YT YouTube2Text

YouTube2Text → Transcripts → Lecture 1: Empirical overview of macro development

Lecture 1: Empirical overview of macro development — Transcript

by CEPR & VideoVox Economics · 14,230 words · 2,051 segments · language en · Watch on YouTube

Open in transcriber Summarize with AI Download SRT/VTT

Full transcript

0:00I wanted to welcome you to the uh this
0:02first lecture of STEG's uh virtual
0:05course on key concepts in macro
0:07development which is an experiment.
0:08We've been overwhelmed by the um number
0:11of uh of people that have signed up for
0:14the course. You can see the participants
0:16flooding in. Um there's a smaller group
0:18of people that uh signed up to be in the
0:21front row there as participants. You can
0:24uh Richard has said that you can
0:26interrupt the um [clears throat]
0:29interrupt him if there's a question that
0:31needs to be answered at the at the at
0:33the time. Um others who are not in the
0:37front row as participants should enter
0:40uh questions into the Q&A. Uh Doug Gan,
0:44by the way, I'm Joe Kabosski.
0:46Doug Gallin is my um partner. were
0:49co-academic leads for STEG structural
0:52transformation and economic growth um
0:54which is the program sponsoring this. Uh
0:57Doug will be trying to answer questions
0:59in the Q&A. Um Richard will leave a
1:01little bit of time for uh questions as
1:04well at the end.
1:05Um, [clears throat] I want to
1:07acknowledge
1:09um the UK's um
1:12um funding the United Kingdom's Foreign
1:16Commonwealth and Development Office uh
1:18is supporting this class as part of the
1:22broader UK aid effort. So, we're very
1:25thankful. Uh hopefully this is a
1:27successful experiment and then I'll just
1:29I'll hand it over to Richard. Um, if you
1:32have any feedback on the course, uh,
1:34feel free to email me. I'll put my, um,
1:38email into the Q&A. Okay. Go ahead,
1:41Richard.
1:43Okay. Uh, thank you, Joe. So, uh, good
1:48morning, good afternoon, good evening to
1:51everybody. Uh, it's a great honor to be
1:55delivering the the opening lecture of
1:58this series. It's a series that I'm uh
2:00very excited about and and it's part of
2:03this of STG which is also something I'm
2:06incredibly excited about.
2:09So today I'm going to kind of introduce
2:12things. So again, this is kind of the
2:15the what I have in mind here is this is
2:18kind of the the first hour of a of the
2:22first class if I were teaching a a class
2:25in in macro development. kind of going
2:27to introduce you to some basics uh and
2:30then kind of lay the groundwork for all
2:32the the lectures that are going to come
2:34later. So, I'll talk a little bit about
2:36those and how those relate to some of
2:38what we talk about today. So, I'm going
2:40to go ahead here and share my screen.
2:48Okay. And I I guess I want to also just
2:50thank uh Joe and and Doug for all their
2:54hard work in kind of leading the effort
2:57of of SPEG. As I say, I think it's a
2:59it's a great thing that they're doing
3:01and I think a lot of good things are
3:03going to come from it. So let's uh get
3:06going. So the big question in macro
3:10development. So, one of the nice things
3:12about macro development relative to lots
3:14of other sub fields inside macro is it's
3:18actually there's kind of a a singular
3:20focus in terms of of what we're after
3:23and and and what we're after. As it says
3:26here, differences in living standards
3:28across country are large and persistent.
3:31Why? So, it's that simple. That's uh the
3:34big question. Let me give you a couple
3:37sub questions kind of that are just
3:39followons to this. We know that some
3:42countries have experienced dramatic
3:44increases in relative living standards
3:46over the last 70 years. So everybody's
3:48familiar with the the Asian miracle
3:50economies. What did they do to achieve
3:53this?
3:54Others have languished. Some have even
3:57experienced dramatic decreases in
3:59relative living standards. Uh case in
4:01point, Venezuela.
4:03what could they have done to change
4:05this? So these are the this as I say is
4:08kind of these are different versions of
4:10the same question but this is the
4:12singular focus for macro development and
4:16many of you uh may have heard a quote
4:19from from Robert Lucas. Uh I always like
4:22to mention this quote to students
4:24because Lucas of course won the Nobel
4:26Prize in economics. He won the Nobel
4:28Prize for his work on business cycles.
4:30business cycles is is is in some ways
4:33one of the defining issues in
4:35macroeconomics.
4:37But for those of you who who study
4:39business cycles and have some knowledge
4:41of of the numbers, uh a big business
4:44cycle episode is is when uh GDP per
4:48capita goes down by something like 5%.
4:50So if you take one as your baseline,
4:53it's going from one to 0.95, that's kind
4:56of a a huge event. uh as I'm going to
5:00talk about today when we talk about
5:01these cross-country differences in
5:03living standard we're talking about if
5:05we think of the US as as one the poor
5:08countries in the world are like at
5:100.03.02.04
5:124. These differences just dwarf business
5:15cycles. And so Lucas, who in the early
5:18part of his career devoted all his time
5:21to thinking about business cycles and
5:22trying to understand what people thought
5:25of as these huge fluctuations in
5:27economic activity in a given economy
5:30over time. Uh later turned to thinking
5:33about development. And as he said, once
5:34you start thinking about those
5:36cross-country differences, it's hard to
5:38think about anything else. Those
5:39differences are just so big and the
5:42potential for
5:44improving lives of people by
5:46understanding the source of those
5:48differences uh just dwarfs other things
5:51that we think about. So all that's kind
5:53of a preamble. Now the first thing I
5:56want to spend some time on in today's
5:59lecture is to talk about the importance
6:01of data. So, as I say here, you don't
6:05need any systematic data to realize that
6:07there are large differences in living
6:09standards. Anybody who has traveled the
6:12world will just uh see firsthand that
6:15there's large differences in living
6:16standards. So, it's not like you need
6:18the data to tell you that there are
6:20large differences. But if you want to
6:24systematically study those differences,
6:26understand them, then you need data. uh
6:29if you don't have data to actually uh
6:32have have measurement of the things
6:34you're trying to explain, you wouldn't
6:36know if you uh have adequately explained
6:39them in the sense that you can
6:42qualitatively tell me a story or write
6:45down a model where one country has
6:47higher living standards than another,
6:50but we actually need to know are the
6:52differences in living standards that
6:54you're explaining similar to the
6:56differences in living standards we see
6:58in the real world. So for that you
7:00actually have to have systematic
7:02measurement. So uh in general there's
7:05lots of data you would like to have as
7:07you think about richer stories of things
7:10that
7:13increases the data demands but kind of
7:15the minimal requirement to kind of get
7:17started is to have data that allows you
7:20to compare uh what we might call real
7:22living standards across countries and
7:25over time. And so uh uh one data set
7:30which is particularly important here the
7:31pen world table. This amounted to as I
7:34say here a revolutionary change in the
7:37widespread availability of the needed
7:39data with broad coverage in terms of
7:42country and time. So you know I'm I
7:45guess an old guy now. So I actually can
7:49think back to when the pen world tables
7:51uh first became available for all the
7:54the graduate students in the crowd. Now
7:56uh this is something you can kind of
7:58take for granted that this this exists.
8:00But this was actually you know effort
8:02that started in the late uh 1970s. So
8:06here's actually just a one of the first
8:09uh published papers. uh Cravens, Hston
8:12and Summers were kind of the the three
8:14people who spearheaded the effort behind
8:16the the pen world table. Here in June
8:2078, they published a paper in the
8:22economic journal. You can see the title
8:24of the paper, real GDP per capita for
8:27more than 100 countries. And what they
8:29did in this paper was basically tell us
8:32in 1970 they gave us a cross-section
8:35comparison of of real GDP per capita uh
8:40for more than 100 countries. So that was
8:42kind of the the early days. But I want
8:45to talk a little bit uh about the
8:48challenge faced by the effort to create
8:50these data because although nowadays you
8:53can just go to the various websites and
8:56and download the data you want to use, I
8:59think it's important to have a little
9:01bit of perspective on what the issues
9:04are in terms of constructing that data
9:06in the first place. So I just want to
9:08spend a little bit of time on that. uh I
9:11think when people use data it's actually
9:13important for them to have some sense of
9:16what the challenges were in creating
9:18that data in order just to be aware of
9:20any uh particular pitfalls that might be
9:23there. So just as a starting point
9:27remember what we're after here is is
9:29measures of real GDP uh across countries
9:34um over time. So as I say here, if we if
9:37if you're just measuring GDP, nominal
9:40GDP in a given country at a given time,
9:43there's already lots of challenges uh
9:46that that come up in the context of
9:48that. Uh so you can go to any kind of
9:51undergraduate textbook on macroeconomics
9:54and there's a chapter on measuring GDP
9:57in the various issues. Uh this is even
10:00these issues are are even uh much more
10:04pronounced in in less developed
10:06countries. Uh for the simple reason that
10:09measuring economic activity uh requires
10:12a certain amount of state capacity and
10:14state capacity tends to be uh much less
10:17in less developed countries. So but
10:19anyways we all we know that there's
10:21already many issues with that. And then
10:24if you have nominal GDP over time for a
10:27particular country and you just want to
10:29turn that into real GDP, we know there's
10:31a bunch of measurement issues there in
10:34terms of how you deal with things like
10:36uh new goods, changes in quality, etc.
10:40But just producing a series for real GDP
10:43in a given country over time, lots of
10:45challenges. As I said here, the object
10:48of the pen world tables was was not to
10:51do a given country or to do comparisons
10:54across countries and and this is
10:57actually a whole another poses a whole
11:00another set of challenges. Now, there
11:02was a traditional strategy and the
11:04traditional strategy uh you might think
11:07about it from a simple-minded
11:09perspective in the following way that if
11:12I go out at a given point in time and
11:14measure GDP in the United States, I'm
11:16going to get some some number of
11:18dollars. If I go to Mexico and measure
11:21GDP at a particular point in time, I'm
11:24going to get some number of pesos. And
11:26the question is, how do we compare
11:29dollars and pesos? And well one simple
11:32way is use exchange rates to
11:34[clears throat] convert pesos into
11:35dollars or dollars into pesos and then
11:38uh once we've converted them to the same
11:40units then we can compare these at that
11:42point in time. So that was the
11:45traditional strategy. I want to make a
11:47few remarks on that traditional
11:50strategy. So first uh if you do that you
11:54get rid of the units problem. Obviously,
11:56you start with uh dollars and pesos, and
11:59this can turn everything into pesos or
12:01everything dollars. But having done
12:04that, it's not clear why the exchange
12:07rate is is in any sense appropriate for
12:10talking about the the standard of
12:12living. If all people did with their
12:15income was buy oil which is traded in an
12:18international market and everybody faces
12:21the same price in dollar then that might
12:23be a fine way for saying how much oil
12:25can people buy. But in fact that's not
12:28the only thing that people purchase. So
12:30using the exchange rate is not necess
12:32there's not really a a good reason for
12:34thinking that that's a good way to think
12:36about uh a standard of living across
12:40countries. Another way to say this is is
12:43although you can put things in the same
12:45unit, kind of the the deeper question is
12:48if I give you a dollar and then I send
12:51you to New York City or Mexico City and
12:55ask what can you get with that dollar,
12:57that's going to vary across countries.
12:59And so the real challenge here is how do
13:02you
13:03assess the purchasing power of a dollar
13:06across countries? And so this is the um
13:09huge challenge uh behind endeavors like
13:12the pen world tables.
13:15And in particular, part of what ma what
13:17makes this a huge challenge is the fact
13:20that we're talking about countries at
13:22very different stages of development.
13:24And so just to to pursue that a little
13:28bit, uh one thing we know is is uh the
13:32standard approach to thinking about uh
13:36how valuable a dollar is in different
13:38places is to take a common basket and
13:42price that basket in the two different
13:44places. Um so that's a common strategy.
13:47This of course is the the essence of the
13:50CPI measurement in countries the United
13:53States where you create a basket of
13:56goods which is the uh in the United
13:59States it's the basket a representative
14:01basket for the average urban consumer
14:05average urban household in the United
14:08States you take that basket and then you
14:10price it in different places and that
14:12gives you a sense as to how far a dollar
14:15goes how much of that basket can you by
14:17in different places. Now, why does this
14:20not translate very easily when we're
14:22talking about crosscountry differences?
14:25It's because the idea of a common basket
14:27doesn't really make any sense. Uh so in
14:31particular uh advanced economy items can
14:35be notoriously expensive in lowincome
14:38countries and and so if you take a
14:41basket that you think represents what
14:43people in advanced economies purchase
14:45and then go and price that basket in a
14:48less developed economy, you're going to
14:50think that that basket is is incredibly
14:52expensive. and and and the issue is that
14:55those advanced economy items are
14:57actually not part of the typical basket
15:00in poor countries and actually there's a
15:02lot of uh thin markets, monopoly power.
15:06A lot of those items are being uh
15:08imported just to be sold for people who
15:11work for international organizations. So
15:13it's it creates a very uh challenging
15:17problem in terms of how to think about
15:18this. Um so there's something called the
15:21ICP, International Comparisons of
15:24Productivity Project. What what this
15:27whole endeavor does is it works to
15:30create these measures of of purchasing
15:32powers. As they say, this is a hugely
15:35challenging endeavor and that they say
15:38the big challenge is is thinking about
15:42very rich and very poor countries and
15:44and and how to do those comparisons.
15:48So this ICP they actually uh redo these
15:52comparisons every few years and actually
15:55just to kind of underscore the
15:58challenges uh that are there uh in 2005
16:02there was a set of revisions and those
16:04revisions in 2005 I think are widely
16:08believed to have aired in
16:10underestimating purchasing power in
16:12low-income countries in a sense they put
16:14too much weight on on the prices of
16:18advanced economy goods in in lowincome
16:21economies. Uh and that to a certain
16:24extent this was undone by the 2011
16:27revisions.
16:29Again, I just want to kind of flag this
16:31issue as something that you be aware of
16:33in the background. that whenever you go
16:36to something like the pen world tables
16:38in the background uh somebody has done
16:41not somebody a whole team of people has
16:43done a lot of work to think deeply and
16:45try and measure these differences in
16:47purchasing power I would just point you
16:50here I I list one paper uh by Angus Deon
16:53uh Angus Deon is is an individual who's
16:57thought very deeply about these sorts of
16:59index problems uh but he has a nice
17:02article in the AEJ macro in 2017. It's
17:06part of a a two paper symposium that was
17:09there uh talking about um the issues and
17:12the the change between the 2005 and the
17:152011 revisions and how that has a huge
17:18impact on on the quantitative assessment
17:21the differences in living stand. So
17:24again, I'm not going to say anything
17:26more about this issue. Uh but the basic
17:28point here is I think if you're going to
17:31use the pen world tables uh it's
17:33important that you understand that in
17:35the background there's this uh very
17:38challenging uh but important effort uh
17:42to to assessing differences in
17:44purchasing power across countries and
17:46that's a an incredibly challenging and
17:49and difficult problem.
17:52So uh I've mentioned the pen world
17:55tables and the pen world tables was
17:57certainly the trailblazer in terms of
17:59providing a source of cross-country
18:02comparisons of real GDP over time. Uh
18:05many of you might have gone to other
18:07websites where you get data. So another
18:10website uh that I go to a lot to
18:13downloads they have a lot of information
18:14there is the growing and growth and
18:16development center the GGDC.
18:19They maintain uh several series
18:21including something called the Madison
18:23historical statistics, the 10 sector
18:26database uh EU cles. There's also uh the
18:31conference board maintains something
18:32called the total economy database
18:36and then the United Nations maintains
18:38something called world development
18:39indicators. Again, these are different
18:42places you can go. I'll say the United
18:44Nations the world development indicators
18:47is a little bit uh distinct in terms of
18:50of
18:52where their data is coming from. Uh all
18:54of the others in some ways have at
18:57various points kind of been involved
19:00with the growing growth and development
19:02center. in some ways the growing and
19:04growth and development centers kind of
19:05become the the the center point uh for a
19:09lot of these data efforts and and people
19:12who've actually been there and gone
19:13elsewhere and moved data sets with them
19:16uh the GGDC uh very important uh
19:21enterprise and and a website with a lot
19:24of uh relevant information.
19:27So, uh, all that's just kind of saying a
19:30little bit about the, uh, data that's
19:33that's now out there and some just some
19:37cautions to be mindful of in terms of of
19:40how that data is created.
19:42Now, the next thing I want to just make
19:44one comment on is about living standards
19:48and productivity. So income per cap if
19:51you just ask somebody what's a a good
19:54measure for uh standard of living I
19:56think they would say income per capita
19:59real income per capita that would be the
20:02most commonly used indicator uh but what
20:05I want to point out here is is that
20:07through first approximation
20:10there's actually a pretty you get pretty
20:12far uh instead of thinking about income
20:15per capita if you actually think about
20:17uh a productivity measure
20:19uh output per worker and and what I want
20:23to do on this slide is just kind of
20:24break out why that is and and the first
20:26thing to realize of course is that in
20:29national income and product accounts
20:31income is equal to GDP. So income per
20:34capita is of course just GDP per capita
20:37but GDP we now are talking about um
20:40output in the economy and the difference
20:42between GDP per capita and GDP per
20:45worker through productivity um measure
20:50uh has to do with the extent to which
20:52workers per capita varies across
20:54countries and in the development context
20:56the key question would be the extent to
20:59which workers per capita varies with
21:01income per capita. Now there is some
21:03variation there and it's certainly not
21:05that all countries line up uh there but
21:09what I say here is that in some ways to
21:12first approximation if you actually just
21:15focus on GDP per worker you actually um
21:19are capturing most of the differences in
21:21income per capita and I think that's a
21:24useful point just to be aware of that
21:28when we're talking about differences in
21:29living standards as I say the the first
21:32order that's basically saying that there
21:35are big differences in output per worker
21:38across the country and you'll see in
21:40some of the slides I'm going to show
21:41today from different sources some people
21:44actually focus entirely on GDP per
21:47worker uh sometimes actually GDP per
21:50hour although getting data on hours
21:53across countries tends to be a more
21:55challenging
21:57effort
21:58so some people report GDP per capita
22:02some GDP per worker. And I guess what I
22:05want to say is is uh don't be thrown off
22:08thinking that oh there's some big
22:10problem because they're reporting very
22:11different things as they say uh to first
22:14approximation. These kind of um paint
22:16the same picture in terms of of what's
22:18going going on.
22:21So uh what I want to talk about next is
22:24just how to organize our thinking. So
22:27now we have uh thanks to the people at
22:29the pen world table uh we have this rich
22:33data set differences in GDP per worker
22:36or GDP per capita I say you know
22:39comparable real measures comparable
22:42across countries for large number of
22:44countries relatively long time period
22:47now I want to ask the question if we
22:50look at that data uh how does it kind of
22:52guide us in terms of how we should think
22:54about things and so that's what I want
22:56turn to next to talk.
22:59So there are of course uh hundreds maybe
23:03thousands of of papers who've documented
23:06various aspects of the uh crosscountry
23:11uh real GDP per capita um series. I have
23:16on this slide just a few of of I'll say
23:19my favorite, maybe I shouldn't say my
23:21favorites, but just ones that uh I've
23:24enjoyed reading over the years. Um, so
23:27I'm not going to uh go through the list
23:30here, but that's just a a bunch of them.
23:32And what I want to do now is I'm going
23:33to kind of go through some of these and
23:35show you some pictures uh and then kind
23:38of talk about where it leads our our
23:40thinking. So here's a type of picture
23:44that you will see presented quite often
23:46when people are talking about
23:48cross-country differences in in living
23:51standards over long time periods. So
23:54this particular picture I took, this
23:56comes from
23:58uh from Chad Jones's uh handbook chapter
24:01which I think was the one reading that I
24:03had listed for today's lecture.
24:07And this is just again going back to
24:091200. So this is using data from the the
24:11Madison project. And what you can see
24:14here is is that up until 1600 or almost
24:181700 everything is the same. the world
24:22is uh is flat both in the sense that
24:25there's no differences across countries
24:27and there's just nothing happening
24:29across countries. So everything is is
24:31just then as as we all know in the 1700s
24:35the industrial revolution happened
24:37starting in the United Kingdom and then
24:40things kind of slowly change and what
24:43you see is is is that uh between 1700
24:46and say 1950 it looks like whereas
24:50you've gone from everybody being the
24:52same you then have this massive
24:54spreading out in terms of of uh standard
24:58of living as measured by GDP.
25:00um per person. And so some people have
25:02called this the the great divergence.
25:05Here's just another um picture uh
25:08showing the same thing. Starting from
25:101870, this idea that that countries are
25:14uh as you go through time, things are
25:17kind of spreading out. One thing I just
25:19want to mention when people show these
25:21pictures is is the scale here is is not
25:24a log scale. It's a it's a level scale.
25:27And I think uh as I always tell even
25:30undergraduates, you have to be very
25:32mindful of of interpreting figures based
25:35on on the scale. And I think when you're
25:37thinking about uh growth rates and
25:40relative growth rates and and
25:42divergence, I think is about differences
25:44in growth rates. It's very deceptive to
25:47to make inference based on looking at a
25:49picture where the scales in in a in
25:52levels as opposed to in in logs.
25:55But having said that, these sorts of
25:58pictures are ones you will commonly see.
26:00Uh, and I had the one article by Lamp
26:03Pritchette on the previous slide and he
26:05his article was called divergence big
26:07time. And this is the kind of stuff he
26:09was talking about that there was if you
26:11go back far enough there's no
26:13differences across countries and as you
26:14go forward you see bigger and bigger
26:17differences. I want to talk a little bit
26:19about
26:21how this guides how we think about what
26:24we're trying to understand. So I call
26:27this slide levels versus growth rate
26:29one. This is an issue I'm going to come
26:31back to again a little bit later in
26:34today's lecture. Uh but what I want to
26:37point out now is the the release of the
26:39pen world table. So this was you know
26:41data that people started um having
26:43access to uh I guess late 1980s
26:47something in the 1980s. Um at the same
26:50time that the pen world tables was
26:53released and people got access to this
26:55data a different development in some
26:58ways completely somewhat independent. I
27:00would say it was the inside
27:02macroeconomics was the emergence of the
27:05field of endogenous uh growth theory.
27:07out you know some of the famous uh
27:11papers from that early literature uh
27:14Paul Romer Bob Lucas uh Gman and Helpman
27:18Aon and Howitt uh so anyways very
27:22exciting time in terms of these theories
27:25the hallmark of indogenous growth theory
27:27was was that country specific policies
27:30can lead to permanent differences in
27:33growth rates across countries that was
27:35kind of a one of the early implications
27:38uh of that literature and so a lot of
27:41people were excited about this because
27:42they said this creates a link between
27:45country specific policies and growth
27:48rate in country now that uh property of
27:52endogenous growth models was in contrast
27:55to kind of the existing uh I'll say solo
28:00type uh growth models call them
28:02exogenous growth models because in those
28:04in in those exogenous growth models,
28:07policies had level effects, not growth
28:10effects. Uh, and so people were very
28:12excited that you could have policies
28:14that have permanent effects on growth
28:16rates, that two countries that have
28:18different policy regimes would have
28:20different permanent growth rates.
28:23Now, just to connect this with with some
28:26of those pictures I showed you, uh I
28:29mean, if you I didn't uh mention it at
28:31the time, but of course, if you go to
28:33the to current period, uh the
28:36differences between rich and poor
28:37countries, you get something like a
28:39factor of 30, 35, or even 40 in terms of
28:43differences in GDP per capita or GDP per
28:47worker. It turned out it was people
28:49found it kind of challenging to try and
28:52get level effects that are that big in
28:54the exogenous growth models. And so uh
28:58in some people thought that the
29:00endogenous growth models were like
29:03particularly promising in terms of
29:04helping us understand those differences
29:07because if you actually have take the
29:11perspective of the indogenous growth
29:12models and say there's country specific
29:15policies which create permanent
29:17differences in growth rate then you can
29:20have differences in growth rates of say
29:238% which you think maybe is not uh so
29:26big and then if you just let things run
29:28long enough you can have massive
29:31differences in terms of the level
29:33differences across countries. So the
29:35idea was that it's that indogenous
29:37growth theory makes it easy for us to
29:39understand large differences across
29:42countries because countries have
29:44different policies different
29:45institutions. This leads to diff
29:49permanent growth rates and then just let
29:52things run over time and things are
29:53going to diverge and you're going to get
29:55very big differences. And this also kind
29:58of fit with this perspective from
30:00looking at these pictures here which
30:02said if you think what describes the
30:05world is that things are diverging then
30:07you actually want a model in which
30:09growth rates are different. And and so
30:12the these two things kind of tied
30:13together and this led to a view that in
30:17fact endogenous growth theory thinking
30:19about permanent differences in growth
30:21rates is the way to think about what's
30:24happening across countries and and the
30:27large part of what I want to do in the
30:29next uh coming minutes is dissuade you
30:32from that perspective. And to do that
30:35I'm going to uh show you some pictures
30:37from some of these other papers I talked
30:39about. So the one paper it's now you
30:43know many years old and so a little bit
30:45out of date and it used data only until
30:481985 but I think it kind of set the
30:50stage for framing how we should think
30:53about uh the crosscountry income
30:56distribution and how it's evolving over
30:58time and so this was a paper by Steve
31:01Pente Ed Prescott
31:03uh changes in the wealth of nations
31:05which actually just there was a
31:06publication in the Minneapolis Fed
31:09quarterly review. But again, I think it
31:12kind of sets the stage for for how uh
31:15people currently think about these
31:17things. And so here's I'm going to show
31:19you a few pictures from this early paper
31:21of theirs. And so here's the first one.
31:25So they were using data from the 10
31:27world tables. And as they say that the
31:30because their paper's older, it covered
31:32the period 1960 to 1985.
31:35And so here's actually um a very
31:37striking picture, very important
31:39picture. What they plotted here was uh
31:43they they said let's call the US
31:45normalize the US level of GDP per worker
31:49equal uh or here I guess GDP per capita
31:53uh equal to one. And then they're going
31:55to compare the five richest countries uh
31:58relative to the United States and the
32:01five poorest countries relative to the
32:03United States. And actually the striking
32:05thing you see here is that these lines
32:07the first approximation are basically
32:10flat. Uh back in in the data they were
32:14using then the poor countries were one
32:16over 32 of the United States. So kind of
32:20uh uh again over around
32:243% of the United States. The richest
32:27countries you can see there's a log
32:29scale here. They're probably around 90%
32:32of the United States. But these lines
32:33are flat. And the key thing about that
32:36is that if these lines are flat, it
32:38means there's actually no tendency over
32:40this period for the distribution to be
32:43spreading out. If there were permanent
32:45differences in growth rates and the
32:48reason that the countries were poor is
32:51because they have low growth rates and
32:53the reason countries are rich is because
32:55they have high growth rates. You'd
32:57expect to see these two lines uh getting
33:00further and further apart. But in fact,
33:02those lines are just kind of holding
33:04steady. Here's just another perspective
33:07on the whole thing. Instead of looking
33:11uh
33:15at at just the relative to the US at
33:18each point in time. Uh this is actually
33:21looking at how the the levels are
33:23changing over time. And what you see
33:24here at the top, the middle and the
33:27bottom, these things are kind of moving
33:29in a parallel fashion over time. Okay,
33:33so those are two looks. Now that should
33:36not be confused with saying that there's
33:39nothing happening to the distribution.
33:43That's kind of saying there's no
33:44tendency for the distribution to become
33:47to to be diverging. But uh as as this
33:50picture showed, there's still lots of
33:53action in inside this distribution. And
33:56in fact um uh if you just look at the
33:59overall dispersion in the distribution,
34:02you actually see that that dispersion is
34:04is changing over time. But that actually
34:08uh represents a few different dynamics.
34:11And so two different dynamics which have
34:14have received lots of attention. And
34:16I'll show you some more pictures on this
34:17later. uh kind of an old textbook
34:20example when people wanted to emphasize
34:23how good the solo growth model is is
34:26people uh talked about what happened in
34:28Europe post World War II and they talked
34:31about the fact that post World War II uh
34:34initially there was big differences
34:35across the countries in Western Europe
34:37and over time uh there was convergence.
34:40So
34:42uh in Europe uh there was the the
34:45dispersion went down uh over time. In
34:49contrast, if you look at Southeast Asia,
34:51you actually see the opposite over this
34:53uh time period. You actually see massive
34:56increases uh in dispersion.
34:58[clears throat] Um but but those are
35:01actually as I say not uh well we'll
35:04we'll come back to that in in a few
35:06minutes.
35:07uh what this is is inside the
35:09distribution there's there's lots of
35:11dynamics uh but the overall tendency in
35:14the distribution is is not for it to
35:16become more spread out here's just
35:17what's happening across economies in
35:20Asia over the in that particular time
35:23period you can see things um getting
35:26more spread out um and I say we'll we'll
35:29revisit that and this is just showing
35:32you what's happening to the distribution
35:34over time and the Key point here again
35:38even though the range of the
35:39distribution in a relative sense was was
35:42relatively stable over this time period.
35:45There is a tendency for the mass to move
35:47around in the distribution and in
35:49particular over the time period they
35:51looked at um there was a little bit of
35:53mass that left the middle and and went
35:56out towards the edges. Now related to
36:00all of this, this was the other thing
36:01that uh parent and Prescott kind of
36:04highlighted and again these are things
36:06we kind of all knew but this was kind of
36:09documenting it in a in a fresh way is
36:12that uh at the same time that the
36:14distribution was not tending to become
36:17particularly more spread out in any
36:19systematic way. There was actually lots
36:21of movement in the distribution and so
36:24they got us thinking about examples of
36:26miracles and disasters. some countries
36:29that moved up a lot in the relative
36:31income distribution and some countries
36:34that have moved down a lot. And over any
36:37given time period, you can find
36:39different examples of miracles and and
36:41disasters. As I said earlier today, the
36:45the most recent striking example of a
36:47disaster is is Venezuela where they just
36:50kind of fell off a cliff in terms of
36:52their um position in the relative income
36:55distribution.
36:57And of course uh back when Kante and
37:00Prescott wrote China was not the big
37:03story. Uh back then it was the other
37:05East Asian miracle economies. Uh
37:08nowadays of course China India are the
37:11some of the big stories in terms of
37:12countries that are moving up in that
37:14distribution.
37:16The the next table I want to show you
37:19here this was another paper I had on the
37:22list a little bit ago by Darte and
37:25Ruchia. And this paper came out in the
37:28Richmond Fed quarterly review. This kind
37:31of uh in some ways updated a lot of what
37:34Frene and Prescott did. But the other
37:36thing they did just to shed a little bit
37:38more light on the movements of countries
37:41within the distribution and to get us
37:43thinking about that dynamic was they
37:46talked about mobility matrices and and
37:49they did this here. I' I've got two of
37:51the mobility matrices that they
37:54presented. They do one of these based in
37:58levels uh relative to the United States
38:00and another one based on quintiles. And
38:03I guess those are obviously somewhat
38:05similar but they're a little bit uh
38:07distinctive because as countries as the
38:10bunching of countries moves the the
38:13mapping from quartiles to the the levels
38:16is changing a little bit. uh but the
38:18first approximation
38:20these matrices kind of tell the same
38:22story and and what I want to point out
38:24here and and I'll show you another
38:27picture um shortly which kind of
38:29emphasizes this is is that the diagonal
38:33elements here are kind of staying in the
38:35same position as you are now and what
38:38you see is all those diagonal elements
38:40are actually pretty big another way to
38:42say that there's a lot of persistence in
38:44the distribution but the other thing I
38:46want to emphas side is the persistence
38:49is actually pretty stable throughout the
38:52distribution. Of course, the two end
38:55points uh because of truncation, you
38:57can't go down if you're at the bottom
38:59and you can't go up if you're at the
39:00top. Those ones of course necessarily
39:02look a little bit bigger than the than
39:05the other entries. But one of the
39:08reasons I wanted to show you this is is
39:10that some of you uh may have been
39:13exposed to kind of a a recent literature
39:16or a recent idea uh called the middle
39:20income trap. There's a idea that some
39:22people have said that countries get
39:25trapped uh in the in the middle income
39:28stage and and what I want to emphasize
39:32with these pictures is I think that that
39:34uh the middle income trap is somewhat of
39:36a of a myth and the reason I say that is
39:39when you look at these mobility matrices
39:41what it you is that there's just a lot
39:44of persistence in the distribution and
39:46the persistence is there at all points
39:48of the distribution. So if you're in the
39:52uh fourth quartile, it's persistent. If
39:54you're in the third quartile, it's
39:56persistent. If you're in the second
39:57quartile, it's persistent. It the
39:59persistence is there everywhere. So
40:02there's no sense in which there's
40:04something about if you're in the middle,
40:06you somehow get stuck there. Uh the data
40:08tells us that countries tend to kind of
40:11stay where they are in the relative
40:13income distribution as as a rule. That's
40:16the tendency. And that tendency is
40:19actually pretty stable throughout the
40:21distribution.
40:22Again, I think uh these matrices are
40:25kind of key information uh for thinking
40:28about that idea.
40:31Okay. What I want to do now is is show
40:34you a few pictures.
40:35Sorry to interrupt.
40:36Oh, sure.
40:38I just wanted to ask uh that table you
40:41just showed us showed pretty obvious
40:43persistence with regards to people
40:45staying in their quartortiles. So, where
40:48did the middle income fallacy come from
40:50given that it that single table seems to
40:53kind of have semi-definitive proof that
40:55it's probably not legit?
40:58Uh,
40:59good question. So I think uh you know if
41:02you read some of this what what people
41:04have done is they've picked some
41:07countries which have transitioned from
41:10lower in the distribution into the
41:12middle of the distribution and then they
41:15say oh and once they do that they tend
41:18to stay there and that must mean there's
41:20a trap there. Um, and as I say, I think
41:23that's the wrong way of thinking because
41:26the reality is wherever somebody is, um,
41:30there's a tendency to stay there. And
41:33so, uh, I think there's this one of
41:37where the myth in some ways comes from
41:39is this idea if somebody goes from the
41:41second to the third, that must mean
41:44they're on their way to the fourth, but
41:46then somehow they got stuck in in the
41:48third. And as I'll I'll kind of paint a
41:52broader picture of this a little bit
41:53later, [snorts] but it kind of has this
41:55presumption that once you start to grow,
41:59you must automatically uh be on a path
42:03to look like rich countries. Um but I
42:06think that's just inherently false. I
42:08mean just to uh to give you know pick I
42:12don't know a random example uh Portugal
42:15is a rich country by all standards but
42:20Portugal in some ways is like one of the
42:22poorest of the uh so-called rich
42:24countries. Um but they kind of are are
42:28where they are. Does that mean there's a
42:30trap where Portugal is? Uh no. I don't
42:32think there's a trap. I think that's
42:34that's the situation that that Portugal
42:37is in. uh as they will talk a little bit
42:40more about that later on. But I, as I
42:43say, I think it's this myth that um once
42:46you start growing, that necessarily puts
42:49you on a path to be like the richest
42:51countries and if you kind of don't get
42:54there all the way, it means that somehow
42:56there was a trap somewhere along the way
42:58as opposed to you instituted some
43:01policies that create a particular
43:04position for you and then you go to that
43:07position. And if you want to go to a
43:09different position, you'd have to make
43:11some further policy changes or
43:12institutional changes.
43:15All right, thanks for that.
43:18On a related note, I mean, one thing
43:21that springs to mind uh for me is that
43:24this this seems like a sort of natural
43:26explanation for the two conflicting
43:27facts that there seems to be sort of
43:29convergence within the OECD and then a
43:32lack of convergence sort of more
43:34globally as presented in the Jones
43:36paper. And so
43:38what's your sort of explanation of how
43:40to how to square the fact that from this
43:43transition matrix there doesn't seem to
43:44be a point at which countries get stuck
43:46and yet there does seem to be a a
43:49stylized fact that past some GDP level
43:52you sort of converge towards the
43:54frontier and outside of that range it's
43:57not clear that that happens.
44:00Well, I guess I want to I'm gonna
44:03postpone some of that just till a little
44:05bit later in the lecture where I think
44:07I'll kind of put describe a framework I
44:10think to kind of reconcile why why I
44:13think there's no issue there and say
44:15first of all I think it's it's not truly
44:19correct to say that you know above a
44:21certain level then you converge because
44:23as I say if you look uh look at Portugal
44:26Portugal is uh in no sense is converging
44:30up to where the US is. Uh we could look
44:32at uh South Korea, which uh you know,
44:36I'm going to show you a picture for
44:38South Korea a little bit later in the
44:40lecture. South Korea made amazing. I
44:43mean, they're one of the miracle
44:44economies. Um but South Korea is is
44:47today around 7 of the US. They didn't
44:50get all the way to one. They certainly
44:52didn't uh they passed through the
44:55so-called middle income trap. So, it's
44:57not like they uh got stuck there. Um,
45:01but Korea made some changes and the
45:04changes that they made are consistent
45:06with being at at 7 and so they end up at
45:09at 7. But it's the data doesn't tell us
45:12that if you get to 7, you're you're on a
45:16path to converging up to one. Uh, that's
45:18just not in the data at all. So the fact
45:21that there was some amount of
45:24convergence among a subset of economies
45:27uh in in Western Europe is not evidence
45:30that once you get to 0.5 or 6, you're
45:34kind of home free and you're going to
45:35converge up to one.
45:38But as I say, let's uh postpone this uh
45:41a little bit more until later in the
45:43lecture. Okay.
45:44Okay. Thank you.
45:48Okay. So here's uh you know what I
45:50showed you before the prena as I say I
45:53think it was kind of early work which
45:55kind of laid the groundwork for how a
45:57lot of people now think about the cross
45:59country income distribution. I'll just
46:01show you some more recent pictures. So
46:04here's a picture from the Chad Jones
46:06handbook and and and he shows two
46:09different things. So the the green line
46:11here is the ratio of the fifth richest
46:15to the uh fifth poorest. And that's the
46:18scale on the right. Uh and what you can
46:22see there is is the early part here.
46:25This was the part that Pente and
46:27Prescott focused on. And you see it is
46:29kind of flat. You then see actually
46:32postpecott
46:33time period. This does actually
46:35increase. But in fact uh to first
46:38approximation although there's been a an
46:41increase there this has again been uh
46:44relatively flat for the last uh 20 or so
46:47years. At the same time as Pente
46:50Prescott noted there is an increase in
46:53standard deviation. So even though the
46:56the the extremes are kind of not
46:59diverging the the way the mass is being
47:01allocated within the distribution is
47:04leading to to higher standard deviation
47:06and that has kind of continued uh up
47:08until around 2000 in which case it's
47:11flattened out. So again this says that
47:13what and Prescott documented here is is
47:18the idea that this is just truly
47:20constant. that's just not borne out with
47:22uh more data. But at the same time, uh
47:26if you go back to the endogenous growth
47:29view and and divergence, you should be
47:32expecting just a a steady upward uh
47:35trend here. And in fact, uh you don't
47:37really see that. So So that's there.
47:40This is uh a picture that you'll see a
47:43lot of places. This is just another take
47:45on persistence. This is uh relative GDP
47:49per person in 1960 again US normalized
47:53to one uh and then in 200
47:5711. I should emphasize just going back
47:59here. Uh
48:02uh this uh the data that uh I think Chad
48:06is using here was was using this 2005
48:11ICP revision that I had talked about
48:14which made some of these uh differences
48:16actually a little bit uh bigger relative
48:19to to the current differences. So here
48:22you can see differences of of a
48:24magnitude 50 and that's I think a little
48:27bit of an artifact of this data issue I
48:29had mentioned. Anyways this picture is
48:32showing is just a way to get some sense
48:35of persistence in the distribution. Uh
48:38so relative position in 1960, relative
48:42position in 2011, this is the 45 degree
48:46line and I think what you see uh and
48:49again consistent with kind of the
48:51previous data uh to first approximation
48:55these points are scattered around the 45
48:57degree line which means there's a lot of
49:00persistence but any point that's above
49:03the 45 degree line that's a country
49:05who's kind of moved up. Any point that's
49:08below the 45 degree line, that's a
49:10country that has has moved down. And
49:13what you can see is there's mass above
49:15and mass below despite at the same time
49:18that the the points kind of track the 45
49:20degree line. And and what's a little bit
49:23harder to see from a picture like this,
49:25which is why I like those mobility
49:27matrices that I showed you from Darte
49:30and Restuchia, [snorts] is is is that
49:33they kind of give you a sense is does
49:34the amount of persistence change across
49:37the distribution.
49:39And you can eyeball this and have some
49:41sense. And I guess you know uh what you
49:45can see is there maybe seems like a
49:46little bit once you're uh rich it seems
49:50maybe it's a little you you stick a
49:52little bit closer to the 45 degree line
49:55although there's are some interesting
49:57examples like uh New Zealand which has
49:59fallen below uh you can see Venezuela
50:02here the the example I've talked about
50:05I'll say one thing when you always look
50:07at these data there's always some
50:09countries which strike you as wow I
50:11wonder what's going on there. And so
50:13when I was putting these slides
50:14together, it was the first time I
50:16noticed Trinidad and Tobago way up here.
50:19So in 1960, they were actually according
50:22to this GDP per person bigger than the
50:26United States. I have to confess I've
50:28never been to Trinidad and Tobago and
50:30don't know much about it, but um that
50:32kind of struck me as a little bit um
50:34surprising, but don't know what the
50:36story is there. Anyways, uh lots of
50:39persistence. uh also people countries
50:42that are making moves the so-called
50:45miracles and disasters [snorts] you can
50:47see that there maybe tends to be a
50:49little bit more disaster at the lower
50:52end. Uh now whether that's uh uh
50:57fundamental is a is a deeper question
51:00but uh that's a picture just to
51:02highlight persistence. Then this is
51:06again showing uh what happened in
51:08Western Europe. Uh I guess it's also got
51:11Japan on this picture and this just is
51:13just showing you the sense in which
51:15there was some convergence across these
51:19countries. uh because this is showing
51:21the position in 1960 and this is showing
51:25the growth rate post 1960
51:29you see the countries that were lower
51:31experienced higher growth rates now
51:33related to I'll just say one thing
51:35related to the earlier question uh
51:38what's one of the stories you would tell
51:40for why it is that these countries
51:42experienced uh such rapid growth uh in a
51:47sense to converge relative to these
51:49countries
51:50Uh so Japan I think is a is a separate
51:53case here but but I think one argument
51:55is the integration of Europe has had a
51:58big effect on a lot of these economies
52:00that bringing them they made changes in
52:03kind of institutions and policies to
52:06integrate into Europe and there were
52:07benefits associated with that. So that's
52:10a particular story. It doesn't say that
52:12that because you were at one quarter
52:15that guaranteed that you were going to
52:16converge. It's more there's actually
52:19something specific that happened here
52:20that was a part of the convergence. So
52:22it's it's not ordained that once you get
52:25to this level then everything is you're
52:28home free from that point on. Uh it just
52:31says that there are examples where we do
52:34see um some convergence. But this is the
52:37uh picture in the Chad Jones chapter
52:40that shows you for the whole set of
52:41countries. He's looking at the
52:43relationship between GDP per person in
52:471960 and growth rates. And what you see
52:50here is basically a scatter plot that
52:52there's there's nothing systematic
52:54between level of GDP in 1960 relative to
52:58the US and subsequent growth rates. So
53:01again emphasizing this point uh that if
53:04you're thinking the the the core cause
53:08of of the position in the distribution
53:11is because of your underlying
53:13differences in permanent growth rates
53:15that just doesn't kind of line up with
53:17what we're seeing in the post 1960
53:20period.
53:22So I showed you a bunch of pictures.
53:24We've already had a little bit of of
53:26discussion and now I kind of want to
53:27return to to the issue I talked about
53:30before levels versus growth rates and
53:32then I also want to talk about frontier
53:35growth versus catchup growth because I
53:37think these are are key in terms of how
53:39we think about things. [snorts] So uh
53:42levels versus growth rates too. So as I
53:46say when you look at the data from 1960
53:49on the last uh 60 years there's really
53:53no evidence for systematic ongoing
53:56spreading out of of levels that does not
53:58seem the the dominant dynamic in terms
54:01of of understanding that distribution.
54:04Instead we see again I'll say
54:07approximately constant growth across the
54:09distribution. That's that last picture I
54:11showed you from Chad Jones. Now, we know
54:14that if you go back in time, things are
54:17a little bit different. Uh, you know,
54:19when I showed you that picture that goes
54:21back to 1200, uh, shows you the
54:23spreading out between 1700 and and 1900.
54:28So, things were obviously a little bit
54:29different then. And I think to
54:32understand that, you have to understand
54:33that the countries began the process of
54:36growth at different times. It started in
54:39the UK, then spread to other countries.
54:42But I think we've now reached a point
54:44where where in some sense growth per se
54:47has has spread across the world. And so
54:51we have to think about that early period
54:53as as a in a different light than we
54:56think of of the last 60 years and in
54:58particular uh where we are now in the
55:00world. And the other thing I want to
55:02emphasize is is that all periods of
55:06sustained high growth were actually
55:08examples of catchup to the frontier. and
55:11and that growth actually decreases as
55:14distance to the frontier lessened. Uh,
55:17another way to say this is is that the
55:20the literature on on divergence and
55:23emphasizing permanent differences in
55:26growth rates kind of said that it's all
55:28about growth rates that we should just
55:30focus on the growth rates. uh and and I
55:34think what the data tells us is is that
55:37that's uh uh misguided to actually
55:42separate growth rates without thinking
55:44about levels that in fact there's no
55:47country that that experienced 7% growth
55:51or 8% growth without having a low level.
55:54Um so in fact the idea that the the
55:58growth rate is the primitive uh I think
56:00is and divorced from levels is actually
56:03a a misguided way to to view the
56:06literature. And so let me just kind of
56:10uh propose the following framework for
56:13kind of organizing the data. So this is
56:16I'm kind of jumping a little bit too far
56:18ahead in terms of the the structure
56:20here, but this is just I think as a as
56:24kind of an organizing device kind of
56:26useful. So think of each country as a
56:29one sector growth model and then there's
56:31going to be something I'm going to call
56:32the world technology frontier that I'm
56:35going to call a hat. uh think of this as
56:38like a technology parameter [snorts] and
56:41then each individual country has a
56:43parameter that index indexes its
56:46position relative to the frontier. I'm
56:48going to call that a tilda. This is
56:50going to be something between zero and
56:52one. So if your a tilda is equal to one,
56:55you're at the frontier. If your a tilda
56:57is equal to 0.03, you're a long ways
57:00away from the frontier. And so then as I
57:03say here TFP in country I at time t is
57:06is um just the product of these two. So
57:10take the frontier and where you are
57:12relative to the frontier.
57:14So a key fact of reality is that a hat
57:20the frontier changes slowly over time.
57:23the to first approximation you can kind
57:26of go back to the when modern growth
57:29started the industrial revolution and
57:31that frontier is kind of growing at
57:33something like 2% a year on the other
57:36hand a tilda can change rapidly uh as a
57:40country changes uh institutions or
57:44policies
57:46miracles and disasters reflect
57:48transition dynamics associated with
57:50large changes in the atilda
57:53And so the key point here is is there's
57:56a key distinction between frontier
58:00growth and and where you are relative to
58:03the frontier. Uh and miracles and
58:06disasters are about transition dynamics
58:09associated with your movement relative
58:12to the frontier. Now of course the the
58:14big question is is what are the things
58:16which introduce a a tilda. So this
58:19framework here does not tell us what
58:22causes them. If you read Asimog, Glue
58:25and Robinson and others, they they'll
58:26obviously uh try to persuade you that
58:29you can think of a tilda as a as a
58:32reduced form for thinking about the
58:34effective institutions. But even if you
58:38uh buy into that approach, the key
58:40question is exactly what what are those
58:42institutions that that are so important
58:44for Ailda? But what I want to emphasize
58:47here is this is kind of a organizing
58:49device for thinking about uh these
58:52movements across countries that there's
58:55something about where a country is
58:57relative to the frontier and that's the
58:59thing that can move rapidly in countries
59:02where we see miracle growth. It's
59:04because countries have changed their
59:06position relative to the frontier and
59:08the miracle growth is just the
59:10transition dynamics associated with that
59:13change in levels.
59:15So here's a picture for Korea that I
59:17mentioned earlier. Korea from 1960 to
59:21the present has kind of moved from 10%
59:24of the US to 70% of the US kind of
59:27massive uh improvement. Uh Korea has to
59:31a large extent kind of flattened out uh
59:34in in recent decades but they had of
59:36course this uh miracle period between
59:391970 and the uh and the time of the
59:42Asian crisis. This is the period for
59:45what uh growth rates looked like in
59:47Korea. This is log differences that you
59:49can read here. Uh so again, things kind
59:52of started uh big in the late60s. You
59:54can see they had growth rates on the
59:56order of of 10%. There's of course lots
59:59of uh v yeartoyear variation but the key
1:00:03point here is that as as Korea caught up
1:00:07moved towards the frontier they made
1:00:09these changes which were moving them
1:00:11from 0.1 to 7 what you saw is rapid
1:00:15growth and as they kind of leveled off
1:00:17the growth decreased here and this is
1:00:20exactly the types of dynamics you get
1:00:23out of the one sector growth model if
1:00:26you have a big increase in a and then
1:00:30you have transition dynamics associated
1:00:32with that and the period of high growth
1:00:35you the period of high growth is just a
1:00:38symptom or a cause of the fact that you
1:00:40made these changes which are are
1:00:42generating a change in the levels and
1:00:44say the key point here is is the the
1:00:48fundamental source of things is the
1:00:50changes in levels and the and the growth
1:00:53rates are just the transition dynamics
1:00:55associated with those changes in levels
1:00:58and so quite different than the uh
1:01:02divergence emphasis on permanent growth
1:01:04rates uh this framework that I'm
1:01:06describing here which is I think the the
1:01:09best way to think about the data
1:01:10emphasizes that we want to think about
1:01:13uh country specific institutions and
1:01:16policies which dictate their level
1:01:19relative to the frontier and it's it's
1:01:21as they change those policies they
1:01:24change their level relative to the
1:01:25frontier and that's how we should be
1:01:27thinking about the the dynamics.
1:01:30Can I ask a question about that?
1:01:32Sure.
1:01:34Um I guess thinking about like
1:01:36prescriptive prescriptions for policies,
1:01:39how like country specific do we want to
1:01:42be in saying like for this very specific
1:01:45just one country, this policy worked
1:01:47well versus this policy seems to work
1:01:50well for lots of different countries.
1:01:52Just how like deep do we want to get
1:01:55there?
1:01:56Uh well that's a excellent question and
1:02:00I'll I'll say I I don't know that I have
1:02:03a excellent answer for your excellent
1:02:05question but let me just say this what's
1:02:08the and if if you read some of what
1:02:11Lucas wrote I mean it kind of relates to
1:02:13your question where he said you know we
1:02:17see some countries have made these huge
1:02:19strides forward some countries have not
1:02:22and the question is is there something
1:02:24the economics profession
1:02:26can say to countries uh that are
1:02:31considering uh some sort of changes.
1:02:33What are the what's the one, two or
1:02:37three things you could do that would
1:02:39actually change your position uh you
1:02:42know uh lead to a a large increase in in
1:02:45your standard of living relative to the
1:02:48relative to the frontier. So certainly
1:02:50the the hope is that there's something
1:02:54systematic we can talk about about what
1:02:56are the things which are good for that
1:02:58as opposed to to not being good for
1:03:01that. But I would say that you know
1:03:03where we are as a profession right now
1:03:06is I don't think we have that much uh
1:03:09concrete to say aside from a little bit
1:03:12at a fairly general level. So again, if
1:03:14you read Asimoglu and Robinson, they
1:03:17talk about good institutions versus bad
1:03:20institutions. And they kind of talk
1:03:22about some of the features that they
1:03:24think are are part of of good
1:03:27institutions, uh, rule of law, property,
1:03:30rights, uh,
1:03:34uh, no confiscation, etc. So, so these
1:03:37are are are somewhat general uh
1:03:41features, but I think we're a little bit
1:03:43uh not there yet in terms of being able
1:03:47to say something more definitive.
1:03:52Thank you.
1:03:53May I ask you a question?
1:03:55Sure.
1:03:55So, is there a sense in which these
1:03:58country specific policies can interact
1:04:00with the outcomes in other countries? So
1:04:02if I say that the US has a policy to get
1:04:05like skilled immigrants from poor
1:04:08countries then this may affect
1:04:11you know how poor countries are doing
1:04:13right.
1:04:14Okay. Uh so I think you your question is
1:04:19kind of a I'll I'll say it's a little
1:04:20bit more general question because I I
1:04:24put forth this kind of framework here
1:04:27and again the framework that I laid out
1:04:29here looks a little bit like a closed
1:04:32economy uh framework and I I think
1:04:35you're I mean I interpret your question
1:04:38more broadly as you know don't we need
1:04:41to think about uh interactions in a
1:04:44world economy might those be important?
1:04:46And I and I certainly didn't with this
1:04:49framework I wrote down, I didn't mean to
1:04:52imply that we don't need to think about
1:04:54uh countries interacting and movements
1:04:57of capital and labor across countries.
1:04:59So, so think of this as as just kind of
1:05:02a first step in terms of organizing our
1:05:06thinking, but it's it's not the uh it's
1:05:09not the whole model with all of the sets
1:05:11of interactions we need to talk about.
1:05:14And and more generally, I would say
1:05:17you're going to hear lectures on this
1:05:19later on if you uh uh uh stick with with
1:05:22the course is, you know, here I' I've
1:05:27actually even overstepped my bounds
1:05:29today because I kind of put everything
1:05:31on TFP. uh the the next lecture in this
1:05:34series is going to talk more about uh uh
1:05:38development accounting and growth
1:05:39accounting and and understanding the
1:05:42differences in productivity across
1:05:44countries and and there's some
1:05:46interesting and open questions there.
1:05:49But once you start thinking about
1:05:51technology and and frontier technology I
1:05:54talked about you have to think about the
1:05:56the process by which technology diffuses
1:05:58across countries and and then you can
1:06:01start thinking about multinationals and
1:06:03where they locate and stuff. So so
1:06:05there's very much a sense in which one
1:06:07needs to have a a global perspective. So
1:06:10this is kind of just kind of uh zeroing
1:06:13in on kind of uh one snapshot inside
1:06:17that global perspective in a in a sense.
1:06:30Okay. So uh just to kind of uh go back
1:06:34to to where I left off. uh if you kind
1:06:38of buy into this as kind of a simple
1:06:41framework for starting to organize our
1:06:44thinking uh and I want to return to this
1:06:46issue about indogenous growth. Where
1:06:48does endogenous growth theory kind of is
1:06:51originally conceived kind of fit into
1:06:53this? And I would say endogenous growth
1:06:55theory is the theory we need to
1:06:58understand what's happening to the
1:07:00frontier. Uh so why is the frontier
1:07:04getting better at 2% uh a year? I say
1:07:07that's what endogenous growth theory is
1:07:09about. Innovation, new ideas, new
1:07:11techniques. That's kind of pushing up
1:07:13the frontier. And so the role of
1:07:15endogenous growth theory is to help us
1:07:18understand why the frontier is moving as
1:07:20it is. But from the perspective of
1:07:23development, the whole issue in
1:07:25development is understanding where
1:07:27countries are relative to the frontier.
1:07:30uh for a country that's close to the
1:07:32frontier, the only way you can have
1:07:35continued growth is if the frontier uh
1:07:38continues to move out. So for countries
1:07:40that are very advanced and near the
1:07:43frontier, frontier growth is kind of
1:07:45critical to their future growth. But for
1:07:48a country that's far from the frontier,
1:07:50the frontier growth is neither necessary
1:07:54nor sufficient for them to grow. they're
1:07:56they have all of this um scope for
1:07:59growth just by moving towards the
1:08:01frontier. And so when one wants to think
1:08:04about the
1:08:06uh frameworks for thinking about the
1:08:08cross-country variation, [clears throat]
1:08:10you have to focus on distance to the
1:08:13frontier and and in some as a first
1:08:16order matter, whether the frontier is
1:08:19going to go out in the next year is kind
1:08:22of uh virtually irrelevant. Um and on
1:08:26top of that I would say for most
1:08:28developing countries I mean China is
1:08:30maybe uh certainly a leading exception
1:08:32to this because China is is massively
1:08:35engaged in R&D and there's lots of and
1:08:38they're a large country. So I think
1:08:40China is in fact influencing having some
1:08:42impact on the frontier but for your
1:08:45typical small developing country uh the
1:08:48the movement in the frontier is
1:08:51basically exogenous to that country and
1:08:54so we need to think about uh the
1:08:56framework we need to think about is
1:08:58where countries are relative to the
1:09:00frontier. the frontier kind of gives you
1:09:02the set of things which are out there
1:09:05that are possible and and the challenges
1:09:07to understand why those why you're so
1:09:10far um why a particular country is so
1:09:13far from the frontier. And so just to
1:09:16kind of uh relate to to some other
1:09:20strands of literature, there's there's
1:09:22obviously a a literature trying to
1:09:25understand why the industrial revolution
1:09:28happened when it did. Why did it happen
1:09:30in the United Kingdom and not somewhere
1:09:32else? Um, that's all very interesting. I
1:09:36mean, I find that literature
1:09:37fascinating. I I I read it a lot and as
1:09:41I say, enjoy it. Um, but that literature
1:09:45is actually not gerine to thinking about
1:09:48uh a poor country today. Um you don't
1:09:51have to understand the industrial why
1:09:53the industrial revolution happened when
1:09:56it did where it did which is about
1:09:58frontier growth. Uh that's in to first
1:10:01approximation orthogonal to the issue of
1:10:04understanding a poor country today. They
1:10:06don't need an industrial revolution.
1:10:08They don't need to invent all these uh
1:10:11new things that have never been done.
1:10:13They actually just need to take
1:10:14advantage of things that have already
1:10:16been adopted. So it's uh uh again this
1:10:20distinction between uh distance to the
1:10:23frontier and and what happens at the
1:10:25frontier and I think this is kind of key
1:10:27in terms of a organizing framework for
1:10:30thinking about things and as I say uh
1:10:33you I I suggested as as the one question
1:10:36said that closed economy maybe is
1:10:39sufficient and I didn't mean to imply
1:10:41that um so you can think about distance
1:10:44to the frontier in a world with with
1:10:46many countries and allowing for rich set
1:10:49of interactions across um countries.
1:10:53So in the last few minutes before we
1:10:55kind of just turn to open questions I
1:10:58kind of want to talk about uh a little
1:11:01bit about stag and and where things are.
1:11:04Uh so as I say on this slide another
1:11:06great divergence and this is micro and
1:11:09macro development. Um, so much of what
1:11:12I've talked about today is is looking at
1:11:15aggregate data because I think the key
1:11:18facts are aggregate facts. There's no
1:11:20doubt about that. The the key fact is
1:11:22the differences in average living
1:11:24standards across countries, which is a
1:11:26an aggregate a macroeconomic fact.
1:11:29That's what we're after.
1:11:31But much of what I've described is is
1:11:34researchers thinking about what's the
1:11:37kind of right aggregate framework for
1:11:40organizing that and thinking about it.
1:11:42Uh and so all of that has been at a at a
1:11:45very aggregate level which is is an
1:11:47important step in the in the whole
1:11:50endeavor. But at the same time that all
1:11:52of that work has been going on kind of
1:11:54at a in terms of macro development,
1:11:57there's been a a huge increase in in
1:12:01work done in microdevelopment where all
1:12:03kinds of researchers uh going out uh
1:12:07doing field work in less developed
1:12:10countries in lots of different settings
1:12:12connect collecting rich data sets on the
1:12:15effect of there's massive literature now
1:12:18on RCTs which generates a huge amount of
1:12:22information about the effect of specific
1:12:25policies on particular outcomes in in in
1:12:28in specific settings. And I think this
1:12:31has led to, as I say here, another great
1:12:34divergence. And this is the divergence
1:12:36between micro and macro development
1:12:39research where macro development has has
1:12:43been very focused on some aggregate data
1:12:46and some aggregate frameworks to
1:12:48organize that data. And the micro
1:12:51literature has been uh in some ways
1:12:55divorced from the aggregate to actually
1:12:58uh go at at kind of the uh the the micro
1:13:04level and just talk about what's
1:13:06actually happening on the ground in
1:13:08these countries. And I think this is a
1:13:11uh sad state of affairs for development
1:13:14overall as a field. And I think I'm
1:13:16particularly excited about STG because I
1:13:18think STG is is trying to to make a
1:13:20change here. [snorts]
1:13:22And I want to say in some ways this is
1:13:24uh all very natural and this is actually
1:13:27happening uh in lots of other fields
1:13:30inside macro. So in fact [snorts] uh if
1:13:33if you follow the macro literature over
1:13:36the last 20 years I think without a
1:13:38doubt the the biggest change in
1:13:41macroeconomics in the last 20 years has
1:13:43been people moving from so-called
1:13:46representative agent models to models
1:13:49that are rich in terms of um
1:13:51heterogeneity in terms of market
1:13:53frictions. And kind of the promise of
1:13:56that literature is that we can
1:13:58simultaneously we can write down a model
1:14:01which can simultaneously
1:14:03connect with aggregate data and rich
1:14:06micro cross-sectional data. So we
1:14:08actually have the the hope of doing
1:14:11aggregate and kind of micro analysis
1:14:14inside coherent frameworks. And I think
1:14:17that's uh again that's what's been
1:14:19happening in macro for the last 20
1:14:21years. And if you look at you know
1:14:24particular sub areas inside macro you
1:14:26see this uh for for macroeconomists who
1:14:30study consumption they're now uh
1:14:33developing those models which can
1:14:34simultaneously talk about the
1:14:36cross-sectional facts about consumption
1:14:38and income at the same time they can
1:14:40talk about the aggregate of consumption
1:14:42and the aggregate of of income. You see
1:14:45this in terms of labor supply. You see
1:14:47this in terms of investment. You see
1:14:49this in terms of models of price setting
1:14:52where we're not just interested in what
1:14:53happens to the overall price level in
1:14:56the economy. We're worried about happens
1:14:58to the distribution of price changes. So
1:15:01macroeconomics has been developing all
1:15:03those models. [snorts] Uh and we're now
1:15:05I think in a position to do this also in
1:15:08the development context. uh develop
1:15:11those sorts of so fill in the pieces of
1:15:14these kind of aggregate structures to
1:15:17add the frictions the heterogeneity
1:15:20uh and this is actually I think critical
1:15:22to making headway because this is going
1:15:24to provide the link between the the on
1:15:28the ground fieldwork
1:15:31uh which talks about how individuals
1:15:34respond how firms respond in particular
1:15:36settings and and to kind of build those
1:15:39up into aggregate outcomes. Now I say
1:15:42here there's a recent paper by Paco
1:15:44Bora, Joe Kabosski, Rob Townsend uh
1:15:47which Joe tells me I think is available
1:15:50uh on at least one of their websites
1:15:52which I think is a great overview of
1:15:55this agenda which is going to develop
1:15:58aggregate models of development but
1:16:01which are rich in terms of the frictions
1:16:03and the heterogeneity
1:16:06and and that I think is is allows for
1:16:09the possibility that the micro and macro
1:16:12development agendas um converge and can
1:16:15speak to each other. And as I say, I
1:16:17think STAG is is actively promoting this
1:16:21agenda. And I think it's one of the most
1:16:23exciting things about STG. And again,
1:16:25it's there's kind of, you know, two
1:16:29uh ways to converge. You can start
1:16:32bottom up and try and aggregate. And you
1:16:34can take the top down, start aggregating
1:16:38and disagregating. and and both of those
1:16:40are going to prove useful. And the last
1:16:42thing I'll say which again very much
1:16:45relates to to STG is is the need to go
1:16:49beyond the one sector model. So I say
1:16:51macroeconomists uh love to use the one
1:16:54sector growth model as kind of a
1:16:56benchmark for organizing
1:16:58uh or thinking about aggregates for even
1:17:02developing uh models for thinking about
1:17:05the forces that affect some of those
1:17:06aggregates. But I think this model as we
1:17:09all know is is might be I always tell my
1:17:12students this is for many macroeconomic
1:17:14issues this might be the first word but
1:17:17it's never the last word. Another way to
1:17:20say that this is a model that has some
1:17:22some stark limitations despite it uh
1:17:25being useful. And I think in the
1:17:28development context, this is like
1:17:29particularly stark because if you use
1:17:32the one sector growth model as kind of
1:17:35your benchmark model for thinking about
1:17:38development, it kind of tells you that
1:17:40that development um rich countries and
1:17:44poor countries kind of are doing the
1:17:46same thing, but they just do it better
1:17:48in in rich countries. Um and I think
1:17:51that's kind of missing the essence of
1:17:54development. uh and in some ways you
1:17:57know STG uh structural transformation
1:18:00economic growth uh the the structural
1:18:04part there I think is key which is that
1:18:06the structure of economic activity is
1:18:09dramatically different between rich and
1:18:11poor economies and it's the structure of
1:18:13economic activity that changes. It's not
1:18:15just that you're doing the same thing
1:18:17and doing it better. You're actually
1:18:19doing different things. You're doing
1:18:21some of the same things differently. And
1:18:23so those changes in structure I think
1:18:25are at the essence of development and
1:18:28there's lots of pieces to that. One is
1:18:30in terms of just the the sectors of
1:18:33activities agriculture versus services
1:18:36versus manufacturing. There's going to
1:18:37be some lectures on that. Uh so that's
1:18:40one element but it's also just the
1:18:43organization of production uh can change
1:18:46how much happens inhouse versus
1:18:48outsourced in terms of intermediate
1:18:50inputs.
1:18:52uh what's the nature of production in
1:18:54terms of capital and labor skilled
1:18:57versus unskilled labor. So these changes
1:18:59in in structure I think are are
1:19:01critical. What's the industrial
1:19:03organization of activity? Is it small
1:19:05firms or big firms? So these are the
1:19:08changes in structure that I think are
1:19:10are critical to thinking about
1:19:12development and requires that we go
1:19:14beyond the one sector growth model to to
1:19:17build in these structural changes. And
1:19:19as I say, this is exactly the type of
1:19:21work that STG is is promoting and
1:19:24advancing. And so, uh, I'm very excited
1:19:26about that. And I see I've I've talked
1:19:30maybe a little bit longer than I plan
1:19:32to, but I'm going to stop there and just
1:19:34open up for some questions at this
1:19:36point.
1:19:38Hi. Uh, excuse me. Can you hear me?
1:19:42Sure. Yeah.
1:19:43I have a question. Uh I'm quite
1:19:45interested in the uh transitional like
1:19:48matrices that you you show about like
1:19:51how countries like moving around in the
1:19:53distribution. I think put in in the
1:19:56framework that you introduced earlier.
1:19:58It seems like it's interesting to to see
1:20:01like what policies or why some countries
1:20:03they're able to like make their a delta
1:20:06all all the way alone to to like reach
1:20:08the frontier. But for many or say a
1:20:11large amount of countries it's only like
1:20:14their catch-up growth is only stay at
1:20:16the specific level like reaching 0.5 and
1:20:19then kind of stays there and could you
1:20:21please explain a bit more like uh in the
1:20:24literature what people already know
1:20:26about what makes the differences of
1:20:28these two types of growth and uh what do
1:20:31we don't know still don't know about.
1:20:34Okay. uh the list of what we don't know
1:20:37is a very long list and the list of what
1:20:40we do know I'm afraid is a is a short
1:20:42list. Um so as I say I think this is
1:20:46what I kind of described I think in a
1:20:49very barebone sense is the right
1:20:51framework for organizing and then what
1:20:53you've asked is of course the key
1:20:56question we need to know what is it that
1:20:59countries do in terms of particular
1:21:03policies or institutions or arrangements
1:21:05or whatever that bundle is uh that maps
1:21:10into these atildas that I that I talked
1:21:13about and I think there again as I say I
1:21:16think Asimoglu and Robinson
1:21:20like to claim I think overclaim a little
1:21:23bit they they claim we now understand
1:21:26that because they say it's it's
1:21:27institutions but I think you know we can
1:21:31call it institutions but I don't think
1:21:33we actually know what what institutions
1:21:36even means exactly and so part of what
1:21:40in terms of the work that that uh people
1:21:43associated with STGA is doing and STAG
1:21:45is promoting is if if you make a a list
1:21:48of all the places where countries can
1:21:51can mess up with things. It's a very
1:21:54long list. They can mess up with credit
1:21:56markets. They can mess up with labor
1:21:58markets. They can misallocate resources.
1:22:01They can make it difficult for firms to
1:22:04adopt technology. And so the list of
1:22:08things where you can mess up is a very
1:22:10long list. And the the agenda in some
1:22:14ways is to figure out which are the
1:22:16things there which are most important.
1:22:19Which things can we actually get um
1:22:22build uh uh cases for that actually
1:22:26these are the things that countries did
1:22:28which actually uh had the big effect.
1:22:31Um, and that's kind of what the the
1:22:34research agenda is all about. And as I
1:22:36say, I I don't I think we have lots of
1:22:39bits of information and hints of
1:22:42different things which matter. And some
1:22:44of those you're going to hear about in
1:22:45future lectures. Um, but I think there's
1:22:49lots of open questions and things uh we
1:22:51just don't know about. I mean, another
1:22:53way to say it is, you know, a country
1:22:55that's a long ways away from the
1:22:57frontier, it seems like there's a lot of
1:22:59benefit from moving towards the
1:23:01frontier.
1:23:03What is it that's preventing that? What
1:23:05are the barriers? Um, that's what we're
1:23:08after. But if you make a list of the
1:23:11potential barriers, it's a long list.
1:23:13And as I say, lot there's lots of
1:23:16individual pieces of work which suggest
1:23:20pieces which are are relevant. As I say,
1:23:22some of that you're going to see in the
1:23:24upcoming lectures, but I think we're uh
1:23:28I'll say for the graduate students, you
1:23:30know, when there's big challenges,
1:23:32that's the most exciting thing for
1:23:34graduate students. So, I think we have
1:23:35lots of graduate students today. So,
1:23:37there's like uh huge opportunity for all
1:23:41of you to kind of figure this out uh
1:23:44move the profession forward and make the
1:23:46world a better place. So, I say this is
1:23:49great for graduate students. If I told
1:23:50you that we already know the answers,
1:23:53that would be really bad for you as
1:23:54graduate students because you say, "What
1:23:56am I going to work on if we already
1:23:58figured that out?" Turns out there's
1:23:59lots of work for you to do here. Uh lots
1:24:02of important work, big question. Uh and
1:24:05the tools are kind of there. So, so it's
1:24:07a great time to be a graduate student
1:24:09for thinking about macro development.
1:24:16If we have no if we have uh any other
1:24:18questions from the ask another question.
1:24:21Oh go ahead.
1:24:23Yeah. So um so we talk about the
1:24:25possibility of catching up for the less
1:24:28developed countries and in terms of the
1:24:30development strategies. Um I've heard of
1:24:32one kind of strategy which called big
1:24:34push uh strategies. How would you
1:24:37evaluate this kind of strategy as I
1:24:39guess it was some some were proposed by
1:24:42hers said in the year of 1950s or 60s
1:24:46has the literature reached any consensus
1:24:48on the idea of big push
1:24:52I mean I'll answer here I'll be happy to
1:24:56also let uh Doug and Joe who are if uh
1:24:59still on the call if they want to chime
1:25:01in and I I guess I would I think at a
1:25:04minimum there's kind two schools of
1:25:07thought on big push. I would say there
1:25:09there are still
1:25:11some economists who think the let's say
1:25:15some variant of big push is critical.
1:25:19Uh, and I'd say there's a whole bunch of
1:25:21other economists, uh, and I would put
1:25:24myself in the other camp that are a bit,
1:25:27I'll say, uh, skeptical that that big
1:25:30push is is kind of the, uh, the be all
1:25:33and end all, uh, in terms of thinking
1:25:35about things. I think in part you know
1:25:39part one of the reasons there's you know
1:25:41two two schools of thought on this I
1:25:44think is that in economics more broadly
1:25:47or in macroeconomics
1:25:49uh whether people believe there are
1:25:51multiple equilibria there tends to be
1:25:53kind of two schools of thought on that
1:25:56and uh at least one way of thinking
1:25:59about big push is that there are
1:26:02multiple equilibria and the reason you
1:26:04need a big push is to move from one
1:26:06equilibria to the other equilibria. And
1:26:08as I say, uh I don't think there's
1:26:12definitive evidence one way or the
1:26:14other. This depends, you know, different
1:26:16people kind of see the world a little
1:26:18bit differently. And uh I tend to to not
1:26:23be so much of a mind that there are
1:26:24multiple equilibria. I don't think that
1:26:28when Korea went from 0.1 to 7, I don't
1:26:31think that was uh multiple equilibria. I
1:26:34kind of think about that as you know
1:26:37there was changes in the Korean economy
1:26:39and and uh I don't think it was that uh
1:26:42nothing changed in some ways but they
1:26:45kind of just moved to this better
1:26:47equilibria. I don't think China's just
1:26:50moving to a better equilibria. I think
1:26:51China is like doing very real things.
1:26:54So, so I personally don't subscribe to
1:26:58the multiple equilibria view, but uh
1:27:01there still are some economists who do.
1:27:06May I jump in here?
1:27:08Sure.
1:27:09Yeah. So, I want to come back to the
1:27:12level versus growth uh point you made
1:27:15earlier. So
1:27:17the idea I think if I understood
1:27:19correctly is that once you change your
1:27:22levels the growth rates would be simply
1:27:24transition dynamics. Uh but I want to
1:27:27ask you to which extent the growth
1:27:30itself leads to the changes in in a you
1:27:34see what I mean like when you're growing
1:27:37you demand others other sort of sort of
1:27:40other changes in a that lead you to grow
1:27:43even more in and in this sense to which
1:27:45extent the growth could actually impact
1:27:47the level.
1:27:52Uh
1:27:55well, I I don't know if we're going to
1:27:57end up with something circular there. I
1:28:00mean,
1:28:02if if we think the fundamental change is
1:28:04a change in the level and then that
1:28:07change in the level, you know, so
1:28:10there's some fundamental change in a
1:28:12level and then there's some effects
1:28:15associated with that and there could be
1:28:17some additional effects associated with
1:28:20that. uh but I wouldn't say that it's
1:28:24the growth affecting level. I would say
1:28:27this is all you know the the the
1:28:29fundamental change is is some change
1:28:32let's call it institutions just to make
1:28:35it use a word that you change the
1:28:38institutions and those change in
1:28:40institutions may have a series of
1:28:43effects and there might be a direct
1:28:45effect and there might be some some some
1:28:48secondord indirect effects uh
1:28:52uh but it's a fundamental change in the
1:28:55level which just putting all of that in
1:28:57motion is but that's that's just again
1:29:01it this is all a little bit abstract in
1:29:04terms of the particular framework. Um
1:29:09but
1:29:09yeah my thought was just like the growth
1:29:11theory is the the growth rate changing
1:29:14it means you're it it sort of requires
1:29:17changes in the institutions that are a
1:29:20result of the growth of the income level
1:29:23but I get I I I hear I get your point.
1:29:26Yeah.
1:29:28We are unfortunately out of time. Uh
1:29:30Richard I want to thank you for a
1:29:32wonderful lecture. I sure I speak for
1:29:35there's many people that have said the
1:29:37same in the in the chat window. Um
1:29:39Richard will be giving another lecture
1:29:41in a couple weeks um on uh
1:29:44misallocation.
1:29:46Um and then uh next week we will have
1:29:48Juliet Kaneedo talking about development
1:29:51accounting.
1:29:53Um so we look forward to that and uh
1:29:56thank you all for attending.
1:30:01Okay. Thank you everybody.
1:30:03I think we want to particularly thank
1:30:05the CR team [clears throat] who have who
1:30:09have managed the hosting and uh we look
1:30:12forward to their help in all the future
1:30:14lectures.

About this transcript

This page contains the full transcript of Lecture 1: Empirical overview of macro development by CEPR & VideoVox Economics, generated from the public captions YouTube serves with the video. The transcript has 14,230 words across 2,051 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.

What you can do with it

Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to choose a caption track available on the video.

Free YouTube transcript tool

YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, caption language selection, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.