean pattern here.

47:42 This is just another way to look at,

47:45 now it's a popular way to look at degree of misallocation.

47:47 It's a coefficient of regress if you know of TFPR and TFPQ

47:51 at the plant level and then aggregating it.

47:54 Same pattern misallocation comes down

47:56 and then

47:57 it takes off in the late 90s and 2000s.

48:01 So,

48:01 overall,

48:02 in the 70s,

48:03 I'm just kind of adding a picture here from a paper of mine.

48:07 Overall,

48:08 the,

48:08 in the 70s,

48:09 we just saw that actually misallocation was coming down

48:12 in in Korea,

48:14 but there's something interesting pattern.

48:15 So this is

48:16 active heavy chemical industry promotion period.

48:19 So what we see is that if you actually look at

48:21 the industries and regions that were targeted by this government.

48:24 The policy,

48:25 there was a lot of increase in

48:27 misallocation within those industries and regions compared to industries

48:30 and regions that were not targeted with the policy.

48:33 So

48:34 the aggregate pattern

48:36 belies some of the more interesting features at the

48:39 more desaggregate level patterns.

48:42 Now,

48:42 in the remaining few minutes,

48:43 I'll talk about the dynamic part.

48:45 This is where really interesting things happen and it connects

48:48 very well with what Wolf was saying in terms of the

48:51 trumpeterian growth models.

48:53 So,

48:54 the first set of pictures I'll show you is this charming picture picture.

48:58 So this is the Davis Hettinger shoeching

49:02 picture.

49:03 So this is a histogram.

49:04 So what is a histogram?

49:05 So it's basically showing the

49:07 employment share of some beans,

49:09 and what are the beans?

49:10 The bean is

49:11 basically the growth rate of a plants.

49:14 So you see the number around 0,

49:15 so the employment cluster around 0.

49:17 So those are workers working in plants,

49:20 that's not really.

49:21 Growing or shrinking.

49:22 Around 11 means this

49:25 plant is growing really fast,

49:28 uh,

49:28 -1 means it's shrinking really fast.

49:30 2 is entry,

49:31 and -2 is exit.

49:33 So it kind of shows the improvement change.

49:35 It's a small change is important,

49:37 plant size

49:38 plant size is important too.

49:39 It's more big

49:41 plant size change is important.

49:43 And in the 80s,

49:44 you see the importers entry really big time,

49:47 right?

49:48 So this is the period where the employment

49:51 average size

49:52 measured by employment of plants was collapsing.

49:55 And that's because there's huge entry and

49:56 entrance obviously is smaller than incumbents.

49:59 So that's when in all sectors of the economy there's a huge spike of entry

50:02 and the average size is coming down.

50:04 But at the same time you see actually there's a lot of firms

50:07 adjusting upward,

50:08 a lot of growing firms.

50:10 But then the next period,

50:12 87 to 92 or 92 to 97,

50:14 slight different pattern.

50:15 You see that things are kind of shifted to the left,

50:18 so entry is not as important.

50:19 Now exit

50:21 actually plays a more important role,

50:22 like a minus 2,

50:23 and then the whole picture is to the left,

50:25 meaning that you see a lot of plants downsizing actually.

50:29 And I think this may be the secret to what was um

50:33 why it was possible for South Korea to sustain the economic growth so long

50:37 that

50:38 In the 70s,

50:39 the government had a really heavy chemical industry policy

50:42 and

50:42 when they stopped it in the 80s,

50:44 you have this fresh batch of new firms coming in

50:47 and that pushed a lot of bad firms,

50:50 big firms to actually

50:51 uh

50:52 downsize and exit in the

50:54 80s,

50:55 the later part of the 80s and early 90s.

50:57 And I think that's really important for understanding how the productivity

51:00 growth measured by valuable worker grew so fast in the 90s.

51:03 So this creative destruction is playing an important role.

51:06 And you look at the last period,

51:09 2008 to 2013 to 2008,

51:12 it's very quiet here.

51:14 What you can see is that while entry and exit is not playing its important role,

51:18 and also there's a lot of implement here around zero,

51:20 meaning that farms are not really growing or shrinking,

51:23 so it's very quiet.

51:24 So this is a reduction in dynamism

51:26 or reduction in churning.

51:29 So the prior set of research we just showing you,

51:31 well this is how plant size is changing,

51:33 measured by employment.

51:35 But for productivity,

51:36 we,

51:36 we really care about the correlation,

51:38 meaning you want the productive farms to grow

51:40 and unproductive farms to shrink.

51:42 So I'm showing you some of the

51:45 regression results following

51:47 uh Dekker Her Jeremy Miranda approach.

51:51 And what I'm showing you is

51:53 whether

51:54 farms grow

51:55 in terms of employment or capital or investment

51:58 when they have a good productivity shock.

52:00 And the first row numbers are positive and significant,

52:03 so that's good news.

52:04 Yes,

52:04 so farms are responding to productivity shocks,

52:07 so

52:07 uh that's good for economic growth.

52:08 But what's really interesting is

52:10 the second panel when we do things

52:12 decade by decade.

52:14 So what you see here is that

52:16 especially the employment growth part,

52:18 1990s,

52:19 that's when

52:20 the productivity is growing the most.

52:22 That's when

52:23 farms are most sensitive in terms of product of shock.

52:26 And then

52:27 the number gets really small over time.

52:29 It's been 2010 it's really quiet,

52:31 so even if you have a good pery shock,

52:33 you're not going to raise your employment and capital very much.

52:36 So there's the same picture,

52:38 the reduction in dynamism.

52:39 Firms are not really responding to perry shocks,

52:41 and I think that really.

52:43 Responds very well with

52:44 what we saw in the overall pattern that things really

52:47 slowed down towards the end of the period 200,

52:50 2010 went from stopping

52:51 uh dynamic in terms of either churning or

52:54 how responsive they are to productivity shocks.

52:57 So I'm about to run out of time,

52:59 so let me wrap up.

53:01 So,

53:01 as you saw,

53:03 at the micro level,

53:04 there was no really simple

53:05 systematic pattern of correlation between what's happening at the micro level

53:09 and

53:10 macro

53:11 transformation of the Korean economy

53:13 because

53:14 uh the economy

53:16 continues to grow,

53:16 but the size decision could move upward and downward,

53:20 and then with misallocation measures,

53:22 there's no simple pattern there either.

53:24 What's really promising was the

53:26 Uh,

53:27 the later part when

53:28 the dynamism is slowing down in terms of churning or responsivity shocks,

53:32 that's when the aggregate economy really slows down in terms of growth.

53:36 And I think that's really relevant for thinking about middle income countries.

53:40 So

53:40 this is a very promising line of work.

53:42 The problem is that

53:43 it presents a very high hurdle

53:45 for data requirements,

53:47 right?

53:47 So now you need a panel dimension

53:49 and you need to cover enough variables to compute

53:52 the productivity reform and input uses and everything.

53:55 So it may not be the best research program for low

53:58 income countries that do not have the right data infrastructure,

54:01 but for middle income countries I think it's a promising avenue.

54:04 Uh because when productivity slows down,

54:06 you don't know where to look at,

54:07 but when you have this micro evidence,

54:09 then you can be guided by economic theory

54:11 and look at

54:12 um

54:13 the sources of these

54:14 barriers and frictions that's slowing farms down and reallocation down

54:18 and that will lead us to

54:19 uh policy prescription.

54:21 Thank you.

54:25 Many thanks Professor Shin.

54:26 It's always a pleasure to hear about the,

54:28 the Korean growth miracle.

54:31 Uh,

54:32 finally,

54:32 we will hear from Professor Grubb.

54:35 Professor Greb will focus on the very important topic of

54:39 economics and policies of energy transition over to you.

54:44 Well

54:44 thank you very much indeed for the invitation and this session.

54:48 It's been absolutely fascinating.

54:50 Uh,

54:51 the only problem is it's left me with about 76 or 7 minutes to coffee.

54:55 Uh,

54:56 I'll,

54:56 I'll do my best.

54:58 Um,

54:58 so I'm going to touch on,

55:01 uh,

55:01 climate change and energy transition,

55:04 um,

55:04 just to a word of introduction that,

55:07 uh,

55:08 first,

55:09 uh,

55:09 I'm also not a development economist.

55:13 I'll be covering more of an energy economist

55:17 and special interest in innovation.

55:19 I will give a brief background,

55:22 uh,

55:22 possibly about half the time on

55:24 the context of climate change scene setting

55:28 because obviously there's a broader conference,

55:30 and then move on to particularly the innovation

55:34 dynamics that we've seen and some implications.

55:38 That's partly because climate change issues are so often

55:41 couched in.

55:43 Everything's going horribly wrong and it's a very doom and gloom scenario.

55:47 I do want to inject a perception that it's more interesting than that,

55:52 not least when we reflect on two things,

55:55 which is that economic development,

55:58 two of the fundamental factors that we've touched on,

56:00 are innovation

56:02 and cheap primary energy.

56:04 The question has always been,

56:05 is there potentially a clash with climate change?

56:08 I will be drawing significantly upon aspects of the IPCC report.

56:14 Many of you,

56:15 I think,

56:15 will know,

56:17 a lot.

56:18 The broad background IPCC is a massive international process.

56:22 These assessments roll around

56:24 every 6 or 7 years,

56:26 uh,

56:26 and we had the 6th assessment,

56:28 uh,

56:28 last

56:29 published last year.

56:31 So I'll draw quite a bit on that.

56:33 I will start with the science

56:35 and risk dimensions of it.

56:38 On the left hand side,

56:40 we have one of those famous graphs of scenarios of global temperature

56:46 suggesting a very wide range

56:49 of possibilities.

56:52 So

56:52 apparently a lot of agency in terms of what happens to the temperature this summer,

56:58 this century.

57:00 On the right hand side

57:02 you have a fairly well known IPCC diagram called the five reasons for Concern,

57:09 which range from loss of

57:11 specific ecosystems on the left hand side

57:13 through on the right hand to globally substantial

57:16 human impact and global tipping points where entire

57:20 Earth systems collapse.

57:23 The good news is on the left hand side,

57:25 I no longer believe any of those high projections.

57:27 They're not going to happen.

57:29 The bad news is on the right hand side,

57:32 all of those reds and yellows have crept downwards every assessment,

57:37 and I should have said really

57:40 intuitively you can see the yellow

57:42 is where there is a significant probability of something somewhat worrying.

57:47 The red it's getting bad and the purple is getting catastrophic,

57:52 and those two roughly.

57:54 seem to sort of cancel out,

57:56 which means climate change

57:58 looks like just as severe a problem as it always has been,

58:01 except that it's a lot more immediate and prescient

58:05 as indeed we see in the the extreme impacts

58:09 we see around the world.

58:11 So we are actually in a zone where probably looking

58:15 at somewhere between 1.5 and 3 degree temperature rise.

58:21 But that is a huge range in terms of what it may imply for humanity.

58:27 It is also a huge range in terms of of

58:29 how demanding or what it implies for the energy system.

58:32 Now that's been the scientific framing around risks.

58:36 The other framing,

58:38 if you like,

58:38 is the economics one.

58:39 What an economist by and large

58:42 is most interested in

58:43 is the trade-offs potentially

58:46 and some numbers.

58:48 Well,

58:48 we've had

58:50 a

58:50 good 2 or 3 decades of economists trying

58:53 to attach the most significant number they think,

58:55 which is

58:56 the social cost of carbon emissions,

58:58 how much damage,

58:59 what's the dollar equivalent

59:00 of a ton of carbon.

59:02 Some,

59:03 I think

59:04 a lot of non-economists find this whole idea somewhat distasteful or dubious,

59:09 and there's been extraordinary ranges in the attempts to attach a number.

59:14 But nevertheless,

59:15 it is striking that although I wouldn't say there's been a lot of convergence,

59:21 there has been a clear trend.

59:23 The number keeps getting higher

59:25 and the range on the upside to some extent gets higher.

59:28 And here in yellow I've highlighted a paper published in Nature

59:34 a few months ago.

59:35 I think relevant not just because it was in

59:38 Nature and a serious effort and serious authors,

59:40 but it was a response

59:42 to the call of the US National Academy of Sciences a few years ago saying,

59:45 look,

59:46 we've learned lots of things

59:48 and that has not been adequately put through into revising estimates of the SCC,

59:53 uh,

59:53 the social cost of carbon.

59:55 Uh,

59:55 and indeed one of the,

59:56 if you like,

59:57 the most sort of

59:59 climate economic skeptics,

1:00:00 Richard Tol,

1:00:01 uh,

1:00:02 published to say,

1:00:02 yeah,

1:00:03 actually he recognizes the cost has gone up quite a lot.

1:00:07 If you fully cost at the kind of levels suggested in that most recent study,

1:00:11 which is

1:00:12 about 3 to 4 times what a lot of economists have been saying only a few years ago,

1:00:17 then basically the cost of climate damages

1:00:19 from burning coal is substantially higher than the cost of buying coal.

1:00:23 And in fact,

1:00:24 in oil,

1:00:25 you may be looking on the order of the equivalent of about $50 per barrel

1:00:30 of oil associated with the climate damages,

1:00:32 plus or minus still a a a sizeable range,

1:00:36 but

1:00:37 that's pretty significant.

1:00:39 In terms of what we then look on in longer term goals

1:00:44 and the framing context on emissions,

1:00:47 well,

1:00:47 arising from the Paris Agreement,

1:00:50 governments agreed two kinds of emission goals.

1:00:52 It's worth saying

1:00:54 one was the objective to stabilize temperatures,

1:00:58 effectively interpreting the original convention's

1:01:01 commitment to stabilize atmospheres.

1:01:04 At a level to avoid dangerous interference,

1:01:07 governments agreed that means well below 2 degrees and striving towards 1.5.

1:01:13 And

1:01:14 the vast majority of major countries around the world have

1:01:17 declared net zero emission goals between 2050 and 2070,

1:01:23 which roughly corresponds with that range.

1:01:26 The other emission

1:01:27 goals was to be expressed nationally

1:01:30 through the nationally determined contributions.

1:01:33 On the left is the chart from the IPCC report which says,

1:01:36 well,

1:01:37 the red is where we think we were going given policies as of 2020.

1:01:43 The purple

1:01:44 is what those NDCs

1:01:47 indicated as national ambitions to 2030.

1:01:52 Obviously way off track,

1:01:55 so the Paris Agreement embodies two

1:01:57 currently fundamentally inconsistent things a long term goal

1:02:02 and a nationally determined

1:02:05 trajectories which are not consistent with that long-term goal.

1:02:09 If we go further and say what

1:02:11 The model suggest to be the

1:02:14 often least cost trajectory?

1:02:16 Well,

1:02:16 it starts by radically changing course,

1:02:19 dropping emissions by

1:02:20 somewhere between

1:02:22 25%

1:02:23 and

1:02:24 at least 45%

1:02:26 by 2030.

1:02:27 That's only 7 years away,

1:02:29 to be honest,

1:02:29 I don't think that's going to happen.

1:02:33 So if I now start moving forward to the analytics,

1:02:35 because some of you may know that background,

1:02:38 um.

1:02:39 This is an economics conference.

1:02:41 There's plenty we can say about climate change economics.

1:02:46 Probably the strap line was best captured by Nicholas Stern,

1:02:49 who described climate change as the biggest market failure in history.

1:02:53 But it has also been described as the perfect moral storm

1:02:58 because of the ethical,

1:02:58 it raises ethic issues about our responsibilities

1:03:02 to other people,

1:03:03 other countries,

1:03:04 other generation,

1:03:05 and human beings do have a sort of

1:03:08 almost infinite capacity to to try and find some way

1:03:11 of saying it's not really my responsibility per se.

1:03:14 It's been described as a super wicked problem,

1:03:16 one that is from a social science perspective,

1:03:19 just really hard

1:03:21 to solve.

1:03:23 And it's also been described in terms of psychological distance.

1:03:26 Human beings kind of evolved,

1:03:28 didn't evolve

1:03:29 to even be able to conceive about something on this

1:03:32 scale,

1:03:33 globally,

1:03:34 intergenerationally,

1:03:36 planetary risks,

1:03:37 etc.

1:03:39 In the IPCC we try and also map out the corresponding analytic frameworks.

1:03:44 We're all here familiar with aggregate efficiency.

1:03:47 I'm going to say a few words about the ethics and equity dimensions more explicitly,

1:03:52 but really focus on the third

1:03:54 innovation and transition in the second half of my talks,

1:03:58 not to forget the psychology and political

1:04:01 dimensions of this.

1:04:03 So

1:04:05 just on the equity front briefly,

1:04:07 the world is unequal.

1:04:09 I don't have time to run you through the details.

1:04:11 That chart on the right shows different regions

1:04:14 vertically per capita emissions against population.

1:04:18 The the dark blue is fossil fuel CO2,

1:04:20 the orange

1:04:22 is land use,

1:04:23 uh,

1:04:23 which is a much more concentrated issue.

1:04:26 Broadly you see substantial variation vertically between countries,

1:04:30 but not just north-south,

1:04:32 also within countries at similar regions of development.

1:04:35 Two other points I'll draw out from this.

1:04:38 Uh,

1:04:38 one is,

1:04:39 yes,

1:04:39 it's partly about trade and offshoring,

1:04:42 but actually even our consumption footprint.

1:04:45 Uh,

1:04:45 that amplifies but doesn't fundamentally change the kind of conclusions there.

1:04:50 And also

1:04:52 it's not just about between countries.

1:04:54 In fact,

1:04:54 a little bit like,

1:04:55 uh,

1:04:56 John's presentation earlier,

1:04:59 there's inequalities at every step of

1:05:00 the system within countries internationally.

1:05:03 You can look at

1:05:03 roughly 10% of the richest households around the world

1:05:08 account for

1:05:09 probably over a third of emissions through their consumption patterns.

1:05:13 I'm not going to describe the chart on the left,

1:05:15 it's a more elaborated version of what I said,

1:05:18 greater disaggregation,

1:05:19 but this time plotted against Human Development Index.

1:05:22 What it really says is,

1:05:24 All

1:05:25 countries and regions will need to change,

1:05:28 but what and how they need to change will vary a lot.

1:05:32 But a lot of the discourse now

1:05:34 is really about in terms of the development,

1:05:38 paradox and challenges,

1:05:39 not just alignment and co-benefits,

1:05:42 but shifting development pathways towards sustainability.

1:05:46 So where does technology come into it?

1:05:49 Well,

1:05:50 we've seen big reductions in the cost of renewable energy,

1:05:53 as I'm sure you'll well know.

1:05:56 These are classic charts now

1:05:58 of solar and wind and below the expansion rates

1:06:02 exponentially at rates of 20 to 40% a year.

1:06:06 Generally now,

1:06:07 both solar and wind,

1:06:08 the two biggest renewables,

1:06:11 uh,

1:06:11 not everywhere,

1:06:12 but as a broad international aggregate,

1:06:15 are cheaper than fossil fuels as a way of generating electricity.

1:06:19 That is a pretty fundamental change

1:06:22 from what we looked at a few years ago.

1:06:25 How long can that exponential growth rate,

1:06:27 and by the way,

1:06:28 even highly numerical experts,

1:06:31 I find,

1:06:32 are not actually that good

1:06:34 at really thinking what does exponential growth mean.

1:06:37 I invite you in the coffee break,

1:06:38 pull out your computer,

1:06:39 stick in exponential growth at these rates,

1:06:41 look what happens within this decade.

1:06:43 It is really quite striking.

1:06:46 And the data suggests as published last Thursday by the IEA,

1:06:50 it ain't slowing down in a hurry.

1:06:52 Expansion from about 200 gigawatts a year pre-COVID,

1:06:58 then 300 this year,

1:07:00 the IEA estimates

1:07:02 400 to 500 gigawatts of renewable energy being installed.

1:07:07 That's a huge growth rate.

1:07:10 Um,

1:07:12 OK,

1:07:12 I'm going to have to skip the audience participation,

1:07:15 but

1:07:17 All technologies,

1:07:19 pretty much in energy have displayed two things.

1:07:22 All of these turned out to be much cheaper than anyone projected.

1:07:26 And they all actually involved government action at scale

1:07:30 over many years on both the technology resource mapping,

1:07:34 development,

1:07:35 and the demand pool and price developments.

1:07:38 So where does this take us in terms of wider transitions?

1:07:41 Uh,

1:07:41 I'll,

1:07:42 I'll draw on,

1:07:42 but I won't have time for the depth,

1:07:44 obviously

1:07:45 quite a major international program looking at these issues,

1:07:48 um,

1:07:49 involve partners,

1:07:50 particularly in India,

1:07:51 China,

1:07:52 and Brazil,

1:07:53 um,

1:07:54 and we looked at a few case studies.

1:07:56 Uh,

1:07:57 wind,

1:07:57 uh,

1:07:58 both Europe and Brazil

1:08:00 moved from 1% to 10% of,

1:08:03 uh,

1:08:03 more than 10%

1:08:04 within a decade of wind energy.

1:08:06 Uh,

1:08:07 India,

1:08:08 light lighting efficiency,

1:08:10 extraordinary transformation partly driven by bulk government procurement,

1:08:14 which resulted in what they call the cheapest lighting in history.

1:08:18 Uh,

1:08:18 that sort of echoes now what the IEA says about solar

1:08:22 PV,

1:08:23 which is the cheapest electricity in history.

1:08:26 So.

1:08:27 Yeah,

1:08:27 lots of good news stories.

1:08:29 What do we learn from those case studies?

1:08:31 Well,

1:08:32 as I said,

1:08:33 significant government action,

1:08:34 largely now self-sustaining,

1:08:37 wouldn't have been pursued by the traditional economic prescription,

1:08:41 competitive markets,

1:08:42 price carbon,

1:08:43 um,

1:08:44 and probably not justified by cost-benefit appraisal if you took all

1:08:47 of the projections of how much these things would cost.

1:08:50 But the common themes cumulative progress,

1:08:53 market-based innovation,

1:08:55 sustained and targeted support beyond the pure R&D.

1:09:00 Substantial uncertainties when when these efforts set out

1:09:05 and strong international dimensions,

1:09:07 the movements of technology frontiers from one region to another,

1:09:11 partly reflecting swings of policies.

1:09:14 The point being,

1:09:15 the traditional simplified stories of either it's all about

1:09:19 government R&D drops the price and then things diffuse,

1:09:22 the A in the bottom left,

1:09:24 nor

1:09:25 the let's just do it through the market,

1:09:27 flat-paced technology.

1:09:29 Neutrality

1:09:31 and then companies will look forward and say that's what I'm going to do.

1:09:35 That's not reality.

1:09:36 Most technologies iterate through repeated cycles of

1:09:40 expansion,

1:09:41 learning,

1:09:42 innovation,

1:09:42 expansion,

1:09:43 learning,

1:09:43 innovation,

1:09:45 cost scale economies,

1:09:46 learning,

1:09:47 etc.

1:09:48 Right,

1:09:49 innovation's a complex journey and with

1:09:51 just tripped 0 time,

1:09:53 I don't have much

1:09:54 chance to go through it.

1:09:56 This slide looks terribly complicated.

1:09:59 It basically says it's not just about technology.

1:10:02 Every stage of this process requires developments in

1:10:05 all of those six factors you can see on the left.

1:10:08 All of those to some degree have to develop along with the technology.

1:10:13 It's more of a sort of a map of a multi,

1:10:15 multi-level journey

1:10:17 to point to examine where might the big blockages occur

1:10:21 that would get in the way of a fuller transition.

1:10:24 So this will bring me towards final remarks.

1:10:29 And I should have said I know

1:10:31 I had a choice to make.

1:10:32 Either I could try and pretend to be a

1:10:34 development economist or I'd stick to what I know.

1:10:37 There are a lot of issues,

1:10:38 and in a second I'll wind up by by noting some of the barriers.

1:10:42 But basically

1:10:43 transitions always start small,

1:10:45 may take years after technology emerges through a phase of

1:10:48 market emergence before they really emerge into the mainstream.

1:10:53 The impact on incumbent industries is often barely visible during that process,

1:10:57 why?

1:10:57 Because exponential growth starts small,

1:10:59 looks small,

1:11:00 and the big,

1:11:00 big,

1:11:00 big boys don't really notice it.

1:11:03 But boy do they start to change things at the margin and then more fundamentally.

1:11:08 I've got another version which shows some of

1:11:09 the incumbent industries as opposed to technologies,

1:11:12 maybe jumping on the rising blue curve,

1:11:14 as was already alluded to.

1:11:16 But a key thing tends to be

1:11:19 at this sort of

1:11:21 phase in the middle that requires a reconfiguration often of market structures,

1:11:26 because of all energy markets ultimately are human creations,

1:11:30 and the kind of markets that suit fossil fuels don't necessarily adapt well

1:11:35 and most cost effectively for renewables.

1:11:37 So this other chart takes uh actually a sort of

1:11:41 concept from a book that we published a few years ago around 3 pillars of policy

1:11:45 and turns it into a dynamic chart

1:11:48 that

1:11:48 one does need phases of strategic investment to foster emerging technologies,

1:11:52 businesses,

1:11:53 etc.

1:11:55 Then one needs to evolve,

1:11:56 refigugure infrastructure

1:11:58 and market structures to adapt suited to new technologies,

1:12:02 build up the scale and technologies,

1:12:04 supply chains,

1:12:05 accelerate global diffusion.

1:12:07 You may well also have issues of attending to standards,

1:12:10 behavior,

1:12:11 uh,

1:12:12 norms,

1:12:13 and maybe measures to address er laggards and and obstacles.

1:12:18 So,

1:12:18 um,

1:12:20 I'm going to finish the slideshow there with,

1:12:21 with one exception,

1:12:23 but let me just say,

1:12:24 of course,

1:12:25 uh,

1:12:26 capital finance are absolutely crucial.

1:12:30 Uh,

1:12:30 if you look generally in aggregate,

1:12:32 you'd say to meet the kind of transitions projected by the IPCC as needed,

1:12:36 we need to expand the pace of international investment or global investment

1:12:41 by a factor of between 3 and 4 compared with 2020 levels.

1:12:45 And if you look differentially,

1:12:48 of course that gap is bigger in much of the developing world.

1:12:52 I won't run you through

1:12:53 those numbers,

1:12:54 but what is required

1:12:56 to actually get such large scale investment

1:13:00 into what look like potentially

1:13:02 attractive

1:13:03 investments is not trivial,

1:13:05 and of course it depends a lot on the cost of capital.

1:13:08 Um,

1:13:09 I think all other presenters have presented econometric studies.

1:13:13 I've sort of got one in my back pocket which we expected to

1:13:16 give a more technology oriented answer to the cost of capital and it,

1:13:21 and you know,

1:13:21 cost of technologies including solar.

1:13:24 It basically said looking across countries,

1:13:26 it's access to capital markets,

1:13:28 the terms and the perceived country risks

1:13:30 that are dominant factors.

1:13:32 So still big issues that obviously the World Bank

1:13:35 and others need to figure their way through.

1:13:38 My concluding slide

1:13:40 is

1:13:41 just to do a simple diagram but conceptually still quite powerful

1:13:46 and and a set of messages.

1:13:48 If you look at this,

1:13:49 it's not just a new technology,

1:13:51 a few green technologies,

1:13:52 it's an completely different structure of our energy systems.

1:13:58 For all kinds of reasons I could go into,

1:14:00 the transition is already underway,

1:14:02 but so far it's driven

1:14:04 more by non-market policies,

1:14:07 and that is proportionately changing

1:14:10 as these mature

1:14:12 and driven more by

1:14:14 the centers of innovation.

1:14:15 That also has to change.

1:14:17 I think one still needs elements of those three pillars of policies that I flagged,

1:14:22 and we are talking about effectively industrial development strategies

1:14:26 that help to shift development pathways.

1:14:29 Um,

1:14:30 we do see growing social in social and innovation and engagement,

1:14:36 um.

1:14:38 There's a long way to go,

1:14:40 but I hope I've given you some signs of hope,

1:14:43 er,

1:14:43 as well as acknowledging we face massive challenges.

1:14:47 Thank you very much.

1:14:49 Many thanks Professor Grubb.

1:14:51 Excellent presentation.

1:14:53 Um,

1:14:54 we are over time,

1:14:55 but I would like to give the opportunity,

1:14:58 uh,

1:14:59 to the ones in the audience.

1:15:00 I don't know if we have questions online,

1:15:02 but

1:15:03 to give the opportunity to one or two questions,

1:15:06 uh,

1:15:06 before we go,

1:15:07 uh,

1:15:07 for the coffee break.

1:15:09 Um,

1:15:10 if you have a question,

1:15:12 please come to the mic,

1:15:14 be succinct,

1:15:15 and,

1:15:15 uh,

1:15:16 indicate who are,

1:15:17 uh,

1:15:18 directing the question to.

1:15:25 If there are no questions,

1:15:26 we can go for coffee right away.

1:15:31 You have a great can you please come to the mic?

1:15:37 Good

1:15:37 morning.

1:15:40 I was impressed by the presentation from.

1:15:44 The gentleman who presented on Korea.

1:15:47 Growth in industry,

1:15:48 but

1:15:49 what I did not

1:15:50 get from that presentation was the state of the political economy,

1:15:54 the culture,

1:15:54 and the dynamism,

1:15:55 because

1:15:57 when Oadr read about

1:15:58 the miraculous growth of South Korea,

1:16:01 they link it to the leadership at the time.

1:16:04 So I would have expected that

1:16:06 although light would have been thrown on a level of leadership,

1:16:10 commitment and trust

1:16:11 that engender growth during that process.

1:16:13 So can you share some light on that,

1:16:15 please?

1:16:15 Thank you.

1:16:17 Thank you for the question.

1:16:18 Uh,

1:16:18 obviously,

1:16:19 there's a very interesting story behind it.

1:16:21 In the 70s,

1:16:22 80s,

1:16:23 the really successful uh outcome

1:16:25 was

1:16:26 really

1:16:27 driven by close collaboration between

1:16:30 uh the government leadership and also some of the leading businessmen,

1:16:33 the founders of Samsung and Hyundai,

1:16:35 they're very special people.

1:16:38 Then,

1:16:38 as you noted,

1:16:39 the political economy problem.

1:16:41 Uh was formed because even though the government formally stopped

1:16:45 doing industrial policy in the 80s,

1:16:48 there was this ongoing implicit

1:16:50 support

1:16:51 and understanding

1:16:52 going on,

1:16:53 which was still fine,

1:16:54 but once you go to 2000,

1:16:56 that became really serious

1:16:58 problem

1:16:58 because as Wufu was saying,

1:17:00 now they became just too big.

1:17:02 They were so dominant in so many different industries

1:17:05 and then without any

1:17:07 uh explicit

1:17:09 Uh,

1:17:09 regulation

1:17:11 They were able to abuse their market power and enter

1:17:13 into many different industries and deter the entry of the

1:17:16 uh the entrepreneurial activities,

1:17:18 and that is a big part of the slowdown in productivity growth towards the end.

1:17:22 So,

1:17:23 it can,

1:17:23 we cannot really say that it's a good thing or a bad thing.

1:17:26 Sometimes a close relationship and cooperation

1:17:28 can help the whole economy lift the boat.

1:17:32 Everybody's sport,

1:17:33 but then sometimes actually if they start to use it in the wrong way,

1:17:36 then it could be very hurtful for the agri-economy,

1:17:39 just become a rent seeking

1:17:40 activity for some.

1:17:41 It's a very

1:17:42 complicated story.

1:17:46 Thank you.

1:17:54 Uh,

1:17:54 my question's for,

1:17:55 uh,

1:17:55 Professor Friedman regarding your presentation,

1:17:59 um,

1:17:59 as you talked about the

1:18:01 children of an earlier age going to the more elite,

1:18:04 um,

1:18:05 neighborhoods per se,

1:18:06 how would you reflect that on the reverse for

1:18:10 putting like

1:18:11 An institute in a lower income,

1:18:14 um,

1:18:15 neighborhood,

1:18:16 would you have the

1:18:18 same effect or an adverse effect?

1:18:22 So,

1:18:23 um,

1:18:24 are you,

1:18:24 are you asking about,

1:18:25 um,

1:18:27 The particular policies in order to help children rise up or more like

1:18:31 downward mobility as a flip of the side of the coin of the

1:18:35 of upward mobility?

1:18:37 OK.

1:18:38 Um,

1:18:38 yeah,

1:18:38 I think what the literature shows is that,

1:18:42 um,

1:18:44 You need two different things I think in order to give children from,

1:18:49 uh,

1:18:50 lower income or lower socioeconomic backgrounds opportunities.

1:18:54 One is just

1:18:56 the institutional pathways to get them into,

1:19:00 um,

1:19:01 better education,

1:19:03 better health,

1:19:04 um,

1:19:06 the types of things that are gonna propel them along a career.

1:19:09 But then also

1:19:11 what you need is the

1:19:13 social and network structure in order to make this something that they want to do,

1:19:18 right?

1:19:18 So for instance,

1:19:19 you know,

1:19:19 one example,

1:19:21 um.

1:19:22 You know,

1:19:22 in India,

1:19:24 the.

1:19:26 Admissions exams to some of the,

1:19:28 you know,

1:19:28 very elite,

1:19:29 um,

1:19:30 IIT IIM,

1:19:31 those types of places,

1:19:32 right?

1:19:32 There's uh,

1:19:33 I don't wanna like

1:19:34 say that that's the optimal system because I think it creates a lot of,

1:19:37 um,

1:19:38 inefficiencies as well,

1:19:39 but there's a sense that like this is something that everybody wants to do

1:19:42 and no matter what your background is,

1:19:43 there's this test.

1:19:44 All you need to do is do well on the test.

1:19:45 You don't need parental connections.

1:19:46 You don't need anything,

1:19:47 just like

1:19:48 do well on the test.

1:19:49 Um,

1:19:50 and

1:19:51 again whether that's the right system or not,

1:19:52 what that does do is it creates,

1:19:54 uh,

1:19:55 a situation where people from all sorts of different backgrounds

1:19:58 feel like this is they,

1:19:59 they know about this,

1:20:00 this is something that they

1:20:02 can think about doing,

1:20:03 um,

1:20:04 it,

1:20:04 it has the appearance of,

1:20:05 of something that anybody can get into,

1:20:07 um,

1:20:08 there may be children in,

1:20:10 you know,

1:20:10 lots of different

1:20:11 villages and neighborhoods that,

1:20:13 that have access to that,

1:20:14 um,

1:20:15 you need that in addition to the institutions that,

1:20:17 that do the training as well.

1:20:21 Thank you.

1:20:22 Uh,

1:20:23 with this,

1:20:23 I would like to conclude this session.

1:20:25 Uh,

1:20:26 I will not attempt to summarize what we

1:20:28 learned here today was a very rich discussion.

1:20:31 I would like just to say that I took,

1:20:34 uh,

1:20:35 This take away I think it's very important that countries uh adopt both policies

1:20:41 to support competition,

1:20:43 innovation,

1:20:44 and

1:20:44 equal opportunities

1:20:46 in order to achieve sustained growth.

1:20:49 Thank you so much to all the panelists.

1:20:53 And for the very good discussion.

1:20:57 And

1:20:58 we'll see you at the,

1:20:59 at the coffee break.

showAllTimestamps
no
transcript
And after this very inspiring presentation by Intermit, uh, we will start the second session part of the session. Um, my name is Manuella Francisco, and I'm the global director of the Global Practice for macro Trade and Investment. I have with me a very distinguished panel of academics. Let me start by presenting them, uh, from my left to the right. Um, we have Professor Ufuk Ax sorry. Professor Jung Sin Shin, a professor at the Washington University of Saint Louis, we have Professor Ufuk Akish, apologies for my bad pronunciation, Professor of Economics at University of Chicago. We have Professor Michael Grubb, professor of Energy and Climate change at UCL University College of London. And we have Professor right next to me, Professor John Friedman, Chair of Economics department at Brown University. The four papers you hear about today present some of the theoretical and policy underpinnings underpinnings that will help us better understand the growth challenge in middle income economies. We, we will hear a presentation that is structured around the 3 themes of the upcoming world development report that is structured around enterprise, social mobility, and energy transition. We will start with Professor Akishit that will focus on the role of creative destruction by generating economic growth through innovation and entrepreneurship, which will form the general framework of the Schumpeterian growth of the of the World Bank development report. Over to you, Professor Akus. So good morning everyone. Uh thanks to the organizers for this fantastic conference. Uh, it's always difficult to talk after Inarmid, who is a great presenter, and, uh, but the nice thing is that he also made a nice transition to what I'm going to talk about. So, uh, today I will talk about economic growth, but from an angle of creative destruction. The literature on creative destruction has made enormous advances over the past 30 years, I would say, and this, you know, once you start thinking about economic growth through the lens of creative destruction, you suddenly start uncovering many, many facts because now you're able to map the economic growth idea to micro-level data and. The microdata analysis is guided by the Schumpeterian theory, as I'm going to illustrate. It's a, it's a very nice and very informative angle that the literature has developed, and this is also the main thrust of the of the next WDR. And so led by Sean Mikal and Indermit, we have a very large team working on the WDR. A lot of them are here, and I would like to give you a little bit of an idea about what we are working on. Uh, at this new WDR, so first of all, we are taking a comprehensive approach. So we, we try to question everything in an economy and we try to say how can we save the middle income countries that are trapped in the middle income trap? How can we save them out of that trap through technology and as Indermet also showed, as you go through the stages of development, as you get closer to the frontier. Rather than capital accumulation, what starts to matter is technology development and productivity. And of course when it comes to productivity and technology, the question is how can we introduce new technologies in an economy. Just focusing on firms would be a mistake because of course firms are some abstract entities, but in those entities there are individuals working and this. The whole approach, the comprehensive approach is quite crucial. So the whole reasoning starts with a pool of individuals in any society, which we are illustrating here at the bottom right. I think the pointer is not working, but so at the bottom right, and of course individuals in any society are ranked according to their talents at the time of the birth. Of course children are born with heterogeneous talents. And societies with their own systems somehow let some of those kids rise up in the society. It could be based on talent or it could be based on some ethnic background or family characteristics. It depends on the system that exists in the society. So then, of course, in an ideal scenario. We would have the most talented kids sent to schools when it comes to innovation-led growth or technology-led growth here. The education starts to become more technical and so then whom are we educating in our technical schools becomes an important question. Who are being sent in the society into those technical schools? So then education becomes a different notion rather than looking at the average years of schooling, now we are trying to look at these elites selected in the society that are sent into those technical schools. Who are they? Are they going there based on their ability or based on some other characteristics? So then once we send the kids into schools and train them, some of them stay in the country and some of them leave. So suddenly brain drain becomes also part of the equation. How do we treat brain drain? Should we turn our back and just complain and whine about the brain drain, or should we somehow embrace it? Indeed today Marta is going to talk about a fantastic paper of hers in more detail about this. But brain drain is also part of the equation. So once we have people trained in the society, some sent abroad and some coming back, now we have the workforce in the society and firms are hiring those trained workforce and start competing among themselves. So what do they invest in? Are they investing? In new technologies or if so through what incentives or through what market mechanisms and then the firms will create some value added and as a result we will have this cycle. Now what is the observation about middle income countries? The observation is that through this pipe the water is coming very, very weak at the end. So clearly something is blocked in the system, and the question is where. It can be anywhere, right? That's why rather than just focusing on one side of the economy, we are just taking this comprehensive approach and try to identify the problems and potential solutions and policy recommendations through this approach. So why the Schumpeterian model? Why the Schumpeterian approach? Of course when it comes to economic growth, there are many, many theories. Indeed we are not denying any other theories. Of course the institutions are important. Capital accumulation through the solar model is important, but now we are also trying to bring in the technology into the equation. When it comes to technology and innovation, there are competing theories. One of them, of course, is the Romer model, which is the Nobel winning model by Paul Romer. But when you think about the Roman model, of course it was a brilliant model, but the way that model operated is that there is this continuum of varieties, and through each innovation we are adding new varieties to the economy, and we just add more and more varieties, and whenever somebody introduces a new variety, that person becomes the Permanent monopolies in that market and produces forever. The same person keeps producing forever and the Roman model. Of course it's a brilliant model endogenis technological change. It doesn't talk about exit and turnover, which is exactly the problem, as I'm going to illustrate in a second, especially in middle income countries, because once you start thinking about turnover and exit, there you suddenly realize there are many, many frictions that are embedded into the economy and the data and that you can only read through the Schumpeterian theory. So what does the Schumpeterian theory do? So there is a firm. Rather than introducing a variety, it introduces a better version of an existing technology. So for instance, firm 1 is the current producer, but then Firm 2 comes in with a better technology. Firm 1 is out and Firm 2 is in. And this way through this turnover in the economy. Firms are just replacing each other and they are introducing new versions of technologies. Of course this was the first version of the Schumpeterian model, but then Since 1992, after Agnonhovi, the literature has evolved immensely. Now creative destruction is not only taking place in our existing theories between entrants and incumbents, but also among incumbents. Indeed, when we look at the data, 50 to 75% of the productivity growth is coming from successful factor reallocation among incumbents, and we cannot avoid this, obviously, and that's why the type of framework that we are. Considering is way beyond Ajonhoitz initial framework, I don't have time unfortunately, so that's why I'm not able to go into those details. But basically here we have a creative destruction model where not only entrants and incumbents compete among themselves, but also incumbents also compete among themselves. So once you start with this logic, suddenly you realize that we might be, we might be living in many different scenarios in different industries, for instance. Entrants through their new technologies are trying to push the frontier forward. What can happen is that of course incumbents will somehow join join the right here. So either incumbents can also react positively to incumbent pressure. In this case there will be the pro-competitive effect. Indeed there were some interesting papers that argued that China's pressure on the US market had some pro-competitive effect or some paper. Also argued that it had a negative effect and then some papers argued that indeed in the continental Europe it was a positive, etc. So as you can see, it's a very empirical question, but we need to be flexible for these types of different hypotheses so that we can read the data better. Alternatively, of course, incumbents can can resist the change, and they are not always going to introduce a better quality when an entrant comes in with a better technology. I might just say, Well, I lost the game. Let me just leave the market. I might try to rely on some other strategies. And indeed, in many, many middle income countries this is a major problem as we see in the data. Incumbents can also collude. It's not only one incumbent that the entrants might be fighting against. It could be that the incumbents might be colluding among themselves. And once you start reading or approaching the data or the whole growth story through this logic, suddenly it opens up the avenue into many interesting questions. Of course we are aware of the fact that innovation might be too early or too soon for some countries or for some sectors. And technology can be upgraded not only through innovation, but sometimes countries can also utilize their advantage of backwardness, and we are not the first ones to say this. Indeed, Gar Shankron in his 1960s book had already made this point very strongly, forcefully. If a country is far from the frontier, Rather than forming these really expensive R&D labs and try to innovate and compete with the frontier countries, they might as well open up to the world and get connected to the outside world and try to bring those technologies that have been invented anywhere around the world to the country first and start learning from it and building from it. Do we see an example of this? Yes, I guess Jungs will talk more about this, but The way we are thinking about the economic growth process is in different stages. So on the x axis you can see the proximity to the frontier. As you go to the right, a country gets closer to the world technology frontier. As we go to the left, it gets far away from the frontier. And if a country is very far from the frontier. You know all they might need at that point is having some more capital and some more resources to build more pipes, more roads, or more bridges. But as we are getting closer to the frontier now, rather than capital accumulation, now technology has to be evolving, and again, it's the basic intuition of the solar model. The Solar model says you can grow just so much with the catch up growth, but once you reach your steady state from that point on, you need to upgrade your technology. How are you going to do it? What we argue here is that if a country is a middle income country, there's an opportunity to bring technologies from outside. And how are you going to do it? Well, now the trade suddenly has a different meaning. Rather than being open means that rather than just trying to sell your goods, it might mean opportunities to learn from the world. And also through diaspora abroad, a country can make use of the knowledge at the world technology frontier and learn from it. And of course as the country starts building on what it learned, then it can start innovating on itself once it's ready to do the innovation. Do we have an example? Yes, Korea's story, of course Korea's story is very, very involved, but one of the things that was shown by Yong Honghim and his co-author in his job market paper this year, they showed they digitized the Korean archives and looked at the government contracts, and what they showed is that The Korean government in the beginning, in the early 70s, when they were making the huge leapfrog, the government was very heavily subsidizing technology adoption from Japan. So the country was of course was not fully ready to do its own innovation, but there were a lot of technologies that were licensed from Japan, and then they learned from those experiences and build on it. So this is a very important observation, I think. Rather than putting an unreasonable target to middle income countries, we need to put reasonable targets first so that they can use it as a stepping stone and then jump. So I highlighted the Schumpeterian dynamics so much, but what do we see once we start thinking about the Schumpeterian dynamics? What do we learn from the microdata as something new? The basic premise of the Schumpeterian theory says that entrants will be more productive and replace the less productive incumbents. Let's look at some middle income countries that's totally randomly selected. So here we are comparing exiters and entrance productivity in Turkey, and the blue line is showing you the productivity of the exiters and the red line is showing you the productivity of the entrants. So clearly the basic premise of the Schumpeterian theory is not working in the case of the Turkish manufacturing sector. So this is something that we have to fix as part of the story. So rather than creating destruction, maybe this is a destructive creation in that sense, because it's lowering the productivity. Indeed, when you look at the total factor productivity evolution in Turkey since the 1970s, there's a steady decline. So let's switch gears and let's look at another country that did well for a while and for quite some time now it's doing quite poorly and here in this study what we did is we looked at the growth strategies of the market leaders. As the firm is evolving and becoming the dominant market leader, does it become more innovative, or does it rely on other alternative strategies? Number 1 here means market leader. Number 20 means that's the 20th largest firm. So as you go to the right, the firm size gets smaller. As you go to the left, firm size increases. What we see is very interesting as firms are becoming the dominant market leader. Innovativeness goes down. The red dots go down, but the number of politicians that they hire relative to the firm size increases, so they are getting more tangled to the political system rather than trying to push the frontier. Again, it's the simple Schumpeterian logic. Once I'm in the market, I need to resist, and that's exactly what we observe here. When we look at the Indian case again, we see a lack of creative destruction, and here on the top left figure you see the life cycle of firms in the US and India. In the US firms are growing very rapidly. There's an up or out type of competitive dynamics when it comes to India. The average firm size conditional on survival is just flat. It means that the small firms, they are not only growing, they are not getting kicked out either. And as a result, the bottom right figure here is showing. The fraction of small firms in the economy as the cohortages. As the cohortages, the dashed line shows that in the US, small firms are eliminated very, very rapidly. In India, small firms are not being eliminated, and as a result, the fraction of small firms remains. So this means that the fact that we are observing so many small firms in the middle. Countries could be a symptom of not only lack of financing for them, but it could be lack of competition by larger firms. Maybe the problem is in the mid-sized or large sized firms, and maybe they are not growing sufficiently to drive out the small firms so that they can create high quality, high paying jobs, so that all these trying to be entrepreneurs. Can get reallocated and earn decent wages in these growing firms, and this is what we need to consider. Indeed, of course, when it comes to middle income countries, informality is a major concern and here once you start thinking through the Schumpeterian lens, informality is not a cause but a result, and informality could be a result of Lack of competition in the market. And here, for instance, again when we look at some regulations and what we see is that regulations that exist in middle income countries can even encourage more informality. But if we could normally push the firms to larger sizes, we would automatically fight against informality. So I'm running out of time. Just a few issues on the human capital side. The report will also focus on human capital. And here we look at the talent allocation. Innovation and technological upgrade cannot happen just by itself or firms' investment. Firms should also have good engineers and inventors. Without having good players, a team cannot win the championship. With that logic, we have to also have a strong workforce. And of course if in a society there is discrimination or financial frictions and not having equal opportunities for people. We might be leaving out important talent out, so this is, this is an important consideration in the report. And then we look at the data again, it's not the average years of schooling that matters for technological progress. It's technical training, technical education, even in the historical US, where we should draw a lot of lessons from, in the historical US we see that college education was extremely important for innovativeness. And not only having college education, also having liquid secondary market for technologies, but today in many of the middle income countries there's barely college graduates. There's barely any regulation for how to protect intellectual property. So we need to talk about these aspects. And in successful countries like Denmark today we see that PhD is the strongest predictor of innovation. It's that technical. It's at that level, basically. And of course when it comes to human capital we don't necessarily need to grow them within the country only. We can also tap on the global talents. China, for instance, 80% of Chinese students go back to their country after studying, and can you imagine what Big spillover there is in the case of China. The number of returnees to China exceeds 1 million a year. And that way, of course, we are talking about a Chinese miracle, but we have to also think about these spillovers, global spillovers that are taking place and what happens in middle income countries right now. Middle income countries, when you look at for instance, who is leaving a country and who is coming back, it's typically the most productive individuals or scientists who are leaving the country. And the least productive ones are coming back because of the bad policies, but rather than trying to force those immigrants to come back, one thing one can do is one can try to connect with them and try to learn from them. Indeed, Martha Prato will talk in great detail about this rather than turning our back to migrants, to people abroad, we can try to build a bridge from them and learn from them to utilize the advantage of backwardness. And finally, there will be also the energy part in the report, but just to give you an idea about how we are thinking about this, you know, we can produce dirty technology in the Schumpeterian sense or non-green technology, let me put it that way. But there can be also green alternatives, green technologies to use. What's the comparison here? The key critical thing here is that today we are living in a world where there is a race between green technologies and non-green technologies, and green technologies by construction are less advanced. We need to reconsider the government policies because some government. Policies that might have well intentions might be slowing down this transition. Indeed we do see some evidence on this. For instance, the subsidies that are being provided to fossil fuels in many countries are just going against the transition, and at best it will just slow down the transition. So that's why thinking about this creative destruction and clearly. Energy transition is the poster child of creative destruction. We are trying to replace an old energy or technology with a green one. That's very much in the realm of Schumpeterian theory, and these are some of the elements that we are considering in the new. Sorry for the limited time, but hopefully it gave you a good idea about what we are working on. Thank you. Many thanks Professor Akhijit. So we will now hear from Professor Friedman. Uh, his presentation focused on intergenerational mobility around the world. His work focused on social mobility and the importance of selecting, developing, and enabling talents, talented individuals. Over to you. Well, thank you very much for the uh invitation to speak here. Um, I should note at the start that I'm not a development economist. Uh, I do not specialize in economic growth and so, uh, I think everyone in this room is gonna have a better sense of the exact types of policies, the exact types of analysis like what Ufuk was talking about that are really going to, uh, help, uh, move development forward across the world. What I wanna talk about is intergenerational mobility. Now intergenerational mobility, uh, featured in that, uh, extremely nice schematic that Ufuk started with as part of economic growth directly in terms of figuring out how to get, uh, the most diverse broad set of talent into the position of being innovators or high productivity employees in a given country through education and I think that's a very important part of what intergenerational mobility does. I also think that intergenerational mobility is uh a broader part of the institutions of a country which if not maintained appropriately can lead to real problems um and I think there are many countries across the world today where the lack of intergenerational mobility or perceived lack of intergenerational mobility, the sense that too many people are born into situations from which they just don't have a shot at moving up in. Uh, the income ladder or, or in the social ladder where parents feel like their kids do not have the opportunity to have a better life than they do, that is not only an economic problem that becomes a social problem that becomes a political problem that can really uh cut the legs out from under what might otherwise be a high growth, high productivity regime. So let me start with just a couple of facts from uh some great data that the World Bank has produced, uh, right, people always think of the American dream as, you know, the possibility that, you know, uh, people, especially immigrants could come to this country, um, in the early 20th century and kind of no matter what your background there was the opportunity for success, and I think. You know, most people know at this point that, you know, that is not something that's really borne out in the day today. Uh, here, uh, the particular statistic being calculated is the probability that a child born into parents who are in the bottom half of the income distribution move into a position by the time that they themselves reach the labor market that's in the top quartile and so, you know, obviously each of those are defined, um, locally. And what you see is that um you know the US is very much in the middle of the distribution of um middle income countries. There are some countries like Brazil and Pakistan who have lower rates of upper mobility, but there are also many countries here like Jordan and Thailand which have uh much higher rates of upper mobility than the United States. Now, these statistics though are still, I think really only glossing the surface of what we can learn about how to increase intergenerational mobility. And so what I wanna talk about today is, uh, what I think are a very exciting set of new tools that have come out of work that, um, my co-authors and I and and many others have done, um, across the world. Um, these are techniques I think that, uh, mostly were pioneered in developed countries, but what's been very exciting is to see a lot of new work applying them to many, especially, uh, middle income, but, but even, um, countries at the, um. Very lowest levels of the development ladder. Um, and I think that it's extremely important, right, especially in this middle income range, to think about, uh, alongside growth and kind of elimination of, of abject poverty, to think about maintaining or increasing or not losing ground on social mobility, um, as we go along. And so, um, over the next 10 minutes I'm just gonna kind of give you a little bit of a, a tour. I'm gonna hit kind of 3 highlights of techniques that have been, um, pioneered that, that I, I think might be helpful, uh, to this end. So first, let me just start with this map of the United States. This is a heat map where in each city area, essentially, which are each of the colored polygons, we're measuring the average incomes in adulthood for children who grew up to families at the 25th percentile, kind of the. Average below average family, uh, and the heat map colors in blue children who are doing very well in red children who are, are not doing as well and what you see across the United States is just there's an enormous amount of variation. Children born in Charlotte, uh, down in the southeast of this country. Um, are only making $26,000 on average as adults. Children born in Dubuque, um, which is part of that very blue region in the upper, uh, Midwest of this country, are earning nearly twice that amount. Now, of course you could look at this map and say fine. The United States is the size of a continent, you know, of course, intergenerational mobility is going to vary at this scale. What I think might be surprising is that even when you zero into a particular city to make it really local, you still see this range of variation in mobility. So this is the same map that I showed you at the national level, but now focusing only on the New York metropolitan area. We've now colored things at the census tract level. So these are essentially neighborhoods. There are about 4000 households per census tract. The colors are essentially the same. Right, so that same variation that you saw between the southeast of the United States and the Upper Midwest, so those are areas that are, you know, still within the same country, but, you know, thousands of miles apart. Here you see that same variation between neighborhoods that are in many cases, literally across the street from one another. Um, and in particular I'm, I'm gonna come back to this example later on, um, you know, you see down in, um, in Brooklyn, right, there's some areas where there's, I'm gonna kind of point it out here, the pointer is not working, but kind of right in, in this region here, um, there's these little blue islands in what is otherwise a very low mobility neighborhood. Uh, the blue areas are, uh, what's called the Nehemiah houses, and they were developed in the 1980s on literally a plot of vacant land. Uh, by a local organization as an attempt to create kind of a new society moving away from some very low mobility areas, uh, that were called the Van Dyke houses. So I think the first lesson that I want people to take away is that mobility is going to vary not only at a national level but in an incredibly local level. And so kind of that's the level of granularity that one has to take when thinking about this problem. Now of course the examples that I'm showing you here from my work are in the United States, but um over the past 10 years we've seen very similar analyses of many countries around the world. So here are 4 examples from developed countries in Europe, um, which show kind of very, very similar patterns. More relevant I think to this audience here are examples from. Um, Asher et al. on the right and from Allesina et al. on the left, um, of course one has to use data creatively in order to generate these types of very local maps. Uh, on the right they're using education as a, as a measure of socioeconomic status. On the left they're using literacy, but, uh, I think the big picture here is that not only is intergenerational mobility varying at a national level, it's varying at uh a subnational and even very local level, uh, within these countries. So second, um, let me just come back, um, to this map here. Uh, this looks at intergenerational mobility overall pooling all different types of, uh, individuals and what analysis in, in this country has shown, this is analysis by my co-authors, um, right. Chetty and Nathan Hendren along with uh Maggie Jones and Sonia Porter, is that this is sharply divided the United States between individuals of different races and so here I'm showing a similar map with black men on the left and white men on the right. Um, and looking at these maps, right, it might look like we're using two different color schemes for the two different maps. Uh, that's actually not the case. It's the same color scheme on the bottom. It's just that upward mobility for, um, black individuals, black men born into poverty on the left is, is basically everywhere less than, uh, the, the same outcomes for, uh, white men born into families at the same level of poverty on the right. And so I, I think it's not just that overall mobility, it's thinking about um how do opportunities exist for children from these uh other dimensions in society, um. Similarly, uh, these, uh, types of findings have come out in work in developing countries. So this is, um, uh, a scatter plot that I made based on results from that same Allesina et al. um, analysis where they've calculated mobility separately for, uh, Christian individuals and Muslim individuals in many different countries in Africa. And so we have uh the upward mobility in each country for Christians on the horizontal axis, for Muslims on the vertical axis, um, and what you see, uh, relative to the 45 degree line is that almost all countries are below and to the right. And so what that's saying is that to a varying degree, uh, you see more upward mobility for Christian individuals than Muslim individuals. So again, you know, that's the type of uh equality of opportunity that I think matters quite a bit, not only from an economic perspective, but from a broader perspective as well. Third, um, we've seen quite a bit of evidence that these types of local variation in upward mobility is not just about sorting of particular people to particular places, but rather it's a causal effect in large part of the. Itself. So how do we know that? Um, we've learned that from an exposure design. And so the idea here is that one's, uh, the, the impact of one's neighborhood varies by how long one's lived in it, right? It's really an exposure-based model. And so let me come back to those two neighborhoods in, in Brooklyn that I started with the Nehemiah houses, which are a very high upper mobility neighborhood. The Van Dyke houses, which are a very low mobility neighborhood, and essentially just across the street from the Nehemiah houses. The dash line at the top is our outcomes for children who grew up their entire lives in the Nehemiah houses, which is much higher than the outcome, uh, the dashed line at the bottom for children who spent their entire lives in the Van Dyke houses. So the idea of this causal design is to separate out, you know, maybe those kids living in those two different, um, uh, sets of houses are very different for other reasons. And so instead of just comparing kids who grew up entirely in one versus another, we're going to look at those who moved from one to another at a particular point in childhood. And so first, let me think about children who were born in the Van Dyke houses and then moved at a very early age, let's say age 2, to the Nehemiah houses. Those children have outcomes that look very similar, uh, slightly below, but very similar to those children who grew up, their entire lives in the Nehemiah houses. And then we're just gonna repeat this, uh, children who moved from the Van Dyke houses to the MEA houses at, at age 3, age 4, and so on. And what you see is that the more time you spend in the Van Dyke houses before moving to the Nehemiah houses, that is the less opportunity that you have to be exposed to that, uh, higher mobility environment, whatever that is in the Nehemiah houses, the worse your outcomes get by. The time you've not moved until you're in your early twenties, moving at that age to the my houses seems to really have no effect. Um, and then in particular, one can then look after age 24. We're measuring incomes at age 24, so this is essentially a placebo check to see, well, maybe the people who move at these different ages are different, and here you basically find no effect. Um, and now this, this mover's design, which has this kind of. Characteristic hockey stick, uh, for those of you who are from North America, um, type of pattern, uh, this too has shown up in a remarkably broad set of circumstances across the world, you know, here's something similar not only in Denmark and Australia on the top, but you see very similar exposure patterns, uh, in Africa from the Allesina work and in uh Brazil, um, on the bottom. And again you see very similar things where it's not just that exposure to these local places is incredibly important, it seems to be exposure during childhood is the most important thing. And of course the exact definitions of what that means are going to differ between places and and how one is um measuring upward mobility, um, but this really does seem to be a very robust fact in many places across the world. Now, the fact that these local exposure to to place are causal, then of course just raises the question of, well, what is it about these local places that that is causal, and here I think, um, you know, again, the answers are going to differ. What do we get from looking in the US at these types of things? It's not just raw growth that seems to matter actually. This slide might be helpful. This is a scatter plot of average growth across US cities on upward mobility in the 30 largest cities. There's just no correlation at all. So just broad-based growth does not necessarily help. What does seem to matter are these features that are much more about not only human capital development and exposure, you know, especially exposure to abject poverty is not a good thing. But you see that the social circumstances and the networks seem to matter quite a bit as well. And so in particular you see the third one here, greater social capital. What seems to be going on here is exposure not only to adults who can not only inspire. you to give you information about the pathways that are available that can help make you make the choices that might maximize your potential as a child, but also exposure to other peers of yours that matters tremendously as well. And so here's a map of the United States from um. Some of Raj Shetty's recent work on social capital, this looks enormously like the picture of upward mobility that I showed you before, um, and so my sense is that these types of uh social exposures are going to matter, uh, in addition to all these more directly human capital focused activities, you know, here's a similar thing looking across regions. From the, the Allesina work, um, again, what pops out here is not only some of those raw economic factors like distance to a railroad, um, but it's things like distance to a Protestant mission which is gonna, gonna be a stand-in for a lot of the cultural differences that, you know, exist, uh, across these institutional differences that exist across these, these places. So, um, kind of three takeaways as, as my time is ending. Um, this is the first talk I've ever given where like the luge, my time is being measured down to 10,000 of a second. It's extremely precise. Um, luge is the only Olympic event where time is measured in the 10,000 of a second. Um, so 3 takeaways. First, it's the local childhood environment, I think that really plays the central role for developing prospects for upward mobility, and it's environment again, local, I think is the key word here. It's really about kind of who you're interacting with on a daily basis. You shouldn't even think about this is at the level of a city. It's more like the, the neighborhood that you're living in. Second. What I think this shows is that one, much like Ufuk was talking about, you can use these large scale observational data to make, uh, to, to learn something about what types of interventions might be most valuable and where, uh, right, this is, you know, not going to be as precise or as direct as an RCT for a particular intervention, but this is going to give you perhaps a better sense of like what bucket to look in. And third, I think it's again we have to move beyond these, uh, you know, very traditional models of human capital development, uh, right? I think it's really about a lot of these broader sociological forces, um, that have a lot to do with economic mobility and inequality as well. Thank you very much. Many thanks Professor Friedman for the very interesting and sobering presentation. We'll hear next from Professor Shin. Uh, Professor Shin will present, uh, a presentation focused on growth through creative destruction from the point of view of firms, uh, focusing on the South Korean growth miracle. Over to you. Thank you. So South Korea is one of the well-known economic success stories. It went from poor to rich in one person's lifetime, actually more like 30 years. And a lot of research has been done, a lot of theories have been expounded on how they made this miracle. What has not been possible until now was an analysis of what is happening at the micro level, at the plant level. Is there a systematic pattern that accompany this macroeconomic transformation at the micro level? So we're able to do that. So this is the motivating figure. The bottom line is the GDP per capita at domestic constant prices. So if you look at it, two things, it's a low scale. From say 1970 to 2000 in 30 years, the GDP per capita uh goes up by a factor of 10, from 700 to 7000. This is remarkableable growth rate. The other thing you notice is that the growth rate kind of slows down over each decade. In the 70s and eighties, it's about 9% to 8% per year growth and 202,010 is 4% and 3% per year and Koreans call that a recession. And then, the blue line is manufacturing value-added per worker, so that's your measure of productivity of the manufacturing sector. Of course, it's higher than the GDP because manufacturing tend to be more productive than service sector. Uh broad patterns are consistent. The subtle differences that manufacturing grew really productive in the 90s, in the value added per worker sense, despite the '97 East Asian financial crisis, and then it really slowed down after 2010. So, we're using newly digitized data, manufacturing census data. We already had data from 1982 onwards, but what is new is the previous period, 196, 1017, 1981. So, one limitation is that the panel dimension of the data is available only from 1982, and then we're gonna actually spend some time in the second half of my presentation about that part. So first, we start with the first half, the cross-section data only. It starts from 1967. Since the cutoff of employment is 5, we're only looking at plants with at least 5 employees. Until 2007, after that, it becomes at least 10 employees. So, you look at the right figure, the average size of the manufacturing plants rises dramatically in the 1970s and collapses dramatically in the 1980s. There's some difference. 1970s is the period of heavy and chemical industry promotion, so a lot of this rise is compositional. So you're moving away from light manufacturing to heavy and chemical industry and chemical industry obviously has much more employees per uh plant. And then the collapse in the 1980s, actually, that's not compositional. So it is happening in basically all industries within having chemical industry, within manufacturing and other industries. So there's some differences in how it goes up and down. There are different ways to look at size. This is what fraction of employment in total employment is in plants with at least 2 250 or 500 employees. Same pattern goes up and down. And then this is log log plot. So some of you are very familiar with this plot. It basically shows the entire distribution, so it's sort of a reverse of the cumulative distribution function. So at any point on the x axis, pick two, that's the logo where it stands, so that's 100, and you ask of all the plants, what fraction of plants are actually larger than 100 employees. And it's, it's, it's moderately declining. So solid line in the middle you see is 1967, which is showing every decade. And then you see that in 1977 it shifts out quite a bit. It's almost like a for short of stochastic dominance. So it's basically what it is is that the entire distribution shifts to the right, so all the plants are basically much larger in 1977 and 196967, but that was the peak of the plant size distribution, and then you see 1987 and 1997, 2007, actually you see these things coming back, so things are getting smaller. So it's very consistent throughout um The entire distribution, that's just something at the top. So what, at this point you take a stock and compare the picture of manufacturing value added per worker and the size distribution changes. What you see is that there's no clear pattern either in levels because value per worker continues to grow, but the size ra goes up and down. Or in terms of growth rate. So we had a fast growth in the 70s and 80s and 90s, but then sometimes science division goes up and comes down, so there's no simpler relationship. So it prompted us to look at it that's over time, we don't see any clear pattern between size and value per worker or the growth rate of that. So we wanted to see whether actually there's any pattern across countries. So more research is definitely needed. uh we just went to OECD database and grabbed whatever we could. On the left panel, these are countries including South Korea that report in terms of plants rather than farms or enterprises. So what you see is that Um, you're attempting to make, uh, oh, there's a downward sloping pattern here. First, only 6 countries here, and also you know that there are many other countries, for example, India, who talked about many other countries with a really small employment and very low value added per worker, so there's no simple relationship. The right hand side is from same OECD data uh looking at countries that report in terms of enterprises and firms, you see that there's no clear pattern here either, at least among these countries in terms of the size of a farm measured by average employment per farm and the value added per worker of the sector. This is a manufacturing. So there doesn't seem to be a very simple relationship between size and productivity or growth. OK. The other thing we can do with this static data, cross section data is do simple calculation of misallocation, sharing clinotype. Um you can see I plot, but let's talk about the right one. So if you look at it really carefully, oh, the data from 69 to 77 is missing because those years we don't have a capital at the plant level, so we cannot do the calculation. So what you see is actually there seems to be a downward slope in degree of missile location. So location efficiency is actually improving between 1968 and up to mid 1980s, but then it gets worse in the 1980s and later part of the 1980s and 191990s, and it gets really bad in the 2000s. I think the last phase is more about the issue of missile specification. I will not read too much into that. But to the extent that we think this is a pattern misallocation, then you again don't see a really clear pattern between the degree of misallocation and either level or growth of the value added per worker, because again, value per worker in manufacturing is growing constantly and then it was growing fast in the 70s when misallocation is coming down, but then it's also growing fast when misallocation is getting worse. So there doesn't seem to be a very clean pattern here. This is just another way to look at, now it's a popular way to look at degree of misallocation. It's a coefficient of regress if you know of TFPR and TFPQ at the plant level and then aggregating it. Same pattern misallocation comes down and then it takes off in the late 90s and 2000s. So, overall, in the 70s, I'm just kind of adding a picture here from a paper of mine. Overall, the, in the 70s, we just saw that actually misallocation was coming down in in Korea, but there's something interesting pattern. So this is active heavy chemical industry promotion period. So what we see is that if you actually look at the industries and regions that were targeted by this government. The policy, there was a lot of increase in misallocation within those industries and regions compared to industries and regions that were not targeted with the policy. So the aggregate pattern belies some of the more interesting features at the more desaggregate level patterns. Now, in the remaining few minutes, I'll talk about the dynamic part. This is where really interesting things happen and it connects very well with what Wolf was saying in terms of the trumpeterian growth models. So, the first set of pictures I'll show you is this charming picture picture. So this is the Davis Hettinger shoeching picture. So this is a histogram. So what is a histogram? So it's basically showing the employment share of some beans, and what are the beans? The bean is basically the growth rate of a plants. So you see the number around 0, so the employment cluster around 0. So those are workers working in plants, that's not really. Growing or shrinking. Around 11 means this plant is growing really fast, uh, -1 means it's shrinking really fast. 2 is entry, and -2 is exit. So it kind of shows the improvement change. It's a small change is important, plant size plant size is important too. It's more big plant size change is important. And in the 80s, you see the importers entry really big time, right? So this is the period where the employment average size measured by employment of plants was collapsing. And that's because there's huge entry and entrance obviously is smaller than incumbents. So that's when in all sectors of the economy there's a huge spike of entry and the average size is coming down. But at the same time you see actually there's a lot of firms adjusting upward, a lot of growing firms. But then the next period, 87 to 92 or 92 to 97, slight different pattern. You see that things are kind of shifted to the left, so entry is not as important. Now exit actually plays a more important role, like a minus 2, and then the whole picture is to the left, meaning that you see a lot of plants downsizing actually. And I think this may be the secret to what was um why it was possible for South Korea to sustain the economic growth so long that In the 70s, the government had a really heavy chemical industry policy and when they stopped it in the 80s, you have this fresh batch of new firms coming in and that pushed a lot of bad firms, big firms to actually uh downsize and exit in the 80s, the later part of the 80s and early 90s. And I think that's really important for understanding how the productivity growth measured by valuable worker grew so fast in the 90s. So this creative destruction is playing an important role. And you look at the last period, 2008 to 2013 to 2008, it's very quiet here. What you can see is that while entry and exit is not playing its important role, and also there's a lot of implement here around zero, meaning that farms are not really growing or shrinking, so it's very quiet. So this is a reduction in dynamism or reduction in churning. So the prior set of research we just showing you, well this is how plant size is changing, measured by employment. But for productivity, we, we really care about the correlation, meaning you want the productive farms to grow and unproductive farms to shrink. So I'm showing you some of the regression results following uh Dekker Her Jeremy Miranda approach. And what I'm showing you is whether farms grow in terms of employment or capital or investment when they have a good productivity shock. And the first row numbers are positive and significant, so that's good news. Yes, so farms are responding to productivity shocks, so uh that's good for economic growth. But what's really interesting is the second panel when we do things decade by decade. So what you see here is that especially the employment growth part, 1990s, that's when the productivity is growing the most. That's when farms are most sensitive in terms of product of shock. And then the number gets really small over time. It's been 2010 it's really quiet, so even if you have a good pery shock, you're not going to raise your employment and capital very much. So there's the same picture, the reduction in dynamism. Firms are not really responding to perry shocks, and I think that really. Responds very well with what we saw in the overall pattern that things really slowed down towards the end of the period 200, 2010 went from stopping uh dynamic in terms of either churning or how responsive they are to productivity shocks. So I'm about to run out of time, so let me wrap up. So, as you saw, at the micro level, there was no really simple systematic pattern of correlation between what's happening at the micro level and macro transformation of the Korean economy because uh the economy continues to grow, but the size decision could move upward and downward, and then with misallocation measures, there's no simple pattern there either. What's really promising was the Uh, the later part when the dynamism is slowing down in terms of churning or responsivity shocks, that's when the aggregate economy really slows down in terms of growth. And I think that's really relevant for thinking about middle income countries. So this is a very promising line of work. The problem is that it presents a very high hurdle for data requirements, right? So now you need a panel dimension and you need to cover enough variables to compute the productivity reform and input uses and everything. So it may not be the best research program for low income countries that do not have the right data infrastructure, but for middle income countries I think it's a promising avenue. Uh because when productivity slows down, you don't know where to look at, but when you have this micro evidence, then you can be guided by economic theory and look at um the sources of these barriers and frictions that's slowing farms down and reallocation down and that will lead us to uh policy prescription. Thank you. Many thanks Professor Shin. It's always a pleasure to hear about the, the Korean growth miracle. Uh, finally, we will hear from Professor Grubb. Professor Greb will focus on the very important topic of economics and policies of energy transition over to you. Well thank you very much indeed for the invitation and this session. It's been absolutely fascinating. Uh, the only problem is it's left me with about 76 or 7 minutes to coffee. Uh, I'll, I'll do my best. Um, so I'm going to touch on, uh, climate change and energy transition, um, just to a word of introduction that, uh, first, uh, I'm also not a development economist. I'll be covering more of an energy economist and special interest in innovation. I will give a brief background, uh, possibly about half the time on the context of climate change scene setting because obviously there's a broader conference, and then move on to particularly the innovation dynamics that we've seen and some implications. That's partly because climate change issues are so often couched in. Everything's going horribly wrong and it's a very doom and gloom scenario. I do want to inject a perception that it's more interesting than that, not least when we reflect on two things, which is that economic development, two of the fundamental factors that we've touched on, are innovation and cheap primary energy. The question has always been, is there potentially a clash with climate change? I will be drawing significantly upon aspects of the IPCC report. Many of you, I think, will know, a lot. The broad background IPCC is a massive international process. These assessments roll around every 6 or 7 years, uh, and we had the 6th assessment, uh, last published last year. So I'll draw quite a bit on that. I will start with the science and risk dimensions of it. On the left hand side, we have one of those famous graphs of scenarios of global temperature suggesting a very wide range of possibilities. So apparently a lot of agency in terms of what happens to the temperature this summer, this century. On the right hand side you have a fairly well known IPCC diagram called the five reasons for Concern, which range from loss of specific ecosystems on the left hand side through on the right hand to globally substantial human impact and global tipping points where entire Earth systems collapse. The good news is on the left hand side, I no longer believe any of those high projections. They're not going to happen. The bad news is on the right hand side, all of those reds and yellows have crept downwards every assessment, and I should have said really intuitively you can see the yellow is where there is a significant probability of something somewhat worrying. The red it's getting bad and the purple is getting catastrophic, and those two roughly. seem to sort of cancel out, which means climate change looks like just as severe a problem as it always has been, except that it's a lot more immediate and prescient as indeed we see in the the extreme impacts we see around the world. So we are actually in a zone where probably looking at somewhere between 1.5 and 3 degree temperature rise. But that is a huge range in terms of what it may imply for humanity. It is also a huge range in terms of of how demanding or what it implies for the energy system. Now that's been the scientific framing around risks. The other framing, if you like, is the economics one. What an economist by and large is most interested in is the trade-offs potentially and some numbers. Well, we've had a good 2 or 3 decades of economists trying to attach the most significant number they think, which is the social cost of carbon emissions, how much damage, what's the dollar equivalent of a ton of carbon. Some, I think a lot of non-economists find this whole idea somewhat distasteful or dubious, and there's been extraordinary ranges in the attempts to attach a number. But nevertheless, it is striking that although I wouldn't say there's been a lot of convergence, there has been a clear trend. The number keeps getting higher and the range on the upside to some extent gets higher. And here in yellow I've highlighted a paper published in Nature a few months ago. I think relevant not just because it was in Nature and a serious effort and serious authors, but it was a response to the call of the US National Academy of Sciences a few years ago saying, look, we've learned lots of things and that has not been adequately put through into revising estimates of the SCC, uh, the social cost of carbon. Uh, and indeed one of the, if you like, the most sort of climate economic skeptics, Richard Tol, uh, published to say, yeah, actually he recognizes the cost has gone up quite a lot. If you fully cost at the kind of levels suggested in that most recent study, which is about 3 to 4 times what a lot of economists have been saying only a few years ago, then basically the cost of climate damages from burning coal is substantially higher than the cost of buying coal. And in fact, in oil, you may be looking on the order of the equivalent of about $50 per barrel of oil associated with the climate damages, plus or minus still a a a sizeable range, but that's pretty significant. In terms of what we then look on in longer term goals and the framing context on emissions, well, arising from the Paris Agreement, governments agreed two kinds of emission goals. It's worth saying one was the objective to stabilize temperatures, effectively interpreting the original convention's commitment to stabilize atmospheres. At a level to avoid dangerous interference, governments agreed that means well below 2 degrees and striving towards 1.5. And the vast majority of major countries around the world have declared net zero emission goals between 2050 and 2070, which roughly corresponds with that range. The other emission goals was to be expressed nationally through the nationally determined contributions. On the left is the chart from the IPCC report which says, well, the red is where we think we were going given policies as of 2020. The purple is what those NDCs indicated as national ambitions to 2030. Obviously way off track, so the Paris Agreement embodies two currently fundamentally inconsistent things a long term goal and a nationally determined trajectories which are not consistent with that long-term goal. If we go further and say what The model suggest to be the often least cost trajectory? Well, it starts by radically changing course, dropping emissions by somewhere between 25% and at least 45% by 2030. That's only 7 years away, to be honest, I don't think that's going to happen. So if I now start moving forward to the analytics, because some of you may know that background, um. This is an economics conference. There's plenty we can say about climate change economics. Probably the strap line was best captured by Nicholas Stern, who described climate change as the biggest market failure in history. But it has also been described as the perfect moral storm because of the ethical, it raises ethic issues about our responsibilities to other people, other countries, other generation, and human beings do have a sort of almost infinite capacity to to try and find some way of saying it's not really my responsibility per se. It's been described as a super wicked problem, one that is from a social science perspective, just really hard to solve. And it's also been described in terms of psychological distance. Human beings kind of evolved, didn't evolve to even be able to conceive about something on this scale, globally, intergenerationally, planetary risks, etc. In the IPCC we try and also map out the corresponding analytic frameworks. We're all here familiar with aggregate efficiency. I'm going to say a few words about the ethics and equity dimensions more explicitly, but really focus on the third innovation and transition in the second half of my talks, not to forget the psychology and political dimensions of this. So just on the equity front briefly, the world is unequal. I don't have time to run you through the details. That chart on the right shows different regions vertically per capita emissions against population. The the dark blue is fossil fuel CO2, the orange is land use, uh, which is a much more concentrated issue. Broadly you see substantial variation vertically between countries, but not just north-south, also within countries at similar regions of development. Two other points I'll draw out from this. Uh, one is, yes, it's partly about trade and offshoring, but actually even our consumption footprint. Uh, that amplifies but doesn't fundamentally change the kind of conclusions there. And also it's not just about between countries. In fact, a little bit like, uh, John's presentation earlier, there's inequalities at every step of the system within countries internationally. You can look at roughly 10% of the richest households around the world account for probably over a third of emissions through their consumption patterns. I'm not going to describe the chart on the left, it's a more elaborated version of what I said, greater disaggregation, but this time plotted against Human Development Index. What it really says is, All countries and regions will need to change, but what and how they need to change will vary a lot. But a lot of the discourse now is really about in terms of the development, paradox and challenges, not just alignment and co-benefits, but shifting development pathways towards sustainability. So where does technology come into it? Well, we've seen big reductions in the cost of renewable energy, as I'm sure you'll well know. These are classic charts now of solar and wind and below the expansion rates exponentially at rates of 20 to 40% a year. Generally now, both solar and wind, the two biggest renewables, uh, not everywhere, but as a broad international aggregate, are cheaper than fossil fuels as a way of generating electricity. That is a pretty fundamental change from what we looked at a few years ago. How long can that exponential growth rate, and by the way, even highly numerical experts, I find, are not actually that good at really thinking what does exponential growth mean. I invite you in the coffee break, pull out your computer, stick in exponential growth at these rates, look what happens within this decade. It is really quite striking. And the data suggests as published last Thursday by the IEA, it ain't slowing down in a hurry. Expansion from about 200 gigawatts a year pre-COVID, then 300 this year, the IEA estimates 400 to 500 gigawatts of renewable energy being installed. That's a huge growth rate. Um, OK, I'm going to have to skip the audience participation, but All technologies, pretty much in energy have displayed two things. All of these turned out to be much cheaper than anyone projected. And they all actually involved government action at scale over many years on both the technology resource mapping, development, and the demand pool and price developments. So where does this take us in terms of wider transitions? Uh, I'll, I'll draw on, but I won't have time for the depth, obviously quite a major international program looking at these issues, um, involve partners, particularly in India, China, and Brazil, um, and we looked at a few case studies. Uh, wind, uh, both Europe and Brazil moved from 1% to 10% of, uh, more than 10% within a decade of wind energy. Uh, India, light lighting efficiency, extraordinary transformation partly driven by bulk government procurement, which resulted in what they call the cheapest lighting in history. Uh, that sort of echoes now what the IEA says about solar PV, which is the cheapest electricity in history. So. Yeah, lots of good news stories. What do we learn from those case studies? Well, as I said, significant government action, largely now self-sustaining, wouldn't have been pursued by the traditional economic prescription, competitive markets, price carbon, um, and probably not justified by cost-benefit appraisal if you took all of the projections of how much these things would cost. But the common themes cumulative progress, market-based innovation, sustained and targeted support beyond the pure R&D. Substantial uncertainties when when these efforts set out and strong international dimensions, the movements of technology frontiers from one region to another, partly reflecting swings of policies. The point being, the traditional simplified stories of either it's all about government R&D drops the price and then things diffuse, the A in the bottom left, nor the let's just do it through the market, flat-paced technology. Neutrality and then companies will look forward and say that's what I'm going to do. That's not reality. Most technologies iterate through repeated cycles of expansion, learning, innovation, expansion, learning, innovation, cost scale economies, learning, etc. Right, innovation's a complex journey and with just tripped 0 time, I don't have much chance to go through it. This slide looks terribly complicated. It basically says it's not just about technology. Every stage of this process requires developments in all of those six factors you can see on the left. All of those to some degree have to develop along with the technology. It's more of a sort of a map of a multi, multi-level journey to point to examine where might the big blockages occur that would get in the way of a fuller transition. So this will bring me towards final remarks. And I should have said I know I had a choice to make. Either I could try and pretend to be a development economist or I'd stick to what I know. There are a lot of issues, and in a second I'll wind up by by noting some of the barriers. But basically transitions always start small, may take years after technology emerges through a phase of market emergence before they really emerge into the mainstream. The impact on incumbent industries is often barely visible during that process, why? Because exponential growth starts small, looks small, and the big, big, big boys don't really notice it. But boy do they start to change things at the margin and then more fundamentally. I've got another version which shows some of the incumbent industries as opposed to technologies, maybe jumping on the rising blue curve, as was already alluded to. But a key thing tends to be at this sort of phase in the middle that requires a reconfiguration often of market structures, because of all energy markets ultimately are human creations, and the kind of markets that suit fossil fuels don't necessarily adapt well and most cost effectively for renewables. So this other chart takes uh actually a sort of concept from a book that we published a few years ago around 3 pillars of policy and turns it into a dynamic chart that one does need phases of strategic investment to foster emerging technologies, businesses, etc. Then one needs to evolve, refigugure infrastructure and market structures to adapt suited to new technologies, build up the scale and technologies, supply chains, accelerate global diffusion. You may well also have issues of attending to standards, behavior, uh, norms, and maybe measures to address er laggards and and obstacles. So, um, I'm going to finish the slideshow there with, with one exception, but let me just say, of course, uh, capital finance are absolutely crucial. Uh, if you look generally in aggregate, you'd say to meet the kind of transitions projected by the IPCC as needed, we need to expand the pace of international investment or global investment by a factor of between 3 and 4 compared with 2020 levels. And if you look differentially, of course that gap is bigger in much of the developing world. I won't run you through those numbers, but what is required to actually get such large scale investment into what look like potentially attractive investments is not trivial, and of course it depends a lot on the cost of capital. Um, I think all other presenters have presented econometric studies. I've sort of got one in my back pocket which we expected to give a more technology oriented answer to the cost of capital and it, and you know, cost of technologies including solar. It basically said looking across countries, it's access to capital markets, the terms and the perceived country risks that are dominant factors. So still big issues that obviously the World Bank and others need to figure their way through. My concluding slide is just to do a simple diagram but conceptually still quite powerful and and a set of messages. If you look at this, it's not just a new technology, a few green technologies, it's an completely different structure of our energy systems. For all kinds of reasons I could go into, the transition is already underway, but so far it's driven more by non-market policies, and that is proportionately changing as these mature and driven more by the centers of innovation. That also has to change. I think one still needs elements of those three pillars of policies that I flagged, and we are talking about effectively industrial development strategies that help to shift development pathways. Um, we do see growing social in social and innovation and engagement, um. There's a long way to go, but I hope I've given you some signs of hope, er, as well as acknowledging we face massive challenges. Thank you very much. Many thanks Professor Grubb. Excellent presentation. Um, we are over time, but I would like to give the opportunity, uh, to the ones in the audience. I don't know if we have questions online, but to give the opportunity to one or two questions, uh, before we go, uh, for the coffee break. Um, if you have a question, please come to the mic, be succinct, and, uh, indicate who are, uh, directing the question to. If there are no questions, we can go for coffee right away. You have a great can you please come to the mic? Good morning. I was impressed by the presentation from. The gentleman who presented on Korea. Growth in industry, but what I did not get from that presentation was the state of the political economy, the culture, and the dynamism, because when Oadr read about the miraculous growth of South Korea, they link it to the leadership at the time. So I would have expected that although light would have been thrown on a level of leadership, commitment and trust that engender growth during that process. So can you share some light on that, please? Thank you. Thank you for the question. Uh, obviously, there's a very interesting story behind it. In the 70s, 80s, the really successful uh outcome was really driven by close collaboration between uh the government leadership and also some of the leading businessmen, the founders of Samsung and Hyundai, they're very special people. Then, as you noted, the political economy problem. Uh was formed because even though the government formally stopped doing industrial policy in the 80s, there was this ongoing implicit support and understanding going on, which was still fine, but once you go to 2000, that became really serious problem because as Wufu was saying, now they became just too big. They were so dominant in so many different industries and then without any uh explicit Uh, regulation They were able to abuse their market power and enter into many different industries and deter the entry of the uh the entrepreneurial activities, and that is a big part of the slowdown in productivity growth towards the end. So, it can, we cannot really say that it's a good thing or a bad thing. Sometimes a close relationship and cooperation can help the whole economy lift the boat. Everybody's sport, but then sometimes actually if they start to use it in the wrong way, then it could be very hurtful for the agri-economy, just become a rent seeking activity for some. It's a very complicated story. Thank you. Uh, my question's for, uh, Professor Friedman regarding your presentation, um, as you talked about the children of an earlier age going to the more elite, um, neighborhoods per se, how would you reflect that on the reverse for putting like An institute in a lower income, um, neighborhood, would you have the same effect or an adverse effect? So, um, are you, are you asking about, um, The particular policies in order to help children rise up or more like downward mobility as a flip of the side of the coin of the of upward mobility? OK. Um, yeah, I think what the literature shows is that, um, You need two different things I think in order to give children from, uh, lower income or lower socioeconomic backgrounds opportunities. One is just the institutional pathways to get them into, um, better education, better health, um, the types of things that are gonna propel them along a career. But then also what you need is the social and network structure in order to make this something that they want to do, right? So for instance, you know, one example, um. You know, in India, the. Admissions exams to some of the, you know, very elite, um, IIT IIM, those types of places, right? There's uh, I don't wanna like say that that's the optimal system because I think it creates a lot of, um, inefficiencies as well, but there's a sense that like this is something that everybody wants to do and no matter what your background is, there's this test. All you need to do is do well on the test. You don't need parental connections. You don't need anything, just like do well on the test. Um, and again whether that's the right system or not, what that does do is it creates, uh, a situation where people from all sorts of different backgrounds feel like this is they, they know about this, this is something that they can think about doing, um, it, it has the appearance of, of something that anybody can get into, um, there may be children in, you know, lots of different villages and neighborhoods that, that have access to that, um, you need that in addition to the institutions that, that do the training as well. Thank you. Uh, with this, I would like to conclude this session. Uh, I will not attempt to summarize what we learned here today was a very rich discussion. I would like just to say that I took, uh, This take away I think it's very important that countries uh adopt both policies to support competition, innovation, and equal opportunities in order to achieve sustained growth. Thank you so much to all the panelists. And for the very good discussion. And we'll see you at the, at the coffee break.
showAllTranscripts
no
duration
PT1H21M5S
scene7File
worldbank/ABCDE23_Session_1
scene7Domain
https://worldbank.scene7.com/
scene7FileAvs
worldbank/ABCDE23_Session_1-AVS
title
ABCDE23 Session 1
description
ABCDE23 Session 1
showTimestampAndTranscript
yes
col-xs-12
col-sm-12
col-md-3
col-lg-3
col-xs-12
col-sm-12
col-md-7
col-lg-7
  • add-style
  • lp-body-content
This video features the following papers presented at the ABCDE 2023 Conference: Economic Growth through Creative Destruction, Intergenerational Mobility Around the World, The Plant-Level View of Korea's Growth Miracle, and The Transition in the Economics and Policies of Energy Transition.
lp-heading-top-medium
lp-heading-bottom-medium
col-xs-12
col-sm-12
col-md-2
col-lg-2