WEBVTT 00:00:00.000 --> 00:00:00.500 align:middle line:90% 00:00:00.500 --> 00:00:01.967 align:middle line:90% [SQUEAKING] 00:00:01.967 --> 00:00:03.923 align:middle line:90% [RUSTLING] 00:00:03.923 --> 00:00:07.346 align:middle line:90% [CLICKING] 00:00:07.346 --> 00:00:10.197 align:middle line:90% 00:00:10.197 --> 00:00:12.280 align:middle line:84% JONATHAN GRUBER: So we're talking about chapter 9, 00:00:12.280 --> 00:00:14.160 align:middle line:90% about political economy. 00:00:14.160 --> 00:00:16.040 align:middle line:84% Let's remember where we've been so far. 00:00:16.040 --> 00:00:19.960 align:middle line:84% We've talked about our first best outcome 00:00:19.960 --> 00:00:22.200 align:middle line:84% or what we called unanimous consent, 00:00:22.200 --> 00:00:25.360 align:middle line:84% which, under certain theoretical conditions that'll never be met, 00:00:25.360 --> 00:00:27.940 align:middle line:84% we could actually meet the Samuelson rule. 00:00:27.940 --> 00:00:30.600 align:middle line:84% We'd actually get everyone to contribute and provide 00:00:30.600 --> 00:00:32.560 align:middle line:84% the optimal level of public goods. 00:00:32.560 --> 00:00:35.180 align:middle line:84% Then we said, well, look, that's not realistic. 00:00:35.180 --> 00:00:37.473 align:middle line:84% What we actually do in society is we vote. 00:00:37.473 --> 00:00:39.640 align:middle line:84% So let's start with society where actually all votes 00:00:39.640 --> 00:00:41.020 align:middle line:90% are done by direct democracy. 00:00:41.020 --> 00:00:42.900 align:middle line:84% And we talked about direct democracy. 00:00:42.900 --> 00:00:47.560 align:middle line:84% And we talked about when direct democracy will or will not 00:00:47.560 --> 00:00:49.680 align:middle line:90% deliver consistent outcomes. 00:00:49.680 --> 00:00:52.960 align:middle line:84% And in particular, we talked about the fact 00:00:52.960 --> 00:00:56.320 align:middle line:84% that if you assume preferences are single-peaked, then 00:00:56.320 --> 00:00:59.220 align:middle line:84% majority voting can deliver consistent outcomes. 00:00:59.220 --> 00:01:01.590 align:middle line:84% And moreover, it delivers this wonderful thing 00:01:01.590 --> 00:01:03.710 align:middle line:84% we call the Median Voter Theorem, which 00:01:03.710 --> 00:01:05.682 align:middle line:84% is to know what society wants, you only 00:01:05.682 --> 00:01:07.390 align:middle line:84% have to know what the median voter wants, 00:01:07.390 --> 00:01:09.150 align:middle line:84% or what we call today, swing voter. 00:01:09.150 --> 00:01:10.830 align:middle line:84% That's all you need to know, which 00:01:10.830 --> 00:01:14.670 align:middle line:84% we said has the advantage that it's incredibly data-efficient. 00:01:14.670 --> 00:01:17.590 align:middle line:84% Has the disadvantage that it is not necessarily the first best 00:01:17.590 --> 00:01:19.710 align:middle line:84% outcome, that what the median voter wants 00:01:19.710 --> 00:01:23.270 align:middle line:84% may not be what meets the Samuelson Rule. 00:01:23.270 --> 00:01:26.310 align:middle line:84% In particular, if the median of the distribution 00:01:26.310 --> 00:01:28.772 align:middle line:84% does not equal the mean of the preference distribution, 00:01:28.772 --> 00:01:30.230 align:middle line:84% it will not meet the Samuelson Rule 00:01:30.230 --> 00:01:32.450 align:middle line:90% with the median voter outcome. 00:01:32.450 --> 00:01:34.510 align:middle line:84% So that's where we've gotten to so far. 00:01:34.510 --> 00:01:38.710 align:middle line:84% Questions about what we covered last time or any of that stuff? 00:01:38.710 --> 00:01:42.643 align:middle line:84% Today we get to the fun stuff, which is well, in reality, 00:01:42.643 --> 00:01:44.310 align:middle line:84% most decisions that are made politically 00:01:44.310 --> 00:01:47.530 align:middle line:84% are not made either by a Lindahl mechanism or direct democracy, 00:01:47.530 --> 00:01:49.345 align:middle line:90% but by representative democracy. 00:01:49.345 --> 00:01:50.970 align:middle line:84% And that's what we'll talk about today, 00:01:50.970 --> 00:01:53.190 align:middle line:84% which is the promise and pitfalls 00:01:53.190 --> 00:01:55.270 align:middle line:90% of representative democracy. 00:01:55.270 --> 00:01:57.350 align:middle line:84% Now what we're going to start with is 00:01:57.350 --> 00:02:01.170 align:middle line:84% another version of the median voter theorem, which 00:02:01.170 --> 00:02:05.370 align:middle line:84% is we're going to show that, in fact, even if you elect 00:02:05.370 --> 00:02:08.710 align:middle line:84% your representatives, under certain conditions, 00:02:08.710 --> 00:02:11.650 align:middle line:84% you will still get the median voter outcome. 00:02:11.650 --> 00:02:13.370 align:middle line:90% Actually, only under one-- 00:02:13.370 --> 00:02:15.210 align:middle line:90% single-peaked preferences still. 00:02:15.210 --> 00:02:16.930 align:middle line:90% And you need one other thing. 00:02:16.930 --> 00:02:20.250 align:middle line:84% You need that politicians are vote maximizers, 00:02:20.250 --> 00:02:24.548 align:middle line:84% that their sole goal is to maximize their votes. 00:02:24.548 --> 00:02:26.090 align:middle line:84% If you take single-peaked preferences 00:02:26.090 --> 00:02:27.910 align:middle line:84% and have vote-maximizing politicians, 00:02:27.910 --> 00:02:30.050 align:middle line:84% it turns out that it doesn't matter 00:02:30.050 --> 00:02:33.050 align:middle line:84% if you have direct democracy or representative democracy. 00:02:33.050 --> 00:02:35.250 align:middle line:84% You will still get the median voter outcome. 00:02:35.250 --> 00:02:37.942 align:middle line:84% And to see that, let's go to figure 9.3. 00:02:37.942 --> 00:02:39.650 align:middle line:84% So imagine these two candidates running-- 00:02:39.650 --> 00:02:40.590 align:middle line:90% Joe and Donald. 00:02:40.590 --> 00:02:42.850 align:middle line:84% Now the next edition would be Kamala and Donald. 00:02:42.850 --> 00:02:47.070 align:middle line:84% And imagine it's a one-issue election. 00:02:47.070 --> 00:02:49.690 align:middle line:84% And the issue is how much to spend on defense. 00:02:49.690 --> 00:02:53.090 align:middle line:84% And there's some range of spending, from 0% to 50%. 00:02:53.090 --> 00:02:57.200 align:middle line:84% And let's assume voters are lined up along this line. 00:02:57.200 --> 00:02:59.480 align:middle line:84% This line represents voters' preferences, 00:02:59.480 --> 00:03:02.600 align:middle line:90% and they're equally distributed. 00:03:02.600 --> 00:03:05.685 align:middle line:84% It doesn't have to be a normal-- it doesn't 00:03:05.685 --> 00:03:07.060 align:middle line:84% have to be a normal distribution, 00:03:07.060 --> 00:03:09.660 align:middle line:84% but it has to be a symmetric distribution. 00:03:09.660 --> 00:03:12.640 align:middle line:84% Imagine it's symmetrically distributed along this line. 00:03:12.640 --> 00:03:14.520 align:middle line:84% And Joe and Donald are considering 00:03:14.520 --> 00:03:16.200 align:middle line:90% what positions to take. 00:03:16.200 --> 00:03:20.440 align:middle line:84% So if Donald says, look, I'm going to advocate 00:03:20.440 --> 00:03:23.260 align:middle line:84% for a lot of defense spending, and Joe says, 00:03:23.260 --> 00:03:25.480 align:middle line:84% I'm going to advocate for little defense spending, 00:03:25.480 --> 00:03:27.240 align:middle line:90% then they'll split the vote. 00:03:27.240 --> 00:03:29.060 align:middle line:84% People go who's ever closest to them. 00:03:29.060 --> 00:03:31.760 align:middle line:84% And the people in the middle will be just indifferent. 00:03:31.760 --> 00:03:35.040 align:middle line:84% In that situation, it is in Joe's interest 00:03:35.040 --> 00:03:36.800 align:middle line:90% to move to the center. 00:03:36.800 --> 00:03:37.520 align:middle line:90% Why? 00:03:37.520 --> 00:03:38.640 align:middle line:84% Because by moving to the center, he 00:03:38.640 --> 00:03:40.040 align:middle line:84% doesn't lose anybody on the left. 00:03:40.040 --> 00:03:41.820 align:middle line:90% He still preferred to Donald. 00:03:41.820 --> 00:03:44.440 align:middle line:84% But he gets some people towards the center 00:03:44.440 --> 00:03:46.000 align:middle line:84% who now are closer to his preferences 00:03:46.000 --> 00:03:48.240 align:middle line:84% than they are to Donald's preferences. 00:03:48.240 --> 00:03:50.280 align:middle line:84% Well Donald, seeing this, could immediately 00:03:50.280 --> 00:03:52.960 align:middle line:84% react by moving even closer to the center. 00:03:52.960 --> 00:03:54.840 align:middle line:90% And then he'll get more votes. 00:03:54.840 --> 00:03:56.860 align:middle line:84% And you can see where this ends up. 00:03:56.860 --> 00:04:00.040 align:middle line:84% This game ends up, clearly, with both people at the center. 00:04:00.040 --> 00:04:01.980 align:middle line:84% And indeed, it's a Nash equilibrium. 00:04:01.980 --> 00:04:04.980 align:middle line:84% There's no incentive to deviate, because the minute you move away 00:04:04.980 --> 00:04:09.180 align:middle line:84% from the perfect center, you cede votes to your opponent. 00:04:09.180 --> 00:04:12.580 align:middle line:84% So this says that the Median Voter Theorem is not just 00:04:12.580 --> 00:04:15.420 align:middle line:84% a theorem for direct democracy, but it's actually 00:04:15.420 --> 00:04:18.220 align:middle line:84% a theorem that applies to representative democracy 00:04:18.220 --> 00:04:19.779 align:middle line:90% as well. 00:04:19.779 --> 00:04:23.460 align:middle line:84% But it turns out it only applies to representative democracy 00:04:23.460 --> 00:04:25.540 align:middle line:84% under a wide variety of assumptions 00:04:25.540 --> 00:04:26.980 align:middle line:90% that I've implicitly made. 00:04:26.980 --> 00:04:29.260 align:middle line:90% So why don't we talk through-- 00:04:29.260 --> 00:04:31.458 align:middle line:90% where's my chalk? 00:04:31.458 --> 00:04:32.500 align:middle line:90% I'll use big chalk today. 00:04:32.500 --> 00:04:33.280 align:middle line:90% Here's one. 00:04:33.280 --> 00:04:35.340 align:middle line:84% Why don't we talk through, what are 00:04:35.340 --> 00:04:39.502 align:middle line:84% the implicit assumptions I've made in laying out this model? 00:04:39.502 --> 00:04:41.460 align:middle line:84% What are some things about reality that deviate 00:04:41.460 --> 00:04:42.460 align:middle line:90% from what I've laid out? 00:04:42.460 --> 00:04:43.008 align:middle line:90% Alec? 00:04:43.008 --> 00:04:44.300 align:middle line:90% AUDIENCE: That we will abstain. 00:04:44.300 --> 00:04:45.217 align:middle line:90% JONATHAN GRUBER: Yeah. 00:04:45.217 --> 00:04:51.660 align:middle line:84% So I've assumed universal voting, universal as opposed 00:04:51.660 --> 00:05:00.570 align:middle line:90% to non-selective voting. 00:05:00.570 --> 00:05:01.810 align:middle line:90% What else? 00:05:01.810 --> 00:05:03.940 align:middle line:90% What else is-- yeah? 00:05:03.940 --> 00:05:06.273 align:middle line:84% AUDIENCE: [INAUDIBLE] even if pick one of the two people 00:05:06.273 --> 00:05:07.482 align:middle line:90% JONATHAN GRUBER: What's that? 00:05:07.482 --> 00:05:10.010 align:middle line:84% AUDIENCE: You're [INAUDIBLE] pick one of those two people. 00:05:10.010 --> 00:05:10.310 align:middle line:90% JONATHAN GRUBER: Right. 00:05:10.310 --> 00:05:12.130 align:middle line:84% So I've assumed only two candidates. 00:05:12.130 --> 00:05:15.570 align:middle line:90% In fact, it turns out-- 00:05:15.570 --> 00:05:18.107 align:middle line:84% and those of you who are game theoretically-minded 00:05:18.107 --> 00:05:19.690 align:middle line:84% will see this immediately-- this model 00:05:19.690 --> 00:05:21.343 align:middle line:84% breaks down with three candidates 00:05:21.343 --> 00:05:23.010 align:middle line:84% because you can always gain market share 00:05:23.010 --> 00:05:24.430 align:middle line:90% by continuing to move. 00:05:24.430 --> 00:05:26.690 align:middle line:84% Once there's three candidates, there's no equilibrium. 00:05:26.690 --> 00:05:29.170 align:middle line:84% There's no Nash equilibrium with three candidates. 00:05:29.170 --> 00:05:31.210 align:middle line:90% So I've assumed two candidates. 00:05:31.210 --> 00:05:33.210 align:middle line:90% What else? 00:05:33.210 --> 00:05:35.550 align:middle line:84% What else have I assumed that's violated in reality? 00:05:35.550 --> 00:05:38.088 align:middle line:90% Yeah? 00:05:38.088 --> 00:05:40.130 align:middle line:84% AUDIENCE: Candidates only care about one and two. 00:05:40.130 --> 00:05:44.530 align:middle line:84% JONATHAN GRUBER: I've assumed, fundamentally, no ideology. 00:05:44.530 --> 00:05:46.850 align:middle line:84% I've assumed that candidates only 00:05:46.850 --> 00:05:48.850 align:middle line:84% are vote-maximizing machines that don't actually 00:05:48.850 --> 00:05:50.290 align:middle line:90% care about anything. 00:05:50.290 --> 00:05:50.790 align:middle line:90% Steven? 00:05:50.790 --> 00:05:52.207 align:middle line:84% AUDIENCE: I was going to say that. 00:05:52.207 --> 00:05:53.180 align:middle line:90% JONATHAN GRUBER: Yeah. 00:05:53.180 --> 00:05:54.600 align:middle line:90% Good. 00:05:54.600 --> 00:05:55.440 align:middle line:90% Others? 00:05:55.440 --> 00:05:55.940 align:middle line:90% What else? 00:05:55.940 --> 00:05:57.250 align:middle line:90% What else have I done here? 00:05:57.250 --> 00:05:58.500 align:middle line:90% What else have I assumed here? 00:05:58.500 --> 00:06:05.100 align:middle line:90% 00:06:05.100 --> 00:06:07.700 align:middle line:84% Well, let's go through these, and then we'll talk. 00:06:07.700 --> 00:06:11.920 align:middle line:84% So first of all, I've assumed universal voting. 00:06:11.920 --> 00:06:14.178 align:middle line:84% Once there's non-universal voting, 00:06:14.178 --> 00:06:15.720 align:middle line:84% people might deviate from the middle. 00:06:15.720 --> 00:06:18.162 align:middle line:84% In particular, if voters on one end of the spectrum-- 00:06:18.162 --> 00:06:20.120 align:middle line:84% if people on one end of the spectrum or another 00:06:20.120 --> 00:06:21.658 align:middle line:84% are more likely to vote, there'll 00:06:21.658 --> 00:06:24.200 align:middle line:84% be an incentive to move away from the middle towards that end 00:06:24.200 --> 00:06:25.240 align:middle line:90% of the spectrum. 00:06:25.240 --> 00:06:27.520 align:middle line:84% I think that's pretty clear to see. 00:06:27.520 --> 00:06:28.900 align:middle line:84% So first of all-- now once again, 00:06:28.900 --> 00:06:31.358 align:middle line:84% if there's non-selective voting that's equally distributed, 00:06:31.358 --> 00:06:32.400 align:middle line:90% that doesn't matter. 00:06:32.400 --> 00:06:36.240 align:middle line:84% So if randomly, half the people vote, the model still works. 00:06:36.240 --> 00:06:39.040 align:middle line:84% It's only if there's differential selective voting. 00:06:39.040 --> 00:06:41.832 align:middle line:84% So it's not that the voting has to be universal. 00:06:41.832 --> 00:06:43.540 align:middle line:84% It's really that has to be non-selective. 00:06:43.540 --> 00:06:44.220 align:middle line:90% That's the key. 00:06:44.220 --> 00:06:46.760 align:middle line:84% It has to be symmetric, the people who aren't voting. 00:06:46.760 --> 00:06:49.218 align:middle line:84% Once they're not, then you're going to not be at the middle 00:06:49.218 --> 00:06:50.360 align:middle line:90% anymore. 00:06:50.360 --> 00:06:52.540 align:middle line:84% Likewise, two candidates, I explained, 00:06:52.540 --> 00:06:53.780 align:middle line:90% the model breaks down. 00:06:53.780 --> 00:06:55.580 align:middle line:84% You just can't find an equilibrium 00:06:55.580 --> 00:06:58.180 align:middle line:84% on the line with three candidates. 00:06:58.180 --> 00:07:01.100 align:middle line:90% No ideology is very interesting. 00:07:01.100 --> 00:07:03.920 align:middle line:84% No ideology is important because remember, 00:07:03.920 --> 00:07:07.060 align:middle line:84% I'm assuming a normal symmetric distribution. 00:07:07.060 --> 00:07:10.605 align:middle line:84% But once there's ideology, then candidates 00:07:10.605 --> 00:07:11.980 align:middle line:84% may not want to be at the middle. 00:07:11.980 --> 00:07:16.835 align:middle line:84% In particular, if voting is selective 00:07:16.835 --> 00:07:19.460 align:middle line:84% and there's one other important element of the political system 00:07:19.460 --> 00:07:22.900 align:middle line:84% I haven't mentioned yet-- and what's that? 00:07:22.900 --> 00:07:25.500 align:middle line:84% What's a key aspect of the US political system in particular 00:07:25.500 --> 00:07:28.620 align:middle line:90% I haven't mentioned yet? 00:07:28.620 --> 00:07:31.460 align:middle line:84% Besides voting, what else matters in US politics? 00:07:31.460 --> 00:07:32.537 align:middle line:90% AUDIENCE: [INAUDIBLE] 00:07:32.537 --> 00:07:33.620 align:middle line:90% JONATHAN GRUBER: Well, no. 00:07:33.620 --> 00:07:36.620 align:middle line:84% Yes, they matter, but what else systematically matters in US 00:07:36.620 --> 00:07:40.580 align:middle line:90% politics besides people's votes? 00:07:40.580 --> 00:07:42.500 align:middle line:90% AUDIENCE: Lobbying? 00:07:42.500 --> 00:07:44.900 align:middle line:84% JONATHAN GRUBER: Lobbying is an aspect that matters. 00:07:44.900 --> 00:07:46.155 align:middle line:90% But more generally than that. 00:07:46.155 --> 00:07:48.780 align:middle line:84% What would some say is the most important factor in US politics 00:07:48.780 --> 00:07:50.450 align:middle line:90% today? 00:07:50.450 --> 00:07:51.377 align:middle line:90% Steven? 00:07:51.377 --> 00:07:52.210 align:middle line:90% AUDIENCE: Attention. 00:07:52.210 --> 00:07:52.870 align:middle line:90% JONATHAN GRUBER: Attention? 00:07:52.870 --> 00:07:54.250 align:middle line:90% And how do you get attention? 00:07:54.250 --> 00:07:54.980 align:middle line:90% AUDIENCE: Media. 00:07:54.980 --> 00:07:56.730 align:middle line:84% JONATHAN GRUBER: And how do you get media? 00:07:56.730 --> 00:07:57.690 align:middle line:90% Money. 00:07:57.690 --> 00:08:00.930 align:middle line:84% So I've ignored the role of money. 00:08:00.930 --> 00:08:03.798 align:middle line:84% I'm delighted that you all are such bright-eyed and optimistic 00:08:03.798 --> 00:08:06.090 align:middle line:84% and believers in democracy that you did not immediately 00:08:06.090 --> 00:08:06.910 align:middle line:90% come up with this answer. 00:08:06.910 --> 00:08:09.630 align:middle line:84% That's a good sign for your lack of skepticism about the world. 00:08:09.630 --> 00:08:11.190 align:middle line:90% Unfortunately, you're wrong. 00:08:11.190 --> 00:08:12.650 align:middle line:90% It is money. 00:08:12.650 --> 00:08:16.710 align:middle line:84% So now imagine a world where money matters. 00:08:16.710 --> 00:08:20.330 align:middle line:84% And money matters because money buys you attention. 00:08:20.330 --> 00:08:25.450 align:middle line:84% Imagine where money matters, voting is not universal, 00:08:25.450 --> 00:08:28.950 align:middle line:84% and ideology-- then ideology can matter. 00:08:28.950 --> 00:08:32.090 align:middle line:84% So for example, you could imagine a world 00:08:32.090 --> 00:08:35.169 align:middle line:84% where you say, look, I am willing to not 00:08:35.169 --> 00:08:36.429 align:middle line:90% be at the median defense. 00:08:36.429 --> 00:08:40.669 align:middle line:84% I will advocate for high defense spending because the people who 00:08:40.669 --> 00:08:43.892 align:middle line:84% like defense spending are very passionate about other things. 00:08:43.892 --> 00:08:45.850 align:middle line:84% And they'll turn out and vote and give me money 00:08:45.850 --> 00:08:48.560 align:middle line:84% if I signal high if I like defense spending. 00:08:48.560 --> 00:08:51.960 align:middle line:84% So basically, once there's-- just like intensity 00:08:51.960 --> 00:08:55.520 align:middle line:84% of preferences violates the efficiency of the Median Voter 00:08:55.520 --> 00:08:59.560 align:middle line:84% Theorem, intensity of preferences here can lead people 00:08:59.560 --> 00:09:02.960 align:middle line:84% to actually not want to settle at the middle if there's more 00:09:02.960 --> 00:09:03.820 align:middle line:90% intense preferences. 00:09:03.820 --> 00:09:05.860 align:middle line:84% Once again, if preferences are equally intense on both sides, 00:09:05.860 --> 00:09:06.902 align:middle line:90% it doesn't change things. 00:09:06.902 --> 00:09:11.140 align:middle line:84% It needs an asymmetry of preference intensity. 00:09:11.140 --> 00:09:13.520 align:middle line:84% But that's sort of related to the other thing, which 00:09:13.520 --> 00:09:15.940 align:middle line:90% is we said only two candidates. 00:09:15.940 --> 00:09:18.800 align:middle line:84% We also didn't talk about, we're assuming this model, 00:09:18.800 --> 00:09:22.548 align:middle line:90% single-dimensional voting. 00:09:22.548 --> 00:09:24.840 align:middle line:84% That's another problem with this model, which is people 00:09:24.840 --> 00:09:27.120 align:middle line:90% don't just vote over one thing. 00:09:27.120 --> 00:09:30.460 align:middle line:84% And so by not staking out the middle on one thing, 00:09:30.460 --> 00:09:32.800 align:middle line:84% you might signal your preference for something else 00:09:32.800 --> 00:09:34.600 align:middle line:90% that people care more about. 00:09:34.600 --> 00:09:39.040 align:middle line:84% So people who like big defense tend to be anti-abortion. 00:09:39.040 --> 00:09:43.658 align:middle line:84% So maybe by saying that-- taking a controversial position 00:09:43.658 --> 00:09:46.200 align:middle line:84% on abortion, maybe if there are other voters who don't really 00:09:46.200 --> 00:09:48.620 align:middle line:84% care about abortion, but really like defense, 00:09:48.620 --> 00:09:50.360 align:middle line:84% then you don't have to be in the middle of an abortion. 00:09:50.360 --> 00:09:51.680 align:middle line:84% That's a reason not to be in the middle, 00:09:51.680 --> 00:09:53.805 align:middle line:84% because you pull in people who really like defense. 00:09:53.805 --> 00:09:54.860 align:middle line:90% And they matter. 00:09:54.860 --> 00:09:57.900 align:middle line:84% So it gets much more complicated. 00:09:57.900 --> 00:10:00.140 align:middle line:84% And then finally, we have the last thing, which we, 00:10:00.140 --> 00:10:03.600 align:middle line:84% once again, cannot forget we're always assuming here, 00:10:03.600 --> 00:10:11.388 align:middle line:84% which is we're assuming full information and, slash, no-- 00:10:11.388 --> 00:10:13.180 align:middle line:84% I mean, I don't want to call them mistakes. 00:10:13.180 --> 00:10:16.620 align:middle line:90% But no bias, if you will. 00:10:16.620 --> 00:10:20.820 align:middle line:84% We're assuming that people truly understand what candidates mean 00:10:20.820 --> 00:10:23.620 align:middle line:84% when they say things and truly understand what 00:10:23.620 --> 00:10:25.700 align:middle line:90% the candidates' positions are. 00:10:25.700 --> 00:10:27.820 align:middle line:84% In a world where moving away from the middle 00:10:27.820 --> 00:10:30.140 align:middle line:84% allows you to obfuscate in certain ways, 00:10:30.140 --> 00:10:31.780 align:middle line:84% you may move away from the middle 00:10:31.780 --> 00:10:33.520 align:middle line:84% because that could be beneficial. 00:10:33.520 --> 00:10:34.500 align:middle line:90% Yeah, Alec? 00:10:34.500 --> 00:10:36.780 align:middle line:84% AUDIENCE: Does the full information assumption 00:10:36.780 --> 00:10:38.280 align:middle line:90% supersede the money assumption? 00:10:38.280 --> 00:10:41.000 align:middle line:84% Because the purpose of money is to provide information. 00:10:41.000 --> 00:10:41.440 align:middle line:90% JONATHAN GRUBER: Excellent. 00:10:41.440 --> 00:10:42.080 align:middle line:90% Excellent question. 00:10:42.080 --> 00:10:43.320 align:middle line:84% Actually, that leads to what I'm talking about next. 00:10:43.320 --> 00:10:44.200 align:middle line:90% So let me hold that question. 00:10:44.200 --> 00:10:45.400 align:middle line:90% Other questions about this? 00:10:45.400 --> 00:10:48.050 align:middle line:90% 00:10:48.050 --> 00:10:50.210 align:middle line:90% So that's a great-- 00:10:50.210 --> 00:10:52.850 align:middle line:84% so the bottom line here is, once again, 00:10:52.850 --> 00:10:54.950 align:middle line:84% as with all these generalized models 00:10:54.950 --> 00:11:00.570 align:middle line:84% we'll teach, from Coase to this model to the Median Voter Model, 00:11:00.570 --> 00:11:03.690 align:middle line:84% there's a lot of interesting aspects 00:11:03.690 --> 00:11:07.150 align:middle line:90% of this that speak to reality. 00:11:07.150 --> 00:11:12.110 align:middle line:84% But it clearly is not a pure description of reality. 00:11:12.110 --> 00:11:13.807 align:middle line:84% And basically, let's talk for a minute. 00:11:13.807 --> 00:11:15.890 align:middle line:84% I'll come back to your question in a minute, Alec. 00:11:15.890 --> 00:11:20.010 align:middle line:84% Let's talk about evidence, which is, does the Median Voter 00:11:20.010 --> 00:11:21.150 align:middle line:90% Model explain reality? 00:11:21.150 --> 00:11:25.050 align:middle line:84% And the answer is, to some extent, yes, to some extent, no. 00:11:25.050 --> 00:11:29.890 align:middle line:84% So how do you actually test the Median Voter Model? 00:11:29.890 --> 00:11:34.130 align:middle line:84% Well basically, you want to ask-- so the Median Voter 00:11:34.130 --> 00:11:36.810 align:middle line:84% Model says that if I'm in a district, 00:11:36.810 --> 00:11:39.450 align:middle line:84% if Paul and I are in two districts, 00:11:39.450 --> 00:11:40.970 align:middle line:84% and my district cares more about x, 00:11:40.970 --> 00:11:42.702 align:middle line:84% and his district cares more about y, 00:11:42.702 --> 00:11:44.910 align:middle line:84% I should vote more towards x, and he should vote more 00:11:44.910 --> 00:11:45.470 align:middle line:90% towards y. 00:11:45.470 --> 00:11:48.330 align:middle line:84% The problem with testing that is, well, 00:11:48.330 --> 00:11:50.990 align:middle line:84% how do you separate that from other things about our districts 00:11:50.990 --> 00:11:51.990 align:middle line:90% that might violate this? 00:11:51.990 --> 00:11:55.043 align:middle line:84% It's the classic bias problem we talked about in chapter 3. 00:11:55.043 --> 00:11:56.710 align:middle line:84% So what studies have done is essentially 00:11:56.710 --> 00:11:59.970 align:middle line:84% looked at random variation in the makeup of a district. 00:11:59.970 --> 00:12:01.830 align:middle line:84% For instance, there's redistricting, 00:12:01.830 --> 00:12:04.350 align:middle line:84% or if there's population shifts, then suddenly, who's 00:12:04.350 --> 00:12:05.910 align:middle line:90% in your district changes. 00:12:05.910 --> 00:12:08.270 align:middle line:84% And what they found is when people 00:12:08.270 --> 00:12:11.890 align:middle line:84% and when the composition of a district changes, politicians, 00:12:11.890 --> 00:12:13.110 align:middle line:90% voters change with it. 00:12:13.110 --> 00:12:14.110 align:middle line:90% Votes change with it. 00:12:14.110 --> 00:12:16.030 align:middle line:84% Let's look at what politicians do. 00:12:16.030 --> 00:12:17.830 align:middle line:84% What they do and their votes in Congress 00:12:17.830 --> 00:12:20.910 align:middle line:84% actually do respond to changing the composition of the people 00:12:20.910 --> 00:12:22.330 align:middle line:90% they're representing. 00:12:22.330 --> 00:12:26.265 align:middle line:84% Basically, you have-- and moreover, 00:12:26.265 --> 00:12:27.890 align:middle line:84% another piece of confirmatory evidence, 00:12:27.890 --> 00:12:30.470 align:middle line:84% if you look at politicians in their last term 00:12:30.470 --> 00:12:32.110 align:middle line:84% when they're going to quit, they become 00:12:32.110 --> 00:12:35.930 align:middle line:84% less responsive to the composition of their district, 00:12:35.930 --> 00:12:38.590 align:middle line:84% which is, once again, consistent with a Median Voter Vote 00:12:38.590 --> 00:12:40.973 align:middle line:90% Maximization model. 00:12:40.973 --> 00:12:43.140 align:middle line:84% On the other hand, there's clear evidence cleverness 00:12:43.140 --> 00:12:45.180 align:middle line:84% the Vote Maximization Model isn't right. 00:12:45.180 --> 00:12:48.820 align:middle line:84% I mean, the most transparent evidence is US senators. 00:12:48.820 --> 00:12:50.700 align:middle line:84% Two US senators from the same state 00:12:50.700 --> 00:12:53.300 align:middle line:90% have the same median voter. 00:12:53.300 --> 00:12:55.000 align:middle line:84% And yet, in a large number of states, 00:12:55.000 --> 00:12:57.380 align:middle line:84% the two senators are often of different parties who 00:12:57.380 --> 00:12:59.900 align:middle line:90% do not agree on every issue. 00:12:59.900 --> 00:13:02.240 align:middle line:84% So clearly, the Median Voter can't explain Everything 00:13:02.240 --> 00:13:03.260 align:middle line:84% when you got two people representing 00:13:03.260 --> 00:13:05.620 align:middle line:84% the exact same population taking sometimes diametrically 00:13:05.620 --> 00:13:06.640 align:middle line:90% opposed positions. 00:13:06.640 --> 00:13:07.528 align:middle line:90% Yes, Steven? 00:13:07.528 --> 00:13:09.820 align:middle line:84% AUDIENCE: [INAUDIBLE] still hold because it's happening 00:13:09.820 --> 00:13:11.172 align:middle line:90% at different points in time? 00:13:11.172 --> 00:13:13.380 align:middle line:84% JONATHAN GRUBER: Yes, that's a good point, if they're 00:13:13.380 --> 00:13:14.360 align:middle line:90% elected at points in time. 00:13:14.360 --> 00:13:16.985 align:middle line:84% But you'd need a pretty radical swing in preferences over, say, 00:13:16.985 --> 00:13:18.127 align:middle line:90% two years, to go from a-- 00:13:18.127 --> 00:13:19.960 align:middle line:84% I mean, if they're right towards the middle, 00:13:19.960 --> 00:13:20.640 align:middle line:90% you can imagine that. 00:13:20.640 --> 00:13:21.600 align:middle line:90% But that's a good point. 00:13:21.600 --> 00:13:23.433 align:middle line:84% But even sometimes elected at the same time, 00:13:23.433 --> 00:13:25.460 align:middle line:90% you get two different parties. 00:13:25.460 --> 00:13:26.440 align:middle line:90% Yeah? 00:13:26.440 --> 00:13:29.620 align:middle line:84% Well, because you have special elections and things. 00:13:29.620 --> 00:13:32.400 align:middle line:84% So basically, that's a rejection. 00:13:32.400 --> 00:13:34.045 align:middle line:84% A more clever rejection actually comes 00:13:34.045 --> 00:13:35.420 align:middle line:84% from researcher, a former student 00:13:35.420 --> 00:13:37.620 align:middle line:84% of mine, Ebonya Washington, did, which 00:13:37.620 --> 00:13:43.770 align:middle line:84% is, she looked at how politicians voted on issues 00:13:43.770 --> 00:13:46.650 align:middle line:84% affecting women's rights based on the gender 00:13:46.650 --> 00:13:49.050 align:middle line:90% composition of their children. 00:13:49.050 --> 00:13:51.830 align:middle line:84% So conditioned on having a certain number of children, 00:13:51.830 --> 00:13:54.010 align:middle line:84% the gender composition is random. 00:13:54.010 --> 00:13:57.810 align:middle line:84% So two politicians, both of which have two kids-- 00:13:57.810 --> 00:14:00.130 align:middle line:84% one has two boys, one has a boy and a girl-- 00:14:00.130 --> 00:14:03.090 align:middle line:90% are functionally identical. 00:14:03.090 --> 00:14:06.850 align:middle line:84% Yet she finds the ones that has the boy and the girl votes 00:14:06.850 --> 00:14:08.850 align:middle line:84% much more positively on women's issues 00:14:08.850 --> 00:14:10.630 align:middle line:90% than the one that has two sons. 00:14:10.630 --> 00:14:12.630 align:middle line:84% That cannot be explained with the Median Voters. 00:14:12.630 --> 00:14:15.850 align:middle line:84% That's got to be ideology, underlying ideology. 00:14:15.850 --> 00:14:20.930 align:middle line:84% So clearly, the Median Voter Model matters. 00:14:20.930 --> 00:14:23.690 align:middle line:84% People respond to the composition 00:14:23.690 --> 00:14:25.010 align:middle line:90% of who their voters are. 00:14:25.010 --> 00:14:29.090 align:middle line:90% But it's also not determining. 00:14:29.090 --> 00:14:31.690 align:middle line:84% I think the most interesting evidence of the Median Voter 00:14:31.690 --> 00:14:34.670 align:middle line:84% Model actually comes from this thing in 2007. 00:14:34.670 --> 00:14:37.290 align:middle line:84% So basically, there's something we do in the US government, 00:14:37.290 --> 00:14:38.123 align:middle line:90% and all governments. 00:14:38.123 --> 00:14:39.050 align:middle line:90% It's called earmarks. 00:14:39.050 --> 00:14:42.550 align:middle line:84% An earmark is when they put together this giant $3 trillion 00:14:42.550 --> 00:14:44.350 align:middle line:90% budget, there'll be a line. 00:14:44.350 --> 00:14:47.510 align:middle line:84% Line 974 will say, including $3 million 00:14:47.510 --> 00:14:49.610 align:middle line:84% for this baby pet project of this politician. 00:14:49.610 --> 00:14:51.130 align:middle line:84% And that's how these deals get made. 00:14:51.130 --> 00:14:52.510 align:middle line:84% There's a bunch of little pet projects 00:14:52.510 --> 00:14:53.927 align:middle line:84% that get rolled in to get people's 00:14:53.927 --> 00:14:55.110 align:middle line:90% votes on the big budget. 00:14:55.110 --> 00:14:57.630 align:middle line:90% Those are called earmarks. 00:14:57.630 --> 00:14:59.330 align:middle line:84% Those are really buried in these bills. 00:14:59.330 --> 00:15:02.772 align:middle line:84% In 2007, the House of Representatives 00:15:02.772 --> 00:15:05.230 align:middle line:84% added a rule to try to fight-- these earmarks are generally 00:15:05.230 --> 00:15:07.510 align:middle line:84% viewed as often very wasteful, things done 00:15:07.510 --> 00:15:10.070 align:middle line:84% to please constituents, and not really what 00:15:10.070 --> 00:15:12.830 align:middle line:84% should be in the interest of the overall US public. 00:15:12.830 --> 00:15:14.830 align:middle line:84% So in 2007, the House of Representatives 00:15:14.830 --> 00:15:16.990 align:middle line:84% put in a rule which said, we've got to make 00:15:16.990 --> 00:15:18.630 align:middle line:90% all these earmarks public. 00:15:18.630 --> 00:15:21.750 align:middle line:84% And the idea was they hoped it would shame politicians 00:15:21.750 --> 00:15:23.990 align:middle line:90% into not creating earmarks. 00:15:23.990 --> 00:15:27.010 align:middle line:84% It had the opposite effect, because they were like, great. 00:15:27.010 --> 00:15:29.458 align:middle line:84% I can brag to my constituents now about all the things 00:15:29.458 --> 00:15:30.250 align:middle line:90% I'm doing for them. 00:15:30.250 --> 00:15:32.655 align:middle line:84% It actually increased earmarks because politicians 00:15:32.655 --> 00:15:34.030 align:middle line:84% didn't want to hide the fact they 00:15:34.030 --> 00:15:35.250 align:middle line:84% were doing good things for their constituents. 00:15:35.250 --> 00:15:36.590 align:middle line:90% They wanted to advertise them. 00:15:36.590 --> 00:15:38.215 align:middle line:84% So it actually had the opposite effect, 00:15:38.215 --> 00:15:40.860 align:middle line:84% which is, once again, consistent with a Median Voter 00:15:40.860 --> 00:15:43.780 align:middle line:90% kind of model. 00:15:43.780 --> 00:15:45.920 align:middle line:84% As one representative from Kansas said, 00:15:45.920 --> 00:15:47.400 align:middle line:90% "Democracy is a contact sport. 00:15:47.400 --> 00:15:49.317 align:middle line:84% I'm not going to be shy about asking for money 00:15:49.317 --> 00:15:51.460 align:middle line:90% for my community." 00:15:51.460 --> 00:15:55.400 align:middle line:84% So that's where we are with the Median Voter Model. 00:15:55.400 --> 00:15:59.860 align:middle line:84% But I want to come back to a question about that. 00:15:59.860 --> 00:16:02.340 align:middle line:84% I want to come back to Alec's point, which is 00:16:02.340 --> 00:16:04.260 align:middle line:90% let's talk about the L word-- 00:16:04.260 --> 00:16:05.240 align:middle line:90% lobbying. 00:16:05.240 --> 00:16:05.740 align:middle line:90% OK. 00:16:05.740 --> 00:16:09.840 align:middle line:84% This is sort of a swear word in common political discourse. 00:16:09.840 --> 00:16:11.745 align:middle line:84% Those lobbyists, they're so terrible. 00:16:11.745 --> 00:16:13.120 align:middle line:84% Lobbyists are ruining Washington. 00:16:13.120 --> 00:16:14.980 align:middle line:84% First of all, where does lobbying come from? 00:16:14.980 --> 00:16:16.522 align:middle line:84% The term lobbying comes from the fact 00:16:16.522 --> 00:16:22.980 align:middle line:84% that President Grant, when he was president, 00:16:22.980 --> 00:16:26.113 align:middle line:84% would, every afternoon, get drunk and sit 00:16:26.113 --> 00:16:27.780 align:middle line:84% in the lobby of the Willard Hotel, which 00:16:27.780 --> 00:16:29.120 align:middle line:84% is across the street from the White House. 00:16:29.120 --> 00:16:30.120 align:middle line:90% And he'd go ask favors of him. 00:16:30.120 --> 00:16:31.703 align:middle line:84% And he'd say yes because he was drunk. 00:16:31.703 --> 00:16:35.362 align:middle line:84% So that's where the term lobbying comes from. 00:16:35.362 --> 00:16:37.820 align:middle line:84% I don't want to impugn the guy-- maybe not every afternoon, 00:16:37.820 --> 00:16:41.300 align:middle line:90% but on many afternoons. 00:16:41.300 --> 00:16:44.480 align:middle line:84% And that history maybe leads to a negative taste in people's 00:16:44.480 --> 00:16:45.500 align:middle line:90% mouths with lobbying. 00:16:45.500 --> 00:16:47.820 align:middle line:84% But in fact, as Alex's question pointed out, 00:16:47.820 --> 00:16:49.680 align:middle line:90% lobbying is not obviously bad. 00:16:49.680 --> 00:16:52.000 align:middle line:84% It's not obviously bad for two reasons. 00:16:52.000 --> 00:16:52.900 align:middle line:90% Why is that? 00:16:52.900 --> 00:16:54.380 align:middle line:84% Why might lobbying be productive? 00:16:54.380 --> 00:16:59.840 align:middle line:84% In fact, why might an optimal government have some lobbying? 00:16:59.840 --> 00:17:00.653 align:middle line:90% Yeah? 00:17:00.653 --> 00:17:02.320 align:middle line:84% AUDIENCE: It helps with the information. 00:17:02.320 --> 00:17:03.580 align:middle line:84% JONATHAN GRUBER: It helps with the information. 00:17:03.580 --> 00:17:05.160 align:middle line:90% So I worked in the government. 00:17:05.160 --> 00:17:07.560 align:middle line:84% We had to deal with an enormous number of issues 00:17:07.560 --> 00:17:09.982 align:middle line:84% all the time on which we were not expert. 00:17:09.982 --> 00:17:11.940 align:middle line:84% Lobbyists would come in, and they'd be experts. 00:17:11.940 --> 00:17:13.880 align:middle line:84% Now, they'd experts with a point of view. 00:17:13.880 --> 00:17:16.005 align:middle line:84% But as long as there are lobbyists from both sides, 00:17:16.005 --> 00:17:17.000 align:middle line:90% I got educated. 00:17:17.000 --> 00:17:20.887 align:middle line:84% So, actually, they help fight the full information problem. 00:17:20.887 --> 00:17:22.220 align:middle line:90% That's one thing they help with. 00:17:22.220 --> 00:17:23.480 align:middle line:90% What else do they help with? 00:17:23.480 --> 00:17:30.432 align:middle line:84% How else could lobbyists help to deliver the efficient outcome? 00:17:30.432 --> 00:17:31.503 align:middle line:90% Yeah, [? Alec? ?] 00:17:31.503 --> 00:17:32.920 align:middle line:84% AUDIENCE: They provide information 00:17:32.920 --> 00:17:33.980 align:middle line:90% about the preferences. 00:17:33.980 --> 00:17:34.960 align:middle line:84% JONATHAN GRUBER: Yeah, they can represent 00:17:34.960 --> 00:17:36.510 align:middle line:90% intensity of preferences. 00:17:36.510 --> 00:17:41.095 align:middle line:84% Remember, the median voter model fails when the median voter 00:17:41.095 --> 00:17:42.470 align:middle line:84% does not represent the mean voter 00:17:42.470 --> 00:17:44.950 align:middle line:84% because there's an intensity of preferences we're missing. 00:17:44.950 --> 00:17:48.070 align:middle line:84% Well, lobbyists represent that intensity of preferences. 00:17:48.070 --> 00:17:51.310 align:middle line:84% The problem, of course, is when they go too far. 00:17:51.310 --> 00:17:55.110 align:middle line:84% The problem, of course, is that basically-- 00:17:55.110 --> 00:17:56.530 align:middle line:84% so I have an example in the book. 00:17:56.530 --> 00:17:57.370 align:middle line:90% You don't have to write it down. 00:17:57.370 --> 00:17:58.710 align:middle line:90% It's just a simple example. 00:17:58.710 --> 00:18:02.750 align:middle line:84% Imagine there's a policy that will benefit a hundred 00:18:02.750 --> 00:18:05.950 align:middle line:84% people in the US by $1 million each, 00:18:05.950 --> 00:18:09.350 align:middle line:84% but the remaining citizens in the US will lose $100 each. 00:18:09.350 --> 00:18:11.070 align:middle line:90% Clearly, that's a loser. 00:18:11.070 --> 00:18:13.870 align:middle line:84% Clearly, we are worse off as a society. 00:18:13.870 --> 00:18:17.162 align:middle line:84% But that project may still pass, often through an earmark, 00:18:17.162 --> 00:18:19.870 align:middle line:84% because basically, the intensity of preference of those 10 voters 00:18:19.870 --> 00:18:21.870 align:middle line:84% is not offset by anyone representing 00:18:21.870 --> 00:18:24.550 align:middle line:84% the nonintense preference of every other voter. 00:18:24.550 --> 00:18:28.630 align:middle line:84% So once again, you can see the balancing act here. 00:18:28.630 --> 00:18:31.490 align:middle line:84% The balancing act is, on the one hand, 00:18:31.490 --> 00:18:35.860 align:middle line:84% if you don't at all represent preference intensity 00:18:35.860 --> 00:18:39.940 align:middle line:84% and there is ideological differences, 00:18:39.940 --> 00:18:43.340 align:middle line:84% then you do not get the optimal outcome with the median voter. 00:18:43.340 --> 00:18:47.020 align:middle line:84% On the other hand, lobbying probably overcorrects for that 00:18:47.020 --> 00:18:50.140 align:middle line:84% by overrepresenting the preferences of small groups 00:18:50.140 --> 00:18:52.620 align:middle line:90% versus large groups. 00:18:52.620 --> 00:18:55.820 align:middle line:84% And that's the tension with lobbying. 00:18:55.820 --> 00:18:57.520 align:middle line:84% Now, there's no better example of that-- 00:18:57.520 --> 00:18:58.460 align:middle line:84% and I have several pages on this. 00:18:58.460 --> 00:18:59.860 align:middle line:84% I'll just go through it quickly-- 00:18:59.860 --> 00:19:02.100 align:middle line:90% than farm policy in the US. 00:19:02.100 --> 00:19:07.420 align:middle line:84% In 1900, 35% of all workers in the US worked on farms. 00:19:07.420 --> 00:19:09.940 align:middle line:90% Today, it's 1%. 00:19:09.940 --> 00:19:14.980 align:middle line:84% And yet, we spend $22.4 billion as a federal government 00:19:14.980 --> 00:19:17.860 align:middle line:84% every year sending money to farms. 00:19:17.860 --> 00:19:22.780 align:middle line:90% Now, think of it this way. 00:19:22.780 --> 00:19:25.920 align:middle line:84% And so basically, each US household, 00:19:25.920 --> 00:19:30.420 align:middle line:84% every household in the US, spends $144 every year 00:19:30.420 --> 00:19:32.463 align:middle line:90% supporting US farmers. 00:19:32.463 --> 00:19:33.880 align:middle line:84% Now, you might think, gee, I don't 00:19:33.880 --> 00:19:35.820 align:middle line:84% want to spend $144 supporting US farmers. 00:19:35.820 --> 00:19:36.700 align:middle line:90% Why does that happen? 00:19:36.700 --> 00:19:40.260 align:middle line:84% Well, that's because each farmer in the US gets, on average, 00:19:40.260 --> 00:19:43.640 align:middle line:90% $15,000 in supports. 00:19:43.640 --> 00:19:45.140 align:middle line:90% So you have your $144. 00:19:45.140 --> 00:19:47.015 align:middle line:84% You don't get together and lobby because it's 00:19:47.015 --> 00:19:48.480 align:middle line:90% a classic free-rider problem. 00:19:48.480 --> 00:19:51.760 align:middle line:84% Who's going to put all the work in to lobby for $144? 00:19:51.760 --> 00:19:55.268 align:middle line:84% Lots of people put the work in to lobby to get $15,000. 00:19:55.268 --> 00:19:57.560 align:middle line:84% So the problem with lobbying is, the free-rider problem 00:19:57.560 --> 00:19:59.560 align:middle line:84% is going to lead to an overweight 00:19:59.560 --> 00:20:02.000 align:middle line:84% on issues that are intensely important to a small number 00:20:02.000 --> 00:20:03.640 align:middle line:84% of people versus issues that aren't 00:20:03.640 --> 00:20:05.880 align:middle line:84% that important to a large number of people. 00:20:05.880 --> 00:20:07.780 align:middle line:84% That's the classic flaw in lobbying. 00:20:07.780 --> 00:20:10.520 align:middle line:90% 00:20:10.520 --> 00:20:14.340 align:middle line:84% Now, if you ask farm folk, they'll say, 00:20:14.340 --> 00:20:17.280 align:middle line:84% well, you're just biased against the family farm, 00:20:17.280 --> 00:20:19.580 align:middle line:84% and family farming is a way of life in America. 00:20:19.580 --> 00:20:20.380 align:middle line:90% We need it. 00:20:20.380 --> 00:20:25.760 align:middle line:84% Well, they'd be wrong because almost all the subsidies 00:20:25.760 --> 00:20:29.840 align:middle line:84% to farms go to large corporate farms, not small family farms. 00:20:29.840 --> 00:20:33.310 align:middle line:84% The top 10% of farms in the US, all of which are owned by big 00:20:33.310 --> 00:20:37.750 align:middle line:84% corporations, received on average $560,000 per year, 00:20:37.750 --> 00:20:40.410 align:middle line:84% while the bottom 80%, which include family farms, 00:20:40.410 --> 00:20:42.310 align:middle line:84% received on average $8,000 a year. 00:20:42.310 --> 00:20:45.650 align:middle line:84% So this is not preserving some way of life of the family farm. 00:20:45.650 --> 00:20:49.470 align:middle line:84% It's large subsidies to large corporations that farm. 00:20:49.470 --> 00:20:53.530 align:middle line:84% And basically, this is not just a US phenomenon. 00:20:53.530 --> 00:20:55.030 align:middle line:84% A lot of things we'll talk about are 00:20:55.030 --> 00:20:56.293 align:middle line:90% how messed up US politics are. 00:20:56.293 --> 00:20:57.710 align:middle line:84% This has got the benefit that it's 00:20:57.710 --> 00:20:59.870 align:middle line:90% messed up all over the world. 00:20:59.870 --> 00:21:03.110 align:middle line:84% For example, the European Union spends about $113 00:21:03.110 --> 00:21:06.670 align:middle line:90% annually supporting its farmers. 00:21:06.670 --> 00:21:10.430 align:middle line:84% The average European cow, the average cow in Europe 00:21:10.430 --> 00:21:13.810 align:middle line:84% gets $2.33 in government subsidy every day. 00:21:13.810 --> 00:21:16.830 align:middle line:90% 00:21:16.830 --> 00:21:19.970 align:middle line:84% Japan spends more than $47 billion on its farmers, 00:21:19.970 --> 00:21:22.950 align:middle line:90% with rice tariffs near 800%. 00:21:22.950 --> 00:21:25.190 align:middle line:84% And a lot of times, these farm subsidies 00:21:25.190 --> 00:21:26.295 align:middle line:90% are just pure corruption. 00:21:26.295 --> 00:21:27.670 align:middle line:84% And we'll come back to corruption 00:21:27.670 --> 00:21:29.010 align:middle line:84% to talk about later in this lecture. 00:21:29.010 --> 00:21:30.665 align:middle line:84% They're not actually even going directly to people 00:21:30.665 --> 00:21:31.980 align:middle line:90% they're supposed to benefit. 00:21:31.980 --> 00:21:35.500 align:middle line:84% So basically, farm policy is a classic example 00:21:35.500 --> 00:21:39.300 align:middle line:84% of how lobbying can work in theory 00:21:39.300 --> 00:21:42.300 align:middle line:84% but in practice will not work when 00:21:42.300 --> 00:21:47.220 align:middle line:84% you have small groups with large interests versus large groups 00:21:47.220 --> 00:21:48.280 align:middle line:90% with small interests. 00:21:48.280 --> 00:21:48.780 align:middle line:90% Yeah? 00:21:48.780 --> 00:21:50.697 align:middle line:84% AUDIENCE: Are those subsidies like [INAUDIBLE] 00:21:50.697 --> 00:21:52.220 align:middle line:90% buy a lot of land [INAUDIBLE]? 00:21:52.220 --> 00:21:53.520 align:middle line:84% JONATHAN GRUBER: It's a very complicated system. 00:21:53.520 --> 00:21:54.770 align:middle line:90% Some of it's a direct subsidy. 00:21:54.770 --> 00:21:57.020 align:middle line:90% Some's loans, subsidized loans. 00:21:57.020 --> 00:21:58.960 align:middle line:90% It varies by crop. 00:21:58.960 --> 00:22:00.977 align:middle line:90% It varies by type of crop. 00:22:00.977 --> 00:22:02.560 align:middle line:84% There's winter wheat and spring wheat. 00:22:02.560 --> 00:22:04.602 align:middle line:84% And winter we consider subsidies of spring wheat. 00:22:04.602 --> 00:22:06.120 align:middle line:84% It varies by all sorts of things. 00:22:06.120 --> 00:22:07.800 align:middle line:84% So there's no simple explanation. 00:22:07.800 --> 00:22:12.460 align:middle line:84% I describe more in the book, but there's no simple explanation. 00:22:12.460 --> 00:22:13.740 align:middle line:90% All right. 00:22:13.740 --> 00:22:18.540 align:middle line:84% So we've done this whole study of, OK, here's 00:22:18.540 --> 00:22:20.580 align:middle line:84% how representative democracy works. 00:22:20.580 --> 00:22:25.260 align:middle line:84% And that's where my lecture ended until about 2016, 00:22:25.260 --> 00:22:28.803 align:middle line:84% when American politics broke, and where we cannot discuss 00:22:28.803 --> 00:22:30.720 align:middle line:84% American politics without discussing the issue 00:22:30.720 --> 00:22:32.360 align:middle line:90% of polarization. 00:22:32.360 --> 00:22:35.680 align:middle line:84% So we have become an unbelievably polarized nation. 00:22:35.680 --> 00:22:40.680 align:middle line:84% And there's no better way to see that than to look at figure 9-4. 00:22:40.680 --> 00:22:43.160 align:middle line:90% So what does figure 9-4 show? 00:22:43.160 --> 00:22:48.280 align:middle line:84% There's an organization which measures how conservative 00:22:48.280 --> 00:22:50.535 align:middle line:84% and how liberal each politician's votes are. 00:22:50.535 --> 00:22:52.660 align:middle line:84% So on most issues, there's a conservative position, 00:22:52.660 --> 00:22:54.000 align:middle line:90% a liberal position. 00:22:54.000 --> 00:22:56.200 align:middle line:84% And they give a score based on essentially 00:22:56.200 --> 00:23:00.375 align:middle line:84% how conservative or liberal the votes are of each politician. 00:23:00.375 --> 00:23:03.000 align:middle line:84% We have two graphs here, one for the Senate, one for the House. 00:23:03.000 --> 00:23:05.220 align:middle line:84% Let's look at the bottom line in the Senate 00:23:05.220 --> 00:23:07.240 align:middle line:90% graph, the line for Democrats. 00:23:07.240 --> 00:23:13.160 align:middle line:84% What this says is, in 1879, the average Democratic senator 00:23:13.160 --> 00:23:19.600 align:middle line:84% was about 45% more liberal than a neutral set of votes, 00:23:19.600 --> 00:23:23.160 align:middle line:84% than if they'd randomly voted up or down on everything. 00:23:23.160 --> 00:23:26.660 align:middle line:84% They then became more centrist, less liberal-- 00:23:26.660 --> 00:23:28.530 align:middle line:84% so as this goes towards the median line, 00:23:28.530 --> 00:23:32.350 align:middle line:84% that's less polarization, more centrism-- 00:23:32.350 --> 00:23:36.990 align:middle line:84% by about the middle of the 20th century and then became more 00:23:36.990 --> 00:23:40.510 align:middle line:84% polarized again, more partisan again. 00:23:40.510 --> 00:23:44.430 align:middle line:84% Same with Democrats in the House-- similar pattern. 00:23:44.430 --> 00:23:48.390 align:middle line:84% Republicans have a similar time pattern but much more extreme. 00:23:48.390 --> 00:23:49.910 align:middle line:84% So what we see is that polarization 00:23:49.910 --> 00:23:51.750 align:middle line:90% has gone up enormously. 00:23:51.750 --> 00:23:54.510 align:middle line:84% So the two lines are as far apart today 00:23:54.510 --> 00:23:56.760 align:middle line:90% as they were in 1879. 00:23:56.760 --> 00:23:58.510 align:middle line:84% The difference is that the Democratic line 00:23:58.510 --> 00:24:00.470 align:middle line:84% hasn't changed a lot, and the Republican line 00:24:00.470 --> 00:24:01.870 align:middle line:90% has gone up a lot. 00:24:01.870 --> 00:24:03.710 align:middle line:84% So essentially, we have polarization 00:24:03.710 --> 00:24:07.470 align:middle line:84% that's driven by both sides but primarily by one side. 00:24:07.470 --> 00:24:11.570 align:middle line:84% But the bottom line is, we have polarization. 00:24:11.570 --> 00:24:14.390 align:middle line:84% This is not a US-specific phenomenon. 00:24:14.390 --> 00:24:16.970 align:middle line:84% Polarization is happening all over the world. 00:24:16.970 --> 00:24:23.830 align:middle line:90% 00:24:23.830 --> 00:24:26.410 align:middle line:84% So all over the world, we see this separation. 00:24:26.410 --> 00:24:28.590 align:middle line:84% But the US is unique in how extreme it's been. 00:24:28.590 --> 00:24:30.262 align:middle line:90% Yeah? 00:24:30.262 --> 00:24:31.970 align:middle line:84% AUDIENCE: For data like this, how do they 00:24:31.970 --> 00:24:34.453 align:middle line:84% keep a standard for what is neutral when the [? line's ?] on 00:24:34.453 --> 00:24:35.286 align:middle line:90% average [INAUDIBLE]? 00:24:35.286 --> 00:24:38.930 align:middle line:84% JONATHAN GRUBER: There's judgment involved, obviously. 00:24:38.930 --> 00:24:40.650 align:middle line:84% Obviously, on every bill, there's 00:24:40.650 --> 00:24:42.970 align:middle line:84% a panel of people who work for this organization who 00:24:42.970 --> 00:24:45.190 align:middle line:90% evaluate, what vote's liberal? 00:24:45.190 --> 00:24:46.890 align:middle line:90% What vote's conservative? 00:24:46.890 --> 00:24:48.730 align:middle line:90% They're nonpartisan experts. 00:24:48.730 --> 00:24:50.230 align:middle line:90% They may get it wrong sometimes. 00:24:50.230 --> 00:24:54.610 align:middle line:84% But the average should be pretty good. 00:24:54.610 --> 00:24:58.550 align:middle line:84% So basically, the whole world is polarized, 00:24:58.550 --> 00:25:01.670 align:middle line:84% but the US actually seems to lead the world in polarization. 00:25:01.670 --> 00:25:05.690 align:middle line:84% We're number one in polarization. 00:25:05.690 --> 00:25:07.570 align:middle line:90% And you see this in-- 00:25:07.570 --> 00:25:09.650 align:middle line:84% COVID is the best example, where when 00:25:09.650 --> 00:25:13.210 align:middle line:84% you ask people of different party, 37% of Republicans 00:25:13.210 --> 00:25:16.370 align:middle line:84% thought COVID was a big problem and 73% of Democrats did. 00:25:16.370 --> 00:25:18.768 align:middle line:84% Now, COVID, it's a scientific thing 00:25:18.768 --> 00:25:20.310 align:middle line:84% about whether COVID is a big problem. 00:25:20.310 --> 00:25:21.810 align:middle line:84% You could imagine a little variation 00:25:21.810 --> 00:25:23.250 align:middle line:90% around the word big problem. 00:25:23.250 --> 00:25:25.960 align:middle line:84% But the bottom line is, there's just enormous disagreement 00:25:25.960 --> 00:25:27.600 align:middle line:90% over basic facts. 00:25:27.600 --> 00:25:28.858 align:middle line:90% And we all see this every day. 00:25:28.858 --> 00:25:29.900 align:middle line:90% We all know this is true. 00:25:29.900 --> 00:25:30.960 align:middle line:84% I don't think it's controversial to say 00:25:30.960 --> 00:25:33.200 align:middle line:84% we've become an incredibly polarized country. 00:25:33.200 --> 00:25:35.520 align:middle line:84% The question is, what has changed? 00:25:35.520 --> 00:25:36.800 align:middle line:90% What's going on? 00:25:36.800 --> 00:25:39.040 align:middle line:90% What's driven this polarization? 00:25:39.040 --> 00:25:42.000 align:middle line:90% So you tell me. 00:25:42.000 --> 00:25:42.792 align:middle line:90% What's caused this? 00:25:42.792 --> 00:25:43.292 align:middle line:90% Yeah? 00:25:43.292 --> 00:25:44.200 align:middle line:90% AUDIENCE: Technology. 00:25:44.200 --> 00:25:45.617 align:middle line:84% JONATHAN GRUBER: What do you mean? 00:25:45.617 --> 00:25:46.800 align:middle line:90% AUDIENCE: Like the internet. 00:25:46.800 --> 00:25:50.720 align:middle line:84% JONATHAN GRUBER: Well, how does the internet cause polarization? 00:25:50.720 --> 00:25:51.845 align:middle line:90% AUDIENCE: More information. 00:25:51.845 --> 00:25:53.428 align:middle line:84% JONATHAN GRUBER: OK, more information. 00:25:53.428 --> 00:25:55.640 align:middle line:84% But in fact, if anything, we'd think more information 00:25:55.640 --> 00:25:56.900 align:middle line:90% will cause less polarization. 00:25:56.900 --> 00:26:00.340 align:middle line:84% More information would be now, if you have biases, 00:26:00.340 --> 00:26:02.320 align:middle line:84% you can defeat them with information. 00:26:02.320 --> 00:26:03.260 align:middle line:90% So let's stay on this. 00:26:03.260 --> 00:26:05.186 align:middle line:90% So flesh out your answer. 00:26:05.186 --> 00:26:07.186 align:middle line:84% It's not just more from-- what do you think when 00:26:07.186 --> 00:26:08.500 align:middle line:90% you speak about the internet? 00:26:08.500 --> 00:26:09.500 align:middle line:84% You're not thinking about all websites. 00:26:09.500 --> 00:26:11.625 align:middle line:84% What websites are you thinking about in particular? 00:26:11.625 --> 00:26:15.160 align:middle line:90% 00:26:15.160 --> 00:26:16.690 align:middle line:90% What about the internet? 00:26:16.690 --> 00:26:17.680 align:middle line:90% Yeah? 00:26:17.680 --> 00:26:20.420 align:middle line:84% AUDIENCE: I think, specifically, misinformation on social media. 00:26:20.420 --> 00:26:21.337 align:middle line:90% JONATHAN GRUBER: Yeah. 00:26:21.337 --> 00:26:24.110 align:middle line:84% So look, let's be honest, the internet's great. 00:26:24.110 --> 00:26:26.510 align:middle line:90% Social media, not so much. 00:26:26.510 --> 00:26:31.390 align:middle line:84% Basically, what's gone on is, essentially, 00:26:31.390 --> 00:26:34.610 align:middle line:84% individuals have been able to create their own-- 00:26:34.610 --> 00:26:38.990 align:middle line:84% what we call echo chambers and their own fact bases based 00:26:38.990 --> 00:26:39.790 align:middle line:90% on the-- 00:26:39.790 --> 00:26:41.750 align:middle line:84% it wouldn't be possible without the internet. 00:26:41.750 --> 00:26:45.710 align:middle line:84% But it's not espn.com that's causing this, OK? 00:26:45.710 --> 00:26:49.410 align:middle line:84% It's social media sites, other news sites. 00:26:49.410 --> 00:26:53.990 align:middle line:84% And that essentially, there's this 00:26:53.990 --> 00:26:56.130 align:middle line:84% been this proliferation of media, 00:26:56.130 --> 00:26:59.590 align:middle line:90% which has caused polarization. 00:26:59.590 --> 00:27:04.010 align:middle line:84% So basically, if you go back to history, after World War II, 00:27:04.010 --> 00:27:05.510 align:middle line:84% most Americans got all their news 00:27:05.510 --> 00:27:07.550 align:middle line:84% from one of two or three sources. 00:27:07.550 --> 00:27:10.710 align:middle line:84% And then, as one author put it, "All that changed with 00:27:10.710 --> 00:27:13.350 align:middle line:84% the advent of cable television in the 1980s and the internet 00:27:13.350 --> 00:27:14.790 align:middle line:90% in the 1990s. 00:27:14.790 --> 00:27:18.150 align:middle line:84% Now Americans can choose from hundreds of partisan news 00:27:18.150 --> 00:27:21.110 align:middle line:84% sources, many of which care more about arousing emotions 00:27:21.110 --> 00:27:23.370 align:middle line:84% than hewing to journalistic standards. 00:27:23.370 --> 00:27:25.610 align:middle line:84% This proliferation of sources interacts with the most 00:27:25.610 --> 00:27:27.530 align:middle line:84% notorious problems in human cognition, 00:27:27.530 --> 00:27:28.710 align:middle line:90% the confirmation bias." 00:27:28.710 --> 00:27:31.450 align:middle line:84% And this is where information can be bad, not good. 00:27:31.450 --> 00:27:33.530 align:middle line:84% People rarely seek out evidence on both sides 00:27:33.530 --> 00:27:35.937 align:middle line:84% before making a decision on moral or political matters. 00:27:35.937 --> 00:27:37.770 align:middle line:84% Rather, they begin with their initial belief 00:27:37.770 --> 00:27:40.210 align:middle line:84% and seek out evidence to confirm it. 00:27:40.210 --> 00:27:45.850 align:middle line:84% And this is the fundamental challenge. 00:27:45.850 --> 00:27:47.850 align:middle line:84% The typical view of economists and scientists 00:27:47.850 --> 00:27:52.170 align:middle line:84% would be truth is the best disinfectant, that basically, 00:27:52.170 --> 00:27:54.850 align:middle line:84% if people are biased, give them the truth and that'll fix it. 00:27:54.850 --> 00:27:56.930 align:middle line:84% But in fact, that is no longer true 00:27:56.930 --> 00:27:58.770 align:middle line:90% when there's no objective truth. 00:27:58.770 --> 00:28:01.210 align:middle line:84% And that is a little bit where we've 00:28:01.210 --> 00:28:07.050 align:middle line:84% been heading, that basically, what we have is a partisan media 00:28:07.050 --> 00:28:10.590 align:middle line:84% environment that's not just confirming voters' beliefs 00:28:10.590 --> 00:28:12.250 align:middle line:90% but actually shaping them. 00:28:12.250 --> 00:28:15.550 align:middle line:84% As some authors say, "describe a polarizing feedback loop, 00:28:15.550 --> 00:28:18.530 align:middle line:84% an echo chamber, where partisans can reinforce and strengthen 00:28:18.530 --> 00:28:20.050 align:middle line:90% their initial biases." 00:28:20.050 --> 00:28:22.360 align:middle line:84% So you have these echo chambers-- the key term. 00:28:22.360 --> 00:28:24.600 align:middle line:84% People are put in these echo chambers, where 00:28:24.600 --> 00:28:26.960 align:middle line:84% they can essentially get initial biases confirmed 00:28:26.960 --> 00:28:27.740 align:middle line:90% and strengthened. 00:28:27.740 --> 00:28:29.198 align:middle line:84% And of course, this is strengthened 00:28:29.198 --> 00:28:31.225 align:middle line:84% by the algorithm used by social media companies, 00:28:31.225 --> 00:28:33.600 align:middle line:84% which, essentially, when they see you like something that 00:28:33.600 --> 00:28:35.613 align:middle line:84% is left wing or right wing, immediately pound 00:28:35.613 --> 00:28:37.280 align:middle line:84% you with more and more lefter and lefter 00:28:37.280 --> 00:28:39.540 align:middle line:84% and lefter and righter and righter and righter things. 00:28:39.540 --> 00:28:40.920 align:middle line:84% That's the way the algorithms are 00:28:40.920 --> 00:28:45.720 align:middle line:90% designed to maximize eyeballs. 00:28:45.720 --> 00:28:46.900 align:middle line:90% So we have this. 00:28:46.900 --> 00:28:49.360 align:middle line:90% 00:28:49.360 --> 00:28:51.360 align:middle line:84% I don't think what I'm saying is new to you all. 00:28:51.360 --> 00:28:54.960 align:middle line:84% The question is, is this just, quite frankly, liberal, usually, 00:28:54.960 --> 00:28:58.040 align:middle line:90% whining, or is this real? 00:28:58.040 --> 00:29:00.720 align:middle line:84% And the evidence is that it's actually quite real. 00:29:00.720 --> 00:29:04.040 align:middle line:84% And in fact, probably the best evidence for this 00:29:04.040 --> 00:29:09.880 align:middle line:84% comes from a really cool study, which asked, how much does 00:29:09.880 --> 00:29:12.680 align:middle line:90% Fox News influence voters? 00:29:12.680 --> 00:29:15.388 align:middle line:84% And look, we know Fox News has a conservative bias, but whatever. 00:29:15.388 --> 00:29:17.722 align:middle line:84% As long as people are watching it, they understand that. 00:29:17.722 --> 00:29:19.560 align:middle line:84% It shouldn't influence their views. 00:29:19.560 --> 00:29:20.560 align:middle line:90% How would you test that? 00:29:20.560 --> 00:29:22.060 align:middle line:84% Well, it's hard because people who choose to watch Fox 00:29:22.060 --> 00:29:23.940 align:middle line:84% are different than people who choose not to. 00:29:23.940 --> 00:29:25.515 align:middle line:84% There's no random assignment there. 00:29:25.515 --> 00:29:27.140 align:middle line:84% But they found a very clever way to get 00:29:27.140 --> 00:29:32.220 align:middle line:84% random assignment, which is, it turns out that most people, when 00:29:32.220 --> 00:29:35.080 align:middle line:84% they turn on the TV, when they used to watch TV, turn on TV, 00:29:35.080 --> 00:29:38.980 align:middle line:90% would start at 2 and click up. 00:29:38.980 --> 00:29:41.540 align:middle line:84% And it turns out that stations that 00:29:41.540 --> 00:29:43.740 align:middle line:84% are at the bottom of the number distribution 00:29:43.740 --> 00:29:46.060 align:middle line:84% were more likely to be watched than stations 00:29:46.060 --> 00:29:47.940 align:middle line:84% at the top of the number distribution. 00:29:47.940 --> 00:29:49.807 align:middle line:84% It also turns out that it's random, 00:29:49.807 --> 00:29:52.140 align:middle line:84% where Fox News was in the number distribution, basically 00:29:52.140 --> 00:29:54.748 align:middle line:84% depends on the timing of when Fox got to town. 00:29:54.748 --> 00:29:56.540 align:middle line:84% So they said, if you're randomly in a place 00:29:56.540 --> 00:29:58.560 align:middle line:84% where Fox is low in the number distribution, 00:29:58.560 --> 00:30:01.408 align:middle line:84% it turns out you watch Fox more than if you're 00:30:01.408 --> 00:30:03.700 align:middle line:84% in a place where Fox is randomly higher than the number 00:30:03.700 --> 00:30:05.060 align:middle line:90% distribution. 00:30:05.060 --> 00:30:07.540 align:middle line:84% And they found that places where Fox was low on the number 00:30:07.540 --> 00:30:10.770 align:middle line:84% distribution were much more likely to vote conservatively, 00:30:10.770 --> 00:30:13.020 align:middle line:84% even though being otherwise identical to a place where 00:30:13.020 --> 00:30:14.180 align:middle line:90% it was higher. 00:30:14.180 --> 00:30:16.020 align:middle line:90% And this is not small. 00:30:16.020 --> 00:30:18.140 align:middle line:90% They actually estimate that-- 00:30:18.140 --> 00:30:19.970 align:middle line:84% I don't have the numbers in the book. 00:30:19.970 --> 00:30:24.650 align:middle line:84% But they estimate that in 2004, John Kerry 00:30:24.650 --> 00:30:27.730 align:middle line:84% would have beaten George Bush if you take out the Fox effect, 00:30:27.730 --> 00:30:31.170 align:middle line:84% and that in 2012, President Obama would have won by 10% more 00:30:31.170 --> 00:30:35.115 align:middle line:84% than he did if you take out the Fox effect. 00:30:35.115 --> 00:30:36.490 align:middle line:84% So this is quantifying and saying 00:30:36.490 --> 00:30:38.850 align:middle line:84% there are enormous, enormous effects here. 00:30:38.850 --> 00:30:41.430 align:middle line:90% Now, that's just Fox News. 00:30:41.430 --> 00:30:43.410 align:middle line:84% But social media has amplified this. 00:30:43.410 --> 00:30:46.970 align:middle line:84% So in 2020, 71% of Americans reported 00:30:46.970 --> 00:30:49.970 align:middle line:84% getting at least some news from social media. 00:30:49.970 --> 00:30:55.370 align:middle line:84% And basically, the echo chamber basically 00:30:55.370 --> 00:31:00.710 align:middle line:84% allows people to confirm their biases and their positions. 00:31:00.710 --> 00:31:02.210 align:middle line:84% Something personally important to me 00:31:02.210 --> 00:31:05.830 align:middle line:84% is how it affected perceptions of the Affordable Care Act. 00:31:05.830 --> 00:31:09.730 align:middle line:84% So in a 2017, poll respondents were 00:31:09.730 --> 00:31:11.970 align:middle line:84% asked whether repeal the Affordable Care Act, 00:31:11.970 --> 00:31:15.690 align:middle line:84% which eliminated expanded Medicaid coverage and private 00:31:15.690 --> 00:31:17.320 align:middle line:84% insurance subsidies-- so basically, 00:31:17.320 --> 00:31:21.640 align:middle line:84% repeal this law, which basically gave a lot of people insurance. 00:31:21.640 --> 00:31:24.180 align:middle line:84% They asked, would this lower insurance coverage? 00:31:24.180 --> 00:31:26.520 align:middle line:90% The answer's unambiguously yes. 00:31:26.520 --> 00:31:28.900 align:middle line:90% 80% of Democrats said it would. 00:31:28.900 --> 00:31:31.680 align:middle line:84% 50% of Republicans said it would. 00:31:31.680 --> 00:31:35.160 align:middle line:84% Here is what's perhaps most disturbing. 00:31:35.160 --> 00:31:37.760 align:middle line:84% If you took people who had pre-existing beliefs 00:31:37.760 --> 00:31:40.240 align:middle line:84% about the Affordable Care Act and gave them 00:31:40.240 --> 00:31:43.360 align:middle line:84% facts about the Affordable Care Act, objective facts 00:31:43.360 --> 00:31:45.640 align:middle line:84% to both sides, that move their opinions 00:31:45.640 --> 00:31:47.760 align:middle line:90% farther apart, not closer. 00:31:47.760 --> 00:31:49.120 align:middle line:90% How can that be? 00:31:49.120 --> 00:31:51.360 align:middle line:84% It's because the confirmation bias principle. 00:31:51.360 --> 00:31:53.658 align:middle line:84% People heard what they wanted to from the facts 00:31:53.658 --> 00:31:55.700 align:middle line:84% and used to confirm their pre-existing positions. 00:31:55.700 --> 00:31:56.965 align:middle line:90% That is incredibly scary. 00:31:56.965 --> 00:31:58.340 align:middle line:84% That's the scariest thing of all. 00:31:58.340 --> 00:31:59.560 align:middle line:84% It says the problem is not necessarily 00:31:59.560 --> 00:32:01.400 align:middle line:84% even more the lack of objective truth, 00:32:01.400 --> 00:32:03.733 align:middle line:84% that basically, there is no objective truth, that if you 00:32:03.733 --> 00:32:06.360 align:middle line:84% give people objective truth, it becomes 00:32:06.360 --> 00:32:08.520 align:middle line:90% twisted into confirmation bias. 00:32:08.520 --> 00:32:12.380 align:middle line:84% So that is a clear cause of polarization. 00:32:12.380 --> 00:32:14.900 align:middle line:84% And it seems like a clear major cause of our polarization. 00:32:14.900 --> 00:32:17.740 align:middle line:90% 00:32:17.740 --> 00:32:19.380 align:middle line:90% So one is media. 00:32:19.380 --> 00:32:21.260 align:middle line:90% So one is media. 00:32:21.260 --> 00:32:25.567 align:middle line:84% The second is the voters, which is absent media. 00:32:25.567 --> 00:32:27.900 align:middle line:84% Obviously, these are going to be impossible to separate. 00:32:27.900 --> 00:32:29.320 align:middle line:84% But there is an important distinction, 00:32:29.320 --> 00:32:30.778 align:middle line:84% which is, even absent media, are we 00:32:30.778 --> 00:32:33.540 align:middle line:84% becoming more polarized as a society? 00:32:33.540 --> 00:32:35.260 align:middle line:84% And we clearly are becoming massively 00:32:35.260 --> 00:32:36.480 align:middle line:90% more polarized society. 00:32:36.480 --> 00:32:38.860 align:middle line:84% Now, we don't know how much of the media channel, 00:32:38.860 --> 00:32:41.180 align:middle line:90% but it's incredible. 00:32:41.180 --> 00:32:46.000 align:middle line:84% Here is my favorite scary fact about this. 00:32:46.000 --> 00:32:49.540 align:middle line:90% 00:32:49.540 --> 00:32:56.260 align:middle line:84% In 1960, the majority of people in America 00:32:56.260 --> 00:32:59.860 align:middle line:84% were opposed to someone marrying someone of a different race. 00:32:59.860 --> 00:33:02.653 align:middle line:84% Today, it's only 10%, still too high but whatever. 00:33:02.653 --> 00:33:04.320 align:middle line:84% 10% of America is opposed to their child 00:33:04.320 --> 00:33:06.060 align:middle line:84% marrying someone of a different race. 00:33:06.060 --> 00:33:10.260 align:middle line:84% In 1960, 5% of voters were opposed to their child 00:33:10.260 --> 00:33:12.300 align:middle line:84% marrying someone of a different political party. 00:33:12.300 --> 00:33:14.890 align:middle line:90% Today, it's 50%. 00:33:14.890 --> 00:33:16.808 align:middle line:84% So people are as opposed to their children 00:33:16.808 --> 00:33:18.850 align:middle line:84% of different political party as they were opposed 00:33:18.850 --> 00:33:21.650 align:middle line:90% to different races in 1960. 00:33:21.650 --> 00:33:25.250 align:middle line:84% And basically, the most thoughtful writer in all of this 00:33:25.250 --> 00:33:26.790 align:middle line:90% is a guy named Ezra Klein. 00:33:26.790 --> 00:33:29.450 align:middle line:84% And he talks about the rise of tribalism, 00:33:29.450 --> 00:33:32.170 align:middle line:84% that essentially, political party 00:33:32.170 --> 00:33:37.090 align:middle line:84% has become a symbol for identity, 00:33:37.090 --> 00:33:40.090 align:middle line:84% that essentially, there's a bundle of things that 00:33:40.090 --> 00:33:42.130 align:middle line:90% make you a Republican voter. 00:33:42.130 --> 00:33:46.090 align:middle line:84% And those things are reinforced, and they go together, 00:33:46.090 --> 00:33:48.390 align:middle line:84% and they become an identity a set of people, 00:33:48.390 --> 00:33:50.970 align:middle line:84% and likewise with the Democratic voter. 00:33:50.970 --> 00:33:54.450 align:middle line:84% We talk about the MAGA, true, real conservatives, 00:33:54.450 --> 00:33:56.490 align:middle line:90% or the woke left. 00:33:56.490 --> 00:33:59.770 align:middle line:84% These are groups where, basically, they 00:33:59.770 --> 00:34:01.430 align:middle line:90% are self-reinforcing. 00:34:01.430 --> 00:34:03.250 align:middle line:90% They hang out with each other-- 00:34:03.250 --> 00:34:05.310 align:middle line:90% media helps with that-- 00:34:05.310 --> 00:34:07.930 align:middle line:84% and where they're becoming less and less likely to either 00:34:07.930 --> 00:34:10.969 align:middle line:84% communicate with or even interact with people 00:34:10.969 --> 00:34:12.293 align:middle line:90% of opposing views. 00:34:12.293 --> 00:34:14.460 align:middle line:84% They don't interact with people with opposing views. 00:34:14.460 --> 00:34:16.210 align:middle line:84% They don't see people with opposing views. 00:34:16.210 --> 00:34:20.360 align:middle line:84% They are vividly opposed to those opposing views. 00:34:20.360 --> 00:34:22.940 align:middle line:84% And it becomes a sense of, everything becomes that way, 00:34:22.940 --> 00:34:25.500 align:middle line:84% that it's not like, well, I don't like them on this issue, 00:34:25.500 --> 00:34:26.345 align:middle line:84% but I'll compromise on this issue. 00:34:26.345 --> 00:34:26.845 align:middle line:90% No. 00:34:26.845 --> 00:34:29.576 align:middle line:84% It becomes, this is your tribe and you can't-- look, 00:34:29.576 --> 00:34:32.159 align:middle line:84% I don't know how many of you-- there's probably no one in this 00:34:32.159 --> 00:34:34.000 align:middle line:84% room as big a sports fan as I am. 00:34:34.000 --> 00:34:35.400 align:middle line:90% I conjecture. 00:34:35.400 --> 00:34:40.440 align:middle line:84% But if you're a sports fan, it's like sport rivalries. 00:34:40.440 --> 00:34:42.320 align:middle line:90% I heat the Yankees. 00:34:42.320 --> 00:34:43.500 align:middle line:90% I'm a Red Sox fan. 00:34:43.500 --> 00:34:44.600 align:middle line:90% I heat the Yankees. 00:34:44.600 --> 00:34:45.500 align:middle line:90% And you know what? 00:34:45.500 --> 00:34:47.148 align:middle line:84% That doesn't do anybody any harm. 00:34:47.148 --> 00:34:48.940 align:middle line:84% I don't do anything physically bad to them. 00:34:48.940 --> 00:34:50.190 align:middle line:90% I might boo them occasionally. 00:34:50.190 --> 00:34:52.580 align:middle line:90% I think they can handle it. 00:34:52.580 --> 00:34:56.760 align:middle line:84% But when people have that same feeling about the opposing 00:34:56.760 --> 00:34:59.440 align:middle line:90% political party, that matters. 00:34:59.440 --> 00:35:02.720 align:middle line:84% That matters because what happens to the Yankees 00:35:02.720 --> 00:35:04.260 align:middle line:84% does not determine my well-being. 00:35:04.260 --> 00:35:06.000 align:middle line:90% They don't rule me yet. 00:35:06.000 --> 00:35:09.240 align:middle line:84% But what happens in the elections does. 00:35:09.240 --> 00:35:12.340 align:middle line:84% And so people have the fandom, super fandom, 00:35:12.340 --> 00:35:15.140 align:middle line:84% about political parties and irrational-- my love for the Red 00:35:15.140 --> 00:35:16.320 align:middle line:90% Sox is totally irrational. 00:35:16.320 --> 00:35:18.167 align:middle line:84% There's nothing rational behind it. 00:35:18.167 --> 00:35:20.500 align:middle line:84% Then that's sort of worrisome about where we're heading. 00:35:20.500 --> 00:35:22.820 align:middle line:90% So that's the second problem. 00:35:22.820 --> 00:35:26.660 align:middle line:84% A third problem that I think is not discussed as much, 00:35:26.660 --> 00:35:30.260 align:middle line:84% but is very important, actually, is essentially 00:35:30.260 --> 00:35:38.580 align:middle line:84% institutional changes, institutional changes, which 00:35:38.580 --> 00:35:40.060 align:middle line:90% have happened at the same time. 00:35:40.060 --> 00:35:42.480 align:middle line:84% As people become polarized, the media is changing. 00:35:42.480 --> 00:35:45.220 align:middle line:90% The institutions have changed. 00:35:45.220 --> 00:35:49.900 align:middle line:84% And there's no better example of this than gerrymandering, 00:35:49.900 --> 00:35:53.900 align:middle line:90% a term many of you have heard. 00:35:53.900 --> 00:35:56.740 align:middle line:84% It's because there was a politician 00:35:56.740 --> 00:35:59.540 align:middle line:84% whose name was Gerry who wanted to create a district where 00:35:59.540 --> 00:36:00.280 align:middle line:90% he could win. 00:36:00.280 --> 00:36:02.260 align:middle line:84% So it was shaped like a salamander. 00:36:02.260 --> 00:36:04.820 align:middle line:90% So basically, here's the deal. 00:36:04.820 --> 00:36:08.340 align:middle line:90% In a rational system-- 00:36:08.340 --> 00:36:10.370 align:middle line:84% senators are elected from each state. 00:36:10.370 --> 00:36:12.890 align:middle line:84% Representatives, all 438 of them, 00:36:12.890 --> 00:36:15.410 align:middle line:90% are elected from districts. 00:36:15.410 --> 00:36:17.830 align:middle line:84% And they're supposed to be population representative. 00:36:17.830 --> 00:36:20.130 align:middle line:84% Senators are state representative. 00:36:20.130 --> 00:36:22.630 align:middle line:84% Congresspeople are supposed to be population representative. 00:36:22.630 --> 00:36:24.130 align:middle line:84% So the idea is to divide the country 00:36:24.130 --> 00:36:27.450 align:middle line:84% into 438 pieces in a sensible way 00:36:27.450 --> 00:36:30.850 align:middle line:84% so they represent the US population. 00:36:30.850 --> 00:36:31.830 align:middle line:90% That is the idea. 00:36:31.830 --> 00:36:33.538 align:middle line:84% So you can imagine how you could do this. 00:36:33.538 --> 00:36:35.078 align:middle line:90% An MIT thesis could do this. 00:36:35.078 --> 00:36:35.870 align:middle line:90% You could sit down. 00:36:35.870 --> 00:36:37.370 align:middle line:84% You could say, look, districts have 00:36:37.370 --> 00:36:39.210 align:middle line:84% to be geographically sensible because people 00:36:39.210 --> 00:36:40.127 align:middle line:90% go to vote physically. 00:36:40.127 --> 00:36:41.877 align:middle line:84% Maybe in 50 years, we'll all vote by mail, 00:36:41.877 --> 00:36:42.910 align:middle line:90% and it won't matter. 00:36:42.910 --> 00:36:44.670 align:middle line:84% But traditionally, people vote physically. 00:36:44.670 --> 00:36:46.753 align:middle line:84% So you've got to have some geographically sensible 00:36:46.753 --> 00:36:48.490 align:middle line:90% configuration of land. 00:36:48.490 --> 00:36:52.050 align:middle line:84% Let's figure out the best way to geographically 00:36:52.050 --> 00:36:55.730 align:middle line:84% divide the country into 438 blocks 00:36:55.730 --> 00:36:58.010 align:middle line:84% that best represent the views of our country. 00:36:58.010 --> 00:36:59.885 align:middle line:84% Now, there'd be some decisions along the way, 00:36:59.885 --> 00:37:02.530 align:middle line:90% but you could get pretty close. 00:37:02.530 --> 00:37:04.850 align:middle line:90% That's how most countries do it. 00:37:04.850 --> 00:37:09.430 align:middle line:84% In the US, in many states, the shapes of districts 00:37:09.430 --> 00:37:14.570 align:middle line:84% are created not by some apolitical process 00:37:14.570 --> 00:37:17.430 align:middle line:84% but by the state legislatures themselves. 00:37:17.430 --> 00:37:20.350 align:middle line:84% State legislatures determine what districts look like. 00:37:20.350 --> 00:37:22.070 align:middle line:84% And state legislatures therefore have 00:37:22.070 --> 00:37:25.470 align:middle line:84% incentive to shape the districts in a way that 00:37:25.470 --> 00:37:28.950 align:middle line:84% maximizes the chance that their party wins. 00:37:28.950 --> 00:37:30.630 align:middle line:90% How do they do that? 00:37:30.630 --> 00:37:32.610 align:middle line:84% They do that by cracking and packing. 00:37:32.610 --> 00:37:33.570 align:middle line:90% I love these terms. 00:37:33.570 --> 00:37:37.630 align:middle line:84% Cracking is divide your opponent's supporters 00:37:37.630 --> 00:37:41.950 align:middle line:84% into multiple districts so in no district are they a majority. 00:37:41.950 --> 00:37:45.930 align:middle line:84% Packing means make one or two districts 00:37:45.930 --> 00:37:48.190 align:middle line:84% which have a ton of your opponents in them, 00:37:48.190 --> 00:37:52.150 align:middle line:84% but then you become majority everywhere else. 00:37:52.150 --> 00:37:53.950 align:middle line:84% So the best way to think about this 00:37:53.950 --> 00:37:56.790 align:middle line:84% is, essentially, you want to maximize 00:37:56.790 --> 00:38:01.670 align:middle line:84% the number of wasted votes that the opposing party has. 00:38:01.670 --> 00:38:05.630 align:middle line:84% You'd like to win every race 51 to 49 00:38:05.630 --> 00:38:08.020 align:middle line:84% because then you win every race and your opponent's 00:38:08.020 --> 00:38:10.440 align:middle line:90% wasted all their votes. 00:38:10.440 --> 00:38:13.260 align:middle line:84% So essentially, every vote, which 00:38:13.260 --> 00:38:18.200 align:middle line:84% is every time you lose or vote where you win by more than 51%, 00:38:18.200 --> 00:38:19.040 align:middle line:90% those are wasted. 00:38:19.040 --> 00:38:21.780 align:middle line:84% The only vote that matters is the vote that gets you to 51%. 00:38:21.780 --> 00:38:23.280 align:middle line:90% Everything above that's a waste. 00:38:23.280 --> 00:38:25.580 align:middle line:90% Everything below that's a waste. 00:38:25.580 --> 00:38:29.100 align:middle line:84% So you can measure how efficient political systems are 00:38:29.100 --> 00:38:34.020 align:middle line:84% or how neutral they are by how many wasted votes there are. 00:38:34.020 --> 00:38:36.660 align:middle line:84% The more wasted votes, the less well 00:38:36.660 --> 00:38:38.525 align:middle line:90% designed a district system is. 00:38:38.525 --> 00:38:39.900 align:middle line:84% You'd like a district system that 00:38:39.900 --> 00:38:43.140 align:middle line:84% would minimize wasted votes, whereas partisans want a system 00:38:43.140 --> 00:38:45.500 align:middle line:90% that maximizes wasted votes. 00:38:45.500 --> 00:38:47.860 align:middle line:84% This is actually pretty cool because this 00:38:47.860 --> 00:38:50.260 align:middle line:84% is the first time I know that computer simulations were 00:38:50.260 --> 00:38:52.300 align:middle line:84% presented in front of the Supreme Court-- 00:38:52.300 --> 00:38:54.960 align:middle line:84% was in 2017, there was a case about this. 00:38:54.960 --> 00:38:57.660 align:middle line:90% 00:38:57.660 --> 00:39:01.140 align:middle line:84% Wisconsin had a massive efficiency gap. 00:39:01.140 --> 00:39:03.740 align:middle line:84% That is, in a typical system, you'd 00:39:03.740 --> 00:39:05.890 align:middle line:84% like to be a zero efficiency gap. 00:39:05.890 --> 00:39:07.830 align:middle line:84% Both parties would be equally efficient. 00:39:07.830 --> 00:39:10.370 align:middle line:84% Both parties have the same number of wasted votes. 00:39:10.370 --> 00:39:12.450 align:middle line:90% Wisconsin, there was a 12% gap. 00:39:12.450 --> 00:39:16.130 align:middle line:84% Democrats had 12% more wasted votes than Republicans. 00:39:16.130 --> 00:39:20.135 align:middle line:84% Now, the Wisconsin Republican Congress state legislature 00:39:20.135 --> 00:39:21.510 align:middle line:84% supported this, said, well, look, 00:39:21.510 --> 00:39:23.610 align:middle line:84% that's just random chance it's going to happen. 00:39:23.610 --> 00:39:25.630 align:middle line:84% So they did a computer simulation. 00:39:25.630 --> 00:39:28.970 align:middle line:84% And he essentially drew 200 different possible 00:39:28.970 --> 00:39:31.650 align:middle line:84% configurations of geographies of Wisconsin 00:39:31.650 --> 00:39:35.410 align:middle line:84% and, in each configuration, measured the efficiency gap. 00:39:35.410 --> 00:39:38.250 align:middle line:84% And all 200, their existing configuration 00:39:38.250 --> 00:39:41.930 align:middle line:84% had a bigger efficiency gap than any other randomly selected 00:39:41.930 --> 00:39:43.952 align:middle line:84% possible configuration of areas, which 00:39:43.952 --> 00:39:45.410 align:middle line:84% clearly showed there was bias going 00:39:45.410 --> 00:39:47.690 align:middle line:90% into selecting those areas. 00:39:47.690 --> 00:39:53.050 align:middle line:84% So basically, we have this problem, fundamentally 00:39:53.050 --> 00:39:55.927 align:middle line:84% put, that voters should choose politicians, 00:39:55.927 --> 00:39:57.510 align:middle line:84% but right now, through gerrymandering, 00:39:57.510 --> 00:39:59.650 align:middle line:84% politicians are choosing their voters 00:39:59.650 --> 00:40:01.490 align:middle line:84% because they're determining where voters go 00:40:01.490 --> 00:40:02.890 align:middle line:90% and who they vote for. 00:40:02.890 --> 00:40:05.630 align:middle line:90% That's the kind of institution. 00:40:05.630 --> 00:40:07.470 align:middle line:84% Now, of course, there's another institution, 00:40:07.470 --> 00:40:09.345 align:middle line:84% which we talk about a lot, which is coming up 00:40:09.345 --> 00:40:10.350 align:middle line:90% in a couple of months. 00:40:10.350 --> 00:40:12.030 align:middle line:90% Remember to vote. 00:40:12.030 --> 00:40:13.330 align:middle line:90% I'll keep reminding you. 00:40:13.330 --> 00:40:15.883 align:middle line:84% Remember, if you don't know what to do, go to MIT TurboVote. 00:40:15.883 --> 00:40:17.050 align:middle line:90% Learn about how to register. 00:40:17.050 --> 00:40:17.930 align:middle line:90% Make sure you vote. 00:40:17.930 --> 00:40:20.482 align:middle line:84% Presidential election's coming up. 00:40:20.482 --> 00:40:23.190 align:middle line:84% And there, we're going to hear a lot about the Electoral College, 00:40:23.190 --> 00:40:26.910 align:middle line:84% which is, once again, another institution-- this is not new-- 00:40:26.910 --> 00:40:29.550 align:middle line:84% but another institution which is affecting the bias 00:40:29.550 --> 00:40:32.870 align:middle line:84% in our system, which is that we have had, 00:40:32.870 --> 00:40:35.110 align:middle line:84% essentially, I think-- let's see. 00:40:35.110 --> 00:40:38.970 align:middle line:84% It's, what's, three of the last six elections, 00:40:38.970 --> 00:40:41.710 align:middle line:84% the winner of the election has not won the popular vote. 00:40:41.710 --> 00:40:45.510 align:middle line:84% So the majority of America is not electing the candidate. 00:40:45.510 --> 00:40:48.870 align:middle line:84% So that's another institution that matters. 00:40:48.870 --> 00:40:52.550 align:middle line:84% The one institution that matters-- yeah? 00:40:52.550 --> 00:40:54.574 align:middle line:84% AUDIENCE: [INAUDIBLE] the United States, 00:40:54.574 --> 00:40:57.010 align:middle line:84% they decided [INAUDIBLE] electoral district 00:40:57.010 --> 00:40:59.790 align:middle line:84% and see who wins this instead of having an aggregation 00:40:59.790 --> 00:41:01.350 align:middle line:90% of individual votes? 00:41:01.350 --> 00:41:02.970 align:middle line:84% JONATHAN GRUBER: You're talking about the Electoral College? 00:41:02.970 --> 00:41:03.360 align:middle line:90% AUDIENCE: Yeah. 00:41:03.360 --> 00:41:03.800 align:middle line:90% JONATHAN GRUBER: Yeah. 00:41:03.800 --> 00:41:05.550 align:middle line:84% That was basically historically because it 00:41:05.550 --> 00:41:06.985 align:middle line:90% was too hard to aggregate. 00:41:06.985 --> 00:41:08.860 align:middle line:84% Everyone was just voting on a piece of paper, 00:41:08.860 --> 00:41:11.582 align:middle line:84% and you had to drive by horses and get them all together. 00:41:11.582 --> 00:41:13.040 align:middle line:84% So it's just too hard to aggregate. 00:41:13.040 --> 00:41:14.415 align:middle line:84% There's no reason to have it now, 00:41:14.415 --> 00:41:17.660 align:middle line:84% except let's, remember, think about the political economy, 00:41:17.660 --> 00:41:21.820 align:middle line:84% which is, as long as more states-- 00:41:21.820 --> 00:41:24.140 align:middle line:84% to repeal this, you would need an amendment 00:41:24.140 --> 00:41:25.377 align:middle line:90% to the Constitution. 00:41:25.377 --> 00:41:27.460 align:middle line:84% The amendment of the Constitution has to be passed 00:41:27.460 --> 00:41:29.340 align:middle line:90% by 2/3 of states. 00:41:29.340 --> 00:41:33.740 align:middle line:84% As long as 1/3 of states, 1/3 plus one state, 00:41:33.740 --> 00:41:38.665 align:middle line:84% benefit more from this, then the other 2/3 lose. 00:41:38.665 --> 00:41:40.040 align:middle line:84% You'll never pass that amendment. 00:41:40.040 --> 00:41:42.180 align:middle line:84% And that'll, by definition, always be true. 00:41:42.180 --> 00:41:44.260 align:middle line:84% By definition, there'll always be at least 1/3 of states that 00:41:44.260 --> 00:41:46.260 align:middle line:84% are better off because of the Electoral College. 00:41:46.260 --> 00:41:48.562 align:middle line:90% That may shift over time. 00:41:48.562 --> 00:41:51.020 align:middle line:84% So if you could manage some period of time, where suddenly, 00:41:51.020 --> 00:41:52.520 align:middle line:84% Electoral College was neutral, maybe 00:41:52.520 --> 00:41:53.700 align:middle line:90% then you could get a change. 00:41:53.700 --> 00:41:55.300 align:middle line:84% But if it's ever not neutral, the people 00:41:55.300 --> 00:41:57.760 align:middle line:84% who win from it, people who get more power because of it-- 00:41:57.760 --> 00:42:08.720 align:middle line:84% so a vote for president cast in Wyoming 00:42:08.720 --> 00:42:11.160 align:middle line:84% currently counts four times as much for president 00:42:11.160 --> 00:42:12.722 align:middle line:90% as one cast in California. 00:42:12.722 --> 00:42:15.180 align:middle line:84% Why would you want to give up that power if you're Wyoming? 00:42:15.180 --> 00:42:18.120 align:middle line:84% What politician would ever visit Wyoming with a national 00:42:18.120 --> 00:42:20.620 align:middle line:90% electoral-- they wouldn't. 00:42:20.620 --> 00:42:23.760 align:middle line:84% So basically, the argument that advocacy of Electoral College 00:42:23.760 --> 00:42:26.260 align:middle line:84% make, which is worth discussing-- 00:42:26.260 --> 00:42:28.040 align:middle line:90% this is not a priori crazy-- 00:42:28.040 --> 00:42:31.080 align:middle line:84% is, look, if we got rid of the Electoral College, 00:42:31.080 --> 00:42:33.080 align:middle line:84% politicians would never visit any but the top 10 00:42:33.080 --> 00:42:33.960 align:middle line:90% most populous states. 00:42:33.960 --> 00:42:34.860 align:middle line:90% Because why bother? 00:42:34.860 --> 00:42:36.160 align:middle line:90% Be inefficient. 00:42:36.160 --> 00:42:38.360 align:middle line:84% So the whole country would not get attention 00:42:38.360 --> 00:42:40.080 align:middle line:84% if it got rid of the Electoral College. 00:42:40.080 --> 00:42:42.040 align:middle line:90% That's the argument. 00:42:42.040 --> 00:42:42.940 align:middle line:90% Yeah, Steven? 00:42:42.940 --> 00:42:44.360 align:middle line:90% AUDIENCE: I was just curious. 00:42:44.360 --> 00:42:46.610 align:middle line:84% When you talk about the Electoral College, the thought 00:42:46.610 --> 00:42:48.567 align:middle line:84% that I have is, at some point, there 00:42:48.567 --> 00:42:50.400 align:middle line:84% might have been concerns about states trying 00:42:50.400 --> 00:42:52.108 align:middle line:84% to increase the amount of votes that they 00:42:52.108 --> 00:42:54.480 align:middle line:84% get relative to other states in the general election 00:42:54.480 --> 00:42:57.640 align:middle line:84% by, for example, lowering the voting age to five or something 00:42:57.640 --> 00:43:01.470 align:middle line:84% like that, where you can make a bunch of changes 00:43:01.470 --> 00:43:02.230 align:middle line:90% to try and get-- 00:43:02.230 --> 00:43:04.272 align:middle line:84% JONATHAN GRUBER: The voting age is set federally. 00:43:04.272 --> 00:43:06.750 align:middle line:84% The voting age is set federally, I believe. 00:43:06.750 --> 00:43:07.770 align:middle line:90% No, is that not true? 00:43:07.770 --> 00:43:08.670 align:middle line:90% AUDIENCE: No. 00:43:08.670 --> 00:43:10.810 align:middle line:84% JONATHAN GRUBER: Can states let 16-year-olds vote if they want? 00:43:10.810 --> 00:43:12.790 align:middle line:84% AUDIENCE: They can also let noncitizens vote if they want. 00:43:12.790 --> 00:43:13.990 align:middle line:90% JONATHAN GRUBER: OK. 00:43:13.990 --> 00:43:15.370 align:middle line:84% I know about that, but I'm not sure about age. 00:43:15.370 --> 00:43:16.470 align:middle line:90% Are you sure about age? 00:43:16.470 --> 00:43:18.965 align:middle line:84% AUDIENCE: Age is maximum limit is 18. 00:43:18.965 --> 00:43:19.590 align:middle line:90% So [INAUDIBLE]. 00:43:19.590 --> 00:43:21.070 align:middle line:90% JONATHAN GRUBER: Maximum, great. 00:43:21.070 --> 00:43:22.050 align:middle line:90% So that's a good point. 00:43:22.050 --> 00:43:23.842 align:middle line:84% That may be another reason why you wouldn't 00:43:23.842 --> 00:43:26.830 align:middle line:84% want to race to the bottom among states of letting newborns vote. 00:43:26.830 --> 00:43:28.950 align:middle line:90% That's a fair point. 00:43:28.950 --> 00:43:32.425 align:middle line:84% But on the other hand, the Electoral College-- it's funny. 00:43:32.425 --> 00:43:34.550 align:middle line:84% Those people who say, well, all the attention would 00:43:34.550 --> 00:43:35.830 align:middle line:84% be paid to California are missing the fact 00:43:35.830 --> 00:43:37.950 align:middle line:84% that now all the attention's paid to Pennsylvania. 00:43:37.950 --> 00:43:39.450 align:middle line:84% So no matter what system you have, 00:43:39.450 --> 00:43:40.950 align:middle line:84% there'll be a small subset of states 00:43:40.950 --> 00:43:43.510 align:middle line:90% that get all the attention. 00:43:43.510 --> 00:43:46.430 align:middle line:84% With our current system, it's the seven swing states, 00:43:46.430 --> 00:43:48.630 align:middle line:84% which are where the median voters live. 00:43:48.630 --> 00:43:50.310 align:middle line:84% So the median voter theorem lives. 00:43:50.310 --> 00:43:53.557 align:middle line:84% And the reason it lives is because Harris and Trump 00:43:53.557 --> 00:43:55.390 align:middle line:84% will be spending a huge amount of their time 00:43:55.390 --> 00:43:58.310 align:middle line:84% over the next few months in just seven states-- 00:43:58.310 --> 00:44:04.060 align:middle line:84% Pennsylvania, Michigan, Wisconsin, Nevada, Arizona, 00:44:04.060 --> 00:44:06.495 align:middle line:90% Georgia, and North Carolina. 00:44:06.495 --> 00:44:08.120 align:middle line:84% They'll be spending the majority-- now, 00:44:08.120 --> 00:44:11.580 align:middle line:84% those seven states only have probably 20% of the US 00:44:11.580 --> 00:44:14.413 align:middle line:84% population, but they'll get 80% of their time. 00:44:14.413 --> 00:44:16.580 align:middle line:84% So it's not like having the Electoral Colleges solve 00:44:16.580 --> 00:44:19.380 align:middle line:84% the problem of politicians concentrating their activity. 00:44:19.380 --> 00:44:20.880 align:middle line:84% If anything, it's made it less good. 00:44:20.880 --> 00:44:23.297 align:middle line:84% At least if they're spending all their time in big states, 00:44:23.297 --> 00:44:25.020 align:middle line:90% it'd represent more Americans. 00:44:25.020 --> 00:44:27.040 align:middle line:90% So that's the Electoral College. 00:44:27.040 --> 00:44:28.740 align:middle line:84% We actually have another feature which 00:44:28.740 --> 00:44:31.060 align:middle line:84% would be even harder to change, which 00:44:31.060 --> 00:44:33.020 align:middle line:90% is the structure of the Senate. 00:44:33.020 --> 00:44:34.840 align:middle line:84% I don't know if we have Hamilton fans here, 00:44:34.840 --> 00:44:36.100 align:middle line:84% but we all know about the compromise 00:44:36.100 --> 00:44:37.260 align:middle line:90% by which America was set up. 00:44:37.260 --> 00:44:39.600 align:middle line:84% It was a compromise between states' rights advocates or not. 00:44:39.600 --> 00:44:41.225 align:middle line:84% And the way they solved that compromise 00:44:41.225 --> 00:44:45.020 align:middle line:84% was by creating a body where the representation is essentially 00:44:45.020 --> 00:44:47.820 align:middle line:90% by state, not by population. 00:44:47.820 --> 00:44:54.020 align:middle line:84% The result is, the current projections are about 70% 00:44:54.020 --> 00:44:59.360 align:middle line:84% of the US population will be represented by 30% of the Senate 00:44:59.360 --> 00:45:01.940 align:middle line:84% because small states get two senators, 00:45:01.940 --> 00:45:03.840 align:middle line:90% just like big states do. 00:45:03.840 --> 00:45:10.200 align:middle line:84% So once again, if you really care about direct democracy, 00:45:10.200 --> 00:45:12.200 align:middle line:84% it's hard to think how that makes sense. 00:45:12.200 --> 00:45:13.800 align:middle line:84% But once again, the argument would 00:45:13.800 --> 00:45:15.640 align:middle line:84% be that there's a unique set of interests 00:45:15.640 --> 00:45:17.000 align:middle line:84% that these states have that would get ignored 00:45:17.000 --> 00:45:18.300 align:middle line:90% if it was popular-- once again. 00:45:18.300 --> 00:45:19.130 align:middle line:90% And they're right. 00:45:19.130 --> 00:45:20.880 align:middle line:84% No one would ever pay attention to Wyoming 00:45:20.880 --> 00:45:22.920 align:middle line:90% if there was no Senate. 00:45:22.920 --> 00:45:24.280 align:middle line:90% They'd have no power. 00:45:24.280 --> 00:45:27.220 align:middle line:84% So basically, that's a fair point. 00:45:27.220 --> 00:45:29.280 align:middle line:90% The question is, why do we care? 00:45:29.280 --> 00:45:32.720 align:middle line:84% And that's an important question to be thinking about. 00:45:32.720 --> 00:45:36.200 align:middle line:84% So those are the kinds of institutions. 00:45:36.200 --> 00:45:41.080 align:middle line:84% But there are other institutional changes 00:45:41.080 --> 00:45:42.140 align:middle line:90% that matter too. 00:45:42.140 --> 00:45:48.343 align:middle line:90% 00:45:48.343 --> 00:45:50.260 align:middle line:84% So one example-- there's a number of examples. 00:45:50.260 --> 00:45:57.190 align:middle line:84% Some recent examples is things like voter fraud concerns 00:45:57.190 --> 00:45:59.930 align:middle line:84% and voter registration, that basically, 00:45:59.930 --> 00:46:01.567 align:middle line:84% once again, in principle, we should 00:46:01.567 --> 00:46:03.150 align:middle line:84% have a national conversation about who 00:46:03.150 --> 00:46:07.313 align:middle line:84% is eligible to elect our representatives. 00:46:07.313 --> 00:46:09.230 align:middle line:84% And then everyone should follow the same rules 00:46:09.230 --> 00:46:10.750 align:middle line:90% and allow those people to vote. 00:46:10.750 --> 00:46:12.590 align:middle line:84% That is no longer happening, or there's 00:46:12.590 --> 00:46:14.190 align:middle line:84% a concern that's no longer happening, 00:46:14.190 --> 00:46:16.815 align:middle line:84% and that different states, based on their partisan composition, 00:46:16.815 --> 00:46:19.830 align:middle line:84% may be restricting who can vote in different ways, 00:46:19.830 --> 00:46:22.553 align:middle line:90% maybe passing laws which say-- 00:46:22.553 --> 00:46:23.970 align:middle line:84% this maybe is another confirmation 00:46:23.970 --> 00:46:25.990 align:middle line:84% of your point, Steven, which is that maybe it's 00:46:25.990 --> 00:46:28.350 align:middle line:84% not playing with age but playing with other things, 00:46:28.350 --> 00:46:32.670 align:middle line:84% like criminal status, access to a polling place, 00:46:32.670 --> 00:46:36.390 align:middle line:84% et cetera that is basically causing 00:46:36.390 --> 00:46:40.710 align:middle line:84% there to be a partisan bias and essentially who can vote. 00:46:40.710 --> 00:46:43.810 align:middle line:84% So basically, that's another institutional change. 00:46:43.810 --> 00:46:48.990 align:middle line:90% 00:46:48.990 --> 00:46:54.230 align:middle line:84% So the point is, we're seeing lots of interesting changes 00:46:54.230 --> 00:46:57.180 align:middle line:90% in how we institutionalize this. 00:46:57.180 --> 00:47:00.060 align:middle line:90% And so the question is-- 00:47:00.060 --> 00:47:02.240 align:middle line:84% we could talk about which of these is the problem. 00:47:02.240 --> 00:47:04.052 align:middle line:90% They're all the problem. 00:47:04.052 --> 00:47:05.260 align:middle line:90% And I'm going to use problem. 00:47:05.260 --> 00:47:07.320 align:middle line:84% I have no problem using a loaded term like that. 00:47:07.320 --> 00:47:10.120 align:middle line:84% It is a problem that we talk past each other, 00:47:10.120 --> 00:47:12.880 align:middle line:84% that we don't have the same set of objective facts, 00:47:12.880 --> 00:47:17.220 align:middle line:84% that we don't have a truly representative democracy 00:47:17.220 --> 00:47:17.980 align:middle line:90% in which-- 00:47:17.980 --> 00:47:20.480 align:middle line:84% I don't mean the Senate, Electoral College thing. 00:47:20.480 --> 00:47:22.480 align:middle line:84% That is just debate between states' rights, not. 00:47:22.480 --> 00:47:25.520 align:middle line:84% And I have my views, but there's no objective truth on that, 00:47:25.520 --> 00:47:26.140 align:middle line:90% I think. 00:47:26.140 --> 00:47:28.140 align:middle line:84% It's more stuff like gerrymandering 00:47:28.140 --> 00:47:30.380 align:middle line:84% or the other biases that we talked about that 00:47:30.380 --> 00:47:32.740 align:middle line:90% are more obviously problematic. 00:47:32.740 --> 00:47:35.860 align:middle line:84% So the question is for you guys to solve, 00:47:35.860 --> 00:47:40.020 align:middle line:84% lest your lives be ruined-- what are we going to do about this? 00:47:40.020 --> 00:47:40.680 align:middle line:90% I'm old. 00:47:40.680 --> 00:47:42.620 align:middle line:90% I'm done. 00:47:42.620 --> 00:47:45.100 align:middle line:84% Basically, I've gotten through this pretty well. 00:47:45.100 --> 00:47:50.320 align:middle line:84% But you guys, this is your life, and it's only getting worse. 00:47:50.320 --> 00:47:52.800 align:middle line:84% So the question is-- and I mean this legitimately. 00:47:52.800 --> 00:47:54.600 align:middle line:90% I'm asking I'm asking the class. 00:47:54.600 --> 00:47:57.580 align:middle line:90% What can we do about this? 00:47:57.580 --> 00:47:59.560 align:middle line:84% What answers are there out there to try 00:47:59.560 --> 00:48:00.800 align:middle line:84% to address these kinds of problems that 00:48:00.800 --> 00:48:02.050 align:middle line:90% are causing more polarization? 00:48:02.050 --> 00:48:02.700 align:middle line:90% Yeah, Alec? 00:48:02.700 --> 00:48:05.127 align:middle line:84% AUDIENCE: Independent districting committees. 00:48:05.127 --> 00:48:06.960 align:middle line:84% JONATHAN GRUBER: So that's a perfect example 00:48:06.960 --> 00:48:10.320 align:middle line:84% of a technical but very valid conclusion, which 00:48:10.320 --> 00:48:11.513 align:middle line:90% some states are doing. 00:48:11.513 --> 00:48:13.680 align:middle line:84% Some states have independent districting committees, 00:48:13.680 --> 00:48:16.320 align:middle line:84% which are trying to really minimize this efficiency gap 00:48:16.320 --> 00:48:19.900 align:middle line:84% and choose districts in the most appropriate way, 00:48:19.900 --> 00:48:21.240 align:middle line:90% not in a political way. 00:48:21.240 --> 00:48:23.760 align:middle line:90% What else can we do? 00:48:23.760 --> 00:48:24.880 align:middle line:90% Alec? 00:48:24.880 --> 00:48:26.160 align:middle line:90% AUDIENCE: Mandatory voting. 00:48:26.160 --> 00:48:28.800 align:middle line:84% It would change the issue [INAUDIBLE]. 00:48:28.800 --> 00:48:33.360 align:middle line:84% JONATHAN GRUBER: So my son is new wife is from Brazil. 00:48:33.360 --> 00:48:36.680 align:middle line:84% They have massively higher voting turnout than we do. 00:48:36.680 --> 00:48:37.423 align:middle line:90% Why? 00:48:37.423 --> 00:48:38.340 align:middle line:90% Well, for two reasons. 00:48:38.340 --> 00:48:40.642 align:middle line:84% First of all, you get fined if you don't vote. 00:48:40.642 --> 00:48:42.100 align:middle line:84% Second of all, voting is a holiday. 00:48:42.100 --> 00:48:43.640 align:middle line:90% Voting day is a holiday. 00:48:43.640 --> 00:48:45.480 align:middle line:84% We're, I think, the only major country where 00:48:45.480 --> 00:48:48.360 align:middle line:90% voting day is on a work day. 00:48:48.360 --> 00:48:51.720 align:middle line:84% Now, once again, how could you be opposed to voting day 00:48:51.720 --> 00:48:52.847 align:middle line:90% not being in a work day? 00:48:52.847 --> 00:48:54.430 align:middle line:84% Well, you're opposed to that if you're 00:48:54.430 --> 00:48:56.805 align:middle line:84% the side that wants low turnout, which, once again, could 00:48:56.805 --> 00:48:58.030 align:middle line:90% vary over time. 00:48:58.030 --> 00:48:59.570 align:middle line:84% Right now it's the Republican side 00:48:59.570 --> 00:49:01.190 align:middle line:90% which would like low turnout. 00:49:01.190 --> 00:49:03.710 align:middle line:84% In 20 years, it could be the Democrats who want low turnout. 00:49:03.710 --> 00:49:05.930 align:middle line:84% The bottom line is, if you want low turnout, 00:49:05.930 --> 00:49:09.510 align:middle line:84% you like voting being on a Tuesday 00:49:09.510 --> 00:49:13.310 align:middle line:84% and you like it not being mandatory. 00:49:13.310 --> 00:49:16.110 align:middle line:84% So those are things we could do-- once again, 00:49:16.110 --> 00:49:18.350 align:middle line:84% all things that other countries do that are sensible. 00:49:18.350 --> 00:49:20.310 align:middle line:90% What else? 00:49:20.310 --> 00:49:21.050 align:middle line:90% Yeah, Enoch? 00:49:21.050 --> 00:49:22.750 align:middle line:84% AUDIENCE: Getting rid of super PACs. 00:49:22.750 --> 00:49:24.870 align:middle line:84% JONATHAN GRUBER: So we could try to tamp down 00:49:24.870 --> 00:49:26.310 align:middle line:90% on money in politics. 00:49:26.310 --> 00:49:29.390 align:middle line:84% So what happened was, the Supreme Court essentially 00:49:29.390 --> 00:49:32.375 align:middle line:84% took a limit of how much money, through various mechanisms, 00:49:32.375 --> 00:49:34.250 align:middle line:84% people could contribute to politics in what-- 00:49:34.250 --> 00:49:36.510 align:middle line:84% 2012, I think was Citizens United-- 00:49:36.510 --> 00:49:37.130 align:middle line:90% 2008. 00:49:37.130 --> 00:49:38.230 align:middle line:90% I don't remember. 00:49:38.230 --> 00:49:40.590 align:middle line:84% They took a limit off how much people can contribute. 00:49:40.590 --> 00:49:42.740 align:middle line:84% And money in politics by and large, 00:49:42.740 --> 00:49:44.990 align:middle line:84% while in principle-- once again, here's the key thing. 00:49:44.990 --> 00:49:47.015 align:middle line:84% In principle, money in politics could be good 00:49:47.015 --> 00:49:49.390 align:middle line:84% because it could represent intensity or preference, which 00:49:49.390 --> 00:49:50.930 align:middle line:84% is missed by the median voter model. 00:49:50.930 --> 00:49:53.530 align:middle line:84% In practice, it's bad because it overly 00:49:53.530 --> 00:49:55.430 align:middle line:84% represents intensity of preferences, 00:49:55.430 --> 00:49:57.610 align:middle line:84% and in particular, overly represents the preferences 00:49:57.610 --> 00:50:00.810 align:middle line:84% of people who happen to have money at that point in time. 00:50:00.810 --> 00:50:04.410 align:middle line:84% And so as a result, you are getting 00:50:04.410 --> 00:50:07.490 align:middle line:84% politicians paying attention to particular donors 00:50:07.490 --> 00:50:10.410 align:middle line:84% rather than the needs of all their constituents. 00:50:10.410 --> 00:50:12.050 align:middle line:84% And you're not necessarily getting 00:50:12.050 --> 00:50:14.350 align:middle line:84% what is in the interest of social welfare. 00:50:14.350 --> 00:50:16.610 align:middle line:84% Remember our social welfare weights. 00:50:16.610 --> 00:50:18.568 align:middle line:84% Remember, we think that, basically, we probably 00:50:18.568 --> 00:50:20.610 align:middle line:84% would be redistributing from those who have a lot 00:50:20.610 --> 00:50:22.390 align:middle line:84% and thus a low marginal utility consumed. 00:50:22.390 --> 00:50:23.890 align:middle line:84% Those who have a little and a high marginal utility 00:50:23.890 --> 00:50:26.390 align:middle line:84% consume, that's the opposite of where the money is currently 00:50:26.390 --> 00:50:27.050 align:middle line:90% held. 00:50:27.050 --> 00:50:30.570 align:middle line:84% And so there'll be a natural bias against social welfare 00:50:30.570 --> 00:50:32.330 align:middle line:90% maximizing redistribution. 00:50:32.330 --> 00:50:34.950 align:middle line:84% So money in politics is another thing we can do. 00:50:34.950 --> 00:50:36.530 align:middle line:90% Yeah? 00:50:36.530 --> 00:50:38.310 align:middle line:84% AUDIENCE: One issue about mass media 00:50:38.310 --> 00:50:43.370 align:middle line:84% echo chambers is that once a news company formed their user 00:50:43.370 --> 00:50:47.370 align:middle line:84% base, which is usually biased, they have no incentive 00:50:47.370 --> 00:50:51.200 align:middle line:84% to deviate from that towards a nonbiased [INAUDIBLE] type 00:50:51.200 --> 00:50:52.120 align:middle line:90% of media. 00:50:52.120 --> 00:50:58.720 align:middle line:84% So if there were a way to either have a greater focus on, quote 00:50:58.720 --> 00:51:02.800 align:middle line:84% unquote, "centrist sources" to re-incentivize the media 00:51:02.800 --> 00:51:03.420 align:middle line:90% companies-- 00:51:03.420 --> 00:51:07.440 align:middle line:84% JONATHAN GRUBER: This is, to me, the fundamental challenge 00:51:07.440 --> 00:51:08.680 align:middle line:90% we face, which is-- 00:51:08.680 --> 00:51:11.912 align:middle line:84% that, to me, of all these problems is the dominant one. 00:51:11.912 --> 00:51:14.120 align:middle line:84% Because I can see answers to a lot of these problems. 00:51:14.120 --> 00:51:15.620 align:middle line:84% Along the way, these were suggested. 00:51:15.620 --> 00:51:18.120 align:middle line:84% That one, I don't see the answer. 00:51:18.120 --> 00:51:21.860 align:middle line:84% Without nationalizing news, how do you get-- 00:51:21.860 --> 00:51:23.700 align:middle line:84% they're profit-maximizing companies. 00:51:23.700 --> 00:51:24.620 align:middle line:90% How do you get them to do that? 00:51:24.620 --> 00:51:25.120 align:middle line:90% Yeah? 00:51:25.120 --> 00:51:30.700 align:middle line:84% AUDIENCE: [INAUDIBLE] they do this type of [INAUDIBLE] reform 00:51:30.700 --> 00:51:34.382 align:middle line:84% [INAUDIBLE] has information from both of the [? position? ?] 00:51:34.382 --> 00:51:36.340 align:middle line:84% JONATHAN GRUBER: Yeah, we could try to do that. 00:51:36.340 --> 00:51:38.440 align:middle line:84% The only problem is, I'm sort of depressed by-- the evidence says 00:51:38.440 --> 00:51:40.720 align:middle line:84% that facts don't matter or could make things worse. 00:51:40.720 --> 00:51:42.660 align:middle line:90% So getting both sides-- 00:51:42.660 --> 00:51:44.560 align:middle line:84% certainly, that would help in theory. 00:51:44.560 --> 00:51:46.120 align:middle line:84% In practice, the Fox viewer is going 00:51:46.120 --> 00:51:50.050 align:middle line:84% to throw away the blue pamphlet and keep the red pamphlet, 00:51:50.050 --> 00:51:53.150 align:middle line:84% or interpret the blue pamphlet as supporting their position. 00:51:53.150 --> 00:51:55.597 align:middle line:84% So in some sense, if you had made that suggestion 00:51:55.597 --> 00:51:57.930 align:middle line:84% 20 years ago, I'd be much more optimistic and said yeah, 00:51:57.930 --> 00:51:58.750 align:middle line:90% we should do that. 00:51:58.750 --> 00:52:00.610 align:middle line:84% Now, I don't know if that moves the needle. 00:52:00.610 --> 00:52:02.310 align:middle line:84% AUDIENCE: So then the provision of information 00:52:02.310 --> 00:52:04.602 align:middle line:84% is not going to change because even though you received 00:52:04.602 --> 00:52:09.313 align:middle line:84% information [INAUDIBLE], you can always change and ignore it. 00:52:09.313 --> 00:52:10.730 align:middle line:84% JONATHAN GRUBER: So the question-- 00:52:10.730 --> 00:52:11.450 align:middle line:90% what do we do about that? 00:52:11.450 --> 00:52:12.325 align:middle line:90% That's the challenge. 00:52:12.325 --> 00:52:13.190 align:middle line:90% Yeah, Steven? 00:52:13.190 --> 00:52:14.670 align:middle line:84% AUDIENCE: This I don't think is actually feasible. 00:52:14.670 --> 00:52:15.310 align:middle line:90% And I'm not sure-- 00:52:15.310 --> 00:52:16.310 align:middle line:90% JONATHAN GRUBER: Oh, no. 00:52:16.310 --> 00:52:18.990 align:middle line:90% We're not doing feasible. 00:52:18.990 --> 00:52:22.207 align:middle line:84% AUDIENCE: But mandating a cap on the price of advertisements. 00:52:22.207 --> 00:52:24.790 align:middle line:84% JONATHAN GRUBER: Basically, you could make it less profitable, 00:52:24.790 --> 00:52:26.550 align:middle line:90% in some sense. 00:52:26.550 --> 00:52:29.370 align:middle line:84% But the problem is-- that's an interesting argument. 00:52:29.370 --> 00:52:32.110 align:middle line:84% But in fact, some economists have explored this issue 00:52:32.110 --> 00:52:33.870 align:middle line:90% of whether competition-- 00:52:33.870 --> 00:52:36.930 align:middle line:84% essentially, whether the money to be made is causing this bias. 00:52:36.930 --> 00:52:38.030 align:middle line:90% And it's not clear. 00:52:38.030 --> 00:52:41.608 align:middle line:84% If it's not profitable to run a news network, 00:52:41.608 --> 00:52:44.150 align:middle line:84% if you can't make a profit on a news network, then basically, 00:52:44.150 --> 00:52:45.760 align:middle line:84% it'll be the really crazy biased people 00:52:45.760 --> 00:52:46.970 align:middle line:90% who run news networks, right? 00:52:46.970 --> 00:52:50.170 align:middle line:84% Because that's the only reason they're going to run it then, 00:52:50.170 --> 00:52:51.610 align:middle line:84% is because they have crazy views. 00:52:51.610 --> 00:52:54.930 align:middle line:84% At least a profit motive can keep them a little honest. 00:52:54.930 --> 00:52:56.250 align:middle line:90% Yeah, Alec? 00:52:56.250 --> 00:52:57.810 align:middle line:84% AUDIENCE: There is a level to which 00:52:57.810 --> 00:52:59.977 align:middle line:84% some of the other things you talked about that we do 00:52:59.977 --> 00:53:03.070 align:middle line:84% have solutions to play into the polarization, 00:53:03.070 --> 00:53:06.890 align:middle line:84% like gerrymandering, because you have politicians who 00:53:06.890 --> 00:53:10.850 align:middle line:84% cater to an extreme base and can do that because their district 00:53:10.850 --> 00:53:12.855 align:middle line:90% isn't competitive. 00:53:12.855 --> 00:53:14.230 align:middle line:84% And when you start to solve that, 00:53:14.230 --> 00:53:16.790 align:middle line:84% you start having politicians who can't do that anymore. 00:53:16.790 --> 00:53:20.807 align:middle line:84% And there is an impact that leaders have on public opinion. 00:53:20.807 --> 00:53:22.390 align:middle line:84% JONATHAN GRUBER: I think that's right. 00:53:22.390 --> 00:53:23.930 align:middle line:90% I think this all interacts. 00:53:23.930 --> 00:53:24.810 align:middle line:90% I agree. 00:53:24.810 --> 00:53:27.370 align:middle line:90% And I agree. 00:53:27.370 --> 00:53:31.650 align:middle line:84% I remain concerned that it's hard to-- that this echo chamber 00:53:31.650 --> 00:53:34.090 align:middle line:90% problem kind of dominates. 00:53:34.090 --> 00:53:37.670 align:middle line:84% But the question is, what do we do about this? 00:53:37.670 --> 00:53:39.925 align:middle line:84% And there's actually a very interesting new case, 00:53:39.925 --> 00:53:41.550 align:middle line:84% which may have a big influence on this. 00:53:41.550 --> 00:53:46.080 align:middle line:84% So I don't know how much you guys know about internet law. 00:53:46.080 --> 00:53:48.520 align:middle line:84% But part of the reason the internet flourished was a law 00:53:48.520 --> 00:53:53.360 align:middle line:84% passed a number of years ago, Section 330 or something-- 00:53:53.360 --> 00:53:57.400 align:middle line:84% 230-- which basically says, platforms cannot be held 00:53:57.400 --> 00:53:59.260 align:middle line:90% responsible for what you post. 00:53:59.260 --> 00:54:02.680 align:middle line:84% So basically, if you post some racist Nazi shit in a platform, 00:54:02.680 --> 00:54:05.080 align:middle line:84% it's not their fault. They can't do that. 00:54:05.080 --> 00:54:07.800 align:middle line:84% So people have been unable to deal with the fact 00:54:07.800 --> 00:54:09.980 align:middle line:90% that [INAUDIBLE] online. 00:54:09.980 --> 00:54:12.080 align:middle line:84% But there's an interesting lawsuit. 00:54:12.080 --> 00:54:16.620 align:middle line:84% And the reason is because they are viewed not as media sources, 00:54:16.620 --> 00:54:19.320 align:middle line:84% but as platforms-- as platforms, not responsible for what's 00:54:19.320 --> 00:54:20.160 align:middle line:90% posted. 00:54:20.160 --> 00:54:22.357 align:middle line:84% But a lot of what's driving the problems today 00:54:22.357 --> 00:54:24.440 align:middle line:84% is not what's being posted but the algorithms that 00:54:24.440 --> 00:54:25.600 align:middle line:90% drive where you look. 00:54:25.600 --> 00:54:28.840 align:middle line:84% And the lawsuit says you can sue them over their algorithms. 00:54:28.840 --> 00:54:30.223 align:middle line:90% That's kind of interesting. 00:54:30.223 --> 00:54:31.640 align:middle line:84% Because if a lot of these problems 00:54:31.640 --> 00:54:34.140 align:middle line:84% are caused by algorithms, then you're not a neutral platform 00:54:34.140 --> 00:54:36.880 align:middle line:84% anymore if you have algorithms driving where people go. 00:54:36.880 --> 00:54:38.698 align:middle line:90% And so if we can start to-- 00:54:38.698 --> 00:54:40.240 align:middle line:84% that's the most optimistic thing I've 00:54:40.240 --> 00:54:42.400 align:middle line:84% heard about trying to deal with this, is, 00:54:42.400 --> 00:54:44.900 align:middle line:84% how do we get at those algorithms 00:54:44.900 --> 00:54:46.260 align:middle line:90% that are driving this? 00:54:46.260 --> 00:54:49.080 align:middle line:84% I think more generally, it's, how do we create-- 00:54:49.080 --> 00:54:50.760 align:middle line:84% and here, I want to be more optimistic. 00:54:50.760 --> 00:54:53.343 align:middle line:84% How do folks like you, which are the next generation of elites 00:54:53.343 --> 00:54:55.060 align:middle line:84% and thought leaders in the country, 00:54:55.060 --> 00:54:58.760 align:middle line:84% how do we deal with this and help solve it? 00:54:58.760 --> 00:55:03.600 align:middle line:84% And one way we do that is by recognizing, in some sense, 00:55:03.600 --> 00:55:05.700 align:middle line:84% the underlying danger of social media 00:55:05.700 --> 00:55:07.405 align:middle line:90% and trying to deal with it. 00:55:07.405 --> 00:55:08.780 align:middle line:84% I don't know how many of you guys 00:55:08.780 --> 00:55:10.907 align:middle line:84% have known my favorite new social media study. 00:55:10.907 --> 00:55:13.240 align:middle line:84% Have you guys heard of the FOMO study with social media? 00:55:13.240 --> 00:55:14.657 align:middle line:84% You guys heard about this one yet? 00:55:14.657 --> 00:55:16.620 align:middle line:90% This is the most amazing study. 00:55:16.620 --> 00:55:19.600 align:middle line:84% So a great economist at Chicago named Leo Bursztyn, 00:55:19.600 --> 00:55:22.100 align:middle line:84% he went to university students at about a dozen universities 00:55:22.100 --> 00:55:24.120 align:middle line:84% and said-- actually, I'll ask this question. 00:55:24.120 --> 00:55:24.760 align:middle line:90% Here we go. 00:55:24.760 --> 00:55:28.620 align:middle line:90% 00:55:28.620 --> 00:55:32.060 align:middle line:84% I don't have a clicker, so we can't take the vote. 00:55:32.060 --> 00:55:34.360 align:middle line:84% Imagine I told you-- you're all busy. 00:55:34.360 --> 00:55:35.580 align:middle line:90% You're not typical, right? 00:55:35.580 --> 00:55:38.660 align:middle line:84% But imagine I was willing to pay you 00:55:38.660 --> 00:55:41.380 align:middle line:90% to give up TikTok for a month. 00:55:41.380 --> 00:55:43.970 align:middle line:84% Imagine willing to pay you to give TikTok for a month. 00:55:43.970 --> 00:55:46.902 align:middle line:84% How much would I have to pay you to give up TikTok for a month? 00:55:46.902 --> 00:55:49.110 align:middle line:84% I'm not going to ask, actually, just be embarrassing. 00:55:49.110 --> 00:55:52.870 align:middle line:84% But I've asked people, kids of people, my friends-- 00:55:52.870 --> 00:55:55.170 align:middle line:84% I get numbers in the range of $100. 00:55:55.170 --> 00:55:58.330 align:middle line:84% And the average in the sample is $60. 00:55:58.330 --> 00:56:01.050 align:middle line:84% They had to pay people $60 to give up TikTok for a month. 00:56:01.050 --> 00:56:03.930 align:middle line:84% Then he said, look, I've reached an agreement with 00:56:03.930 --> 00:56:08.910 align:middle line:84% your university that if 2/3 of you agree to it, 00:56:08.910 --> 00:56:14.170 align:middle line:84% the US will ban TikTok for everyone on campus for a month. 00:56:14.170 --> 00:56:17.090 align:middle line:84% How much would I have to pay you to do that? 00:56:17.090 --> 00:56:19.430 align:middle line:84% Or how much would you pay for that? 00:56:19.430 --> 00:56:22.170 align:middle line:84% And the answer is, people said they would pay $40. 00:56:22.170 --> 00:56:23.490 align:middle line:90% Think about that for a second. 00:56:23.490 --> 00:56:26.110 align:middle line:84% They have to be paid to give up their own TikTok account, 00:56:26.110 --> 00:56:28.430 align:middle line:84% but they would pay $40 for everyone to give up, 00:56:28.430 --> 00:56:31.370 align:middle line:84% including themselves, to give up their TikTok account. 00:56:31.370 --> 00:56:33.430 align:middle line:84% That is proof of FOMO right there. 00:56:33.430 --> 00:56:34.370 align:middle line:90% It's perfect evidence. 00:56:34.370 --> 00:56:35.995 align:middle line:84% Basically, the only people using TikTok 00:56:35.995 --> 00:56:38.490 align:middle line:84% is because they're afraid of not. 00:56:38.490 --> 00:56:41.908 align:middle line:84% In other words, TikTok is making us worse off, not better off, 00:56:41.908 --> 00:56:43.700 align:middle line:84% because we'd all be happier if it was gone. 00:56:43.700 --> 00:56:44.800 align:middle line:90% Now, that's extreme. 00:56:44.800 --> 00:56:46.360 align:middle line:84% But the bottom line is, how can we 00:56:46.360 --> 00:56:47.777 align:middle line:84% use studies like that and evidence 00:56:47.777 --> 00:56:49.320 align:middle line:84% like that to really come at the root 00:56:49.320 --> 00:56:50.920 align:middle line:90% source of what's driving this? 00:56:50.920 --> 00:56:55.025 align:middle line:84% And here is where I will put in my pitch for where you can help. 00:56:55.025 --> 00:56:56.400 align:middle line:84% And it comes a bit to the comment 00:56:56.400 --> 00:56:58.640 align:middle line:84% here about sending that information, which 00:56:58.640 --> 00:57:02.360 align:middle line:84% is, the other trend that's gone on that's making this worse 00:57:02.360 --> 00:57:07.420 align:middle line:84% is a view of, to put it in elitist terms, 00:57:07.420 --> 00:57:09.840 align:middle line:84% lack of respect for elites, to put it 00:57:09.840 --> 00:57:13.440 align:middle line:84% in nonelitist terms, a trend towards people 00:57:13.440 --> 00:57:15.120 align:middle line:90% doing their own research. 00:57:15.120 --> 00:57:16.700 align:middle line:90% I did my own research. 00:57:16.700 --> 00:57:17.200 align:middle line:90% OK. 00:57:17.200 --> 00:57:20.040 align:middle line:90% Well, that's just not healthy. 00:57:20.040 --> 00:57:23.360 align:middle line:84% There are experts in things, and experts know stuff, 00:57:23.360 --> 00:57:25.520 align:middle line:90% and that should be respected. 00:57:25.520 --> 00:57:27.142 align:middle line:90% But that's going away. 00:57:27.142 --> 00:57:29.100 align:middle line:84% I don't think that's a controversial statement. 00:57:29.100 --> 00:57:30.933 align:middle line:84% I mean, it's controversial in today's world. 00:57:30.933 --> 00:57:32.480 align:middle line:84% But it's not factually controversial 00:57:32.480 --> 00:57:34.855 align:middle line:84% that some people know more about things than other people 00:57:34.855 --> 00:57:36.560 align:middle line:84% and that we can learn more about things 00:57:36.560 --> 00:57:39.040 align:middle line:90% by consulting those who know. 00:57:39.040 --> 00:57:43.360 align:middle line:84% Why has this lack of respect for expertise exploded? 00:57:43.360 --> 00:57:46.620 align:middle line:84% Well, all these reasons plus one more, 00:57:46.620 --> 00:57:48.900 align:middle line:84% elites separating themselves, intellectuals 00:57:48.900 --> 00:57:51.240 align:middle line:84% separating themselves from communicating with everybody, 00:57:51.240 --> 00:57:53.500 align:middle line:84% talking only to each other in language we only 00:57:53.500 --> 00:57:56.940 align:middle line:84% understand, and not using terminology 00:57:56.940 --> 00:58:00.060 align:middle line:84% and expressing ourselves in ways that other people can 00:58:00.060 --> 00:58:01.000 align:middle line:90% understand. 00:58:01.000 --> 00:58:04.660 align:middle line:84% In fact, the one piece of really good evidence 00:58:04.660 --> 00:58:08.380 align:middle line:84% of how you can move people is through personal connections. 00:58:08.380 --> 00:58:10.980 align:middle line:84% The one piece of evidence on how you change people's opinions 00:58:10.980 --> 00:58:14.200 align:middle line:84% is, when people have friends and connections from the other side, 00:58:14.200 --> 00:58:16.660 align:middle line:90% you can move their opinions. 00:58:16.660 --> 00:58:20.260 align:middle line:84% But when experts stand up and just say, look, 00:58:20.260 --> 00:58:23.260 align:middle line:84% you're wrong-- one of my favorite-- you guys 00:58:23.260 --> 00:58:25.920 align:middle line:84% know The Onion-- favorite Onion articles ever, 00:58:25.920 --> 00:58:28.593 align:middle line:84% it was a fake editorial saying, you know, 00:58:28.593 --> 00:58:30.260 align:middle line:84% I totally changed my views when that guy 00:58:30.260 --> 00:58:32.380 align:middle line:90% called me an ignorant asshole. 00:58:32.380 --> 00:58:35.880 align:middle line:84% And it's like, no, that's not going to change anybody's views. 00:58:35.880 --> 00:58:37.620 align:middle line:84% And you may feel that way, and you're 00:58:37.620 --> 00:58:39.130 align:middle line:90% entitled to your feelings. 00:58:39.130 --> 00:58:43.830 align:middle line:84% But the bottom line is, we need to engage with where people are. 00:58:43.830 --> 00:58:48.570 align:middle line:84% And we, as the experts, need to go and engage and build 00:58:48.570 --> 00:58:49.857 align:middle line:90% our communication skills. 00:58:49.857 --> 00:58:51.690 align:middle line:84% I realize MIT is not a communication-focused 00:58:51.690 --> 00:58:54.065 align:middle line:84% university, and you all hate your communications courses. 00:58:54.065 --> 00:58:55.230 align:middle line:90% I understand that. 00:58:55.230 --> 00:58:58.510 align:middle line:84% But it's important that we don't just talk to each other 00:58:58.510 --> 00:59:01.570 align:middle line:84% but that we can talk in terms that everyone understands 00:59:01.570 --> 00:59:04.410 align:middle line:84% and that we can basically-- because that's how we actually 00:59:04.410 --> 00:59:05.530 align:middle line:90% change things. 00:59:05.530 --> 00:59:08.330 align:middle line:84% So my message to you is to urge you, 00:59:08.330 --> 00:59:09.938 align:middle line:84% as you become an expert in something, 00:59:09.938 --> 00:59:11.730 align:middle line:84% to not buy into the bullshit that expertise 00:59:11.730 --> 00:59:14.070 align:middle line:84% is invaluable-- expertise is incredibly valuable-- 00:59:14.070 --> 00:59:15.390 align:middle line:84% but to recognize people have that view 00:59:15.390 --> 00:59:16.890 align:middle line:84% because we're not good at expressing 00:59:16.890 --> 00:59:19.530 align:middle line:84% our expertise in understandable and humane ways. 00:59:19.530 --> 00:59:22.208 align:middle line:84% And that's, I think, the place we can make a difference, 00:59:22.208 --> 00:59:23.750 align:middle line:84% and I'm hoping a positive difference. 00:59:23.750 --> 00:59:26.630 align:middle line:84% Hoping that someday, if it's done in that way-- 00:59:26.630 --> 00:59:27.910 align:middle line:90% oh, here they come for me. 00:59:27.910 --> 00:59:29.190 align:middle line:90% [LAUGHTER] 00:59:29.190 --> 00:59:31.290 align:middle line:84% Someday, if it's done in that way, 00:59:31.290 --> 00:59:33.210 align:middle line:84% that you can get those red and blue pamphlets, 00:59:33.210 --> 00:59:34.770 align:middle line:90% and you will pay attention. 00:59:34.770 --> 00:59:38.660 align:middle line:84% So that's what I want to say about polarization. 00:59:38.660 --> 00:59:40.800 align:middle line:90% Any other questions about that? 00:59:40.800 --> 00:59:44.840 align:middle line:90% Let me come to one last topic. 00:59:44.840 --> 00:59:46.800 align:middle line:84% Let's go even further, which is-- 00:59:46.800 --> 00:59:52.642 align:middle line:84% I've talked about a system of unanimous consent. 00:59:52.642 --> 00:59:54.100 align:middle line:84% I've talked about direct democracy. 00:59:54.100 --> 00:59:55.300 align:middle line:84% I've talked representative democracy. 00:59:55.300 --> 00:59:56.592 align:middle line:90% I've talked about polarization. 00:59:56.592 --> 00:59:59.240 align:middle line:84% But I have yet to talk about a system 00:59:59.240 --> 01:00:02.960 align:middle line:84% where politicians don't actually care at all about what's 01:00:02.960 --> 01:00:04.920 align:middle line:90% good for the country. 01:00:04.920 --> 01:00:07.280 align:middle line:84% I've talked about a system where politicians may not 01:00:07.280 --> 01:00:10.810 align:middle line:84% do what's best for the country in aggregate or may bias them, 01:00:10.810 --> 01:00:12.560 align:middle line:84% but I still assume politicians are kind of 01:00:12.560 --> 01:00:13.800 align:middle line:90% trying to do what's right. 01:00:13.800 --> 01:00:15.360 align:middle line:90% And the whole course does. 01:00:15.360 --> 01:00:17.020 align:middle line:84% But in fact, that may not be true. 01:00:17.020 --> 01:00:20.640 align:middle line:84% And that leads to a whole area of corruption, which 01:00:20.640 --> 01:00:24.160 align:middle line:84% is basically a fundamental challenge to, 01:00:24.160 --> 01:00:27.120 align:middle line:84% how do we think about modeling government 01:00:27.120 --> 01:00:30.183 align:middle line:84% and designing policy in a world where, 01:00:30.183 --> 01:00:31.600 align:middle line:84% really, the goal of the government 01:00:31.600 --> 01:00:34.100 align:middle line:84% is not at all to do social welfare maximization, 01:00:34.100 --> 01:00:37.260 align:middle line:84% but just to maximize their own power and wealth? 01:00:37.260 --> 01:00:40.640 align:middle line:84% Now, this is obviously a huge problem-- 01:00:40.640 --> 01:00:42.260 align:middle line:84% to be fair, worse around the world 01:00:42.260 --> 01:00:45.040 align:middle line:84% than it is here, especially in developing countries. 01:00:45.040 --> 01:00:48.860 align:middle line:84% So as fun examples, the former prime minister of Malaysia 01:00:48.860 --> 01:00:54.780 align:middle line:84% was arrested in July 2018 because, since 2015, the US 01:00:54.780 --> 01:00:57.340 align:middle line:84% government had been investigating $4.5 billion 01:00:57.340 --> 01:00:58.980 align:middle line:84% stolen from the Malaysian government 01:00:58.980 --> 01:01:01.900 align:middle line:84% and laundered through US corporations. 01:01:01.900 --> 01:01:04.040 align:middle line:84% More than $1 billion of that was spent on US goods, 01:01:04.040 --> 01:01:07.250 align:middle line:84% including a 22-carat pink diamond necklace that cost $27 01:01:07.250 --> 01:01:12.220 align:middle line:84% million, necklaces that cost $1.3 million, 01:01:12.220 --> 01:01:17.220 align:middle line:84% the production of the movie The Wolf of Wall Street, 01:01:17.220 --> 01:01:20.340 align:middle line:84% and his son's lavish parties, including $12 million in gifts 01:01:20.340 --> 01:01:23.060 align:middle line:90% of art to Leo DiCaprio. 01:01:23.060 --> 01:01:25.580 align:middle line:90% So that is one example. 01:01:25.580 --> 01:01:27.740 align:middle line:90% Now, the US is not alone. 01:01:27.740 --> 01:01:31.320 align:middle line:84% Now, probably the king of corruption in the world-- 01:01:31.320 --> 01:01:32.700 align:middle line:90% I think the title-- 01:01:32.700 --> 01:01:37.230 align:middle line:84% probably goes to Joseph Mobutu of the Congo, 01:01:37.230 --> 01:01:41.930 align:middle line:84% which was renamed Zaire and then renamed back to Congo. 01:01:41.930 --> 01:01:47.410 align:middle line:84% Between 1967 and 1997, when he was chased out of power, 01:01:47.410 --> 01:01:50.970 align:middle line:84% the entire worldwide incomes grew over that period. 01:01:50.970 --> 01:01:53.730 align:middle line:84% In real terms, they probably doubled or tripled. 01:01:53.730 --> 01:01:57.587 align:middle line:84% In the Congo, income per capita fell by 2/3. 01:01:57.587 --> 01:02:00.170 align:middle line:84% Over a 30-year period while the rest of the world was becoming 01:02:00.170 --> 01:02:03.427 align:middle line:84% two to three times richer, income per capita fell by 2/3. 01:02:03.427 --> 01:02:05.510 align:middle line:84% And he became one of the richest men in the world, 01:02:05.510 --> 01:02:08.530 align:middle line:84% with wealth of more than $5 billion. 01:02:08.530 --> 01:02:11.233 align:middle line:84% Now, in a situation like that, how can you 01:02:11.233 --> 01:02:12.650 align:middle line:84% think they're doing anything we're 01:02:12.650 --> 01:02:14.353 align:middle line:90% talking about in this class? 01:02:14.353 --> 01:02:16.770 align:middle line:84% They're not worrying about optimal public goods provision. 01:02:16.770 --> 01:02:20.010 align:middle line:84% They're worrying optimal yacht provision to them. 01:02:20.010 --> 01:02:23.823 align:middle line:90% Now, we're no exempt from this. 01:02:23.823 --> 01:02:26.490 align:middle line:84% And one of the fun examples is-- we have any folks from Illinois 01:02:26.490 --> 01:02:27.810 align:middle line:90% here? 01:02:27.810 --> 01:02:28.780 align:middle line:90% No Illinois folks, OK. 01:02:28.780 --> 01:02:30.530 align:middle line:84% Well, Illinois is the great-- maybe you're 01:02:30.530 --> 01:02:33.310 align:middle line:84% afraid to admit it-- corruption capital. 01:02:33.310 --> 01:02:35.990 align:middle line:90% George Ryan was the governor. 01:02:35.990 --> 01:02:38.830 align:middle line:84% He was indicted by a federal grand jury 01:02:38.830 --> 01:02:43.430 align:middle line:84% because it turned out that when he was secretary of state, 01:02:43.430 --> 01:02:45.630 align:middle line:84% he allowed people to buy their way into driver's m 01:02:45.630 --> 01:02:47.430 align:middle line:90% even if they weren't qualified. 01:02:47.430 --> 01:02:49.190 align:middle line:84% And result-- at least 20 people died 01:02:49.190 --> 01:02:51.890 align:middle line:84% in accidents involving drivers who were unqualified 01:02:51.890 --> 01:02:54.310 align:middle line:90% who bought their licenses. 01:02:54.310 --> 01:02:57.070 align:middle line:84% And so he was sent to jail and replaced 01:02:57.070 --> 01:03:00.230 align:middle line:84% by someone who campaigned on clean government, a man whose 01:03:00.230 --> 01:03:02.590 align:middle line:84% name I heard, named Rod Blagojevich. 01:03:02.590 --> 01:03:04.290 align:middle line:84% Rod Blagojevich, you might say, God, 01:03:04.290 --> 01:03:06.910 align:middle line:84% that name sounds maybe a little familiar when I was a kid. 01:03:06.910 --> 01:03:11.150 align:middle line:84% He went to jail because he auctioned off 01:03:11.150 --> 01:03:12.990 align:middle line:90% Barack Obama's Senate seat. 01:03:12.990 --> 01:03:16.370 align:middle line:84% So Barack Obama went to president, ran for president. 01:03:16.370 --> 01:03:17.990 align:middle line:84% His Senate seat became open, and there 01:03:17.990 --> 01:03:19.198 align:middle line:90% should be an election for it. 01:03:19.198 --> 01:03:21.903 align:middle line:84% Blagojevich secretly took bribes to decide who would-- 01:03:21.903 --> 01:03:23.570 align:middle line:84% the governor got to appoint the senator. 01:03:23.570 --> 01:03:25.990 align:middle line:84% He took bribes to decide who would get that seat. 01:03:25.990 --> 01:03:27.670 align:middle line:90% And so he went down. 01:03:27.670 --> 01:03:30.790 align:middle line:90% So corruption is everywhere. 01:03:30.790 --> 01:03:32.460 align:middle line:90% It's endemic. 01:03:32.460 --> 01:03:34.000 align:middle line:90% So what do we think about this? 01:03:34.000 --> 01:03:36.480 align:middle line:84% How do we think about the role of corruption? 01:03:36.480 --> 01:03:38.900 align:middle line:84% Well, there's a couple points I want to make. 01:03:38.900 --> 01:03:43.700 align:middle line:84% Point one is, we should not have a black-and-white view 01:03:43.700 --> 01:03:45.900 align:middle line:90% of thinking about corruption. 01:03:45.900 --> 01:03:47.500 align:middle line:84% That is, we should never have the view 01:03:47.500 --> 01:03:50.193 align:middle line:84% that any corruption means the whole system's broken. 01:03:50.193 --> 01:03:51.860 align:middle line:84% This is a problem we have in US politics 01:03:51.860 --> 01:03:56.420 align:middle line:84% today, which is that any example of corruption 01:03:56.420 --> 01:04:02.820 align:middle line:84% causes a uniform loss of belief in the whole system. 01:04:02.820 --> 01:04:07.740 align:middle line:84% And that's a problem, that basically, people 01:04:07.740 --> 01:04:11.500 align:middle line:84% lose their faith in the system at even a hint of corruption. 01:04:11.500 --> 01:04:14.500 align:middle line:84% We're going to talk in the next section of the course 01:04:14.500 --> 01:04:17.920 align:middle line:84% all about public programs like Social Security, Medicare, 01:04:17.920 --> 01:04:19.340 align:middle line:90% unemployment insurance. 01:04:19.340 --> 01:04:21.140 align:middle line:84% There has never been a public program 01:04:21.140 --> 01:04:23.460 align:middle line:84% in the history of the world in any country that 01:04:23.460 --> 01:04:25.340 align:middle line:90% has not had some fraud-- 01:04:25.340 --> 01:04:26.580 align:middle line:90% never. 01:04:26.580 --> 01:04:29.000 align:middle line:84% Does that mean they shouldn't exist? 01:04:29.000 --> 01:04:29.500 align:middle line:90% No. 01:04:29.500 --> 01:04:31.250 align:middle line:84% It just means you got to balance the fraud 01:04:31.250 --> 01:04:32.485 align:middle line:90% against the good they do. 01:04:32.485 --> 01:04:35.110 align:middle line:84% Likewise, the fact that there's some corruption in the system-- 01:04:35.110 --> 01:04:36.450 align:middle line:84% does that mean the system is fundamentally broken 01:04:36.450 --> 01:04:37.810 align:middle line:90% and has to be ripped apart? 01:04:37.810 --> 01:04:39.730 align:middle line:84% Only if you've got a better system with less corruption 01:04:39.730 --> 01:04:40.438 align:middle line:90% should you do it. 01:04:40.438 --> 01:04:42.610 align:middle line:84% It may be that a certain amount of corruption's 01:04:42.610 --> 01:04:44.650 align:middle line:90% inevitable in a system. 01:04:44.650 --> 01:04:46.697 align:middle line:84% And one should not think that any corruption-- 01:04:46.697 --> 01:04:48.530 align:middle line:84% it's sort of like we talked about pollution, 01:04:48.530 --> 01:04:50.710 align:middle line:84% said that optimal pollution is never zero. 01:04:50.710 --> 01:04:53.030 align:middle line:84% I'm not saying optimal corruption's never zero. 01:04:53.030 --> 01:04:55.850 align:middle line:84% I'm just saying realistic corruption is never zero. 01:04:55.850 --> 01:04:57.997 align:middle line:84% And so the first point is, we cannot get in this 01:04:57.997 --> 01:05:00.330 align:middle line:84% black-and-white system of thinking like people got about 01:05:00.330 --> 01:05:02.650 align:middle line:84% pollution, that any means the whole thing needs to be torn 01:05:02.650 --> 01:05:03.150 align:middle line:90% down. 01:05:03.150 --> 01:05:03.890 align:middle line:90% That's not true. 01:05:03.890 --> 01:05:05.845 align:middle line:84% We have to think about trading off 01:05:05.845 --> 01:05:07.970 align:middle line:84% how bad the corruption is versus the other features 01:05:07.970 --> 01:05:10.270 align:middle line:90% of the system. 01:05:10.270 --> 01:05:12.250 align:middle line:90% So that's point 1. 01:05:12.250 --> 01:05:16.610 align:middle line:84% Point 2 is related to that, which 01:05:16.610 --> 01:05:19.750 align:middle line:84% is, corruption can lead to a dangerous cycle. 01:05:19.750 --> 01:05:21.802 align:middle line:84% Now, I've traveled all over the world. 01:05:21.802 --> 01:05:24.010 align:middle line:84% And whenever I'm in a country, I try to talk politics 01:05:24.010 --> 01:05:25.810 align:middle line:90% with the local folk. 01:05:25.810 --> 01:05:29.170 align:middle line:84% And in most countries in the world, 01:05:29.170 --> 01:05:31.690 align:middle line:84% in every developing country I've been to, you always get-- 01:05:31.690 --> 01:05:33.610 align:middle line:84% I'll say, OK, who are your political parties? 01:05:33.610 --> 01:05:34.860 align:middle line:90% Who are you going to vote for? 01:05:34.860 --> 01:05:37.857 align:middle line:84% You always get the same answer, everyone I ask. 01:05:37.857 --> 01:05:38.690 align:middle line:90% They're all corrupt. 01:05:38.690 --> 01:05:39.732 align:middle line:90% It doesn't matter anyway. 01:05:39.732 --> 01:05:40.950 align:middle line:90% I'm not going to vote. 01:05:40.950 --> 01:05:44.270 align:middle line:84% Everywhere, every developing country gives the same answer. 01:05:44.270 --> 01:05:46.190 align:middle line:90% Think about what that means. 01:05:46.190 --> 01:05:51.950 align:middle line:84% That means that corruption can create a self-reinforcing cycle. 01:05:51.950 --> 01:05:54.270 align:middle line:84% And the reason is my third point, 01:05:54.270 --> 01:05:58.550 align:middle line:84% which is that the best barrier to corruption 01:05:58.550 --> 01:06:02.450 align:middle line:84% is democracy, that basically, the evidence exists 01:06:02.450 --> 01:06:04.510 align:middle line:84% that the more democratic a society, 01:06:04.510 --> 01:06:07.055 align:middle line:90% the less corruption there is. 01:06:07.055 --> 01:06:08.430 align:middle line:84% There's a number of studies which 01:06:08.430 --> 01:06:10.770 align:middle line:84% show that increasing democratic institutions, 01:06:10.770 --> 01:06:15.630 align:middle line:84% the democratic supports, leads to a fallen corruption. 01:06:15.630 --> 01:06:19.390 align:middle line:84% So this means that we have a possibility of a vicious cycle, 01:06:19.390 --> 01:06:22.310 align:middle line:84% where corrupt politicians cause people to lose faith 01:06:22.310 --> 01:06:25.270 align:middle line:84% in government, cause them to not participate, 01:06:25.270 --> 01:06:27.110 align:middle line:90% which allows more corruption. 01:06:27.110 --> 01:06:29.620 align:middle line:84% And it continues 'til you're Zaire. 01:06:29.620 --> 01:06:31.397 align:middle line:84% And that is the risk with corruption 01:06:31.397 --> 01:06:32.980 align:middle line:84% that you move to-- we talk in economic 01:06:32.980 --> 01:06:34.105 align:middle line:90% about multiple equilibrium. 01:06:34.105 --> 01:06:35.700 align:middle line:84% We'll talk about some macroeconomics, 01:06:35.700 --> 01:06:38.260 align:middle line:84% different equilibriums that are equally likely possible, 01:06:38.260 --> 01:06:40.300 align:middle line:84% depending on random things happening. 01:06:40.300 --> 01:06:42.620 align:middle line:84% Well, we are in the US, have traditionally 01:06:42.620 --> 01:06:44.540 align:middle line:84% been a very good equilibrium, where 01:06:44.540 --> 01:06:47.147 align:middle line:84% we have relatively little corruption and relatively high 01:06:47.147 --> 01:06:49.480 align:middle line:84% participation in democracy-- not as high as it could be. 01:06:49.480 --> 01:06:51.355 align:middle line:84% We're not as good as Europe, but we're better 01:06:51.355 --> 01:06:52.700 align:middle line:90% than developing countries. 01:06:52.700 --> 01:06:57.460 align:middle line:84% That can go away if a corrupt president erodes the faith 01:06:57.460 --> 01:06:58.960 align:middle line:90% in the whole political system. 01:06:58.960 --> 01:07:01.180 align:middle line:84% People just throw up their hands and say, forget it. 01:07:01.180 --> 01:07:03.100 align:middle line:90% I'm not interested. 01:07:03.100 --> 01:07:05.560 align:middle line:84% And that is the other message on my soapbox 01:07:05.560 --> 01:07:08.580 align:middle line:84% I'm delivering to you all, which is, 01:07:08.580 --> 01:07:11.580 align:middle line:84% many of my friends on my side of the political spectrum 01:07:11.580 --> 01:07:13.340 align:middle line:84% got very depressed after 2016 and said, 01:07:13.340 --> 01:07:14.160 align:middle line:90% oh, the whole thing's broken. 01:07:14.160 --> 01:07:15.368 align:middle line:90% I'm not going to participate. 01:07:15.368 --> 01:07:16.980 align:middle line:90% And that's wrong. 01:07:16.980 --> 01:07:19.540 align:middle line:84% It's actually the wrong reaction. 01:07:19.540 --> 01:07:21.760 align:middle line:84% If the side you don't like is winning, 01:07:21.760 --> 01:07:23.720 align:middle line:90% the reaction is not to drop out. 01:07:23.720 --> 01:07:25.600 align:middle line:90% Participate more. 01:07:25.600 --> 01:07:28.720 align:middle line:84% But the problem is, it can feel it 01:07:28.720 --> 01:07:30.038 align:middle line:90% can feel like an uphill battle. 01:07:30.038 --> 01:07:31.580 align:middle line:84% It can feel like Sisyphus doing that. 01:07:31.580 --> 01:07:33.240 align:middle line:84% And that situation, many developing countries 01:07:33.240 --> 01:07:33.900 align:middle line:90% find them in. 01:07:33.900 --> 01:07:36.067 align:middle line:84% And it's a risk that any country can find themselves 01:07:36.067 --> 01:07:37.140 align:middle line:90% in that situation. 01:07:37.140 --> 01:07:37.640 align:middle line:90% OK. 01:07:37.640 --> 01:07:38.360 align:middle line:90% Yeah? 01:07:38.360 --> 01:07:40.520 align:middle line:90% AUDIENCE: Quick comment. 01:07:40.520 --> 01:07:42.960 align:middle line:84% You mentioned that democracy is associated 01:07:42.960 --> 01:07:44.200 align:middle line:90% with lower corruption. 01:07:44.200 --> 01:07:46.760 align:middle line:84% How about democracy as well correlating 01:07:46.760 --> 01:07:48.287 align:middle line:90% with economic development? 01:07:48.287 --> 01:07:50.120 align:middle line:84% JONATHAN GRUBER: Yeah, they've done studies. 01:07:50.120 --> 01:07:51.380 align:middle line:90% It's a great point. 01:07:51.380 --> 01:07:52.380 align:middle line:90% It's a great point. 01:07:52.380 --> 01:07:53.255 align:middle line:90% AUDIENCE: [INAUDIBLE] 01:07:53.255 --> 01:07:56.080 align:middle line:84% JONATHAN GRUBER: It's a bias problem, right? 01:07:56.080 --> 01:07:59.260 align:middle line:84% Democracy also correlates with-- now, 01:07:59.260 --> 01:08:01.260 align:middle line:84% democracy also correlates with economic success, 01:08:01.260 --> 01:08:01.940 align:middle line:90% say, to separate them. 01:08:01.940 --> 01:08:04.160 align:middle line:84% But there's various clever ways they've done that. 01:08:04.160 --> 01:08:06.150 align:middle line:84% It's not a pure-- it's not a randomized trial. 01:08:06.150 --> 01:08:08.400 align:middle line:84% But generally, the view is, from the bulk of evidence, 01:08:08.400 --> 01:08:11.200 align:middle line:84% that democracy is correlated with lower corruption. 01:08:11.200 --> 01:08:11.920 align:middle line:90% AUDIENCE: OK. 01:08:11.920 --> 01:08:15.020 align:middle line:84% I guess my other comment was, there are outliers in this. 01:08:15.020 --> 01:08:18.240 align:middle line:84% For example, China is not really a democracy. 01:08:18.240 --> 01:08:22.859 align:middle line:84% But I guess, over time, given what some the leaders are doing, 01:08:22.859 --> 01:08:25.390 align:middle line:84% corruption may be also going down in the country. 01:08:25.390 --> 01:08:26.390 align:middle line:84% JONATHAN GRUBER: Once again, we don't know. 01:08:26.390 --> 01:08:27.050 align:middle line:90% That's the problem. 01:08:27.050 --> 01:08:29.217 align:middle line:84% We don't know how much corruption there is in China. 01:08:29.217 --> 01:08:31.430 align:middle line:90% It could be enormous. 01:08:31.430 --> 01:08:33.955 align:middle line:84% We've got no-- I mean, look at how little Putin 01:08:33.955 --> 01:08:35.830 align:middle line:84% knew about the corruption in his own military 01:08:35.830 --> 01:08:37.590 align:middle line:84% and the decrepit state of his own military 01:08:37.590 --> 01:08:38.750 align:middle line:90% before he went to war. 01:08:38.750 --> 01:08:40.208 align:middle line:84% He thought he had a great military. 01:08:40.208 --> 01:08:42.350 align:middle line:84% Turns out, he doesn't, because it's all corrupt. 01:08:42.350 --> 01:08:43.910 align:middle line:84% So the problem with a nondemocracy 01:08:43.910 --> 01:08:45.750 align:middle line:84% is, you don't have the institutions that 01:08:45.750 --> 01:08:47.790 align:middle line:84% let you know how much corruption there is. 01:08:47.790 --> 01:08:49.850 align:middle line:84% So as a result, it could be self-reinforcing. 01:08:49.850 --> 01:08:51.683 align:middle line:84% I don't know how much corruption's in China. 01:08:51.683 --> 01:08:52.950 align:middle line:90% I'm not sure anybody does. 01:08:52.950 --> 01:08:57.069 align:middle line:84% So that's the other point I want to make. 01:08:57.069 --> 01:09:02.229 align:middle line:84% Finally, let me come to the meta point of this chapter. 01:09:02.229 --> 01:09:03.750 align:middle line:84% The meta point of this chapter is 01:09:03.750 --> 01:09:06.149 align:middle line:84% to go back to the four questions of public finance, 01:09:06.149 --> 01:09:10.229 align:middle line:84% and to remember that just because the private market fails 01:09:10.229 --> 01:09:13.630 align:middle line:84% doesn't mean the government will succeed. 01:09:13.630 --> 01:09:16.430 align:middle line:84% What we technically are saying when we prove private market 01:09:16.430 --> 01:09:19.390 align:middle line:84% failure, like an externality or a public good, 01:09:19.390 --> 01:09:22.990 align:middle line:84% we're saying there is the scope for the government 01:09:22.990 --> 01:09:25.538 align:middle line:84% to improve things but not the, necessarily, will. 01:09:25.538 --> 01:09:27.080 align:middle line:84% And I'll be sloppy on that discourse. 01:09:27.080 --> 01:09:28.220 align:middle line:84% I'll sometimes slip into language 01:09:28.220 --> 01:09:29.571 align:middle line:90% like, OK, the market's broken. 01:09:29.571 --> 01:09:30.779 align:middle line:90% The government should fix it. 01:09:30.779 --> 01:09:32.580 align:middle line:90% But the government might not. 01:09:32.580 --> 01:09:34.160 align:middle line:84% Just like there's market failure, 01:09:34.160 --> 01:09:36.460 align:middle line:90% there is government failure. 01:09:36.460 --> 01:09:39.460 align:middle line:84% And so the key thing to remember, 01:09:39.460 --> 01:09:43.740 align:middle line:84% even if I'm sloppy with my language, is that, 01:09:43.740 --> 01:09:47.240 align:middle line:84% basically, when we prove market failure, 01:09:47.240 --> 01:09:49.620 align:middle line:84% we're not proving the government can make it better. 01:09:49.620 --> 01:09:51.370 align:middle line:84% We're proving the government has the scope 01:09:51.370 --> 01:09:52.979 align:middle line:90% to potentially make it better. 01:09:52.979 --> 01:09:54.646 align:middle line:84% When the government will make it better, 01:09:54.646 --> 01:09:57.180 align:middle line:84% it depends on the scope of government failure. 01:09:57.180 --> 01:10:00.565 align:middle line:84% And in particular, let me end with one 01:10:00.565 --> 01:10:02.940 align:middle line:84% important point about that, which comes, as I mentioned-- 01:10:02.940 --> 01:10:05.190 align:middle line:84% I realized, like I said, we only get a lecture in half 01:10:05.190 --> 01:10:06.815 align:middle line:84% on political economy, which is too bad. 01:10:06.815 --> 01:10:07.607 align:middle line:90% It's a great topic. 01:10:07.607 --> 01:10:10.190 align:middle line:84% But we have maybe the world's most famous political economist, 01:10:10.190 --> 01:10:11.440 align:middle line:90% Daron Acemoglu, teaching here. 01:10:11.440 --> 01:10:12.900 align:middle line:84% And he has a famous book called Why 01:10:12.900 --> 01:10:16.980 align:middle line:84% Nations Fail, which is all about the role 01:10:16.980 --> 01:10:21.580 align:middle line:84% that institutions play in determining the success 01:10:21.580 --> 01:10:25.080 align:middle line:84% or failure of countries, that it's not just about whether you 01:10:25.080 --> 01:10:27.360 align:middle line:84% got some corrupt guy in charge or not, 01:10:27.360 --> 01:10:30.020 align:middle line:84% but basically, how robust the institutions are. 01:10:30.020 --> 01:10:31.640 align:middle line:84% And there's no way to see that more 01:10:31.640 --> 01:10:33.800 align:middle line:90% than look at the 2020 election. 01:10:33.800 --> 01:10:38.200 align:middle line:84% The US electoral system held up under enormous pressure from one 01:10:38.200 --> 01:10:40.760 align:middle line:90% side to change the results. 01:10:40.760 --> 01:10:42.200 align:middle line:90% It held up. 01:10:42.200 --> 01:10:43.705 align:middle line:90% Institutions matter. 01:10:43.705 --> 01:10:45.580 align:middle line:84% So there's a longstanding debate about, well, 01:10:45.580 --> 01:10:47.260 align:middle line:84% is it institutions that matter or people that matter? 01:10:47.260 --> 01:10:48.100 align:middle line:90% What really matters? 01:10:48.100 --> 01:10:51.040 align:middle line:84% And Daron's lifeblood has been a series 01:10:51.040 --> 01:10:54.280 align:middle line:84% of articles showing how critical institutions are, the underlying 01:10:54.280 --> 01:10:55.780 align:middle line:84% institutions of a country's success. 01:10:55.780 --> 01:10:58.840 align:middle line:84% And maybe his most famous article is a super cool article 01:10:58.840 --> 01:11:04.160 align:middle line:84% he did, where he and his co-authors noted that, 01:11:04.160 --> 01:11:09.960 align:middle line:84% basically, there were two sets, two types of countries colonized 01:11:09.960 --> 01:11:14.560 align:middle line:84% by the great powers back in the 1600s and 1700s. 01:11:14.560 --> 01:11:18.360 align:middle line:84% And basically, [MIC INTERFERENCE] 01:11:18.360 --> 01:11:21.568 align:middle line:84% some countries were [MIC INTERFERENCE]. 01:11:21.568 --> 01:11:27.990 align:middle line:90% 01:11:27.990 --> 01:11:31.630 align:middle line:84% Some countries were colonized from afar, countries 01:11:31.630 --> 01:11:33.310 align:middle line:90% in the Caribbean and Africa. 01:11:33.310 --> 01:11:35.690 align:middle line:84% The Europeans generally didn't go live there. 01:11:35.690 --> 01:11:38.590 align:middle line:84% They set up proxy governments and went back to live in Europe. 01:11:38.590 --> 01:11:42.790 align:middle line:84% Other countries, like in South America or Australia, 01:11:42.790 --> 01:11:45.550 align:middle line:84% the Europeans went and lived there. 01:11:45.550 --> 01:11:48.170 align:middle line:84% It turns out that if you look today-- 01:11:48.170 --> 01:11:50.450 align:middle line:84% and it turns out, from the perspective of that time, 01:11:50.450 --> 01:11:54.350 align:middle line:84% those countries were all about pretty comparable economically. 01:11:54.350 --> 01:11:57.892 align:middle line:84% Today, there's an incredible correlation 01:11:57.892 --> 01:11:59.350 align:middle line:84% between whether they were countries 01:11:59.350 --> 01:12:02.110 align:middle line:84% where the Europeans actually went and lived 01:12:02.110 --> 01:12:03.030 align:middle line:90% and economic outcomes. 01:12:03.030 --> 01:12:05.110 align:middle line:84% In particular, the ones where Europeans went and colonized 01:12:05.110 --> 01:12:07.750 align:middle line:84% and lived there had much better economic outcomes than the one 01:12:07.750 --> 01:12:09.772 align:middle line:90% they ruled from afar. 01:12:09.772 --> 01:12:11.230 align:middle line:84% They might say, well, I don't know. 01:12:11.230 --> 01:12:12.470 align:middle line:84% Maybe they want to live there because things 01:12:12.470 --> 01:12:13.220 align:middle line:90% about the country. 01:12:13.220 --> 01:12:14.678 align:middle line:84% So how did they actually test that? 01:12:14.678 --> 01:12:16.750 align:middle line:84% Did a really cool thing-- was they said, look, 01:12:16.750 --> 01:12:19.300 align:middle line:84% what's the best predictor of whether Europeans 01:12:19.300 --> 01:12:20.420 align:middle line:90% would live there? 01:12:20.420 --> 01:12:23.460 align:middle line:84% The incidence of malaria, that basically, Europeans 01:12:23.460 --> 01:12:26.503 align:middle line:84% didn't want to go live where there was malaria. 01:12:26.503 --> 01:12:27.920 align:middle line:84% Locals aren't affected by malaria. 01:12:27.920 --> 01:12:29.042 align:middle line:90% They were immune. 01:12:29.042 --> 01:12:30.500 align:middle line:84% So it's basically whether Europeans 01:12:30.500 --> 01:12:31.600 align:middle line:90% would get sick by living there. 01:12:31.600 --> 01:12:32.980 align:middle line:84% And they found that places where you 01:12:32.980 --> 01:12:35.100 align:middle line:84% get more sick by living there, the Europeans were 01:12:35.100 --> 01:12:36.240 align:middle line:90% less likely to live there. 01:12:36.240 --> 01:12:37.160 align:middle line:90% What did they do? 01:12:37.160 --> 01:12:39.180 align:middle line:84% They did not set up strong institutions 01:12:39.180 --> 01:12:40.000 align:middle line:90% because they didn't live there. 01:12:40.000 --> 01:12:40.480 align:middle line:90% They didn't care. 01:12:40.480 --> 01:12:41.880 align:middle line:84% They were just extracting resources. 01:12:41.880 --> 01:12:43.240 align:middle line:84% They didn't care if there was crime. 01:12:43.240 --> 01:12:45.080 align:middle line:84% They didn't care what life was like there, 01:12:45.080 --> 01:12:45.880 align:middle line:90% how awful life was there. 01:12:45.880 --> 01:12:46.960 align:middle line:90% They said, give me my diamonds. 01:12:46.960 --> 01:12:47.660 align:middle line:90% I'm out. 01:12:47.660 --> 01:12:49.120 align:middle line:84% Whereas when they lived there, they cared. 01:12:49.120 --> 01:12:50.745 align:middle line:84% They didn't want to be victim to crime. 01:12:50.745 --> 01:12:52.340 align:middle line:90% They wanted good schools. 01:12:52.340 --> 01:12:53.820 align:middle line:84% So essentially, where there wasn't 01:12:53.820 --> 01:12:55.560 align:middle line:84% a risk of malaria, a low risk of malaria, 01:12:55.560 --> 01:12:57.360 align:middle line:84% Europeans actually settled there, 01:12:57.360 --> 01:13:01.420 align:middle line:84% created good institutions, and those countries grew rapidly. 01:13:01.420 --> 01:13:03.340 align:middle line:84% Where there was a risk of malaria-- 01:13:03.340 --> 01:13:04.960 align:middle line:90% now, it's not one to one. 01:13:04.960 --> 01:13:06.480 align:middle line:90% It's a relationship. 01:13:06.480 --> 01:13:08.140 align:middle line:84% But as the risk of malaria went up, 01:13:08.140 --> 01:13:10.180 align:middle line:84% Europeans were less likely to settle there. 01:13:10.180 --> 01:13:12.340 align:middle line:84% They just became extractive powers. 01:13:12.340 --> 01:13:14.860 align:middle line:84% And weak institutions got set up. 01:13:14.860 --> 01:13:17.235 align:middle line:84% So that's why-- and that's, they argue, 01:13:17.235 --> 01:13:19.360 align:middle line:84% a argue fundamental reason why Africa was much less 01:13:19.360 --> 01:13:21.200 align:middle line:90% successful than South America. 01:13:21.200 --> 01:13:22.800 align:middle line:84% If you look at two continents, which 01:13:22.800 --> 01:13:25.800 align:middle line:84% were in pretty similar places in 1600-- 01:13:25.800 --> 01:13:29.492 align:middle line:84% indeed, Africa's resource richer than South America is. 01:13:29.492 --> 01:13:30.200 align:middle line:90% Africa is a mess. 01:13:30.200 --> 01:13:32.160 align:middle line:84% And South America is, at least historically, 01:13:32.160 --> 01:13:34.118 align:middle line:84% been largely successful, certainly a lot richer 01:13:34.118 --> 01:13:35.080 align:middle line:90% than Africa. 01:13:35.080 --> 01:13:37.205 align:middle line:84% And they would argue it's because Africa was set up 01:13:37.205 --> 01:13:39.880 align:middle line:84% as an extractive culture, whereas South America was set up 01:13:39.880 --> 01:13:42.297 align:middle line:84% as a culture which had good institutions because Europeans 01:13:42.297 --> 01:13:43.600 align:middle line:90% wanted to live there. 01:13:43.600 --> 01:13:47.153 align:middle line:84% So it's a pretty cool study that says that this stuff matters. 01:13:47.153 --> 01:13:48.820 align:middle line:84% It doesn't matter for our everyday life, 01:13:48.820 --> 01:13:52.280 align:middle line:84% but literally, it matters to orders of magnitude of GDP, 01:13:52.280 --> 01:13:53.660 align:middle line:90% setting up proper institutions. 01:13:53.660 --> 01:13:56.120 align:middle line:84% And that's why, once again, it comes back to this point, 01:13:56.120 --> 01:13:59.980 align:middle line:84% that it's critical that we keep the institutions in the-- 01:13:59.980 --> 01:14:00.940 align:middle line:90% it's critical. 01:14:00.940 --> 01:14:04.227 align:middle line:84% It matters less who's in charge, who's president. 01:14:04.227 --> 01:14:06.560 align:middle line:84% What matters is the underlying institutions that support 01:14:06.560 --> 01:14:08.185 align:middle line:90% a democracy stay in place. 01:14:08.185 --> 01:14:09.560 align:middle line:84% When those start to erode, that's 01:14:09.560 --> 01:14:11.820 align:middle line:90% when things really go south. 01:14:11.820 --> 01:14:12.320 align:middle line:90% OK. 01:14:12.320 --> 01:14:14.840 align:middle line:90% Questions about that? 01:14:14.840 --> 01:14:15.692 align:middle line:90% Yeah? 01:14:15.692 --> 01:14:17.150 align:middle line:84% AUDIENCE: Because I think we talked 01:14:17.150 --> 01:14:19.800 align:middle line:84% about [INAUDIBLE] institutions versus scarcity 01:14:19.800 --> 01:14:21.550 align:middle line:84% of institutions-- are [INAUDIBLE] building 01:14:21.550 --> 01:14:23.450 align:middle line:84% institutions like ingrained in themselves, 01:14:23.450 --> 01:14:24.950 align:middle line:90% like incentivize corruption? 01:14:24.950 --> 01:14:26.950 align:middle line:84% JONATHAN GRUBER: Yeah, so what [INAUDIBLE] 01:14:26.950 --> 01:14:28.430 align:middle line:90% incentivize corruption? 01:14:28.430 --> 01:14:29.530 align:middle line:90% What do you mean by that? 01:14:29.530 --> 01:14:31.722 align:middle line:84% AUDIENCE: Or it was like [INAUDIBLE] were not 01:14:31.722 --> 01:14:32.430 align:middle line:90% well thought out. 01:14:32.430 --> 01:14:33.990 align:middle line:90% And they were like [INAUDIBLE]. 01:14:33.990 --> 01:14:34.907 align:middle line:90% JONATHAN GRUBER: Yeah. 01:14:34.907 --> 01:14:37.270 align:middle line:84% No, obviously, you can build good and bad institutions. 01:14:37.270 --> 01:14:40.117 align:middle line:84% But I think the point is that legitimate institutions are 01:14:40.117 --> 01:14:41.950 align:middle line:84% generally more observable, and observability 01:14:41.950 --> 01:14:44.310 align:middle line:90% is the opposite of corruption. 01:14:44.310 --> 01:14:47.233 align:middle line:84% So basically, when things are observable, 01:14:47.233 --> 01:14:48.650 align:middle line:84% they're less likely to be corrupt. 01:14:48.650 --> 01:14:50.150 align:middle line:84% So I think that would be the answer. 01:14:50.150 --> 01:14:51.965 align:middle line:84% Once again, if you find this interesting, 01:14:51.965 --> 01:14:54.090 align:middle line:84% Why Nations Fail-- sold more than a million copies, 01:14:54.090 --> 01:14:55.325 align:middle line:90% amazing book. 01:14:55.325 --> 01:14:57.950 align:middle line:84% So you can read all about this set of really interesting issues 01:14:57.950 --> 01:15:01.390 align:middle line:84% from one of our nation's leading intellectual stars. 01:15:01.390 --> 01:15:02.110 align:middle line:90% Yeah? 01:15:02.110 --> 01:15:03.870 align:middle line:84% AUDIENCE: Do you think that paying politicians more 01:15:03.870 --> 01:15:06.010 align:middle line:84% for cracking down on corruption would do anything? 01:15:06.010 --> 01:15:08.135 align:middle line:84% JONATHAN GRUBER: It's a really interesting question 01:15:08.135 --> 01:15:11.650 align:middle line:84% about the proper reimbursement of public officials. 01:15:11.650 --> 01:15:13.108 align:middle line:84% Indeed, there's a lot of incentive, 01:15:13.108 --> 01:15:15.192 align:middle line:84% that part of the reason politicians are so corrupt 01:15:15.192 --> 01:15:17.550 align:middle line:84% and local officials are corrupt in low-income countries, 01:15:17.550 --> 01:15:19.570 align:middle line:84% they're paid so little [INAUDIBLE] can survive, 01:15:19.570 --> 01:15:21.230 align:middle line:84% that basically, they take bribes. 01:15:21.230 --> 01:15:24.130 align:middle line:84% And think of a simple model, where 01:15:24.130 --> 01:15:27.410 align:middle line:84% I have consumption and honesty, my utility function. 01:15:27.410 --> 01:15:29.850 align:middle line:90% And I have a budget constraint. 01:15:29.850 --> 01:15:32.490 align:middle line:84% Well, in that model, the more income I have, the more honest 01:15:32.490 --> 01:15:36.550 align:middle line:84% I'll be, because honesty is a good something I want. 01:15:36.550 --> 01:15:38.250 align:middle line:90% But I got to survive too. 01:15:38.250 --> 01:15:40.970 align:middle line:84% So in that simple model, paying politicians more 01:15:40.970 --> 01:15:42.150 align:middle line:90% will fight corruption. 01:15:42.150 --> 01:15:44.250 align:middle line:84% There's some evidence that's true. 01:15:44.250 --> 01:15:46.050 align:middle line:90% I think it's second order. 01:15:46.050 --> 01:15:47.010 align:middle line:90% I'm not an expert. 01:15:47.010 --> 01:15:48.987 align:middle line:84% Ben Olken, who's another incredibly wonderful 01:15:48.987 --> 01:15:50.570 align:middle line:84% political economist in our department, 01:15:50.570 --> 01:15:51.810 align:middle line:90% has done work on that. 01:15:51.810 --> 01:15:54.470 align:middle line:84% I think the evidence is that it matters, 01:15:54.470 --> 01:15:59.410 align:middle line:84% but it's not a primary determinant. 01:15:59.410 --> 01:16:00.270 align:middle line:90% All right. 01:16:00.270 --> 01:16:01.250 align:middle line:90% Great. 01:16:01.250 --> 01:16:01.750 align:middle line:90% Good. 01:16:01.750 --> 01:16:03.917 align:middle line:84% Hopefully I didn't get myself into too much trouble. 01:16:03.917 --> 01:16:05.070 align:middle line:90% Let's stop there. 01:16:05.070 --> 01:16:10.120 align:middle line:84% And we will come back on Monday, and we'll talk about chapter 10. 01:16:10.120 --> 01:16:15.000 align:middle line:90%