WEBVTT 00:00:00.000 --> 00:00:00.500 align:middle line:90% 00:00:00.500 --> 00:00:01.949 align:middle line:90% [SQUEAKING] 00:00:01.949 --> 00:00:03.881 align:middle line:90% [RUSTLING] 00:00:03.881 --> 00:00:07.262 align:middle line:90% [CLICKING] 00:00:07.262 --> 00:00:09.677 align:middle line:90% 00:00:09.677 --> 00:00:11.880 align:middle line:84% JON GRUBER: Education economics-- so 00:00:11.880 --> 00:00:12.920 align:middle line:90% let's remind ourselves. 00:00:12.920 --> 00:00:16.480 align:middle line:84% We showed last time that basically, it 00:00:16.480 --> 00:00:17.920 align:middle line:84% doesn't look like across countries 00:00:17.920 --> 00:00:21.440 align:middle line:84% what you spend on education really correlates with outcomes. 00:00:21.440 --> 00:00:23.282 align:middle line:84% We then ask the question, well, why should-- 00:00:23.282 --> 00:00:24.740 align:middle line:84% this is what the government spends? 00:00:24.740 --> 00:00:25.360 align:middle line:90% I'm sorry. 00:00:25.360 --> 00:00:27.193 align:middle line:84% We then ask, well, why should the government 00:00:27.193 --> 00:00:28.280 align:middle line:90% be involved in education? 00:00:28.280 --> 00:00:30.160 align:middle line:84% Why not just have a private market where everyone just 00:00:30.160 --> 00:00:32.410 align:middle line:84% pays for it themselves and take loans if they need to? 00:00:32.410 --> 00:00:35.760 align:middle line:84% And we came up with a number of answers 00:00:35.760 --> 00:00:39.280 align:middle line:84% from the fact there's minimum standards of citizenship we want 00:00:39.280 --> 00:00:41.760 align:middle line:84% to meet, to improvements in productivity 00:00:41.760 --> 00:00:48.143 align:middle line:84% that spill over to others, to credit market failures. 00:00:48.143 --> 00:00:49.560 align:middle line:84% And I think an important one here, 00:00:49.560 --> 00:00:51.960 align:middle line:84% is this failure to maximize family utility. 00:00:51.960 --> 00:00:56.600 align:middle line:84% Let's remember that it's kids who ultimately 00:00:56.600 --> 00:00:59.260 align:middle line:84% bear the consequences of their parents' decisions to educate 00:00:59.260 --> 00:01:01.200 align:middle line:90% them in important ways. 00:01:01.200 --> 00:01:04.660 align:middle line:84% Income mobility is a critical aspect of success in life. 00:01:04.660 --> 00:01:07.540 align:middle line:84% And income mobility depends critically on education. 00:01:07.540 --> 00:01:10.347 align:middle line:84% So only if we really think parents are really 00:01:10.347 --> 00:01:12.180 align:middle line:84% doing what's in society's best interest, not 00:01:12.180 --> 00:01:14.140 align:middle line:84% just their own best interest, do we really 00:01:14.140 --> 00:01:16.460 align:middle line:84% want to leave a total in the hands of families about 00:01:16.460 --> 00:01:18.080 align:middle line:84% whether kids go to school or not? 00:01:18.080 --> 00:01:19.980 align:middle line:84% We probably want some government involvement 00:01:19.980 --> 00:01:22.020 align:middle line:90% in thinking about that at least. 00:01:22.020 --> 00:01:24.598 align:middle line:84% OK, so now, with that as background, now 00:01:24.598 --> 00:01:27.140 align:middle line:84% that we've answered the first question-- which is when should 00:01:27.140 --> 00:01:28.182 align:middle line:90% the government intervene? 00:01:28.182 --> 00:01:31.478 align:middle line:84% The answer is generally it should in cases like education. 00:01:31.478 --> 00:01:33.020 align:middle line:84% We come to the second question, which 00:01:33.020 --> 00:01:35.395 align:middle line:84% is how should the government intervene? 00:01:35.395 --> 00:01:37.020 align:middle line:84% And here we're going to start with what 00:01:37.020 --> 00:01:39.780 align:middle line:84% the government actually does, and then think about 00:01:39.780 --> 00:01:41.500 align:middle line:90% whether it makes sense. 00:01:41.500 --> 00:01:43.500 align:middle line:90% So what does the government do? 00:01:43.500 --> 00:01:45.140 align:middle line:84% Well, the government, in practice, 00:01:45.140 --> 00:01:49.700 align:middle line:84% provides free public education of a certain amount. 00:01:49.700 --> 00:01:55.140 align:middle line:84% To understand how that works, let's think about a simple model 00:01:55.140 --> 00:01:59.600 align:middle line:84% where individuals, much like our model of grants 00:01:59.600 --> 00:02:01.520 align:middle line:84% at the end of the last chapter, but now it's 00:02:01.520 --> 00:02:03.300 align:middle line:90% an individual making a decision. 00:02:03.300 --> 00:02:05.792 align:middle line:84% They're deciding between how much to spend on education 00:02:05.792 --> 00:02:07.500 align:middle line:84% and how much to spend on everything else. 00:02:07.500 --> 00:02:11.152 align:middle line:84% So this is the figure 11-2 in your handout. 00:02:11.152 --> 00:02:13.360 align:middle line:84% The government's deciding how much spend in education 00:02:13.360 --> 00:02:15.110 align:middle line:84% and how much to spend on everything else-- 00:02:15.110 --> 00:02:16.360 align:middle line:90% the individual, I'm sorry. 00:02:16.360 --> 00:02:18.200 align:middle line:84% And they face a budget constraint. 00:02:18.200 --> 00:02:20.080 align:middle line:84% They can either spend an amount-- 00:02:20.080 --> 00:02:21.180 align:middle line:90% they have a budget. 00:02:21.180 --> 00:02:23.720 align:middle line:84% They can either spend it all in education, in which case 00:02:23.720 --> 00:02:26.800 align:middle line:84% they'll get B of amount of education, 00:02:26.800 --> 00:02:30.720 align:middle line:84% or all the other goods, in which case they'll get G2. 00:02:30.720 --> 00:02:34.400 align:middle line:84% And let's assume, despite the cross-country graph 00:02:34.400 --> 00:02:36.880 align:middle line:84% we saw a few minutes ago, that education spending 00:02:36.880 --> 00:02:37.857 align:middle line:90% is productive. 00:02:37.857 --> 00:02:39.940 align:middle line:84% Let's assume that the more you spend in education, 00:02:39.940 --> 00:02:41.872 align:middle line:84% the better education your kid gets. 00:02:41.872 --> 00:02:43.580 align:middle line:84% We're just going to make that assumption. 00:02:43.580 --> 00:02:45.330 align:middle line:84% We'll come back later to evidence on that. 00:02:45.330 --> 00:02:46.520 align:middle line:90% Let's assume that here. 00:02:46.520 --> 00:02:50.720 align:middle line:84% So the x-axis is both spending and quality 00:02:50.720 --> 00:02:52.680 align:middle line:84% So parents have to make some decision. 00:02:52.680 --> 00:02:54.482 align:middle line:84% And absent government intervention, 00:02:54.482 --> 00:02:56.940 align:middle line:84% they'll make some decisions with three illustrative parents 00:02:56.940 --> 00:02:57.440 align:middle line:90% here-- 00:02:57.440 --> 00:03:01.060 align:middle line:84% X, Y, and Z. They choose different tradeoffs absent 00:03:01.060 --> 00:03:02.780 align:middle line:90% government intervention. 00:03:02.780 --> 00:03:05.820 align:middle line:84% So the ex-ante, we end up with three different education 00:03:05.820 --> 00:03:06.940 align:middle line:90% levels-- 00:03:06.940 --> 00:03:11.378 align:middle line:84% E1, E2, and E3 for three different parents. 00:03:11.378 --> 00:03:12.920 align:middle line:84% Now the government comes in and says, 00:03:12.920 --> 00:03:15.820 align:middle line:84% we're going to provide free public education up 00:03:15.820 --> 00:03:23.460 align:middle line:84% to an amount E sub F. So anybody can have E sub F for free. 00:03:23.460 --> 00:03:26.220 align:middle line:84% But we're going to make a key assumption here. 00:03:26.220 --> 00:03:29.860 align:middle line:84% If you go to public school, there's no supplementation. 00:03:29.860 --> 00:03:30.920 align:middle line:90% There's no tutoring. 00:03:30.920 --> 00:03:34.820 align:middle line:84% There's no, like, extra Russian math school 00:03:34.820 --> 00:03:35.903 align:middle line:90% you can send your kids to. 00:03:35.903 --> 00:03:38.570 align:middle line:84% It's a world where basically, if you go to public school, that's 00:03:38.570 --> 00:03:39.420 align:middle line:90% all you get. 00:03:39.420 --> 00:03:41.500 align:middle line:84% So your choice is either you go to public school 00:03:41.500 --> 00:03:42.667 align:middle line:90% or you go to private school. 00:03:42.667 --> 00:03:45.620 align:middle line:84% If you go to public school, you get E sub F for free, 00:03:45.620 --> 00:03:47.260 align:middle line:90% but no more. 00:03:47.260 --> 00:03:51.563 align:middle line:84% If you go to private school, you still get what you pay for. 00:03:51.563 --> 00:03:53.480 align:middle line:84% So how does this change the budget constraint? 00:03:53.480 --> 00:03:55.470 align:middle line:84% Well, the new budget constraint, much 00:03:55.470 --> 00:03:57.490 align:middle line:84% like we saw with the grant example last time, 00:03:57.490 --> 00:04:01.790 align:middle line:84% is a flat segment that goes all the way to an amount E sub F, 00:04:01.790 --> 00:04:04.190 align:middle line:84% all the way up to point C, because you 00:04:04.190 --> 00:04:05.910 align:middle line:84% can get that much education for free 00:04:05.910 --> 00:04:08.830 align:middle line:84% without giving up any other private goods. 00:04:08.830 --> 00:04:11.650 align:middle line:84% At that point, you stop getting free public education. 00:04:11.650 --> 00:04:13.910 align:middle line:84% The budget constraint drops from C to D. 00:04:13.910 --> 00:04:17.910 align:middle line:84% And you return to the original private budget constraint. 00:04:17.910 --> 00:04:23.590 align:middle line:84% Student budget constraint goes G2, C, D, B. 00:04:23.590 --> 00:04:26.230 align:middle line:84% That's the new budget constraint. 00:04:26.230 --> 00:04:28.910 align:middle line:90% Questions about that? 00:04:28.910 --> 00:04:31.470 align:middle line:84% Now let's ask, how does that affect people's choices? 00:04:31.470 --> 00:04:37.390 align:middle line:84% Well, let's start with person X. Person X 00:04:37.390 --> 00:04:39.710 align:middle line:84% used to not want a lot of education 00:04:39.710 --> 00:04:41.130 align:middle line:90% and a lot of private goods. 00:04:41.130 --> 00:04:43.427 align:middle line:90% Now they can get more of both. 00:04:43.427 --> 00:04:45.510 align:middle line:84% They can get a lot more education and more private 00:04:45.510 --> 00:04:49.150 align:middle line:84% goods by dropping out of their crappy private school 00:04:49.150 --> 00:04:51.510 align:middle line:90% and joining this public school. 00:04:51.510 --> 00:04:53.890 align:middle line:90% So person X, it's a no-brainer. 00:04:53.890 --> 00:04:57.010 align:middle line:84% They've made themselves unambiguously better off. 00:04:57.010 --> 00:05:01.290 align:middle line:90% For person Z, nothing's changed. 00:05:01.290 --> 00:05:04.890 align:middle line:84% The public schooling level is so much lower 00:05:04.890 --> 00:05:06.930 align:middle line:84% than their desired level of education 00:05:06.930 --> 00:05:09.850 align:middle line:90% that they're not interested. 00:05:09.850 --> 00:05:11.550 align:middle line:84% Person Y is the interesting case. 00:05:11.550 --> 00:05:13.850 align:middle line:84% And what is particularly notable about person Y? 00:05:13.850 --> 00:05:19.130 align:middle line:84% What's notable about person Y is that their indifference 00:05:19.130 --> 00:05:21.650 align:middle line:84% curve that's tangent to the original private budget 00:05:21.650 --> 00:05:27.090 align:middle line:84% constraint passes through the vertical segment from C to D. 00:05:27.090 --> 00:05:27.950 align:middle line:90% What does that mean? 00:05:27.950 --> 00:05:32.090 align:middle line:84% That means that graphically, they are better off at point C 00:05:32.090 --> 00:05:34.330 align:middle line:84% than they are at their current choice. 00:05:34.330 --> 00:05:37.810 align:middle line:84% So they will move from their current choice of E sub 2 00:05:37.810 --> 00:05:38.510 align:middle line:90% to E sub F. 00:05:38.510 --> 00:05:39.510 align:middle line:90% And how do we know that? 00:05:39.510 --> 00:05:41.710 align:middle line:84% You can just prove geometrically that since there's 00:05:41.710 --> 00:05:44.650 align:middle line:84% a difference across that vertical segment, at point C, 00:05:44.650 --> 00:05:47.210 align:middle line:84% they'll be in a higher indifference curve. 00:05:47.210 --> 00:05:52.310 align:middle line:84% In other words, person y is willing to get less education 00:05:52.310 --> 00:05:56.430 align:middle line:84% to get a lot more other consumption. 00:05:56.430 --> 00:06:02.750 align:middle line:84% This is a case of more than 100% crowd out. 00:06:02.750 --> 00:06:07.190 align:middle line:84% This is a case where basically, the government providing it 00:06:07.190 --> 00:06:10.415 align:middle line:84% doesn't just reduce private provision to 0. 00:06:10.415 --> 00:06:12.790 align:middle line:84% It produces private provision by more than the government 00:06:12.790 --> 00:06:15.030 align:middle line:90% actually provided. 00:06:15.030 --> 00:06:21.070 align:middle line:84% That is total educational attainment for person y falls. 00:06:21.070 --> 00:06:24.270 align:middle line:84% So on net, what happens to society? 00:06:24.270 --> 00:06:25.790 align:middle line:84% Well that depends on the proportions 00:06:25.790 --> 00:06:30.430 align:middle line:84% of people who are X, Y, and Z. If society's all like X, 00:06:30.430 --> 00:06:32.750 align:middle line:84% education unambiguously increases. 00:06:32.750 --> 00:06:35.030 align:middle line:84% If society is all like Z, education unambiguously 00:06:35.030 --> 00:06:36.110 align:middle line:90% doesn't change. 00:06:36.110 --> 00:06:39.350 align:middle line:84% But if it's like Y, education can actually fall. 00:06:39.350 --> 00:06:42.810 align:middle line:84% This is a key point made by Sam Peltzman in 1973. 00:06:42.810 --> 00:06:45.510 align:middle line:84% It's a famous model-- another Chicago economist, 00:06:45.510 --> 00:06:47.650 align:middle line:84% showing that the government can make things worse, 00:06:47.650 --> 00:06:51.250 align:middle line:84% and that in principle, free public education can 00:06:51.250 --> 00:06:52.510 align:middle line:90% lower educational attainment. 00:06:52.510 --> 00:06:53.802 align:middle line:90% It's a pretty striking finding. 00:06:53.802 --> 00:06:55.692 align:middle line:84% But it's pretty clear from this graph 00:06:55.692 --> 00:06:57.650 align:middle line:84% that a free public education can actually lower 00:06:57.650 --> 00:06:59.155 align:middle line:90% how much education people get. 00:06:59.155 --> 00:07:00.530 align:middle line:84% And whether it will happen or not 00:07:00.530 --> 00:07:04.770 align:middle line:84% depends on how big Y is compared to X, basically. 00:07:04.770 --> 00:07:07.270 align:middle line:84% X is getting a lot more, y is getting somewhat less. 00:07:07.270 --> 00:07:09.170 align:middle line:84% The size of those groups will determine 00:07:09.170 --> 00:07:12.490 align:middle line:84% whether total education goes up or not. 00:07:12.490 --> 00:07:15.250 align:middle line:84% So this is really a quite striking finding, 00:07:15.250 --> 00:07:18.050 align:middle line:84% that free public education could actually lead 00:07:18.050 --> 00:07:21.050 align:middle line:90% to less educational attainment. 00:07:21.050 --> 00:07:23.930 align:middle line:84% Now, Peltzman said, look, there's 00:07:23.930 --> 00:07:27.130 align:middle line:84% an easy solution to this problem, which 00:07:27.130 --> 00:07:30.050 align:middle line:84% is what if instead of telling people 00:07:30.050 --> 00:07:34.570 align:middle line:84% they have to go to public school, we gave them a voucher? 00:07:34.570 --> 00:07:39.090 align:middle line:84% And that voucher was for the amount E sub F. 00:07:39.090 --> 00:07:42.010 align:middle line:84% And we said, look, there will be public schools 00:07:42.010 --> 00:07:43.153 align:middle line:90% that will provide E sub F. 00:07:43.153 --> 00:07:45.070 align:middle line:84% But if you don't want to go to public schools, 00:07:45.070 --> 00:07:46.970 align:middle line:84% you can take your voucher E sub F 00:07:46.970 --> 00:07:49.830 align:middle line:84% and apply it to any private school as well. 00:07:49.830 --> 00:07:51.928 align:middle line:84% So we're going to give you a piece of paper. 00:07:51.928 --> 00:07:53.470 align:middle line:84% You can just turn that piece of paper 00:07:53.470 --> 00:07:55.250 align:middle line:84% back into the public school and be done with it. 00:07:55.250 --> 00:07:57.375 align:middle line:84% You turn that piece of paper into a private school, 00:07:57.375 --> 00:07:59.510 align:middle line:84% and then pay some extra on top of that 00:07:59.510 --> 00:08:00.690 align:middle line:90% and go to private school. 00:08:00.690 --> 00:08:05.630 align:middle line:84% So this is a coupon off of private school for the amount 00:08:05.630 --> 00:08:07.790 align:middle line:90% that we spend on public school. 00:08:07.790 --> 00:08:10.630 align:middle line:84% What would that do to the budget constraint? 00:08:10.630 --> 00:08:13.350 align:middle line:84% Well, that moves the budget constraint to something like 00:08:13.350 --> 00:08:20.150 align:middle line:84% figure 11-3, where now there's still the same flat segment 00:08:20.150 --> 00:08:20.990 align:middle line:90% along the top. 00:08:20.990 --> 00:08:23.490 align:middle line:84% But now below that, the whole budget constraint shifted out. 00:08:23.490 --> 00:08:25.910 align:middle line:84% This is like the conditional block grant 00:08:25.910 --> 00:08:27.990 align:middle line:90% example from last lecture. 00:08:27.990 --> 00:08:29.130 align:middle line:90% The whole budget shrinks. 00:08:29.130 --> 00:08:34.409 align:middle line:84% So basically what happens is, person X continues to go to C. 00:08:34.409 --> 00:08:36.230 align:middle line:90% They continue to get education. 00:08:36.230 --> 00:08:41.150 align:middle line:84% But now person Y also gets more education and person 00:08:41.150 --> 00:08:43.190 align:middle line:90% Z gets more education. 00:08:43.190 --> 00:08:47.490 align:middle line:84% So now you unambiguously increase educational attainment. 00:08:47.490 --> 00:08:53.530 align:middle line:84% So relative to a simple, limited public education system, 00:08:53.530 --> 00:08:57.130 align:middle line:84% by giving a voucher, you can unambiguously 00:08:57.130 --> 00:08:59.730 align:middle line:90% increase educational attainment. 00:08:59.730 --> 00:09:02.250 align:middle line:84% And this is one of three arguments supporting 00:09:02.250 --> 00:09:04.250 align:middle line:84% what has become a very significant voucher 00:09:04.250 --> 00:09:05.850 align:middle line:90% movement in the US-- 00:09:05.850 --> 00:09:08.690 align:middle line:84% a movement in the US and around the world 00:09:08.690 --> 00:09:10.490 align:middle line:84% to allow students to essentially, instead 00:09:10.490 --> 00:09:12.750 align:middle line:84% of being forced to go to the public schools, 00:09:12.750 --> 00:09:14.970 align:middle line:84% to be given a choice where they can literally 00:09:14.970 --> 00:09:17.090 align:middle line:84% direct their dollars to the public school 00:09:17.090 --> 00:09:19.090 align:middle line:90% or to other alternatives. 00:09:19.090 --> 00:09:22.530 align:middle line:84% And there are three arguments people make for vouchers. 00:09:22.530 --> 00:09:26.170 align:middle line:84% One is the crowd out argument, that it 00:09:26.170 --> 00:09:29.210 align:middle line:84% ends this crowd out of group y, and it actually increases 00:09:29.210 --> 00:09:32.370 align:middle line:90% attainment for group Z as well. 00:09:32.370 --> 00:09:36.530 align:middle line:84% The second argument that people make 00:09:36.530 --> 00:09:44.830 align:middle line:84% is a consumer sovereignty argument, that basically, look-- 00:09:44.830 --> 00:09:47.430 align:middle line:90% 00:09:47.430 --> 00:09:50.590 align:middle line:90% I spelled sovereignty, whatever. 00:09:50.590 --> 00:09:53.790 align:middle line:84% Basically, look, we think we should respect 00:09:53.790 --> 00:09:55.170 align:middle line:90% the consumer of lots of things. 00:09:55.170 --> 00:09:57.295 align:middle line:84% Why do we think the government knows better than we 00:09:57.295 --> 00:09:59.410 align:middle line:90% do what the best school is? 00:09:59.410 --> 00:10:03.090 align:middle line:84% The government may just stick kids in some crappy school. 00:10:03.090 --> 00:10:07.390 align:middle line:84% Let's let people decide what's best for their kids. 00:10:07.390 --> 00:10:09.110 align:middle line:84% And an advantage of allowing that 00:10:09.110 --> 00:10:15.110 align:middle line:84% is a third benefit, which is efficiency, which is basically 00:10:15.110 --> 00:10:18.310 align:middle line:90% efficiency from shopping. 00:10:18.310 --> 00:10:22.870 align:middle line:84% If people can shop, they'll choose the best schools. 00:10:22.870 --> 00:10:25.110 align:middle line:90% What are local public schools? 00:10:25.110 --> 00:10:27.770 align:middle line:84% They're government-sanctioned monopolies. 00:10:27.770 --> 00:10:29.770 align:middle line:84% When you have a government-sanctioned monopoly, 00:10:29.770 --> 00:10:33.570 align:middle line:84% it's by definition going to be operating inefficiently. 00:10:33.570 --> 00:10:35.800 align:middle line:90% It's a nonprofit monopoly. 00:10:35.800 --> 00:10:37.550 align:middle line:84% A for-profit monopoly operates efficiently 00:10:37.550 --> 00:10:39.870 align:middle line:90% and sucks up all the profits. 00:10:39.870 --> 00:10:41.750 align:middle line:84% But a nonprofit monopoly has no incentive 00:10:41.750 --> 00:10:45.370 align:middle line:84% to operate efficiently because it can't keep the money. 00:10:45.370 --> 00:10:48.170 align:middle line:84% So the last thing you want is a nonprofit monopoly. 00:10:48.170 --> 00:10:49.710 align:middle line:84% It's not the last thing you want, 00:10:49.710 --> 00:10:51.543 align:middle line:84% but it's costly to have a nonprofit monopoly 00:10:51.543 --> 00:10:53.710 align:middle line:84% because there's no incentive to operate efficiently. 00:10:53.710 --> 00:10:56.650 align:middle line:84% If operating efficiently is stressful, why bother? 00:10:56.650 --> 00:10:59.370 align:middle line:90% You get the kids no matter what. 00:10:59.370 --> 00:11:03.050 align:middle line:84% So that's the clear arguments for a voucher system. 00:11:03.050 --> 00:11:06.650 align:middle line:84% And that is why there's a strong advocacy typically 00:11:06.650 --> 00:11:08.430 align:middle line:84% on the conservative side of the aisle. 00:11:08.430 --> 00:11:09.730 align:middle line:84% But this is a little more bipartisan 00:11:09.730 --> 00:11:11.310 align:middle line:84% than some other things we'll talk about. 00:11:11.310 --> 00:11:12.977 align:middle line:84% It's typically conservative, but there's 00:11:12.977 --> 00:11:17.598 align:middle line:84% a little more bipartisan appeal to this for vouchers. 00:11:17.598 --> 00:11:19.390 align:middle line:84% So what are the arguments against vouchers? 00:11:19.390 --> 00:11:22.810 align:middle line:90% 00:11:22.810 --> 00:11:23.670 align:middle line:90% Why not vouchers? 00:11:23.670 --> 00:11:24.470 align:middle line:90% Enoch. 00:11:24.470 --> 00:11:26.130 align:middle line:84% AUDIENCE: In the transition phase, 00:11:26.130 --> 00:11:32.242 align:middle line:84% it would harm [INAUDIBLE] things that [INAUDIBLE]. 00:11:32.242 --> 00:11:34.310 align:middle line:84% JON GRUBER: What about the kids left behind? 00:11:34.310 --> 00:11:36.718 align:middle line:84% Remember, what vouchers do is they say-- basically, 00:11:36.718 --> 00:11:38.510 align:middle line:84% you take your voucher go to another school. 00:11:38.510 --> 00:11:39.710 align:middle line:90% And the idea the efficiency-- 00:11:39.710 --> 00:11:41.960 align:middle line:84% I should have said this, by the way, a little clearer. 00:11:41.960 --> 00:11:44.560 align:middle line:84% The way the efficiency works is, when you go to the school, 00:11:44.560 --> 00:11:46.220 align:middle line:90% you bring your dollars with you. 00:11:46.220 --> 00:11:47.720 align:middle line:84% So the good school get more dollars. 00:11:47.720 --> 00:11:49.095 align:middle line:84% The bad schools get less dollars. 00:11:49.095 --> 00:11:51.710 align:middle line:84% And ultimately, if it's a company, goes out of business. 00:11:51.710 --> 00:11:52.960 align:middle line:90% That's the idea of efficiency. 00:11:52.960 --> 00:11:54.900 align:middle line:84% Efficiency works through exit and entry. 00:11:54.900 --> 00:11:56.900 align:middle line:84% The problem Enoch is pointing out 00:11:56.900 --> 00:11:58.900 align:middle line:84% is, if somebody's got an absent parent who's 00:11:58.900 --> 00:12:01.280 align:middle line:84% not paying attention and the school goes out of business, 00:12:01.280 --> 00:12:02.820 align:middle line:90% they're screwed. 00:12:02.820 --> 00:12:05.780 align:middle line:84% So basically, in some sense, the problem 00:12:05.780 --> 00:12:09.120 align:middle line:84% is you are essentially allowing people to make the best choice. 00:12:09.120 --> 00:12:11.340 align:middle line:84% But you're relying on family utility maximization 00:12:11.340 --> 00:12:12.930 align:middle line:84% that they will make the best choice. 00:12:12.930 --> 00:12:15.180 align:middle line:84% And the ones who don't make that choice end up getting 00:12:15.180 --> 00:12:19.220 align:middle line:84% left behind in increasingly poorer and poorer schools. 00:12:19.220 --> 00:12:21.300 align:middle line:90% So that's one problem. 00:12:21.300 --> 00:12:23.520 align:middle line:84% What other what other problems are there? 00:12:23.520 --> 00:12:28.380 align:middle line:90% 00:12:28.380 --> 00:12:30.920 align:middle line:84% Well, this is a related one to this, to this one. 00:12:30.920 --> 00:12:31.880 align:middle line:90% Let's come to this. 00:12:31.880 --> 00:12:34.267 align:middle line:84% What's the other problem with this argument? 00:12:34.267 --> 00:12:35.600 align:middle line:90% What's wrong with this argument? 00:12:35.600 --> 00:12:38.700 align:middle line:90% 00:12:38.700 --> 00:12:39.220 align:middle line:90% Yeah. 00:12:39.220 --> 00:12:40.860 align:middle line:90% AUDIENCE: [INAUDIBLE] 00:12:40.860 --> 00:12:42.680 align:middle line:84% JON GRUBER: It's hard to shop people. 00:12:42.680 --> 00:12:43.780 align:middle line:90% There's bad information. 00:12:43.780 --> 00:12:45.320 align:middle line:90% People don't do it well. 00:12:45.320 --> 00:12:50.140 align:middle line:84% So there's basically an imperfect market for schools. 00:12:50.140 --> 00:12:53.117 align:middle line:84% Both market imperfections and market power 00:12:53.117 --> 00:12:55.200 align:middle line:84% are going to make it that you won't get efficiency 00:12:55.200 --> 00:12:56.640 align:middle line:90% from shopping. 00:12:56.640 --> 00:12:59.347 align:middle line:84% So you may do better from shopping. 00:12:59.347 --> 00:13:01.680 align:middle line:84% But the notion is, the first best is clearly wrong given 00:13:01.680 --> 00:13:03.560 align:middle line:90% imperfections in this market. 00:13:03.560 --> 00:13:05.740 align:middle line:84% So the second thing is market imperfections. 00:13:05.740 --> 00:13:10.560 align:middle line:90% 00:13:10.560 --> 00:13:12.760 align:middle line:84% So that's what's wrong with that argument. 00:13:12.760 --> 00:13:14.400 align:middle line:84% What's wrong with this argument-- 00:13:14.400 --> 00:13:18.037 align:middle line:90% consumer sovereignty? 00:13:18.037 --> 00:13:19.620 align:middle line:84% How would you argue back against that? 00:13:19.620 --> 00:13:23.280 align:middle line:90% 00:13:23.280 --> 00:13:23.800 align:middle line:90% Steven. 00:13:23.800 --> 00:13:27.220 align:middle line:84% AUDIENCE: [INAUDIBLE] consumers are able to switch schools, 00:13:27.220 --> 00:13:29.907 align:middle line:84% so if the nicer school is further away. 00:13:29.907 --> 00:13:32.240 align:middle line:84% JON GRUBER: No, but that would be a market imperfection, 00:13:32.240 --> 00:13:33.800 align:middle line:90% or what gets left behind. 00:13:33.800 --> 00:13:36.020 align:middle line:84% But to relate to that, what else concerns you? 00:13:36.020 --> 00:13:36.820 align:middle line:90% Yeah. 00:13:36.820 --> 00:13:39.480 align:middle line:84% AUDIENCE: How do we know that parents know 00:13:39.480 --> 00:13:40.920 align:middle line:90% what's best for their children? 00:13:40.920 --> 00:13:42.260 align:middle line:84% JON GRUBER: Yeah, how do we know the parents are going 00:13:42.260 --> 00:13:44.500 align:middle line:84% to choose a school that delivers the socially 00:13:44.500 --> 00:13:46.420 align:middle line:90% optimal outcome to their kids? 00:13:46.420 --> 00:13:49.355 align:middle line:84% I mean, if you're in Texas and you 00:13:49.355 --> 00:13:50.980 align:middle line:84% want to compete for the best resources, 00:13:50.980 --> 00:13:52.480 align:middle line:84% you offer the best football team. 00:13:52.480 --> 00:13:54.340 align:middle line:84% You don't offer the best math classes, 00:13:54.340 --> 00:13:57.260 align:middle line:84% because basically the high schools with good football teams 00:13:57.260 --> 00:13:59.080 align:middle line:90% get all the rich parents. 00:13:59.080 --> 00:14:00.940 align:middle line:90% They get a lot of money. 00:14:00.940 --> 00:14:03.540 align:middle line:84% Is that what we want, we're in a market where high schools are 00:14:03.540 --> 00:14:06.620 align:middle line:84% chosen based on their football team, not their educational 00:14:06.620 --> 00:14:07.980 align:middle line:90% attainment? 00:14:07.980 --> 00:14:13.220 align:middle line:84% So basically, the problem is that you 00:14:13.220 --> 00:14:17.780 align:middle line:84% could have excessive specialization, that basically, 00:14:17.780 --> 00:14:23.780 align:middle line:84% parents say, look, I care about my kid spending 00:14:23.780 --> 00:14:31.280 align:middle line:84% the entire class learning that evolution is fake 00:14:31.280 --> 00:14:33.860 align:middle line:90% and creationism is the truth. 00:14:33.860 --> 00:14:35.300 align:middle line:90% Do we want that? 00:14:35.300 --> 00:14:37.480 align:middle line:90% This is not just a vague idea. 00:14:37.480 --> 00:14:39.280 align:middle line:84% There was actually a massive exposé 00:14:39.280 --> 00:14:44.612 align:middle line:84% of the worst schools in America by objective school tests. 00:14:44.612 --> 00:14:46.320 align:middle line:84% Certainly, the worst schools in the state 00:14:46.320 --> 00:14:49.120 align:middle line:84% of New York by school tests are a set 00:14:49.120 --> 00:14:51.680 align:middle line:84% of specialized Hasidic high schools-- 00:14:51.680 --> 00:14:56.120 align:middle line:84% Hasidic is a sect of Judaism, in New York City, which 00:14:56.120 --> 00:14:59.720 align:middle line:84% are set up as essentially religious schools masquerading 00:14:59.720 --> 00:15:02.560 align:middle line:90% as educational institutions. 00:15:02.560 --> 00:15:04.000 align:middle line:84% The students in those schools are 00:15:04.000 --> 00:15:06.680 align:middle line:84% uniformly, on every single test, the very worst students 00:15:06.680 --> 00:15:07.640 align:middle line:90% in New York State. 00:15:07.640 --> 00:15:09.280 align:middle line:84% Students who have been those schools 00:15:09.280 --> 00:15:11.420 align:middle line:84% say they come out totally unprepared. 00:15:11.420 --> 00:15:13.300 align:middle line:84% They don't take math or reading or anything. 00:15:13.300 --> 00:15:15.800 align:middle line:84% They just learn religious studies, 00:15:15.800 --> 00:15:19.400 align:middle line:84% and they come out unprepared to live in modern society. 00:15:19.400 --> 00:15:22.520 align:middle line:84% But these are private schools and they 00:15:22.520 --> 00:15:24.320 align:middle line:90% are sanctioned by the state. 00:15:24.320 --> 00:15:28.040 align:middle line:84% And basically, you have excessive specialization. 00:15:28.040 --> 00:15:30.500 align:middle line:84% Now, in principle-- yeah, Steven. 00:15:30.500 --> 00:15:35.320 align:middle line:84% AUDIENCE: [INAUDIBLE] is there a way to open up [INAUDIBLE]? 00:15:35.320 --> 00:15:36.820 align:middle line:84% JON GRUBER: Yeah, so in principle, 00:15:36.820 --> 00:15:39.060 align:middle line:84% you could try to address that by saying, 00:15:39.060 --> 00:15:41.580 align:middle line:84% OK, there's certain things you have to teach, 00:15:41.580 --> 00:15:44.540 align:middle line:84% certain standards you have to reach. 00:15:44.540 --> 00:15:46.778 align:middle line:84% But in practice, in the limit, it 00:15:46.778 --> 00:15:48.820 align:middle line:84% might just end up being easier to run the school. 00:15:48.820 --> 00:15:50.140 align:middle line:84% I mean, in some sense, given how difficult 00:15:50.140 --> 00:15:52.240 align:middle line:84% it is to maintain those standards and monitor them, 00:15:52.240 --> 00:15:54.320 align:middle line:84% you got at some point say you really want to just do that. 00:15:54.320 --> 00:15:57.000 align:middle line:84% And that's, in some sense, the whole privatization debate 00:15:57.000 --> 00:15:59.860 align:middle line:84% we got into chapter 7, which is it better to let 00:15:59.860 --> 00:16:01.700 align:middle line:84% a private company do it and try to monitor 00:16:01.700 --> 00:16:03.367 align:middle line:84% the quality and the product, or you have 00:16:03.367 --> 00:16:06.620 align:middle line:90% the government do it itself? 00:16:06.620 --> 00:16:07.120 align:middle line:90% What else? 00:16:07.120 --> 00:16:08.662 align:middle line:84% What's another problem with vouchers? 00:16:08.662 --> 00:16:13.260 align:middle line:90% 00:16:13.260 --> 00:16:13.760 align:middle line:90% Yeah. 00:16:13.760 --> 00:16:15.740 align:middle line:84% AUDIENCE: [INAUDIBLE] equity issue. 00:16:15.740 --> 00:16:18.320 align:middle line:84% So you're paying those who are already going [INAUDIBLE]. 00:16:18.320 --> 00:16:20.860 align:middle line:84% JON GRUBER: Awesome, great point. 00:16:20.860 --> 00:16:22.980 align:middle line:84% Think about the equity implications. 00:16:22.980 --> 00:16:26.260 align:middle line:90% Who is likely to be person Z? 00:16:26.260 --> 00:16:27.780 align:middle line:90% A rich person. 00:16:27.780 --> 00:16:29.740 align:middle line:90% Who's likely to person X? 00:16:29.740 --> 00:16:31.058 align:middle line:90% A poor person. 00:16:31.058 --> 00:16:33.100 align:middle line:84% Now, in this case, it's not true because they all 00:16:33.100 --> 00:16:34.200 align:middle line:90% have the same budget. 00:16:34.200 --> 00:16:37.040 align:middle line:84% But in general, who sends their kids to private schools 00:16:37.040 --> 00:16:38.320 align:middle line:90% is richer people. 00:16:38.320 --> 00:16:40.440 align:middle line:84% And for those people, we are just 00:16:40.440 --> 00:16:42.860 align:middle line:84% paying them to do something they were already going to do. 00:16:42.860 --> 00:16:45.400 align:middle line:84% We're inducing extra education, but not 00:16:45.400 --> 00:16:48.920 align:middle line:84% through changing their behavior just by giving them more money. 00:16:48.920 --> 00:16:51.640 align:middle line:84% We are going to spend a lot of the taxation 00:16:51.640 --> 00:16:56.000 align:middle line:84% section of this course talking about the concept of marginal 00:16:56.000 --> 00:16:58.680 align:middle line:90% versus inframarginal consumers. 00:16:58.680 --> 00:17:01.520 align:middle line:84% A marginal consumer is one whose behavior 00:17:01.520 --> 00:17:05.599 align:middle line:84% is changed by a change in government policy. 00:17:05.599 --> 00:17:07.901 align:middle line:84% An inframarginal consumer is one who 00:17:07.901 --> 00:17:09.359 align:middle line:84% benefits from the government policy 00:17:09.359 --> 00:17:11.493 align:middle line:90% without their behavior changing. 00:17:11.493 --> 00:17:13.160 align:middle line:84% When you put in a policy that's designed 00:17:13.160 --> 00:17:16.400 align:middle line:84% to change behavior, i.e., to increase education, 00:17:16.400 --> 00:17:19.258 align:middle line:84% then all the money you give to rich people who are already 00:17:19.258 --> 00:17:21.300 align:middle line:84% send their kids to private schools inframarginal. 00:17:21.300 --> 00:17:23.467 align:middle line:84% It's just a waste from the government's perspective. 00:17:23.467 --> 00:17:25.480 align:middle line:90% It's only redistributional. 00:17:25.480 --> 00:17:29.042 align:middle line:84% Marginal changes have efficiency consequences. 00:17:29.042 --> 00:17:30.500 align:middle line:84% If I induce someone to go to school 00:17:30.500 --> 00:17:33.100 align:middle line:84% who isn't going to school, that's a marginal change. 00:17:33.100 --> 00:17:35.220 align:middle line:84% That is an efficiency consequence. 00:17:35.220 --> 00:17:37.860 align:middle line:84% If I take someone who's already going to school 00:17:37.860 --> 00:17:41.060 align:middle line:84% and just give them money, that is an inframarginal change. 00:17:41.060 --> 00:17:45.900 align:middle line:84% It's redistributive, but it has no efficiency consequences. 00:17:45.900 --> 00:17:48.045 align:middle line:84% Look, more people in group Z go to school, 00:17:48.045 --> 00:17:49.920 align:middle line:84% but they go to school because they're richer. 00:17:49.920 --> 00:17:51.140 align:middle line:84% We could have more people go to Z school 00:17:51.140 --> 00:17:52.973 align:middle line:84% by just redistributing money to rich people. 00:17:52.973 --> 00:17:54.860 align:middle line:84% We wouldn't have to have vouchers. 00:17:54.860 --> 00:17:58.090 align:middle line:84% So basically, if we think about marginal, 00:17:58.090 --> 00:18:00.940 align:middle line:84% inframarginal the marginal piece is about efficiency. 00:18:00.940 --> 00:18:04.220 align:middle line:84% The inframarginal piece is about redistribution. 00:18:04.220 --> 00:18:07.240 align:middle line:84% So you have to ask yourself, well, who are redistributing to? 00:18:07.240 --> 00:18:08.757 align:middle line:84% Well, vouchers, we're redistributing 00:18:08.757 --> 00:18:11.340 align:middle line:84% towards rich people because they were the ones already sending 00:18:11.340 --> 00:18:13.860 align:middle line:90% their kids to private school. 00:18:13.860 --> 00:18:17.740 align:middle line:84% So it becomes an inefficient use of public dollars 00:18:17.740 --> 00:18:20.700 align:middle line:84% to offer a voucher, because yes, you 00:18:20.700 --> 00:18:23.900 align:middle line:84% might induce some poor people to get to go to private school. 00:18:23.900 --> 00:18:26.420 align:middle line:84% But you're also buying out rich people already sending 00:18:26.420 --> 00:18:27.900 align:middle line:90% their kids to private school. 00:18:27.900 --> 00:18:30.400 align:middle line:84% Now, in practice, once again, you could deal with this. 00:18:30.400 --> 00:18:34.470 align:middle line:84% You could have income-related vouchers. 00:18:34.470 --> 00:18:36.930 align:middle line:84% And in practice, that is something you could approach. 00:18:36.930 --> 00:18:38.430 align:middle line:84% But in some sense, then the question 00:18:38.430 --> 00:18:40.190 align:middle line:90% is, where's the income cutoff? 00:18:40.190 --> 00:18:44.310 align:middle line:84% And really, remember, on this x-axis, it is not income. 00:18:44.310 --> 00:18:46.083 align:middle line:84% It's desired educational spending. 00:18:46.083 --> 00:18:47.750 align:middle line:84% So you really like to target the voucher 00:18:47.750 --> 00:18:50.870 align:middle line:84% by who really needs to push to send 00:18:50.870 --> 00:18:53.790 align:middle line:84% their kid to private school versus who's already sending 00:18:53.790 --> 00:18:54.970 align:middle line:90% their kid to private school. 00:18:54.970 --> 00:18:58.510 align:middle line:84% Well, that's correlated with income, but not perfectly. 00:18:58.510 --> 00:19:01.090 align:middle line:84% So cutting by income doesn't really solve the problem. 00:19:01.090 --> 00:19:03.750 align:middle line:84% You really want to cut by underlying educational spending 00:19:03.750 --> 00:19:04.710 align:middle line:90% desire. 00:19:04.710 --> 00:19:08.750 align:middle line:84% And that's immeasurable So another problem 00:19:08.750 --> 00:19:09.810 align:middle line:90% is the equity problem. 00:19:09.810 --> 00:19:10.310 align:middle line:90% Yeah. 00:19:10.310 --> 00:19:12.150 align:middle line:84% AUDIENCE: Aside from, like, whether or not 00:19:12.150 --> 00:19:14.730 align:middle line:84% their children are currently enrolled in a private school, 00:19:14.730 --> 00:19:17.150 align:middle line:84% are there any [INAUDIBLE] preference metrics [INAUDIBLE] 00:19:17.150 --> 00:19:19.170 align:middle line:90% might approach that? 00:19:19.170 --> 00:19:20.530 align:middle line:90% JON GRUBER: Not really. 00:19:20.530 --> 00:19:23.990 align:middle line:84% I mean, no, because actually, the truth is private school 00:19:23.990 --> 00:19:27.670 align:middle line:84% attendance-- while I parodied the private school 00:19:27.670 --> 00:19:31.430 align:middle line:84% attendance that we think about is Andover and Exeter. 00:19:31.430 --> 00:19:33.330 align:middle line:84% But 90% of kids in private school in America 00:19:33.330 --> 00:19:35.955 align:middle line:84% go to parochial schools, which are actually not that expensive, 00:19:35.955 --> 00:19:38.030 align:middle line:84% and actually attended not by the poorest, 00:19:38.030 --> 00:19:40.450 align:middle line:84% but typically often not by the richest either. 00:19:40.450 --> 00:19:44.170 align:middle line:84% So in some sense it is hard to actually income is actually 00:19:44.170 --> 00:19:46.390 align:middle line:90% a reason is not a great proxy. 00:19:46.390 --> 00:19:48.850 align:middle line:84% It's correlated with private school demand and education, 00:19:48.850 --> 00:19:51.993 align:middle line:84% but it's not actually an outstanding proxy for it. 00:19:51.993 --> 00:19:54.410 align:middle line:84% And so basically, the problem is you deal with the equity, 00:19:54.410 --> 00:19:55.850 align:middle line:84% but you don't really get the vouchers 00:19:55.850 --> 00:19:57.010 align:middle line:90% who you want them to go to. 00:19:57.010 --> 00:20:00.122 align:middle line:84% You want to induce the people who-- you want to get them to x. 00:20:00.122 --> 00:20:02.330 align:middle line:84% You want the people who you're going to most increase 00:20:02.330 --> 00:20:04.490 align:middle line:84% their educational attainment, but you don't know 00:20:04.490 --> 00:20:06.210 align:middle line:90% how to target them that way. 00:20:06.210 --> 00:20:07.148 align:middle line:90% Does that make sense? 00:20:07.148 --> 00:20:08.690 align:middle line:84% That's a very important concept we'll 00:20:08.690 --> 00:20:11.232 align:middle line:84% come back to over and over again when we talk about taxation. 00:20:11.232 --> 00:20:13.450 align:middle line:90% 00:20:13.450 --> 00:20:15.930 align:middle line:84% One other problem that's very important 00:20:15.930 --> 00:20:21.170 align:middle line:90% here is, what about segregation? 00:20:21.170 --> 00:20:24.530 align:middle line:84% And segregation is interesting because it's 00:20:24.530 --> 00:20:25.950 align:middle line:90% on both sides of the line. 00:20:25.950 --> 00:20:31.510 align:middle line:90% 00:20:31.510 --> 00:20:34.050 align:middle line:84% Voucher opponents say, look, what's going to happen? 00:20:34.050 --> 00:20:37.253 align:middle line:84% Well, all the motivated families got their shit together 00:20:37.253 --> 00:20:38.170 align:middle line:90% and pulled their kids. 00:20:38.170 --> 00:20:41.470 align:middle line:90% It's related to the first point. 00:20:41.470 --> 00:20:44.390 align:middle line:84% All the motivated families that had their shit 00:20:44.390 --> 00:20:46.690 align:middle line:84% together pulled their kids out of the public schools, 00:20:46.690 --> 00:20:48.670 align:middle line:90% put them in the private schools. 00:20:48.670 --> 00:20:51.370 align:middle line:84% The ones who are left behind are the less motivated, 00:20:51.370 --> 00:20:54.270 align:middle line:84% the people with less odds of succeeding anyway. 00:20:54.270 --> 00:20:56.190 align:middle line:84% And so you're further segregating society 00:20:56.190 --> 00:20:59.270 align:middle line:84% into those who are likely to succeed and those who are not. 00:20:59.270 --> 00:21:03.910 align:middle line:84% Now, supporters of vouchers actually recognized 00:21:03.910 --> 00:21:06.310 align:middle line:84% that argument and said, well, that's actually bullshit 00:21:06.310 --> 00:21:08.518 align:middle line:84% because we already have incredibly segregated schools 00:21:08.518 --> 00:21:09.910 align:middle line:90% in America today. 00:21:09.910 --> 00:21:16.870 align:middle line:84% So the fact is that the typical white student attends 00:21:16.870 --> 00:21:22.110 align:middle line:84% a school that's 75% white, but 40% of minority students 00:21:22.110 --> 00:21:25.990 align:middle line:84% attend schools that are 90% to 100% minority. 00:21:25.990 --> 00:21:32.990 align:middle line:84% So basically, this is really the idea that, look, 00:21:32.990 --> 00:21:35.730 align:middle line:84% we're already segregated by race. 00:21:35.730 --> 00:21:37.650 align:middle line:84% Maybe segregation by ability isn't so bad 00:21:37.650 --> 00:21:39.787 align:middle line:84% if it allows a poor Black kid who's 00:21:39.787 --> 00:21:41.370 align:middle line:84% stuck in a minority school-- but who's 00:21:41.370 --> 00:21:43.430 align:middle line:84% really able, his parents want him to succeed, 00:21:43.430 --> 00:21:46.875 align:middle line:84% to go to a private school with other white kids. 00:21:46.875 --> 00:21:48.750 align:middle line:84% Then yeah, you're separating more by ability, 00:21:48.750 --> 00:21:50.570 align:middle line:84% but you're separating less by race. 00:21:50.570 --> 00:21:55.570 align:middle line:84% So this is a tough one to evaluate. 00:21:55.570 --> 00:21:57.730 align:middle line:84% That's the pros and cons of vouchers. 00:21:57.730 --> 00:22:00.677 align:middle line:84% This is how I want you to learn to think about problems. 00:22:00.677 --> 00:22:03.010 align:middle line:84% There's almost nothing we learn about in this class that 00:22:03.010 --> 00:22:04.770 align:middle line:90% has an easy answer. 00:22:04.770 --> 00:22:07.832 align:middle line:90% This is not E equals mc squared. 00:22:07.832 --> 00:22:09.790 align:middle line:84% There's almost nothing-- not that it's so easy. 00:22:09.790 --> 00:22:11.488 align:middle line:90% But there's a formula. 00:22:11.488 --> 00:22:13.030 align:middle line:84% So it's not [INAUDIBLE] just formula. 00:22:13.030 --> 00:22:14.072 align:middle line:90% There's always tradeoffs. 00:22:14.072 --> 00:22:15.593 align:middle line:90% And we want to think about how-- 00:22:15.593 --> 00:22:17.010 align:middle line:84% want to get you trained up and how 00:22:17.010 --> 00:22:20.170 align:middle line:84% to think about those tradeoffs OK. 00:22:20.170 --> 00:22:20.752 align:middle line:90% Yeah, Enoch. 00:22:20.752 --> 00:22:23.210 align:middle line:84% AUDIENCE: Is there any evidence, outside of moral argument, 00:22:23.210 --> 00:22:25.730 align:middle line:90% segregation is a bad thing? 00:22:25.730 --> 00:22:31.510 align:middle line:84% JON GRUBER: Actually, it's not clear-- well, 00:22:31.510 --> 00:22:33.410 align:middle line:84% two types of literature on segregation. 00:22:33.410 --> 00:22:34.990 align:middle line:84% So one type is you might specifically 00:22:34.990 --> 00:22:38.570 align:middle line:84% say, segregation in education, is that bad? 00:22:38.570 --> 00:22:40.730 align:middle line:84% The evidence there is mixed on that one, 00:22:40.730 --> 00:22:43.630 align:middle line:84% because sometimes putting students who are very high 00:22:43.630 --> 00:22:46.498 align:middle line:84% achieving together has positive externalities. 00:22:46.498 --> 00:22:48.790 align:middle line:84% And having a low-achieving kid with high-achieving kids 00:22:48.790 --> 00:22:50.915 align:middle line:84% can really have a big negative externality on them, 00:22:50.915 --> 00:22:52.550 align:middle line:90% so that evidence is mixed. 00:22:52.550 --> 00:22:55.072 align:middle line:84% There's more general evidence of segregation in society, 00:22:55.072 --> 00:22:56.530 align:middle line:84% and that evidence is pretty uniform 00:22:56.530 --> 00:22:58.630 align:middle line:84% that's bad for the people who get left behind. 00:22:58.630 --> 00:23:00.030 align:middle line:84% There's not great evidence on how good 00:23:00.030 --> 00:23:01.950 align:middle line:84% it is for people who don't get left behind. 00:23:01.950 --> 00:23:03.950 align:middle line:84% The trick with all this stuff is you really want 00:23:03.950 --> 00:23:05.190 align:middle line:90% to care about social welfare. 00:23:05.190 --> 00:23:07.732 align:middle line:84% It's pretty clear it's bad for the folks who get left behind. 00:23:07.732 --> 00:23:10.430 align:middle line:84% It's less clear if it's overall bad for society or not. 00:23:10.430 --> 00:23:12.430 align:middle line:84% One sense in which it is overall bad for society 00:23:12.430 --> 00:23:14.563 align:middle line:84% is a lot of political economy research, which 00:23:14.563 --> 00:23:16.230 align:middle line:84% shows that more segregated societies are 00:23:16.230 --> 00:23:18.950 align:middle line:84% less likely to provide public goods that benefit everyone. 00:23:18.950 --> 00:23:20.990 align:middle line:84% Because if you don't know the other folk, 00:23:20.990 --> 00:23:21.970 align:middle line:84% if you're not with the other folk, 00:23:21.970 --> 00:23:23.303 align:middle line:90% you don't want to help them out. 00:23:23.303 --> 00:23:27.610 align:middle line:84% So we talked about social capital and altruism. 00:23:27.610 --> 00:23:29.730 align:middle line:84% That is, there's less social capital 00:23:29.730 --> 00:23:32.570 align:middle line:90% and more segregated societies. 00:23:32.570 --> 00:23:34.850 align:middle line:90% Other questions? 00:23:34.850 --> 00:23:39.230 align:middle line:84% Now, when you're in a situation like this, what do you do? 00:23:39.230 --> 00:23:40.772 align:middle line:84% You call up the empirical economists. 00:23:40.772 --> 00:23:43.355 align:middle line:84% And you say, well, we don't know what's right or what's wrong. 00:23:43.355 --> 00:23:43.890 align:middle line:90% You tell us. 00:23:43.890 --> 00:23:45.723 align:middle line:84% So that's what empirical economists come in. 00:23:45.723 --> 00:23:47.390 align:middle line:90% So we have a lot of evidence. 00:23:47.390 --> 00:23:49.730 align:middle line:84% There's been a huge literature built up 00:23:49.730 --> 00:23:54.370 align:middle line:84% on thinking about the effects of reform 00:23:54.370 --> 00:23:56.530 align:middle line:90% in education markets writ large. 00:23:56.530 --> 00:24:01.610 align:middle line:84% Now, we don't have a ton of US evidence on vouchers. 00:24:01.610 --> 00:24:02.950 align:middle line:90% We have a little evidence. 00:24:02.950 --> 00:24:06.690 align:middle line:84% The evidence we have actually comes from a very clever, 00:24:06.690 --> 00:24:09.210 align:middle line:84% generalized approach that we use in empirical economics 00:24:09.210 --> 00:24:12.010 align:middle line:84% these days, which is what's called the waiting list design 00:24:12.010 --> 00:24:14.810 align:middle line:90% or randomized lottery design. 00:24:14.810 --> 00:24:17.970 align:middle line:84% The way vouchers work when they're assigned 00:24:17.970 --> 00:24:21.130 align:middle line:84% is essentially, too many kids want the voucher relative to how 00:24:21.130 --> 00:24:22.378 align:middle line:90% many vouchers are available. 00:24:22.378 --> 00:24:23.170 align:middle line:90% They'll limit them. 00:24:23.170 --> 00:24:27.430 align:middle line:84% To limit the fiscal cost to the government, they'll limit them. 00:24:27.430 --> 00:24:28.800 align:middle line:90% Too many kids want it. 00:24:28.800 --> 00:24:30.550 align:middle line:84% So what they'll do is decide who gets them 00:24:30.550 --> 00:24:33.585 align:middle line:90% by lottery, to be fair. 00:24:33.585 --> 00:24:35.210 align:middle line:84% Once you decide who gets it by lottery, 00:24:35.210 --> 00:24:36.918 align:middle line:84% you set up a randomized trial of the type 00:24:36.918 --> 00:24:39.150 align:middle line:90% we discussed in chapter 3. 00:24:39.150 --> 00:24:42.190 align:middle line:90% So you have a lottery. 00:24:42.190 --> 00:24:44.990 align:middle line:84% And Paul wins and Valerie loses, then 00:24:44.990 --> 00:24:48.630 align:middle line:84% you have two identical people, one of whom won and one of whom 00:24:48.630 --> 00:24:49.487 align:middle line:90% lost. 00:24:49.487 --> 00:24:51.070 align:middle line:84% By random chance, you can compare them 00:24:51.070 --> 00:24:52.987 align:middle line:84% and see what happens to them five years later. 00:24:52.987 --> 00:24:54.910 align:middle line:84% Was it good for them or bad for them? 00:24:54.910 --> 00:24:57.710 align:middle line:84% So this randomized lottery design 00:24:57.710 --> 00:24:59.810 align:middle line:84% is essentially a way that society gives you 00:24:59.810 --> 00:25:02.310 align:middle line:84% a true randomized control trial without you having to set it 00:25:02.310 --> 00:25:03.840 align:middle line:90% up and run it yourself. 00:25:03.840 --> 00:25:05.590 align:middle line:84% And that method has been used in many ways 00:25:05.590 --> 00:25:07.790 align:middle line:90% to study voucher programs. 00:25:07.790 --> 00:25:13.912 align:middle line:84% I think the evidence in the US is positive but modest. 00:25:13.912 --> 00:25:15.870 align:middle line:84% I think the evidence [INAUDIBLE] be interpreted 00:25:15.870 --> 00:25:19.950 align:middle line:90% as positive but modest. 00:25:19.950 --> 00:25:22.650 align:middle line:84% The more compelling evidence comes from other countries 00:25:22.650 --> 00:25:24.923 align:middle line:84% where voucher programs are much, much larger. 00:25:24.923 --> 00:25:27.090 align:middle line:84% My colleague Josh Angrist did a very important paper 00:25:27.090 --> 00:25:31.590 align:middle line:84% on Colombia, which covered a large share of the population, 00:25:31.590 --> 00:25:33.850 align:middle line:84% and he found very large positive effects 00:25:33.850 --> 00:25:38.450 align:middle line:84% on the lottery winners versus the lottery losers. 00:25:38.450 --> 00:25:40.918 align:middle line:84% So there is some evidence from other countries 00:25:40.918 --> 00:25:43.210 align:middle line:84% where the lotteries are quite large where it's a more-- 00:25:43.210 --> 00:25:45.022 align:middle line:84% and the US voucher advocates would argue, 00:25:45.022 --> 00:25:46.730 align:middle line:84% well, you don't see big effects in the US 00:25:46.730 --> 00:25:49.770 align:middle line:84% because you can't get this effect when just a few people 00:25:49.770 --> 00:25:50.900 align:middle line:90% are getting vouchers. 00:25:50.900 --> 00:25:52.650 align:middle line:84% The only way to test vouchers is give them 00:25:52.650 --> 00:25:55.890 align:middle line:84% to a huge number of people, so you let the market react. 00:25:55.890 --> 00:25:58.150 align:middle line:84% We have what we call in economics partial equilibrium 00:25:58.150 --> 00:25:59.692 align:middle line:84% and general equilibrium, another term 00:25:59.692 --> 00:26:00.900 align:middle line:90% we'll talk about in taxation. 00:26:00.900 --> 00:26:02.567 align:middle line:84% Partial equilibria is holding everything 00:26:02.567 --> 00:26:04.872 align:middle line:84% else constant, what happens general equilibrium is 00:26:04.872 --> 00:26:06.830 align:middle line:84% what happens not holding other things constant. 00:26:06.830 --> 00:26:08.497 align:middle line:84% And voucher advocates would argue, well, 00:26:08.497 --> 00:26:10.130 align:middle line:84% if you give 100 guys vouchers and look 00:26:10.130 --> 00:26:11.930 align:middle line:84% at what happens, that doesn't tell you 00:26:11.930 --> 00:26:13.305 align:middle line:84% whether it's good or bad for them 00:26:13.305 --> 00:26:15.570 align:middle line:84% because you don't get the market response. 00:26:15.570 --> 00:26:16.410 align:middle line:90% Yeah, question. 00:26:16.410 --> 00:26:19.230 align:middle line:84% AUDIENCE: From the [? global ?] literature, 00:26:19.230 --> 00:26:21.740 align:middle line:84% is there any a correlation that you're 00:26:21.740 --> 00:26:27.180 align:middle line:84% aware of from [INAUDIBLE] size of the-- the magnitude 00:26:27.180 --> 00:26:29.682 align:middle line:84% of the voucher compared to an individual income? 00:26:29.682 --> 00:26:32.140 align:middle line:84% JON GRUBER: Yeah, I mean, I think they're not at that level 00:26:32.140 --> 00:26:32.640 align:middle line:90% yet. 00:26:32.640 --> 00:26:35.300 align:middle line:84% This is still just voucher, no voucher. 00:26:35.300 --> 00:26:42.540 align:middle line:84% Now, this does, however, raise a very interesting question, which 00:26:42.540 --> 00:26:45.980 align:middle line:84% is, when you use this lottery study, 00:26:45.980 --> 00:26:49.040 align:middle line:84% you get the effect on the winners versus the losers. 00:26:49.040 --> 00:26:51.620 align:middle line:84% But you don't get the effect on society, 00:26:51.620 --> 00:26:53.328 align:middle line:84% just like I was saying about segregation. 00:26:53.328 --> 00:26:55.370 align:middle line:84% What you care about is whether society as a whole 00:26:55.370 --> 00:26:56.080 align:middle line:90% is better off. 00:26:56.080 --> 00:26:57.320 align:middle line:84% So yeah, the winners may be better off. 00:26:57.320 --> 00:26:58.580 align:middle line:90% But are the losers worse off? 00:26:58.580 --> 00:27:00.020 align:middle line:84% Are the losers worse off in a way 00:27:00.020 --> 00:27:03.300 align:middle line:84% that offsets the gains to the winners? 00:27:03.300 --> 00:27:05.620 align:middle line:84% There's starting to be some literature on this. 00:27:05.620 --> 00:27:08.020 align:middle line:84% In particular, there's a reasonably sizable voucher 00:27:08.020 --> 00:27:11.280 align:middle line:84% program in Florida, and the evidence suggests no. 00:27:11.280 --> 00:27:14.380 align:middle line:84% Overall, the whole cohort is better off [INAUDIBLE] 00:27:14.380 --> 00:27:16.700 align:middle line:84% school choice, maybe because of the competition, 00:27:16.700 --> 00:27:17.440 align:middle line:90% however it works. 00:27:17.440 --> 00:27:19.480 align:middle line:90% Maybe because of better sorting. 00:27:19.480 --> 00:27:21.320 align:middle line:84% In other words, segregation is sorting. 00:27:21.320 --> 00:27:22.920 align:middle line:90% And that can be good or bad. 00:27:22.920 --> 00:27:25.240 align:middle line:84% Whatever the reason, the evidence 00:27:25.240 --> 00:27:29.600 align:middle line:84% seems to be that overall, the system improves. 00:27:29.600 --> 00:27:32.560 align:middle line:84% So with vouchers, the existing evidence 00:27:32.560 --> 00:27:36.760 align:middle line:84% seems to suggest there be improved educational outcomes, 00:27:36.760 --> 00:27:40.240 align:middle line:84% but at a cost of clearly lower equity 00:27:40.240 --> 00:27:42.797 align:middle line:84% and the other concerns that are raised. 00:27:42.797 --> 00:27:44.880 align:middle line:84% The question is, are there other alternatives that 00:27:44.880 --> 00:27:45.880 align:middle line:90% might do better? 00:27:45.880 --> 00:27:48.080 align:middle line:84% And we've tried a number of alternatives. 00:27:48.080 --> 00:27:51.920 align:middle line:84% One alternative is, instead of having 00:27:51.920 --> 00:27:55.800 align:middle line:84% vouchers, have free choice, but only among public schools. 00:27:55.800 --> 00:27:58.780 align:middle line:84% That is, try to get these benefits, 00:27:58.780 --> 00:28:01.080 align:middle line:84% but use it to send incentives across public schools, 00:28:01.080 --> 00:28:04.280 align:middle line:84% rather than subsidizing private entities. 00:28:04.280 --> 00:28:08.460 align:middle line:84% That's been done a lot, and there's two forms of that. 00:28:08.460 --> 00:28:12.160 align:middle line:84% One form is cities like Chicago or Cambridge, 00:28:12.160 --> 00:28:14.560 align:middle line:84% where they let you go to any elementary school you 00:28:14.560 --> 00:28:16.935 align:middle line:84% want in the city or any high school you want in the city. 00:28:16.935 --> 00:28:18.760 align:middle line:84% So where my kids grew up in Lexington, 00:28:18.760 --> 00:28:20.680 align:middle line:84% you go to the local elementary school and the middle school, 00:28:20.680 --> 00:28:21.060 align:middle line:90% and the high school. 00:28:21.060 --> 00:28:21.940 align:middle line:90% It's all assigned to you. 00:28:21.940 --> 00:28:23.860 align:middle line:84% In these cities, you go to any one you want. 00:28:23.860 --> 00:28:25.943 align:middle line:84% And you apply and you go to the best one you want. 00:28:25.943 --> 00:28:27.568 align:middle line:84% And once again, there's a waiting list, 00:28:27.568 --> 00:28:28.860 align:middle line:90% so we can study what happens. 00:28:28.860 --> 00:28:30.937 align:middle line:90% And the answer is not a lot. 00:28:30.937 --> 00:28:33.020 align:middle line:84% It doesn't seem like letting people shuffle around 00:28:33.020 --> 00:28:36.220 align:middle line:84% like that makes that much difference. 00:28:36.220 --> 00:28:38.220 align:middle line:84% The other big change here that's made 00:28:38.220 --> 00:28:41.660 align:middle line:84% is the rise of charter and magnet schools, which 00:28:41.660 --> 00:28:44.300 align:middle line:84% are essentially public schools kind 00:28:44.300 --> 00:28:48.860 align:middle line:84% of outside the existing regulatory and union guidelines. 00:28:48.860 --> 00:28:53.020 align:middle line:84% So essentially, a charter school is typically 00:28:53.020 --> 00:28:55.540 align:middle line:90% private, often for-profit. 00:28:55.540 --> 00:28:57.180 align:middle line:84% But they are funded by the state, 00:28:57.180 --> 00:29:00.640 align:middle line:84% and they're still considered public schools, 00:29:00.640 --> 00:29:04.020 align:middle line:84% although in some sense they're halfway in between. 00:29:04.020 --> 00:29:05.980 align:middle line:84% So they're not private in the sense 00:29:05.980 --> 00:29:07.660 align:middle line:84% they're heavily regulated by the state 00:29:07.660 --> 00:29:11.140 align:middle line:84% and they meet certain state standards. 00:29:11.140 --> 00:29:14.337 align:middle line:84% They are private in the sense that they are not 00:29:14.337 --> 00:29:16.920 align:middle line:84% subject to some other rules that public schools are subbed to, 00:29:16.920 --> 00:29:22.480 align:middle line:84% in particular unionization of teachers and other standards. 00:29:22.480 --> 00:29:26.520 align:middle line:84% And once again, people get into charter schools by lottery. 00:29:26.520 --> 00:29:27.760 align:middle line:90% There's a huge literature. 00:29:27.760 --> 00:29:29.468 align:middle line:84% And actually, the world leaders from that 00:29:29.468 --> 00:29:31.560 align:middle line:84% are my colleagues Josh Angrist and Parag Pathak. 00:29:31.560 --> 00:29:33.227 align:middle line:84% At Blueprint Labs here at MIT, they've 00:29:33.227 --> 00:29:35.560 align:middle line:84% done dozens of studies looking at the effect-- maybe not 00:29:35.560 --> 00:29:37.977 align:middle line:84% dozens-- but many studies looking at the effect of charter 00:29:37.977 --> 00:29:38.640 align:middle line:90% schools. 00:29:38.640 --> 00:29:43.900 align:middle line:84% And I would say the evidence is once again, on average, 00:29:43.900 --> 00:29:45.360 align:middle line:90% positive. 00:29:45.360 --> 00:29:48.360 align:middle line:84% The evidence seems to be that charter schools benefit 00:29:48.360 --> 00:29:49.580 align:middle line:90% the kids who go there. 00:29:49.580 --> 00:29:51.160 align:middle line:90% And here's the exciting part. 00:29:51.160 --> 00:29:54.440 align:middle line:84% The benefits are the largest for the least advantaged kids. 00:29:54.440 --> 00:29:58.520 align:middle line:84% That's why it's more exciting than the voucher literature, 00:29:58.520 --> 00:30:00.560 align:middle line:84% because without giving extra money to people 00:30:00.560 --> 00:30:02.660 align:middle line:84% who are already going to private school, you're not giving away. 00:30:02.660 --> 00:30:04.500 align:middle line:84% You're just taking people and shuffling them across schools. 00:30:04.500 --> 00:30:06.182 align:middle line:84% So no one's getting extra money here. 00:30:06.182 --> 00:30:08.640 align:middle line:84% But you're doing so in a way which actually improves equity 00:30:08.640 --> 00:30:10.280 align:middle line:84% because the charter schools are better for the more 00:30:10.280 --> 00:30:11.420 align:middle line:90% disadvantaged students. 00:30:11.420 --> 00:30:12.800 align:middle line:90% Why is that? 00:30:12.800 --> 00:30:16.600 align:middle line:84% Well, the evidence once again is not determinant. 00:30:16.600 --> 00:30:19.540 align:middle line:84% But it seems to be, there's a set of things that correlate 00:30:19.540 --> 00:30:23.580 align:middle line:84% with success in school, most importantly things like stronger 00:30:23.580 --> 00:30:28.595 align:middle line:84% discipline, stronger standards, wearing uniforms, more norms, 00:30:28.595 --> 00:30:30.220 align:middle line:84% that basically things seem to correlate 00:30:30.220 --> 00:30:32.260 align:middle line:90% with success in school. 00:30:32.260 --> 00:30:36.420 align:middle line:84% Basically, there's more zero tolerance strategies. 00:30:36.420 --> 00:30:41.060 align:middle line:84% Now, that raises separate issues of equity and fairness, 00:30:41.060 --> 00:30:43.660 align:middle line:90% et cetera. 00:30:43.660 --> 00:30:46.320 align:middle line:84% But that is what seems to be the reason that they do better. 00:30:46.320 --> 00:30:46.820 align:middle line:90% Yeah. 00:30:46.820 --> 00:30:48.900 align:middle line:84% AUDIENCE: And those strategies aren't imposed 00:30:48.900 --> 00:30:50.172 align:middle line:90% in normal public schools. 00:30:50.172 --> 00:30:52.380 align:middle line:84% JON GRUBER: No, they're not imposed in public schools 00:30:52.380 --> 00:30:55.500 align:middle line:84% because parents would bitch about it or whatever. 00:30:55.500 --> 00:30:56.870 align:middle line:90% AUDIENCE: But it's not illegal. 00:30:56.870 --> 00:30:58.120 align:middle line:90% JON GRUBER: It could be legal. 00:30:58.120 --> 00:31:00.965 align:middle line:84% The school could be worried about getting sued, cetera. 00:31:00.965 --> 00:31:01.840 align:middle line:90% It's a good question. 00:31:01.840 --> 00:31:03.740 align:middle line:84% I don't know why-- but some public schools 00:31:03.740 --> 00:31:05.580 align:middle line:90% are moving towards that. 00:31:05.580 --> 00:31:08.440 align:middle line:84% But really, the charter schools have been the lead on that. 00:31:08.440 --> 00:31:08.940 align:middle line:90% Yeah. 00:31:08.940 --> 00:31:11.280 align:middle line:84% AUDIENCE: [INAUDIBLE] like teacher quality channel, 00:31:11.280 --> 00:31:12.740 align:middle line:90% is there anything [INAUDIBLE]? 00:31:12.740 --> 00:31:14.203 align:middle line:84% JON GRUBER: That's a great point. 00:31:14.203 --> 00:31:16.120 align:middle line:84% The other thing that charter schools have done 00:31:16.120 --> 00:31:20.320 align:middle line:84% is moved towards more innovative ways of hiring and recognizing 00:31:20.320 --> 00:31:22.040 align:middle line:90% teacher quality. 00:31:22.040 --> 00:31:25.180 align:middle line:84% So once again, because they're free, they're more flexible. 00:31:25.180 --> 00:31:26.180 align:middle line:90% There's no unionization. 00:31:26.180 --> 00:31:28.180 align:middle line:84% They're more flexible in terms of hire and fire. 00:31:28.180 --> 00:31:30.240 align:middle line:84% They'll do things like hiring people 00:31:30.240 --> 00:31:32.298 align:middle line:84% who are not trained as teachers, but have shown 00:31:32.298 --> 00:31:33.840 align:middle line:84% great aptitude for that set of skills 00:31:33.840 --> 00:31:36.760 align:middle line:84% and give them a shot as a teacher. 00:31:36.760 --> 00:31:38.555 align:middle line:84% So there's more flexibility and more effort 00:31:38.555 --> 00:31:40.680 align:middle line:84% to try to find people who'll be better in teaching, 00:31:40.680 --> 00:31:43.263 align:middle line:84% rather than people who've just traditionally been in teaching. 00:31:43.263 --> 00:31:45.140 align:middle line:90% So that's a good point. 00:31:45.140 --> 00:31:46.840 align:middle line:84% It's another thing they're doing. 00:31:46.840 --> 00:31:49.160 align:middle line:90% So this has been successful. 00:31:49.160 --> 00:31:53.260 align:middle line:84% Now, once again, here's a critical thing-- 00:31:53.260 --> 00:31:55.120 align:middle line:90% another big lesson for today. 00:31:55.120 --> 00:32:00.000 align:middle line:84% We always have to think about two views of evidence. 00:32:00.000 --> 00:32:02.920 align:middle line:90% There's data and anecdote. 00:32:02.920 --> 00:32:05.980 align:middle line:84% There are also plenty of shitty charter schools. 00:32:05.980 --> 00:32:08.540 align:middle line:84% If you're John Oliver fans, he's the king of the anecdote. 00:32:08.540 --> 00:32:10.540 align:middle line:84% If you saw John Oliver piece on charter schools, 00:32:10.540 --> 00:32:12.748 align:middle line:84% you think they're the worst thing that ever happened. 00:32:12.748 --> 00:32:15.540 align:middle line:84% It's because there are some bad ones. 00:32:15.540 --> 00:32:18.620 align:middle line:84% That's not the way we, as social welfare maximizers, 00:32:18.620 --> 00:32:20.500 align:middle line:90% should think about the problem. 00:32:20.500 --> 00:32:21.900 align:middle line:90% I think he's hilarious guy. 00:32:21.900 --> 00:32:23.580 align:middle line:84% But I've gotten kind of sick of him, 00:32:23.580 --> 00:32:25.080 align:middle line:84% because basically, he's not thinking 00:32:25.080 --> 00:32:27.480 align:middle line:84% about the way-- social impacts should be in aggregate, 00:32:27.480 --> 00:32:28.360 align:middle line:90% it's a good thing or not. 00:32:28.360 --> 00:32:30.460 align:middle line:84% Now, there may be things you can do to bring up the bottom. 00:32:30.460 --> 00:32:32.002 align:middle line:84% The lessons of those anecdotes, maybe 00:32:32.002 --> 00:32:33.580 align:middle line:90% here's some regulations we need. 00:32:33.580 --> 00:32:36.300 align:middle line:84% But you can't take those anecdotes of some bad charter 00:32:36.300 --> 00:32:38.820 align:middle line:84% schools and say we shouldn't have charter schools. 00:32:38.820 --> 00:32:41.540 align:middle line:84% Basically what that says is, we need to change the regulations 00:32:41.540 --> 00:32:42.360 align:middle line:90% around them. 00:32:42.360 --> 00:32:44.300 align:middle line:90% But we deal in data. 00:32:44.300 --> 00:32:47.100 align:middle line:84% And so for example, my colleagues 00:32:47.100 --> 00:32:49.660 align:middle line:84% have shown that the charter school program in Louisiana 00:32:49.660 --> 00:32:50.500 align:middle line:90% was not successful. 00:32:50.500 --> 00:32:53.000 align:middle line:84% Here's a particular charter school program that didn't work. 00:32:53.000 --> 00:32:55.660 align:middle line:84% It didn't work because it was badly run, et cetera. 00:32:55.660 --> 00:33:00.180 align:middle line:84% So they're not always good, but the bulk of the evidence 00:33:00.180 --> 00:33:01.980 align:middle line:90% seems to be positive. 00:33:01.980 --> 00:33:06.300 align:middle line:84% So that's the second thing we've done. 00:33:06.300 --> 00:33:10.000 align:middle line:84% So we've done two things to try to fall short of vouchers, 00:33:10.000 --> 00:33:12.460 align:middle line:84% but go in that direction, which is, public school choice, 00:33:12.460 --> 00:33:14.040 align:middle line:90% doesn't seem to really matter. 00:33:14.040 --> 00:33:16.520 align:middle line:84% Allowing students these quasi-public alternatives does 00:33:16.520 --> 00:33:17.620 align:middle line:90% seem to be a good thing. 00:33:17.620 --> 00:33:20.280 align:middle line:90% 00:33:20.280 --> 00:33:23.120 align:middle line:84% Finally, the last thing we've done 00:33:23.120 --> 00:33:37.640 align:middle line:84% is massively increased the way we use testing and incentives 00:33:37.640 --> 00:33:40.760 align:middle line:90% to reward and penalize schools. 00:33:40.760 --> 00:33:42.880 align:middle line:84% So remember, I said that the way schools are 00:33:42.880 --> 00:33:49.177 align:middle line:84% financed is probably, like, 47% state, 47% town, 5% federal. 00:33:49.177 --> 00:33:50.760 align:middle line:84% Well, the federal government, starting 00:33:50.760 --> 00:33:53.218 align:middle line:84% with George Bush and the No Child Left Behind Act, and then 00:33:53.218 --> 00:33:55.320 align:middle line:84% subsequent administrations as well, 00:33:55.320 --> 00:33:57.480 align:middle line:84% has started to put more of a thumb on the scale. 00:33:57.480 --> 00:33:59.420 align:middle line:84% But the way they've done it is said, 00:33:59.420 --> 00:34:01.925 align:middle line:84% we're going to reward schools based on their performance. 00:34:01.925 --> 00:34:03.800 align:middle line:84% So it's been a massive increase in the amount 00:34:03.800 --> 00:34:05.020 align:middle line:90% of standardized testing. 00:34:05.020 --> 00:34:06.978 align:middle line:84% You guys went to public school, took a lot more 00:34:06.978 --> 00:34:08.693 align:middle line:90% standardized tests than I did. 00:34:08.693 --> 00:34:10.610 align:middle line:84% So the massive increased standardized testing. 00:34:10.610 --> 00:34:14.290 align:middle line:84% And that testing is being used to reward or penalize schools. 00:34:14.290 --> 00:34:16.290 align:middle line:84% So it's not quite the "put them out of business" 00:34:16.290 --> 00:34:17.306 align:middle line:90% model of vouchers. 00:34:17.306 --> 00:34:18.889 align:middle line:84% But it is a little bit take money away 00:34:18.889 --> 00:34:20.050 align:middle line:84% from the schools that are doing badly 00:34:20.050 --> 00:34:21.967 align:middle line:84% and give money to schools that are doing well. 00:34:21.967 --> 00:34:25.929 align:middle line:84% It's like a smaller version of this efficiency point. 00:34:25.929 --> 00:34:29.170 align:middle line:84% So basically that's caused a huge growth, 00:34:29.170 --> 00:34:33.370 align:middle line:84% and once again, evidence versus anecdote. 00:34:33.370 --> 00:34:34.889 align:middle line:84% The evidence is that overall, this 00:34:34.889 --> 00:34:38.570 align:middle line:84% has been quite positive for the US educational system. 00:34:38.570 --> 00:34:42.489 align:middle line:84% Basically, overall student outcomes 00:34:42.489 --> 00:34:46.210 align:middle line:84% have improved from this testing regime. 00:34:46.210 --> 00:34:51.290 align:middle line:84% However, it's not been uniformly positive in two senses. 00:34:51.290 --> 00:34:53.650 align:middle line:84% First of all, if you talk to anyone 00:34:53.650 --> 00:34:56.170 align:middle line:84% from a high-quality school district, 00:34:56.170 --> 00:34:57.702 align:middle line:90% they'll say this is unnecessary. 00:34:57.702 --> 00:34:58.910 align:middle line:90% Well, sure, it's unnecessary. 00:34:58.910 --> 00:35:00.978 align:middle line:84% They're a high-quality school district, 00:35:00.978 --> 00:35:02.770 align:middle line:84% That's why it's the disadvantaged kids that 00:35:02.770 --> 00:35:04.110 align:middle line:90% benefit from charter schools. 00:35:04.110 --> 00:35:06.130 align:middle line:84% This has really been a change that's mostly 00:35:06.130 --> 00:35:07.490 align:middle line:90% about bringing the bottom up. 00:35:07.490 --> 00:35:08.650 align:middle line:90% It's not really helping the top. 00:35:08.650 --> 00:35:09.690 align:middle line:84% They were already providing good schools. 00:35:09.690 --> 00:35:10.927 align:middle line:90% It's a pain in their ass. 00:35:10.927 --> 00:35:13.010 align:middle line:84% So ideal system, you'd figure who's a good school, 00:35:13.010 --> 00:35:18.230 align:middle line:84% not put them in the system, but you can't So that's 00:35:18.230 --> 00:35:19.170 align:middle line:90% one complaint. 00:35:19.170 --> 00:35:20.830 align:middle line:84% The other problem is actually, there's 00:35:20.830 --> 00:35:24.630 align:middle line:84% a whole literature on funky ways people have tried 00:35:24.630 --> 00:35:26.710 align:middle line:90% to get around these tests. 00:35:26.710 --> 00:35:32.330 align:middle line:84% For example, I don't know how many of you guys-- 00:35:32.330 --> 00:35:34.330 align:middle line:84% I hate to say this because I'll hate the answer. 00:35:34.330 --> 00:35:36.455 align:middle line:84% How many of you guys have watched The Simpsons more 00:35:36.455 --> 00:35:38.970 align:middle line:90% than 10 times? 00:35:38.970 --> 00:35:39.470 align:middle line:90% Fuck. 00:35:39.470 --> 00:35:40.430 align:middle line:90% [LAUGHTER] 00:35:40.430 --> 00:35:40.930 align:middle line:90% People. 00:35:40.930 --> 00:35:44.270 align:middle line:84% This is the most important show in the history of television. 00:35:44.270 --> 00:35:47.830 align:middle line:84% This is where you should get your lessons today, your humor, 00:35:47.830 --> 00:35:49.730 align:middle line:90% is from The Simpsons. 00:35:49.730 --> 00:35:50.230 align:middle line:90% [LAUGHTER] 00:35:50.230 --> 00:35:52.230 align:middle line:84% And The Simpsons did a great episode, 00:35:52.230 --> 00:35:56.550 align:middle line:84% which actually mirrors reality, where it's tested at the school. 00:35:56.550 --> 00:35:58.990 align:middle line:84% And they take a certain set of kids like Bart 00:35:58.990 --> 00:36:00.730 align:middle line:84% and say, hey, this is a field trip. 00:36:00.730 --> 00:36:02.330 align:middle line:84% And [INAUDIBLE] they just want to take them out of the school, 00:36:02.330 --> 00:36:04.663 align:middle line:84% so they don't take the test to make the school look bad. 00:36:04.663 --> 00:36:06.210 align:middle line:90% Turns out that actually happens. 00:36:06.210 --> 00:36:09.210 align:middle line:84% Schools would raise the rate at which they kick kids out 00:36:09.210 --> 00:36:11.432 align:middle line:84% for disciplinary problems on testing day. 00:36:11.432 --> 00:36:12.890 align:middle line:84% So the kids who were troublemakers, 00:36:12.890 --> 00:36:14.515 align:middle line:84% do badly in the test, wouldn't be there 00:36:14.515 --> 00:36:16.210 align:middle line:90% and wouldn't take the test. 00:36:16.210 --> 00:36:18.850 align:middle line:84% One of my favorite papers on this-- schools 00:36:18.850 --> 00:36:20.890 align:middle line:84% would actually manipulate school menus 00:36:20.890 --> 00:36:23.890 align:middle line:84% to sugar up the kids right before tests. 00:36:23.890 --> 00:36:25.898 align:middle line:84% Literally, they looked at school menus. 00:36:25.898 --> 00:36:27.690 align:middle line:84% And literally, the days that certain grades 00:36:27.690 --> 00:36:30.410 align:middle line:84% had tests, their school lunches had more carbs and sugars 00:36:30.410 --> 00:36:32.930 align:middle line:90% in them. 00:36:32.930 --> 00:36:34.190 align:middle line:90% Now, those are cool studies. 00:36:34.190 --> 00:36:34.870 align:middle line:90% I love them. 00:36:34.870 --> 00:36:37.370 align:middle line:84% They don't mean testing is a bad thing to do. 00:36:37.370 --> 00:36:41.810 align:middle line:84% They just mean that there's a trade-off to everything. 00:36:41.810 --> 00:36:43.310 align:middle line:84% So there's been these fun reactions. 00:36:43.310 --> 00:36:46.350 align:middle line:84% But the bottom line is, where does this leave us? 00:36:46.350 --> 00:36:48.930 align:middle line:84% When I look at this whole thing, where does it lead me? 00:36:48.930 --> 00:36:52.010 align:middle line:84% Well, I guess the way I like to think about it is the following. 00:36:52.010 --> 00:36:54.070 align:middle line:84% Imagine that over on this extreme, 00:36:54.070 --> 00:36:57.290 align:middle line:90% we have no school choice. 00:36:57.290 --> 00:37:00.930 align:middle line:84% And over in this extreme, we have pure Chicago dream 00:37:00.930 --> 00:37:03.710 align:middle line:90% vouchers. 00:37:03.710 --> 00:37:07.950 align:middle line:90% US education has been here. 00:37:07.950 --> 00:37:10.950 align:middle line:90% It's moving a bit to here. 00:37:10.950 --> 00:37:13.870 align:middle line:84% All the evidence suggests that's been positive, 00:37:13.870 --> 00:37:17.110 align:middle line:84% which suggests continued movement in this direction. 00:37:17.110 --> 00:37:18.757 align:middle line:90% Should we go all the way here? 00:37:18.757 --> 00:37:20.090 align:middle line:90% That's a pretty extreme outcome. 00:37:20.090 --> 00:37:21.890 align:middle line:84% And I think we need more evidence to see that. 00:37:21.890 --> 00:37:22.990 align:middle line:84% But certainly the existing evidence 00:37:22.990 --> 00:37:24.823 align:middle line:84% says we should be moving towards more school 00:37:24.823 --> 00:37:27.110 align:middle line:90% choice along the way. 00:37:27.110 --> 00:37:28.770 align:middle line:84% That would be my read of the evidence. 00:37:28.770 --> 00:37:29.270 align:middle line:90% Yeah. 00:37:29.270 --> 00:37:31.790 align:middle line:84% AUDIENCE: [INAUDIBLE] be properly read 00:37:31.790 --> 00:37:34.562 align:middle line:84% as dishonest or more like you're encouraging kids 00:37:34.562 --> 00:37:36.270 align:middle line:84% to get more sleep before a test or making 00:37:36.270 --> 00:37:37.312 align:middle line:90% sure they're eating well? 00:37:37.312 --> 00:37:37.950 align:middle line:90% JON GRUBER: No. 00:37:37.950 --> 00:37:40.510 align:middle line:84% I mean, if kids should eat well, they should eat well. 00:37:40.510 --> 00:37:42.503 align:middle line:84% They shouldn't just do it before a test. 00:37:42.503 --> 00:37:44.170 align:middle line:84% So, no, I think that's pretty dishonest. 00:37:44.170 --> 00:37:47.930 align:middle line:84% Not dishonest, but profit-maximizing, 00:37:47.930 --> 00:37:50.070 align:middle line:84% which some people called dishonest. 00:37:50.070 --> 00:37:53.630 align:middle line:84% All right, so that's all I want to say about that. 00:37:53.630 --> 00:37:55.830 align:middle line:84% Let me go to the second topic of the lecture, 00:37:55.830 --> 00:38:00.750 align:middle line:84% which is the field that we call "returns to education," 00:38:00.750 --> 00:38:02.930 align:middle line:90% which is a field of studying-- 00:38:02.930 --> 00:38:07.385 align:middle line:84% OK, we know we want government involved in education. 00:38:07.385 --> 00:38:09.010 align:middle line:84% But government involvement in education 00:38:09.010 --> 00:38:09.950 align:middle line:90% can mean lots of things. 00:38:09.950 --> 00:38:11.825 align:middle line:84% In particular, how much should the government 00:38:11.825 --> 00:38:13.850 align:middle line:90% spend on education? 00:38:13.850 --> 00:38:16.890 align:middle line:84% And that's going to depend on what is the rate of return 00:38:16.890 --> 00:38:18.410 align:middle line:90% to that investment. 00:38:18.410 --> 00:38:21.793 align:middle line:84% Remember, government is always doing cost-benefit analysis. 00:38:21.793 --> 00:38:23.210 align:middle line:84% If the rate of return to education 00:38:23.210 --> 00:38:25.043 align:middle line:84% exceeds the rate at which the government has 00:38:25.043 --> 00:38:27.790 align:middle line:84% to spend to borrow the money, it should make the investment. 00:38:27.790 --> 00:38:29.393 align:middle line:90% If it doesn't, it shouldn't. 00:38:29.393 --> 00:38:30.810 align:middle line:84% So the rate of return to education 00:38:30.810 --> 00:38:32.490 align:middle line:84% is a critical parameter in thinking 00:38:32.490 --> 00:38:36.170 align:middle line:84% about optimal government involvement. 00:38:36.170 --> 00:38:39.210 align:middle line:84% Now, the typical estimate-- this is a very rough. 00:38:39.210 --> 00:38:41.570 align:middle line:84% typical estimate is that another year of education 00:38:41.570 --> 00:38:44.650 align:middle line:90% raises your earnings by 7%. 00:38:44.650 --> 00:38:46.750 align:middle line:84% Every year of education, you get a raise of 7%. 00:38:46.750 --> 00:38:47.795 align:middle line:90% Now, that's not linear. 00:38:47.795 --> 00:38:49.170 align:middle line:84% And obviously, your MIT education 00:38:49.170 --> 00:38:50.910 align:middle line:84% is different than your other education. 00:38:50.910 --> 00:38:53.690 align:middle line:84% But that's a number to have in mind. 00:38:53.690 --> 00:38:56.970 align:middle line:84% That would suggest a fairly high rate of return to education. 00:38:56.970 --> 00:39:00.010 align:middle line:84% A 7% boost in lifetime earnings is pretty large 00:39:00.010 --> 00:39:04.310 align:middle line:84% compared to the cost of hiring another teacher. 00:39:04.310 --> 00:39:08.230 align:middle line:90% However, we have a problem. 00:39:08.230 --> 00:39:11.390 align:middle line:84% It's not clear how to interpret that result. 00:39:11.390 --> 00:39:14.030 align:middle line:90% There's two interpretations. 00:39:14.030 --> 00:39:19.150 align:middle line:84% One interpretation, which is called the "human capital view," 00:39:19.150 --> 00:39:22.390 align:middle line:84% is that people are like machines. 00:39:22.390 --> 00:39:24.290 align:middle line:84% And just like as we spend more in a machine, 00:39:24.290 --> 00:39:25.393 align:middle line:90% we get more capital. 00:39:25.393 --> 00:39:26.810 align:middle line:84% If we spend more educating people, 00:39:26.810 --> 00:39:28.335 align:middle line:90% we get more human capital. 00:39:28.335 --> 00:39:30.710 align:middle line:84% The more we spend in your education, the more the smarter 00:39:30.710 --> 00:39:32.350 align:middle line:84% you are, the more productive you are, 00:39:32.350 --> 00:39:34.600 align:middle line:84% the more you're producing valuable things for society. 00:39:34.600 --> 00:39:37.470 align:middle line:84% In other words, we are raising your marginal product of labor. 00:39:37.470 --> 00:39:40.690 align:middle line:84% So there's a 7% rise in not just your wage, 00:39:40.690 --> 00:39:42.210 align:middle line:84% but your marginal product of labor. 00:39:42.210 --> 00:39:43.870 align:middle line:84% That's why your wage is going up. 00:39:43.870 --> 00:39:45.230 align:middle line:84% Your marginal product of labor is actually 00:39:45.230 --> 00:39:46.570 align:middle line:84% your value, marginal product of labor. 00:39:46.570 --> 00:39:48.028 align:middle line:84% It's the value of what you produce. 00:39:48.028 --> 00:39:49.710 align:middle line:84% And that's why you get paid more. 00:39:49.710 --> 00:39:52.030 align:middle line:84% That would be the human capital view. 00:39:52.030 --> 00:39:55.190 align:middle line:84% But the evidence that people are more educated 00:39:55.190 --> 00:39:58.390 align:middle line:84% are paid more is absolutely consistent with another story, 00:39:58.390 --> 00:39:59.910 align:middle line:90% which is the screening story. 00:39:59.910 --> 00:40:02.450 align:middle line:90% 00:40:02.450 --> 00:40:06.770 align:middle line:84% Let's say that the way all educational systems worked 00:40:06.770 --> 00:40:09.850 align:middle line:84% is they consisted of memorizing baseball statistics. 00:40:09.850 --> 00:40:12.917 align:middle line:84% And let us posit that memorizing baseball statistics is totally 00:40:12.917 --> 00:40:15.250 align:middle line:84% irrelevant to producing what you're all going to produce 00:40:15.250 --> 00:40:17.167 align:middle line:84% in your jobs. [? It's ?] [? not ?] [? easy, ?] 00:40:17.167 --> 00:40:19.010 align:middle line:90% but let's imagine it is. 00:40:19.010 --> 00:40:21.757 align:middle line:84% And furthermore, let's say that how far you get in education 00:40:21.757 --> 00:40:24.090 align:middle line:84% is determined by how good you are at memorizing baseball 00:40:24.090 --> 00:40:26.250 align:middle line:90% statistics. 00:40:26.250 --> 00:40:29.250 align:middle line:84% You will find that the people who get the most education 00:40:29.250 --> 00:40:31.390 align:middle line:90% earn the most. 00:40:31.390 --> 00:40:31.890 align:middle line:90% Why? 00:40:31.890 --> 00:40:34.515 align:middle line:84% Because the smartest people are the best at memorizing baseball 00:40:34.515 --> 00:40:36.090 align:middle line:90% statistics. 00:40:36.090 --> 00:40:38.910 align:middle line:84% So they go on the most and they earn the most. 00:40:38.910 --> 00:40:41.158 align:middle line:84% But you didn't actually learn anything in school. 00:40:41.158 --> 00:40:43.450 align:middle line:84% School just separated people who are good at memorizing 00:40:43.450 --> 00:40:45.665 align:middle line:84% statistics from those who aren't. 00:40:45.665 --> 00:40:47.790 align:middle line:84% So you didn't raise your marginal product of labor. 00:40:47.790 --> 00:40:51.170 align:middle line:84% It didn't make you any better at making computers. 00:40:51.170 --> 00:40:54.490 align:middle line:84% It just was a way of screening out who was smart and who 00:40:54.490 --> 00:40:56.050 align:middle line:90% wasn't. 00:40:56.050 --> 00:40:58.300 align:middle line:84% The problem is that would also show 00:40:58.300 --> 00:41:01.240 align:middle line:84% that those with higher education would earn more. 00:41:01.240 --> 00:41:01.740 align:middle line:90% Why? 00:41:01.740 --> 00:41:03.940 align:middle line:84% Because what you're doing is actually, 00:41:03.940 --> 00:41:06.140 align:middle line:84% if we go back to chapter 3 terms, 00:41:06.140 --> 00:41:09.580 align:middle line:84% there's an omitted bias in our regressions. 00:41:09.580 --> 00:41:13.560 align:middle line:84% We regress education on years of earnings, on years of education. 00:41:13.560 --> 00:41:16.140 align:middle line:84% There's an omitted variable, called ability. 00:41:16.140 --> 00:41:18.860 align:middle line:84% And ability is correlated with both what you earn 00:41:18.860 --> 00:41:20.700 align:middle line:84% and how many years of education you get. 00:41:20.700 --> 00:41:21.920 align:middle line:90% So we have a non-comparable. 00:41:21.920 --> 00:41:24.420 align:middle line:84% The treatment group that got more education is not 00:41:24.420 --> 00:41:27.167 align:middle line:84% comparable to the control group that got less education, 00:41:27.167 --> 00:41:29.500 align:middle line:84% as well as you can't just compare them and say education 00:41:29.500 --> 00:41:30.760 align:middle line:90% improved your human capital. 00:41:30.760 --> 00:41:31.260 align:middle line:90% Yeah. 00:41:31.260 --> 00:41:33.020 align:middle line:84% AUDIENCE: Is screening [INAUDIBLE] 00:41:33.020 --> 00:41:35.268 align:middle line:84% if it's a more efficient way of helping people 00:41:35.268 --> 00:41:37.060 align:middle line:84% navigate their way through the labor market 00:41:37.060 --> 00:41:41.100 align:middle line:84% to jobs where they're actually, like, [INAUDIBLE]? 00:41:41.100 --> 00:41:43.460 align:middle line:84% JON GRUBER: Screening, could be valuable. 00:41:43.460 --> 00:41:45.700 align:middle line:84% The point is, it's not clear that you'd 00:41:45.700 --> 00:41:48.180 align:middle line:84% want to spend 35% of the entire state and local government 00:41:48.180 --> 00:41:49.847 align:middle line:84% budget in the US helping people navigate 00:41:49.847 --> 00:41:51.740 align:middle line:90% which job is best for them. 00:41:51.740 --> 00:41:56.260 align:middle line:84% We could probably have a lot more efficient system. 00:41:56.260 --> 00:41:58.243 align:middle line:84% In other words, the other name for screening 00:41:58.243 --> 00:41:59.660 align:middle line:84% is called Harvard Business School. 00:41:59.660 --> 00:42:01.660 align:middle line:90% [LAUGHTER] 00:42:01.660 --> 00:42:03.160 align:middle line:90% You don't learn anything. 00:42:03.160 --> 00:42:07.760 align:middle line:84% But basically, people who go to Harvard Business School 00:42:07.760 --> 00:42:08.880 align:middle line:90% are the cream of the crop. 00:42:08.880 --> 00:42:09.380 align:middle line:90% They get in. 00:42:09.380 --> 00:42:11.422 align:middle line:84% They have a great way of finding the best people. 00:42:11.422 --> 00:42:13.300 align:middle line:84% Now, it's not true it's not productive. 00:42:13.300 --> 00:42:15.260 align:middle line:84% It's productive mostly making connections. 00:42:15.260 --> 00:42:16.660 align:middle line:90% I'm being extreme, obviously. 00:42:16.660 --> 00:42:20.060 align:middle line:84% But the point is that basically, these are two different views. 00:42:20.060 --> 00:42:21.300 align:middle line:90% How do you distinguish them? 00:42:21.300 --> 00:42:23.800 align:middle line:84% Well, you distinguish them using the tools of chapter 3 00:42:23.800 --> 00:42:26.972 align:middle line:84% to develop experimental or quasi-experimental methods 00:42:26.972 --> 00:42:28.680 align:middle line:84% to look at what happens when you randomly 00:42:28.680 --> 00:42:30.720 align:middle line:90% give some more education. 00:42:30.720 --> 00:42:35.560 align:middle line:84% And there's a massive literature replete with incredibly clever 00:42:35.560 --> 00:42:36.678 align:middle line:90% approaches. 00:42:36.678 --> 00:42:38.720 align:middle line:84% So one approach, my Nobel Prize-winning colleague 00:42:38.720 --> 00:42:41.478 align:middle line:84% Esther Duflo, in her thesis, she looked 00:42:41.478 --> 00:42:43.520 align:middle line:84% at what happened when the government of Indonesia 00:42:43.520 --> 00:42:45.270 align:middle line:84% randomly built a bunch more schools in one 00:42:45.270 --> 00:42:46.800 align:middle line:90% place than another. 00:42:46.800 --> 00:42:48.360 align:middle line:90% The two places were identical. 00:42:48.360 --> 00:42:50.017 align:middle line:84% One got more schools, one didn't. 00:42:50.017 --> 00:42:51.600 align:middle line:84% She found that years later, the ones-- 00:42:51.600 --> 00:42:53.183 align:middle line:84% place that got more schools, that kids 00:42:53.183 --> 00:42:54.440 align:middle line:90% were earning a lot more. 00:42:54.440 --> 00:42:57.460 align:middle line:84% That's evidence it was a human capital improvement. 00:42:57.460 --> 00:43:00.340 align:middle line:84% Another great study is by Adriana Ileras-Muney 00:43:00.340 --> 00:43:02.060 align:middle line:90% out at UCLA. 00:43:02.060 --> 00:43:05.980 align:middle line:84% In the early 1900s in the US, kids didn't have to go 00:43:05.980 --> 00:43:06.940 align:middle line:90% to school. 00:43:06.940 --> 00:43:09.680 align:middle line:84% And then they put in minimum school-leaving ages. 00:43:09.680 --> 00:43:12.660 align:middle line:90% But these ages varied by state. 00:43:12.660 --> 00:43:17.060 align:middle line:84% So compare two 17-year-olds one who's in a state where you have 00:43:17.060 --> 00:43:20.060 align:middle line:84% to go to school till you're 17, and one a state where you can 00:43:20.060 --> 00:43:21.517 align:middle line:90% leave when you're 16? 00:43:21.517 --> 00:43:23.100 align:middle line:84% Well, the only difference between them 00:43:23.100 --> 00:43:25.380 align:middle line:84% is one was forced to stay in school one more year. 00:43:25.380 --> 00:43:26.960 align:middle line:90% Do they earn more later in life? 00:43:26.960 --> 00:43:28.540 align:middle line:90% The answer is yes. 00:43:28.540 --> 00:43:31.620 align:middle line:84% So there's a lot of evidence for the human capital story. 00:43:31.620 --> 00:43:33.920 align:middle line:84% There is some evidence for the screening story. 00:43:33.920 --> 00:43:40.780 align:middle line:90% 00:43:40.780 --> 00:43:43.640 align:middle line:84% Basically, there was this great study in Sweden, 00:43:43.640 --> 00:43:46.500 align:middle line:84% where they found that teachers discretionarily 00:43:46.500 --> 00:43:48.798 align:middle line:84% were moving students from not passing to passing. 00:43:48.798 --> 00:43:51.340 align:middle line:84% And they could look at teachers who were using the discretion 00:43:51.340 --> 00:43:53.640 align:middle line:84% to move students and those who weren't. 00:43:53.640 --> 00:43:55.460 align:middle line:90% Now, it was the same score. 00:43:55.460 --> 00:43:56.840 align:middle line:90% Let me back up. 00:43:56.840 --> 00:43:59.060 align:middle line:84% You had two students, same score on the test. 00:43:59.060 --> 00:44:00.860 align:middle line:84% One's with the teacher that's a soft heart. 00:44:00.860 --> 00:44:02.700 align:middle line:84% And she'd pass a lot more students. 00:44:02.700 --> 00:44:05.920 align:middle line:84% The other teacher was a hard ass and wouldn't pass the students. 00:44:05.920 --> 00:44:08.000 align:middle line:84% Those students had the same human capital, 00:44:08.000 --> 00:44:10.107 align:middle line:84% but the ones who passed earned a lot more. 00:44:10.107 --> 00:44:12.440 align:middle line:84% So that suggests, well, there is some screening function 00:44:12.440 --> 00:44:13.560 align:middle line:90% here too. 00:44:13.560 --> 00:44:15.163 align:middle line:90% So both are operating. 00:44:15.163 --> 00:44:16.580 align:middle line:84% But primarily, it's human capital. 00:44:16.580 --> 00:44:21.100 align:middle line:84% There is a real return to more education 00:44:21.100 --> 00:44:23.520 align:middle line:84% There is a real return to more education. 00:44:23.520 --> 00:44:26.080 align:middle line:90% Now, let me stop there. 00:44:26.080 --> 00:44:29.720 align:middle line:84% Does the fact that there is a high rate of return to education 00:44:29.720 --> 00:44:33.360 align:middle line:84% mean the government should invest in it? 00:44:33.360 --> 00:44:34.672 align:middle line:90% Why not necessarily? 00:44:34.672 --> 00:44:36.880 align:middle line:84% Let's say that it's true that every year of education 00:44:36.880 --> 00:44:38.957 align:middle line:90% makes earnings 7% higher. 00:44:38.957 --> 00:44:41.040 align:middle line:84% Why should that necessarily justify the government 00:44:41.040 --> 00:44:43.960 align:middle line:90% getting involved? 00:44:43.960 --> 00:44:44.460 align:middle line:90% Yeah. 00:44:44.460 --> 00:44:45.572 align:middle line:90% AUDIENCE: It could be-- 00:44:45.572 --> 00:44:48.280 align:middle line:84% it could be the externalities in that your marginal product could 00:44:48.280 --> 00:44:50.340 align:middle line:84% go up by more than 7% in [? taxes. ?] 00:44:50.340 --> 00:44:52.905 align:middle line:84% JON GRUBER: Well, no, you made it too hard. 00:44:52.905 --> 00:44:54.280 align:middle line:84% Let's say there's no externality. 00:44:54.280 --> 00:44:56.165 align:middle line:84% Then the government shouldn't get involved. 00:44:56.165 --> 00:44:57.540 align:middle line:84% I mean, basically, the government 00:44:57.540 --> 00:45:00.285 align:middle line:84% should say, look, I should just-- why should I 00:45:00.285 --> 00:45:01.160 align:middle line:90% bribe public schools? 00:45:01.160 --> 00:45:03.660 align:middle line:84% I should just give loans to everybody. 00:45:03.660 --> 00:45:06.683 align:middle line:84% If basically, all I'm doing is increasing your wages, 00:45:06.683 --> 00:45:08.100 align:middle line:84% what's the case for the government 00:45:08.100 --> 00:45:09.660 align:middle line:90% to invest money in this? 00:45:09.660 --> 00:45:11.340 align:middle line:84% Now, there's a case for all the other reasons-- making 00:45:11.340 --> 00:45:14.007 align:middle line:84% sure you're a good citizen, that you don't commit crimes, things 00:45:14.007 --> 00:45:14.560 align:middle line:90% like that. 00:45:14.560 --> 00:45:16.380 align:middle line:84% But basically, if literally all education 00:45:16.380 --> 00:45:18.660 align:middle line:84% is doing-- so once again, let's imagine education 00:45:18.660 --> 00:45:20.758 align:middle line:84% through ninth grade is making you a good citizen, 00:45:20.758 --> 00:45:21.800 align:middle line:90% makes you don't do crime. 00:45:21.800 --> 00:45:23.420 align:middle line:84% Let's imagine all this stuff in high school beyond it's 00:45:23.420 --> 00:45:25.100 align:middle line:84% just about increasing your wages. 00:45:25.100 --> 00:45:27.140 align:middle line:84% Then the government shouldn't be doing it. 00:45:27.140 --> 00:45:28.820 align:middle line:84% The government should make it easier 00:45:28.820 --> 00:45:31.320 align:middle line:84% for you to do it, address the credit market failures, 00:45:31.320 --> 00:45:33.740 align:middle line:90% address the parental failures. 00:45:33.740 --> 00:45:35.260 align:middle line:84% But the government, unless there's 00:45:35.260 --> 00:45:38.500 align:middle line:84% an externality, an effect on others, 00:45:38.500 --> 00:45:40.343 align:middle line:84% the government shouldn't be doing it. 00:45:40.343 --> 00:45:42.760 align:middle line:84% So the question is, are there externalities for education? 00:45:42.760 --> 00:45:45.468 align:middle line:84% And the answer is there's been a huge literature which has shown, 00:45:45.468 --> 00:45:47.820 align:middle line:84% yes, education leads to less crime. 00:45:47.820 --> 00:45:51.400 align:middle line:84% Education leads to better health for you and your children. 00:45:51.400 --> 00:45:54.000 align:middle line:84% Education leads to more informed voting and higher voter 00:45:54.000 --> 00:45:55.440 align:middle line:90% participation. 00:45:55.440 --> 00:45:58.480 align:middle line:84% There's a lot of studies which show that education matters. 00:45:58.480 --> 00:46:00.200 align:middle line:90% It raises wages. 00:46:00.200 --> 00:46:03.720 align:middle line:84% But also, education has productivity spillovers. 00:46:03.720 --> 00:46:07.120 align:middle line:84% If you're around more educated people, you yourself earn more. 00:46:07.120 --> 00:46:09.960 align:middle line:84% So there's clearly massive positive externalities 00:46:09.960 --> 00:46:11.160 align:middle line:90% for education. 00:46:11.160 --> 00:46:13.560 align:middle line:84% So the return is positive, not just 00:46:13.560 --> 00:46:17.760 align:middle line:84% privately for the individuals, but publicly for society too. 00:46:17.760 --> 00:46:23.480 align:middle line:84% Now, that says we should invest in education. 00:46:23.480 --> 00:46:26.120 align:middle line:84% But the question then is, do we invest in that 00:46:26.120 --> 00:46:28.560 align:middle line:84% by just letting people go to school more? 00:46:28.560 --> 00:46:32.280 align:middle line:84% Or should we actually care about the quality of those schools? 00:46:32.280 --> 00:46:35.480 align:middle line:84% How should we measure whether better schools actually 00:46:35.480 --> 00:46:37.760 align:middle line:84% lead to better student outcomes, as opposed 00:46:37.760 --> 00:46:38.802 align:middle line:90% to just getting educated? 00:46:38.802 --> 00:46:41.218 align:middle line:84% We've just said getting educated leads to better outcomes. 00:46:41.218 --> 00:46:43.100 align:middle line:84% People go to school commit less crimes. 00:46:43.100 --> 00:46:44.420 align:middle line:84% People who go to school are healthier. 00:46:44.420 --> 00:46:45.837 align:middle line:84% People who go to school earn more. 00:46:45.837 --> 00:46:48.600 align:middle line:84% But there's going to school and there's going to school. 00:46:48.600 --> 00:46:51.960 align:middle line:84% There's a rich private school and a crappy public school. 00:46:51.960 --> 00:46:53.280 align:middle line:90% We just said going to school. 00:46:53.280 --> 00:46:54.840 align:middle line:84% What about the quality dimension? 00:46:54.840 --> 00:46:56.780 align:middle line:90% How much does that matter? 00:46:56.780 --> 00:47:01.220 align:middle line:84% And here, once again, there's been an enormous body 00:47:01.220 --> 00:47:04.900 align:middle line:84% of work looking at this, looking at things like teacher 00:47:04.900 --> 00:47:06.883 align:middle line:90% salaries and class size. 00:47:06.883 --> 00:47:08.300 align:middle line:84% One of most interesting literature 00:47:08.300 --> 00:47:11.745 align:middle line:84% has been on class size, where basically, this 00:47:11.745 --> 00:47:14.120 align:middle line:84% is in some sense, a little bit of a weakness of my field. 00:47:14.120 --> 00:47:15.820 align:middle line:90% I think, if I can be honest. 00:47:15.820 --> 00:47:17.940 align:middle line:84% For years there was this view, well, 00:47:17.940 --> 00:47:20.440 align:middle line:84% we don't have good evidence that smaller class size matters, 00:47:20.440 --> 00:47:21.800 align:middle line:90% so we shouldn't care about it. 00:47:21.800 --> 00:47:22.960 align:middle line:84% But if you ask those same people, 00:47:22.960 --> 00:47:25.127 align:middle line:84% do you want your kids in a big class or small class, 00:47:25.127 --> 00:47:26.480 align:middle line:90% they'd say a small class. 00:47:26.480 --> 00:47:27.600 align:middle line:90% We knew it mattered. 00:47:27.600 --> 00:47:30.140 align:middle line:84% We just hadn't come up with a convincing way to get at it. 00:47:30.140 --> 00:47:32.640 align:middle line:84% It's not like we didn't think it mattered. 00:47:32.640 --> 00:47:33.560 align:middle line:90% We knew it mattered. 00:47:33.560 --> 00:47:35.935 align:middle line:84% We tried to come up with a convincing way to but we have. 00:47:35.935 --> 00:47:38.400 align:middle line:84% And there's been a number of fascinating studies. 00:47:38.400 --> 00:47:41.580 align:middle line:84% One really cool study, once again by Josh Angrist-- 00:47:41.580 --> 00:47:44.500 align:middle line:84% in Israel, they follow the biblical rule 00:47:44.500 --> 00:47:46.220 align:middle line:90% called Maimonides law. 00:47:46.220 --> 00:47:48.760 align:middle line:84% And Maimonides law says there is a certain number 00:47:48.760 --> 00:47:52.320 align:middle line:84% of kids, which is the biggest a classroom should be. 00:47:52.320 --> 00:47:55.340 align:middle line:84% And when your cohort size exceeds that number by one, 00:47:55.340 --> 00:47:56.715 align:middle line:90% they split you into two classes. 00:47:56.715 --> 00:47:58.840 align:middle line:84% It's a random cutoff that determines whether you're 00:47:58.840 --> 00:48:00.090 align:middle line:90% in a big class or small class. 00:48:00.090 --> 00:48:02.900 align:middle line:84% If your cutoff is one below that number, you're in a big class. 00:48:02.900 --> 00:48:04.600 align:middle line:84% If you're one above that number, you're in a small class 00:48:04.600 --> 00:48:06.047 align:middle line:90% because they split into two. 00:48:06.047 --> 00:48:07.880 align:middle line:84% And he finds that being in the smaller class 00:48:07.880 --> 00:48:09.960 align:middle line:90% improves your outcomes. 00:48:09.960 --> 00:48:11.495 align:middle line:90% There's also real trials. 00:48:11.495 --> 00:48:14.120 align:middle line:84% There was something called the Tennessee STAR experiment, where 00:48:14.120 --> 00:48:17.800 align:middle line:84% they literally randomized kids across class sizes in Tennessee. 00:48:17.800 --> 00:48:23.605 align:middle line:84% And they found the kids who were in smaller classes did better, 00:48:23.605 --> 00:48:25.980 align:middle line:84% not only in the short run, but actually, in the long run. 00:48:25.980 --> 00:48:27.440 align:middle line:84% They followed those kids up and looked at their earnings 00:48:27.440 --> 00:48:29.700 align:middle line:84% later in life, and they earned more. 00:48:29.700 --> 00:48:31.800 align:middle line:90% So it's really powerful. 00:48:31.800 --> 00:48:34.640 align:middle line:84% Here's a fascinating fact, though. 00:48:34.640 --> 00:48:40.400 align:middle line:84% California, in 1996, passed a law mandating a maximum class 00:48:40.400 --> 00:48:43.178 align:middle line:84% size, which would be justified based on this evidence. 00:48:43.178 --> 00:48:44.720 align:middle line:84% So they basically said the class size 00:48:44.720 --> 00:48:48.290 align:middle line:84% had to be reduced to no more than 20 00:48:48.290 --> 00:48:50.210 align:middle line:90% a year for grades K to 3. 00:48:50.210 --> 00:48:53.130 align:middle line:84% It cost the state $1 billion to do this. 00:48:53.130 --> 00:48:55.490 align:middle line:84% And studies of it found that it did not 00:48:55.490 --> 00:48:58.050 align:middle line:90% improve student outcomes. 00:48:58.050 --> 00:49:00.050 align:middle line:84% Now, this is very strange, but it's 00:49:00.050 --> 00:49:02.770 align:middle line:84% a perfect example of partial equilibrium 00:49:02.770 --> 00:49:05.530 align:middle line:84% versus general equilibrium economics. 00:49:05.530 --> 00:49:09.068 align:middle line:84% Holding all else constant, is smaller classes good for you. 00:49:09.068 --> 00:49:11.110 align:middle line:84% But now let's say you're the state of California. 00:49:11.110 --> 00:49:12.350 align:middle line:90% You're a city in California. 00:49:12.350 --> 00:49:15.730 align:middle line:84% You've just been told you have to overnight, or very rapidly, 00:49:15.730 --> 00:49:17.450 align:middle line:90% reduce your class sizes. 00:49:17.450 --> 00:49:18.690 align:middle line:90% What do you do? 00:49:18.690 --> 00:49:20.370 align:middle line:90% You hire crappy teachers. 00:49:20.370 --> 00:49:22.690 align:middle line:84% You put the first and second grade together 00:49:22.690 --> 00:49:24.270 align:middle line:84% so you can split them more equally, 00:49:24.270 --> 00:49:26.478 align:middle line:84% which has actually been shown to be bad educationally 00:49:26.478 --> 00:49:27.490 align:middle line:90% to mixed grades. 00:49:27.490 --> 00:49:29.650 align:middle line:84% You do a bunch of stuff to meet this mandate, which 00:49:29.650 --> 00:49:32.067 align:middle line:84% is reducing class size through other actions which are not 00:49:32.067 --> 00:49:33.950 align:middle line:90% productive for education. 00:49:33.950 --> 00:49:35.450 align:middle line:84% So this is a fascinating example of, 00:49:35.450 --> 00:49:37.367 align:middle line:84% when we think about government policy changes, 00:49:37.367 --> 00:49:39.450 align:middle line:84% we have to think about all the reactions to them. 00:49:39.450 --> 00:49:41.242 align:middle line:84% And while the evidence says smaller classes 00:49:41.242 --> 00:49:43.090 align:middle line:84% are better, simply mandating California 00:49:43.090 --> 00:49:45.423 align:middle line:84% to have smaller classes without putting the resources in 00:49:45.423 --> 00:49:48.430 align:middle line:84% to make it feasible led to actually not 00:49:48.430 --> 00:49:49.830 align:middle line:90% an improvement in outcomes. 00:49:49.830 --> 00:49:50.990 align:middle line:90% Yeah. 00:49:50.990 --> 00:49:53.550 align:middle line:84% AUDIENCE: I have a question about the [INAUDIBLE]. 00:49:53.550 --> 00:49:55.510 align:middle line:84% Is the evidence about the class sizes 00:49:55.510 --> 00:49:58.710 align:middle line:84% or about the teacher-student ratios? 00:49:58.710 --> 00:50:00.210 align:middle line:84% JON GRUBER: That's a great question. 00:50:00.210 --> 00:50:02.910 align:middle line:84% I think it's typically viewed as teacher-student ratio. 00:50:02.910 --> 00:50:05.790 align:middle line:90% Yeah. 00:50:05.790 --> 00:50:06.550 align:middle line:90% All right? 00:50:06.550 --> 00:50:09.010 align:middle line:84% So once again, where are we in education? 00:50:09.010 --> 00:50:11.747 align:middle line:84% First of all government should be involved. 00:50:11.747 --> 00:50:13.330 align:middle line:84% How should the government be involved? 00:50:13.330 --> 00:50:14.830 align:middle line:84% Well, basically, there's a trade-off 00:50:14.830 --> 00:50:16.410 align:middle line:90% with how much choice to leave. 00:50:16.410 --> 00:50:18.410 align:middle line:84% My view is there probably should be more choice, 00:50:18.410 --> 00:50:20.630 align:middle line:90% but we understand the trade-off. 00:50:20.630 --> 00:50:25.330 align:middle line:84% Then the question is, well, is investing education worth it? 00:50:25.330 --> 00:50:26.710 align:middle line:90% The answer is yes. 00:50:26.710 --> 00:50:29.210 align:middle line:84% Is investing in higher quality education worth it? 00:50:29.210 --> 00:50:32.350 align:middle line:90% The answer is yes. 00:50:32.350 --> 00:50:35.110 align:middle line:84% And that's where we are, except for one topic, which 00:50:35.110 --> 00:50:36.730 align:middle line:84% is a little bit germane to all of you, 00:50:36.730 --> 00:50:38.130 align:middle line:90% which is higher education. 00:50:38.130 --> 00:50:41.350 align:middle line:84% And that's what I spend the last 25 minutes on, 00:50:41.350 --> 00:50:43.497 align:middle line:84% which is the system of higher education, 00:50:43.497 --> 00:50:44.830 align:middle line:90% particularly focusing on the US. 00:50:44.830 --> 00:50:47.210 align:middle line:84% The reason I'm focusing on the US in particular 00:50:47.210 --> 00:50:49.590 align:middle line:84% is that we are by far the best in the world. 00:50:49.590 --> 00:50:59.170 align:middle line:84% The US, every year, 1.2 million foreign students 00:50:59.170 --> 00:51:01.890 align:middle line:84% a year spend more than $39 billion 00:51:01.890 --> 00:51:03.690 align:middle line:90% to enroll in American colleges. 00:51:03.690 --> 00:51:06.750 align:middle line:84% At the same time, 60,000 Americans study abroad, 00:51:06.750 --> 00:51:09.450 align:middle line:84% typically just to go party in Europe. 00:51:09.450 --> 00:51:12.210 align:middle line:84% So basically, the market has spoken. 00:51:12.210 --> 00:51:16.810 align:middle line:84% And we have by far the world's best higher education system. 00:51:16.810 --> 00:51:18.990 align:middle line:84% And if you are a free market advocate, 00:51:18.990 --> 00:51:22.210 align:middle line:84% you say, well, here's an interesting fact. 00:51:22.210 --> 00:51:27.490 align:middle line:84% If you look at non-higher education, we are terrible. 00:51:27.490 --> 00:51:29.150 align:middle line:84% We're way down in world rankings. 00:51:29.150 --> 00:51:31.570 align:middle line:90% 90% of it is public. 00:51:31.570 --> 00:51:35.970 align:middle line:84% If you look at higher education-- 00:51:35.970 --> 00:51:41.190 align:middle line:84% if you look at higher education, 40% of it is private. 00:51:41.190 --> 00:51:43.430 align:middle line:90% Only 60% is public. 00:51:43.430 --> 00:51:45.450 align:middle line:84% So here's an interesting pair of facts. 00:51:45.450 --> 00:51:47.350 align:middle line:84% The system where we're the best in the world 00:51:47.350 --> 00:51:49.030 align:middle line:90% is where we have 40% private. 00:51:49.030 --> 00:51:51.690 align:middle line:84% The system where we're one of the worst in the world's 00:51:51.690 --> 00:51:55.830 align:middle line:84% where we're less than 10% private. 00:51:55.830 --> 00:51:57.870 align:middle line:84% So that suggests that gee, maybe competition 00:51:57.870 --> 00:52:00.010 align:middle line:84% and private incentives, there's something to it. 00:52:00.010 --> 00:52:01.250 align:middle line:84% Now, I wanted more to say about that. 00:52:01.250 --> 00:52:02.792 align:middle line:84% It's just an interesting observation. 00:52:02.792 --> 00:52:04.715 align:middle line:84% It's not obviously the reason we're best 00:52:04.715 --> 00:52:06.090 align:middle line:84% in the world at higher education. 00:52:06.090 --> 00:52:07.830 align:middle line:84% But it is an interesting observation 00:52:07.830 --> 00:52:14.568 align:middle line:84% that where we have more private sector involvement, 00:52:14.568 --> 00:52:16.110 align:middle line:84% we're doing better, at least relative 00:52:16.110 --> 00:52:18.430 align:middle line:90% to the rest of the world. 00:52:18.430 --> 00:52:21.670 align:middle line:84% Now, what is the government's role in higher education? 00:52:21.670 --> 00:52:23.430 align:middle line:84% Now, once again, here, we have to go back 00:52:23.430 --> 00:52:24.830 align:middle line:90% to the first question. 00:52:24.830 --> 00:52:27.250 align:middle line:84% It's not as obvious the government should be involved. 00:52:27.250 --> 00:52:30.018 align:middle line:84% I mean, all our citizenship stuff 00:52:30.018 --> 00:52:31.310 align:middle line:90% should have been taken care of. 00:52:31.310 --> 00:52:33.602 align:middle line:84% Our crime stuff should have been taken care, et cetera. 00:52:33.602 --> 00:52:35.727 align:middle line:84% So it's not entirely clear the government should be 00:52:35.727 --> 00:52:36.970 align:middle line:90% involved in higher education. 00:52:36.970 --> 00:52:42.090 align:middle line:84% Indeed, what is the main market failure with higher education? 00:52:42.090 --> 00:52:45.830 align:middle line:84% The main market failure is credit market failures. 00:52:45.830 --> 00:52:48.047 align:middle line:84% It is the fact that in principle, in a perfect world, 00:52:48.047 --> 00:52:50.630 align:middle line:84% we'd say, look, we're going to educate you through 12th grade. 00:52:50.630 --> 00:52:51.590 align:middle line:90% You want to go on, that's great. 00:52:51.590 --> 00:52:52.870 align:middle line:90% Just take a loan and go on. 00:52:52.870 --> 00:52:56.410 align:middle line:84% Most of the returns will be private anyway, 00:52:56.410 --> 00:52:58.530 align:middle line:90% so why not let you just go on? 00:52:58.530 --> 00:52:59.750 align:middle line:90% And the answer would be two. 00:52:59.750 --> 00:53:02.610 align:middle line:84% One is there are spillovers even from education. 00:53:02.610 --> 00:53:05.090 align:middle line:84% So there is evidence that even educating people in college 00:53:05.090 --> 00:53:06.870 align:middle line:90% has positive externalities. 00:53:06.870 --> 00:53:08.730 align:middle line:84% So that's one argument for why we'd still 00:53:08.730 --> 00:53:10.390 align:middle line:84% want the government involved in higher education. 00:53:10.390 --> 00:53:11.973 align:middle line:84% There are still positive externalities 00:53:11.973 --> 00:53:13.290 align:middle line:90% for higher education. 00:53:13.290 --> 00:53:15.447 align:middle line:90% Two is credit market failures. 00:53:15.447 --> 00:53:18.030 align:middle line:84% Like I said last time, you can't offer yourself as collateral. 00:53:18.030 --> 00:53:19.925 align:middle line:90% We call that slavery. 00:53:19.925 --> 00:53:21.550 align:middle line:84% You can't offer yourself as collateral. 00:53:21.550 --> 00:53:23.490 align:middle line:84% As a result, you can't necessarily 00:53:23.490 --> 00:53:26.930 align:middle line:84% get loans to better education, even if that would be 00:53:26.930 --> 00:53:29.250 align:middle line:90% a good productive investment. 00:53:29.250 --> 00:53:31.410 align:middle line:84% So credit market failures is another reason 00:53:31.410 --> 00:53:34.490 align:middle line:84% why we think the government should be involved. 00:53:34.490 --> 00:53:36.990 align:middle line:84% So how is the government involved? 00:53:36.990 --> 00:53:42.590 align:middle line:84% Well, the government involvement is illustrated in figure 11-4. 00:53:42.590 --> 00:53:47.530 align:middle line:84% The major way the government's involved is public universities. 00:53:47.530 --> 00:53:49.470 align:middle line:84% Like I said, 60% of university students 00:53:49.470 --> 00:53:51.750 align:middle line:90% do attend public universities. 00:53:51.750 --> 00:53:55.430 align:middle line:84% And by far, the largest expenditure 00:53:55.430 --> 00:54:00.270 align:middle line:84% of the public sector on higher education 00:54:00.270 --> 00:54:02.950 align:middle line:90% is these public universities. 00:54:02.950 --> 00:54:08.630 align:middle line:84% The second biggest is what we call Pell grants. 00:54:08.630 --> 00:54:12.750 align:middle line:84% Pell grants is a program that provides grants to attend school 00:54:12.750 --> 00:54:14.478 align:middle line:90% for lower-income families. 00:54:14.478 --> 00:54:16.270 align:middle line:84% So it's a grant of several thousand dollars 00:54:16.270 --> 00:54:18.670 align:middle line:84% you can use to offset the cost of college 00:54:18.670 --> 00:54:20.430 align:middle line:90% for low-income families. 00:54:20.430 --> 00:54:23.270 align:middle line:84% We spend $28 billion a year on that. 00:54:23.270 --> 00:54:27.670 align:middle line:84% The third and the smallest is student loans. 00:54:27.670 --> 00:54:29.550 align:middle line:90% Student loans are interesting. 00:54:29.550 --> 00:54:33.790 align:middle line:84% We used to have a system where we had two sources of loans. 00:54:33.790 --> 00:54:36.037 align:middle line:84% Direct student loans, which we're 00:54:36.037 --> 00:54:38.370 align:middle line:84% borrowing from the government-- which is what all of you 00:54:38.370 --> 00:54:39.230 align:middle line:90% do now if you have loans. 00:54:39.230 --> 00:54:40.610 align:middle line:90% It's the only program available. 00:54:40.610 --> 00:54:42.985 align:middle line:84% We also have something called the Guaranteed Student Loan 00:54:42.985 --> 00:54:46.170 align:middle line:84% program, the GSL program, where basically the government 00:54:46.170 --> 00:54:48.338 align:middle line:84% subsidized private banks to loan you money. 00:54:48.338 --> 00:54:50.130 align:middle line:84% Fascinating story, I'll divert for a second 00:54:50.130 --> 00:54:51.990 align:middle line:84% to talk about this, because once again, 00:54:51.990 --> 00:54:54.130 align:middle line:84% it comes to a privatization issue. 00:54:54.130 --> 00:54:56.690 align:middle line:84% The way this program works is the government 00:54:56.690 --> 00:55:01.610 align:middle line:84% said to banks, you loan money to students at a certain interest 00:55:01.610 --> 00:55:02.250 align:middle line:90% rate. 00:55:02.250 --> 00:55:06.250 align:middle line:84% We will guarantee if the students default. It 00:55:06.250 --> 00:55:09.450 align:middle line:84% said to banks, essentially, we are giving you a risk-free part 00:55:09.450 --> 00:55:11.410 align:middle line:90% of your portfolio. 00:55:11.410 --> 00:55:17.010 align:middle line:84% Now, the banks were like, great, we'll make tons of money 00:55:17.010 --> 00:55:17.590 align:middle line:90% on that. 00:55:17.590 --> 00:55:19.465 align:middle line:84% Basically, why do banks charge high interest? 00:55:19.465 --> 00:55:20.450 align:middle line:90% Because of the risk. 00:55:20.450 --> 00:55:21.908 align:middle line:84% Well, the government was saying you 00:55:21.908 --> 00:55:25.530 align:middle line:84% can charge your regular interest rate with no risk. 00:55:25.530 --> 00:55:26.710 align:middle line:90% The banks loved it. 00:55:26.710 --> 00:55:28.550 align:middle line:84% Now, people realized that was a problem. 00:55:28.550 --> 00:55:31.050 align:middle line:84% And they passed a law which said by a date certain, 00:55:31.050 --> 00:55:33.858 align:middle line:84% the government had to reset the interest rate for student loans. 00:55:33.858 --> 00:55:36.400 align:middle line:84% That date certain was while I was in the Treasury Department, 00:55:36.400 --> 00:55:38.180 align:middle line:84% so I was the guy in charge of this. 00:55:38.180 --> 00:55:40.638 align:middle line:84% I was in charge of resetting the interest on student loans. 00:55:40.638 --> 00:55:42.740 align:middle line:90% This is 1998. 00:55:42.740 --> 00:55:46.540 align:middle line:84% And we looked at the data and decided 00:55:46.540 --> 00:55:49.260 align:middle line:84% that banks were being massively overpaid, 00:55:49.260 --> 00:55:51.020 align:middle line:90% given the lack of risk. 00:55:51.020 --> 00:55:54.055 align:middle line:84% And they should receive about a percentage and a half less. 00:55:54.055 --> 00:55:55.180 align:middle line:90% I forget the exact numbers. 00:55:55.180 --> 00:55:57.740 align:middle line:84% It's about 1 and 1/2% less in interest for these student 00:55:57.740 --> 00:55:58.240 align:middle line:90% loans. 00:55:58.240 --> 00:55:59.900 align:middle line:90% We should cut it. 00:55:59.900 --> 00:56:00.680 align:middle line:90% We announced this. 00:56:00.680 --> 00:56:01.960 align:middle line:84% I was there the day they announced it. 00:56:01.960 --> 00:56:03.790 align:middle line:84% Al Gore and I even had the same tie on. 00:56:03.790 --> 00:56:05.040 align:middle line:90% I was very excited about that. 00:56:05.040 --> 00:56:06.080 align:middle line:90% There they announced it. 00:56:06.080 --> 00:56:08.262 align:middle line:84% We're going to cut student loan interest rates. 00:56:08.262 --> 00:56:09.720 align:middle line:84% All the students were very excited. 00:56:09.720 --> 00:56:11.252 align:middle line:90% The banks were super-pissed. 00:56:11.252 --> 00:56:11.960 align:middle line:90% Oh it's terrible. 00:56:11.960 --> 00:56:12.752 align:middle line:90% We can't afford it. 00:56:12.752 --> 00:56:13.820 align:middle line:90% We'll go out of business. 00:56:13.820 --> 00:56:16.860 align:middle line:84% Well, it turns out the banks have very good lobbyists. 00:56:16.860 --> 00:56:19.820 align:middle line:84% And the lobbyists started really putting pressure on Congress 00:56:19.820 --> 00:56:21.460 align:middle line:90% to not approve this. 00:56:21.460 --> 00:56:25.590 align:middle line:84% Well, meanwhile, the Clinton administration, the student 00:56:25.590 --> 00:56:26.840 align:middle line:90% groups were like, we love you. 00:56:26.840 --> 00:56:27.870 align:middle line:90% You can't back off this. 00:56:27.870 --> 00:56:29.620 align:middle line:84% So Congress and the Clinton administration 00:56:29.620 --> 00:56:31.640 align:middle line:90% were caught in a tough position. 00:56:31.640 --> 00:56:32.640 align:middle line:90% They put this out there. 00:56:32.640 --> 00:56:34.980 align:middle line:84% Student groups were excited about paying lower interest. 00:56:34.980 --> 00:56:36.680 align:middle line:90% But the banks were mad. 00:56:36.680 --> 00:56:39.760 align:middle line:84% So here's the interesting, depressing chapter 9 lesson. 00:56:39.760 --> 00:56:42.840 align:middle line:84% Between an immovable force-- between an unstoppable force 00:56:42.840 --> 00:56:44.480 align:middle line:84% and immovable object, there's actually 00:56:44.480 --> 00:56:45.660 align:middle line:90% something in between them. 00:56:45.660 --> 00:56:46.860 align:middle line:90% It's called the deficit. 00:56:46.860 --> 00:56:48.000 align:middle line:90% [LAUGHTER] 00:56:48.000 --> 00:56:49.380 align:middle line:90% So what did the government do? 00:56:49.380 --> 00:56:50.880 align:middle line:84% They said, OK, we'll continue to pay 00:56:50.880 --> 00:56:52.380 align:middle line:90% the banks who are paying them. 00:56:52.380 --> 00:56:53.340 align:middle line:90% Students will pay less. 00:56:53.340 --> 00:56:55.160 align:middle line:84% And the government will pay the difference. 00:56:55.160 --> 00:56:57.120 align:middle line:84% So let's be clear about what they decided. 00:56:57.120 --> 00:56:59.920 align:middle line:84% They decided the government would send checks to the banks 00:56:59.920 --> 00:57:02.640 align:middle line:84% to make profitable loans to students. 00:57:02.640 --> 00:57:06.240 align:middle line:84% That is one of the saddest episodes 00:57:06.240 --> 00:57:08.760 align:middle line:90% of my time in Washington. 00:57:08.760 --> 00:57:11.380 align:middle line:84% Fortunately, under the Obama administration, 00:57:11.380 --> 00:57:13.940 align:middle line:84% they actually got rid of the Guaranteed Student Loan program. 00:57:13.940 --> 00:57:15.523 align:middle line:84% And now the government makes the loans 00:57:15.523 --> 00:57:18.400 align:middle line:84% directly under the Direct Loan program from the government. 00:57:18.400 --> 00:57:22.360 align:middle line:90% And these don't cost a ton. 00:57:22.360 --> 00:57:25.200 align:middle line:90% These don't cost a ton. 00:57:25.200 --> 00:57:27.580 align:middle line:84% We'll come back later to the major issues with them. 00:57:27.580 --> 00:57:32.990 align:middle line:84% Now let's look at this chart and ask, how is the government 00:57:32.990 --> 00:57:35.240 align:middle line:84% spending related to the market failures we identified? 00:57:35.240 --> 00:57:37.020 align:middle line:84% We identified two market failures. 00:57:37.020 --> 00:57:40.320 align:middle line:84% One is that there may be positive spillovers. 00:57:40.320 --> 00:57:42.420 align:middle line:84% Well, that would be about the size of the pie. 00:57:42.420 --> 00:57:43.720 align:middle line:84% It's not really about how we divide it up. 00:57:43.720 --> 00:57:44.960 align:middle line:84% That would be like the bigger the spillovers, 00:57:44.960 --> 00:57:47.740 align:middle line:84% the bigger the pie should be of government spending. 00:57:47.740 --> 00:57:50.257 align:middle line:84% The second is credit market failures. 00:57:50.257 --> 00:57:52.340 align:middle line:84% Well, that would suggest that the big slice should 00:57:52.340 --> 00:57:55.780 align:middle line:90% be loans, not state spending. 00:57:55.780 --> 00:57:58.580 align:middle line:84% So in fact, just like we talked about Tiebout 00:57:58.580 --> 00:58:01.180 align:middle line:84% and talked about how the way federal and local spending 00:58:01.180 --> 00:58:04.940 align:middle line:84% lines up with the model, this doesn't line up with the model. 00:58:04.940 --> 00:58:07.580 align:middle line:84% The model would say we should be making big loans, not just 00:58:07.580 --> 00:58:10.940 align:middle line:84% spending tons of money to provide this education for free. 00:58:10.940 --> 00:58:14.300 align:middle line:90% So what is the justification? 00:58:14.300 --> 00:58:17.520 align:middle line:90% Well, it's hard, to be honest. 00:58:17.520 --> 00:58:21.420 align:middle line:84% I mean, to some extent, some of the justification 00:58:21.420 --> 00:58:24.260 align:middle line:84% is that students who get loans may not 00:58:24.260 --> 00:58:26.200 align:middle line:84% choose the socially optimal career paths. 00:58:26.200 --> 00:58:28.810 align:middle line:90% So some really cool studies-- 00:58:28.810 --> 00:58:30.560 align:middle line:84% because you also forget the state schools. 00:58:30.560 --> 00:58:31.518 align:middle line:90% Let's leave that alone. 00:58:31.518 --> 00:58:33.680 align:middle line:84% Just look at the grants versus loan size. 00:58:33.680 --> 00:58:36.740 align:middle line:84% We spend almost 10 times as much on grants and loans. 00:58:36.740 --> 00:58:39.073 align:middle line:84% Whereas the credit market failure, you want the reverse. 00:58:39.073 --> 00:58:40.657 align:middle line:84% But there's actually some cool studies 00:58:40.657 --> 00:58:43.320 align:middle line:84% which have compared people who randomly had grants versus loans 00:58:43.320 --> 00:58:44.960 align:middle line:84% based on school policies and stuff. 00:58:44.960 --> 00:58:47.280 align:middle line:84% And they find that when you get a loan, 00:58:47.280 --> 00:58:52.240 align:middle line:84% you're much more likely to pursue a higher-earning career, 00:58:52.240 --> 00:58:55.520 align:middle line:84% as opposed to what we might call a lower reward 00:58:55.520 --> 00:58:58.480 align:middle line:84% or more socially valuable relative to dollars career. 00:58:58.480 --> 00:59:02.020 align:middle line:84% So for instance, at NYU Law School, 00:59:02.020 --> 00:59:05.435 align:middle line:84% they changed from grants to loans, or vice versa. 00:59:05.435 --> 00:59:07.560 align:middle line:84% And students stopped going into public interest law 00:59:07.560 --> 00:59:10.000 align:middle line:84% and start going to corporate law. 00:59:10.000 --> 00:59:12.500 align:middle line:84% Now, I don't want to diss the corporate lawyers. 00:59:12.500 --> 00:59:13.760 align:middle line:90% There's social value there. 00:59:13.760 --> 00:59:17.720 align:middle line:84% I'm just saying that those things do affect choice. 00:59:17.720 --> 00:59:19.720 align:middle line:90% So that's one argument. 00:59:19.720 --> 00:59:21.960 align:middle line:84% That still isn't really a terrific argument 00:59:21.960 --> 00:59:25.100 align:middle line:84% for why we spend so much on state and local colleges. 00:59:25.100 --> 00:59:27.800 align:middle line:84% And that is, I think, an open and interesting question. 00:59:27.800 --> 00:59:30.152 align:middle line:84% The return to college is still high. 00:59:30.152 --> 00:59:31.860 align:middle line:84% So there's been nice randomized studies-- 00:59:31.860 --> 00:59:34.670 align:middle line:84% there's a really cool randomized study recently super cool study. 00:59:34.670 --> 00:59:35.420 align:middle line:90% What the guy did-- 00:59:35.420 --> 00:59:36.380 align:middle line:90% I saw it this summer. 00:59:36.380 --> 00:59:39.860 align:middle line:90% 00:59:39.860 --> 00:59:42.260 align:middle line:84% Basically what he did is said, there's 00:59:42.260 --> 00:59:44.420 align:middle line:90% all these colleges in Texas. 00:59:44.420 --> 00:59:46.708 align:middle line:84% And essentially, he could look at the data. 00:59:46.708 --> 00:59:48.500 align:middle line:84% And they each have somewhat different rules 00:59:48.500 --> 00:59:50.960 align:middle line:84% about the score you need to get accepted-- your GPA cutoff. 00:59:50.960 --> 00:59:53.740 align:middle line:84% And you can actually see the GPA cutoff in the data. 00:59:53.740 --> 00:59:56.860 align:middle line:84% So he did what's called-- we'll cover this later. 00:59:56.860 --> 00:59:59.995 align:middle line:84% Valerie, did you do regression discontinuity in chapter 3? 00:59:59.995 --> 01:00:00.620 align:middle line:90% AUDIENCE: Yeah. 01:00:00.620 --> 01:00:01.340 align:middle line:90% JON GRUBER: OK. 01:00:01.340 --> 01:00:03.460 align:middle line:84% Basically, this is a technique we'll spend 01:00:03.460 --> 01:00:04.840 align:middle line:90% more time in the semester. 01:00:04.840 --> 01:00:07.580 align:middle line:84% But essentially, the idea is the following. 01:00:07.580 --> 01:00:11.500 align:middle line:84% Let's say you have some score distribution You have some score 01:00:11.500 --> 01:00:17.060 align:middle line:84% distribution, and there's some random cutoff put in here. 01:00:17.060 --> 01:00:21.180 align:middle line:84% Well, these people are really different than these people. 01:00:21.180 --> 01:00:23.880 align:middle line:84% But these people aren't so different from those people. 01:00:23.880 --> 01:00:27.500 align:middle line:84% So if you know if there's a random cutoff, 01:00:27.500 --> 01:00:29.300 align:middle line:90% it's called a discontinuity. 01:00:29.300 --> 01:00:31.720 align:middle line:84% You can compare people right around that cutoff who 01:00:31.720 --> 01:00:33.500 align:middle line:90% are pretty likely identical. 01:00:33.500 --> 01:00:35.520 align:middle line:90% But one got in and one didn't. 01:00:35.520 --> 01:00:37.823 align:middle line:90% So it's like the lottery. 01:00:37.823 --> 01:00:39.740 align:middle line:84% But it doesn't come from a randomized lottery. 01:00:39.740 --> 01:00:41.320 align:middle line:84% It comes from a cutoff and comparing people 01:00:41.320 --> 01:00:41.880 align:middle line:90% right around that. 01:00:41.880 --> 01:00:43.422 align:middle line:84% Now, it's not as good as the lottery. 01:00:43.422 --> 01:00:45.088 align:middle line:84% You can say, well, people are different. 01:00:45.088 --> 01:00:46.040 align:middle line:90% They're not identical. 01:00:46.040 --> 01:00:48.207 align:middle line:84% We have methods to try to see how identical they are 01:00:48.207 --> 01:00:50.320 align:middle line:84% and there's lots of statistical tests you can do. 01:00:50.320 --> 01:00:52.237 align:middle line:84% But in principle, if you get arbitrarily close 01:00:52.237 --> 01:00:54.648 align:middle line:90% to this cutoff, it is random. 01:00:54.648 --> 01:00:56.440 align:middle line:84% Literally, if you have a million people who 01:00:56.440 --> 01:01:00.200 align:middle line:84% score 3.49 and a million people who score 3.51, 01:01:00.200 --> 01:01:03.560 align:middle line:84% and the cutoff is 3.5, those are pretty much the same people 01:01:03.560 --> 01:01:05.240 align:middle line:90% and you can compare them. 01:01:05.240 --> 01:01:07.120 align:middle line:84% And what this guy did is, he found 01:01:07.120 --> 01:01:09.968 align:middle line:84% that this line moved for different schools around Texas. 01:01:09.968 --> 01:01:11.760 align:middle line:84% There's all these little tests he could do. 01:01:11.760 --> 01:01:13.540 align:middle line:84% At every school, he do a little test. 01:01:13.540 --> 01:01:14.560 align:middle line:84% And what he found doing that was, 01:01:14.560 --> 01:01:16.060 align:middle line:84% there was a very high rate of return 01:01:16.060 --> 01:01:17.460 align:middle line:90% to public education in Texas. 01:01:17.460 --> 01:01:21.600 align:middle line:90% 01:01:21.600 --> 01:01:23.240 align:middle line:84% That just says the size of the circle 01:01:23.240 --> 01:01:24.780 align:middle line:90% may be, if anything, too small. 01:01:24.780 --> 01:01:27.280 align:middle line:84% It still doesn't say why it should be public schools and not 01:01:27.280 --> 01:01:28.860 align:middle line:90% more loans to private schools. 01:01:28.860 --> 01:01:32.180 align:middle line:84% Now, the answer to that, to come to that, 01:01:32.180 --> 01:01:35.800 align:middle line:84% we turn to the last big topic I want to cover, 01:01:35.800 --> 01:01:37.088 align:middle line:90% which is student loans. 01:01:37.088 --> 01:01:38.880 align:middle line:84% And why don't we just all do student loans? 01:01:38.880 --> 01:01:41.088 align:middle line:84% And here, we have the big issue of student loan debt. 01:01:41.088 --> 01:01:44.300 align:middle line:90% As you know, we have about-- 01:01:44.300 --> 01:01:47.040 align:middle line:84% I think it's currently several trillion dollars. 01:01:47.040 --> 01:01:50.100 align:middle line:84% I don't remember what the stock of student loan debt is now. 01:01:50.100 --> 01:01:52.080 align:middle line:84% It's $1.4 trillion as of a few years ago. 01:01:52.080 --> 01:01:53.760 align:middle line:90% It's bigger than that now. 01:01:53.760 --> 01:01:57.220 align:middle line:84% It's a huge stock of debt that students are bearing. 01:01:57.220 --> 01:02:00.260 align:middle line:84% And a lot of students aren't paying it back. 01:02:00.260 --> 01:02:03.020 align:middle line:84% 24% of students are delinquent at least one student loan 01:02:03.020 --> 01:02:04.340 align:middle line:90% payment. 01:02:04.340 --> 01:02:07.780 align:middle line:84% 5% have defaulted altogether on student loan payments. 01:02:07.780 --> 01:02:12.180 align:middle line:84% And 22% of workers have student loan payments 01:02:12.180 --> 01:02:14.360 align:middle line:84% that amount to more than 12% of their income. 01:02:14.360 --> 01:02:15.740 align:middle line:84% That is, more than one in eight of every dollar 01:02:15.740 --> 01:02:17.448 align:middle line:84% is going to pay back their student loans. 01:02:17.448 --> 01:02:19.180 align:middle line:90% That's a huge loan burden. 01:02:19.180 --> 01:02:22.043 align:middle line:84% So one argument for why we should rely less on loans 01:02:22.043 --> 01:02:24.460 align:middle line:84% and more on spending-- although again, it doesn't tell you 01:02:24.460 --> 01:02:25.860 align:middle line:90% the purple versus the green. 01:02:25.860 --> 01:02:28.373 align:middle line:84% It's just the purple and the green versus the orange. 01:02:28.373 --> 01:02:30.040 align:middle line:84% Essentially, these student loans, people 01:02:30.040 --> 01:02:32.080 align:middle line:90% can't afford to pay them back. 01:02:32.080 --> 01:02:37.000 align:middle line:84% And that has led to a big debate over allowing people 01:02:37.000 --> 01:02:39.458 align:middle line:84% to default on their student loans, forgiving student loans. 01:02:39.458 --> 01:02:41.375 align:middle line:84% This is something that Elizabeth Warren really 01:02:41.375 --> 01:02:43.140 align:middle line:84% raised people's attention as a candidate. 01:02:43.140 --> 01:02:45.723 align:middle line:84% And the Biden administration has made a focus of their efforts 01:02:45.723 --> 01:02:47.505 align:middle line:90% to forgive student loans. 01:02:47.505 --> 01:02:49.630 align:middle line:84% They had a big plan to forgive student loans, which 01:02:49.630 --> 01:02:51.547 align:middle line:84% would have forgiven like a huge share of them. 01:02:51.547 --> 01:02:53.320 align:middle line:84% The Supreme Court turned it down, 01:02:53.320 --> 01:02:54.900 align:middle line:84% and then they put in a smaller plan. 01:02:54.900 --> 01:02:56.200 align:middle line:84% But still, they released student loans of something 01:02:56.200 --> 01:02:57.680 align:middle line:90% like four million people. 01:02:57.680 --> 01:03:02.560 align:middle line:84% Four million people got their student loans paid for. 01:03:02.560 --> 01:03:04.800 align:middle line:84% They didn't have to pay back their student loans. 01:03:04.800 --> 01:03:07.420 align:middle line:84% So the question is, what is the right strategy here? 01:03:07.420 --> 01:03:10.563 align:middle line:84% There's a lot of arguments against just allowing people 01:03:10.563 --> 01:03:11.980 align:middle line:84% to default on their student loans. 01:03:11.980 --> 01:03:14.000 align:middle line:84% And one is the moral hazard argument. 01:03:14.000 --> 01:03:15.260 align:middle line:84% If you're going to let some people to follow, 01:03:15.260 --> 01:03:16.843 align:middle line:84% other people won't pay because they'll 01:03:16.843 --> 01:03:18.560 align:middle line:90% think they'll get to default. 01:03:18.560 --> 01:03:20.755 align:middle line:84% There's also an equity argument, which is, 01:03:20.755 --> 01:03:22.880 align:middle line:84% what about people who paid back their student loans 01:03:22.880 --> 01:03:24.630 align:middle line:84% and struggled and worked really hard to pay them back? 01:03:24.630 --> 01:03:27.090 align:middle line:84% And someone else just had the government pay them back. 01:03:27.090 --> 01:03:29.650 align:middle line:84% That seems a little bit inequitable. 01:03:29.650 --> 01:03:31.930 align:middle line:84% So there's some arguments against that. 01:03:31.930 --> 01:03:34.770 align:middle line:84% The argument for it is, a lot of these loans 01:03:34.770 --> 01:03:37.350 align:middle line:84% are based on students who are essentially 01:03:37.350 --> 01:03:38.630 align:middle line:90% fraudulently educated. 01:03:38.630 --> 01:03:40.890 align:middle line:84% I'm not just talking about Trump University here. 01:03:40.890 --> 01:03:44.650 align:middle line:84% But I'm talking about a broad sector of for-profit colleges 01:03:44.650 --> 01:03:49.850 align:middle line:84% that have actually shown to be very bad actors in many cases. 01:03:49.850 --> 01:03:52.270 align:middle line:84% For example, the ITT Technical Institute-- 01:03:52.270 --> 01:03:54.570 align:middle line:84% do you guys ever see ITT Technical Institute ads on TV? 01:03:54.570 --> 01:03:58.010 align:middle line:84% They ran an aggressive campaign ad in radio. 01:03:58.010 --> 01:04:00.230 align:middle line:84% They found these commercials were dishonest. 01:04:00.230 --> 01:04:02.530 align:middle line:84% The government study-- were dishonest and targeted 01:04:02.530 --> 01:04:04.170 align:middle line:90% low-income students. 01:04:04.170 --> 01:04:06.685 align:middle line:84% Basically, ITT would reach out to you and say, 01:04:06.685 --> 01:04:07.810 align:middle line:90% we're going to educate you. 01:04:07.810 --> 01:04:09.268 align:middle line:84% Here, it's going to be free to you, 01:04:09.268 --> 01:04:12.230 align:middle line:84% take out this big student loan to pay for it. 01:04:12.230 --> 01:04:13.890 align:middle line:90% And then they would give you-- 01:04:13.890 --> 01:04:16.410 align:middle line:84% and then they wouldn't give you a really valuable education. 01:04:16.410 --> 01:04:20.050 align:middle line:84% And basically, more generally, if you 01:04:20.050 --> 01:04:22.710 align:middle line:84% look at all students heading to for-profit colleges, 01:04:22.710 --> 01:04:26.085 align:middle line:84% one third graduate within eight years. 01:04:26.085 --> 01:04:28.210 align:middle line:84% So you define graduating graduate with eight years. 01:04:28.210 --> 01:04:30.610 align:middle line:84% 2/3 don't even graduate within eight years of attending. 01:04:30.610 --> 01:04:34.550 align:middle line:90% 01:04:34.550 --> 01:04:36.110 align:middle line:84% A really cool study looked at what 01:04:36.110 --> 01:04:38.190 align:middle line:84% happened when for-profit colleges closed 01:04:38.190 --> 01:04:40.350 align:middle line:84% and the kids moved to community colleges. 01:04:40.350 --> 01:04:44.310 align:middle line:84% The graduation rates went up from 34% to 60%. 01:04:44.310 --> 01:04:47.450 align:middle line:84% So schools on average are clearly doing a terrible job. 01:04:47.450 --> 01:04:49.920 align:middle line:84% And there's fun anecdotes like ITT. 01:04:49.920 --> 01:04:51.670 align:middle line:84% But there's actually the data, like I just 01:04:51.670 --> 01:04:53.962 align:middle line:84% said to you, which is on average for-profit schools are 01:04:53.962 --> 01:04:57.790 align:middle line:84% doing a bad job, and in particular, a misleading job. 01:04:57.790 --> 01:04:59.810 align:middle line:90% Now, as always, it's not easy. 01:04:59.810 --> 01:05:01.310 align:middle line:84% On the other hand, there are studies 01:05:01.310 --> 01:05:04.190 align:middle line:84% which show that many disadvantaged students who 01:05:04.190 --> 01:05:07.510 align:middle line:84% did not have the ability to access traditional education 01:05:07.510 --> 01:05:09.500 align:middle line:84% succeeded in for-profit colleges. 01:05:09.500 --> 01:05:11.250 align:middle line:84% So there are students who succeeded there, 01:05:11.250 --> 01:05:12.630 align:middle line:84% and it provided a road to success 01:05:12.630 --> 01:05:15.590 align:middle line:84% for some students who didn't otherwise have access. 01:05:15.590 --> 01:05:19.550 align:middle line:84% But on average, they don't seem to be performing very well. 01:05:19.550 --> 01:05:24.470 align:middle line:84% And so the argument was, these people went to really scummy, 01:05:24.470 --> 01:05:26.450 align:middle line:84% for-profit colleges who misled them and made 01:05:26.450 --> 01:05:27.830 align:middle line:90% them take big student loans. 01:05:27.830 --> 01:05:30.250 align:middle line:84% They now have this big student loan debt. 01:05:30.250 --> 01:05:32.810 align:middle line:84% We shouldn't make them pay it back. 01:05:32.810 --> 01:05:39.370 align:middle line:90% Now, that is an argument. 01:05:39.370 --> 01:05:40.810 align:middle line:84% But it turns out nobody's talking 01:05:40.810 --> 01:05:42.650 align:middle line:84% about forgiving loan debt just for those students, 01:05:42.650 --> 01:05:43.990 align:middle line:84% forgiving loan debt for everyone. 01:05:43.990 --> 01:05:45.823 align:middle line:84% So it's hard to explain, that's the minority 01:05:45.823 --> 01:05:47.290 align:middle line:90% of students with debt. 01:05:47.290 --> 01:05:49.770 align:middle line:84% So the question is, are there better things we can do? 01:05:49.770 --> 01:05:52.470 align:middle line:84% And the Obama administration tried to do some better things. 01:05:52.470 --> 01:05:56.550 align:middle line:84% For example, they increased the quality available to students, 01:05:56.550 --> 01:05:59.690 align:middle line:84% so you understood what you're getting into. 01:05:59.690 --> 01:06:01.890 align:middle line:84% Also, a more aggressive strategy was 01:06:01.890 --> 01:06:05.090 align:middle line:84% to basically deny eligibility for student loans 01:06:05.090 --> 01:06:08.027 align:middle line:84% to schools that didn't show a good enough graduation rate. 01:06:08.027 --> 01:06:10.610 align:middle line:84% Basically, unless your schools showed a good enough graduation 01:06:10.610 --> 01:06:12.310 align:middle line:84% rate, you couldn't get student loans. 01:06:12.310 --> 01:06:13.970 align:middle line:90% This is an interesting example. 01:06:13.970 --> 01:06:16.570 align:middle line:84% We talked here about a disadvantage 01:06:16.570 --> 01:06:19.910 align:middle line:84% of vouchers as excessive specialization, 01:06:19.910 --> 01:06:22.035 align:middle line:90% or bad performance in this case. 01:06:22.035 --> 01:06:23.910 align:middle line:84% This is a way you could try to regulate that. 01:06:23.910 --> 01:06:27.350 align:middle line:84% You say, look, we are going to regulate it to make sure 01:06:27.350 --> 01:06:30.150 align:middle line:84% that you're a decent school by only giving you only-- we're not 01:06:30.150 --> 01:06:31.552 align:middle line:90% going to shut you down. 01:06:31.552 --> 01:06:33.010 align:middle line:84% We're going to let the market work. 01:06:33.010 --> 01:06:34.760 align:middle line:84% But you can't get student loans unless you 01:06:34.760 --> 01:06:37.710 align:middle line:90% can show you're decent. 01:06:37.710 --> 01:06:40.350 align:middle line:84% The Trump administration killed that regulation. 01:06:40.350 --> 01:06:44.390 align:middle line:84% And it never, as far as I know, never went into effect. 01:06:44.390 --> 01:06:46.670 align:middle line:90% So that's one approach. 01:06:46.670 --> 01:06:48.670 align:middle line:84% Another approach that I like a lot is 01:06:48.670 --> 01:06:52.790 align:middle line:84% to do what the UK and Australia do, 01:06:52.790 --> 01:06:54.310 align:middle line:84% which is to replace our current loan 01:06:54.310 --> 01:06:59.050 align:middle line:84% program with what's called an income-contingent loan program. 01:06:59.050 --> 01:07:02.030 align:middle line:84% Basically, we'll loan you money and you pay it back 01:07:02.030 --> 01:07:04.670 align:middle line:90% as a fraction of your income. 01:07:04.670 --> 01:07:08.150 align:middle line:84% So if you end up not making much money, you don't owe much back. 01:07:08.150 --> 01:07:10.550 align:middle line:84% You get rich, you owe a lot back. 01:07:10.550 --> 01:07:13.390 align:middle line:84% And just the way it works in the UK is basically, 01:07:13.390 --> 01:07:17.290 align:middle line:84% they exempt the first, like, 15,000 pounds of your earnings. 01:07:17.290 --> 01:07:21.850 align:middle line:84% And every pound above 15,000, you pay 9% back in student loan 01:07:21.850 --> 01:07:26.310 align:middle line:84% payments until you've paid back your loan or until 20 years, 01:07:26.310 --> 01:07:29.142 align:middle line:90% at which point it just stops. 01:07:29.142 --> 01:07:31.350 align:middle line:84% Basically, the government is still losing some money. 01:07:31.350 --> 01:07:32.850 align:middle line:84% They're not making them pay back enough to make up 01:07:32.850 --> 01:07:35.130 align:middle line:84% if people default. But it's not nearly as much 01:07:35.130 --> 01:07:36.670 align:middle line:90% as just paying for college. 01:07:36.670 --> 01:07:39.170 align:middle line:84% And it's doing it in a more equitable way because people pay 01:07:39.170 --> 01:07:41.850 align:middle line:90% back a share of their earnings. 01:07:41.850 --> 01:07:46.650 align:middle line:84% Now, the US has a program like this. 01:07:46.650 --> 01:07:49.370 align:middle line:84% And in fact, it's fairly popular. 01:07:49.370 --> 01:07:52.450 align:middle line:84% The problem is that basically, people are still defaulting 01:07:52.450 --> 01:07:53.890 align:middle line:90% on it, not paying back. 01:07:53.890 --> 01:07:56.130 align:middle line:84% So one change the Biden administration made 01:07:56.130 --> 01:07:57.850 align:middle line:84% was to make the income-contingent loan 01:07:57.850 --> 01:08:00.050 align:middle line:90% program much more generous. 01:08:00.050 --> 01:08:02.950 align:middle line:84% Basically, it used to be you had to pay up to 10% of your income. 01:08:02.950 --> 01:08:04.810 align:middle line:84% Now it's 5% of your income or something like that. 01:08:04.810 --> 01:08:06.190 align:middle line:84% I don't know exactly what they did. 01:08:06.190 --> 01:08:07.982 align:middle line:84% But they made the income-contingent program 01:08:07.982 --> 01:08:12.530 align:middle line:84% much more generous to try to encourage use of it. 01:08:12.530 --> 01:08:15.210 align:middle line:84% There's two problems with this program. 01:08:15.210 --> 01:08:17.310 align:middle line:84% Once again, you'll get used to me saying that. 01:08:17.310 --> 01:08:20.710 align:middle line:84% Harry Truman once said, I wish I had a one-handed economist so he 01:08:20.710 --> 01:08:22.790 align:middle line:84% couldn't say "on the other hand." 01:08:22.790 --> 01:08:25.810 align:middle line:84% The problem with this is twofold. 01:08:25.810 --> 01:08:28.750 align:middle line:84% First of all, as we'll learn later in the semester, 01:08:28.750 --> 01:08:31.370 align:middle line:84% taxes have distortionary consequences. 01:08:31.370 --> 01:08:33.870 align:middle line:84% And if you're going to have to pay an extra 9% in everything 01:08:33.870 --> 01:08:36.830 align:middle line:84% you earn, you might decide not to work as hard. 01:08:36.830 --> 01:08:37.850 align:middle line:90% I have a paper on that. 01:08:37.850 --> 01:08:38.683 align:middle line:90% We looked in the UK. 01:08:38.683 --> 01:08:39.810 align:middle line:90% There's no effect there. 01:08:39.810 --> 01:08:42.350 align:middle line:84% It's meaningless in terms of people don't really 01:08:42.350 --> 01:08:44.430 align:middle line:90% respond to that. 01:08:44.430 --> 01:08:47.510 align:middle line:84% The second concern is, does this really 01:08:47.510 --> 01:08:49.590 align:middle line:90% work when it's not mandatory? 01:08:49.590 --> 01:08:51.990 align:middle line:84% After all, if you're going to college 01:08:51.990 --> 01:08:53.689 align:middle line:84% and you're anticipating your earnings, 01:08:53.689 --> 01:08:55.981 align:middle line:84% who's going to want to take the income-contingent loan? 01:08:55.981 --> 01:08:58.024 align:middle line:90% 01:08:58.024 --> 01:08:59.649 align:middle line:84% And who's going to not want to take it? 01:08:59.649 --> 01:09:00.149 align:middle line:90% Steven. 01:09:00.149 --> 01:09:01.053 align:middle line:90% AUDIENCE: [INAUDIBLE] 01:09:01.053 --> 01:09:02.470 align:middle line:84% JON GRUBER: Yeah, people are going 01:09:02.470 --> 01:09:04.210 align:middle line:84% to want to take it those of lower income. 01:09:04.210 --> 01:09:05.710 align:middle line:84% The ones who aren't going to take it 01:09:05.710 --> 01:09:06.918 align:middle line:90% are the ones who participate. 01:09:06.918 --> 01:09:08.830 align:middle line:90% They're going to go to get rich. 01:09:08.830 --> 01:09:11.229 align:middle line:84% This is a problem that we'll discuss next lecture 01:09:11.229 --> 01:09:13.470 align:middle line:84% called adverse selection, a common problem 01:09:13.470 --> 01:09:15.210 align:middle line:90% in insurance markets. 01:09:15.210 --> 01:09:17.649 align:middle line:84% Basically, the problem is that if only the people who 01:09:17.649 --> 01:09:19.566 align:middle line:84% know they're going to have low income take it, 01:09:19.566 --> 01:09:21.729 align:middle line:84% you'll lose a lot of money on the program. 01:09:21.729 --> 01:09:24.271 align:middle line:84% So the other problem is it's hard to have a program like this 01:09:24.271 --> 01:09:28.729 align:middle line:84% be successful unless it's mandatory, as it is in the UK 01:09:28.729 --> 01:09:30.290 align:middle line:90% and in Australia. 01:09:30.290 --> 01:09:32.729 align:middle line:90% You cannot take a loan. 01:09:32.729 --> 01:09:35.069 align:middle line:84% We're not saying you could not take a loan. 01:09:35.069 --> 01:09:37.915 align:middle line:84% But if you want to take a loan, it has to be income-contingent. 01:09:37.915 --> 01:09:40.290 align:middle line:84% So it wouldn't fully solve the adverse selection problem, 01:09:40.290 --> 01:09:41.529 align:middle line:84% because people who knew they were rich, 01:09:41.529 --> 01:09:43.810 align:middle line:84% who were already rich, could just not take loans. 01:09:43.810 --> 01:09:45.630 align:middle line:84% But they probably, by and large, already don't take loans. 01:09:45.630 --> 01:09:46.350 align:middle line:90% So it's OK. 01:09:46.350 --> 01:09:46.850 align:middle line:90% Yeah. 01:09:46.850 --> 01:09:49.441 align:middle line:84% AUDIENCE: [INAUDIBLE] are people allowed to [INAUDIBLE] take 01:09:49.441 --> 01:09:52.149 align:middle line:84% private loans if they think they will [? have ?] a higher income? 01:09:52.149 --> 01:09:54.250 align:middle line:84% JON GRUBER: No, you cannot take a loan to pay back your college, 01:09:54.250 --> 01:09:55.930 align:middle line:90% unless it's from this program. 01:09:55.930 --> 01:09:57.972 align:middle line:84% And also private loans, they wouldn't loan to you 01:09:57.972 --> 01:09:59.748 align:middle line:90% because there's no collateral. 01:09:59.748 --> 01:10:02.290 align:middle line:84% I mean, the private market can't exist, and probably wouldn't 01:10:02.290 --> 01:10:04.050 align:middle line:90% exist, otherwise. 01:10:04.050 --> 01:10:06.190 align:middle line:90% OK, other questions about that. 01:10:06.190 --> 01:10:07.890 align:middle line:90% Yeah, Enoch. 01:10:07.890 --> 01:10:10.970 align:middle line:84% AUDIENCE: You mentioned use of colleges and basically 01:10:10.970 --> 01:10:12.710 align:middle line:90% giving you useless education. 01:10:12.710 --> 01:10:17.150 align:middle line:84% Is there anything about useless degrees or degree paths that-- 01:10:17.150 --> 01:10:19.050 align:middle line:84% JON GRUBER: It's a fantastic question. 01:10:19.050 --> 01:10:21.790 align:middle line:84% So in some Scandinavian countries, 01:10:21.790 --> 01:10:24.473 align:middle line:84% where they had a certain number of slots in higher education, 01:10:24.473 --> 01:10:26.390 align:middle line:84% they would actually change the number of slots 01:10:26.390 --> 01:10:27.765 align:middle line:84% to reflect where they felt people 01:10:27.765 --> 01:10:30.610 align:middle line:84% should be educated-- more STEM slots, fewer English slots. 01:10:30.610 --> 01:10:32.790 align:middle line:84% I mean, I'm preaching to the choir here. 01:10:32.790 --> 01:10:37.110 align:middle line:84% But basically, that is a really deep and hard question, 01:10:37.110 --> 01:10:40.430 align:middle line:84% which is fundamentally, what is the nature of the education 01:10:40.430 --> 01:10:41.790 align:middle line:90% being delivered? 01:10:41.790 --> 01:10:44.265 align:middle line:84% So the Obama administration said, 01:10:44.265 --> 01:10:45.890 align:middle line:84% we'll target loans on graduation rates. 01:10:45.890 --> 01:10:47.030 align:middle line:90% You could go further. 01:10:47.030 --> 01:10:48.990 align:middle line:84% You could target loans on who's ending up 01:10:48.990 --> 01:10:50.470 align:middle line:90% in a successful career. 01:10:50.470 --> 01:10:52.972 align:middle line:84% The problem is, you have to define "success." 01:10:52.972 --> 01:10:54.430 align:middle line:84% And in some sense, the problem with 01:10:54.430 --> 01:10:56.055 align:middle line:84% the income-contingent system is that it 01:10:56.055 --> 01:10:57.710 align:middle line:90% has the opposite incentive. 01:10:57.710 --> 01:10:59.950 align:middle line:84% If you really think there's a lot of education that's 01:10:59.950 --> 01:11:00.950 align:middle line:90% useless-- 01:11:00.950 --> 01:11:03.790 align:middle line:84% education leads to useless jobs, then you don't want to do that. 01:11:03.790 --> 01:11:06.630 align:middle line:84% Once again, do you think wages reflect social marginal value 01:11:06.630 --> 01:11:09.050 align:middle line:84% product or private marginal value product? 01:11:09.050 --> 01:11:12.200 align:middle line:84% We think wages reflect social value marginal product, 01:11:12.200 --> 01:11:14.400 align:middle line:84% then we should basically reward education leads 01:11:14.400 --> 01:11:16.040 align:middle line:90% to the highest wages. 01:11:16.040 --> 01:11:19.060 align:middle line:84% If we think wages reflect private value marginal product, 01:11:19.060 --> 01:11:21.820 align:middle line:84% but that there's externalities that are aside from wages, 01:11:21.820 --> 01:11:23.460 align:middle line:84% we shouldn't necessarily target it. 01:11:23.460 --> 01:11:26.640 align:middle line:84% So that's a really interesting and hard question. 01:11:26.640 --> 01:11:28.960 align:middle line:90% Other questions? 01:11:28.960 --> 01:11:32.840 align:middle line:84% OK, there's a cool study which tried to decide, 01:11:32.840 --> 01:11:35.280 align:middle line:90% is going to MIT worth it? 01:11:35.280 --> 01:11:38.160 align:middle line:84% Is going to better schools worth it 01:11:38.160 --> 01:11:39.980 align:middle line:84% compared to going to worse schools? 01:11:39.980 --> 01:11:42.605 align:middle line:90% How would you test that? 01:11:42.605 --> 01:11:44.480 align:middle line:84% I mean, obviously, you're different than kids 01:11:44.480 --> 01:11:45.647 align:middle line:90% who go to different schools. 01:11:45.647 --> 01:11:46.977 align:middle line:90% How would you test? 01:11:46.977 --> 01:11:48.060 align:middle line:90% How would you set that up? 01:11:48.060 --> 01:11:50.890 align:middle line:84% Let's say you wanted to test, is going to school 01:11:50.890 --> 01:11:52.140 align:middle line:90% A better than going to school? 01:11:52.140 --> 01:11:52.680 align:middle line:90% B? 01:11:52.680 --> 01:11:53.220 align:middle line:90% Yeah. 01:11:53.220 --> 01:11:55.582 align:middle line:84% AUDIENCE: You take a look at a sample 01:11:55.582 --> 01:11:57.540 align:middle line:84% of people that have been accepted to [? MIT. ?] 01:11:57.540 --> 01:11:58.582 align:middle line:90% JON GRUBER: There you go. 01:11:58.582 --> 01:12:00.960 align:middle line:84% There you have one of the most famous education studies, 01:12:00.960 --> 01:12:02.960 align:middle line:90% the Dale-Krueger study. 01:12:02.960 --> 01:12:04.785 align:middle line:90% What they did is exactly that. 01:12:04.785 --> 01:12:07.160 align:middle line:84% They found a bunch of people accepted to multiple schools 01:12:07.160 --> 01:12:08.660 align:middle line:84% and asked, if you went to the better 01:12:08.660 --> 01:12:11.300 align:middle line:84% school you were accepted to do, did you end up doing better? 01:12:11.300 --> 01:12:14.860 align:middle line:84% And the answer was actually remarkably little. 01:12:14.860 --> 01:12:17.473 align:middle line:90% A lot of it is screening. 01:12:17.473 --> 01:12:19.140 align:middle line:84% The answer is, the students who are good 01:12:19.140 --> 01:12:21.660 align:middle line:84% did pretty well wherever they went. 01:12:21.660 --> 01:12:24.380 align:middle line:84% And indeed, there's a really cool study done 01:12:24.380 --> 01:12:26.180 align:middle line:90% by [? Rothschild ?] at Harvard. 01:12:26.180 --> 01:12:30.540 align:middle line:84% They got data on parental income and subsequent children income 01:12:30.540 --> 01:12:33.540 align:middle line:84% and ranked schools by their income mobility. 01:12:33.540 --> 01:12:35.580 align:middle line:84% That is, how much did schools improve your rank 01:12:35.580 --> 01:12:37.268 align:middle line:90% in the income distribution? 01:12:37.268 --> 01:12:38.060 align:middle line:90% And the answer is-- 01:12:38.060 --> 01:12:39.310 align:middle line:90% I don't remember what MIT did. 01:12:39.310 --> 01:12:40.727 align:middle line:84% But the answer is the best schools 01:12:40.727 --> 01:12:43.340 align:middle line:84% are schools like CCNY in New York-- the City College of New 01:12:43.340 --> 01:12:45.360 align:middle line:84% York, which takes a lot of disadvantaged kids, 01:12:45.360 --> 01:12:46.818 align:middle line:84% gives them a really good education, 01:12:46.818 --> 01:12:49.500 align:middle line:84% and promotes them up the income distribution. 01:12:49.500 --> 01:12:53.140 align:middle line:84% Of the major universities, Berkeley was the best. 01:12:53.140 --> 01:12:56.220 align:middle line:84% And I don't remember who did badly, 01:12:56.220 --> 01:12:58.032 align:middle line:84% but I'm sure MIT didn't do that well. 01:12:58.032 --> 01:12:59.240 align:middle line:90% I don't know how well we did. 01:12:59.240 --> 01:13:01.120 align:middle line:84% We probably did better than a lot of the Ivies, 01:13:01.120 --> 01:13:02.537 align:middle line:84% because once again, we're training 01:13:02.537 --> 01:13:04.700 align:middle line:90% in high-wage professions. 01:13:04.700 --> 01:13:08.360 align:middle line:84% So anyway, bottom line-- higher education, fascinating topic, 01:13:08.360 --> 01:13:10.840 align:middle line:84% lots of work to be done, very relevant to all of your lives 01:13:10.840 --> 01:13:11.500 align:middle line:90% right now. 01:13:11.500 --> 01:13:14.660 align:middle line:84% A-- The bottom line is the government 01:13:14.660 --> 01:13:15.660 align:middle line:90% should clearly be in it. 01:13:15.660 --> 01:13:17.220 align:middle line:90% There's positive externalities. 01:13:17.220 --> 01:13:19.240 align:middle line:90% How much is still unclear. 01:13:19.240 --> 01:13:23.200 align:middle line:84% B-- it's really unclear why we spend our money the way we do, 01:13:23.200 --> 01:13:24.502 align:middle line:90% as opposed to more loans. 01:13:24.502 --> 01:13:25.960 align:middle line:84% But C, if we rely more on loans, we 01:13:25.960 --> 01:13:27.335 align:middle line:84% need to think about the structure 01:13:27.335 --> 01:13:30.020 align:middle line:84% of who gets those loans and how they're paid back. 01:13:30.020 --> 01:13:31.960 align:middle line:90% Any other questions on that? 01:13:31.960 --> 01:13:32.880 align:middle line:90% Yeah, Enoch, 01:13:32.880 --> 01:13:33.880 align:middle line:90% AUDIENCE: Why don't we-- 01:13:33.880 --> 01:13:37.080 align:middle line:84% I know a couple years back, they did free college. 01:13:37.080 --> 01:13:39.280 align:middle line:84% JON GRUBER: Free college is another idea. 01:13:39.280 --> 01:13:41.040 align:middle line:84% Once again, first of all, the main reason, 01:13:41.040 --> 01:13:43.720 align:middle line:84% it would be unbelievably expensive. 01:13:43.720 --> 01:13:46.260 align:middle line:84% And second of all comes to your earlier question, 01:13:46.260 --> 01:13:49.440 align:middle line:84% which is not obvious college is the right thing for everyone. 01:13:49.440 --> 01:13:50.960 align:middle line:84% There's a big debate right now about 01:13:50.960 --> 01:13:52.860 align:middle line:84% whether people are wasting their money in college, 01:13:52.860 --> 01:13:55.382 align:middle line:84% getting needlessly in debt to something that isn't actually 01:13:55.382 --> 01:13:57.840 align:middle line:84% making them do any better than not going to college at all. 01:13:57.840 --> 01:14:01.440 align:middle line:84% So I think that's an interesting debate to be had. 01:14:01.440 --> 01:14:03.950 align:middle line:90% All right, let's stop there. 01:14:03.950 --> 01:14:08.000 align:middle line:90%