WEBVTT 00:00:00.000 --> 00:00:07.380 align:middle line:84% [SQUEAKING] [CLICKING] [RUSTLING] [CLICKING] 00:00:07.380 --> 00:00:19.730 align:middle line:90% 00:00:19.730 --> 00:00:21.260 align:middle line:90% ESTHER DUFLO: I am Esther Duflo. 00:00:21.260 --> 00:00:24.650 align:middle line:84% And this is Ben Olken, who is going to quickly introduce 00:00:24.650 --> 00:00:26.120 align:middle line:90% himself in a minute. 00:00:26.120 --> 00:00:29.060 align:middle line:84% This is part of a two-course sequence 00:00:29.060 --> 00:00:35.960 align:middle line:84% on what we think are the core things to know in development 00:00:35.960 --> 00:00:38.060 align:middle line:90% economics. 00:00:38.060 --> 00:00:39.982 align:middle line:84% I'll describe a little bit the full syllabus 00:00:39.982 --> 00:00:41.690 align:middle line:84% in the course of this lecture, so I'm not 00:00:41.690 --> 00:00:43.610 align:middle line:90% going to say more just now. 00:00:43.610 --> 00:00:46.370 align:middle line:84% But it's basically a sequence that we've 00:00:46.370 --> 00:00:48.500 align:middle line:84% worked on, developed, over the years, 00:00:48.500 --> 00:00:55.490 align:middle line:84% between Abhijit Banerjee, Rob Townsend, David Atkin, Ben, 00:00:55.490 --> 00:00:56.450 align:middle line:90% and myself. 00:00:56.450 --> 00:01:01.280 align:middle line:84% So we are super excited to teach it-- happy to have all of you 00:01:01.280 --> 00:01:01.850 align:middle line:90% here. 00:01:01.850 --> 00:01:05.072 align:middle line:84% And Ben will say a few words before he has to head out. 00:01:05.072 --> 00:01:05.780 align:middle line:90% BEN OLKEN: Great. 00:01:05.780 --> 00:01:09.800 align:middle line:84% I think I have to stand next to you, to use the microphone. 00:01:09.800 --> 00:01:11.010 align:middle line:90% It's great to see all of you. 00:01:11.010 --> 00:01:14.390 align:middle line:84% And it's terrific to be back here in person. 00:01:14.390 --> 00:01:16.640 align:middle line:84% For those who don't know me, I'm Ben Olken. 00:01:16.640 --> 00:01:19.010 align:middle line:84% I'm a development economist here at MIT. 00:01:19.010 --> 00:01:24.980 align:middle line:84% And my own areas of research are primarily, in the last 10 years 00:01:24.980 --> 00:01:26.195 align:middle line:90% or so, on the public sector-- 00:01:26.195 --> 00:01:28.070 align:middle line:84% the roles of the public sector in development 00:01:28.070 --> 00:01:30.498 align:middle line:84% and understanding the different ways in which we think 00:01:30.498 --> 00:01:32.540 align:middle line:84% about public-sector issues in developing context. 00:01:32.540 --> 00:01:34.520 align:middle line:84% So I'm going to be teaching a variety of topics 00:01:34.520 --> 00:01:35.000 align:middle line:90% this semester. 00:01:35.000 --> 00:01:36.667 align:middle line:84% I'm sure Esther will go over them today. 00:01:36.667 --> 00:01:40.865 align:middle line:84% But that will be what I'm covering, at the end. 00:01:40.865 --> 00:01:42.740 align:middle line:84% I think that's enough for now, and I'm really 00:01:42.740 --> 00:01:46.120 align:middle line:84% looking forward to seeing all of you this semester. 00:01:46.120 --> 00:01:49.530 align:middle line:84% ESTHER DUFLO: So maybe a little practicality, 00:01:49.530 --> 00:01:52.050 align:middle line:84% just getting the practicality out of the way-- 00:01:52.050 --> 00:01:55.290 align:middle line:90% this is a class that-- 00:01:55.290 --> 00:01:56.110 align:middle line:90% how does it work? 00:01:56.110 --> 00:01:58.530 align:middle line:84% It works in a fairly classical way. 00:01:58.530 --> 00:02:00.930 align:middle line:90% We are going to-- 00:02:00.930 --> 00:02:05.460 align:middle line:84% you have received the syllabus, on the Canvas website. 00:02:05.460 --> 00:02:08.350 align:middle line:84% The Canvas, I think, should be open to the entire community. 00:02:08.350 --> 00:02:11.039 align:middle line:84% So, for the time being, anybody who is here, registered or not, 00:02:11.039 --> 00:02:12.870 align:middle line:84% you can have access to the Canvas. 00:02:12.870 --> 00:02:15.000 align:middle line:84% As the enrollment stabilize, people 00:02:15.000 --> 00:02:19.470 align:middle line:84% know where they, we are going to lock it to the membership only. 00:02:19.470 --> 00:02:23.190 align:middle line:84% But, for now, anybody who is here, or even who is not, 00:02:23.190 --> 00:02:27.420 align:middle line:84% should have access to the syllabus. 00:02:27.420 --> 00:02:29.830 align:middle line:84% I'll then describe a bit the syllabus. 00:02:29.830 --> 00:02:32.400 align:middle line:84% But otherwise, we'll have lectures 00:02:32.400 --> 00:02:36.600 align:middle line:84% on Mondays and Wednesdays and a recitation once a week. 00:02:36.600 --> 00:02:40.560 align:middle line:84% In the lecture, it's going to be-- so, about once a week, 00:02:40.560 --> 00:02:44.130 align:middle line:84% we'll ask you to read a paper before the lecture. 00:02:44.130 --> 00:02:51.210 align:middle line:84% And there are about 24 to 48 hours before the lecture, 00:02:51.210 --> 00:02:55.200 align:middle line:84% to write down on the Piazza website 00:02:55.200 --> 00:03:00.120 align:middle line:84% a comment or a question that you have about the paper. 00:03:00.120 --> 00:03:02.770 align:middle line:84% This is something we have found useful-- 00:03:02.770 --> 00:03:04.530 align:middle line:84% first of all, as a commitment device, 00:03:04.530 --> 00:03:06.480 align:middle line:84% to make sure that you're actually reading the papers. 00:03:06.480 --> 00:03:07.710 align:middle line:84% It's like you have to write the comment. 00:03:07.710 --> 00:03:10.140 align:middle line:84% [INAUDIBLE] because you're doing that a little bit 00:03:10.140 --> 00:03:11.280 align:middle line:90% before the class. 00:03:11.280 --> 00:03:14.190 align:middle line:84% I will have read the comments, and then I 00:03:14.190 --> 00:03:16.990 align:middle line:84% can incorporate the comments and the question when I teach, 00:03:16.990 --> 00:03:21.591 align:middle line:84% which is a first way to get your input-- 00:03:21.591 --> 00:03:23.490 align:middle line:84% in addition, of course, to anything that you 00:03:23.490 --> 00:03:26.970 align:middle line:90% would want to say in class. 00:03:26.970 --> 00:03:29.970 align:middle line:84% Other than that, we'll have a problem set 00:03:29.970 --> 00:03:32.840 align:middle line:84% that are going over in particular the empirical 00:03:32.840 --> 00:03:33.340 align:middle line:90% method. 00:03:33.340 --> 00:03:36.330 align:middle line:84% So this is a class that's focused a lot obviously 00:03:36.330 --> 00:03:40.290 align:middle line:84% on content and how poor people live their lives 00:03:40.290 --> 00:03:42.450 align:middle line:90% and how to understand them. 00:03:42.450 --> 00:03:45.660 align:middle line:84% But also, as we go along on methods, 00:03:45.660 --> 00:03:48.340 align:middle line:84% a little bit like the labor class, 00:03:48.340 --> 00:03:49.920 align:middle line:84% which is both about labor markets 00:03:49.920 --> 00:03:52.740 align:middle line:84% and about methods in labor economics. 00:03:52.740 --> 00:03:55.950 align:middle line:84% In particular, this first sequence is, at the same time 00:03:55.950 --> 00:03:57.690 align:middle line:84% I'm going to introduce you-- especially 00:03:57.690 --> 00:04:00.150 align:middle line:84% in the first half of the class that I'll be teaching, 00:04:00.150 --> 00:04:02.880 align:middle line:84% I'm going to introduce you, as I go along, 00:04:02.880 --> 00:04:06.270 align:middle line:84% to the workhorse methods in empirical economics 00:04:06.270 --> 00:04:08.480 align:middle line:90% that we'll use. 00:04:08.480 --> 00:04:10.860 align:middle line:84% So, for some of you, if you're second-year or third-year, 00:04:10.860 --> 00:04:13.200 align:middle line:90% that's going to be review. 00:04:13.200 --> 00:04:15.370 align:middle line:84% For some of you, it's going to be brand-new, 00:04:15.370 --> 00:04:18.060 align:middle line:84% but it's taught at a level that, in principle, everybody 00:04:18.060 --> 00:04:22.380 align:middle line:84% should find something stimulating or interesting 00:04:22.380 --> 00:04:25.810 align:middle line:84% and you shouldn't get lost too much. 00:04:25.810 --> 00:04:28.770 align:middle line:84% And if you are, then you should make sure to talk to Ed 00:04:28.770 --> 00:04:29.520 align:middle line:90% or talk to us. 00:04:29.520 --> 00:04:31.740 align:middle line:84% Ed is, going to, in the recitation, 00:04:31.740 --> 00:04:35.370 align:middle line:84% emphasize some of the point-- go over some papers in more 00:04:35.370 --> 00:04:36.390 align:middle line:90% details. 00:04:36.390 --> 00:04:39.750 align:middle line:84% In particular, one thing we really 00:04:39.750 --> 00:04:43.350 align:middle line:84% would like to get out of the first and second years 00:04:43.350 --> 00:04:47.490 align:middle line:84% is not only that you have read content and you know papers 00:04:47.490 --> 00:04:50.160 align:middle line:84% but that you have a sense of how one builds 00:04:50.160 --> 00:04:52.740 align:middle line:84% a paper, one builds a results a research agenda, 00:04:52.740 --> 00:04:54.900 align:middle line:84% one constructs an empirical specification, 00:04:54.900 --> 00:04:58.990 align:middle line:84% what sense it makes, and so on and so forth. 00:04:58.990 --> 00:05:01.980 align:middle line:84% So in terms of requirement, besides coming to class, 00:05:01.980 --> 00:05:05.820 align:middle line:84% participating, et cetera, we'll have a few problem sets 00:05:05.820 --> 00:05:09.930 align:middle line:84% that helps you where you'll get your hands dirty on data 00:05:09.930 --> 00:05:10.860 align:middle line:90% and some such. 00:05:10.860 --> 00:05:14.730 align:middle line:84% Hopefully they are interesting and not make-work. 00:05:14.730 --> 00:05:19.740 align:middle line:84% Then we'll have a replication or an empirical exercise 00:05:19.740 --> 00:05:23.826 align:middle line:90% and a research proposal that-- 00:05:23.826 --> 00:05:26.280 align:middle line:84% you know, you can start thinking about ideas 00:05:26.280 --> 00:05:28.530 align:middle line:90% that you can take forward-- 00:05:28.530 --> 00:05:32.050 align:middle line:84% assuming you want to continue in this field. 00:05:32.050 --> 00:05:36.870 align:middle line:84% This class is really for anybody [LAUGHS] who is interested 00:05:36.870 --> 00:05:37.560 align:middle line:90% in getting-- 00:05:37.560 --> 00:05:39.750 align:middle line:84% so first of all, of course, anybody who wants 00:05:39.750 --> 00:05:42.300 align:middle line:90% to do development as a field. 00:05:42.300 --> 00:05:46.080 align:middle line:84% But I also think that it's a good class for any economist 00:05:46.080 --> 00:05:51.390 align:middle line:84% to take, because the development issues are cross-cutting 00:05:51.390 --> 00:05:52.980 align:middle line:90% across all fields. 00:05:52.980 --> 00:05:55.350 align:middle line:84% So of all the field of economics, 00:05:55.350 --> 00:05:58.890 align:middle line:84% development is the one that is the most 00:05:58.890 --> 00:06:01.860 align:middle line:90% wide tent, where people-- 00:06:01.860 --> 00:06:05.010 align:middle line:84% You might think that, oh, at MIT, we associate development 00:06:05.010 --> 00:06:07.770 align:middle line:84% with [INAUDIBLE],, but that's not true. 00:06:07.770 --> 00:06:12.750 align:middle line:84% Development economics research use a whole range application 00:06:12.750 --> 00:06:15.420 align:middle line:84% of methods and a whole range of insights, 00:06:15.420 --> 00:06:17.250 align:middle line:84% from behavioral, from public finance, 00:06:17.250 --> 00:06:20.430 align:middle line:90% from labor, from macro-- 00:06:20.430 --> 00:06:21.810 align:middle line:90% econometrics, obviously. 00:06:21.810 --> 00:06:26.670 align:middle line:84% There is really not a cultural tradition 00:06:26.670 --> 00:06:27.810 align:middle line:90% of development economics. 00:06:27.810 --> 00:06:29.730 align:middle line:84% It's super open, which you will also 00:06:29.730 --> 00:06:31.290 align:middle line:84% find if you continue in the field, 00:06:31.290 --> 00:06:33.623 align:middle line:84% as you go to conference, that there are papers going all 00:06:33.623 --> 00:06:36.000 align:middle line:84% around the place, which means that, 00:06:36.000 --> 00:06:39.590 align:middle line:84% as a young economist, as a budding economist, 00:06:39.590 --> 00:06:43.220 align:middle line:84% what you will find in the class is application of everything 00:06:43.220 --> 00:06:47.870 align:middle line:84% else that you're learning in other classes 00:06:47.870 --> 00:06:52.460 align:middle line:84% to the topic of how the poorest people live their lives. 00:06:52.460 --> 00:06:55.790 align:middle line:84% And you might wonder why it's an interesting topic, why 00:06:55.790 --> 00:06:58.040 align:middle line:84% it's an important topic, why we should think about it. 00:06:58.040 --> 00:07:00.373 align:middle line:84% But maybe you don't wonder, because you're already here. 00:07:00.373 --> 00:07:03.320 align:middle line:84% But it's always worth thinking of-- 00:07:03.320 --> 00:07:05.840 align:middle line:90% I like to start, every year-- 00:07:05.840 --> 00:07:07.850 align:middle line:84% regardless of whether I teach or not-- 00:07:07.850 --> 00:07:11.480 align:middle line:84% to remind me of the enormity of the fact. 00:07:11.480 --> 00:07:15.440 align:middle line:84% So in the handout that Ed distributed, 00:07:15.440 --> 00:07:21.470 align:middle line:84% you will find the gross-national-income tables 00:07:21.470 --> 00:07:24.500 align:middle line:84% produced by the World Bank before the COVID-19 crisis. 00:07:24.500 --> 00:07:29.240 align:middle line:84% I will start talking about the COVID-19 crisis in a minute. 00:07:29.240 --> 00:07:31.730 align:middle line:84% They use gross national income as opposed to GDP. 00:07:31.730 --> 00:07:36.380 align:middle line:84% Do you happen to know the difference between the two? 00:07:36.380 --> 00:07:37.940 align:middle line:90% You go ahead. 00:07:37.940 --> 00:07:40.790 align:middle line:84% AUDIENCE: Gross national income is 00:07:40.790 --> 00:07:42.750 align:middle line:84% that produced by residents of the country, 00:07:42.750 --> 00:07:44.553 align:middle line:90% regardless of where they are. 00:07:44.553 --> 00:07:45.470 align:middle line:90% ESTHER DUFLO: Exactly. 00:07:45.470 --> 00:07:50.300 align:middle line:84% It's income that residents receive, 00:07:50.300 --> 00:07:52.880 align:middle line:84% no matter where they received it. 00:07:52.880 --> 00:07:54.560 align:middle line:84% So that's a slightly different measure, 00:07:54.560 --> 00:07:57.410 align:middle line:90% but it's relatively similar. 00:07:57.410 --> 00:08:01.670 align:middle line:84% And maybe it's the better view of standard of living. 00:08:01.670 --> 00:08:04.640 align:middle line:90% Then it's given in PPP. 00:08:04.640 --> 00:08:08.210 align:middle line:84% So, just a reminder of what's the PPP adjustment-- 00:08:08.210 --> 00:08:11.215 align:middle line:84% what does PPP stands for, and how do we-- 00:08:11.215 --> 00:08:13.090 align:middle line:84% AUDIENCE: [INAUDIBLE] purchasing [? power? ?] 00:08:13.090 --> 00:08:13.930 align:middle line:90% ESTHER DUFLO: Yep. 00:08:13.930 --> 00:08:16.850 align:middle line:84% AUDIENCE: And it's an adjustment for how expensive 00:08:16.850 --> 00:08:19.733 align:middle line:84% a particular type of [? goods ?] is. 00:08:19.733 --> 00:08:20.650 align:middle line:90% ESTHER DUFLO: Exactly. 00:08:20.650 --> 00:08:24.460 align:middle line:84% So it's more expensive to get a haircut in New York 00:08:24.460 --> 00:08:26.440 align:middle line:90% than in Delhi. 00:08:26.440 --> 00:08:29.110 align:middle line:84% That's the extreme version, because it's only labor. 00:08:29.110 --> 00:08:30.940 align:middle line:84% But it's more expensive in general 00:08:30.940 --> 00:08:34.929 align:middle line:84% to live your life in Norway than it is in France-- and in France 00:08:34.929 --> 00:08:38.470 align:middle line:90% than it is in Burundi. 00:08:38.470 --> 00:08:41.929 align:middle line:84% So that's the [? effort. ?] I should emphasize that-- 00:08:41.929 --> 00:08:45.070 align:middle line:84% so how do you think we do that-- or they do that-- 00:08:45.070 --> 00:08:46.030 align:middle line:90% the PPP adjustment? 00:08:46.030 --> 00:08:48.255 align:middle line:84% Because once we've said, oh, it's PPP-adjusted, 00:08:48.255 --> 00:08:50.380 align:middle line:84% I think it's useful to think for a minute about how 00:08:50.380 --> 00:08:53.720 align:middle line:90% one would do such a thing. 00:08:53.720 --> 00:08:55.210 align:middle line:84% AUDIENCE: Do they send people out 00:08:55.210 --> 00:08:58.485 align:middle line:84% to basically try to assess if there's a set of goods that 00:08:58.485 --> 00:09:02.540 align:middle line:84% could be considered the same quality, 00:09:02.540 --> 00:09:05.320 align:middle line:84% across many different countries and that determines the price? 00:09:05.320 --> 00:09:06.070 align:middle line:90% ESTHER DUFLO: Yes. 00:09:06.070 --> 00:09:08.040 align:middle line:90% So let's unpack that. 00:09:08.040 --> 00:09:10.830 align:middle line:84% So [? Racquel ?] is saying you send people to find goods 00:09:10.830 --> 00:09:12.420 align:middle line:90% that have the same quality. 00:09:12.420 --> 00:09:13.710 align:middle line:90% First of all, you need to-- 00:09:13.710 --> 00:09:17.160 align:middle line:90% which goods would you-- 00:09:17.160 --> 00:09:18.360 align:middle line:90% one might pick? 00:09:18.360 --> 00:09:19.845 align:middle line:90% Which goods might one pick? 00:09:19.845 --> 00:09:23.570 align:middle line:90% 00:09:23.570 --> 00:09:25.610 align:middle line:90% AUDIENCE: Like, food often? 00:09:25.610 --> 00:09:28.520 align:middle line:84% ESTHER DUFLO: So what you want-- yeah, food, but what you want 00:09:28.520 --> 00:09:30.410 align:middle line:90% is basically a basket. 00:09:30.410 --> 00:09:34.383 align:middle line:84% And you want a basket that people consume. 00:09:34.383 --> 00:09:36.050 align:middle line:84% And now you immediately have the problem 00:09:36.050 --> 00:09:39.260 align:middle line:84% that people do not consume the same basket 00:09:39.260 --> 00:09:42.090 align:middle line:84% in Burundi and in Norway in the first place. 00:09:42.090 --> 00:09:43.933 align:middle line:84% So the entire exercise-- it sounds 00:09:43.933 --> 00:09:46.100 align:middle line:84% very easy to say "we adjust for standard of living," 00:09:46.100 --> 00:09:48.700 align:middle line:90% but in fact it's very tricky. 00:09:48.700 --> 00:09:53.140 align:middle line:84% Because this is the [INAUDIBLE] indices of inflation 00:09:53.140 --> 00:09:58.120 align:middle line:84% if it rings some bell from undergraduate economics. 00:09:58.120 --> 00:10:02.260 align:middle line:84% But very, very on steroids, in a sense, 00:10:02.260 --> 00:10:04.730 align:middle line:90% that problem is, which is-- 00:10:04.730 --> 00:10:07.720 align:middle line:84% In fact, it's very nice to think that one could do PPP 00:10:07.720 --> 00:10:10.090 align:middle line:84% adjustment, but one cannot very easily, 00:10:10.090 --> 00:10:13.340 align:middle line:84% because the baskets are not even the same. 00:10:13.340 --> 00:10:16.330 align:middle line:84% And then there is the question of the quality 00:10:16.330 --> 00:10:17.780 align:middle line:90% which [? Racquel ?] emphasized. 00:10:17.780 --> 00:10:19.780 align:middle line:84% And then there is a question of actually getting 00:10:19.780 --> 00:10:21.100 align:middle line:90% hold of the prices. 00:10:21.100 --> 00:10:23.500 align:middle line:84% In rich countries, there are scanners and people 00:10:23.500 --> 00:10:24.130 align:middle line:90% use scanners. 00:10:24.130 --> 00:10:25.930 align:middle line:84% In poor countries, you have indeed 00:10:25.930 --> 00:10:28.240 align:middle line:90% to send people out to markets. 00:10:28.240 --> 00:10:30.580 align:middle line:84% Where you send the people to the market 00:10:30.580 --> 00:10:32.050 align:middle line:90% is making a huge difference. 00:10:32.050 --> 00:10:34.240 align:middle line:84% So if you're only sending them to the cities, 00:10:34.240 --> 00:10:36.037 align:middle line:84% you get the prices in the cities and you 00:10:36.037 --> 00:10:37.120 align:middle line:90% get the goods in the city. 00:10:37.120 --> 00:10:41.980 align:middle line:84% But you go to the countryside, you go far from the city, 00:10:41.980 --> 00:10:43.340 align:middle line:90% the goods are not the same. 00:10:43.340 --> 00:10:45.370 align:middle line:84% Some goods are going to be entirely missing. 00:10:45.370 --> 00:10:47.360 align:middle line:90% And the prices are not the same. 00:10:47.360 --> 00:10:49.840 align:middle line:84% You're very far from the place where 00:10:49.840 --> 00:10:51.610 align:middle line:84% the things are being produced or shipped, 00:10:51.610 --> 00:10:55.060 align:middle line:90% and so the prices have gone up. 00:10:55.060 --> 00:10:57.710 align:middle line:84% Or maybe the demand is very low, so the prices have gone down. 00:10:57.710 --> 00:10:58.210 align:middle line:90% Who knows? 00:10:58.210 --> 00:11:01.120 align:middle line:84% But all that said, it's not the same. 00:11:01.120 --> 00:11:04.200 align:middle line:84% And so there is a project at the bank 00:11:04.200 --> 00:11:06.960 align:middle line:84% called the International Price Comparison Project, 00:11:06.960 --> 00:11:08.490 align:middle line:84% which is a big team of people which 00:11:08.490 --> 00:11:10.900 align:middle line:90% are doing that all the time. 00:11:10.900 --> 00:11:18.990 align:middle line:84% And from time to time, they give you revised PPP adjustments. 00:11:18.990 --> 00:11:21.870 align:middle line:84% And then they immediately get under fire 00:11:21.870 --> 00:11:25.230 align:middle line:90% for not having done it right. 00:11:25.230 --> 00:11:27.480 align:middle line:84% And it's not completely anecdotal, 00:11:27.480 --> 00:11:30.150 align:middle line:84% because it becomes very, very political, 00:11:30.150 --> 00:11:33.240 align:middle line:84% how many poor people there are in a particular country. 00:11:33.240 --> 00:11:37.290 align:middle line:84% At some moments, countries want to have a lot of poor people, 00:11:37.290 --> 00:11:39.383 align:middle line:84% because that helps them in getting assistance. 00:11:39.383 --> 00:11:40.800 align:middle line:84% And some people, they want to show 00:11:40.800 --> 00:11:42.550 align:middle line:84% that they have eliminated poverty, so they 00:11:42.550 --> 00:11:43.590 align:middle line:90% want to get rid of the-- 00:11:43.590 --> 00:11:47.790 align:middle line:84% So the moment these price lists come up, 00:11:47.790 --> 00:11:50.220 align:middle line:84% there is a big political hullabaloo. 00:11:50.220 --> 00:11:51.830 align:middle line:84% So all that to say that, when we look 00:11:51.830 --> 00:11:54.330 align:middle line:84% at [LAUGHS] the piece of paper-- just on this piece of paper 00:11:54.330 --> 00:11:56.790 align:middle line:84% of [? the ?] gross national income, 00:11:56.790 --> 00:11:58.530 align:middle line:84% there is a lot of assumption behind. 00:11:58.530 --> 00:12:01.830 align:middle line:84% There is a lot of politics behind. 00:12:01.830 --> 00:12:03.370 align:middle line:84% And you can make sure, for example, 00:12:03.370 --> 00:12:05.215 align:middle line:84% that Angus Deaton will go after you for not 00:12:05.215 --> 00:12:06.090 align:middle line:90% having done it right. 00:12:06.090 --> 00:12:08.820 align:middle line:84% But they are doing the best they can. 00:12:08.820 --> 00:12:14.020 align:middle line:90% And they produce this synthesis. 00:12:14.020 --> 00:12:15.350 align:middle line:90% And we are using them. 00:12:15.350 --> 00:12:18.740 align:middle line:84% So, conceptually, we are trying to say that it adjusts 00:12:18.740 --> 00:12:20.810 align:middle line:90% for the price of living. 00:12:20.810 --> 00:12:23.180 align:middle line:84% With all of that said, what's the richest country 00:12:23.180 --> 00:12:29.200 align:middle line:84% in the world, according to those tables? 00:12:29.200 --> 00:12:31.405 align:middle line:84% Or richest few countries in the world. 00:12:31.405 --> 00:12:43.372 align:middle line:90% 00:12:43.372 --> 00:12:45.980 align:middle line:90% AUDIENCE: Singapore? 00:12:45.980 --> 00:12:47.463 align:middle line:90% ESTHER DUFLO: Singapore, maybe? 00:12:47.463 --> 00:12:49.130 align:middle line:84% I think there is one that's even richer. 00:12:49.130 --> 00:12:49.640 align:middle line:90% Sorry? 00:12:49.640 --> 00:12:51.068 align:middle line:90% AUDIENCE: [INAUDIBLE] 00:12:51.068 --> 00:12:53.110 align:middle line:84% ESTHER DUFLO: Yeah, that's definitely in the top. 00:12:53.110 --> 00:12:56.500 align:middle line:90% 00:12:56.500 --> 00:12:58.570 align:middle line:84% I should have the piece of paper, to verify. 00:12:58.570 --> 00:13:03.560 align:middle line:90% 00:13:03.560 --> 00:13:09.410 align:middle line:84% I think it might be Macau, with a gross national income 00:13:09.410 --> 00:13:14.030 align:middle line:90% per capita of $123,000. 00:13:14.030 --> 00:13:16.850 align:middle line:84% And then Singapore, Qatar, Bermuda, Luxembourg, 00:13:16.850 --> 00:13:19.650 align:middle line:90% Switzerland come close, below. 00:13:19.650 --> 00:13:24.300 align:middle line:84% Now, of course, what's wrong with these countries, all 00:13:24.300 --> 00:13:29.762 align:middle line:84% of these countries, really, in terms of thinking 00:13:29.762 --> 00:13:30.845 align:middle line:90% of the standard of living? 00:13:30.845 --> 00:13:34.480 align:middle line:90% 00:13:34.480 --> 00:13:34.980 align:middle line:90% Yep. 00:13:34.980 --> 00:13:37.080 align:middle line:84% AUDIENCE: They all consist of essentially one 00:13:37.080 --> 00:13:37.950 align:middle line:90% or two large cities? 00:13:37.950 --> 00:13:38.992 align:middle line:90% ESTHER DUFLO: Yeah, yeah. 00:13:38.992 --> 00:13:44.550 align:middle line:84% And they are [INAUDIBLE],, like tax shelters and [INAUDIBLE].. 00:13:44.550 --> 00:13:45.540 align:middle line:90% Maybe that's not real. 00:13:45.540 --> 00:13:48.060 align:middle line:90% 00:13:48.060 --> 00:13:51.480 align:middle line:84% So next, if we remove them, who comes next? 00:13:51.480 --> 00:13:52.920 align:middle line:90% Sort of a real country. 00:13:52.920 --> 00:14:07.410 align:middle line:90% 00:14:07.410 --> 00:14:07.910 align:middle line:90% Oops. 00:14:07.910 --> 00:14:20.420 align:middle line:90% 00:14:20.420 --> 00:14:21.860 align:middle line:84% AUDIENCE: Is it the United States? 00:14:21.860 --> 00:14:23.510 align:middle line:90% ESTHER DUFLO: Yeah, not bad. 00:14:23.510 --> 00:14:24.830 align:middle line:90% Quite close. 00:14:24.830 --> 00:14:29.090 align:middle line:90% There's one more that's just-- 00:14:29.090 --> 00:14:30.500 align:middle line:90% Norway? 00:14:30.500 --> 00:14:31.190 align:middle line:90% Yep. 00:14:31.190 --> 00:14:34.950 align:middle line:90% We have Norway and then the US. 00:14:34.950 --> 00:14:36.450 align:middle line:84% And then what's the poorest country? 00:14:36.450 --> 00:14:38.492 align:middle line:84% And of course, life is super expensive in Norway. 00:14:38.492 --> 00:14:40.680 align:middle line:90% But remember, it's PPP-adjusted. 00:14:40.680 --> 00:14:42.270 align:middle line:84% The poorest country in the world-- 00:14:42.270 --> 00:14:45.180 align:middle line:84% I already put this slide up, so I'm going 00:14:45.180 --> 00:14:47.730 align:middle line:90% to not let the suspense hold. 00:14:47.730 --> 00:14:53.940 align:middle line:84% It's Burundi, according to this list, at $790. 00:14:53.940 --> 00:15:00.090 align:middle line:84% So if we do the ratio of the two, that's the large-- 00:15:00.090 --> 00:15:04.840 align:middle line:84% your take your Norway to Burundi, 00:15:04.840 --> 00:15:07.380 align:middle line:90% it's a ratio of almost 100-- 00:15:07.380 --> 00:15:08.340 align:middle line:90% 1 to 100. 00:15:08.340 --> 00:15:13.040 align:middle line:84% And remember that we've tried the PPP adjustment. 00:15:13.040 --> 00:15:18.390 align:middle line:84% So it's hard to think-- it's hard to fully visualize them, 00:15:18.390 --> 00:15:19.740 align:middle line:90% even though we've no-- 00:15:19.740 --> 00:15:22.740 align:middle line:84% you're thinking still of-- thinking of Burundi and living. 00:15:22.740 --> 00:15:25.840 align:middle line:84% But we've, in principle, done the PPP adjustment, 00:15:25.840 --> 00:15:31.440 align:middle line:84% so this is as close as we have, from this very simple measure 00:15:31.440 --> 00:15:35.100 align:middle line:84% that's an imperfect measure that is income of what 00:15:35.100 --> 00:15:37.510 align:middle line:90% standard of living are present. 00:15:37.510 --> 00:15:38.860 align:middle line:90% Now, that's a lot. 00:15:38.860 --> 00:15:44.740 align:middle line:84% And therefore, understanding where this comes from, 00:15:44.740 --> 00:15:51.280 align:middle line:84% and understanding how people with so little money 00:15:51.280 --> 00:15:54.220 align:middle line:84% live their life and manage and all of that, 00:15:54.220 --> 00:15:58.030 align:middle line:90% is of prime importance. 00:15:58.030 --> 00:16:03.510 align:middle line:84% It's hard to think of something that's more key, compared to-- 00:16:03.510 --> 00:16:05.080 align:middle line:84% I love public finance, but we are 00:16:05.080 --> 00:16:07.720 align:middle line:90% thinking about improving the-- 00:16:07.720 --> 00:16:11.860 align:middle line:84% reducing some inefficiency in something, 00:16:11.860 --> 00:16:13.960 align:middle line:90% reducing the triangle by 2%. 00:16:13.960 --> 00:16:17.470 align:middle line:84% [LAUGHS] I'm talking about differences in income of 1 00:16:17.470 --> 00:16:18.370 align:middle line:90% in 200. 00:16:18.370 --> 00:16:26.420 align:middle line:84% So that has to be taking a lot of space in our brain. 00:16:26.420 --> 00:16:28.700 align:middle line:84% And of course, you could say, well, it's just money. 00:16:28.700 --> 00:16:30.710 align:middle line:84% Money's not everything there is to poverty. 00:16:30.710 --> 00:16:32.540 align:middle line:84% But these income differences, of course, 00:16:32.540 --> 00:16:36.703 align:middle line:84% matter for well-being outcomes, more generally, as well. 00:16:36.703 --> 00:16:38.120 align:middle line:84% It's often a mistake we're making, 00:16:38.120 --> 00:16:40.220 align:middle line:90% of equating GDP with well-being. 00:16:40.220 --> 00:16:43.100 align:middle line:84% We're certainly not going to do that mistake in this class, 00:16:43.100 --> 00:16:50.460 align:middle line:84% but we can see that it is related. 00:16:50.460 --> 00:16:52.970 align:middle line:84% So for example, if you look at the under-five mortality 00:16:52.970 --> 00:16:58.840 align:middle line:84% in Norway, it's 2 deaths per 1,000 live births. 00:16:58.840 --> 00:17:03.010 align:middle line:84% If you look at the infant mortality in Burundi, 00:17:03.010 --> 00:17:10.109 align:middle line:84% a similar number-- it's 57 deaths per 1,000 live births. 00:17:10.109 --> 00:17:14.040 align:middle line:84% So this factor of 1 in 200 in standard of living 00:17:14.040 --> 00:17:16.770 align:middle line:84% is also a factor is 1 into 25 in the chance 00:17:16.770 --> 00:17:20.297 align:middle line:84% to make it past your first birthday 00:17:20.297 --> 00:17:21.255 align:middle line:90% or your fifth birthday. 00:17:21.255 --> 00:17:24.599 align:middle line:90% 00:17:24.599 --> 00:17:29.650 align:middle line:84% And although income is neither necessary nor sufficient-- 00:17:29.650 --> 00:17:31.470 align:middle line:84% so for example, if you compare, what 00:17:31.470 --> 00:17:34.590 align:middle line:84% is under-five mortality in the US? 00:17:34.590 --> 00:17:48.730 align:middle line:90% 00:17:48.730 --> 00:17:50.917 align:middle line:90% AUDIENCE: It's 7 per 1,000. 00:17:50.917 --> 00:17:52.000 align:middle line:90% ESTHER DUFLO: 7 per 1,000. 00:17:52.000 --> 00:17:54.430 align:middle line:84% So it's, like, three and a half times Norway, 00:17:54.430 --> 00:17:57.580 align:middle line:84% with a very similar GDP per capita. 00:17:57.580 --> 00:18:01.000 align:middle line:84% And now if we compare Sri Lanka and Guatemala-- 00:18:01.000 --> 00:18:06.736 align:middle line:90% 00:18:06.736 --> 00:18:08.660 align:middle line:90% AUDIENCE: 7? 00:18:08.660 --> 00:18:10.305 align:middle line:84% ESTHER DUFLO: So Sri Lanka is how much? 00:18:10.305 --> 00:18:10.805 align:middle line:90% AUDIENCE: 7? 00:18:10.805 --> 00:18:12.950 align:middle line:90% ESTHER DUFLO: 7, like the US. 00:18:12.950 --> 00:18:14.290 align:middle line:90% And Guatemala? 00:18:14.290 --> 00:18:15.230 align:middle line:90% AUDIENCE: 25. 00:18:15.230 --> 00:18:15.980 align:middle line:90% ESTHER DUFLO: 25. 00:18:15.980 --> 00:18:19.010 align:middle line:84% And Guatemala is, if I'm not wrong, richer 00:18:19.010 --> 00:18:23.820 align:middle line:90% than Sri Lanka in this table. 00:18:23.820 --> 00:18:27.000 align:middle line:90% So income is not sufficient. 00:18:27.000 --> 00:18:27.860 align:middle line:90% It's not necessary. 00:18:27.860 --> 00:18:30.480 align:middle line:84% Sri Lanka is just as good as US, in this table. 00:18:30.480 --> 00:18:33.470 align:middle line:84% I should say, this is pre-COVID numbers. 00:18:33.470 --> 00:18:35.720 align:middle line:84% Sri Lanka is falling apart right now. 00:18:35.720 --> 00:18:39.080 align:middle line:90% But this is pre-COVID number. 00:18:39.080 --> 00:18:41.270 align:middle line:84% And on the other hand, you have countries 00:18:41.270 --> 00:18:44.570 align:middle line:84% that are very similar level who are doing very differently-- 00:18:44.570 --> 00:18:48.950 align:middle line:84% showing the importance of probably policy 00:18:48.950 --> 00:18:51.700 align:middle line:84% but also maybe culture or maybe the way people organize 00:18:51.700 --> 00:18:54.200 align:middle line:84% their lives, et cetera-- so, showing that there is certainly 00:18:54.200 --> 00:18:57.740 align:middle line:84% a lot to understand about what can be done even 00:18:57.740 --> 00:18:59.330 align:middle line:84% with very little money, with very 00:18:59.330 --> 00:19:02.633 align:middle line:84% little income in a country, and what can be wasted even 00:19:02.633 --> 00:19:03.800 align:middle line:90% with a fair amount of money. 00:19:03.800 --> 00:19:06.450 align:middle line:90% 00:19:06.450 --> 00:19:11.190 align:middle line:84% Of course, the COVID-19 crisis, in a way, 00:19:11.190 --> 00:19:14.308 align:middle line:84% it makes you think that you should listen-- 00:19:14.308 --> 00:19:15.600 align:middle line:90% you should do health economics. 00:19:15.600 --> 00:19:17.940 align:middle line:84% But, for me, it's one more reason 00:19:17.940 --> 00:19:22.766 align:middle line:84% why we need to do development economics. 00:19:22.766 --> 00:19:25.860 align:middle line:84% You all know that the rate of vaccination 00:19:25.860 --> 00:19:29.310 align:middle line:90% is extraordinarily unequal-- 00:19:29.310 --> 00:19:32.100 align:middle line:90% again, by income level. 00:19:32.100 --> 00:19:33.960 align:middle line:90% So you're seeing-- 00:19:33.960 --> 00:19:40.080 align:middle line:84% The colors here is according to the number of doses received. 00:19:40.080 --> 00:19:41.790 align:middle line:90% This is relatively recent data. 00:19:41.790 --> 00:19:46.170 align:middle line:84% I think I retrieved the data towards the end of July. 00:19:46.170 --> 00:19:49.470 align:middle line:84% And you're seeing the US, Europe-- 00:19:49.470 --> 00:19:52.350 align:middle line:90% 00:19:52.350 --> 00:19:56.010 align:middle line:84% dark blue, as a lot of people get vaccinated; 00:19:56.010 --> 00:20:00.670 align:middle line:90% Africa, almost white. 00:20:00.670 --> 00:20:03.480 align:middle line:84% So as we are thinking about booster shots 00:20:03.480 --> 00:20:08.225 align:middle line:84% and immunizing children and things like that, 00:20:08.225 --> 00:20:10.410 align:middle line:84% the number of people who are vaccinated 00:20:10.410 --> 00:20:18.138 align:middle line:84% against COVID in Africa is less than a few percent. 00:20:18.138 --> 00:20:20.830 align:middle line:84% Another thing that was very different during the COVID 00:20:20.830 --> 00:20:24.700 align:middle line:84% crisis, and sets apart what it is to be poor and really just 00:20:24.700 --> 00:20:26.530 align:middle line:90% makes-- 00:20:26.530 --> 00:20:30.700 align:middle line:84% is probably a direct consequences of poverty, 00:20:30.700 --> 00:20:35.020 align:middle line:84% is the budgetary and fiscal support 00:20:35.020 --> 00:20:38.260 align:middle line:90% to people and [INAUDIBLE]. 00:20:38.260 --> 00:20:40.555 align:middle line:84% Across the rich countries, do you 00:20:40.555 --> 00:20:43.180 align:middle line:84% know-- and this is so small that you probably can't [? read, ?] 00:20:43.180 --> 00:20:44.930 align:middle line:84% and therefore I can ask you the question-- 00:20:44.930 --> 00:20:48.100 align:middle line:84% do you know, more or less, what fraction of GDP rich countries 00:20:48.100 --> 00:20:50.410 align:middle line:84% spent on fiscal stimulus measures 00:20:50.410 --> 00:20:54.280 align:middle line:84% during the COVID crisis, on average? 00:20:54.280 --> 00:20:57.410 align:middle line:90% 00:20:57.410 --> 00:20:59.885 align:middle line:90% Try a ballpark number. 00:20:59.885 --> 00:21:07.670 align:middle line:90% 00:21:07.670 --> 00:21:09.230 align:middle line:90% More than 5% or less than 5%? 00:21:09.230 --> 00:21:12.690 align:middle line:90% 00:21:12.690 --> 00:21:15.250 align:middle line:90% More than 5%. 00:21:15.250 --> 00:21:17.110 align:middle line:90% More than 50% or less than 50%? 00:21:17.110 --> 00:21:18.076 align:middle line:90% [LAUGHTER] 00:21:18.076 --> 00:21:23.270 align:middle line:90% [LAUGHS] Less than 50%. 00:21:23.270 --> 00:21:26.600 align:middle line:84% [LAUGHS] So somewhere between 5% and 50%. 00:21:26.600 --> 00:21:27.530 align:middle line:90% About 20%. 00:21:27.530 --> 00:21:28.800 align:middle line:90% About 20%. 00:21:28.800 --> 00:21:34.040 align:middle line:84% So the EU, the US, the rich countries, Korea, 00:21:34.040 --> 00:21:39.070 align:middle line:84% spent about 20% of their GDP on fiscal stimulus measures. 00:21:39.070 --> 00:21:44.110 align:middle line:84% Now, how much money did the poorest countries spent-- 00:21:44.110 --> 00:21:46.644 align:middle line:90% ballpark? 00:21:46.644 --> 00:21:48.653 align:middle line:90% As a fraction of their GDP-- 00:21:48.653 --> 00:21:50.570 align:middle line:84% which, of course, is much smaller, as we know, 00:21:50.570 --> 00:21:52.695 align:middle line:84% because they are the poorest country by definition. 00:21:52.695 --> 00:21:56.210 align:middle line:90% 00:21:56.210 --> 00:21:58.980 align:middle line:90% About 2%. 00:21:58.980 --> 00:22:00.920 align:middle line:84% So the poorest countries spent about 2% 00:22:00.920 --> 00:22:04.880 align:middle line:84% of their much smaller GDP to prop up their economy 00:22:04.880 --> 00:22:06.140 align:middle line:90% during the COVID crisis. 00:22:06.140 --> 00:22:11.030 align:middle line:84% And then the emerging market spent about 6%. 00:22:11.030 --> 00:22:14.270 align:middle line:90% So that's how different it was. 00:22:14.270 --> 00:22:17.300 align:middle line:84% Of course, that's the direct consequence of money. 00:22:17.300 --> 00:22:20.060 align:middle line:84% The rich countries could borrow essentially 00:22:20.060 --> 00:22:26.510 align:middle line:84% at 0 interest rate with no impact on their credit rating 00:22:26.510 --> 00:22:27.620 align:middle line:90% or anything. 00:22:27.620 --> 00:22:31.160 align:middle line:90% It was basically magic money-- 00:22:31.160 --> 00:22:32.540 align:middle line:90% and understood as such-- 00:22:32.540 --> 00:22:34.400 align:middle line:84% which allowed them to do whatever 00:22:34.400 --> 00:22:36.380 align:middle line:90% it takes that propped us up. 00:22:36.380 --> 00:22:38.480 align:middle line:84% But the rich country didn't have that option. 00:22:38.480 --> 00:22:40.205 align:middle line:90% If you are the country of-- 00:22:40.205 --> 00:22:41.360 align:middle line:90% you know, even India. 00:22:41.360 --> 00:22:43.820 align:middle line:84% It's not that you can go out on the international market 00:22:43.820 --> 00:22:47.242 align:middle line:84% and borrow, in order to spend on your citizens. 00:22:47.242 --> 00:22:48.200 align:middle line:90% It's just not possible. 00:22:48.200 --> 00:22:53.060 align:middle line:90% 00:22:53.060 --> 00:22:57.050 align:middle line:84% So the only way for the poor countries to have spent more 00:22:57.050 --> 00:23:03.170 align:middle line:84% on COVID-19 stimulus measures would have to receive either 00:23:03.170 --> 00:23:07.220 align:middle line:84% direct aid in the form of a budget support or something 00:23:07.220 --> 00:23:10.820 align:middle line:84% that the IMF has really worked on-- but it's not even fully 00:23:10.820 --> 00:23:12.410 align:middle line:90% finalized now-- 00:23:12.410 --> 00:23:15.380 align:middle line:84% to, in a sense, get the authorization 00:23:15.380 --> 00:23:19.970 align:middle line:84% to create money via special drawing rights, 00:23:19.970 --> 00:23:21.350 align:middle line:84% which would be the equivalent to, 00:23:21.350 --> 00:23:24.170 align:middle line:84% the whole world can decide to emit more money-- 00:23:24.170 --> 00:23:27.670 align:middle line:84% which the IMF calls "special drawing rights." 00:23:27.670 --> 00:23:32.300 align:middle line:84% But when we do that, first of all, 00:23:32.300 --> 00:23:35.140 align:middle line:84% everyone has to agree to emit more special drawing rights. 00:23:35.140 --> 00:23:38.050 align:middle line:84% Second of all, once the special drawing rights are emitted, 00:23:38.050 --> 00:23:42.640 align:middle line:84% they are allocated according to contribution to the IMF, 00:23:42.640 --> 00:23:44.480 align:middle line:84% which is roughly according to the GDP. 00:23:44.480 --> 00:23:47.800 align:middle line:84% Which is that, this money was created, finally, 00:23:47.800 --> 00:23:49.630 align:middle line:90% sometime in May or June-- 00:23:49.630 --> 00:23:51.640 align:middle line:84% but anyways, allocated to country 00:23:51.640 --> 00:23:53.000 align:middle line:90% as a fraction of their money. 00:23:53.000 --> 00:23:57.190 align:middle line:84% So it went again in priority to the rich country. 00:23:57.190 --> 00:23:59.770 align:middle line:84% And then they needed to do some complicated manipulation, 00:23:59.770 --> 00:24:03.460 align:middle line:84% to get it to give up their allocation, 00:24:03.460 --> 00:24:07.000 align:middle line:84% to give it to the poor countries, but they didn't. 00:24:07.000 --> 00:24:08.110 align:middle line:90% The US refused to. 00:24:08.110 --> 00:24:10.370 align:middle line:90% It didn't really happen. 00:24:10.370 --> 00:24:15.610 align:middle line:84% So that means that this is something that we don't fully 00:24:15.610 --> 00:24:17.920 align:middle line:84% know, of course, the ramifications of the fact 00:24:17.920 --> 00:24:19.900 align:middle line:84% that poor countries were unable to spend 00:24:19.900 --> 00:24:21.610 align:middle line:90% as much to prop their economy. 00:24:21.610 --> 00:24:28.720 align:middle line:84% But it means that it ended up being a much bigger shock 00:24:28.720 --> 00:24:31.750 align:middle line:84% on the lives of people who are much more 00:24:31.750 --> 00:24:34.000 align:middle line:90% fragile to start with. 00:24:34.000 --> 00:24:36.880 align:middle line:84% Even though, at the beginning of the crisis in particular, 00:24:36.880 --> 00:24:40.960 align:middle line:84% the health shock per se wasn't that 00:24:40.960 --> 00:24:42.730 align:middle line:84% large in the poorer countries compared 00:24:42.730 --> 00:24:44.680 align:middle line:90% to the richer countries. 00:24:44.680 --> 00:24:48.790 align:middle line:84% But that shows you the huge difference it means, to be-- 00:24:48.790 --> 00:24:50.260 align:middle line:90% a huge difference in life. 00:24:50.260 --> 00:24:52.210 align:middle line:84% Because of course, then it has implications, 00:24:52.210 --> 00:24:54.700 align:middle line:84% because countries that could not support their citizens 00:24:54.700 --> 00:24:59.020 align:middle line:84% during lockdown couldn't really do lockdown beyond a point. 00:24:59.020 --> 00:25:04.720 align:middle line:84% You remember the terrible third wave in India. 00:25:04.720 --> 00:25:06.880 align:middle line:84% For them, it was more like a second wave in India. 00:25:06.880 --> 00:25:08.320 align:middle line:90% And then in Indonesia. 00:25:08.320 --> 00:25:11.470 align:middle line:84% And in both cases-- and we have now one in Sri Lanka. 00:25:11.470 --> 00:25:14.620 align:middle line:90% In all cases, it is known. 00:25:14.620 --> 00:25:16.370 align:middle line:84% It's not that these governments don't know 00:25:16.370 --> 00:25:17.580 align:middle line:90% that they should lock down. 00:25:17.580 --> 00:25:21.260 align:middle line:84% But they cannot, because they cannot afford it. 00:25:21.260 --> 00:25:22.980 align:middle line:90% Their citizens cannot afford it. 00:25:22.980 --> 00:25:27.150 align:middle line:90% So the two things are connected. 00:25:27.150 --> 00:25:30.980 align:middle line:84% So these differences in GDP are coming also from-- 00:25:30.980 --> 00:25:33.920 align:middle line:84% have ramifications are important and that are really 00:25:33.920 --> 00:25:36.860 align:middle line:84% matters of life and death, in normal time 00:25:36.860 --> 00:25:41.780 align:middle line:84% and in the not-normal time we are living today. 00:25:41.780 --> 00:25:44.180 align:middle line:84% There are also massive differences in growth rates. 00:25:44.180 --> 00:25:46.940 align:middle line:90% 00:25:46.940 --> 00:25:50.150 align:middle line:84% If you take, for example, a difference between-- 00:25:50.150 --> 00:25:52.820 align:middle line:90% China grew at 6.5%. 00:25:52.820 --> 00:25:54.950 align:middle line:84% Burundi grew, in the same period, 00:25:54.950 --> 00:26:01.070 align:middle line:84% at about 0.15%, Pakistan at about 3% in the middle, 00:26:01.070 --> 00:26:05.660 align:middle line:84% like in Indonesia, and Kenya at 1.18%. 00:26:05.660 --> 00:26:09.350 align:middle line:84% And you can see that these differences in growth rates 00:26:09.350 --> 00:26:13.920 align:middle line:84% are for countries that start at relatively similar level. 00:26:13.920 --> 00:26:18.290 align:middle line:84% So this is something that has, of course, 00:26:18.290 --> 00:26:24.090 align:middle line:84% puzzled and concerned economists across fields. 00:26:24.090 --> 00:26:28.310 align:middle line:84% There is a whole study of economic growth. 00:26:28.310 --> 00:26:33.260 align:middle line:84% And in a minute I'll go to various ways 00:26:33.260 --> 00:26:38.240 align:middle line:84% that people have marshaled to try 00:26:38.240 --> 00:26:42.290 align:middle line:84% to understand this difference rate in growth rate. 00:26:42.290 --> 00:26:44.690 align:middle line:84% One thing I want to also point out, especially 00:26:44.690 --> 00:26:50.180 align:middle line:84% as I mentioned the difference in the quality of life 00:26:50.180 --> 00:26:53.840 align:middle line:84% between people in poorer and richer countries, 00:26:53.840 --> 00:26:58.190 align:middle line:84% is the fact that we shouldn't either 00:26:58.190 --> 00:27:05.220 align:middle line:84% be too depressed about the situation 00:27:05.220 --> 00:27:08.850 align:middle line:84% of the poor countries-- or maybe we can, now, post-COVID. 00:27:08.850 --> 00:27:12.150 align:middle line:84% But pre-COVID, in some sense, what 00:27:12.150 --> 00:27:15.780 align:middle line:84% happened in emerging economy over the last few decades 00:27:15.780 --> 00:27:19.740 align:middle line:84% is perhaps the best geopolitical news that there was. 00:27:19.740 --> 00:27:23.460 align:middle line:84% In an environment where there is a lot of discontent in rich 00:27:23.460 --> 00:27:26.160 align:middle line:90% and middle-income countries-- 00:27:26.160 --> 00:27:29.250 align:middle line:84% growing inequality, et cetera, one thing 00:27:29.250 --> 00:27:31.140 align:middle line:84% that happened to the poorest of the world 00:27:31.140 --> 00:27:34.150 align:middle line:90% is that they became less poor. 00:27:34.150 --> 00:27:37.290 align:middle line:90% If you're looking at-- 00:27:37.290 --> 00:27:39.870 align:middle line:84% This is what people called the sometimes the "elephant 00:27:39.870 --> 00:27:45.060 align:middle line:84% curve," because it has kind of a big back and then maybe 00:27:45.060 --> 00:27:47.430 align:middle line:84% that's the head and that's the trunk-- 00:27:47.430 --> 00:27:48.910 align:middle line:90% something like that. 00:27:48.910 --> 00:27:54.030 align:middle line:84% And it's looking at per-adult real income growth rate, 00:27:54.030 --> 00:27:55.710 align:middle line:90% over the last-- 00:27:55.710 --> 00:27:57.960 align:middle line:84% since 1950, per-adult income growth 00:27:57.960 --> 00:28:02.580 align:middle line:84% rate by income group in the world. 00:28:02.580 --> 00:28:06.750 align:middle line:84% And if you notice the axis, the axis is not-- 00:28:06.750 --> 00:28:07.800 align:middle line:90% it's logarithmic. 00:28:07.800 --> 00:28:09.900 align:middle line:84% So we are looking at the very, very, very rich 00:28:09.900 --> 00:28:13.710 align:middle line:84% here, and then, towards the beginning of the graph, 00:28:13.710 --> 00:28:14.820 align:middle line:90% we are comparing 10-- 00:28:14.820 --> 00:28:18.330 align:middle line:84% it's not even logarithmic, but it sort of stretches out 00:28:18.330 --> 00:28:20.710 align:middle line:90% towards the end of the graph. 00:28:20.710 --> 00:28:24.930 align:middle line:84% So this is a part that you surely have heard about, 00:28:24.930 --> 00:28:29.530 align:middle line:84% in many settings, which is, over that time period, 00:28:29.530 --> 00:28:33.490 align:middle line:84% the top 1% have captured 27 of global growth. 00:28:33.490 --> 00:28:36.870 align:middle line:84% So your growth rate per adult real income-- 00:28:36.870 --> 00:28:42.960 align:middle line:84% real-income growth rate of the richest was 200% 00:28:42.960 --> 00:28:47.970 align:middle line:90% for the 0.00001% of the world. 00:28:47.970 --> 00:28:54.690 align:middle line:84% And really it's a story of the top percent-- and even more, 00:28:54.690 --> 00:28:59.347 align:middle line:84% perhaps, the top 1/10 of 1% of the world's population-- 00:28:59.347 --> 00:29:00.930 align:middle line:84% becoming richer and richer and richer. 00:29:00.930 --> 00:29:06.020 align:middle line:84% So we know that, just from living in rich countries. 00:29:06.020 --> 00:29:09.920 align:middle line:84% But the other part of this graph is also interesting, which is, 00:29:09.920 --> 00:29:13.430 align:middle line:84% the income of the poor actually, in proportion, 00:29:13.430 --> 00:29:17.510 align:middle line:84% also went up more than that of the squeezed-up bottom-- 00:29:17.510 --> 00:29:20.490 align:middle line:84% 90% in the US and Western Europe. 00:29:20.490 --> 00:29:22.400 align:middle line:84% So the poorest people in the world, 00:29:22.400 --> 00:29:26.420 align:middle line:84% the people on the 20th percentile, 30th percent-- 00:29:26.420 --> 00:29:30.590 align:middle line:84% their income actually increased by 100% over this period, 00:29:30.590 --> 00:29:31.980 align:middle line:90% on average. 00:29:31.980 --> 00:29:35.150 align:middle line:84% Which is about as much as the growth 00:29:35.150 --> 00:29:37.970 align:middle line:90% of income of the top 0.001%. 00:29:37.970 --> 00:29:41.690 align:middle line:84% So this was actually, of course, you know the-- 00:29:41.690 --> 00:29:44.840 align:middle line:84% China becoming much richer, India growing very fast, 00:29:44.840 --> 00:29:50.540 align:middle line:84% Brazil growing somewhat fast, that has caused that. 00:29:50.540 --> 00:29:53.750 align:middle line:84% Even pre-COVID, this was flattening out, 00:29:53.750 --> 00:29:58.220 align:middle line:84% because of China and India being now sufficiently rich. 00:29:58.220 --> 00:30:00.590 align:middle line:84% Yet you don't see so many of their citizens 00:30:00.590 --> 00:30:03.620 align:middle line:84% in the bottom 10%, so that elephant is slowly, 00:30:03.620 --> 00:30:08.630 align:middle line:90% slowly flattening itself out. 00:30:08.630 --> 00:30:12.680 align:middle line:90% But it's a very important story. 00:30:12.680 --> 00:30:13.370 align:middle line:90% Yeah. 00:30:13.370 --> 00:30:14.828 align:middle line:84% AUDIENCE: Do you know if this graph 00:30:14.828 --> 00:30:18.500 align:middle line:84% is based on what their income-group percentile was 00:30:18.500 --> 00:30:22.525 align:middle line:84% a few decades ago or their income percentile as it is now? 00:30:22.525 --> 00:30:23.900 align:middle line:84% ESTHER DUFLO: From the beginning. 00:30:23.900 --> 00:30:26.240 align:middle line:84% So it's the income of-- over this time period, 00:30:26.240 --> 00:30:31.631 align:middle line:84% the income of the people who were at the beginning from-- 00:30:31.631 --> 00:30:33.442 align:middle line:90% at each-- 00:30:33.442 --> 00:30:34.150 align:middle line:90% AUDIENCE: Thanks. 00:30:34.150 --> 00:30:39.000 align:middle line:90% 00:30:39.000 --> 00:30:41.590 align:middle line:84% ESTHER DUFLO: So that's something to remember. 00:30:41.590 --> 00:30:44.770 align:middle line:84% And so you could say, well, this is China and India. 00:30:44.770 --> 00:30:46.900 align:middle line:84% This is a story mainly of China and India. 00:30:46.900 --> 00:30:51.510 align:middle line:84% But the story of the world getting better pre-COVID 00:30:51.510 --> 00:30:55.338 align:middle line:84% is a story that was not only China and India. 00:30:55.338 --> 00:30:57.630 align:middle line:84% First of all, there were a lot of extremely poor people 00:30:57.630 --> 00:31:00.540 align:middle line:84% in China and India, so that's a good thing even if it was just 00:31:00.540 --> 00:31:01.080 align:middle line:90% [? there. ?] 00:31:01.080 --> 00:31:04.950 align:middle line:84% But the story of things getting better in the world is actually 00:31:04.950 --> 00:31:07.440 align:middle line:84% a little bit everywhere-- not necessarily in terms 00:31:07.440 --> 00:31:08.490 align:middle line:90% of income-- 00:31:08.490 --> 00:31:11.730 align:middle line:84% in African countries, I've had very diverse experiences 00:31:11.730 --> 00:31:13.470 align:middle line:84% from country to country; some have gone 00:31:13.470 --> 00:31:16.230 align:middle line:90% very fast and some not much-- 00:31:16.230 --> 00:31:18.120 align:middle line:90% but in terms of welfare. 00:31:18.120 --> 00:31:21.150 align:middle line:84% For example, if you look at the average years of schooling 00:31:21.150 --> 00:31:25.620 align:middle line:84% between 1950 and 2010, you can see a huge increase 00:31:25.620 --> 00:31:27.680 align:middle line:90% in years of schooling. 00:31:27.680 --> 00:31:31.170 align:middle line:84% And this is the average years of schooling, so, of course, 00:31:31.170 --> 00:31:34.850 align:middle line:84% this takes into account all the adults, so this moves slowly. 00:31:34.850 --> 00:31:38.300 align:middle line:84% But you can see, as the map has more and more kids getting 00:31:38.300 --> 00:31:40.260 align:middle line:90% into school, today-- 00:31:40.260 --> 00:31:45.290 align:middle line:84% or, again, pre-COVID-- the vast majority of 00:31:45.290 --> 00:31:48.542 align:middle line:84% children were enrolled in primary school, 00:31:48.542 --> 00:31:50.750 align:middle line:84% vast majority of primary-school children in the world 00:31:50.750 --> 00:31:52.880 align:middle line:90% were enrolled in primary school. 00:31:52.880 --> 00:31:57.720 align:middle line:84% Literacy rate as a consequence has also changed. 00:31:57.720 --> 00:31:59.510 align:middle line:84% So this is looking at recent data, 00:31:59.510 --> 00:32:02.665 align:middle line:84% but by comparing the older to the youngest, 00:32:02.665 --> 00:32:05.330 align:middle line:84% where, among the old people, it's very red, not 00:32:05.330 --> 00:32:07.490 align:middle line:90% very many people are literate. 00:32:07.490 --> 00:32:12.000 align:middle line:84% But among the young people, it's much bluer. 00:32:12.000 --> 00:32:16.470 align:middle line:84% Still along the getting better, the global malaria deaths 00:32:16.470 --> 00:32:20.260 align:middle line:84% by world region is on this picture. 00:32:20.260 --> 00:32:22.500 align:middle line:84% You can see that most of them are in Africa, 00:32:22.500 --> 00:32:26.970 align:middle line:84% but it's falling down at incredible speed. 00:32:26.970 --> 00:32:30.420 align:middle line:84% And those gains, they are gains that 00:32:30.420 --> 00:32:34.260 align:middle line:84% happened even in places where GDP per capita 00:32:34.260 --> 00:32:37.170 align:middle line:90% almost didn't change. 00:32:37.170 --> 00:32:40.050 align:middle line:84% So there have been gains in the welfare of people 00:32:40.050 --> 00:32:43.290 align:middle line:90% coming from better policies-- 00:32:43.290 --> 00:32:44.350 align:middle line:90% coming-- 00:32:44.350 --> 00:32:47.715 align:middle line:84% And so that's something that's super interesting to understand 00:32:47.715 --> 00:32:50.700 align:middle line:84% why did it come about, from a geopolitical perspective, 00:32:50.700 --> 00:32:54.180 align:middle line:84% that there was more focus on these issues. 00:32:54.180 --> 00:32:59.530 align:middle line:84% And more ability to deliver, as we move along. 00:32:59.530 --> 00:33:03.260 align:middle line:84% As a result, you have a decline in children's mortality, 00:33:03.260 --> 00:33:06.840 align:middle line:84% particularly fast in low-income country. 00:33:06.840 --> 00:33:10.900 align:middle line:84% But unfortunately the question now is, what's going to happen? 00:33:10.900 --> 00:33:12.640 align:middle line:90% And that's what we don't know. 00:33:12.640 --> 00:33:15.600 align:middle line:84% That's also why it's a particularly maybe fascinating 00:33:15.600 --> 00:33:21.977 align:middle line:84% time to study development, because the next-- 00:33:21.977 --> 00:33:24.560 align:middle line:84% I think we have a better sense of what the next few years will 00:33:24.560 --> 00:33:25.410 align:middle line:90% be in this country. 00:33:25.410 --> 00:33:28.400 align:middle line:84% I don't think we have any sense of what the next few years will 00:33:28.400 --> 00:33:30.620 align:middle line:90% be in poor countries. 00:33:30.620 --> 00:33:32.640 align:middle line:90% I don't have [LAUGHS] any sense. 00:33:32.640 --> 00:33:36.175 align:middle line:84% And understanding what is happening now 00:33:36.175 --> 00:33:37.550 align:middle line:84% and what has happened in the past 00:33:37.550 --> 00:33:40.213 align:middle line:84% and how people react to different situations 00:33:40.213 --> 00:33:41.630 align:middle line:84% is something that is going to help 00:33:41.630 --> 00:33:46.430 align:middle line:84% us say something about this-- and also maybe make it better. 00:33:46.430 --> 00:33:51.920 align:middle line:84% So, pre-COVID, the projection was that the poverty rate was 00:33:51.920 --> 00:33:55.640 align:middle line:90% going to continue to diminish-- 00:33:55.640 --> 00:34:00.440 align:middle line:84% of course, at a lower rate, simply because the South Asian 00:34:00.440 --> 00:34:06.500 align:middle line:84% countries are becoming too rich, so they cannot contribute much 00:34:06.500 --> 00:34:08.179 align:middle line:90% to the poverty rate. 00:34:08.179 --> 00:34:10.550 align:middle line:84% They don't have very many poor people to contribute. 00:34:10.550 --> 00:34:13.894 align:middle line:90% 00:34:13.894 --> 00:34:15.650 align:middle line:84% And they were already a little bit 00:34:15.650 --> 00:34:17.960 align:middle line:84% pessimistic about sub-Saharan Africa managing 00:34:17.960 --> 00:34:20.960 align:middle line:84% to further go down, in terms of number of poor people. 00:34:20.960 --> 00:34:25.190 align:middle line:84% And this is where we were before COVID. 00:34:25.190 --> 00:34:28.340 align:middle line:84% After COVID, of course, this has changed. 00:34:28.340 --> 00:34:30.860 align:middle line:84% These are also projections for the World Bank of what 00:34:30.860 --> 00:34:33.690 align:middle line:90% they called the "nowcasting." 00:34:33.690 --> 00:34:36.920 align:middle line:84% Which is, of course, the way we normally count 00:34:36.920 --> 00:34:40.040 align:middle line:90% the number of poor people-- 00:34:40.040 --> 00:34:42.710 align:middle line:90% Do you know how we count-- 00:34:42.710 --> 00:34:44.520 align:middle line:84% how they count the number of poor people? 00:34:44.520 --> 00:34:47.699 align:middle line:84% So if we calculate the poverty rate, 00:34:47.699 --> 00:34:51.110 align:middle line:90% how do they go about doing that? 00:34:51.110 --> 00:34:52.429 align:middle line:90% Yep. 00:34:52.429 --> 00:34:54.380 align:middle line:84% AUDIENCE: You define a poverty line which 00:34:54.380 --> 00:34:59.660 align:middle line:84% is based on a basket of goods that people can afford 00:34:59.660 --> 00:35:00.590 align:middle line:90% and the price of that. 00:35:00.590 --> 00:35:04.010 align:middle line:84% If the household earns less than that money, 00:35:04.010 --> 00:35:05.510 align:middle line:84% that depends on the number of people 00:35:05.510 --> 00:35:07.670 align:middle line:90% that live in the household-- 00:35:07.670 --> 00:35:10.303 align:middle line:84% then that household is qualified as poor. 00:35:10.303 --> 00:35:11.220 align:middle line:90% ESTHER DUFLO: Exactly. 00:35:11.220 --> 00:35:12.928 align:middle line:84% So at the international level, right now, 00:35:12.928 --> 00:35:16.310 align:middle line:84% the poverty line is 1.90, in PPP dollar. 00:35:16.310 --> 00:35:19.880 align:middle line:90% It keeps moving up, by the way. 00:35:19.880 --> 00:35:22.580 align:middle line:90% It used to be one, and then-- 00:35:22.580 --> 00:35:24.200 align:middle line:84% so that, again-- that's the same team 00:35:24.200 --> 00:35:29.060 align:middle line:84% that does PPP calculation, who does it. 00:35:29.060 --> 00:35:33.240 align:middle line:84% And then you need to send people out to survey people. 00:35:33.240 --> 00:35:36.110 align:middle line:84% And what happened during the COVID-19 crisis is, 00:35:36.110 --> 00:35:38.850 align:middle line:90% all surveying stopped. 00:35:38.850 --> 00:35:41.460 align:middle line:84% So we don't really know how many poor people there are. 00:35:41.460 --> 00:35:43.528 align:middle line:84% [LAUGHS] The truth is, we have no idea 00:35:43.528 --> 00:35:44.820 align:middle line:90% how many poor people there are. 00:35:44.820 --> 00:35:49.650 align:middle line:84% Because it hasn't been possible to go survey people in person. 00:35:49.650 --> 00:35:52.170 align:middle line:84% Most statistics-- for example, in India, 00:35:52.170 --> 00:35:56.100 align:middle line:84% the national survey organization, 00:35:56.100 --> 00:35:58.680 align:middle line:84% national sample survey organization, that is normally 00:35:58.680 --> 00:36:01.680 align:middle line:84% conducting large-scale survey every five years, 00:36:01.680 --> 00:36:05.220 align:middle line:84% and slightly smaller surveys every year, 00:36:05.220 --> 00:36:06.537 align:middle line:90% has stopped doing it. 00:36:06.537 --> 00:36:08.370 align:middle line:84% And even if they're doing it, the government 00:36:08.370 --> 00:36:10.440 align:middle line:84% is not reporting what they are doing anyway. 00:36:10.440 --> 00:36:16.230 align:middle line:84% So it's been replaced by private companies that will provide you 00:36:16.230 --> 00:36:22.620 align:middle line:84% with household survey data that, who knows how they are sampled, 00:36:22.620 --> 00:36:24.240 align:middle line:90% et cetera. 00:36:24.240 --> 00:36:26.520 align:middle line:84% So that's not really a replacement. 00:36:26.520 --> 00:36:28.620 align:middle line:84% But we have completely lost the ability 00:36:28.620 --> 00:36:32.330 align:middle line:84% to even measure how many poor people there are. 00:36:32.330 --> 00:36:35.760 align:middle line:84% So, because we hate to not be able to say things, 00:36:35.760 --> 00:36:38.690 align:middle line:84% even if we don't have the data, what the World Bank does 00:36:38.690 --> 00:36:42.890 align:middle line:84% is that they do what they call "nowcasting," which is, 00:36:42.890 --> 00:36:46.070 align:middle line:84% they predict how many poor people there should be, 00:36:46.070 --> 00:36:48.570 align:middle line:90% based on the country's GDP. 00:36:48.570 --> 00:36:51.290 align:middle line:84% So in a sense, these numbers are just a translation of the fact 00:36:51.290 --> 00:36:53.400 align:middle line:84% that we could have looked directly-- 00:36:53.400 --> 00:36:55.880 align:middle line:84% which is, the GDP of this country has collapsed. 00:36:55.880 --> 00:36:58.910 align:middle line:84% But with that in mind, so you don't-- 00:36:58.910 --> 00:37:01.640 align:middle line:84% one shouldn't take these numbers too, too seriously-- 00:37:01.640 --> 00:37:03.230 align:middle line:84% but with that in mind, of course, 00:37:03.230 --> 00:37:05.960 align:middle line:84% there has been a huge increase in the number 00:37:05.960 --> 00:37:10.970 align:middle line:84% of poor people in 2020, which hasn't let down in 2021. 00:37:10.970 --> 00:37:15.040 align:middle line:84% Of course, we don't know what's happening next. 00:37:15.040 --> 00:37:17.500 align:middle line:84% And why we don't know what's happening next is that, 00:37:17.500 --> 00:37:19.210 align:middle line:90% at least at this point-- 00:37:19.210 --> 00:37:21.040 align:middle line:84% and that's one thing that I'm hoping, 00:37:21.040 --> 00:37:24.880 align:middle line:84% that some of the thinking in this class 00:37:24.880 --> 00:37:27.550 align:middle line:84% will help us think through is, we 00:37:27.550 --> 00:37:30.460 align:middle line:84% don't know whether this change in GDP 00:37:30.460 --> 00:37:34.480 align:middle line:84% and the change in individual standard of living 00:37:34.480 --> 00:37:37.630 align:middle line:90% will be transitory or permanent. 00:37:37.630 --> 00:37:38.542 align:middle line:90% In many cases-- 00:37:38.542 --> 00:37:40.000 align:middle line:84% For example, in rich countries, you 00:37:40.000 --> 00:37:44.200 align:middle line:84% can be fairly confident that, even if people are poor today, 00:37:44.200 --> 00:37:49.370 align:middle line:84% they'll bounce up and go back in the swing of things. 00:37:49.370 --> 00:37:52.090 align:middle line:84% But one of the things we are going to talk about a lot 00:37:52.090 --> 00:37:56.290 align:middle line:84% here is poverty trap, and the question, or the idea, 00:37:56.290 --> 00:38:01.960 align:middle line:84% that, possibly, being poor keeps you poor, and therefore that-- 00:38:01.960 --> 00:38:05.380 align:middle line:90% 00:38:05.380 --> 00:38:08.040 align:middle line:84% So we have no idea whether, at the individual level, 00:38:08.040 --> 00:38:09.040 align:middle line:90% there are poverty traps. 00:38:09.040 --> 00:38:11.530 align:middle line:84% So some of the people who were kind of just 00:38:11.530 --> 00:38:14.890 align:middle line:84% edged out of the poverty trap will kind of fall back up. 00:38:14.890 --> 00:38:18.340 align:middle line:84% And at a country level, things like education system, 00:38:18.340 --> 00:38:21.550 align:middle line:84% health system, will be able to catch up the generation 00:38:21.550 --> 00:38:23.050 align:middle line:84% that they have missed so far or not. 00:38:23.050 --> 00:38:24.160 align:middle line:90% Yep? 00:38:24.160 --> 00:38:27.280 align:middle line:84% AUDIENCE: In terms of the historical improvement, 00:38:27.280 --> 00:38:31.480 align:middle line:84% how much do you think came from internal policy improvements? 00:38:31.480 --> 00:38:34.295 align:middle line:84% And how much came from external-- 00:38:34.295 --> 00:38:36.770 align:middle line:84% or external pressure on the countries 00:38:36.770 --> 00:38:39.870 align:middle line:90% to improve those outcomes? 00:38:39.870 --> 00:38:41.730 align:middle line:84% ESTHER DUFLO: I think a fair summary 00:38:41.730 --> 00:38:45.210 align:middle line:84% is that nothing can come just from external pressure. 00:38:45.210 --> 00:38:47.010 align:middle line:90% If you're looking at-- 00:38:47.010 --> 00:38:49.140 align:middle line:84% from most countries in the world, 00:38:49.140 --> 00:38:56.310 align:middle line:84% the fraction of aid given as a fraction of GDP is minuscule. 00:38:56.310 --> 00:38:58.350 align:middle line:84% There are some very small countries 00:38:58.350 --> 00:39:01.360 align:middle line:84% that are also very poor, where it's a lot more important. 00:39:01.360 --> 00:39:03.900 align:middle line:84% But in the world as a whole, even in Africa, 00:39:03.900 --> 00:39:05.963 align:middle line:84% the fraction of aid as a fraction of the budget-- 00:39:05.963 --> 00:39:07.380 align:middle line:84% what do you think it is, actually? 00:39:07.380 --> 00:39:09.570 align:middle line:84% If we take Africa minus South Africa, which 00:39:09.570 --> 00:39:14.470 align:middle line:84% is a rich economy that doesn't depend on aid, 00:39:14.470 --> 00:39:17.630 align:middle line:90% how much do you think it is? 00:39:17.630 --> 00:39:18.267 align:middle line:90% AUDIENCE: 20%? 00:39:18.267 --> 00:39:19.100 align:middle line:90% ESTHER DUFLO: Sorry? 00:39:19.100 --> 00:39:20.030 align:middle line:90% AUDIENCE: 20%? 00:39:20.030 --> 00:39:21.405 align:middle line:84% ESTHER DUFLO: It's more like 12%. 00:39:21.405 --> 00:39:23.440 align:middle line:90% 00:39:23.440 --> 00:39:25.060 align:middle line:90% And that's without South Africa. 00:39:25.060 --> 00:39:28.180 align:middle line:84% So it's actually not that important, quantitatively, 00:39:28.180 --> 00:39:29.665 align:middle line:90% even for Africa. 00:39:29.665 --> 00:39:32.625 align:middle line:90% 00:39:32.625 --> 00:39:34.360 align:middle line:84% So the first part of the answer is 00:39:34.360 --> 00:39:37.360 align:middle line:84% that it has to come from individual policy 00:39:37.360 --> 00:39:38.410 align:middle line:90% and individual money. 00:39:38.410 --> 00:39:42.250 align:middle line:84% Fundamentally, this is where it has to have come from. 00:39:42.250 --> 00:39:47.470 align:middle line:84% That said, it is true that there has been a global effort, 00:39:47.470 --> 00:39:51.010 align:middle line:84% starting with the Millennium Development Goals followed 00:39:51.010 --> 00:39:53.470 align:middle line:84% by the Sustainable Development Goals, 00:39:53.470 --> 00:39:59.920 align:middle line:84% to of focus policy attention on some core welfare issues-- 00:39:59.920 --> 00:40:02.260 align:middle line:84% like basic health measures, basic education 00:40:02.260 --> 00:40:03.580 align:middle line:90% measures, et cetera. 00:40:03.580 --> 00:40:08.650 align:middle line:84% And I do think that this has helped countries, governments, 00:40:08.650 --> 00:40:12.850 align:middle line:84% in many countries to become a bit pragmatic in terms 00:40:12.850 --> 00:40:16.180 align:middle line:84% of their policy goals, in particular 00:40:16.180 --> 00:40:18.530 align:middle line:84% as far as health and education are concerned. 00:40:18.530 --> 00:40:22.360 align:middle line:84% So I wouldn't call it "pressure" in the sense of the UK coming 00:40:22.360 --> 00:40:27.800 align:middle line:84% to Ghana and say, you do this, or else, 00:40:27.800 --> 00:40:33.320 align:middle line:84% but more of a softer social norm at the level of the world, 00:40:33.320 --> 00:40:35.890 align:middle line:84% which is that these are important issues and a bit more 00:40:35.890 --> 00:40:37.780 align:middle line:90% pragmatism. 00:40:37.780 --> 00:40:40.870 align:middle line:84% I would also say-- and this is something 00:40:40.870 --> 00:40:42.700 align:middle line:84% we will discover as we move along-- 00:40:42.700 --> 00:40:47.050 align:middle line:84% that the quality of policymaking has improved, where people 00:40:47.050 --> 00:40:49.510 align:middle line:84% understand-- governments understand a little bit better 00:40:49.510 --> 00:40:50.800 align:middle line:90% what needs to be done? 00:40:50.800 --> 00:40:51.470 align:middle line:90% What works? 00:40:51.470 --> 00:40:55.750 align:middle line:84% What doesn't work and what they can do within their budget, 00:40:55.750 --> 00:40:56.830 align:middle line:90% and so on and so forth. 00:40:56.830 --> 00:41:03.920 align:middle line:90% 00:41:03.920 --> 00:41:07.300 align:middle line:84% So the point of this class is to try and understand 00:41:07.300 --> 00:41:11.470 align:middle line:84% these massive income differences in three respects. 00:41:11.470 --> 00:41:15.810 align:middle line:84% Why are some countries so poor and some countries so rich? 00:41:15.810 --> 00:41:18.240 align:middle line:84% What are the particular economic problems 00:41:18.240 --> 00:41:21.720 align:middle line:84% that are specific to the poor countries 00:41:21.720 --> 00:41:23.940 align:middle line:90% and how we understand them? 00:41:23.940 --> 00:41:26.370 align:middle line:84% And, from a policy perspective, what 00:41:26.370 --> 00:41:28.620 align:middle line:84% could be done to solve a particular-- 00:41:28.620 --> 00:41:30.360 align:middle line:84% these issues that we just discussed, 00:41:30.360 --> 00:41:32.400 align:middle line:84% these particular market failures that arise 00:41:32.400 --> 00:41:34.657 align:middle line:90% because countries are poor? 00:41:34.657 --> 00:41:36.990 align:middle line:84% So of course, some of the problems of the poor countries 00:41:36.990 --> 00:41:38.657 align:middle line:84% need to be solved in the rich countries. 00:41:38.657 --> 00:41:42.300 align:middle line:84% Like, COVID vaccination doses is something that, it's not really 00:41:42.300 --> 00:41:43.710 align:middle line:90% in their power. 00:41:43.710 --> 00:41:48.660 align:middle line:84% But a lot of the things that countries do is-- 00:41:48.660 --> 00:41:50.790 align:middle line:84% and this discussion we were just having-- 00:41:50.790 --> 00:41:53.880 align:middle line:84% is about what countries should do for themselves 00:41:53.880 --> 00:41:58.210 align:middle line:84% and what went right, what went wrong, and how we can do that. 00:41:58.210 --> 00:42:00.450 align:middle line:84% And of course, one question that I 00:42:00.450 --> 00:42:05.250 align:middle line:84% think we should keep thinking in the back of our mind-- 00:42:05.250 --> 00:42:07.230 align:middle line:84% not necessarily because we have an answer, 00:42:07.230 --> 00:42:10.260 align:middle line:84% but because it will help us think about [LAUGHS] 00:42:10.260 --> 00:42:14.070 align:middle line:84% the problems, is whether the COVID-crisis reversal is here 00:42:14.070 --> 00:42:21.222 align:middle line:84% to stay and what can countries do, what is within their power, 00:42:21.222 --> 00:42:22.680 align:middle line:84% to make sure it's not here to stay. 00:42:22.680 --> 00:42:25.550 align:middle line:90% 00:42:25.550 --> 00:42:29.083 align:middle line:84% So as I said, this is a question that 00:42:29.083 --> 00:42:31.500 align:middle line:84% has been with us, understanding this these differences has 00:42:31.500 --> 00:42:32.750 align:middle line:90% been with us, for a long time. 00:42:32.750 --> 00:42:39.410 align:middle line:84% And for a long time, it was more of a macro field. 00:42:39.410 --> 00:42:42.290 align:middle line:84% Development economics was more about macroeconomics. 00:42:42.290 --> 00:42:46.130 align:middle line:84% And the idea was to focus on the aggregate income 00:42:46.130 --> 00:42:48.410 align:middle line:84% differences themselves and to try to understand 00:42:48.410 --> 00:42:50.100 align:middle line:90% where they come from. 00:42:50.100 --> 00:42:52.580 align:middle line:84% So if you start with an aggregate-production function, 00:42:52.580 --> 00:42:56.300 align:middle line:84% you have, income is produced by some technology 00:42:56.300 --> 00:42:57.650 align:middle line:90% that we don't really understand. 00:42:57.650 --> 00:42:59.840 align:middle line:84% That's, like, the measure of our ignorance, 00:42:59.840 --> 00:43:01.820 align:middle line:90% according to a Bob Solow. 00:43:01.820 --> 00:43:07.490 align:middle line:84% And then a function of capital, and human capital, 00:43:07.490 --> 00:43:09.080 align:middle line:84% expressed here in little letters, 00:43:09.080 --> 00:43:11.375 align:middle line:84% because it's human capital per capita. 00:43:11.375 --> 00:43:15.490 align:middle line:90% 00:43:15.490 --> 00:43:18.880 align:middle line:84% The idea of field called "development accounting" 00:43:18.880 --> 00:43:21.880 align:middle line:84% was to see how much we can explain 00:43:21.880 --> 00:43:24.235 align:middle line:84% from this simple macro-production function. 00:43:24.235 --> 00:43:27.400 align:middle line:90% 00:43:27.400 --> 00:43:29.500 align:middle line:84% If we can explain a lot, then that's great. 00:43:29.500 --> 00:43:31.930 align:middle line:84% For example, if we find that differences are mostly k, 00:43:31.930 --> 00:43:33.190 align:middle line:90% then it makes the-- 00:43:33.190 --> 00:43:35.740 align:middle line:84% one view is that the answer to poverty is simple. 00:43:35.740 --> 00:43:41.140 align:middle line:84% Just give more k, and if it's h, then give them more h, 00:43:41.140 --> 00:43:42.670 align:middle line:90% and we'll be good. 00:43:42.670 --> 00:43:44.260 align:middle line:84% If it's A, of course, we don't know, 00:43:44.260 --> 00:43:48.010 align:middle line:84% because A is what we don't know, so it's much harder what to do. 00:43:48.010 --> 00:43:50.350 align:middle line:84% So that's related to the philosophy of the World 00:43:50.350 --> 00:43:52.795 align:middle line:84% Bank in the 1960s, where the view was 00:43:52.795 --> 00:43:54.920 align:middle line:84% the problem that poor countries don't have enough k 00:43:54.920 --> 00:43:57.170 align:middle line:84% so we need to build roads and we need to build bridges 00:43:57.170 --> 00:44:00.310 align:middle line:90% and we need to build dams and-- 00:44:00.310 --> 00:44:02.543 align:middle line:84% because the poor countries can't borrow. 00:44:02.543 --> 00:44:03.710 align:middle line:90% They are credit-constrained. 00:44:03.710 --> 00:44:09.490 align:middle line:90% So that will help them catch up. 00:44:09.490 --> 00:44:12.880 align:middle line:84% And then, maybe a little bit later [INAUDIBLE] 00:44:12.880 --> 00:44:15.438 align:middle line:84% the idea that it's more h, and so we can build more schools, 00:44:15.438 --> 00:44:17.230 align:middle line:84% and then that's going to solve the problem. 00:44:17.230 --> 00:44:20.070 align:middle line:90% 00:44:20.070 --> 00:44:23.090 align:middle line:84% There's a famous paper from the early 1990s 00:44:23.090 --> 00:44:28.700 align:middle line:84% that showed that was very optimistic, about this model. 00:44:28.700 --> 00:44:31.130 align:middle line:84% They looked at-- they used data from-- 00:44:31.130 --> 00:44:36.220 align:middle line:84% cross-country data with capital and enrollment rates. 00:44:36.220 --> 00:44:39.820 align:middle line:84% And they found that they could explain 00:44:39.820 --> 00:44:43.540 align:middle line:84% 80% of the difference in income per capita across country. 00:44:43.540 --> 00:44:45.580 align:middle line:84% So if that's the case, then that's great. 00:44:45.580 --> 00:44:48.040 align:middle line:84% You know, everything-- there's so 00:44:48.040 --> 00:44:52.660 align:middle line:84% much you can do with the things you can measure and act upon. 00:44:52.660 --> 00:44:54.550 align:middle line:84% Unfortunately, that has been revised, 00:44:54.550 --> 00:44:58.450 align:middle line:84% and subsequent work suggests that this was way 00:44:58.450 --> 00:45:01.850 align:middle line:90% too optimistic. 00:45:01.850 --> 00:45:04.320 align:middle line:84% Klenow and Rodriguez-Clare brought it down to 50%-- 00:45:04.320 --> 00:45:07.150 align:middle line:90% 00:45:07.150 --> 00:45:08.350 align:middle line:90% or (EMPHASIS) by 50%-- 00:45:08.350 --> 00:45:10.840 align:middle line:90% and then by 90% for growth. 00:45:10.840 --> 00:45:15.370 align:middle line:84% Caselli also went down to using-- 00:45:15.370 --> 00:45:18.790 align:middle line:84% probably has the best paper on development accounting-- 00:45:18.790 --> 00:45:21.100 align:middle line:84% said that at least two third is explained by A, 00:45:21.100 --> 00:45:26.030 align:middle line:84% which is precisely what we cannot explain. 00:45:26.030 --> 00:45:28.360 align:middle line:84% So that's just in terms of accounting. 00:45:28.360 --> 00:45:30.430 align:middle line:84% That's just in terms of what's correlated 00:45:30.430 --> 00:45:32.140 align:middle line:84% with what, without trying to give 00:45:32.140 --> 00:45:36.340 align:middle line:84% any causal interpretation-- just as a matter of, 00:45:36.340 --> 00:45:38.350 align:middle line:84% is it the case that, when countries have more k, 00:45:38.350 --> 00:45:39.910 align:middle line:90% they are richer? 00:45:39.910 --> 00:45:43.270 align:middle line:84% But of course, this is also getting complicated by the fact 00:45:43.270 --> 00:45:46.570 align:middle line:84% that all of these things are collated and are jointly 00:45:46.570 --> 00:45:47.920 align:middle line:90% determined. 00:45:47.920 --> 00:45:50.775 align:middle line:90% So, in steady state-- 00:45:50.775 --> 00:45:52.900 align:middle line:84% and this is the point that Solow made, in a sense-- 00:45:52.900 --> 00:45:55.840 align:middle line:84% that in steady state, it has to be that the return to capital 00:45:55.840 --> 00:45:58.500 align:middle line:90% is equal to the discount rate. 00:45:58.500 --> 00:46:03.940 align:middle line:84% Because otherwise people would buy more capital, 00:46:03.940 --> 00:46:06.320 align:middle line:84% and so people would save more, otherwise. 00:46:06.320 --> 00:46:09.400 align:middle line:84% So it has to be that the return to capital equalize. 00:46:09.400 --> 00:46:13.420 align:middle line:84% So if you give a lot of k to a country, 00:46:13.420 --> 00:46:15.100 align:middle line:84% then they are just going to save-- 00:46:15.100 --> 00:46:18.220 align:middle line:84% that's going to draw down the return to saving. 00:46:18.220 --> 00:46:22.070 align:middle line:84% People are going to stop saving and [INAUDIBLE] adjust to that. 00:46:22.070 --> 00:46:27.880 align:middle line:84% So that would suggest that only A can really durably affect 00:46:27.880 --> 00:46:29.950 align:middle line:90% growth. 00:46:29.950 --> 00:46:34.800 align:middle line:84% So if that's the case, then you can't just do k. 00:46:34.800 --> 00:46:37.772 align:middle line:84% In you increase k below the steady-state level, 00:46:37.772 --> 00:46:38.730 align:middle line:90% the return will be low. 00:46:38.730 --> 00:46:42.150 align:middle line:84% People will stop improving-- and likewise for h. 00:46:42.150 --> 00:46:45.330 align:middle line:90% 00:46:45.330 --> 00:46:48.790 align:middle line:84% Instead, you need to increase A, and that's all you need to do. 00:46:48.790 --> 00:46:52.570 align:middle line:84% And you will see that, even though that's the macro view, 00:46:52.570 --> 00:46:56.942 align:middle line:84% that idea is actually quite pervasive in sort 00:46:56.942 --> 00:47:01.180 align:middle line:84% of the mental model that people have, 00:47:01.180 --> 00:47:02.570 align:middle line:90% even in micro development. 00:47:02.570 --> 00:47:05.650 align:middle line:84% So we'll discuss education, where 00:47:05.650 --> 00:47:07.510 align:middle line:84% many people of a certain generation 00:47:07.510 --> 00:47:11.290 align:middle line:84% think that the best education policy is no education policy. 00:47:11.290 --> 00:47:13.210 align:middle line:84% Because if you invest more in education, 00:47:13.210 --> 00:47:14.920 align:middle line:90% people are going to reduce-- 00:47:14.920 --> 00:47:16.612 align:middle line:84% the returns are going to fall enough 00:47:16.612 --> 00:47:18.820 align:middle line:84% that people are going to reduce their own investment, 00:47:18.820 --> 00:47:21.620 align:middle line:90% and it's self-defeating. 00:47:21.620 --> 00:47:24.350 align:middle line:84% So instead, you need to increase A, because A will increase 00:47:24.350 --> 00:47:26.030 align:middle line:90% the return to education. 00:47:26.030 --> 00:47:28.940 align:middle line:84% It will increase the return to investing in capital. 00:47:28.940 --> 00:47:33.800 align:middle line:84% And people will respond accordingly. 00:47:33.800 --> 00:47:37.420 align:middle line:90% So where does A come from? 00:47:37.420 --> 00:47:39.760 align:middle line:90% One answer is that it's history. 00:47:39.760 --> 00:47:44.350 align:middle line:84% So that's what you're getting, say, from Daron's work 00:47:44.350 --> 00:47:49.240 align:middle line:84% and famously, in particular, the Acemoglu-Johnson-Robinson paper 00:47:49.240 --> 00:47:53.230 align:middle line:84% looking at the quality of institution. 00:47:53.230 --> 00:47:56.890 align:middle line:84% It depends-- but also the Shleifer-Vishny agenda. 00:47:56.890 --> 00:47:59.980 align:middle line:84% It depends on who colonized you, what they decided to do, 00:47:59.980 --> 00:48:03.100 align:middle line:84% whether [INAUDIBLE],, whether they set up good institution 00:48:03.100 --> 00:48:06.010 align:middle line:84% or bad institution, et cetera-- whether you have the right kind 00:48:06.010 --> 00:48:08.930 align:middle line:84% of [? law ?] or the wrong kind of [? law. ?] 00:48:08.930 --> 00:48:13.570 align:middle line:90% Another is that it's geography. 00:48:13.570 --> 00:48:15.680 align:middle line:84% That's a Jeff Sachs view of the world. 00:48:15.680 --> 00:48:19.780 align:middle line:84% So that's basically bad luck, if you're in the middle of Africa. 00:48:19.780 --> 00:48:23.830 align:middle line:84% The climate is very conducive to mosquitoes and all the disease 00:48:23.830 --> 00:48:24.550 align:middle line:90% they bring. 00:48:24.550 --> 00:48:27.670 align:middle line:90% You're very far from the sea. 00:48:27.670 --> 00:48:29.170 align:middle line:90% It's too hot. 00:48:29.170 --> 00:48:31.630 align:middle line:84% So it's going to be very difficult to go. 00:48:31.630 --> 00:48:33.950 align:middle line:90% That's it. 00:48:33.950 --> 00:48:36.760 align:middle line:84% So that's why, in the view of Jeff Sachs, all you can do 00:48:36.760 --> 00:48:38.200 align:middle line:90% is to really help-- 00:48:38.200 --> 00:48:41.148 align:middle line:84% you have to compensate people for the bad luck of being 00:48:41.148 --> 00:48:42.190 align:middle line:90% born where they are born. 00:48:42.190 --> 00:48:45.050 align:middle line:90% 00:48:45.050 --> 00:48:50.210 align:middle line:84% So either of these, just from a descriptive perspective, 00:48:50.210 --> 00:48:53.450 align:middle line:84% there is one problem with all of these views-- is that, fine, 00:48:53.450 --> 00:48:57.290 align:middle line:84% but then those factors are fixed. 00:48:57.290 --> 00:49:00.300 align:middle line:84% Geography, of course, doesn't change much. 00:49:00.300 --> 00:49:03.960 align:middle line:84% And then who colonized you hasn't changed in a long time. 00:49:03.960 --> 00:49:06.830 align:middle line:84% So how do we explain the fact that Bangladesh is sometimes 00:49:06.830 --> 00:49:09.110 align:middle line:84% doing great, in terms of growth, and sometimes 00:49:09.110 --> 00:49:12.330 align:middle line:90% doing badly, in term of growth? 00:49:12.330 --> 00:49:14.880 align:middle line:84% And this is something which is really true. 00:49:14.880 --> 00:49:17.940 align:middle line:90% 00:49:17.940 --> 00:49:20.790 align:middle line:84% There are huge fluctuations in country growth rates, 00:49:20.790 --> 00:49:24.000 align:middle line:84% from time to time, which can't be related to these factors 00:49:24.000 --> 00:49:27.580 align:middle line:84% because they are fixed over time. 00:49:27.580 --> 00:49:31.770 align:middle line:84% So maybe it's about variation in macroeconomic policies. 00:49:31.770 --> 00:49:33.450 align:middle line:84% If it's not just long-run factors, 00:49:33.450 --> 00:49:36.240 align:middle line:84% then maybe we can look at macro policies 00:49:36.240 --> 00:49:39.630 align:middle line:84% and see whether that varies growth rate of A. 00:49:39.630 --> 00:49:43.310 align:middle line:84% So what are those macro policies that we could look at? 00:49:43.310 --> 00:49:43.950 align:middle line:90% All of you. 00:49:43.950 --> 00:49:53.260 align:middle line:90% 00:49:53.260 --> 00:49:54.980 align:middle line:90% Don't be shy. 00:49:54.980 --> 00:49:57.042 align:middle line:84% Usually, when I ask a question, I wait. 00:49:57.042 --> 00:49:58.690 align:middle line:90% AUDIENCE: Trade policies? 00:49:58.690 --> 00:49:59.800 align:middle line:90% ESTHER DUFLO: Trade? 00:49:59.800 --> 00:50:01.032 align:middle line:90% [? Certainly. ?] 00:50:01.032 --> 00:50:03.270 align:middle line:90% AUDIENCE: Poverty [INAUDIBLE]? 00:50:03.270 --> 00:50:04.770 align:middle line:84% ESTHER DUFLO: Poverty rates is more, 00:50:04.770 --> 00:50:06.480 align:middle line:90% I would say, an outcome than a-- 00:50:06.480 --> 00:50:06.980 align:middle line:90% but-- 00:50:06.980 --> 00:50:09.905 align:middle line:90% 00:50:09.905 --> 00:50:10.780 align:middle line:90% AUDIENCE: [INAUDIBLE] 00:50:10.780 --> 00:50:11.350 align:middle line:90% ESTHER DUFLO: Sorry? 00:50:11.350 --> 00:50:12.750 align:middle line:84% AUDIENCE: Quality of [? courts. ?] 00:50:12.750 --> 00:50:13.860 align:middle line:90% ESTHER DUFLO: Quality of ports? 00:50:13.860 --> 00:50:14.640 align:middle line:90% AUDIENCE: Courts. 00:50:14.640 --> 00:50:15.060 align:middle line:90% ESTHER DUFLO: Courts. 00:50:15.060 --> 00:50:16.140 align:middle line:90% Yeah, quality of courts. 00:50:16.140 --> 00:50:17.280 align:middle line:90% That could change, yeah. 00:50:17.280 --> 00:50:19.192 align:middle line:84% AUDIENCE: Amount of government expenditures? 00:50:19.192 --> 00:50:21.150 align:middle line:84% ESTHER DUFLO: Government expenditure, or budget 00:50:21.150 --> 00:50:22.710 align:middle line:90% macroprudential policy-- 00:50:22.710 --> 00:50:23.880 align:middle line:90% that kind of things. 00:50:23.880 --> 00:50:26.230 align:middle line:84% How many schools people are building. 00:50:26.230 --> 00:50:29.820 align:middle line:84% You can have, in principal, a budget in education, 00:50:29.820 --> 00:50:31.370 align:middle line:90% budget in health-- 00:50:31.370 --> 00:50:34.860 align:middle line:90% you can have a long list. 00:50:34.860 --> 00:50:37.170 align:middle line:84% And of course, property rights, corruption-- 00:50:37.170 --> 00:50:38.380 align:middle line:90% [INAUDIBLE] 00:50:38.380 --> 00:50:42.180 align:middle line:84% So in the 1990s, there was a huge industry 00:50:42.180 --> 00:50:43.740 align:middle line:90% of funding growth regression. 00:50:43.740 --> 00:50:46.860 align:middle line:90% 00:50:46.860 --> 00:50:48.660 align:middle line:84% There are a few issues with this approach. 00:50:48.660 --> 00:50:51.950 align:middle line:90% 00:50:51.950 --> 00:50:52.800 align:middle line:90% What are the issues? 00:50:52.800 --> 00:50:55.806 align:middle line:90% 00:50:55.806 --> 00:50:56.970 align:middle line:90% Yeah, Erin? 00:50:56.970 --> 00:50:59.220 align:middle line:84% AUDIENCE: If you're trying to determine whether or not 00:50:59.220 --> 00:51:05.160 align:middle line:84% the policy needs to [INAUDIBLE] or net growth [INAUDIBLE] 00:51:05.160 --> 00:51:06.000 align:middle line:90% sustainable? 00:51:06.000 --> 00:51:06.750 align:middle line:90% ESTHER DUFLO: Yes. 00:51:06.750 --> 00:51:10.470 align:middle line:90% There are issues of causalities. 00:51:10.470 --> 00:51:12.390 align:middle line:90% There are issues of-- 00:51:12.390 --> 00:51:14.340 align:middle line:84% because, for example, a country that 00:51:14.340 --> 00:51:16.320 align:middle line:84% certainly has much better governance 00:51:16.320 --> 00:51:17.640 align:middle line:90% can start building schools. 00:51:17.640 --> 00:51:20.560 align:middle line:84% They can also start controlling malaria. 00:51:20.560 --> 00:51:24.000 align:middle line:84% So you don't know whether this is malaria, the school, 00:51:24.000 --> 00:51:27.090 align:middle line:84% or the better governance that in itself led to better business 00:51:27.090 --> 00:51:29.620 align:middle line:84% climate that is harder to measure. 00:51:29.620 --> 00:51:31.920 align:middle line:84% So all of these things typically, they 00:51:31.920 --> 00:51:35.010 align:middle line:84% move together in ways that makes it very difficult 00:51:35.010 --> 00:51:36.990 align:middle line:90% to interpret that. 00:51:36.990 --> 00:51:39.990 align:middle line:84% Even before you get to this problem, 00:51:39.990 --> 00:51:44.050 align:middle line:84% there are only 180 countries, and many, many, many variables, 00:51:44.050 --> 00:51:45.210 align:middle line:90% more than 180. 00:51:45.210 --> 00:51:48.030 align:middle line:90% So you quickly run out of power. 00:51:48.030 --> 00:51:49.510 align:middle line:90% You get into this-- 00:51:49.510 --> 00:51:53.595 align:middle line:84% so maybe machine learning can help you pick something or-- 00:51:53.595 --> 00:51:55.890 align:middle line:84% but one can see that, logically, it 00:51:55.890 --> 00:51:58.230 align:middle line:84% doesn't make much sense to try to estimate 00:51:58.230 --> 00:52:02.520 align:middle line:84% the effect of so many policies with so few countries. 00:52:02.520 --> 00:52:07.646 align:middle line:84% In fact-- Sala-I-Martin, who was a little bit 00:52:07.646 --> 00:52:10.020 align:middle line:84% the king or the pope of that literature, 00:52:10.020 --> 00:52:14.520 align:middle line:84% brought it in a sense to its conclusion and its end, 00:52:14.520 --> 00:52:16.710 align:middle line:84% with a paper that's called "I Just Ran 4 Million 00:52:16.710 --> 00:52:18.240 align:middle line:90% Regressions." 00:52:18.240 --> 00:52:19.940 align:middle line:84% And so you run 4 million regressions, 00:52:19.940 --> 00:52:23.130 align:middle line:84% and you concluded I think that only savings rates really 00:52:23.130 --> 00:52:25.308 align:middle line:90% matter. 00:52:25.308 --> 00:52:27.060 align:middle line:84% And as I already said, these policies 00:52:27.060 --> 00:52:29.400 align:middle line:84% are all endogenously determined, and therefore it's 00:52:29.400 --> 00:52:30.900 align:middle line:90% very difficult to separate. 00:52:30.900 --> 00:52:34.030 align:middle line:90% 00:52:34.030 --> 00:52:38.800 align:middle line:84% And even beyond that, just at a descriptive level, 00:52:38.800 --> 00:52:40.870 align:middle line:90% if you look decade by decade-- 00:52:40.870 --> 00:52:43.480 align:middle line:84% go back to my example of Bangladesh-- 00:52:43.480 --> 00:52:47.530 align:middle line:84% the policies themselves are very similar, from year to year, 00:52:47.530 --> 00:52:48.730 align:middle line:90% but the policy-- 00:52:48.730 --> 00:52:50.290 align:middle line:90% but the growth rate changes-- 00:52:50.290 --> 00:52:52.810 align:middle line:84% fluctuate a lot from decade to decade. 00:52:52.810 --> 00:52:58.030 align:middle line:84% So, just as a purely descriptive level, in growth rate 00:52:58.030 --> 00:53:01.500 align:middle line:84% there is very little explanatory power. 00:53:01.500 --> 00:53:04.490 align:middle line:84% So the bottom line is that the growth-- 00:53:04.490 --> 00:53:08.540 align:middle line:84% [LAUGHS] it's not just us development economists who 00:53:08.540 --> 00:53:11.790 align:middle line:84% look at the macro literature and say, you failed in that issue. 00:53:11.790 --> 00:53:16.340 align:middle line:84% I think it's also the conclusion of the macro community itself, 00:53:16.340 --> 00:53:20.570 align:middle line:84% to say that they just cannot explain economic growth, 00:53:20.570 --> 00:53:23.930 align:middle line:90% based only on aggregate facts. 00:53:23.930 --> 00:53:25.100 align:middle line:90% That's all we have. 00:53:25.100 --> 00:53:27.890 align:middle line:84% We have to get within the hood, which 00:53:27.890 --> 00:53:31.820 align:middle line:84% is why we are starting the whole sequence in development 00:53:31.820 --> 00:53:34.940 align:middle line:84% by looking at micro and then we'll re-aggregate it 00:53:34.940 --> 00:53:37.640 align:middle line:84% from the ground up, in the second half of the semester, 00:53:37.640 --> 00:53:40.340 align:middle line:84% to think about a macro from sound micro principle. 00:53:40.340 --> 00:53:43.610 align:middle line:90% 00:53:43.610 --> 00:53:45.240 align:middle line:90% So what's our issue? 00:53:45.240 --> 00:53:47.000 align:middle line:84% Well, one of the issues is that what 00:53:47.000 --> 00:53:51.930 align:middle line:84% justifies the aggregate production function is not 00:53:51.930 --> 00:53:56.080 align:middle line:84% the idea that there is one giant machine in the economy 00:53:56.080 --> 00:54:00.900 align:middle line:84% and one very smart worker with all of the human capital. 00:54:00.900 --> 00:54:02.010 align:middle line:90% That's not the image. 00:54:02.010 --> 00:54:03.040 align:middle line:90% Right? 00:54:03.040 --> 00:54:05.920 align:middle line:84% We realize that there are a lot of people who are each 00:54:05.920 --> 00:54:08.330 align:middle line:90% going at their own business. 00:54:08.330 --> 00:54:12.250 align:middle line:84% So why can you even aggregate them? 00:54:12.250 --> 00:54:14.440 align:middle line:90% Why can you even write-- 00:54:14.440 --> 00:54:17.620 align:middle line:84% why is it even OK to write an aggregate-production function 00:54:17.620 --> 00:54:18.980 align:middle line:90% to describe the economy? 00:54:18.980 --> 00:54:22.390 align:middle line:84% What's the assumption that's under it, so that it's not just 00:54:22.390 --> 00:54:26.590 align:middle line:84% an image but it's something that actually makes sense? 00:54:26.590 --> 00:54:29.640 align:middle line:90% 00:54:29.640 --> 00:54:37.290 align:middle line:84% Why can I go from MIT and Harvard and [? Twinkie ?] 00:54:37.290 --> 00:54:40.680 align:middle line:84% store, shop, and all of that and aggregate them 00:54:40.680 --> 00:54:43.120 align:middle line:90% into one US economy? 00:54:43.120 --> 00:54:46.010 align:middle line:90% 00:54:46.010 --> 00:54:46.510 align:middle line:90% Yep. 00:54:46.510 --> 00:54:48.760 align:middle line:84% AUDIENCE: [INAUDIBLE] everyone is perfectly optimizing 00:54:48.760 --> 00:54:51.290 align:middle line:84% [INAUDIBLE] efficient [INAUDIBLE] because that way 00:54:51.290 --> 00:54:54.113 align:middle line:84% you're-- that gets your return [INAUDIBLE]?? 00:54:54.113 --> 00:54:55.030 align:middle line:90% ESTHER DUFLO: Exactly. 00:54:55.030 --> 00:54:57.850 align:middle line:84% It's that each factor gets allocated 00:54:57.850 --> 00:55:00.100 align:middle line:90% to its most productive use. 00:55:00.100 --> 00:55:03.730 align:middle line:84% So the marginal return of this investment is equalized. 00:55:03.730 --> 00:55:05.740 align:middle line:84% So even though there is not one big machine, 00:55:05.740 --> 00:55:07.780 align:middle line:84% we can think of it as one big machine 00:55:07.780 --> 00:55:10.550 align:middle line:84% because the economy is allocating the capital. 00:55:10.550 --> 00:55:14.960 align:middle line:84% So there is such a thing as the return to human capital. 00:55:14.960 --> 00:55:18.557 align:middle line:84% And there is such a thing as the return to physical capital. 00:55:18.557 --> 00:55:19.840 align:middle line:90% OK? 00:55:19.840 --> 00:55:23.510 align:middle line:84% If, for some reason, that's not true, 00:55:23.510 --> 00:55:26.710 align:middle line:84% then we'll see some sector of the economy 00:55:26.710 --> 00:55:29.260 align:middle line:84% with a much higher return to capital 00:55:29.260 --> 00:55:33.340 align:middle line:84% and some sector of the economy with higher return 00:55:33.340 --> 00:55:35.360 align:middle line:90% to human capital. 00:55:35.360 --> 00:55:38.080 align:middle line:84% And this is actually entirely testable. 00:55:38.080 --> 00:55:40.690 align:middle line:84% Because what I just said, well, we can look. 00:55:40.690 --> 00:55:43.540 align:middle line:84% We can look at whether it is, in fact, the case 00:55:43.540 --> 00:55:47.470 align:middle line:84% that, within an economy, the return to investments 00:55:47.470 --> 00:55:49.510 align:middle line:84% are the same from place to place. 00:55:49.510 --> 00:55:52.060 align:middle line:90% 00:55:52.060 --> 00:55:55.120 align:middle line:84% And the point is that it's not true at all. 00:55:55.120 --> 00:55:57.760 align:middle line:84% The return to capital and human capital 00:55:57.760 --> 00:56:01.400 align:middle line:84% vary enormously within countries. 00:56:01.400 --> 00:56:03.970 align:middle line:84% And if you remember the famous [? Lucas ?] paper, 00:56:03.970 --> 00:56:07.420 align:middle line:84% he was marveling at the difference in return 00:56:07.420 --> 00:56:10.720 align:middle line:84% to capital in the US and in India 00:56:10.720 --> 00:56:13.630 align:middle line:84% and said, why isn't all the capital going to the US-- 00:56:13.630 --> 00:56:16.018 align:middle line:90% to-- sorry. 00:56:16.018 --> 00:56:18.310 align:middle line:84% He was marveling at the difference in return to capital 00:56:18.310 --> 00:56:20.500 align:middle line:84% and thinking it's smaller than it should be. 00:56:20.500 --> 00:56:22.930 align:middle line:84% Because there is less capital in India. 00:56:22.930 --> 00:56:26.410 align:middle line:84% But the point is that he shouldn't 00:56:26.410 --> 00:56:29.530 align:middle line:84% have been so surprised, because, within India, there 00:56:29.530 --> 00:56:32.020 align:middle line:84% are huge differences in return to capital. 00:56:32.020 --> 00:56:36.760 align:middle line:84% And yet the capital doesn't move from one place to another, let 00:56:36.760 --> 00:56:41.305 align:middle line:84% alone-- not even within India-- within one town in India 00:56:41.305 --> 00:56:44.740 align:middle line:84% [LAUGHS] do you have huge difference in return 00:56:44.740 --> 00:56:45.700 align:middle line:90% to capital. 00:56:45.700 --> 00:56:49.300 align:middle line:84% Within one industry of one town in India, 00:56:49.300 --> 00:56:51.910 align:middle line:84% you have huge difference in return to capital . 00:56:51.910 --> 00:56:57.730 align:middle line:84% Abhijit has a paper with Kaivan Munshi on the town of Tirupur, 00:56:57.730 --> 00:57:02.620 align:middle line:84% which is the garment manufacturing-- knitted 00:57:02.620 --> 00:57:04.150 align:middle line:90% garment manufacturer. 00:57:04.150 --> 00:57:07.240 align:middle line:84% And they showed that there are two groups of-- 00:57:07.240 --> 00:57:10.260 align:middle line:84% they are sort of the local boys and the-- 00:57:10.260 --> 00:57:13.120 align:middle line:90% 00:57:13.120 --> 00:57:17.920 align:middle line:84% who are the descendants of a rich agricultural community 00:57:17.920 --> 00:57:19.610 align:middle line:90% and got into this business. 00:57:19.610 --> 00:57:22.990 align:middle line:84% And there are a lot of people from other towns 00:57:22.990 --> 00:57:26.650 align:middle line:84% who came to Tirupur when it became a huge capital. 00:57:26.650 --> 00:57:29.740 align:middle line:84% And when you look at the return to capital, 00:57:29.740 --> 00:57:32.050 align:middle line:84% it is much lower for the local boys 00:57:32.050 --> 00:57:35.937 align:middle line:84% than for the people who have come to the city. 00:57:35.937 --> 00:57:37.770 align:middle line:84% So you might think, of course, it's obvious. 00:57:37.770 --> 00:57:41.022 align:middle line:84% The local boys are not particularly good at garments. 00:57:41.022 --> 00:57:43.480 align:middle line:84% They just happen to be there, and they do it like everyone, 00:57:43.480 --> 00:57:46.900 align:middle line:84% whereas the other people have been drawn to this place. 00:57:46.900 --> 00:57:48.400 align:middle line:84% But once you've said that, well, you 00:57:48.400 --> 00:57:50.118 align:middle line:84% say, well, if you live in Tirupur 00:57:50.118 --> 00:57:51.910 align:middle line:84% and you have money to invest, why don't you 00:57:51.910 --> 00:57:53.820 align:middle line:90% give it to the strangers? 00:57:53.820 --> 00:57:55.440 align:middle line:90% Somehow, something stops you. 00:57:55.440 --> 00:57:56.930 align:middle line:90% Right? 00:57:56.930 --> 00:58:01.310 align:middle line:90% Here, what would stop them? 00:58:01.310 --> 00:58:05.150 align:middle line:84% Why, as a well-established Tirupurian 00:58:05.150 --> 00:58:08.240 align:middle line:84% with some amount of money, you would 00:58:08.240 --> 00:58:14.120 align:middle line:84% prefer to set up your son with a little company versus 00:58:14.120 --> 00:58:15.830 align:middle line:84% to give it to someone who just came 00:58:15.830 --> 00:58:17.750 align:middle line:84% and who is evidently super good at this job? 00:58:17.750 --> 00:58:21.052 align:middle line:90% 00:58:21.052 --> 00:58:21.760 align:middle line:90% Yes, [INAUDIBLE]. 00:58:21.760 --> 00:58:22.895 align:middle line:84% AUDIENCE: You don't because of trust? 00:58:22.895 --> 00:58:23.670 align:middle line:90% [INAUDIBLE]-- 00:58:23.670 --> 00:58:24.420 align:middle line:90% ESTHER DUFLO: Yes. 00:58:24.420 --> 00:58:27.660 align:middle line:84% You might not trust that you're going to get it back. 00:58:27.660 --> 00:58:29.210 align:middle line:90% So that's the return-- 00:58:29.210 --> 00:58:32.090 align:middle line:84% the market for capital is very imperfect. 00:58:32.090 --> 00:58:36.050 align:middle line:84% And that means that there are huge differences in return 00:58:36.050 --> 00:58:40.280 align:middle line:90% to capital that stays. 00:58:40.280 --> 00:58:41.990 align:middle line:84% This is, of course, something that-- 00:58:41.990 --> 00:58:44.210 align:middle line:84% I'm giving one example from one town in India, 00:58:44.210 --> 00:58:48.110 align:middle line:84% but it is something that has been documented much more 00:58:48.110 --> 00:58:49.040 align:middle line:90% broadly-- 00:58:49.040 --> 00:58:52.340 align:middle line:84% and in particular by a classic paper by Hsieh and Klenow 00:58:52.340 --> 00:58:57.500 align:middle line:84% which shows that there is huge heterogeneity in return 00:58:57.500 --> 00:58:59.070 align:middle line:90% and that it originates-- 00:58:59.070 --> 00:59:01.080 align:middle line:90% So this is the total factor-- 00:59:01.080 --> 00:59:05.780 align:middle line:84% this is TFPQ, Total Factor Productivity-- so, an idea 00:59:05.780 --> 00:59:09.600 align:middle line:90% of how productive a firm is. 00:59:09.600 --> 00:59:12.090 align:middle line:84% And you see that there is dispersion. 00:59:12.090 --> 00:59:13.050 align:middle line:90% It's not one. 00:59:13.050 --> 00:59:14.080 align:middle line:90% That's the first point. 00:59:14.080 --> 00:59:18.390 align:middle line:84% And second, that dispersion is actually greater in India 00:59:18.390 --> 00:59:20.370 align:middle line:90% than it is in China and the US. 00:59:20.370 --> 00:59:24.810 align:middle line:90% 00:59:24.810 --> 00:59:28.080 align:middle line:84% This is also something that we need to explain-- 00:59:28.080 --> 00:59:30.570 align:middle line:84% that, why-- that we can think about, 00:59:30.570 --> 00:59:34.980 align:middle line:84% why are the markets not working as well to allocate the capital 00:59:34.980 --> 00:59:39.030 align:middle line:84% to its best use, so that only the firm-- 00:59:39.030 --> 00:59:42.450 align:middle line:84% only the most productive firms stay and the other ones 00:59:42.450 --> 00:59:44.212 align:middle line:90% disappear. 00:59:44.212 --> 00:59:44.920 align:middle line:90% That makes sense? 00:59:44.920 --> 00:59:47.860 align:middle line:90% 00:59:47.860 --> 00:59:49.800 align:middle line:90% So where are we now? 00:59:49.800 --> 00:59:52.560 align:middle line:84% A fundamental source in the difference in productivity 00:59:52.560 --> 00:59:56.940 align:middle line:84% across countries is the fact that both h and k 00:59:56.940 --> 00:59:59.670 align:middle line:84% are misallocated, due to a number of factors. 00:59:59.670 --> 01:00:02.220 align:middle line:84% And a number of factors is [INAUDIBLE] 01:00:02.220 --> 01:00:06.060 align:middle line:84% what we are going to concern ourselves with, 01:00:06.060 --> 01:00:07.560 align:middle line:90% in this entire sequence. 01:00:07.560 --> 01:00:11.040 align:middle line:90% 01:00:11.040 --> 01:00:14.370 align:middle line:84% First we are going to look at whether, to some extent, 01:00:14.370 --> 01:00:16.470 align:middle line:90% it comes from poverty itself-- 01:00:16.470 --> 01:00:18.150 align:middle line:84% whether there is something fundamentally 01:00:18.150 --> 01:00:22.980 align:middle line:84% different about poverty which means that you can't accumulate 01:00:22.980 --> 01:00:25.090 align:middle line:90% your way out of poverty. 01:00:25.090 --> 01:00:27.710 align:middle line:84% So that's the idea of poverty trap. 01:00:27.710 --> 01:00:31.790 align:middle line:84% Then we look at where h comes from and why 01:00:31.790 --> 01:00:33.750 align:middle line:84% this heterogeneity might persist-- 01:00:33.750 --> 01:00:35.990 align:middle line:84% why people are not investing optimally. 01:00:35.990 --> 01:00:38.390 align:middle line:84% So we look at education, health, nutrition-- 01:00:38.390 --> 01:00:41.120 align:middle line:84% so that's a lot of what I am going to be doing. 01:00:41.120 --> 01:00:43.370 align:middle line:84% Then we'll look at heterogeneity in k. 01:00:43.370 --> 01:00:47.870 align:middle line:84% So we'll look at credit and savings and then land. 01:00:47.870 --> 01:00:51.080 align:middle line:84% And then allocation of the labor market. 01:00:51.080 --> 01:00:55.700 align:middle line:84% That's another-- we only put h, but h is really l times 01:00:55.700 --> 01:00:58.220 align:middle line:84% the education and health of each person-- 01:00:58.220 --> 01:00:59.840 align:middle line:84% the labor market itself, which is 01:00:59.840 --> 01:01:03.050 align:middle line:84% a super active area in development these days 01:01:03.050 --> 01:01:05.690 align:middle line:90% and super fun. 01:01:05.690 --> 01:01:08.890 align:middle line:84% And then finally, we'll look at A, 01:01:08.890 --> 01:01:11.435 align:middle line:84% in the second semester specifically, 01:01:11.435 --> 01:01:13.310 align:middle line:84% looking at technology, organization of firms. 01:01:13.310 --> 01:01:15.830 align:middle line:90% 01:01:15.830 --> 01:01:17.955 align:middle line:84% We'll look at policies along the way-- 01:01:17.955 --> 01:01:19.580 align:middle line:84% Of course, when I talk about education, 01:01:19.580 --> 01:01:21.560 align:middle line:84% I'm going to be talking about education policy. 01:01:21.560 --> 01:01:22.760 align:middle line:84% When I talk about health, I'm going 01:01:22.760 --> 01:01:24.177 align:middle line:84% to be talking about health policy. 01:01:24.177 --> 01:01:27.050 align:middle line:84% But as Ben mentioned, there will be a special sequence 01:01:27.050 --> 01:01:29.930 align:middle line:84% at the end which will be specifically 01:01:29.930 --> 01:01:34.080 align:middle line:90% about the public sector. 01:01:34.080 --> 01:01:38.735 align:middle line:84% So that gives you a sense of very broad road 01:01:38.735 --> 01:01:40.235 align:middle line:90% map for the course. 01:01:40.235 --> 01:01:43.670 align:middle line:90% 01:01:43.670 --> 01:01:49.860 align:middle line:84% Before I go in more detail-- we have about five minutes left-- 01:01:49.860 --> 01:01:52.010 align:middle line:84% I was going to give you some teasers. 01:01:52.010 --> 01:01:55.190 align:middle line:84% But before that, let me see if there is any question, 01:01:55.190 --> 01:01:57.320 align:middle line:90% comments, point of debate. 01:01:57.320 --> 01:02:01.080 align:middle line:84% Because what's coming now is not particular. 01:02:01.080 --> 01:02:03.440 align:middle line:84% We can go to it or not go to it, so I would rather 01:02:03.440 --> 01:02:06.785 align:middle line:84% make sure you have a chance to each express yourself. 01:02:06.785 --> 01:02:15.740 align:middle line:90% 01:02:15.740 --> 01:02:18.530 align:middle line:90% Going, going? 01:02:18.530 --> 01:02:19.970 align:middle line:90% Then I will speak. 01:02:19.970 --> 01:02:22.430 align:middle line:84% So another view of the course, and a few fun facts. 01:02:22.430 --> 01:02:24.470 align:middle line:84% I'm not going to give you the entire overview, 01:02:24.470 --> 01:02:26.850 align:middle line:90% but I'm going to start. 01:02:26.850 --> 01:02:29.630 align:middle line:84% So we'll start with the idea of poverty trap. 01:02:29.630 --> 01:02:31.640 align:middle line:84% And the other question is simple. 01:02:31.640 --> 01:02:35.200 align:middle line:84% Is the fact that you're poor keeping you poor? 01:02:35.200 --> 01:02:37.400 align:middle line:90% So why would it be the case? 01:02:37.400 --> 01:02:42.640 align:middle line:84% Well, because of what we'll come to know and love 01:02:42.640 --> 01:02:44.918 align:middle line:90% as the S-shaped curve. 01:02:44.918 --> 01:02:46.210 align:middle line:90% [? Well, ?] the S-shaped curve. 01:02:46.210 --> 01:02:49.900 align:middle line:84% This is from a paper by Clare Balboni, who is teaching here, 01:02:49.900 --> 01:02:54.880 align:middle line:84% and many other coauthors, looking at people 01:02:54.880 --> 01:02:57.070 align:middle line:90% who are extremely poor-- 01:02:57.070 --> 01:02:58.840 align:middle line:84% because they were part of a program 01:02:58.840 --> 01:03:01.360 align:middle line:90% to help the extremely poor. 01:03:01.360 --> 01:03:04.240 align:middle line:84% And some of them received a transfer 01:03:04.240 --> 01:03:07.780 align:middle line:90% that make them less poor. 01:03:07.780 --> 01:03:10.810 align:middle line:90% And they are looking at-- 01:03:10.810 --> 01:03:12.950 align:middle line:84% So this is not the impact of the transfer at all. 01:03:12.950 --> 01:03:16.540 align:middle line:84% This is something that is purely descriptive, where, 01:03:16.540 --> 01:03:21.460 align:middle line:84% on the x-axis, you have how much money in total 01:03:21.460 --> 01:03:26.170 align:middle line:84% they had in assets, including the transfer, in 2007, 01:03:26.170 --> 01:03:30.590 align:middle line:84% and then productive assets in 2011. 01:03:30.590 --> 01:03:34.940 align:middle line:84% So it's how much they grew between the two periods. 01:03:34.940 --> 01:03:42.460 align:middle line:84% And you will notice that there is an interesting curve, here, 01:03:42.460 --> 01:03:44.570 align:middle line:90% which is S-shaped. 01:03:44.570 --> 01:03:45.890 align:middle line:90% So what does this mean? 01:03:45.890 --> 01:03:50.640 align:middle line:90% Suppose that you continue to-- 01:03:50.640 --> 01:03:55.790 align:middle line:84% you do a next curve that goes from 2011, 2015. 01:03:55.790 --> 01:03:57.960 align:middle line:90% Suppose the curve is the same. 01:03:57.960 --> 01:04:00.030 align:middle line:90% And you're trying to find out-- 01:04:00.030 --> 01:04:02.250 align:middle line:90% someone who started in 2007-- 01:04:02.250 --> 01:04:06.570 align:middle line:84% by 5 years later, 10 years later, 15 years later, 01:04:06.570 --> 01:04:08.450 align:middle line:90% where will they end up? 01:04:08.450 --> 01:04:11.960 align:middle line:84% And suppose that's the law-- is that, if you 01:04:11.960 --> 01:04:14.480 align:middle line:84% start with this much money, you'll 01:04:14.480 --> 01:04:16.260 align:middle line:90% end up with this much money. 01:04:16.260 --> 01:04:16.760 align:middle line:90% OK? 01:04:16.760 --> 01:04:18.660 align:middle line:90% You understand the exercise? 01:04:18.660 --> 01:04:22.190 align:middle line:84% So take someone who starts with 2.6-- 01:04:22.190 --> 01:04:26.910 align:middle line:84% whatever it means, some 2.6 monies. 01:04:26.910 --> 01:04:33.990 align:middle line:84% In 2011, they have 2.7 maybe monies. 01:04:33.990 --> 01:04:36.060 align:middle line:84% How do I know how much they are-- 01:04:36.060 --> 01:04:42.654 align:middle line:84% how do I continue this to know what they'll have in 2015? 01:04:42.654 --> 01:04:44.120 align:middle line:90% AUDIENCE: [INAUDIBLE] 01:04:44.120 --> 01:04:45.140 align:middle line:90% ESTHER DUFLO: Yes? 01:04:45.140 --> 01:04:45.650 align:middle line:90% And then? 01:04:45.650 --> 01:04:47.920 align:middle line:90% AUDIENCE: [INAUDIBLE] 01:04:47.920 --> 01:04:49.010 align:middle line:90% 01:04:49.010 --> 01:04:49.760 align:middle line:90% ESTHER DUFLO: Yes. 01:04:49.760 --> 01:04:50.990 align:middle line:90% Going back to the x-axis. 01:04:50.990 --> 01:04:53.760 align:middle line:90% And from the x-axis, back up. 01:04:53.760 --> 01:04:56.570 align:middle line:84% So we can even skip the going down. 01:04:56.570 --> 01:05:05.120 align:middle line:84% But we'll have-- so in 2007, they start at 2.6. 01:05:05.120 --> 01:05:06.680 align:middle line:90% They end up here. 01:05:06.680 --> 01:05:08.810 align:middle line:90% And then to know-- 01:05:08.810 --> 01:05:10.010 align:middle line:90% so this is the level. 01:05:10.010 --> 01:05:12.290 align:middle line:84% So, to know how they have in 2015, 01:05:12.290 --> 01:05:14.000 align:middle line:90% I go here and then back up. 01:05:14.000 --> 01:05:17.720 align:middle line:90% And then 2000-- what was it? 01:05:17.720 --> 01:05:23.130 align:middle line:84% Even five years later, I continue-- 01:05:23.130 --> 01:05:24.180 align:middle line:90% continue. 01:05:24.180 --> 01:05:26.980 align:middle line:90% So where does this end? 01:05:26.980 --> 01:05:28.410 align:middle line:90% Here, where they cross. 01:05:28.410 --> 01:05:31.030 align:middle line:90% All right? 01:05:31.030 --> 01:05:38.540 align:middle line:84% And then someone who started here, in 2007, 01:05:38.540 --> 01:05:44.763 align:middle line:84% they go here, then they go here, then they go here, 01:05:44.763 --> 01:05:47.180 align:middle line:84% then they go here, and then there is a little thingy here, 01:05:47.180 --> 01:05:48.620 align:middle line:90% but then they end up over here. 01:05:48.620 --> 01:05:50.950 align:middle line:90% That's the steady state. 01:05:50.950 --> 01:05:53.260 align:middle line:90% So that's the poverty trap. 01:05:53.260 --> 01:05:56.590 align:middle line:84% It's the idea that there might be multiple steady states-- 01:05:56.590 --> 01:06:04.790 align:middle line:84% that where you start from has an impact of where you end up. 01:06:04.790 --> 01:06:06.370 align:middle line:84% OK, we're going to go back to this. 01:06:06.370 --> 01:06:08.980 align:middle line:84% We're going to spend a lot of time to this graph. 01:06:08.980 --> 01:06:11.440 align:middle line:90% But that's something which is-- 01:06:11.440 --> 01:06:17.170 align:middle line:84% So why we don't have poverty traps in most of the models 01:06:17.170 --> 01:06:21.220 align:middle line:84% that we work with that are appropriate for rich countries 01:06:21.220 --> 01:06:25.660 align:middle line:90% is that we assume concavity. 01:06:25.660 --> 01:06:28.630 align:middle line:84% So this part of the curves goes away. 01:06:28.630 --> 01:06:34.200 align:middle line:84% So when people are poorer, the capital 01:06:34.200 --> 01:06:36.840 align:middle line:90% is very productive for them. 01:06:36.840 --> 01:06:40.140 align:middle line:84% So they're making more money out of the little money they have, 01:06:40.140 --> 01:06:41.670 align:middle line:90% proportionally. 01:06:41.670 --> 01:06:45.420 align:middle line:84% Then, therefore, they can progressively, 01:06:45.420 --> 01:06:48.480 align:middle line:84% little by little, build their nest eggs 01:06:48.480 --> 01:06:52.890 align:middle line:84% to end up all in the same steady state. 01:06:52.890 --> 01:06:55.440 align:middle line:84% With the poverty trap, some people 01:06:55.440 --> 01:06:58.380 align:middle line:84% are so poor that they end up there, 01:06:58.380 --> 01:07:00.750 align:middle line:90% and they end up staying poor. 01:07:00.750 --> 01:07:01.890 align:middle line:90% And so the game-- 01:07:01.890 --> 01:07:04.650 align:middle line:84% or not the "game"-- the exercise-- is to think, 01:07:04.650 --> 01:07:06.060 align:middle line:90% where does it come from? 01:07:06.060 --> 01:07:07.960 align:middle line:90% And why-- 01:07:07.960 --> 01:07:11.010 align:middle line:84% This is, in a sense, a very reduced-form picture, 01:07:11.010 --> 01:07:12.900 align:middle line:84% but from real data, telling you in, 01:07:12.900 --> 01:07:16.870 align:middle line:84% Bangladesh there appears to be a poverty trap. 01:07:16.870 --> 01:07:22.090 align:middle line:84% Which, by the way, is the reasoning for this program 01:07:22.090 --> 01:07:25.360 align:middle line:84% in the first place, which is a program that was started 01:07:25.360 --> 01:07:28.900 align:middle line:84% by an NGO called BRAC and was expanded since then in many 01:07:28.900 --> 01:07:33.910 align:middle line:84% countries, with the idea that you cannot go big push. 01:07:33.910 --> 01:07:37.060 align:middle line:84% You give an asset to people, and you help them for a while, 01:07:37.060 --> 01:07:38.740 align:middle line:90% to help them push-- 01:07:38.740 --> 01:07:43.560 align:middle line:84% If a lot of people have been on this part of the curve, 01:07:43.560 --> 01:07:45.740 align:middle line:90% you try and push them. 01:07:45.740 --> 01:07:49.010 align:middle line:84% And what's underlying this picture, in a sense, 01:07:49.010 --> 01:07:53.090 align:middle line:84% is that maybe the transfer was not large enough 01:07:53.090 --> 01:07:57.620 align:middle line:84% to push the poorest of the poor who were in these villages 01:07:57.620 --> 01:08:01.340 align:middle line:84% even away from the poverty trap, so they fell back. 01:08:01.340 --> 01:08:04.280 align:middle line:84% While some of the people that pushed them up, 01:08:04.280 --> 01:08:09.030 align:middle line:84% and that produced long-term impact of this program. 01:08:09.030 --> 01:08:10.560 align:middle line:90% So we'll start with that. 01:08:10.560 --> 01:08:13.230 align:middle line:84% We'll spend two lectures on poverty trap, one 01:08:13.230 --> 01:08:17.624 align:middle line:84% on thinking about theory, one on going back 01:08:17.624 --> 01:08:20.689 align:middle line:84% to this program and similar programs. 01:08:20.689 --> 01:08:21.189 align:middle line:90% Oops. 01:08:21.189 --> 01:08:33.410 align:middle line:90% 01:08:33.410 --> 01:08:35.930 align:middle line:84% When we are done with that, we'll-- so first we'll try 01:08:35.930 --> 01:08:38.990 align:middle line:90% to spend some time-- 01:08:38.990 --> 01:08:43.670 align:middle line:84% so-- I've already showed my spoiler alert. 01:08:43.670 --> 01:08:45.290 align:middle line:90% There are poverty traps. 01:08:45.290 --> 01:08:47.670 align:middle line:90% On the other hand, why? 01:08:47.670 --> 01:08:48.899 align:middle line:90% Where are they coming from? 01:08:48.899 --> 01:08:52.050 align:middle line:84% That is something where there is still-- we still don't know. 01:08:52.050 --> 01:08:54.569 align:middle line:84% So it's like a super interesting, exciting area 01:08:54.569 --> 01:08:57.779 align:middle line:84% of research is, what causes poverty trap? 01:08:57.779 --> 01:09:00.630 align:middle line:90% There are many possible reasons. 01:09:00.630 --> 01:09:04.290 align:middle line:84% Then we'll talk about human capital. 01:09:04.290 --> 01:09:08.279 align:middle line:84% We'll talk about the quality of education, the supply 01:09:08.279 --> 01:09:10.529 align:middle line:84% side of education-- in particular, the fact 01:09:10.529 --> 01:09:13.590 align:middle line:84% that the quality of education is terrible in developing 01:09:13.590 --> 01:09:14.500 align:middle line:90% countries. 01:09:14.500 --> 01:09:18.240 align:middle line:84% So I said at the beginning that a lot of people go to school. 01:09:18.240 --> 01:09:21.060 align:middle line:84% That's great, but they go to school and they learn nothing. 01:09:21.060 --> 01:09:23.729 align:middle line:84% This is a graph that some of you might have seen. 01:09:23.729 --> 01:09:26.189 align:middle line:84% It's from the Annual Status of Education Report, 01:09:26.189 --> 01:09:30.040 align:middle line:84% in India, that looks at the fraction of kids 01:09:30.040 --> 01:09:31.170 align:middle line:90% in standard five-- 01:09:31.170 --> 01:09:32.819 align:middle line:84% so, end of primary school-- who can 01:09:32.819 --> 01:09:34.300 align:middle line:84% read at the level of standard two-- 01:09:34.300 --> 01:09:36.479 align:middle line:90% so, beginning of primary school. 01:09:36.479 --> 01:09:40.080 align:middle line:84% And you can see that not only it's not improving over time, 01:09:40.080 --> 01:09:43.710 align:middle line:84% it's, if anything, worsening over time. 01:09:43.710 --> 01:09:46.700 align:middle line:84% Of course, we have a lot of-- you can say it's-- and it's not 01:09:46.700 --> 01:09:50.330 align:middle line:84% just a composition effect, in the sense that it's not just 01:09:50.330 --> 01:09:52.970 align:middle line:84% because there are more kids in school and therefore 01:09:52.970 --> 01:09:54.690 align:middle line:90% the results are worse. 01:09:54.690 --> 01:09:56.880 align:middle line:90% So we'll talk about education. 01:09:56.880 --> 01:10:01.150 align:middle line:84% We'll talk about health, obviously, 01:10:01.150 --> 01:10:03.970 align:middle line:90% both the supply and the demand. 01:10:03.970 --> 01:10:11.530 align:middle line:84% Then Ben will talk about labor markets, land markets, credit. 01:10:11.530 --> 01:10:14.300 align:middle line:84% Technology will actually be in the second part of the course. 01:10:14.300 --> 01:10:16.300 align:middle line:84% I'll talk about savings on the other end. 01:10:16.300 --> 01:10:22.420 align:middle line:84% And finally Ben will finish with the public sector. 01:10:22.420 --> 01:10:25.090 align:middle line:84% What are the problems with redistribution? 01:10:25.090 --> 01:10:27.250 align:middle line:84% What's the optimal shape of transfer? 01:10:27.250 --> 01:10:30.070 align:middle line:84% Are there problems in providing public goods? 01:10:30.070 --> 01:10:32.350 align:middle line:84% Are there problems, labor-market problems, 01:10:32.350 --> 01:10:35.350 align:middle line:84% specific to the public sectors, with incentives and things 01:10:35.350 --> 01:10:36.770 align:middle line:90% like that? 01:10:36.770 --> 01:10:41.000 align:middle line:84% So that's going to be the game plan for this semester. 01:10:41.000 --> 01:10:45.985 align:middle line:84% So I hope that you will continue to be here. 01:10:45.985 --> 01:10:48.620 align:middle line:90% 01:10:48.620 --> 01:10:53.420 align:middle line:84% I wish you an excellent semester, back here in person. 01:10:53.420 --> 01:10:57.790 align:middle line:84% And we'll have fun in this class. 01:10:57.790 --> 01:11:10.000 align:middle line:90%