WEBVTT 00:00:00.000 --> 00:00:05.769 align:middle line:84% [SQUEAKING] [RUSTLING] [CLICKING] 00:00:05.769 --> 00:00:10.080 align:middle line:90% 00:00:10.080 --> 00:00:11.760 align:middle line:84% JON GRUBER: Today, we are going to be 00:00:11.760 --> 00:00:15.480 align:middle line:84% talking about continuing our discussion of health 00:00:15.480 --> 00:00:16.683 align:middle line:90% externalities. 00:00:16.683 --> 00:00:19.100 align:middle line:84% Now, if you remember last time we started with the example 00:00:19.100 --> 00:00:21.840 align:middle line:84% smoking, to remind you we talked about smoking 00:00:21.840 --> 00:00:25.320 align:middle line:84% as a negative consumption externality, something 00:00:25.320 --> 00:00:28.440 align:middle line:84% where the consumption of cigarettes by one person 00:00:28.440 --> 00:00:30.240 align:middle line:90% makes other people worse off. 00:00:30.240 --> 00:00:32.897 align:middle line:84% We talked about where the externalities arise. 00:00:32.897 --> 00:00:34.480 align:middle line:84% First of all, we talked about the fact 00:00:34.480 --> 00:00:37.520 align:middle line:84% that smoking in standard economic models 00:00:37.520 --> 00:00:40.145 align:middle line:84% should only be regulated if it creates externalities on others. 00:00:40.145 --> 00:00:42.520 align:middle line:84% So if smokers just sit on a rock in the middle of nowhere 00:00:42.520 --> 00:00:44.640 align:middle line:84% and kill themselves that's their problem. 00:00:44.640 --> 00:00:47.400 align:middle line:84% In economics, the reason we regulate goods 00:00:47.400 --> 00:00:49.220 align:middle line:84% is the damage it will do to others. 00:00:49.220 --> 00:00:52.440 align:middle line:84% The question is, where do the externalities arise for smoking? 00:00:52.440 --> 00:00:54.520 align:middle line:84% And we started-- and if you look at the-- 00:00:54.520 --> 00:00:57.680 align:middle line:84% we talked last time about some of the externalities for smoking 00:00:57.680 --> 00:01:00.970 align:middle line:84% and when they actually are return externalities. 00:01:00.970 --> 00:01:03.010 align:middle line:84% And so the first thing we talked about 00:01:03.010 --> 00:01:07.010 align:middle line:90% was increased health care costs. 00:01:07.010 --> 00:01:09.650 align:middle line:90% Health care costs. 00:01:09.650 --> 00:01:11.630 align:middle line:84% And we said, is that an externality? 00:01:11.630 --> 00:01:13.810 align:middle line:84% We said, well, it's only externality if the smoker 00:01:13.810 --> 00:01:15.370 align:middle line:90% themselves doesn't bear it. 00:01:15.370 --> 00:01:17.210 align:middle line:84% If smokers pay for the increased health 00:01:17.210 --> 00:01:19.907 align:middle line:84% care costs through higher premiums and other things, 00:01:19.907 --> 00:01:20.990 align:middle line:90% that's not an externality. 00:01:20.990 --> 00:01:22.150 align:middle line:90% They've internalized it. 00:01:22.150 --> 00:01:25.330 align:middle line:84% It's only externality if smokers are, say, uninsured 00:01:25.330 --> 00:01:26.750 align:middle line:90% or paid for by the government. 00:01:26.750 --> 00:01:27.930 align:middle line:90% They don't bear it. 00:01:27.930 --> 00:01:30.448 align:middle line:84% We can talk about other externalities in the same way. 00:01:30.448 --> 00:01:31.990 align:middle line:84% Let's think about another externality 00:01:31.990 --> 00:01:33.948 align:middle line:84% to smoking, which is work potential and effects 00:01:33.948 --> 00:01:36.530 align:middle line:90% on workplace productivity. 00:01:36.530 --> 00:01:39.650 align:middle line:84% So smokers have been measured to be 00:01:39.650 --> 00:01:43.590 align:middle line:84% less productive in the workplace than non-smokers. 00:01:43.590 --> 00:01:45.770 align:middle line:84% They take smoke breaks, et cetera. 00:01:45.770 --> 00:01:48.890 align:middle line:90% Is that an externality or not? 00:01:48.890 --> 00:01:49.730 align:middle line:90% Yeah. 00:01:49.730 --> 00:01:50.530 align:middle line:90% AUDIENCE: Depends. 00:01:50.530 --> 00:01:54.580 align:middle line:84% If they're [INAUDIBLE] and they [INAUDIBLE]. 00:01:54.580 --> 00:01:55.330 align:middle line:90% JON GRUBER: Right. 00:01:55.330 --> 00:01:58.830 align:middle line:84% If it's reflected in their wages, it's internalized. 00:01:58.830 --> 00:02:00.070 align:middle line:90% It's not an externality. 00:02:00.070 --> 00:02:02.490 align:middle line:84% It's only an externality if they're not paying for it. 00:02:02.490 --> 00:02:04.250 align:middle line:84% The best estimates are they're not. 00:02:04.250 --> 00:02:05.288 align:middle line:90% So it is an externality. 00:02:05.288 --> 00:02:07.330 align:middle line:84% Moreover, there's a lot of good economic research 00:02:07.330 --> 00:02:09.250 align:middle line:84% that the productivity of one worker 00:02:09.250 --> 00:02:11.432 align:middle line:84% spills over to the productivity of other workers. 00:02:11.432 --> 00:02:13.390 align:middle line:84% And as a result, even if they're paying for it, 00:02:13.390 --> 00:02:15.170 align:middle line:84% if they're less productive, that could make their coworkers 00:02:15.170 --> 00:02:16.130 align:middle line:90% less productive. 00:02:16.130 --> 00:02:18.210 align:middle line:84% That would also be an externality. 00:02:18.210 --> 00:02:19.650 align:middle line:90% So that's another example. 00:02:19.650 --> 00:02:24.010 align:middle line:84% A third example that's quite striking is fires. 00:02:24.010 --> 00:02:26.990 align:middle line:84% Turns out that smokers cause a lot of fires in the world. 00:02:26.990 --> 00:02:30.530 align:middle line:84% In 2011, fires started by smokers in the US 00:02:30.530 --> 00:02:35.930 align:middle line:84% caused 540 deaths and $620 million of property damage. 00:02:35.930 --> 00:02:38.910 align:middle line:84% Now, once again, if a smoker lives in a house by themselves 00:02:38.910 --> 00:02:40.910 align:middle line:84% in the middle of nowhere with no fire department 00:02:40.910 --> 00:02:43.077 align:middle line:84% and it burns their house down, that's their problem. 00:02:43.077 --> 00:02:44.113 align:middle line:90% But that doesn't happen. 00:02:44.113 --> 00:02:45.530 align:middle line:84% The externalities arise because A, 00:02:45.530 --> 00:02:47.197 align:middle line:84% you can burn your neighbor's house down. 00:02:47.197 --> 00:02:48.490 align:middle line:90% That's an externality. 00:02:48.490 --> 00:02:50.850 align:middle line:84% B, the fire department has to come, 00:02:50.850 --> 00:02:52.310 align:middle line:90% raising everybody else's taxes. 00:02:52.310 --> 00:02:53.650 align:middle line:90% That's an externality. 00:02:53.650 --> 00:02:55.730 align:middle line:84% So once again, we want to think about what's 00:02:55.730 --> 00:02:57.900 align:middle line:84% internalized is what people pay for. 00:02:57.900 --> 00:03:00.580 align:middle line:84% What's externalized is what others bear 00:03:00.580 --> 00:03:04.140 align:middle line:90% without them paying for it. 00:03:04.140 --> 00:03:05.900 align:middle line:84% Now, there's an interesting twist 00:03:05.900 --> 00:03:09.220 align:middle line:84% on all of this, which is-- we focused on the negative, 00:03:09.220 --> 00:03:12.380 align:middle line:84% but smoking actually has a big positive externality. 00:03:12.380 --> 00:03:15.140 align:middle line:84% Turns out there's a big positive externality for smoking. 00:03:15.140 --> 00:03:17.460 align:middle line:84% And it's because smokers have a convenient habit 00:03:17.460 --> 00:03:19.780 align:middle line:90% of dying at age 65. 00:03:19.780 --> 00:03:21.000 align:middle line:90% Why is that convenient? 00:03:21.000 --> 00:03:23.220 align:middle line:84% It's convenient because they've paid Social Security 00:03:23.220 --> 00:03:25.720 align:middle line:84% taxes their whole life, and they don't collect the benefits. 00:03:25.720 --> 00:03:27.012 align:middle line:90% They even have a name for this. 00:03:27.012 --> 00:03:30.020 align:middle line:90% They call it the death benefit. 00:03:30.020 --> 00:03:32.640 align:middle line:84% The death benefit is a positive externality, 00:03:32.640 --> 00:03:34.760 align:middle line:84% which is smokers essentially pay into pensions. 00:03:34.760 --> 00:03:36.540 align:middle line:84% A lot of what we do in our life is 00:03:36.540 --> 00:03:38.920 align:middle line:84% paying when we're young to get benefits when we're old. 00:03:38.920 --> 00:03:41.540 align:middle line:84% Well, if die before you're old, you don't get those benefits. 00:03:41.540 --> 00:03:43.140 align:middle line:84% You're paying those to other people. 00:03:43.140 --> 00:03:44.480 align:middle line:90% That a positive externality. 00:03:44.480 --> 00:03:48.100 align:middle line:84% Indeed, it's so big that many have argued that on net, smoking 00:03:48.100 --> 00:03:49.940 align:middle line:90% is good for society. 00:03:49.940 --> 00:03:53.140 align:middle line:84% And indeed, the best estimates are the negative externalities 00:03:53.140 --> 00:03:56.900 align:middle line:84% of smoking are pretty small, maybe around $0.50 a pack. 00:03:56.900 --> 00:03:59.820 align:middle line:84% So about $0.50 a pack is the net damage done to the society, 00:03:59.820 --> 00:04:02.450 align:middle line:84% which is actually, current cigarette tax at the federal 00:04:02.450 --> 00:04:06.540 align:middle line:84% and state level are running between $2 and $5 a pack. 00:04:06.540 --> 00:04:09.780 align:middle line:84% That seems like, wow, that's a lot smaller than the taxes. 00:04:09.780 --> 00:04:11.400 align:middle line:90% But that misses two things. 00:04:11.400 --> 00:04:13.900 align:middle line:84% The first thing, and this is one thing we haven't discussed, 00:04:13.900 --> 00:04:16.459 align:middle line:84% which is secondhand smoke, the fact that when I smoke, 00:04:16.459 --> 00:04:17.899 align:middle line:90% that can affect you. 00:04:17.899 --> 00:04:21.500 align:middle line:84% Now let's think about why that's an externality. 00:04:21.500 --> 00:04:24.360 align:middle line:90% It's an externality because-- 00:04:24.360 --> 00:04:26.860 align:middle line:84% well, let me give an example of it not being an externality. 00:04:26.860 --> 00:04:28.360 align:middle line:84% Imagine instead of secondhand smoke, 00:04:28.360 --> 00:04:31.380 align:middle line:84% I smoked and I just jammed my cigarette in your arm. 00:04:31.380 --> 00:04:32.895 align:middle line:84% That would not be an externalities. 00:04:32.895 --> 00:04:34.520 align:middle line:84% If that's what caused secondhand smoke, 00:04:34.520 --> 00:04:36.687 align:middle line:84% why would that not be externality? 00:04:36.687 --> 00:04:38.020 align:middle line:90% AUDIENCE: Why would you do that? 00:04:38.020 --> 00:04:38.600 align:middle line:90% JON GRUBER: Someone else. 00:04:38.600 --> 00:04:40.100 align:middle line:84% Let's get some other folks involved. 00:04:40.100 --> 00:04:41.820 align:middle line:84% Yeah, why is it not an externality? 00:04:41.820 --> 00:04:44.200 align:middle line:84% AUDIENCE: Because that's a very concrete-- 00:04:44.200 --> 00:04:46.040 align:middle line:90% you are damaging this person. 00:04:46.040 --> 00:04:47.860 align:middle line:90% JON GRUBER: Yeah. 00:04:47.860 --> 00:04:49.400 align:middle line:84% But what does he have over his arm? 00:04:49.400 --> 00:04:50.692 align:middle line:90% What's important about his arm? 00:04:50.692 --> 00:04:51.620 align:middle line:90% AUDIENCE: His arm? 00:04:51.620 --> 00:04:53.620 align:middle line:84% JON GRUBER: He has property rights over his arm. 00:04:53.620 --> 00:04:55.510 align:middle line:90% So Coase should work. 00:04:55.510 --> 00:04:57.630 align:middle line:84% Basically, if I jab him, he can sue me. 00:04:57.630 --> 00:04:58.910 align:middle line:90% He can punch me. 00:04:58.910 --> 00:05:02.150 align:middle line:84% He can make me bear the consequences of that damage. 00:05:02.150 --> 00:05:04.130 align:middle line:84% Why is secondhand smoke not like that? 00:05:04.130 --> 00:05:06.030 align:middle line:90% Because no one owns the air. 00:05:06.030 --> 00:05:08.070 align:middle line:84% So when I smoke and that smoke goes in the air 00:05:08.070 --> 00:05:11.230 align:middle line:84% and the people get damaged, there's no property right. 00:05:11.230 --> 00:05:14.870 align:middle line:84% And that's why there can be no internal solution to that. 00:05:14.870 --> 00:05:17.510 align:middle line:84% So that's one reason why there's secondhand smoke-- damage 00:05:17.510 --> 00:05:18.930 align:middle line:90% with secondhand smoke. 00:05:18.930 --> 00:05:20.590 align:middle line:84% But there's another wrinkle too, which 00:05:20.590 --> 00:05:23.070 align:middle line:84% is most of the damage done from secondhand smoke is 00:05:23.070 --> 00:05:24.830 align:middle line:90% done to family members. 00:05:24.830 --> 00:05:25.330 align:middle line:90% OK? 00:05:25.330 --> 00:05:26.872 align:middle line:84% I've been talking about this example. 00:05:26.872 --> 00:05:28.630 align:middle line:84% My wife, who has never smoked, but her mom 00:05:28.630 --> 00:05:30.297 align:middle line:84% was a three pack a day cigarette smoker. 00:05:30.297 --> 00:05:33.910 align:middle line:84% And she now has COPD, pulmonary disease, 00:05:33.910 --> 00:05:35.910 align:middle line:84% from her mom smoking in the house. 00:05:35.910 --> 00:05:37.870 align:middle line:90% Is that an externality? 00:05:37.870 --> 00:05:41.230 align:middle line:84% When is it and when is it not an externality? 00:05:41.230 --> 00:05:41.790 align:middle line:90% Yeah. 00:05:41.790 --> 00:05:43.207 align:middle line:84% AUDIENCE: It is not an externality 00:05:43.207 --> 00:05:44.412 align:middle line:90% if it's paid for by her mom. 00:05:44.412 --> 00:05:45.370 align:middle line:90% JON GRUBER: Well, yeah. 00:05:45.370 --> 00:05:46.890 align:middle line:84% Her mom could pay for it, but her mom's dead. 00:05:46.890 --> 00:05:48.723 align:middle line:84% She's not paying her health insurance bills. 00:05:48.723 --> 00:05:50.843 align:middle line:84% Why else might it not be an externality? 00:05:50.843 --> 00:05:52.760 align:middle line:84% Why else might it not be an be an externality? 00:05:52.760 --> 00:05:53.320 align:middle line:90% Yeah. 00:05:53.320 --> 00:05:55.240 align:middle line:84% AUDIENCE: Because your decision to smoke 00:05:55.240 --> 00:05:58.447 align:middle line:84% takes into account the cost that you cost your family member. 00:05:58.447 --> 00:05:59.280 align:middle line:90% JON GRUBER: Exactly. 00:05:59.280 --> 00:06:01.680 align:middle line:84% If you have a family utility function 00:06:01.680 --> 00:06:03.560 align:middle line:84% where you're maximizing not your own utility 00:06:03.560 --> 00:06:06.620 align:middle line:84% but your family's utility, then that's not an externality. 00:06:06.620 --> 00:06:10.460 align:middle line:84% So in a family utility function, my wife's mother said, look, 00:06:10.460 --> 00:06:13.280 align:middle line:84% I know when I smoke, I'm creating this externality 00:06:13.280 --> 00:06:14.637 align:middle line:90% on my daughter. 00:06:14.637 --> 00:06:16.220 align:middle line:84% That doesn't mean she shouldn't smoke. 00:06:16.220 --> 00:06:18.040 align:middle line:84% It just means that she should really, really want 00:06:18.040 --> 00:06:19.340 align:middle line:90% to smoke before she does so. 00:06:19.340 --> 00:06:21.000 align:middle line:90% Doesn't mean smoking's wrong. 00:06:21.000 --> 00:06:23.360 align:middle line:84% It just means she-- just like addiction doesn't mean 00:06:23.360 --> 00:06:25.240 align:middle line:90% smoking's necessarily wrong. 00:06:25.240 --> 00:06:27.040 align:middle line:84% It just means that that's something 00:06:27.040 --> 00:06:28.600 align:middle line:90% she should account for. 00:06:28.600 --> 00:06:31.000 align:middle line:84% So if people have family utility functions, 00:06:31.000 --> 00:06:33.480 align:middle line:84% that secondhand smoke is not an externality. 00:06:33.480 --> 00:06:35.700 align:middle line:84% Turns out people don't have family utility functions, 00:06:35.700 --> 00:06:37.040 align:middle line:90% especially men. 00:06:37.040 --> 00:06:38.680 align:middle line:84% And we know this because there are 00:06:38.680 --> 00:06:40.320 align:middle line:84% dozens of studies in the developing 00:06:40.320 --> 00:06:45.480 align:middle line:84% world which test the key prediction of a family utility 00:06:45.480 --> 00:06:46.280 align:middle line:90% model. 00:06:46.280 --> 00:06:48.480 align:middle line:84% What's the key prediction is that $1 00:06:48.480 --> 00:06:51.720 align:middle line:84% in the hands of any family member should be spent the same. 00:06:51.720 --> 00:06:53.580 align:middle line:84% If we all have a family utility function, 00:06:53.580 --> 00:06:55.420 align:middle line:84% then if mom or dad gets the dollar, 00:06:55.420 --> 00:06:57.160 align:middle line:84% they should spend it the same, because they're maximizing 00:06:57.160 --> 00:06:58.540 align:middle line:90% the same objective function. 00:06:58.540 --> 00:07:00.255 align:middle line:90% But in fact, that's not true. 00:07:00.255 --> 00:07:02.380 align:middle line:84% When women get the money, they spend on their kids. 00:07:02.380 --> 00:07:04.760 align:middle line:84% When dads get the money, they spend it on beer. 00:07:04.760 --> 00:07:05.320 align:middle line:90% OK? 00:07:05.320 --> 00:07:07.880 align:middle line:84% So basically there's a strict rejection 00:07:07.880 --> 00:07:09.640 align:middle line:90% of family utility functions. 00:07:09.640 --> 00:07:12.600 align:middle line:84% And as a result, we don't think people are maximizing 00:07:12.600 --> 00:07:14.160 align:middle line:90% family utility functions. 00:07:14.160 --> 00:07:16.320 align:middle line:84% And so secondhand smoke would be an externality 00:07:16.320 --> 00:07:18.760 align:middle line:84% even if it's damaging family members as well. 00:07:18.760 --> 00:07:20.733 align:middle line:90% This is all kind of esoteric. 00:07:20.733 --> 00:07:22.400 align:middle line:84% But it's just to raise the point that we 00:07:22.400 --> 00:07:24.358 align:middle line:84% think critically about when externalities arise 00:07:24.358 --> 00:07:26.360 align:middle line:90% and when they do not. 00:07:26.360 --> 00:07:29.720 align:middle line:84% Now, once again, measuring externality of secondhand smoke 00:07:29.720 --> 00:07:30.720 align:middle line:90% is very hard. 00:07:30.720 --> 00:07:33.080 align:middle line:84% The estimates are in the narrow range of $0.01 per pack 00:07:33.080 --> 00:07:34.760 align:middle line:90% to $2 a pack. 00:07:34.760 --> 00:07:38.320 align:middle line:84% We really have no idea how big these effects are. 00:07:38.320 --> 00:07:41.480 align:middle line:84% But even at the upper end, even if you add that 00:07:41.480 --> 00:07:44.020 align:middle line:84% to the other estimates, it's still not that big. 00:07:44.020 --> 00:07:45.440 align:middle line:90% It's maybe $2 a pack. 00:07:45.440 --> 00:07:48.280 align:middle line:84% That's still not that big compared to the number of taxes. 00:07:48.280 --> 00:07:50.450 align:middle line:90% So are we overtaxing cigarettes? 00:07:50.450 --> 00:07:53.850 align:middle line:84% Well, as I mentioned, I worked in the Clinton administration. 00:07:53.850 --> 00:07:55.310 align:middle line:84% And in the Clinton administration, 00:07:55.310 --> 00:07:56.790 align:middle line:84% I was very-- when I worked in government, 00:07:56.790 --> 00:07:58.373 align:middle line:84% I was very proud of being an economist 00:07:58.373 --> 00:08:01.050 align:middle line:84% because we are so much smarter than lawyers. 00:08:01.050 --> 00:08:04.290 align:middle line:84% And when I was in the room, it was me and a bunch of lawyers. 00:08:04.290 --> 00:08:06.393 align:middle line:84% And I always would make points that-- it's just, 00:08:06.393 --> 00:08:08.810 align:middle line:84% the framework you get in 14.01, the framework you get here 00:08:08.810 --> 00:08:10.890 align:middle line:84% helps us think critically about these problems the way 00:08:10.890 --> 00:08:11.730 align:middle line:90% lawyers don't get. 00:08:11.730 --> 00:08:13.580 align:middle line:84% Except for one time, when you want 00:08:13.580 --> 00:08:15.330 align:middle line:84% to stand up in front of a group of lawyers 00:08:15.330 --> 00:08:17.090 align:middle line:84% and explain the death benefit and say 00:08:17.090 --> 00:08:19.400 align:middle line:84% we should tax cigarettes less because they kill people. 00:08:19.400 --> 00:08:21.150 align:middle line:84% And they were like, that's kind of stupid. 00:08:21.150 --> 00:08:23.010 align:middle line:84% And I was like you know what, that's right. 00:08:23.010 --> 00:08:23.870 align:middle line:90% That is kind of stupid. 00:08:23.870 --> 00:08:25.328 align:middle line:84% And so I came back from Washington, 00:08:25.328 --> 00:08:26.993 align:middle line:84% embarked on a research agenda, which 00:08:26.993 --> 00:08:28.910 align:middle line:84% was trying to understand what's wrong with it. 00:08:28.910 --> 00:08:30.650 align:middle line:84% And I realized what's wrong with that 00:08:30.650 --> 00:08:34.530 align:middle line:84% is that it's a fundamental flaw in the model we use 00:08:34.530 --> 00:08:36.890 align:middle line:84% to model smoking decisions, which is 00:08:36.890 --> 00:08:39.570 align:middle line:90% we ignore two critical things. 00:08:39.570 --> 00:08:42.250 align:middle line:84% The first thing we ignore is that smoking decisions 00:08:42.250 --> 00:08:44.650 align:middle line:90% are made by young people. 00:08:44.650 --> 00:08:48.690 align:middle line:84% Now, even if you're the most rigid Chicago economist 00:08:48.690 --> 00:08:51.710 align:middle line:84% and think that at age 21, we're all perfectly-formed economists, 00:08:51.710 --> 00:08:54.190 align:middle line:84% we almost agree at age zero, we're not. 00:08:54.190 --> 00:08:56.690 align:middle line:84% And therefore, there's got to be some glide path along which 00:08:56.690 --> 00:08:58.927 align:middle line:84% we start making decisions the way we should. 00:08:58.927 --> 00:09:01.510 align:middle line:84% And that means that teenagers are probably not on that glide-- 00:09:01.510 --> 00:09:04.750 align:middle line:84% not fully glid, if that's a word. 00:09:04.750 --> 00:09:06.730 align:middle line:84% That means that teenagers are not-- teenagers 00:09:06.730 --> 00:09:09.570 align:middle line:84% making the vast majority of smoking decisions 00:09:09.570 --> 00:09:12.470 align:middle line:84% are maybe not acting in their own long-run utility, 00:09:12.470 --> 00:09:14.850 align:middle line:90% maximizing self-interest. 00:09:14.850 --> 00:09:16.970 align:middle line:84% My favorite fact of any research project 00:09:16.970 --> 00:09:19.610 align:middle line:84% I ever did was the fact that I discovered doing research 00:09:19.610 --> 00:09:23.050 align:middle line:84% on smoking, which is there's a survey in Michigan that asked 00:09:23.050 --> 00:09:26.570 align:middle line:84% high school students who smoke a pack a day, 00:09:26.570 --> 00:09:29.370 align:middle line:84% do you think you'll be smoking in five years? 00:09:29.370 --> 00:09:31.530 align:middle line:84% And they followed about five years later. 00:09:31.530 --> 00:09:33.410 align:middle line:84% What they found was among those who 00:09:33.410 --> 00:09:37.170 align:middle line:84% said, yes, I'll be smoking in five years, 72% were smoking. 00:09:37.170 --> 00:09:40.550 align:middle line:84% Among those who said, no, I will not be smoking in five years, 00:09:40.550 --> 00:09:42.370 align:middle line:90% 74% were smoking. 00:09:42.370 --> 00:09:47.300 align:middle line:84% And this is the key thing about irrationality around smoking. 00:09:47.300 --> 00:09:50.240 align:middle line:84% It's not that people don't know smoking is bad for you. 00:09:50.240 --> 00:09:52.300 align:middle line:84% Indeed, if you ask smokers how bad smoking 00:09:52.300 --> 00:09:54.500 align:middle line:84% is for you, on average, smokers think it's actually 00:09:54.500 --> 00:09:56.470 align:middle line:90% worse for you than it is. 00:09:56.470 --> 00:09:58.220 align:middle line:84% It's that smokers think they'll all escape 00:09:58.220 --> 00:09:59.667 align:middle line:90% the consequences by quitting. 00:09:59.667 --> 00:10:01.500 align:middle line:84% They don't understate the damage to smoking, 00:10:01.500 --> 00:10:03.860 align:middle line:84% they understate the addictiveness. 00:10:03.860 --> 00:10:06.160 align:middle line:84% OK, so if you ask smokers-- here's a great fact. 00:10:06.160 --> 00:10:09.500 align:middle line:84% If you ask smokers how long the typical smoker will live, 00:10:09.500 --> 00:10:10.860 align:middle line:90% they get the answer right. 00:10:10.860 --> 00:10:14.240 align:middle line:84% You ask them how long they'll live, they say seven more years, 00:10:14.240 --> 00:10:16.740 align:middle line:84% because they think they're going to escape the consequences. 00:10:16.740 --> 00:10:19.360 align:middle line:84% OK, so that's the first problem with the standard model. 00:10:19.360 --> 00:10:21.420 align:middle line:90% It's that kind of irrationality. 00:10:21.420 --> 00:10:23.180 align:middle line:84% The second problem with that model 00:10:23.180 --> 00:10:26.220 align:middle line:84% is that even among adult smokers, 00:10:26.220 --> 00:10:31.660 align:middle line:84% they face what we call a self-control problem. 00:10:31.660 --> 00:10:32.880 align:middle line:90% A self-control problem. 00:10:32.880 --> 00:10:34.145 align:middle line:90% And this is really the core. 00:10:34.145 --> 00:10:36.020 align:middle line:84% I cover this a bit in 14.01, if you took that 00:10:36.020 --> 00:10:39.893 align:middle line:84% with me, the core of the field of behavioral economics. 00:10:39.893 --> 00:10:41.560 align:middle line:84% If you remember, just a bit of a primer. 00:10:41.560 --> 00:10:43.100 align:middle line:84% Remember, in behavioral economics, 00:10:43.100 --> 00:10:45.990 align:middle line:84% the key trick is, how do we minimally 00:10:45.990 --> 00:10:50.150 align:middle line:84% adapt our models to capture some of the realities of psychology 00:10:50.150 --> 00:10:52.330 align:middle line:84% that are ignored by standard economic models? 00:10:52.330 --> 00:10:54.955 align:middle line:84% And perhaps the most significant article in this literature was 00:10:54.955 --> 00:10:58.070 align:middle line:84% an article written by David Laibson in the 1990s at Harvard, 00:10:58.070 --> 00:11:01.710 align:middle line:84% where he talked about a model of what we call hyperbolic 00:11:01.710 --> 00:11:05.190 align:middle line:84% discounting, which is a model where basically, 00:11:05.190 --> 00:11:09.750 align:middle line:84% people don't think rationally about the distant future 00:11:09.750 --> 00:11:11.188 align:middle line:90% relative to the near future. 00:11:11.188 --> 00:11:13.230 align:middle line:84% That essentially, when we write down our standard 00:11:13.230 --> 00:11:15.350 align:middle line:84% over time utility model, we typically 00:11:15.350 --> 00:11:18.550 align:middle line:84% write as utility at a point in time 00:11:18.550 --> 00:11:22.310 align:middle line:84% equals the sum of the discounted stream of future consumption. 00:11:22.310 --> 00:11:26.670 align:middle line:84% So it's basically consumption today 00:11:26.670 --> 00:11:31.110 align:middle line:84% times a discount rate with delta less than 1 sum from t 00:11:31.110 --> 00:11:33.890 align:middle line:90% equals 0 to cap t. 00:11:33.890 --> 00:11:37.052 align:middle line:84% The sum from t equals 0 to cap t. 00:11:37.052 --> 00:11:38.010 align:middle line:90% That's how we write it. 00:11:38.010 --> 00:11:40.550 align:middle line:84% That's called an exponential discounting function. 00:11:40.550 --> 00:11:44.150 align:middle line:84% You assume that-- so if delta is, say, 0.9, then next year 00:11:44.150 --> 00:11:46.030 align:middle line:90% is worth 0.9 of today. 00:11:46.030 --> 00:11:48.350 align:middle line:84% Two years is worth 0.81 of today. 00:11:48.350 --> 00:11:49.750 align:middle line:90% Three years, et cetera. 00:11:49.750 --> 00:11:53.550 align:middle line:84% We assume every period we discount by 90%. 00:11:53.550 --> 00:11:57.230 align:middle line:84% However, that's not the reality of how people behave. 00:11:57.230 --> 00:11:58.930 align:middle line:84% So they ran the following experiment. 00:11:58.930 --> 00:12:07.870 align:middle line:84% They asked people, imagine I can give you $100 today at t 00:12:07.870 --> 00:12:14.070 align:middle line:84% equals 0, or $200 in one year at t equals 1. 00:12:14.070 --> 00:12:15.210 align:middle line:90% Would you take that deal? 00:12:15.210 --> 00:12:19.270 align:middle line:84% And most people said, I'll take the $100 today. 00:12:19.270 --> 00:12:23.710 align:middle line:84% Then they said, what if I offered you $100 at t 00:12:23.710 --> 00:12:29.430 align:middle line:90% equals 5 or $200 at t equals 6? 00:12:29.430 --> 00:12:32.030 align:middle line:84% Now, you should see with this mathematical function, that's 00:12:32.030 --> 00:12:33.208 align:middle line:90% an identical choice. 00:12:33.208 --> 00:12:34.750 align:middle line:84% Because all that matters is the delta 00:12:34.750 --> 00:12:36.370 align:middle line:90% between two consecutive periods. 00:12:36.370 --> 00:12:38.190 align:middle line:90% And that delta is always 0.9. 00:12:38.190 --> 00:12:42.680 align:middle line:84% And yet in this case, everybody chose 200. 00:12:42.680 --> 00:12:43.680 align:middle line:90% Why? 00:12:43.680 --> 00:12:47.260 align:middle line:84% Well, because we don't discount exponentially. 00:12:47.260 --> 00:12:49.160 align:middle line:84% We discount hyperbolically, which 00:12:49.160 --> 00:12:52.080 align:middle line:84% is essentially we have an ever-decreasing discount 00:12:52.080 --> 00:12:54.240 align:middle line:84% rate the further we get in the future. 00:12:54.240 --> 00:12:57.360 align:middle line:84% We're more and more patient the further you get in the future. 00:12:57.360 --> 00:12:59.860 align:middle line:84% Now, that's incredibly hard mathematically to model. 00:12:59.860 --> 00:13:01.840 align:middle line:84% So Laibson's insight was to write down 00:13:01.840 --> 00:13:03.600 align:middle line:84% what we call the quasi-hyperbolic utility 00:13:03.600 --> 00:13:08.360 align:middle line:84% function, which is that u equals today's consumption 00:13:08.360 --> 00:13:15.080 align:middle line:84% Ct plus some factor beta times the discounted sum 00:13:15.080 --> 00:13:21.960 align:middle line:84% consumption from t plus 1 to the future from t equals 1 to t. 00:13:21.960 --> 00:13:23.320 align:middle line:90% Sorry, just t. 00:13:23.320 --> 00:13:26.360 align:middle line:84% C to t delta t, but starting at t equals 1. 00:13:26.360 --> 00:13:28.940 align:middle line:84% So in other words, you have today's decision. 00:13:28.940 --> 00:13:31.480 align:middle line:84% And then the entire future is discounted by the extra factor 00:13:31.480 --> 00:13:32.600 align:middle line:90% beta. 00:13:32.600 --> 00:13:34.640 align:middle line:84% So once you get to tomorrow, we're 00:13:34.640 --> 00:13:36.320 align:middle line:90% back to the standard model. 00:13:36.320 --> 00:13:39.080 align:middle line:84% But today versus tomorrow-- today versus that whole future-- 00:13:39.080 --> 00:13:40.280 align:middle line:90% is an extra factor beta. 00:13:40.280 --> 00:13:42.360 align:middle line:90% And beta is less than 1. 00:13:42.360 --> 00:13:44.040 align:middle line:84% And that explains the kind of trade-offs 00:13:44.040 --> 00:13:48.080 align:middle line:84% we just described about the $100, $200 example. 00:13:48.080 --> 00:13:52.000 align:middle line:84% And what's cool about this model is this simple change, 00:13:52.000 --> 00:13:58.600 align:middle line:84% literally pulling the C out and adding this extra beta factor 00:13:58.600 --> 00:14:02.585 align:middle line:84% can explain a ton of the decisions we see in reality. 00:14:02.585 --> 00:14:03.960 align:middle line:84% In particular, it can handle what 00:14:03.960 --> 00:14:06.040 align:middle line:84% we call the self-control problem, which 00:14:06.040 --> 00:14:08.640 align:middle line:84% is the notion that basically people are 00:14:08.640 --> 00:14:10.840 align:middle line:84% unable to take actions which have 00:14:10.840 --> 00:14:13.580 align:middle line:84% short-term costs and long-run benefits. 00:14:13.580 --> 00:14:15.280 align:middle line:90% And that's true for all of us. 00:14:15.280 --> 00:14:17.740 align:middle line:84% We all face things with self-control problems. 00:14:17.740 --> 00:14:19.600 align:middle line:90% For me, it's biting my nails. 00:14:19.600 --> 00:14:20.120 align:middle line:90% OK? 00:14:20.120 --> 00:14:22.000 align:middle line:84% I know I should stop biting my nails. 00:14:22.000 --> 00:14:25.200 align:middle line:84% I know that that is a long-term detrimental effect. 00:14:25.200 --> 00:14:27.000 align:middle line:90% OK, it looks ugly, whatever. 00:14:27.000 --> 00:14:27.560 align:middle line:90% OK? 00:14:27.560 --> 00:14:29.360 align:middle line:84% I know I should stop, but if there's ever 00:14:29.360 --> 00:14:33.080 align:middle line:84% a close sports game or a tight deadline, I bite my nails. 00:14:33.080 --> 00:14:34.460 align:middle line:90% I've tried to stop. 00:14:34.460 --> 00:14:36.968 align:middle line:84% I've tried to get that stuff you paint on your nails. 00:14:36.968 --> 00:14:39.260 align:middle line:84% I can't swear, because my niece and nephew in the back. 00:14:39.260 --> 00:14:42.010 align:middle line:84% So I'm going to try to stop swearing here. 00:14:42.010 --> 00:14:45.622 align:middle line:84% That stuff you paint on your nails 00:14:45.622 --> 00:14:46.830 align:middle line:90% so you don't bite your nails. 00:14:46.830 --> 00:14:49.305 align:middle line:84% Just makes it taste bad when I bite my nails. 00:14:49.305 --> 00:14:51.430 align:middle line:84% I have a self-control problem when I bite my nails. 00:14:51.430 --> 00:14:52.805 align:middle line:84% We all have self-control problems 00:14:52.805 --> 00:14:56.910 align:middle line:84% around issues with a short-run cost and a long-run benefit. 00:14:56.910 --> 00:14:57.410 align:middle line:90% OK? 00:14:57.410 --> 00:14:59.970 align:middle line:90% Dieting, exercise. 00:14:59.970 --> 00:15:01.710 align:middle line:84% I always say before I get on a plane, 00:15:01.710 --> 00:15:02.770 align:middle line:84% I'm not going to eat the crappy food 00:15:02.770 --> 00:15:03.750 align:middle line:90% they're serving on the plane. 00:15:03.750 --> 00:15:05.010 align:middle line:90% But they put it in front of you. 00:15:05.010 --> 00:15:05.810 align:middle line:90% What are you going to do? 00:15:05.810 --> 00:15:07.350 align:middle line:84% You're going to eat the crappy food they serve on the plane. 00:15:07.350 --> 00:15:10.290 align:middle line:84% We all have these kind of self-control problems. 00:15:10.290 --> 00:15:13.210 align:middle line:84% Once you have a self-control problem, 00:15:13.210 --> 00:15:16.290 align:middle line:84% the economic model changes radically. 00:15:16.290 --> 00:15:19.490 align:middle line:84% And it changes radically because we allow for the fact 00:15:19.490 --> 00:15:22.570 align:middle line:84% that smokers might not be making decisions 00:15:22.570 --> 00:15:26.330 align:middle line:84% in their own long-run best interests. 00:15:26.330 --> 00:15:29.370 align:middle line:84% Now, Coase would say, well, that's ridiculous. 00:15:29.370 --> 00:15:31.730 align:middle line:84% If smokers face a self-control problem, 00:15:31.730 --> 00:15:34.647 align:middle line:84% they should just go ahead and implement processes to fight 00:15:34.647 --> 00:15:35.730 align:middle line:90% that self-control problem. 00:15:35.730 --> 00:15:37.210 align:middle line:90% And smokers do. 00:15:37.210 --> 00:15:40.380 align:middle line:84% If you look at-- in fact, the best evidence 00:15:40.380 --> 00:15:42.500 align:middle line:84% that folks suffer self-control problems 00:15:42.500 --> 00:15:44.860 align:middle line:84% is to look at all the recommended ways 00:15:44.860 --> 00:15:46.620 align:middle line:90% to quit smoking. 00:15:46.620 --> 00:15:50.340 align:middle line:84% They're all "make your life miserable if you smoke". 00:15:50.340 --> 00:15:53.580 align:middle line:84% They're "put your old cigarette butts in a jar of water", 00:15:53.580 --> 00:15:55.973 align:middle line:84% "give someone $1 every time you smoke". 00:15:55.973 --> 00:15:58.640 align:middle line:84% Now, those are things where in a rational model, you'd never do. 00:15:58.640 --> 00:16:01.340 align:middle line:84% Why would you ever make yourself worse off? 00:16:01.340 --> 00:16:03.120 align:middle line:90% But in this model, you would. 00:16:03.120 --> 00:16:05.418 align:middle line:84% Because remember, the payoff is long-term benefits-- 00:16:05.418 --> 00:16:07.460 align:middle line:84% you live longer-- against short-term costs-- it's 00:16:07.460 --> 00:16:08.820 align:middle line:90% really painful to quit. 00:16:08.820 --> 00:16:10.960 align:middle line:84% So I bring the cost to the short term. 00:16:10.960 --> 00:16:12.500 align:middle line:90% You balance it out. 00:16:12.500 --> 00:16:15.260 align:middle line:84% When my kids were in elementary school, 00:16:15.260 --> 00:16:19.160 align:middle line:84% they got an anti-smoking bookmark. 00:16:19.160 --> 00:16:20.940 align:middle line:84% Did the bookmark say you'll die sooner? 00:16:20.940 --> 00:16:21.660 align:middle line:90% No. 00:16:21.660 --> 00:16:22.437 align:middle line:90% What did it say? 00:16:22.437 --> 00:16:23.520 align:middle line:90% You'll be worse at sports. 00:16:23.520 --> 00:16:24.800 align:middle line:90% Your teeth will look gross. 00:16:24.800 --> 00:16:26.382 align:middle line:90% You'll get acne, whatever. 00:16:26.382 --> 00:16:28.340 align:middle line:84% All stuff which is quite frankly pretty trivial 00:16:28.340 --> 00:16:30.160 align:middle line:90% relative to dying younger. 00:16:30.160 --> 00:16:32.060 align:middle line:84% But they knew the kids wouldn't care. 00:16:32.060 --> 00:16:34.460 align:middle line:84% So they focused on bringing the cost to short run. 00:16:34.460 --> 00:16:37.980 align:middle line:84% And the fact that A, we know we have 00:16:37.980 --> 00:16:40.900 align:middle line:84% to do that means that these problems are real, and B, 00:16:40.900 --> 00:16:43.580 align:middle line:84% it dramatically changes things by introducing 00:16:43.580 --> 00:16:46.300 align:middle line:84% the concept of what I labeled internalities. 00:16:46.300 --> 00:16:48.740 align:middle line:84% In my work with Botond Koszegi, we 00:16:48.740 --> 00:16:51.020 align:middle line:84% set up a new literature discussing all these things 00:16:51.020 --> 00:16:53.580 align:middle line:84% called internalities, which are the damage you 00:16:53.580 --> 00:16:59.260 align:middle line:84% do to yourself through behavior which is not 00:16:59.260 --> 00:17:01.540 align:middle line:90% captured by the standard model. 00:17:01.540 --> 00:17:03.560 align:middle line:84% So we ignored the whole you smoking thing. 00:17:03.560 --> 00:17:04.400 align:middle line:90% Put that aside. 00:17:04.400 --> 00:17:06.380 align:middle line:84% We just didn't focus on this aspect, the fact 00:17:06.380 --> 00:17:09.500 align:middle line:84% that people have self-control problems. 00:17:09.500 --> 00:17:14.339 align:middle line:84% Now, we said that if you take this standard model-- forget 00:17:14.339 --> 00:17:15.619 align:middle line:90% the secondhand smoke. 00:17:15.619 --> 00:17:18.720 align:middle line:84% The externalities due to smoking are $0.50 a pack. 00:17:18.720 --> 00:17:22.442 align:middle line:84% Therefore, the optimal tax tau is $0.50 a pack. 00:17:22.442 --> 00:17:23.859 align:middle line:84% If you use this model and you take 00:17:23.859 --> 00:17:26.300 align:middle line:84% a beta of 0.6, which would typically 00:17:26.300 --> 00:17:31.140 align:middle line:84% be estimated in the laboratory, the optimal tax is $10 a pack. 00:17:31.140 --> 00:17:32.660 align:middle line:90% Why does it change so much? 00:17:32.660 --> 00:17:35.460 align:middle line:84% Because smoking is really, really bad for you. 00:17:35.460 --> 00:17:38.670 align:middle line:84% Every cigarette shortens your life by 7 minutes. 00:17:38.670 --> 00:17:41.910 align:middle line:84% If you take that estimate and you multiply it by typical 00:17:41.910 --> 00:17:44.550 align:middle line:84% values of a life-- once again, we'll cover that in chapter 8-- 00:17:44.550 --> 00:17:46.190 align:middle line:84% you get that a pack of cigarettes 00:17:46.190 --> 00:17:49.350 align:middle line:90% damage your health by about $40. 00:17:49.350 --> 00:17:52.430 align:middle line:84% So if you're even making a small mistake, 00:17:52.430 --> 00:17:56.270 align:middle line:84% that's a huge number relative to the $0.50. 00:17:56.270 --> 00:18:02.470 align:middle line:84% So even if beta is 0.9, we estimate a nominal tax of $5, 00:18:02.470 --> 00:18:03.010 align:middle line:90% OK? 00:18:03.010 --> 00:18:06.670 align:middle line:84% Because the fact that damage of smoking-- the internal damage 00:18:06.670 --> 00:18:09.910 align:middle line:84% is so large that if any internalities matter, 00:18:09.910 --> 00:18:11.410 align:middle line:84% cigarette tax should be much higher. 00:18:11.410 --> 00:18:14.030 align:middle line:84% Another way to put this is if you're going to argue cigarette 00:18:14.030 --> 00:18:15.830 align:middle line:84% tax should only reflect externalities, 00:18:15.830 --> 00:18:18.310 align:middle line:84% you're going to have to argue that smokers truly 00:18:18.310 --> 00:18:21.070 align:middle line:84% are making no mistakes of any kind, either irrational youth 00:18:21.070 --> 00:18:23.170 align:middle line:84% smoking decisions or self-control problems. 00:18:23.170 --> 00:18:27.430 align:middle line:84% Once you permit that a little bit, you get a huge optimal tax. 00:18:27.430 --> 00:18:29.630 align:middle line:84% And this is the insight of our work, 00:18:29.630 --> 00:18:35.430 align:middle line:84% that basically these externalities are real. 00:18:35.430 --> 00:18:37.230 align:middle line:84% And once again, the other point is 00:18:37.230 --> 00:18:39.350 align:middle line:84% there's no Coasian solution because there's no way 00:18:39.350 --> 00:18:41.270 align:middle line:90% to perfectly stop yourself. 00:18:41.270 --> 00:18:43.390 align:middle line:84% You can say, well, look, I'll put cigarette butts 00:18:43.390 --> 00:18:43.932 align:middle line:90% in the water. 00:18:43.932 --> 00:18:44.670 align:middle line:90% It doesn't help. 00:18:44.670 --> 00:18:45.950 align:middle line:84% You could say, for example, what if you 00:18:45.950 --> 00:18:47.350 align:middle line:84% had a bracelet that shocked you every time 00:18:47.350 --> 00:18:48.250 align:middle line:90% you went to buy cigarettes? 00:18:48.250 --> 00:18:49.792 align:middle line:84% Well, you could cut the bracelet off. 00:18:49.792 --> 00:18:52.110 align:middle line:84% You put it on yourself, you could take it off. 00:18:52.110 --> 00:18:54.990 align:middle line:84% But what is a good device you can't avoid, at least 00:18:54.990 --> 00:18:56.710 align:middle line:90% without illegal behavior? 00:18:56.710 --> 00:18:58.670 align:middle line:90% Cigarette taxes. 00:18:58.670 --> 00:19:00.270 align:middle line:84% So cigarette taxes are essentially 00:19:00.270 --> 00:19:03.550 align:middle line:84% government-imposed self control device, 00:19:03.550 --> 00:19:06.270 align:middle line:84% because smokers are price sensitive. 00:19:06.270 --> 00:19:08.430 align:middle line:84% The typical elasticity of demand for cigarettes 00:19:08.430 --> 00:19:10.670 align:middle line:90% is about minus 0.4. 00:19:10.670 --> 00:19:14.190 align:middle line:84% And for kids, it's close to minus 1. 00:19:14.190 --> 00:19:16.670 align:middle line:84% So basically the point is, you can't 00:19:16.670 --> 00:19:19.570 align:middle line:84% impose your own self-control because it doesn't work. 00:19:19.570 --> 00:19:21.485 align:middle line:84% But the government can impose self-control 00:19:21.485 --> 00:19:23.110 align:middle line:84% until the point where smuggling becomes 00:19:23.110 --> 00:19:24.880 align:middle line:84% a big enough deal and black market cigarettes 00:19:24.880 --> 00:19:27.005 align:middle line:84% are a big enough deal that you can't get around it. 00:19:27.005 --> 00:19:30.430 align:middle line:84% But until you get to that point, higher cigarette taxes 00:19:30.430 --> 00:19:34.920 align:middle line:84% serve as exactly the kind of public self control tool 00:19:34.920 --> 00:19:37.560 align:middle line:84% that you need to actually fight this problem. 00:19:37.560 --> 00:19:39.640 align:middle line:90% Questions about that? 00:19:39.640 --> 00:19:41.760 align:middle line:90% Yeah, you look uncertain. 00:19:41.760 --> 00:19:43.980 align:middle line:84% AUDIENCE: I have a question about cigarette taxes. 00:19:43.980 --> 00:19:45.560 align:middle line:84% Don't they tend to be regressive? 00:19:45.560 --> 00:19:46.685 align:middle line:90% JON GRUBER: Great question. 00:19:46.685 --> 00:19:51.080 align:middle line:84% So remember, we're only discussing efficiency this far 00:19:51.080 --> 00:19:51.980 align:middle line:90% into the course. 00:19:51.980 --> 00:19:53.740 align:middle line:90% But you guys are all good folks. 00:19:53.740 --> 00:19:55.400 align:middle line:90% You care about equity too. 00:19:55.400 --> 00:19:59.023 align:middle line:84% And the concern is that cigarette taxes are-- 00:19:59.023 --> 00:20:01.440 align:middle line:84% the term regressive, we'll talk about this more in the tax 00:20:01.440 --> 00:20:02.025 align:middle line:90% chapter. 00:20:02.025 --> 00:20:03.400 align:middle line:84% It's the notion that people spend 00:20:03.400 --> 00:20:04.880 align:middle line:84% a higher percentage of their income 00:20:04.880 --> 00:20:07.560 align:middle line:84% on something the poorer they are. 00:20:07.560 --> 00:20:09.985 align:middle line:84% It's not that people spend more when they're poor, 00:20:09.985 --> 00:20:12.360 align:middle line:84% it's that they spend a higher percentage of their income. 00:20:12.360 --> 00:20:14.383 align:middle line:84% And for cigarettes, that's absolutely true. 00:20:14.383 --> 00:20:16.800 align:middle line:84% Poor people spend a much higher percentage of their income 00:20:16.800 --> 00:20:18.095 align:middle line:90% on cigarettes than rich people. 00:20:18.095 --> 00:20:19.220 align:middle line:90% And that's for two reasons. 00:20:19.220 --> 00:20:21.080 align:middle line:84% The first reason I want to discuss last time, 00:20:21.080 --> 00:20:22.955 align:middle line:84% that rich people smoking has all disappeared. 00:20:22.955 --> 00:20:25.120 align:middle line:84% It's only poor people who smoke now. 00:20:25.120 --> 00:20:30.260 align:middle line:84% The second reason is because as a fraction of income, 00:20:30.260 --> 00:20:31.220 align:middle line:90% it's much larger. 00:20:31.220 --> 00:20:34.530 align:middle line:84% A fixed cigarette tax, a $5 pack cigarette tax 00:20:34.530 --> 00:20:35.790 align:middle line:90% means nothing to me. 00:20:35.790 --> 00:20:38.130 align:middle line:84% It means something to a poor person. 00:20:38.130 --> 00:20:40.650 align:middle line:84% So for that reason, we will say we shouldn't tax cigarettes 00:20:40.650 --> 00:20:41.790 align:middle line:90% because they're regressive. 00:20:41.790 --> 00:20:42.250 align:middle line:90% Yeah. 00:20:42.250 --> 00:20:43.990 align:middle line:84% AUDIENCE: What is the current cigarette tax? 00:20:43.990 --> 00:20:46.240 align:middle line:84% JON GRUBER: The current cigarette tax varies by state. 00:20:46.240 --> 00:20:49.650 align:middle line:84% But like I said, it's between about $2 and $5 00:20:49.650 --> 00:20:52.770 align:middle line:90% depending on the state. 00:20:52.770 --> 00:20:55.790 align:middle line:84% However, my co-author and I, Botond Koszegi, 00:20:55.790 --> 00:20:57.610 align:middle line:90% we argue that's wrong. 00:20:57.610 --> 00:20:59.915 align:middle line:84% And it's wrong once you enter this model. 00:20:59.915 --> 00:21:01.290 align:middle line:84% Because once you're in this model 00:21:01.290 --> 00:21:03.170 align:middle line:84% and recognize that the poor people who smoke 00:21:03.170 --> 00:21:06.890 align:middle line:84% might be making mistakes, then the fact-- 00:21:06.890 --> 00:21:08.730 align:middle line:84% and you also recognize poor people 00:21:08.730 --> 00:21:11.730 align:middle line:84% are much more price-elastic than rich people. 00:21:11.730 --> 00:21:13.890 align:middle line:84% For rich people, the price elasticity is very low. 00:21:13.890 --> 00:21:16.330 align:middle line:84% For poor people, it's like minus 1. 00:21:16.330 --> 00:21:18.370 align:middle line:84% So in fact, when you raise cigarette taxes, 00:21:18.370 --> 00:21:22.190 align:middle line:84% on net poor people spend less, because so many of them quit, 00:21:22.190 --> 00:21:25.130 align:middle line:84% that overall poor person's spending on cigarettes falls. 00:21:25.130 --> 00:21:27.170 align:middle line:84% Now, in a standard model, that wouldn't matter. 00:21:27.170 --> 00:21:28.545 align:middle line:84% But in our model, it does matter. 00:21:28.545 --> 00:21:31.115 align:middle line:84% If you actually think people are not smoking optimally, 00:21:31.115 --> 00:21:32.490 align:middle line:84% the fact that cigarettes actually 00:21:32.490 --> 00:21:34.650 align:middle line:84% cause poor people to quit is a benefit, 00:21:34.650 --> 00:21:36.830 align:middle line:84% and it's much larger for poor people to rich people. 00:21:36.830 --> 00:21:38.247 align:middle line:84% So we actually argue cigarette tax 00:21:38.247 --> 00:21:41.010 align:middle line:90% is progressive, not regressive. 00:21:41.010 --> 00:21:41.810 align:middle line:90% Yeah. 00:21:41.810 --> 00:21:43.310 align:middle line:84% AUDIENCE: So I'm thinking like there 00:21:43.310 --> 00:21:46.090 align:middle line:84% are ways that people could get nicotine 00:21:46.090 --> 00:21:47.770 align:middle line:90% that are not through smoking. 00:21:47.770 --> 00:21:50.130 align:middle line:84% And some of those are more discreet, 00:21:50.130 --> 00:21:52.570 align:middle line:84% so they don't have the fire issue 00:21:52.570 --> 00:21:54.428 align:middle line:84% and they don't have the smoke break issue 00:21:54.428 --> 00:21:56.470 align:middle line:84% and they don't have a lot of those externalities. 00:21:56.470 --> 00:21:59.770 align:middle line:84% So for taxes on goods like those-- 00:21:59.770 --> 00:22:02.030 align:middle line:84% JON GRUBER: Yeah, this is a fascinating issue. 00:22:02.030 --> 00:22:03.650 align:middle line:84% Once again, it's amazing teaching 00:22:03.650 --> 00:22:06.090 align:middle line:84% this course for the 137 years I've taught it, 00:22:06.090 --> 00:22:09.290 align:middle line:90% which is that how things change. 00:22:09.290 --> 00:22:11.070 align:middle line:84% This is not a topic I discussed before. 00:22:11.070 --> 00:22:12.795 align:middle line:84% So in fact, let's go back to packs. 00:22:12.795 --> 00:22:14.420 align:middle line:84% If had said to me 15 years ago, there's 00:22:14.420 --> 00:22:16.130 align:middle line:84% going to be a new thing called vaping, 00:22:16.130 --> 00:22:18.690 align:middle line:84% it's going to live in the nicotine without the tar. 00:22:18.690 --> 00:22:22.170 align:middle line:84% Now, we know tar is fundamentally health damaging. 00:22:22.170 --> 00:22:23.670 align:middle line:84% We're not quite sure about nicotine. 00:22:23.670 --> 00:22:24.890 align:middle line:84% The evidence on the health effect of nicotine 00:22:24.890 --> 00:22:25.770 align:middle line:90% isn't as clear. 00:22:25.770 --> 00:22:27.650 align:middle line:90% I would have said awesome. 00:22:27.650 --> 00:22:29.580 align:middle line:84% Basically, if you could take a smoker 00:22:29.580 --> 00:22:32.540 align:middle line:84% and convert them from a cigarette to a vape, 00:22:32.540 --> 00:22:33.820 align:middle line:90% that's a good thing. 00:22:33.820 --> 00:22:35.820 align:middle line:84% What I would have missed is the fact 00:22:35.820 --> 00:22:37.700 align:middle line:84% that for every smoker switching, there's 00:22:37.700 --> 00:22:40.140 align:middle line:90% 10 kids who are starting vaping. 00:22:40.140 --> 00:22:42.560 align:middle line:84% So this is a fascinating question no one's asked, 00:22:42.560 --> 00:22:45.140 align:middle line:84% which is vaping on net good or bad? 00:22:45.140 --> 00:22:47.180 align:middle line:84% And the answer is unclear because it 00:22:47.180 --> 00:22:49.680 align:middle line:90% depends on two things. 00:22:49.680 --> 00:22:51.140 align:middle line:84% It's all about costs and benefits. 00:22:51.140 --> 00:22:53.140 align:middle line:84% The benefit of introducing vaping 00:22:53.140 --> 00:22:55.820 align:middle line:84% is it lowers the damage due to smoking. 00:22:55.820 --> 00:22:58.220 align:middle line:84% The cost is it introduces damage due to nicotine 00:22:58.220 --> 00:23:00.180 align:middle line:90% to 10 extra people. 00:23:00.180 --> 00:23:03.280 align:middle line:84% Now, we don't really know how bad nicotine addiction is. 00:23:03.280 --> 00:23:05.860 align:middle line:84% We don't really know how bad vaping is for your lungs. 00:23:05.860 --> 00:23:08.020 align:middle line:84% So we don't really know how to make that trade-off. 00:23:08.020 --> 00:23:11.620 align:middle line:84% We're pretty sure that even with potential damage to lungs 00:23:11.620 --> 00:23:13.938 align:middle line:84% with nicotine, vaping is still better than smoking. 00:23:13.938 --> 00:23:15.980 align:middle line:84% So we're pretty sure that if literally all we did 00:23:15.980 --> 00:23:19.140 align:middle line:84% is replace smokers with vapers, we'd be better off. 00:23:19.140 --> 00:23:22.140 align:middle line:84% What we don't know is to replace every smoker with 10 vapers 00:23:22.140 --> 00:23:23.478 align:middle line:90% if we're better off or not. 00:23:23.478 --> 00:23:25.520 align:middle line:84% And that's exactly the kind of fascinating thing. 00:23:25.520 --> 00:23:26.820 align:middle line:84% That's why this literature continues 00:23:26.820 --> 00:23:28.020 align:middle line:90% to evolve and be interesting. 00:23:28.020 --> 00:23:29.478 align:middle line:84% It's exactly that kind of question. 00:23:29.478 --> 00:23:32.500 align:middle line:84% Same about ZINs and all the other stuff folks are using now, 00:23:32.500 --> 00:23:35.020 align:middle line:90% we just don't know. 00:23:35.020 --> 00:23:35.680 align:middle line:90% Good question. 00:23:35.680 --> 00:23:36.371 align:middle line:90% Yeah. 00:23:36.371 --> 00:23:38.913 align:middle line:84% AUDIENCE: Because you guys are advocating the increased taxes 00:23:38.913 --> 00:23:42.120 align:middle line:84% on cigarettes, what's the barriers to increasing? 00:23:42.120 --> 00:23:42.903 align:middle line:90% Is it political? 00:23:42.903 --> 00:23:44.320 align:middle line:84% JON GRUBER: Great, great question. 00:23:44.320 --> 00:23:46.500 align:middle line:84% I mean, this will be an important question 00:23:46.500 --> 00:23:48.458 align:middle line:84% to go on the lecture to talk about other health 00:23:48.458 --> 00:23:51.420 align:middle line:84% externalities, which is that basically, the barriers are 00:23:51.420 --> 00:23:52.900 align:middle line:90% really political. 00:23:52.900 --> 00:23:57.080 align:middle line:84% In fact, someone made a great point, which is 100 years ago, 00:23:57.080 --> 00:23:59.397 align:middle line:84% tobacco leaf was grown all over the country. 00:23:59.397 --> 00:24:01.480 align:middle line:84% Today it's grown in a very small number of states, 00:24:01.480 --> 00:24:04.022 align:middle line:84% and only once it's grown in a small number of states who were 00:24:04.022 --> 00:24:05.575 align:middle line:90% willing to regulate tobacco. 00:24:05.575 --> 00:24:07.700 align:middle line:84% Basically, there was a large political constituency 00:24:07.700 --> 00:24:10.100 align:middle line:90% to not regulate smoking. 00:24:10.100 --> 00:24:12.967 align:middle line:84% We knew starting in the 1950s it was bad for you. 00:24:12.967 --> 00:24:15.300 align:middle line:84% But we didn't really start raising cigarette taxes until 00:24:15.300 --> 00:24:17.460 align:middle line:90% the 1990s. 00:24:17.460 --> 00:24:21.740 align:middle line:84% And basically, it was just-- it's a fascinating issue of what 00:24:21.740 --> 00:24:23.303 align:middle line:90% finally led to the change. 00:24:23.303 --> 00:24:24.720 align:middle line:84% How did things really turn around? 00:24:24.720 --> 00:24:26.213 align:middle line:84% And this is actually-- there's so 00:24:26.213 --> 00:24:28.630 align:middle line:84% many things we could follow up on but don't have time for. 00:24:28.630 --> 00:24:30.290 align:middle line:84% This actually raises a great question. 00:24:30.290 --> 00:24:32.748 align:middle line:84% We don't have a course at MIT in law and economics anymore, 00:24:32.748 --> 00:24:33.790 align:middle line:90% but we should. 00:24:33.790 --> 00:24:35.910 align:middle line:84% Great question of what's the optimal role 00:24:35.910 --> 00:24:37.470 align:middle line:90% of the legal system. 00:24:37.470 --> 00:24:40.230 align:middle line:84% What's the optimal role of the legal system in terms 00:24:40.230 --> 00:24:42.490 align:middle line:90% of the other economic outcomes? 00:24:42.490 --> 00:24:44.910 align:middle line:84% So for years, we couldn't deal with the externalities 00:24:44.910 --> 00:24:46.270 align:middle line:90% of cigarettes politically. 00:24:46.270 --> 00:24:47.710 align:middle line:90% Then what happened? 00:24:47.710 --> 00:24:50.350 align:middle line:84% This US states came op with a clever strategy. 00:24:50.350 --> 00:24:52.350 align:middle line:84% The US states said, look, smoking 00:24:52.350 --> 00:24:54.390 align:middle line:90% is costing us a lot of money. 00:24:54.390 --> 00:24:56.630 align:middle line:84% We are going to sue the tobacco companies 00:24:56.630 --> 00:24:59.150 align:middle line:84% for an enormous amount of money to make them pay back 00:24:59.150 --> 00:25:00.950 align:middle line:90% what they're costing us. 00:25:00.950 --> 00:25:04.550 align:middle line:84% Tobacco companies suddenly faced a massive potential risk. 00:25:04.550 --> 00:25:06.510 align:middle line:84% And then they negotiated a deal where 00:25:06.510 --> 00:25:08.810 align:middle line:84% they would agree to a voluntary tax, 00:25:08.810 --> 00:25:12.430 align:middle line:84% essentially, in return for ending these lawsuits. 00:25:12.430 --> 00:25:14.778 align:middle line:84% Now, I'm not a fan of a lot of the legal system 00:25:14.778 --> 00:25:16.570 align:middle line:84% and some of the excessive lawsuits we have. 00:25:16.570 --> 00:25:18.230 align:middle line:84% But here's an example where the legal system got 00:25:18.230 --> 00:25:19.710 align:middle line:90% a political process unstuck. 00:25:19.710 --> 00:25:23.230 align:middle line:84% Once they started that, then states started saying, oh wow, 00:25:23.230 --> 00:25:24.713 align:middle line:84% now we can start raising our taxes. 00:25:24.713 --> 00:25:26.380 align:middle line:84% The federal government raised the taxes. 00:25:26.380 --> 00:25:27.860 align:middle line:84% So it started this dynamic effect. 00:25:27.860 --> 00:25:28.360 align:middle line:90% Oops. 00:25:28.360 --> 00:25:34.160 align:middle line:90% 00:25:34.160 --> 00:25:37.440 align:middle line:84% It started this dynamic effect, where basically it 00:25:37.440 --> 00:25:40.680 align:middle line:84% became OK politically to start going after cigarettes. 00:25:40.680 --> 00:25:41.303 align:middle line:90% Yes. 00:25:41.303 --> 00:25:42.720 align:middle line:84% AUDIENCE: By a voluntary tax, that 00:25:42.720 --> 00:25:44.595 align:middle line:84% just meant like their lobbyists took the foot 00:25:44.595 --> 00:25:46.520 align:middle line:84% off the gas a little bit, or that they just 00:25:46.520 --> 00:25:47.980 align:middle line:84% started handing money over to the states? 00:25:47.980 --> 00:25:49.105 align:middle line:90% JON GRUBER: Great question. 00:25:49.105 --> 00:25:52.600 align:middle line:84% So if I said to you as students in this class, 00:25:52.600 --> 00:25:53.955 align:middle line:90% why should we raise the price? 00:25:53.955 --> 00:25:56.580 align:middle line:84% If I said to you, we're going to raise the price of cigarettes, 00:25:56.580 --> 00:25:59.600 align:middle line:84% you should say-- well, according to what I taught you-- great. 00:25:59.600 --> 00:26:02.120 align:middle line:84% And you shouldn't really care how it happens. 00:26:02.120 --> 00:26:04.540 align:middle line:84% But there's two different ways to make it happen. 00:26:04.540 --> 00:26:07.280 align:middle line:84% One way to make it happen is a tax. 00:26:07.280 --> 00:26:09.140 align:middle line:84% That has the advantage of raising revenues. 00:26:09.140 --> 00:26:11.015 align:middle line:84% It also has the advantage that it potentially 00:26:11.015 --> 00:26:12.012 align:middle line:90% punishes the companies. 00:26:12.012 --> 00:26:13.720 align:middle line:84% The other is just to say, well, companies 00:26:13.720 --> 00:26:15.360 align:middle line:84% are going to have a coordinated price 00:26:15.360 --> 00:26:18.240 align:middle line:90% rise among all the companies. 00:26:18.240 --> 00:26:20.220 align:middle line:84% That has the same effect as a tax, 00:26:20.220 --> 00:26:23.280 align:middle line:84% but only if anything increases their profits, 00:26:23.280 --> 00:26:26.040 align:middle line:84% since smoking is inelastic, is relatively inelastic. 00:26:26.040 --> 00:26:28.760 align:middle line:84% So the public health community was not necessarily crazy 00:26:28.760 --> 00:26:31.135 align:middle line:84% about this, because they wanted to not just stop smoking, 00:26:31.135 --> 00:26:32.920 align:middle line:84% but to punish the tobacco companies. 00:26:32.920 --> 00:26:34.473 align:middle line:84% If you're just about public health, 00:26:34.473 --> 00:26:36.140 align:middle line:84% we just care what the prices are higher. 00:26:36.140 --> 00:26:38.280 align:middle line:84% So I don't really care how it happened. 00:26:38.280 --> 00:26:38.800 align:middle line:90% Yeah. 00:26:38.800 --> 00:26:41.200 align:middle line:84% AUDIENCE: Was there any question about antitrust 00:26:41.200 --> 00:26:42.982 align:middle line:90% talk surrounding that? 00:26:42.982 --> 00:26:44.440 align:middle line:84% JON GRUBER: I mean, we're not going 00:26:44.440 --> 00:26:46.898 align:middle line:84% to talk a lot about imperfect markets and oligopolies here. 00:26:46.898 --> 00:26:50.360 align:middle line:84% But obviously, it was effectively a coordinated price 00:26:50.360 --> 00:26:53.840 align:middle line:90% increase, but not in actuality. 00:26:53.840 --> 00:26:54.460 align:middle line:90% OK? 00:26:54.460 --> 00:26:57.160 align:middle line:90% Other questions, comments? 00:26:57.160 --> 00:26:57.800 align:middle line:90% OK. 00:26:57.800 --> 00:27:00.440 align:middle line:90% So that's how we think about-- 00:27:00.440 --> 00:27:02.080 align:middle line:84% that's how we think about smoking. 00:27:02.080 --> 00:27:04.875 align:middle line:84% Now, let's go on with that framework 00:27:04.875 --> 00:27:06.750 align:middle line:84% and talk about the other health externalities 00:27:06.750 --> 00:27:08.880 align:middle line:90% we deal with in our society. 00:27:08.880 --> 00:27:13.080 align:middle line:84% Another big one is alcohol consumption or drinking. 00:27:13.080 --> 00:27:14.200 align:middle line:90% OK? 00:27:14.200 --> 00:27:17.560 align:middle line:84% Alcohol consumption has massive negative external effects 00:27:17.560 --> 00:27:19.720 align:middle line:90% on society. 00:27:19.720 --> 00:27:21.540 align:middle line:84% The numbers are really astonishing. 00:27:21.540 --> 00:27:22.850 align:middle line:90% These are out of date. 00:27:22.850 --> 00:27:25.370 align:middle line:84% But every day, 30 people in the US 00:27:25.370 --> 00:27:27.330 align:middle line:84% die in a motor vehicle crash that involves 00:27:27.330 --> 00:27:29.930 align:middle line:90% an alcohol-impaired driver. 00:27:29.930 --> 00:27:32.610 align:middle line:84% This amounts to one death every 48 minutes. 00:27:32.610 --> 00:27:36.452 align:middle line:84% 10,000 people a year are killed in alcohol crashes. 00:27:36.452 --> 00:27:38.410 align:middle line:84% Alcohol-related car crashes are about one third 00:27:38.410 --> 00:27:41.050 align:middle line:84% of all car crashes in this country. 00:27:41.050 --> 00:27:45.010 align:middle line:84% The estimates are that the external costs of drinking just 00:27:45.010 --> 00:27:48.460 align:middle line:84% from driving alone are about $250 billion a year, 00:27:48.460 --> 00:27:49.710 align:middle line:90% just from drunk driving alone. 00:27:49.710 --> 00:27:51.490 align:middle line:84% And that's a massive underestimate, 00:27:51.490 --> 00:27:53.690 align:middle line:84% because it turns out, alcohol is also 00:27:53.690 --> 00:27:57.730 align:middle line:84% incredibly associated with criminal activities and abuse. 00:27:57.730 --> 00:28:02.170 align:middle line:84% Estimates suggest that much criminal activity is committed 00:28:02.170 --> 00:28:05.873 align:middle line:84% while people are inebriated, and a lot of spousal abuse 00:28:05.873 --> 00:28:07.290 align:middle line:84% is committed on people inebriated. 00:28:07.290 --> 00:28:09.010 align:middle line:84% There's very clever studies showing 00:28:09.010 --> 00:28:11.210 align:middle line:84% that when alcohol is harder to get, crime goes down 00:28:11.210 --> 00:28:13.330 align:middle line:90% and spousal abuse goes down. 00:28:13.330 --> 00:28:14.510 align:middle line:90% These are also external. 00:28:14.510 --> 00:28:16.350 align:middle line:84% I mean, once again, unless it's a family utility function. 00:28:16.350 --> 00:28:17.267 align:middle line:90% I won't get into that. 00:28:17.267 --> 00:28:17.990 align:middle line:90% That's dangerous. 00:28:17.990 --> 00:28:21.350 align:middle line:84% But I think we can assume spousal abuse is bad. 00:28:21.350 --> 00:28:23.890 align:middle line:84% We can certainly see crime as negative externalities. 00:28:23.890 --> 00:28:28.437 align:middle line:84% So as a result, they're basically-- 00:28:28.437 --> 00:28:30.770 align:middle line:84% and in particular, there's this really interesting study 00:28:30.770 --> 00:28:33.402 align:middle line:84% in India, which found that when they prohibited alcohol 00:28:33.402 --> 00:28:35.610 align:middle line:84% in one state, there was a massive decline in violence 00:28:35.610 --> 00:28:37.810 align:middle line:90% against women in that state. 00:28:37.810 --> 00:28:39.730 align:middle line:90% There's another. 00:28:39.730 --> 00:28:45.530 align:middle line:84% And so basically, the externalities are huge. 00:28:45.530 --> 00:28:48.518 align:middle line:84% So unlike smoking, which has relatively modest externalities, 00:28:48.518 --> 00:28:49.810 align:middle line:90% alcohol has huge externalities. 00:28:49.810 --> 00:28:51.850 align:middle line:84% Indeed, if you perform those externalities, 00:28:51.850 --> 00:28:54.370 align:middle line:84% the optimal tax on alcohol should be roughly four times 00:28:54.370 --> 00:28:57.410 align:middle line:90% what it is today. 00:28:57.410 --> 00:28:58.890 align:middle line:90% Guys, alcohol is super cheap. 00:28:58.890 --> 00:29:01.550 align:middle line:90% 00:29:01.550 --> 00:29:03.730 align:middle line:84% It's not sort of nuts that alcohol 00:29:03.730 --> 00:29:05.070 align:middle line:90% is less than bottled water. 00:29:05.070 --> 00:29:07.770 align:middle line:90% 00:29:07.770 --> 00:29:10.330 align:middle line:84% A can of beer is less than a bottle of water. 00:29:10.330 --> 00:29:15.370 align:middle line:84% So basically, we have alcohol should be taxed much higher. 00:29:15.370 --> 00:29:17.170 align:middle line:84% Now, what about the externalities? 00:29:17.170 --> 00:29:20.047 align:middle line:84% And by the way, that's something Chicago economists 00:29:20.047 --> 00:29:20.880 align:middle line:90% couldn't argue with. 00:29:20.880 --> 00:29:22.005 align:middle line:90% They argue the measurement. 00:29:22.005 --> 00:29:23.900 align:middle line:84% But the bottom line is all economists 00:29:23.900 --> 00:29:26.460 align:middle line:84% agree that we should tax externality-producing 00:29:26.460 --> 00:29:27.380 align:middle line:90% activities. 00:29:27.380 --> 00:29:27.880 align:middle line:90% OK? 00:29:27.880 --> 00:29:29.220 align:middle line:84% We all agree that externalities are large, 00:29:29.220 --> 00:29:30.900 align:middle line:90% alcohol taxes should be higher. 00:29:30.900 --> 00:29:33.020 align:middle line:84% But what about the internalities? 00:29:33.020 --> 00:29:35.540 align:middle line:84% Well here, traditionally when I taught this course, 00:29:35.540 --> 00:29:37.660 align:middle line:84% I said well, this is sort of the flip of smoking. 00:29:37.660 --> 00:29:39.300 align:middle line:84% The internalities of alcohol aren't that bad. 00:29:39.300 --> 00:29:41.420 align:middle line:84% But in fact, there's a lot of evidence suggesting 00:29:41.420 --> 00:29:43.180 align:middle line:84% that the internalities are huge, too-- 00:29:43.180 --> 00:29:45.380 align:middle line:84% that there's huge negative underappreciated health 00:29:45.380 --> 00:29:47.182 align:middle line:90% effects on people of drinking. 00:29:47.182 --> 00:29:49.140 align:middle line:84% We've basically debunked the myth at this point 00:29:49.140 --> 00:29:50.820 align:middle line:84% that a little drinking is good for you. 00:29:50.820 --> 00:29:53.400 align:middle line:84% We've now basically found that no drinking is good for you. 00:29:53.400 --> 00:29:54.478 align:middle line:90% It's like cigarettes. 00:29:54.478 --> 00:29:56.520 align:middle line:84% Basically, no amount of drinking is good for you. 00:29:56.520 --> 00:29:59.103 align:middle line:84% Now, it's not like I said, that every cigarette's bad for you. 00:29:59.103 --> 00:30:01.040 align:middle line:84% Probably modest drinking is not bad for you, 00:30:01.040 --> 00:30:02.380 align:middle line:90% but it's not good for you. 00:30:02.380 --> 00:30:05.380 align:middle line:84% And serious drinking causes enormous health consequences 00:30:05.380 --> 00:30:06.380 align:middle line:90% later in life. 00:30:06.380 --> 00:30:08.385 align:middle line:84% And by the same internalities model 00:30:08.385 --> 00:30:09.760 align:middle line:84% I talked about a few minutes ago, 00:30:09.760 --> 00:30:13.420 align:middle line:84% if people are not incorporating those, 00:30:13.420 --> 00:30:15.500 align:middle line:90% that's putting costs on them. 00:30:15.500 --> 00:30:17.100 align:middle line:84% And as a result, those internalities 00:30:17.100 --> 00:30:19.913 align:middle line:84% should be added as well when thinking about an optimal tax. 00:30:19.913 --> 00:30:21.330 align:middle line:84% But here's the trick with alcohol. 00:30:21.330 --> 00:30:22.670 align:middle line:90% Here's why it's harder. 00:30:22.670 --> 00:30:25.590 align:middle line:84% Smoking is the easiest case to start with because it's 00:30:25.590 --> 00:30:26.990 align:middle line:90% basically linear. 00:30:26.990 --> 00:30:28.870 align:middle line:84% Every cigarette you smoke shortens your life 00:30:28.870 --> 00:30:31.510 align:middle line:84% by 7 minutes in a roughly linear fashion. 00:30:31.510 --> 00:30:34.350 align:middle line:84% Alcohol is harder because basically the damage 00:30:34.350 --> 00:30:35.910 align:middle line:90% is much more concentrated. 00:30:35.910 --> 00:30:39.230 align:middle line:84% Most drinking is done by people who don't create damage 00:30:39.230 --> 00:30:41.270 align:middle line:90% to themselves or society. 00:30:41.270 --> 00:30:43.190 align:middle line:84% The vast majority of alcohol consumption 00:30:43.190 --> 00:30:45.310 align:middle line:84% is done by people who won't be driving 00:30:45.310 --> 00:30:48.250 align:middle line:84% and won't commit crimes or abuse their spouses. 00:30:48.250 --> 00:30:50.230 align:middle line:84% They're just having a good time or whatever. 00:30:50.230 --> 00:30:51.510 align:middle line:90% OK? 00:30:51.510 --> 00:30:54.590 align:middle line:84% So that raises an interesting difficulty, which is, well, gee, 00:30:54.590 --> 00:30:57.110 align:middle line:84% if I'm going to tax alcohol-- if I tax cigarettes, 00:30:57.110 --> 00:30:58.027 align:middle line:90% every cigarette's bad. 00:30:58.027 --> 00:30:59.485 align:middle line:84% It's clear I should tax cigarettes. 00:30:59.485 --> 00:31:00.990 align:middle line:84% With alcohol, should I really tax it 00:31:00.990 --> 00:31:03.990 align:middle line:84% when most people and most alcohol consumption 00:31:03.990 --> 00:31:06.230 align:middle line:90% is not actually damaging? 00:31:06.230 --> 00:31:06.730 align:middle line:90% OK? 00:31:06.730 --> 00:31:10.170 align:middle line:84% And the answer here is in a perfect world, 00:31:10.170 --> 00:31:11.710 align:middle line:90% we would use other tools. 00:31:11.710 --> 00:31:16.670 align:middle line:84% For example, a perfect world, we would 00:31:16.670 --> 00:31:18.370 align:middle line:84% try to use tools other than taxes 00:31:18.370 --> 00:31:20.512 align:middle line:90% which could target the high-- 00:31:20.512 --> 00:31:21.970 align:middle line:84% so in a perfect world, for example, 00:31:21.970 --> 00:31:23.512 align:middle line:84% you'd only tax people who bought more 00:31:23.512 --> 00:31:25.770 align:middle line:84% than a certain amount of alcohol per day. 00:31:25.770 --> 00:31:26.270 align:middle line:90% OK? 00:31:26.270 --> 00:31:27.978 align:middle line:84% That's impossible, because you can always 00:31:27.978 --> 00:31:29.270 align:middle line:90% send someone to buy for you. 00:31:29.270 --> 00:31:31.550 align:middle line:84% In a perfect world, we would actually target the tax 00:31:31.550 --> 00:31:32.650 align:middle line:90% to those doing damage. 00:31:32.650 --> 00:31:33.870 align:middle line:90% We can't. 00:31:33.870 --> 00:31:34.410 align:middle line:90% OK? 00:31:34.410 --> 00:31:37.090 align:middle line:84% We also could use other tools like penalties on drunk driving. 00:31:37.090 --> 00:31:38.730 align:middle line:84% Indeed, penalties on drunk driving are pretty bad. 00:31:38.730 --> 00:31:40.630 align:middle line:84% You guys really should not drive drunk. 00:31:40.630 --> 00:31:45.390 align:middle line:84% By the time you get your second DUI, busted for driving twice, 00:31:45.390 --> 00:31:47.410 align:middle line:84% I'm informed-- not personal experience-- 00:31:47.410 --> 00:31:49.850 align:middle line:84% I'm informed that by someone who got their second DUI, 00:31:49.850 --> 00:31:52.558 align:middle line:84% that by that time, they lost the license for three years and paid 00:31:52.558 --> 00:31:54.390 align:middle line:90% $50,000 in legal bills. 00:31:54.390 --> 00:31:56.892 align:middle line:84% So they're extreme, but it's apparently not extreme enough, 00:31:56.892 --> 00:31:59.350 align:middle line:84% because we're having a lot of people dying in drunk driving 00:31:59.350 --> 00:32:00.670 align:middle line:90% accidents. 00:32:00.670 --> 00:32:03.870 align:middle line:84% So this is where we get into what the economists would call 00:32:03.870 --> 00:32:05.198 align:middle line:90% the theory of the second best. 00:32:05.198 --> 00:32:06.990 align:middle line:84% The theory of the second best is the notion 00:32:06.990 --> 00:32:10.150 align:middle line:84% that, well, the ideal would be let's just tax the guys causing 00:32:10.150 --> 00:32:11.190 align:middle line:90% the damage. 00:32:11.190 --> 00:32:15.680 align:middle line:84% If we can't do that, should we resort to no taxes or some tax 00:32:15.680 --> 00:32:16.620 align:middle line:90% on everyone? 00:32:16.620 --> 00:32:18.920 align:middle line:84% And the answer is, some tax on everyone. 00:32:18.920 --> 00:32:24.120 align:middle line:84% And the reason is because the disutility to a modest drinker 00:32:24.120 --> 00:32:27.440 align:middle line:84% from a tax is small relative to the benefit 00:32:27.440 --> 00:32:30.000 align:middle line:90% of reducing excessive drinking. 00:32:30.000 --> 00:32:33.220 align:middle line:84% So it's not as good as just taxing excessive drinking. 00:32:33.220 --> 00:32:36.880 align:middle line:84% But the benefits of excessive drinking exceed the disutility. 00:32:36.880 --> 00:32:39.840 align:middle line:84% Because remember, deadweight loss are small near the optimum. 00:32:39.840 --> 00:32:42.200 align:middle line:90% Think about a small alcohol tax. 00:32:42.200 --> 00:32:44.662 align:middle line:84% If you're drinking and not causing damage, 00:32:44.662 --> 00:32:47.120 align:middle line:84% that small alcohol tax will not cause much deadweight loss. 00:32:47.120 --> 00:32:49.682 align:middle line:84% If you're drinking and causing a lot of damage, 00:32:49.682 --> 00:32:51.640 align:middle line:84% that small alcohol tax could reduce that damage 00:32:51.640 --> 00:32:53.760 align:middle line:84% and really reduce that deadweight loss wedge. 00:32:53.760 --> 00:32:56.980 align:middle line:84% So by that logic, we always want some alcohol tax. 00:32:56.980 --> 00:32:59.800 align:middle line:84% The question, how big can it be before the damage to those 00:32:59.800 --> 00:33:03.840 align:middle line:84% who aren't causing cost to society 00:33:03.840 --> 00:33:06.720 align:middle line:84% exceed the benefits of reducing excessive drinking? 00:33:06.720 --> 00:33:10.560 align:middle line:90% OK, questions about that? 00:33:10.560 --> 00:33:11.220 align:middle line:90% All right. 00:33:11.220 --> 00:33:14.040 align:middle line:90% 00:33:14.040 --> 00:33:15.960 align:middle line:90% Cigarettes are easy. 00:33:15.960 --> 00:33:17.240 align:middle line:90% Drinking is harder. 00:33:17.240 --> 00:33:19.880 align:middle line:84% Now let's talk about illicit drugs. 00:33:19.880 --> 00:33:21.720 align:middle line:90% This is a real challenging case. 00:33:21.720 --> 00:33:23.680 align:middle line:84% And why is this a challenging case? 00:33:23.680 --> 00:33:25.920 align:middle line:84% It's a challenging case because here, 00:33:25.920 --> 00:33:30.680 align:middle line:84% the externalities largely arise from the illegality itself. 00:33:30.680 --> 00:33:32.385 align:middle line:90% Think about it. 00:33:32.385 --> 00:33:33.760 align:middle line:84% Why do drugs cause externalities? 00:33:33.760 --> 00:33:35.960 align:middle line:84% They cause crime because people are 00:33:35.960 --> 00:33:39.760 align:middle line:84% trying to pay for the expensive drugs, or the criminal market. 00:33:39.760 --> 00:33:42.272 align:middle line:84% If drugs were legal, that wouldn't exist. 00:33:42.272 --> 00:33:44.480 align:middle line:84% They'd be cheap because they'd be able to flow freely 00:33:44.480 --> 00:33:45.740 align:middle line:90% in and out of the country. 00:33:45.740 --> 00:33:48.282 align:middle line:84% And there wouldn't be criminal activity associated with them. 00:33:48.282 --> 00:33:51.445 align:middle line:84% So in fact, the biggest externalities of drugs 00:33:51.445 --> 00:33:52.820 align:middle line:84% would go away if they were legal. 00:33:52.820 --> 00:33:54.528 align:middle line:84% And that's why for years, economists have 00:33:54.528 --> 00:33:56.440 align:middle line:90% argued for legalizing drugs. 00:33:56.440 --> 00:33:58.240 align:middle line:84% By and large, they've said, look, 00:33:58.240 --> 00:34:00.752 align:middle line:84% the externality argument cannot justify illegal drugs. 00:34:00.752 --> 00:34:02.960 align:middle line:84% In fact, it's pretty hard to justify illegal anything 00:34:02.960 --> 00:34:04.885 align:middle line:90% under externalities. 00:34:04.885 --> 00:34:07.260 align:middle line:84% The economists say if it's really an externality problem, 00:34:07.260 --> 00:34:10.719 align:middle line:84% we should basically make them legal and tax them. 00:34:10.719 --> 00:34:14.210 align:middle line:84% But the question is, are the internalities especially large 00:34:14.210 --> 00:34:17.530 align:middle line:84% here, because of the addictiveness 00:34:17.530 --> 00:34:20.850 align:middle line:84% combined with the extreme consequences of mistakes? 00:34:20.850 --> 00:34:25.590 align:middle line:84% If I make a mistake and start smoking and I'm unable to quit, 00:34:25.590 --> 00:34:27.810 align:middle line:84% I will live seven years less on average. 00:34:27.810 --> 00:34:31.969 align:middle line:84% But if I make a mistake and get hooked on heroin or fentanyl, 00:34:31.969 --> 00:34:34.210 align:middle line:90% I could die 60 years less. 00:34:34.210 --> 00:34:35.449 align:middle line:90% I could die tomorrow. 00:34:35.449 --> 00:34:38.850 align:middle line:84% So since the consequences are more extreme, 00:34:38.850 --> 00:34:43.170 align:middle line:84% people are more worried about considering internalities. 00:34:43.170 --> 00:34:46.362 align:middle line:84% This comes up with the only political issue my wife 00:34:46.362 --> 00:34:48.570 align:middle line:84% and I have ever had a fundamental disagreement, which 00:34:48.570 --> 00:34:51.290 align:middle line:84% is the legalization of marijuana. 00:34:51.290 --> 00:34:53.210 align:middle line:84% She felt marijuana shouldn't be legalized 00:34:53.210 --> 00:34:55.010 align:middle line:90% because it's bad for people. 00:34:55.010 --> 00:34:58.010 align:middle line:84% I said basically, no, the externalities of marijuana 00:34:58.010 --> 00:34:59.730 align:middle line:90% are due to the illegality. 00:34:59.730 --> 00:35:01.930 align:middle line:84% There's no evidence it's that bad for people. 00:35:01.930 --> 00:35:03.290 align:middle line:90% We should legalize it. 00:35:03.290 --> 00:35:06.330 align:middle line:84% Now, it has been legalized in a number of states, as we know. 00:35:06.330 --> 00:35:08.570 align:middle line:84% The evidence so far is that certainly 00:35:08.570 --> 00:35:11.140 align:middle line:84% marijuana-related crime is way down. 00:35:11.140 --> 00:35:14.380 align:middle line:90% It's not illegal anymore. 00:35:14.380 --> 00:35:16.645 align:middle line:84% Overall, traffic deaths are not worse. 00:35:16.645 --> 00:35:18.020 align:middle line:84% Although it's a cool study by one 00:35:18.020 --> 00:35:20.060 align:middle line:84% of my students showing that traffic deaths go up 00:35:20.060 --> 00:35:23.060 align:middle line:90% a lot around dispensaries. 00:35:23.060 --> 00:35:25.040 align:middle line:84% And so overall traffic death numbers, 00:35:25.040 --> 00:35:27.300 align:middle line:84% they go up a lot around dispensaries. 00:35:27.300 --> 00:35:31.060 align:middle line:84% There is much, much more marijuana consumption. 00:35:31.060 --> 00:35:34.540 align:middle line:84% But it looks like most of that is among adults, not kids. 00:35:34.540 --> 00:35:36.700 align:middle line:84% So far, I would say the evidence of legalization 00:35:36.700 --> 00:35:39.622 align:middle line:84% is fairly positive, but the jury's still out. 00:35:39.622 --> 00:35:42.080 align:middle line:84% This is the kind of thing-- in a perfect world, once again, 00:35:42.080 --> 00:35:45.400 align:middle line:84% a lot of the policies we talk about where we need evidence, 00:35:45.400 --> 00:35:46.663 align:middle line:90% we'd put it temporarily. 00:35:46.663 --> 00:35:49.080 align:middle line:84% It's going to be hard to put this genie back in the bottle 00:35:49.080 --> 00:35:50.720 align:middle line:84% if we decide marijuana legalization is bad. 00:35:50.720 --> 00:35:52.387 align:middle line:84% It's going to be hard to deal with that. 00:35:52.387 --> 00:35:57.020 align:middle line:84% We're never banning alcohol or cigarettes. 00:35:57.020 --> 00:36:02.200 align:middle line:84% So the question is, how do we feel about that? 00:36:02.200 --> 00:36:04.440 align:middle line:84% And then in particular, if marijuana should be legal, 00:36:04.440 --> 00:36:05.820 align:middle line:90% what else should it be? 00:36:05.820 --> 00:36:08.280 align:middle line:84% And how do we think about these trade-offs? 00:36:08.280 --> 00:36:10.980 align:middle line:84% And basically, it really comes down to a lot of the fact 00:36:10.980 --> 00:36:12.820 align:middle line:84% that people are really worried about 00:36:12.820 --> 00:36:16.500 align:middle line:84% these internal damages people do for the more serious drugs. 00:36:16.500 --> 00:36:19.700 align:middle line:84% Now, of course, the biggest drug problem in our country today 00:36:19.700 --> 00:36:23.620 align:middle line:84% is opioids, and currently the fentanyl crisis. 00:36:23.620 --> 00:36:28.100 align:middle line:84% Here's a fascinating case where public policy backfired. 00:36:28.100 --> 00:36:29.080 align:middle line:90% What happened? 00:36:29.080 --> 00:36:34.100 align:middle line:84% What happened was in this country before the 21st century, 00:36:34.100 --> 00:36:36.820 align:middle line:84% we used to systematically undertreat pain. 00:36:36.820 --> 00:36:39.800 align:middle line:84% If someone came in and was in pain, we basically ignored them. 00:36:39.800 --> 00:36:41.550 align:middle line:84% There was a realization we were doing that 00:36:41.550 --> 00:36:43.780 align:middle line:84% and it was bad for people at the same time 00:36:43.780 --> 00:36:48.060 align:middle line:84% as this miraculous new drug called OxyContin was invented. 00:36:48.060 --> 00:36:52.180 align:middle line:84% Doctors knew for 150 years that opioids were addictive and bad 00:36:52.180 --> 00:36:52.680 align:middle line:90% for you. 00:36:52.680 --> 00:36:54.520 align:middle line:84% So doctors didn't treat pain with opioids 00:36:54.520 --> 00:36:55.980 align:middle line:84% because people would be addicted. 00:36:55.980 --> 00:36:57.620 align:middle line:84% OxyContin came along and said, we 00:36:57.620 --> 00:36:59.840 align:middle line:84% have a new formulation that's slow release, 00:36:59.840 --> 00:37:01.200 align:middle line:90% so you won't get addicted. 00:37:01.200 --> 00:37:04.100 align:middle line:84% And there's even a New England Journal of Medicine article 00:37:04.100 --> 00:37:05.140 align:middle line:90% to prove it. 00:37:05.140 --> 00:37:07.390 align:middle line:84% Well, it turns out the New England Journal of Medicine 00:37:07.390 --> 00:37:09.910 align:middle line:84% article, much like our case with our autism article, 00:37:09.910 --> 00:37:12.150 align:middle line:84% was some letter written by some guy who 00:37:12.150 --> 00:37:14.310 align:middle line:84% treated 10 guys at a hospital in Boston 00:37:14.310 --> 00:37:16.870 align:middle line:84% and said they didn't get addicted, so OxyContin's OK. 00:37:16.870 --> 00:37:20.310 align:middle line:84% But that was enough for Purdue Pharma to go to the doctors. 00:37:20.310 --> 00:37:22.585 align:middle line:84% And the doctors said, look, well geez, 00:37:22.585 --> 00:37:24.210 align:middle line:84% I'd love to make people's pain go away. 00:37:24.210 --> 00:37:25.250 align:middle line:84% If you're telling me it's non-addictive 00:37:25.250 --> 00:37:27.333 align:middle line:84% and I can make people's pain go away, let's do it. 00:37:27.333 --> 00:37:29.470 align:middle line:90% And OxyContin exploded. 00:37:29.470 --> 00:37:33.470 align:middle line:84% The problem is, OxyContin was overused and prone 00:37:33.470 --> 00:37:35.830 align:middle line:84% to abuse, which is you could crack the pills in half 00:37:35.830 --> 00:37:37.350 align:middle line:90% and snort it. 00:37:37.350 --> 00:37:40.030 align:middle line:84% And it gave you real high and became abused. 00:37:40.030 --> 00:37:42.150 align:middle line:84% So once we started realizing more people 00:37:42.150 --> 00:37:45.470 align:middle line:90% were dying from drug overdoses-- 00:37:45.470 --> 00:37:46.290 align:middle line:90% and we see this. 00:37:46.290 --> 00:37:50.990 align:middle line:84% You can actually see this if you look at figure 6-6. 00:37:50.990 --> 00:37:54.330 align:middle line:84% This shows deaths of car crashes, deaths from overdoses. 00:37:54.330 --> 00:37:56.890 align:middle line:84% I remember the first time I saw this graph in 2010. 00:37:56.890 --> 00:37:57.870 align:middle line:90% I couldn't believe it. 00:37:57.870 --> 00:38:01.790 align:middle line:84% Starting in about 2008, you see the deaths from overdoses 00:38:01.790 --> 00:38:04.270 align:middle line:84% starts to exceed deaths of car crashes. 00:38:04.270 --> 00:38:06.150 align:middle line:84% That was exceptional to think that there 00:38:06.150 --> 00:38:09.550 align:middle line:84% were more overdose deaths than-- and it continues to go up. 00:38:09.550 --> 00:38:13.310 align:middle line:84% So that basically you had a situation where there 00:38:13.310 --> 00:38:16.370 align:middle line:90% was excessive deaths from this. 00:38:16.370 --> 00:38:18.790 align:middle line:84% So the government realizes that we have to end this abuse. 00:38:18.790 --> 00:38:21.270 align:middle line:84% We have to really create an abuse-deterrent formulation 00:38:21.270 --> 00:38:24.030 align:middle line:84% of OxyContin where you couldn't crack it in half. 00:38:24.030 --> 00:38:25.890 align:middle line:84% That was a well-intentioned public policy. 00:38:25.890 --> 00:38:27.030 align:middle line:90% What was the problem? 00:38:27.030 --> 00:38:28.905 align:middle line:84% The problem was you had hundreds of thousands 00:38:28.905 --> 00:38:31.670 align:middle line:84% to millions of people addicted to opioids. 00:38:31.670 --> 00:38:33.570 align:middle line:84% They said, fine, we can't get OxyContin. 00:38:33.570 --> 00:38:37.090 align:middle line:84% We'll go to the next thing, which is heroin to meet our fix. 00:38:37.090 --> 00:38:39.590 align:middle line:84% The problem is heroin really kills you. 00:38:39.590 --> 00:38:41.570 align:middle line:84% And then from there, they went to fentanyl, 00:38:41.570 --> 00:38:43.430 align:middle line:90% which really, really kills you. 00:38:43.430 --> 00:38:46.190 align:middle line:84% So the problem is we created this addiction. 00:38:46.190 --> 00:38:48.047 align:middle line:84% And in trying to fight one aspect, 00:38:48.047 --> 00:38:50.130 align:middle line:84% we shifted people onto a more dangerous substance. 00:38:50.130 --> 00:38:52.755 align:middle line:84% And you see the incredible peak in the number of people who die 00:38:52.755 --> 00:38:57.510 align:middle line:84% from overdose deaths, reaching more than 100,000 people a year. 00:38:57.510 --> 00:39:00.830 align:middle line:84% More than 100,000 a year, basically all caused in many 00:39:00.830 --> 00:39:05.000 align:middle line:84% ways because of this initial prescription drug. 00:39:05.000 --> 00:39:07.160 align:middle line:84% And that raises-- we talked in the first lecture 00:39:07.160 --> 00:39:09.240 align:middle line:84% about the role of the FDA-regulated drugs. 00:39:09.240 --> 00:39:12.440 align:middle line:84% This is the kind of challenge we face in our society. 00:39:12.440 --> 00:39:14.937 align:middle line:84% Once again, this is why we have to constantly-- 00:39:14.937 --> 00:39:17.270 align:middle line:84% I talk about a well-meaning, well-functioning government 00:39:17.270 --> 00:39:18.220 align:middle line:90% in this class. 00:39:18.220 --> 00:39:19.240 align:middle line:84% But that's why we have to constantly 00:39:19.240 --> 00:39:21.448 align:middle line:84% be vigilant about the consequence of the government's 00:39:21.448 --> 00:39:22.360 align:middle line:90% actions. 00:39:22.360 --> 00:39:23.460 align:middle line:90% Yeah. 00:39:23.460 --> 00:39:25.520 align:middle line:90% AUDIENCE: I think it was that-- 00:39:25.520 --> 00:39:28.720 align:middle line:84% I'm not completely familiar-- but I think that the doctors had 00:39:28.720 --> 00:39:33.760 align:middle line:84% to prescribe the opioid, and then couldn't the doctors say 00:39:33.760 --> 00:39:36.560 align:middle line:84% if they knew it was bad, that you don't have to take-- 00:39:36.560 --> 00:39:39.080 align:middle line:84% you don't have to get the prescription filled? 00:39:39.080 --> 00:39:40.083 align:middle line:90% Would that be-- 00:39:40.083 --> 00:39:42.000 align:middle line:84% JON GRUBER: Yeah, I mean, there was a lot of-- 00:39:42.000 --> 00:39:45.015 align:middle line:84% I was involved with some legal cases with opioids. 00:39:45.015 --> 00:39:47.140 align:middle line:84% There's a lot of-- it's unclear who the bad guy is. 00:39:47.140 --> 00:39:48.660 align:middle line:90% There's a lot of bad guys here. 00:39:48.660 --> 00:39:50.320 align:middle line:90% Purdue Pharma is really bad. 00:39:50.320 --> 00:39:51.460 align:middle line:90% They pushed it too much. 00:39:51.460 --> 00:39:53.250 align:middle line:84% The doctors, some doctors were trying 00:39:53.250 --> 00:39:55.500 align:middle line:84% to do right by their patients, saying, you're in pain. 00:39:55.500 --> 00:39:56.417 align:middle line:90% I'm going to help you. 00:39:56.417 --> 00:39:57.560 align:middle line:90% Some were not. 00:39:57.560 --> 00:39:59.260 align:middle line:84% Some were like, great, I prescribe this 00:39:59.260 --> 00:40:00.260 align:middle line:90% and make a lot of money. 00:40:00.260 --> 00:40:03.240 align:middle line:84% I'm just going to prescribe it and you don't have to take it. 00:40:03.240 --> 00:40:04.800 align:middle line:84% Some doctors went worse and literally 00:40:04.800 --> 00:40:08.160 align:middle line:84% were making a prescription for what they call pill mills. 00:40:08.160 --> 00:40:11.680 align:middle line:84% But there's a lot of blame to go around here. 00:40:11.680 --> 00:40:12.380 align:middle line:90% Other questions? 00:40:12.380 --> 00:40:12.760 align:middle line:90% Yeah. 00:40:12.760 --> 00:40:14.427 align:middle line:84% AUDIENCE: What about policy that perhaps 00:40:14.427 --> 00:40:16.560 align:middle line:84% would stop health insurance companies 00:40:16.560 --> 00:40:18.880 align:middle line:90% from paying for opioids? 00:40:18.880 --> 00:40:21.620 align:middle line:84% JON GRUBER: Well, I mean, once again, delivered appropriately, 00:40:21.620 --> 00:40:22.760 align:middle line:90% we wouldn't want that. 00:40:22.760 --> 00:40:26.100 align:middle line:84% Opioids are miraculous post-surgery. 00:40:26.100 --> 00:40:28.860 align:middle line:84% It's a simple policy, which is they should have sent you 00:40:28.860 --> 00:40:30.800 align:middle line:84% home with-- when you get a surgery now, 00:40:30.800 --> 00:40:32.380 align:middle line:84% they send you home with 10 pills. 00:40:32.380 --> 00:40:35.240 align:middle line:84% They send you home with 30 and a refill for six months. 00:40:35.240 --> 00:40:36.840 align:middle line:90% So it's not that hard to change. 00:40:36.840 --> 00:40:39.515 align:middle line:84% It's just it was in no one's financial interest to change it. 00:40:39.515 --> 00:40:41.640 align:middle line:84% So that's why these lawsuits are targeting everyone 00:40:41.640 --> 00:40:43.760 align:middle line:84% around the chain, because everyone was bad. 00:40:43.760 --> 00:40:45.880 align:middle line:84% Like the distributors, they're the people 00:40:45.880 --> 00:40:48.480 align:middle line:84% who distributed the drugs to stores. 00:40:48.480 --> 00:40:50.880 align:middle line:84% They should have noticed that there were areas in Florida 00:40:50.880 --> 00:40:54.080 align:middle line:84% where stores were wearing 10,000 pills a day in little 00:40:54.080 --> 00:40:55.340 align:middle line:90% communities in Florida. 00:40:55.340 --> 00:40:57.798 align:middle line:84% They should know, wait a second, there's no way people need 00:40:57.798 --> 00:41:00.160 align:middle line:84% 10,000 OxyContins a day in podunk Florida. 00:41:00.160 --> 00:41:04.170 align:middle line:84% But they let it go because they were making money, et cetera. 00:41:04.170 --> 00:41:05.330 align:middle line:90% OK? 00:41:05.330 --> 00:41:09.330 align:middle line:84% Now, that is all child's play compared 00:41:09.330 --> 00:41:10.750 align:middle line:84% to the hardest one of these at all 00:41:10.750 --> 00:41:13.410 align:middle line:84% and the number one public policy issue, probably 00:41:13.410 --> 00:41:16.650 align:middle line:84% in the-- public health issue probably in the world today, 00:41:16.650 --> 00:41:19.130 align:middle line:90% which is obesity. 00:41:19.130 --> 00:41:23.025 align:middle line:84% Obesity is a really hard problem because obesity is incredibly 00:41:23.025 --> 00:41:24.650 align:middle line:84% difficult. Obesity is typically defined 00:41:24.650 --> 00:41:28.090 align:middle line:84% as having a body mass index which is ratio of weight 00:41:28.090 --> 00:41:31.810 align:middle line:84% to height above a certain level-- ratio of height 00:41:31.810 --> 00:41:34.690 align:middle line:84% to weight, a function of that above a certain level. 00:41:34.690 --> 00:41:39.100 align:middle line:84% The share of Americans who are obese rose from 13% in 1960 00:41:39.100 --> 00:41:42.570 align:middle line:90% to 45% today. 00:41:42.570 --> 00:41:44.183 align:middle line:90% 45% of Americans are obese. 00:41:44.183 --> 00:41:45.850 align:middle line:84% Now, you might hear that number and say, 00:41:45.850 --> 00:41:46.950 align:middle line:90% that sounds strange to me. 00:41:46.950 --> 00:41:47.910 align:middle line:90% And indeed, it should. 00:41:47.910 --> 00:41:51.970 align:middle line:84% I have not taught an obese student in at least 15 years. 00:41:51.970 --> 00:41:53.850 align:middle line:90% I teach a lot of students. 00:41:53.850 --> 00:41:54.570 align:middle line:90% OK? 00:41:54.570 --> 00:41:57.210 align:middle line:84% Obesity, much like smoking, has been 00:41:57.210 --> 00:41:59.890 align:middle line:84% concentrated in certain segments of the population. 00:41:59.890 --> 00:42:03.720 align:middle line:84% The achievement-focused segment of the population-- 00:42:03.720 --> 00:42:06.820 align:middle line:84% how many of you guys drink either Diet Coke or Coke 00:42:06.820 --> 00:42:09.180 align:middle line:90% or any soda product? 00:42:09.180 --> 00:42:09.880 align:middle line:90% Don't be shy. 00:42:09.880 --> 00:42:11.260 align:middle line:90% I do all the time. 00:42:11.260 --> 00:42:11.780 align:middle line:90% OK? 00:42:11.780 --> 00:42:12.447 align:middle line:90% This is amazing. 00:42:12.447 --> 00:42:14.863 align:middle line:84% I remember the first time I asked that question about Coke 00:42:14.863 --> 00:42:17.292 align:middle line:84% Zero in class and I thought I'd take a survey of Coke Zero 00:42:17.292 --> 00:42:18.000 align:middle line:90% versus Diet Coke. 00:42:18.000 --> 00:42:19.875 align:middle line:84% I said, how many of you guys drink Coke Zero? 00:42:19.875 --> 00:42:20.840 align:middle line:90% 10% raised their hand. 00:42:20.840 --> 00:42:21.640 align:middle line:90% How many drink Diet Coke? 00:42:21.640 --> 00:42:22.520 align:middle line:90% 10% raise their hand. 00:42:22.520 --> 00:42:23.500 align:middle line:84% I'm like, what about the rest of you? 00:42:23.500 --> 00:42:24.280 align:middle line:84% They were like, we don't drink soda. 00:42:24.280 --> 00:42:25.447 align:middle line:90% And I'm like, what the fuck? 00:42:25.447 --> 00:42:26.700 align:middle line:90% You don't drink soda? 00:42:26.700 --> 00:42:27.580 align:middle line:90% What do you mean? 00:42:27.580 --> 00:42:31.752 align:middle line:84% And it's just a health-conscious thing. 00:42:31.752 --> 00:42:32.960 align:middle line:90% You're more health conscious. 00:42:32.960 --> 00:42:34.480 align:middle line:84% They drink a lot of soda in America. 00:42:34.480 --> 00:42:36.580 align:middle line:84% Soda industry is doing fine, just not you guys. 00:42:36.580 --> 00:42:40.980 align:middle line:84% Obesity is a major, major problem in this nation. 00:42:40.980 --> 00:42:43.180 align:middle line:90% And not just in this nation. 00:42:43.180 --> 00:42:47.060 align:middle line:84% There are about on the order of almost 1 billion adults 00:42:47.060 --> 00:42:49.060 align:middle line:90% worldwide who are obese. 00:42:49.060 --> 00:42:51.620 align:middle line:84% And fundamentally, the problem in the world 00:42:51.620 --> 00:42:54.060 align:middle line:84% is moving from hunger to obesity. 00:42:54.060 --> 00:42:55.780 align:middle line:84% Fundamentally, we've conquered hunger, 00:42:55.780 --> 00:42:58.660 align:middle line:90% and the problem is obesity. 00:42:58.660 --> 00:43:01.420 align:middle line:90% Now, why is that happening? 00:43:01.420 --> 00:43:03.820 align:middle line:90% Well really, it's two things. 00:43:03.820 --> 00:43:05.300 align:middle line:84% The first and less important part 00:43:05.300 --> 00:43:07.460 align:middle line:84% is the decline in physical activity. 00:43:07.460 --> 00:43:08.400 align:middle line:90% As someone said once-- 00:43:08.400 --> 00:43:11.200 align:middle line:84% I love this-- we used to get paid to work out. 00:43:11.200 --> 00:43:14.100 align:middle line:90% Now we have to pay to work out. 00:43:14.100 --> 00:43:17.180 align:middle line:84% We're fundamentally not active anymore, and therefore 00:43:17.180 --> 00:43:19.368 align:middle line:90% not getting physical activity. 00:43:19.368 --> 00:43:21.660 align:middle line:84% But that's less important than the second factor, which 00:43:21.660 --> 00:43:25.020 align:middle line:84% is the availability of cheap and easy-to-prepare but bad-for-you 00:43:25.020 --> 00:43:26.460 align:middle line:90% foods. 00:43:26.460 --> 00:43:29.060 align:middle line:84% Preparing a dinner used to be a major time-consuming 00:43:29.060 --> 00:43:32.680 align:middle line:84% undertaking that involved healthy ingredients. 00:43:32.680 --> 00:43:39.020 align:middle line:84% Now it's Doritos, which are cheap and quickly available. 00:43:39.020 --> 00:43:42.700 align:middle line:84% It is astonishing how much more expensive fruits and vegetables 00:43:42.700 --> 00:43:45.020 align:middle line:84% are than things that are bad for you. 00:43:45.020 --> 00:43:48.140 align:middle line:84% And as a result, there's relative price incentives 00:43:48.140 --> 00:43:49.120 align:middle line:90% to eat badly. 00:43:49.120 --> 00:43:52.820 align:middle line:84% And that's what's also caused the obesity. 00:43:52.820 --> 00:43:53.820 align:middle line:90% OK? 00:43:53.820 --> 00:43:55.900 align:middle line:90% This has caused obesity. 00:43:55.900 --> 00:43:58.930 align:middle line:84% This has an enormous external consequence, 00:43:58.930 --> 00:44:00.190 align:middle line:90% which is the rise in diabetes. 00:44:00.190 --> 00:44:02.510 align:middle line:84% Or as my daughter said to me, "diabeetus", 00:44:02.510 --> 00:44:05.150 align:middle line:84% from that meme from many years ago. 00:44:05.150 --> 00:44:08.350 align:middle line:84% From diabetes, which is a progressive, chronic disease 00:44:08.350 --> 00:44:14.910 align:middle line:84% that's incurable, that is now affecting about 10.5% 00:44:14.910 --> 00:44:16.370 align:middle line:84% of the US population is diabetic. 00:44:16.370 --> 00:44:17.950 align:middle line:84% About one third of the US population 00:44:17.950 --> 00:44:20.550 align:middle line:84% is pre-diabetic, which means they're at risk of diabetes. 00:44:20.550 --> 00:44:21.630 align:middle line:90% That's incredible. 00:44:21.630 --> 00:44:24.310 align:middle line:84% Basically, every child born today, one out of three 00:44:24.310 --> 00:44:26.170 align:middle line:90% will develop diabetes. 00:44:26.170 --> 00:44:26.670 align:middle line:90% OK? 00:44:26.670 --> 00:44:28.310 align:middle line:90% That's an astonishing rate. 00:44:28.310 --> 00:44:30.050 align:middle line:84% And that is enormous consequence, 00:44:30.050 --> 00:44:32.870 align:middle line:90% both external and internal. 00:44:32.870 --> 00:44:38.950 align:middle line:84% The US spends about $350 billion a year on obesity-related costs. 00:44:38.950 --> 00:44:42.590 align:middle line:84% And if someone's obese, it shortens their life span 00:44:42.590 --> 00:44:45.150 align:middle line:84% by an average of about four years. 00:44:45.150 --> 00:44:47.550 align:middle line:84% So we're talking about 40% of Americans 00:44:47.550 --> 00:44:50.190 align:middle line:84% are shortening their lifespan an average of four years. 00:44:50.190 --> 00:44:52.350 align:middle line:90% That is astonishing. 00:44:52.350 --> 00:44:53.910 align:middle line:84% This is an enormous, enormous problem 00:44:53.910 --> 00:44:56.430 align:middle line:90% that we're facing worldwide. 00:44:56.430 --> 00:44:59.590 align:middle line:84% The problem is, this is really hard to deal with, 00:44:59.590 --> 00:45:01.890 align:middle line:84% because remember I talked about, smoking is easy. 00:45:01.890 --> 00:45:03.510 align:middle line:90% Every cigarette is bad for you. 00:45:03.510 --> 00:45:05.710 align:middle line:84% Alcohol is harder because for a while, it's not bad, 00:45:05.710 --> 00:45:06.890 align:middle line:90% and then it's bad for you. 00:45:06.890 --> 00:45:11.310 align:middle line:84% Eating is really good for you until it isn't. 00:45:11.310 --> 00:45:13.450 align:middle line:84% So you don't want to stop people from eating. 00:45:13.450 --> 00:45:14.908 align:middle line:84% We'd be great if you stopped people 00:45:14.908 --> 00:45:17.430 align:middle line:84% from smoking, actually not terrible if you stopped people 00:45:17.430 --> 00:45:20.150 align:middle line:84% from drinking, but really bad if you stopped people from eating. 00:45:20.150 --> 00:45:23.330 align:middle line:84% So it's a way harder problem to deal with, because once again, 00:45:23.330 --> 00:45:25.490 align:middle line:90% nonlinear taxes are really hard. 00:45:25.490 --> 00:45:28.910 align:middle line:84% It's hard to tax heavy eating and not non-heavy eating. 00:45:28.910 --> 00:45:30.675 align:middle line:90% So what do we do about that? 00:45:30.675 --> 00:45:32.550 align:middle line:84% Well, there's a number of public policy tools 00:45:32.550 --> 00:45:34.230 align:middle line:84% that have been introduced to try to deal 00:45:34.230 --> 00:45:36.670 align:middle line:90% with the obesity problem. 00:45:36.670 --> 00:45:39.610 align:middle line:84% The first is to try to change the nature of the food supply. 00:45:39.610 --> 00:45:41.808 align:middle line:84% And the term you'll often hear is food deserts. 00:45:41.808 --> 00:45:43.850 align:middle line:84% These are areas where there's bad food available, 00:45:43.850 --> 00:45:46.110 align:middle line:90% but not good food. 00:45:46.110 --> 00:45:48.170 align:middle line:84% The evidence on that is kind of mixed. 00:45:48.170 --> 00:45:51.170 align:middle line:84% There's some evidence that if you live closer to fast food, 00:45:51.170 --> 00:45:52.268 align:middle line:90% you're in worse health. 00:45:52.268 --> 00:45:54.310 align:middle line:84% But basically, the easy availability of fast food 00:45:54.310 --> 00:45:55.378 align:middle line:90% puts you in worse health. 00:45:55.378 --> 00:45:56.920 align:middle line:84% But there's unfortunately no evidence 00:45:56.920 --> 00:45:59.680 align:middle line:84% that living closer to good food makes you healthier, 00:45:59.680 --> 00:46:02.478 align:middle line:84% probably because it's just too damn expensive. 00:46:02.478 --> 00:46:04.020 align:middle line:84% So when you solve these food deserts, 00:46:04.020 --> 00:46:05.560 align:middle line:84% it doesn't make people healthier. 00:46:05.560 --> 00:46:08.200 align:middle line:84% That solution by itself isn't really working. 00:46:08.200 --> 00:46:12.160 align:middle line:84% Another approach is to tax unhealthy foods. 00:46:12.160 --> 00:46:15.840 align:middle line:84% In particular, there's a lot of interest now in soda taxes. 00:46:15.840 --> 00:46:20.000 align:middle line:84% A major source of obesity in the US is sugar sodas. 00:46:20.000 --> 00:46:22.560 align:middle line:90% It's a major source of obesity. 00:46:22.560 --> 00:46:25.000 align:middle line:90% And basically the idea is-- 00:46:25.000 --> 00:46:27.600 align:middle line:84% and soft drink consumption increased by 300% 00:46:27.600 --> 00:46:30.480 align:middle line:90% from 1998 to 2008. 00:46:30.480 --> 00:46:32.340 align:middle line:84% Now, I don't know what's happened recently. 00:46:32.340 --> 00:46:33.840 align:middle line:90% I should update that. 00:46:33.840 --> 00:46:37.100 align:middle line:84% Now, there is a number of cities in the US-- 00:46:37.100 --> 00:46:41.400 align:middle line:84% Philly, Berkeley, other cities-- that put in sugary soda taxes 00:46:41.400 --> 00:46:46.920 align:middle line:84% and put in taxes on sodas to try to reduce soda consumption. 00:46:46.920 --> 00:46:49.400 align:middle line:84% And they are working to some extent. 00:46:49.400 --> 00:46:50.900 align:middle line:90% They're not a huge effect. 00:46:50.900 --> 00:46:52.280 align:middle line:84% There's no real good evidence yet 00:46:52.280 --> 00:46:54.322 align:middle line:84% that they're significantly affecting body weight. 00:46:54.322 --> 00:46:56.870 align:middle line:84% They do seem to be deterring soda consumption to some extent, 00:46:56.870 --> 00:47:00.370 align:middle line:84% but not really lowering body weight. 00:47:00.370 --> 00:47:02.290 align:middle line:84% Now, other countries are more aggressive. 00:47:02.290 --> 00:47:05.730 align:middle line:84% Mexico is taxing sugar-- all sugar products. 00:47:05.730 --> 00:47:07.790 align:middle line:84% I've worked on diabetes in Mexico. 00:47:07.790 --> 00:47:10.330 align:middle line:84% Mexico has the world's worst diabetes problem. 00:47:10.330 --> 00:47:14.530 align:middle line:90% Mexico spends something like-- 00:47:14.530 --> 00:47:19.210 align:middle line:84% it's something like 1.5% of their entire GDP just dealing 00:47:19.210 --> 00:47:21.250 align:middle line:84% with the complications of diabetes. 00:47:21.250 --> 00:47:25.750 align:middle line:84% And it's partly biological, the makeup of the citizens. 00:47:25.750 --> 00:47:27.250 align:middle line:90% It's also their diets. 00:47:27.250 --> 00:47:30.610 align:middle line:84% So Mexico put a sugar tax on all sugars, 00:47:30.610 --> 00:47:34.610 align:middle line:84% not just on sodas, but on other things. 00:47:34.610 --> 00:47:36.370 align:middle line:90% So that's one approach. 00:47:36.370 --> 00:47:38.090 align:middle line:84% But the problem with that is it doesn't 00:47:38.090 --> 00:47:41.650 align:middle line:84% get all the causes of obesity and people can substitute. 00:47:41.650 --> 00:47:43.410 align:middle line:90% Fat's really bad for you too. 00:47:43.410 --> 00:47:45.360 align:middle line:84% You tax sugar, people switch to fat. 00:47:45.360 --> 00:47:46.610 align:middle line:90% Tax fat, they switch to sugar. 00:47:46.610 --> 00:47:48.370 align:middle line:84% Tax both, they switch to something else. 00:47:48.370 --> 00:47:51.430 align:middle line:84% So it's a really hard problem to deal with just by taxes. 00:47:51.430 --> 00:47:55.810 align:middle line:84% Probably the least amenable to taxes among these problems. 00:47:55.810 --> 00:47:59.010 align:middle line:84% A third approach is more information, 00:47:59.010 --> 00:48:02.670 align:middle line:84% explaining to people how bad certain things are for you. 00:48:02.670 --> 00:48:05.130 align:middle line:84% Indeed, I knew someone who was a partner of Dunkin' Donuts. 00:48:05.130 --> 00:48:06.730 align:middle line:84% And he told me that when they put calorie information 00:48:06.730 --> 00:48:09.330 align:middle line:84% on Dunkin' Donuts, donut consumption did not change 00:48:09.330 --> 00:48:13.010 align:middle line:84% at all, but muffin consumption fell through the floor, 00:48:13.010 --> 00:48:15.090 align:middle line:84% because everybody thought muffins were OK. 00:48:15.090 --> 00:48:16.630 align:middle line:84% Turns out they're worse than donuts. 00:48:16.630 --> 00:48:17.450 align:middle line:90% Who knew? 00:48:17.450 --> 00:48:19.750 align:middle line:84% So basically, information can help. 00:48:19.750 --> 00:48:21.250 align:middle line:84% And there's studies of what happens. 00:48:21.250 --> 00:48:23.542 align:middle line:84% The Affordable Care Act actually included a requirement 00:48:23.542 --> 00:48:25.890 align:middle line:84% that large restaurants include calorie information 00:48:25.890 --> 00:48:26.552 align:middle line:90% on their menus. 00:48:26.552 --> 00:48:28.010 align:middle line:84% And studies that have shown that it 00:48:28.010 --> 00:48:30.930 align:middle line:84% has impacted eating habits modestly, 00:48:30.930 --> 00:48:32.330 align:middle line:90% that information can help. 00:48:32.330 --> 00:48:35.570 align:middle line:84% Just once again, the effect is modest. 00:48:35.570 --> 00:48:39.290 align:middle line:84% Now finally, what may be most effective in battling 00:48:39.290 --> 00:48:41.930 align:middle line:84% obesity is thinking about self-control strategies. 00:48:41.930 --> 00:48:44.490 align:middle line:84% Can we actually engage with self-control problems people 00:48:44.490 --> 00:48:45.690 align:middle line:90% face? 00:48:45.690 --> 00:48:49.090 align:middle line:84% And here, there is an enormous amount of literature 00:48:49.090 --> 00:48:52.620 align:middle line:84% which suggests that small behavioral interventions can 00:48:52.620 --> 00:48:54.543 align:middle line:84% actually massively change how people eat. 00:48:54.543 --> 00:48:56.460 align:middle line:84% So for example, my favorite study in this area 00:48:56.460 --> 00:48:58.780 align:middle line:84% is they took people who were going to a movie who'd all 00:48:58.780 --> 00:49:01.020 align:middle line:84% eaten lunch at the same time, and they gave them 00:49:01.020 --> 00:49:03.620 align:middle line:84% buckets of stale popcorn of different sizes-- 00:49:03.620 --> 00:49:04.503 align:middle line:90% small, medium, large. 00:49:04.503 --> 00:49:06.420 align:middle line:84% Basically, the bigger bucket of stale popcorn, 00:49:06.420 --> 00:49:07.502 align:middle line:90% the more they ate. 00:49:07.502 --> 00:49:09.460 align:middle line:84% Even though they thought they were equally full 00:49:09.460 --> 00:49:11.540 align:middle line:84% and the popcorn was stale, if you put more in front of them, 00:49:11.540 --> 00:49:12.360 align:middle line:90% they eat more. 00:49:12.360 --> 00:49:12.860 align:middle line:90% OK? 00:49:12.860 --> 00:49:15.420 align:middle line:84% Another fun study someone did up at Dartmouth. 00:49:15.420 --> 00:49:18.680 align:middle line:84% So when you go to buffet, you walk down the buffet, 00:49:18.680 --> 00:49:19.740 align:middle line:90% pick what you like. 00:49:19.740 --> 00:49:22.220 align:middle line:84% And it turns out, at a typical buffet, about 15% of food 00:49:22.220 --> 00:49:23.080 align:middle line:90% is wasted. 00:49:23.080 --> 00:49:23.980 align:middle line:84% Because by the time you get to the end, 00:49:23.980 --> 00:49:25.140 align:middle line:84% like oh, I didn't know this was going to be here, 00:49:25.140 --> 00:49:26.165 align:middle line:90% or I already over-- 00:49:26.165 --> 00:49:28.040 align:middle line:84% I'm going to pile it on the edge of my plate. 00:49:28.040 --> 00:49:29.022 align:middle line:90% You don't eat it. 00:49:29.022 --> 00:49:30.980 align:middle line:84% So someone switched a buffet so the plates were 00:49:30.980 --> 00:49:32.940 align:middle line:84% at the end rather than the beginning, 00:49:32.940 --> 00:49:35.320 align:middle line:84% and consumption from the buffet went down 15%. 00:49:35.320 --> 00:49:36.820 align:middle line:84% Because you walk by and you're like, 00:49:36.820 --> 00:49:37.720 align:middle line:90% oh, that's going to be there. 00:49:37.720 --> 00:49:39.700 align:middle line:84% I know I shouldn't take so much at the beginning. 00:49:39.700 --> 00:49:41.325 align:middle line:84% These sorts of behavioral interventions 00:49:41.325 --> 00:49:42.620 align:middle line:90% can be very effective. 00:49:42.620 --> 00:49:45.900 align:middle line:84% And so that's another tool we can use. 00:49:45.900 --> 00:49:47.100 align:middle line:90% OK? 00:49:47.100 --> 00:49:50.500 align:middle line:84% Finally, we have perhaps what's the ultimate tool, which 00:49:50.500 --> 00:49:54.460 align:middle line:84% you say, look, what really is the damage that's done here-- 00:49:54.460 --> 00:49:58.580 align:middle line:84% the damage that's done here is really done by excessive body 00:49:58.580 --> 00:49:59.080 align:middle line:90% weight. 00:49:59.080 --> 00:50:01.540 align:middle line:90% So why don't we just tax that? 00:50:01.540 --> 00:50:03.140 align:middle line:90% We'll have a skinny subsidy. 00:50:03.140 --> 00:50:05.540 align:middle line:84% We can subsidize people to be skinny. 00:50:05.540 --> 00:50:09.180 align:middle line:84% Now, obviously I'm not running for office. 00:50:09.180 --> 00:50:12.460 align:middle line:84% That would be clearly a dramatic challenge. 00:50:12.460 --> 00:50:14.828 align:middle line:84% But it's also not obviously a great policy-- 00:50:14.828 --> 00:50:16.120 align:middle line:90% not obviously the right policy. 00:50:16.120 --> 00:50:17.340 align:middle line:90% Why is that? 00:50:17.340 --> 00:50:23.193 align:middle line:84% If the problem is self-control, then basically how 00:50:23.193 --> 00:50:25.860 align:middle line:84% are you going to deal with it by something that taxes people are 00:50:25.860 --> 00:50:27.900 align:middle line:90% going to pay in 15 months? 00:50:27.900 --> 00:50:29.900 align:middle line:84% So it's January, and next April I'm 00:50:29.900 --> 00:50:31.900 align:middle line:84% going to pay a higher tax because I got fat. 00:50:31.900 --> 00:50:32.920 align:middle line:90% I'm not going to focus on that. 00:50:32.920 --> 00:50:34.500 align:middle line:84% I'm focused on what's in front of me. 00:50:34.500 --> 00:50:36.420 align:middle line:84% So it's also not obviously the right policy 00:50:36.420 --> 00:50:38.900 align:middle line:84% in terms of if the problem is self-control. 00:50:38.900 --> 00:50:41.340 align:middle line:84% The bottom line is obesity is the single hardest of all 00:50:41.340 --> 00:50:43.100 align:middle line:90% these problems to deal with. 00:50:43.100 --> 00:50:45.340 align:middle line:84% The answer is probably a broad combination 00:50:45.340 --> 00:50:48.018 align:middle line:84% of all these policies need to be put in place to deal with it. 00:50:48.018 --> 00:50:50.310 align:middle line:84% But it's really the fundamental public health challenge 00:50:50.310 --> 00:50:51.830 align:middle line:90% of our time. 00:50:51.830 --> 00:50:52.590 align:middle line:90% OK? 00:50:52.590 --> 00:50:55.270 align:middle line:90% Questions about that? 00:50:55.270 --> 00:50:56.030 align:middle line:90% Yeah. 00:50:56.030 --> 00:50:58.670 align:middle line:84% AUDIENCE: You said exercising [INAUDIBLE] exercise now 00:50:58.670 --> 00:50:59.770 align:middle line:90% before we have it. 00:50:59.770 --> 00:51:03.070 align:middle line:84% Is there thoughts on subsidizing exercise in some way? 00:51:03.070 --> 00:51:05.670 align:middle line:84% JON GRUBER: Well, I mean, the trick with exercising 00:51:05.670 --> 00:51:09.890 align:middle line:84% is it's way less effective than lowering eating So basically, 00:51:09.890 --> 00:51:13.390 align:middle line:84% roughly speaking, walking for an hour is kind 00:51:13.390 --> 00:51:15.750 align:middle line:90% of like a muffin and a donut. 00:51:15.750 --> 00:51:17.450 align:middle line:90% So that's a lot of exercise. 00:51:17.450 --> 00:51:20.110 align:middle line:84% And it's not that much food that you're offsetting. 00:51:20.110 --> 00:51:23.790 align:middle line:84% So basically, exercise is harder. 00:51:23.790 --> 00:51:25.243 align:middle line:90% There's a lot of policies. 00:51:25.243 --> 00:51:26.910 align:middle line:84% So one thing the Affordable Care Act did 00:51:26.910 --> 00:51:29.118 align:middle line:84% is it allowed companies to actually value your health 00:51:29.118 --> 00:51:31.110 align:middle line:84% insurance premium by your participation 00:51:31.110 --> 00:51:32.615 align:middle line:84% in things like exercise programs. 00:51:32.615 --> 00:51:34.990 align:middle line:84% So there are now financial incentives to exercise and get 00:51:34.990 --> 00:51:35.698 align:middle line:90% people to commit. 00:51:35.698 --> 00:51:37.073 align:middle line:84% And that's clearly going to help. 00:51:37.073 --> 00:51:38.770 align:middle line:84% Basically, there's no magic bullet here. 00:51:38.770 --> 00:51:40.187 align:middle line:84% We're going to have to keep trying 00:51:40.187 --> 00:51:42.070 align:middle line:90% all these different approaches. 00:51:42.070 --> 00:51:42.590 align:middle line:90% Yeah. 00:51:42.590 --> 00:51:45.470 align:middle line:84% AUDIENCE: What about the weight loss drugs like Ozempic? 00:51:45.470 --> 00:51:46.410 align:middle line:90% JON GRUBER: Ah, great. 00:51:46.410 --> 00:51:47.240 align:middle line:90% Excellent point. 00:51:47.240 --> 00:51:49.323 align:middle line:84% I actually have an editorial on the New York Times 00:51:49.323 --> 00:51:52.600 align:middle line:84% that I published a few weeks ago about this drug. 00:51:52.600 --> 00:51:56.235 align:middle line:84% So we're going to talk about the incredibly new miraculous 00:51:56.235 --> 00:51:57.860 align:middle line:84% drugs we're developing in this country. 00:51:57.860 --> 00:51:58.920 align:middle line:90% This is one of them. 00:51:58.920 --> 00:52:01.295 align:middle line:84% We'll talk about all this when we talk about health care. 00:52:01.295 --> 00:52:02.840 align:middle line:90% This is an incredible drug. 00:52:02.840 --> 00:52:04.720 align:middle line:84% Now, we don't quite know how it works. 00:52:04.720 --> 00:52:06.870 align:middle line:84% But basically, if you talk to people on it, 00:52:06.870 --> 00:52:09.700 align:middle line:84% it essentially makes you feel satiated. 00:52:09.700 --> 00:52:11.480 align:middle line:84% In the past, we've tried to fight obesity 00:52:11.480 --> 00:52:13.180 align:middle line:90% with chemical interventions. 00:52:13.180 --> 00:52:15.880 align:middle line:84% We've done so in a way that makes you unhappy. 00:52:15.880 --> 00:52:17.460 align:middle line:84% This doesn't make-- feels like, meh, 00:52:17.460 --> 00:52:19.320 align:middle line:90% I just don't want it anymore. 00:52:19.320 --> 00:52:20.720 align:middle line:90% That's great. 00:52:20.720 --> 00:52:22.500 align:middle line:84% It basically works on the margin. 00:52:22.500 --> 00:52:24.960 align:middle line:84% We talked about the margin we need and want. 00:52:24.960 --> 00:52:28.440 align:middle line:84% Basically, it used to be that basically other drugs would 00:52:28.440 --> 00:52:29.980 align:middle line:90% effect both need and want. 00:52:29.980 --> 00:52:32.360 align:middle line:84% This would effects want, but not need. 00:52:32.360 --> 00:52:33.660 align:middle line:90% So basically, it's incredible. 00:52:33.660 --> 00:52:35.660 align:middle line:84% It also looks like-- there's early evidence that 00:52:35.660 --> 00:52:38.760 align:middle line:84% may affect other socially undesirable behaviors 00:52:38.760 --> 00:52:41.600 align:middle line:84% like excessive drinking and gambling and things like that. 00:52:41.600 --> 00:52:44.780 align:middle line:84% So it seems like something that's a really miraculous drug. 00:52:44.780 --> 00:52:48.200 align:middle line:84% It's also $15,000 and you have to be on it forever. 00:52:48.200 --> 00:52:52.280 align:middle line:84% So my coauthors and I estimated that if every obese person 00:52:52.280 --> 00:52:55.300 align:middle line:84% in the US was made eligible for Ozempic and the government paid 00:52:55.300 --> 00:52:56.800 align:middle line:84% for it, it would cost the government 00:52:56.800 --> 00:52:59.240 align:middle line:84% $800 billion every single year, which 00:52:59.240 --> 00:53:02.400 align:middle line:84% is the size of the Medicare program we'll talk about later. 00:53:02.400 --> 00:53:05.560 align:middle line:84% That said, Ozempic is now in these-- the GLP-1s are now 00:53:05.560 --> 00:53:07.940 align:middle line:90% underused, quite frankly. 00:53:07.940 --> 00:53:09.666 align:middle line:84% Because they're so expensive, they're 00:53:09.666 --> 00:53:12.660 align:middle line:84% only prescribed for certain very well-targeted cases. 00:53:12.660 --> 00:53:13.680 align:middle line:90% Here's a great example. 00:53:13.680 --> 00:53:15.607 align:middle line:84% You can get Ozempic if you're diabetic. 00:53:15.607 --> 00:53:17.940 align:middle line:84% Well, if you're diabetic, it's sort of already too late. 00:53:17.940 --> 00:53:19.580 align:middle line:84% I mean, it's great to control your weight. 00:53:19.580 --> 00:53:21.663 align:middle line:84% But the truth is, the guys who aren't yet diabetic 00:53:21.663 --> 00:53:24.667 align:middle line:84% are who we really want on these drugs and they can't get them. 00:53:24.667 --> 00:53:26.500 align:middle line:84% So in some sense, these drugs are underused. 00:53:26.500 --> 00:53:27.740 align:middle line:84% On the other hand, they're really expensive. 00:53:27.740 --> 00:53:29.310 align:middle line:84% And this is something we'll talk a lot about when 00:53:29.310 --> 00:53:31.860 align:middle line:84% we talk about health care, which is drug price regulation. 00:53:31.860 --> 00:53:32.360 align:middle line:90% Yeah. 00:53:32.360 --> 00:53:33.485 align:middle line:90% AUDIENCE: I'm just curious. 00:53:33.485 --> 00:53:35.240 align:middle line:84% So with [INAUDIBLE], how much would 00:53:35.240 --> 00:53:37.240 align:middle line:84% the government save if you subsidized 00:53:37.240 --> 00:53:40.720 align:middle line:84% for every obese patient in the US? 00:53:40.720 --> 00:53:42.180 align:middle line:90% JON GRUBER: If you subsidized? 00:53:42.180 --> 00:53:43.180 align:middle line:90% AUDIENCE: These drugs-- 00:53:43.180 --> 00:53:43.920 align:middle line:84% JON GRUBER: No, no, no, this is the thing. 00:53:43.920 --> 00:53:45.308 align:middle line:90% That includes the cost. 00:53:45.308 --> 00:53:46.850 align:middle line:84% This is what was said in the article, 00:53:46.850 --> 00:53:49.590 align:middle line:84% which is that everybody's like-- so the CEO of Novo Nordisk said, 00:53:49.590 --> 00:53:51.690 align:middle line:84% this is a great deal, $15,000, because it saves money down 00:53:51.690 --> 00:53:52.310 align:middle line:90% the road. 00:53:52.310 --> 00:53:53.210 align:middle line:90% He's wrong. 00:53:53.210 --> 00:53:56.610 align:middle line:84% Basically, if everyone was obese, 00:53:56.610 --> 00:53:58.210 align:middle line:84% this drug would cost us a trillion 00:53:58.210 --> 00:54:01.930 align:middle line:84% and we'd save $200 billion for a net of $800 billion. 00:54:01.930 --> 00:54:04.710 align:middle line:84% So if the drug was cheap enough, it could pay for itself, 00:54:04.710 --> 00:54:06.850 align:middle line:84% but nowhere near the current price. 00:54:06.850 --> 00:54:07.410 align:middle line:90% Yeah. 00:54:07.410 --> 00:54:10.170 align:middle line:84% AUDIENCE: How do you reconcile that some individuals who 00:54:10.170 --> 00:54:16.330 align:middle line:84% are obese may have preferences or choices from before they're 00:54:16.330 --> 00:54:17.383 align:middle line:90% 18 years old? 00:54:17.383 --> 00:54:19.050 align:middle line:84% JON GRUBER: It's a great question, which 00:54:19.050 --> 00:54:22.250 align:middle line:84% this is in some sense one of the most fascinating things we had 00:54:22.250 --> 00:54:24.130 align:middle line:84% to deal with in health care reform, 00:54:24.130 --> 00:54:28.850 align:middle line:84% was the line between self-control and lack 00:54:28.850 --> 00:54:29.810 align:middle line:90% of control. 00:54:29.810 --> 00:54:32.570 align:middle line:84% So when we passed the Affordable Care Act, part of what we did 00:54:32.570 --> 00:54:35.473 align:middle line:84% is we made health insurance communal. 00:54:35.473 --> 00:54:36.890 align:middle line:84% We got everybody health insurance. 00:54:36.890 --> 00:54:40.050 align:middle line:84% And every email that didn't contain a swear 00:54:40.050 --> 00:54:44.410 align:middle line:84% that I got was about, why should I pay more for my fat neighbor? 00:54:44.410 --> 00:54:45.550 align:middle line:90% Why should I pay more? 00:54:45.550 --> 00:54:47.883 align:middle line:84% And the answer is, well, because maybe your fat neighbor 00:54:47.883 --> 00:54:49.810 align:middle line:84% is fat for reasons they couldn't control. 00:54:49.810 --> 00:54:52.250 align:middle line:84% But if that neighbor's fat because they just ate a lot, 00:54:52.250 --> 00:54:53.690 align:middle line:84% maybe we do want to think about how much we 00:54:53.690 --> 00:54:54.710 align:middle line:90% want to subsidize that. 00:54:54.710 --> 00:54:56.377 align:middle line:84% So that's a really interesting question, 00:54:56.377 --> 00:54:58.090 align:middle line:84% of how you address sin behaviors is 00:54:58.090 --> 00:55:00.370 align:middle line:84% how do you think about biological propensities 00:55:00.370 --> 00:55:02.390 align:middle line:84% versus bad behaviors is really hard. 00:55:02.390 --> 00:55:04.050 align:middle line:90% It's a hard question. 00:55:04.050 --> 00:55:04.850 align:middle line:90% Yeah. 00:55:04.850 --> 00:55:06.730 align:middle line:84% AUDIENCE: Michelle Obama's, I don't 00:55:06.730 --> 00:55:08.167 align:middle line:90% know, crusade against lunches-- 00:55:08.167 --> 00:55:10.250 align:middle line:84% JON GRUBER: I don't think she was against lunches. 00:55:10.250 --> 00:55:12.010 align:middle line:90% AUDIENCE: Oh, yeah. 00:55:12.010 --> 00:55:16.270 align:middle line:84% I know that was a talking point probably about a decade ago. 00:55:16.270 --> 00:55:16.770 align:middle line:90% Yeah. 00:55:16.770 --> 00:55:18.050 align:middle line:90% I'm older, I guess. 00:55:18.050 --> 00:55:20.970 align:middle line:84% Do we have any evidence that was effective or not effective? 00:55:20.970 --> 00:55:21.970 align:middle line:90% Or if that did anything? 00:55:21.970 --> 00:55:23.220 align:middle line:90% JON GRUBER: No, we don't know. 00:55:23.220 --> 00:55:25.810 align:middle line:84% I mean, I don't think public good interest campaigns really 00:55:25.810 --> 00:55:27.370 align:middle line:90% moved the needle. 00:55:27.370 --> 00:55:30.230 align:middle line:84% I think she had a campaign against poor eating, 00:55:30.230 --> 00:55:31.430 align:middle line:90% against poor nutrition. 00:55:31.430 --> 00:55:33.790 align:middle line:84% I think to move the needle until we move it massively. 00:55:33.790 --> 00:55:35.130 align:middle line:84% That was the thing with smoking, which 00:55:35.130 --> 00:55:36.922 align:middle line:84% is we had these things going, going, going. 00:55:36.922 --> 00:55:39.610 align:middle line:84% All of a sudden, boom, it became OK to be against smoking, 00:55:39.610 --> 00:55:40.950 align:middle line:90% and things changed overnight. 00:55:40.950 --> 00:55:43.020 align:middle line:84% So it's maybe building a movement 00:55:43.020 --> 00:55:45.900 align:middle line:84% that eventually matters, but it hasn't moved the needle so far. 00:55:45.900 --> 00:55:46.780 align:middle line:90% Yeah. 00:55:46.780 --> 00:55:49.113 align:middle line:84% AUDIENCE: I was wondering, wouldn't it be more effective 00:55:49.113 --> 00:55:51.860 align:middle line:84% if health insurance companies would pay for, I guess, 00:55:51.860 --> 00:55:54.555 align:middle line:90% groceries and exercise versus-- 00:55:54.555 --> 00:55:55.680 align:middle line:90% JON GRUBER: Great question. 00:55:55.680 --> 00:55:58.860 align:middle line:84% This is actually a big movement in health care 00:55:58.860 --> 00:56:01.500 align:middle line:84% is what's called the social terms of health movement, which 00:56:01.500 --> 00:56:04.260 align:middle line:84% is actually having health insurance plans pay 00:56:04.260 --> 00:56:07.528 align:middle line:84% for your food, your housing, all these things which ultimately 00:56:07.528 --> 00:56:08.320 align:middle line:90% affect your health. 00:56:08.320 --> 00:56:09.920 align:middle line:84% Quite frankly, probably more than health insurance. 00:56:09.920 --> 00:56:11.460 align:middle line:84% I hate to say it as a health economist. 00:56:11.460 --> 00:56:13.140 align:middle line:84% But look, the truth is, where you live and what you eat 00:56:13.140 --> 00:56:14.515 align:middle line:84% and whether you get shot probably 00:56:14.515 --> 00:56:16.598 align:middle line:84% matters a whole lot more for your health than what 00:56:16.598 --> 00:56:18.220 align:middle line:90% health insurance plan you have. 00:56:18.220 --> 00:56:21.720 align:middle line:84% The problem is, for a health insurance perspective, 00:56:21.720 --> 00:56:23.820 align:middle line:84% it only makes sense to pay for it if it's actually 00:56:23.820 --> 00:56:24.997 align:middle line:90% going to save them money. 00:56:24.997 --> 00:56:27.080 align:middle line:84% So it's very important to distinguish two things-- 00:56:27.080 --> 00:56:28.300 align:middle line:90% and we'll come back to this-- 00:56:28.300 --> 00:56:31.540 align:middle line:84% cost saving versus cost effective. 00:56:31.540 --> 00:56:35.980 align:middle line:84% If I took every one of you today and sent you to the doctor, 00:56:35.980 --> 00:56:38.260 align:middle line:84% that would probably be cost effective, 00:56:38.260 --> 00:56:41.430 align:middle line:84% in the sense that the doctor's visit-- 00:56:41.430 --> 00:56:43.390 align:middle line:84% we'd probably find one or two of you something 00:56:43.390 --> 00:56:45.265 align:middle line:84% wrong that you wouldn't have found otherwise. 00:56:45.265 --> 00:56:46.150 align:middle line:90% Nope, don't worry. 00:56:46.150 --> 00:56:47.430 align:middle line:90% Wouldn't have found otherwise. 00:56:47.430 --> 00:56:51.732 align:middle line:84% And that would be good to find it. 00:56:51.732 --> 00:56:53.190 align:middle line:84% But you probably on net would still 00:56:53.190 --> 00:56:55.230 align:middle line:84% spend more on all the doctors than finding the one or two 00:56:55.230 --> 00:56:56.147 align:middle line:90% things that are wrong. 00:56:56.147 --> 00:56:58.830 align:middle line:84% So it wouldn't be cost saving, but it might be cost effective. 00:56:58.830 --> 00:57:00.455 align:middle line:84% It might be a good thing to do in terms 00:57:00.455 --> 00:57:03.423 align:middle line:84% of the value of your health, but it wouldn't be cost saving. 00:57:03.423 --> 00:57:05.090 align:middle line:84% And the question of social determinants, 00:57:05.090 --> 00:57:07.010 align:middle line:84% I they're pretty clearly about cost savings. 00:57:07.010 --> 00:57:08.450 align:middle line:84% The question is if it's cost effective. 00:57:08.450 --> 00:57:10.450 align:middle line:84% The insurance companies aren't going to do them. 00:57:10.450 --> 00:57:12.590 align:middle line:84% The government would have to subsidize. 00:57:12.590 --> 00:57:13.415 align:middle line:90% Yeah. 00:57:13.415 --> 00:57:15.290 align:middle line:84% AUDIENCE: How is this different from alcohol? 00:57:15.290 --> 00:57:16.090 align:middle line:90% Like you said-- 00:57:16.090 --> 00:57:18.465 align:middle line:84% JON GRUBER: It's different alcohol in a number of senses. 00:57:18.465 --> 00:57:20.790 align:middle line:84% One sense is, remember, alcohol is never good for you. 00:57:20.790 --> 00:57:22.950 align:middle line:84% Now, people enjoy it, but it's never good for you. 00:57:22.950 --> 00:57:24.190 align:middle line:90% Eating is good for you. 00:57:24.190 --> 00:57:26.310 align:middle line:84% So you've got this nonmonotonicity 00:57:26.310 --> 00:57:27.830 align:middle line:84% that you don't have with alcohol. 00:57:27.830 --> 00:57:28.970 align:middle line:90% That's point one. 00:57:28.970 --> 00:57:31.190 align:middle line:84% Point two is, with alcohol, I think 00:57:31.190 --> 00:57:32.482 align:middle line:90% the externalities are enormous. 00:57:32.482 --> 00:57:34.732 align:middle line:84% And obviously, the internalities are a little tougher, 00:57:34.732 --> 00:57:35.690 align:middle line:90% are less obvious here. 00:57:35.690 --> 00:57:38.270 align:middle line:84% There's obvious externalities and internalities. 00:57:38.270 --> 00:57:40.630 align:middle line:84% And the third thing is with alcohol, we pretty much know 00:57:40.630 --> 00:57:42.270 align:middle line:84% how to fix it, which is a higher tax-- 00:57:42.270 --> 00:57:43.410 align:middle line:84% fix it or know how to address it. 00:57:43.410 --> 00:57:44.690 align:middle line:90% A higher tax would address it. 00:57:44.690 --> 00:57:48.750 align:middle line:84% Whereas here, higher taxes don't solve the problem. 00:57:48.750 --> 00:57:49.390 align:middle line:90% Yeah. 00:57:49.390 --> 00:57:52.030 align:middle line:84% AUDIENCE: With the way SNAP, EBT are structured, 00:57:52.030 --> 00:57:55.830 align:middle line:84% the amount of money for food, is that limit-- 00:57:55.830 --> 00:58:00.030 align:middle line:84% the limited amount of cash-- also pushing people 00:58:00.030 --> 00:58:01.950 align:middle line:90% to the cheapest possible option? 00:58:01.950 --> 00:58:03.290 align:middle line:84% JON GRUBER: It's a really interesting question. 00:58:03.290 --> 00:58:04.490 align:middle line:84% We'll come back and talk about SNAP. 00:58:04.490 --> 00:58:05.830 align:middle line:84% That's what we used to call food stamps, which 00:58:05.830 --> 00:58:07.610 align:middle line:84% are subsidies for-- it's what you're talking about, right? 00:58:07.610 --> 00:58:08.270 align:middle line:90% AUDIENCE: Yeah. 00:58:08.270 --> 00:58:09.645 align:middle line:84% JON GRUBER: Which are basically-- 00:58:09.645 --> 00:58:12.550 align:middle line:84% we'll talk in chapter 17 about how we distribute sometimes 00:58:12.550 --> 00:58:14.070 align:middle line:90% in cash, sometimes in goods. 00:58:14.070 --> 00:58:15.928 align:middle line:84% And one form of goods is to give people 00:58:15.928 --> 00:58:18.470 align:middle line:84% essentially debit cards that they can use to buy food cheaply 00:58:18.470 --> 00:58:19.630 align:middle line:90% at the supermarket. 00:58:19.630 --> 00:58:21.950 align:middle line:84% And the question is, well gee, should those 00:58:21.950 --> 00:58:25.342 align:middle line:84% debit cards, for example, only be usable for healthy food? 00:58:25.342 --> 00:58:27.550 align:middle line:84% That comes to the fundamental question of paternalism 00:58:27.550 --> 00:58:30.050 align:middle line:84% that we'll talk about in chapter 17, which is to what extent 00:58:30.050 --> 00:58:33.470 align:middle line:84% is the government going to be paternalistic about what 00:58:33.470 --> 00:58:34.670 align:middle line:90% you use your money for? 00:58:34.670 --> 00:58:37.680 align:middle line:90% And that's a hard question. 00:58:37.680 --> 00:58:38.220 align:middle line:90% Yeah. 00:58:38.220 --> 00:58:40.720 align:middle line:84% AUDIENCE: With alcohol, the whole reason 00:58:40.720 --> 00:58:42.293 align:middle line:84% that it is parabolic is the sense 00:58:42.293 --> 00:58:43.960 align:middle line:84% that if you drink a little, it gives you 00:58:43.960 --> 00:58:45.600 align:middle line:84% a lot of social benefits, such as you 00:58:45.600 --> 00:58:47.120 align:middle line:90% would socialize with friends. 00:58:47.120 --> 00:58:48.870 align:middle line:84% JON GRUBER: Well, the question is, though, 00:58:48.870 --> 00:58:50.680 align:middle line:90% is that internal or external? 00:58:50.680 --> 00:58:54.225 align:middle line:84% I'm saying the externalities are not nonmonotonic. 00:58:54.225 --> 00:58:55.600 align:middle line:84% Now, you're right, there could be 00:58:55.600 --> 00:58:57.920 align:middle line:84% some nonmonotonic externalities if basically we 00:58:57.920 --> 00:59:00.440 align:middle line:84% have social anxiety and alcohol is the way we 00:59:00.440 --> 00:59:01.920 align:middle line:90% fight our social anxiety. 00:59:01.920 --> 00:59:05.080 align:middle line:84% But then there's a question of why would we want-- wouldn't we 00:59:05.080 --> 00:59:07.360 align:middle line:84% want our better tools for fighting social anxiety 00:59:07.360 --> 00:59:08.980 align:middle line:90% than getting drunk? 00:59:08.980 --> 00:59:11.720 align:middle line:90% I'm often fascinated. 00:59:11.720 --> 00:59:14.400 align:middle line:84% I play with kids like my little niece and nephew. 00:59:14.400 --> 00:59:15.213 align:middle line:90% And it's super fun. 00:59:15.213 --> 00:59:16.380 align:middle line:90% And why are little kids fun? 00:59:16.380 --> 00:59:17.780 align:middle line:84% Because the don't have social anxiety, 00:59:17.780 --> 00:59:18.780 align:middle line:90% they're just having fun. 00:59:18.780 --> 00:59:22.560 align:middle line:84% Then somehow we hit puberty and we all become shy and unable-- 00:59:22.560 --> 00:59:24.440 align:middle line:90% unless we're drunk. 00:59:24.440 --> 00:59:26.022 align:middle line:90% And then we're like kids again. 00:59:26.022 --> 00:59:27.480 align:middle line:84% So aren't there better ways that we 00:59:27.480 --> 00:59:29.440 align:middle line:84% could get people to fight their social anxiety than getting 00:59:29.440 --> 00:59:29.940 align:middle line:90% drunk? 00:59:29.940 --> 00:59:33.480 align:middle line:90% 00:59:33.480 --> 00:59:34.180 align:middle line:90% All right. 00:59:34.180 --> 00:59:36.840 align:middle line:84% OK, that's all I want to say about chapter 6. 00:59:36.840 --> 00:59:39.960 align:middle line:84% We're going to turn now and start chapter 7 on public goods. 00:59:39.960 --> 00:59:42.940 align:middle line:84% Remember that this section of the class, 00:59:42.940 --> 00:59:45.160 align:middle line:84% we're talking about externalities on public goods. 00:59:45.160 --> 00:59:49.440 align:middle line:84% Now, what's tricky about public goods is 00:59:49.440 --> 00:59:52.600 align:middle line:84% they're related but somewhat subtly different 00:59:52.600 --> 00:59:55.120 align:middle line:90% than externalities. 00:59:55.120 --> 01:00:00.272 align:middle line:84% And to think about that, let's think about two aspects 01:00:00.272 --> 01:00:01.480 align:middle line:90% along which goods can differ. 01:00:01.480 --> 01:00:04.283 align:middle line:84% You should have a chapter 7 handout as well. 01:00:04.283 --> 01:00:05.700 align:middle line:84% The same handout as the chapter 7. 01:00:05.700 --> 01:00:08.160 align:middle line:90% Look at table 7-1. 01:00:08.160 --> 01:00:12.000 align:middle line:84% Let's think about two aspects of goods we want to focus on. 01:00:12.000 --> 01:00:17.000 align:middle line:84% The first is rivalry and the second is excludability. 01:00:17.000 --> 01:00:21.240 align:middle line:84% Rivalry is whether my consumption of the good 01:00:21.240 --> 01:00:24.400 align:middle line:84% interferes with your consumption of the good. 01:00:24.400 --> 01:00:26.120 align:middle line:84% So if I have the good, does that deter 01:00:26.120 --> 01:00:28.320 align:middle line:90% your enjoyment of the good? 01:00:28.320 --> 01:00:33.720 align:middle line:84% Excludability is, can I stop you from having the good? 01:00:33.720 --> 01:00:37.890 align:middle line:84% A pure private good is one which is both rival and excludable, 01:00:37.890 --> 01:00:39.960 align:middle line:90% like ice cream. 01:00:39.960 --> 01:00:42.210 align:middle line:84% If I have ice cream, you enjoy it less because there's 01:00:42.210 --> 01:00:43.330 align:middle line:90% less there for you. 01:00:43.330 --> 01:00:45.190 align:middle line:84% And I could just not sell you ice cream. 01:00:45.190 --> 01:00:48.530 align:middle line:84% It's rivalrous if every bite I take is one less bite for you. 01:00:48.530 --> 01:00:51.730 align:middle line:84% And it's excludable if I can just not sell it to you. 01:00:51.730 --> 01:00:54.090 align:middle line:90% That is a pure private good. 01:00:54.090 --> 01:00:56.830 align:middle line:84% A pure public good is a good which 01:00:56.830 --> 01:00:59.330 align:middle line:84% is neither rival nor excludable, the ultimate example 01:00:59.330 --> 01:01:01.210 align:middle line:90% being national defense. 01:01:01.210 --> 01:01:03.850 align:middle line:84% If a missile is protecting me, it doesn't deter the missile 01:01:03.850 --> 01:01:04.990 align:middle line:90% from protecting you. 01:01:04.990 --> 01:01:07.730 align:middle line:84% There's no sense in which having a missile protecting an area 01:01:07.730 --> 01:01:08.590 align:middle line:90% is rival. 01:01:08.590 --> 01:01:10.090 align:middle line:84% Once somebody is protected, everybody is protected. 01:01:10.090 --> 01:01:11.070 align:middle line:90% And it's not excludable. 01:01:11.070 --> 01:01:13.112 align:middle line:84% You can't say the missile can protect your house, 01:01:13.112 --> 01:01:14.210 align:middle line:90% but not your house. 01:01:14.210 --> 01:01:20.730 align:middle line:84% OK, national defense is an example of a pure public good. 01:01:20.730 --> 01:01:23.490 align:middle line:84% But we also have what we call impure public goods. 01:01:23.490 --> 01:01:28.370 align:middle line:84% So for example, things which are rival but not excludable, 01:01:28.370 --> 01:01:30.490 align:middle line:84% like, say, a crowded city sidewalk. 01:01:30.490 --> 01:01:32.158 align:middle line:84% A crowded city sidewalk, it's rival. 01:01:32.158 --> 01:01:33.700 align:middle line:84% When I'm on the sidewalk, I interview 01:01:33.700 --> 01:01:34.740 align:middle line:84% if your enjoyment of the sidewalk 01:01:34.740 --> 01:01:36.160 align:middle line:90% because I'm getting in the way. 01:01:36.160 --> 01:01:39.700 align:middle line:84% But you can't not let me on the sidewalk. 01:01:39.700 --> 01:01:41.005 align:middle line:90% That is rival, not excludable. 01:01:41.005 --> 01:01:42.380 align:middle line:84% We have goods that are excludable 01:01:42.380 --> 01:01:45.660 align:middle line:84% but not rival, the best being TV streaming, 01:01:45.660 --> 01:01:49.180 align:middle line:84% which is to first approximation if my neighbor and I 01:01:49.180 --> 01:01:50.592 align:middle line:90% both stream, it does not rival. 01:01:50.592 --> 01:01:52.800 align:middle line:84% At some point, the load gets high enough, it's rival. 01:01:52.800 --> 01:01:54.080 align:middle line:90% But it's generally not rival. 01:01:54.080 --> 01:01:55.500 align:middle line:84% But the cable company is not just delivered 01:01:55.500 --> 01:01:56.458 align:middle line:90% to my neighbor's house. 01:01:56.458 --> 01:01:57.620 align:middle line:90% It's delivered to my house. 01:01:57.620 --> 01:02:00.020 align:middle line:84% So that's another kind of impure public good. 01:02:00.020 --> 01:02:03.900 align:middle line:84% So we basically have these different kinds 01:02:03.900 --> 01:02:06.140 align:middle line:90% of public and private goods. 01:02:06.140 --> 01:02:11.940 align:middle line:84% And it turns out that there's a very different implication 01:02:11.940 --> 01:02:15.700 align:middle line:84% for optimality when talk about purely public and purely private 01:02:15.700 --> 01:02:17.100 align:middle line:90% goods. 01:02:17.100 --> 01:02:18.960 align:middle line:90% So to do this, let's go back. 01:02:18.960 --> 01:02:19.620 align:middle line:90% Yeah. 01:02:19.620 --> 01:02:22.037 align:middle line:84% AUDIENCE: Satellite would fall under purely public though, 01:02:22.037 --> 01:02:22.620 align:middle line:90% right? 01:02:22.620 --> 01:02:23.620 align:middle line:90% JON GRUBER: A satellite? 01:02:23.620 --> 01:02:25.573 align:middle line:84% AUDIENCE: Like satellite television, where-- 01:02:25.573 --> 01:02:26.740 align:middle line:90% JON GRUBER: It's excludable. 01:02:26.740 --> 01:02:29.040 align:middle line:84% AUDIENCE: Or what's the one where it's just around, 01:02:29.040 --> 01:02:30.420 align:middle line:84% and if you have a receiver, you can pick it up 01:02:30.420 --> 01:02:31.420 align:middle line:90% because it's in the air? 01:02:31.420 --> 01:02:36.000 align:middle line:84% JON GRUBER: I mean, I guess in some sense, nothing, right? 01:02:36.000 --> 01:02:41.660 align:middle line:84% I mean, basically-- I guess ham radio or whatever. 01:02:41.660 --> 01:02:43.860 align:middle line:84% Yeah, anything like that would be pure public good. 01:02:43.860 --> 01:02:45.740 align:middle line:90% That's right. 01:02:45.740 --> 01:02:48.048 align:middle line:84% OK, so to do this, what I want to do 01:02:48.048 --> 01:02:50.340 align:middle line:84% is I want to go back and review a little bit our theory 01:02:50.340 --> 01:02:51.940 align:middle line:90% of pure private goods. 01:02:51.940 --> 01:02:54.700 align:middle line:84% And then what I'm going to do is go on and talk 01:02:54.700 --> 01:02:56.160 align:middle line:90% about pure public goods. 01:02:56.160 --> 01:03:00.960 align:middle line:84% So to review this, let's look at figure 7. 01:03:00.960 --> 01:03:06.700 align:middle line:90% Let's look at figure 7-1, OK? 01:03:06.700 --> 01:03:10.620 align:middle line:84% And let's imagine two goods, Ben and Jerry-- two people, Ben 01:03:10.620 --> 01:03:11.620 align:middle line:90% and Jerry. 01:03:11.620 --> 01:03:15.740 align:middle line:84% And Ben and Jerry are deciding between two private goods, ice 01:03:15.740 --> 01:03:19.380 align:middle line:90% cream and ice cream cones. 01:03:19.380 --> 01:03:21.500 align:middle line:90% No, I should say-- 01:03:21.500 --> 01:03:25.090 align:middle line:90% oh, ice cream cones and cookies. 01:03:25.090 --> 01:03:26.840 align:middle line:84% OK, I'm looking at the wrong-- one second. 01:03:26.840 --> 01:03:29.700 align:middle line:90% Sorry about that. 01:03:29.700 --> 01:03:31.750 align:middle line:90% Ice cream cones and cookies. 01:03:31.750 --> 01:03:36.230 align:middle line:84% And basically, they're deciding between consuming these two 01:03:36.230 --> 01:03:38.110 align:middle line:90% pure private goods. 01:03:38.110 --> 01:03:39.530 align:middle line:90% They have a certain budget. 01:03:39.530 --> 01:03:42.090 align:middle line:84% And they can spend that on ice cream cones or on cookies. 01:03:42.090 --> 01:03:44.990 align:middle line:84% And for simplicity, we're going to assume 01:03:44.990 --> 01:03:46.850 align:middle line:90% that the price of cookies is $1. 01:03:46.850 --> 01:03:48.150 align:middle line:84% Remember, we're talking about this modeling trick 01:03:48.150 --> 01:03:49.430 align:middle line:90% called the numeraire good. 01:03:49.430 --> 01:03:52.510 align:middle line:84% We'll assume the price of cookies is $1. 01:03:52.510 --> 01:03:55.050 align:middle line:84% So in this world, given their budget, 01:03:55.050 --> 01:03:57.810 align:middle line:84% as we derive in chapter 2 and derived in 14.01, 01:03:57.810 --> 01:04:00.150 align:middle line:84% you can derive a demand curve for ice cream. 01:04:00.150 --> 01:04:03.140 align:middle line:84% Given a budget, given a utility function, 01:04:03.140 --> 01:04:04.890 align:middle line:84% you can have a demand curve for ice cream. 01:04:04.890 --> 01:04:09.930 align:middle line:84% That demand curve for ice cream is shown for Ben in panel A 01:04:09.930 --> 01:04:13.910 align:middle line:84% and for Jerry in panel B. So Ben's demand for ice cream 01:04:13.910 --> 01:04:19.230 align:middle line:84% is such that at a price of $2, he wants two ice cream cones. 01:04:19.230 --> 01:04:19.910 align:middle line:90% OK? 01:04:19.910 --> 01:04:21.910 align:middle line:90% Jerry likes ice cream less. 01:04:21.910 --> 01:04:24.710 align:middle line:84% At a price of $2, he only wants one ice cream cone. 01:04:24.710 --> 01:04:28.030 align:middle line:84% And they're well-defined demand curves for ice cream. 01:04:28.030 --> 01:04:32.750 align:middle line:84% Now, what is the market demand for ice cream? 01:04:32.750 --> 01:04:35.870 align:middle line:84% Well, the market demand curve for ice cream 01:04:35.870 --> 01:04:41.590 align:middle line:84% is just the sum of Ben and Jerry's demand. 01:04:41.590 --> 01:04:44.250 align:middle line:84% The market demand is the sum of Ben and Jerry's demand. 01:04:44.250 --> 01:04:48.490 align:middle line:84% That's in panel C. In aggregate, if you charge a price of $2, 01:04:48.490 --> 01:04:50.710 align:middle line:84% there's three ice cream cones demanded. 01:04:50.710 --> 01:04:52.750 align:middle line:84% And assuming no externalities, that 01:04:52.750 --> 01:04:55.510 align:middle line:84% is the social marginal benefit of ice cream 01:04:55.510 --> 01:04:57.630 align:middle line:90% in this two-person society. 01:04:57.630 --> 01:05:00.050 align:middle line:84% Assume there's some social marginal cost of producing ice 01:05:00.050 --> 01:05:00.550 align:middle line:90% cream. 01:05:00.550 --> 01:05:05.550 align:middle line:84% Where they intersect is the optimal quantity of ice cream. 01:05:05.550 --> 01:05:09.230 align:middle line:84% That's the first best, the social welfare maximizing 01:05:09.230 --> 01:05:10.950 align:middle line:84% optimum, assuming no externalities, 01:05:10.950 --> 01:05:14.470 align:middle line:84% would be to produce three ice cream cones at a price of $2. 01:05:14.470 --> 01:05:16.670 align:middle line:90% That would be the first best. 01:05:16.670 --> 01:05:22.590 align:middle line:84% OK, now, let's actually write this down mathematically. 01:05:22.590 --> 01:05:28.440 align:middle line:84% If we recall from chapter 2, the optimality condition 01:05:28.440 --> 01:05:35.160 align:middle line:84% for a private good would be that the marginal utility of Ben 01:05:35.160 --> 01:05:44.920 align:middle line:84% for ice cream over the marginal utility of Ben for cookies 01:05:44.920 --> 01:05:51.080 align:middle line:84% is going to be equal to the marginal rate of substitution 01:05:51.080 --> 01:05:52.145 align:middle line:90% for Ben-- 01:05:52.145 --> 01:05:53.520 align:middle line:84% the marginal rate of substitution 01:05:53.520 --> 01:05:54.920 align:middle line:90% between ice cream and cookies. 01:05:54.920 --> 01:05:56.503 align:middle line:84% It's the marginal rate of substitution 01:05:56.503 --> 01:05:59.000 align:middle line:90% for ice cream and cookies. 01:05:59.000 --> 01:06:00.960 align:middle line:90% Ice cream and cookies. 01:06:00.960 --> 01:06:04.760 align:middle line:84% It's the ratio of marginal utilities. 01:06:04.760 --> 01:06:10.200 align:middle line:84% Now, in equilibrium, since everyone is in equilibrium, 01:06:10.200 --> 01:06:13.640 align:middle line:84% that is then equal at the optimality to the price 01:06:13.640 --> 01:06:17.160 align:middle line:90% ratio of ice cream to cookies. 01:06:17.160 --> 01:06:20.600 align:middle line:84% That's the optimality condition derived in 14.01. 01:06:20.600 --> 01:06:23.820 align:middle line:84% Moreover, since there's good equilibrium, 01:06:23.820 --> 01:06:25.650 align:middle line:84% we also know that it's equal to people 01:06:25.650 --> 01:06:29.436 align:middle line:84% to Jerry's marginal rate of substitution between ice cream 01:06:29.436 --> 01:06:30.088 align:middle line:90% and cookies. 01:06:30.088 --> 01:06:32.630 align:middle line:84% Because if Ben and Jerry had different marginal substitution, 01:06:32.630 --> 01:06:33.650 align:middle line:84% they should shift their consumption 01:06:33.650 --> 01:06:35.150 align:middle line:84% to the marginal rate of substitution 01:06:35.150 --> 01:06:36.770 align:middle line:90% or equal to the price ratio. 01:06:36.770 --> 01:06:38.930 align:middle line:84% So the marginal rate of substitution for Jerry 01:06:38.930 --> 01:06:40.722 align:middle line:84% must deal the marginal rate of substitution 01:06:40.722 --> 01:06:41.850 align:middle line:90% for Ben in equilibrium. 01:06:41.850 --> 01:06:44.650 align:middle line:84% And they're both equal to that market price ratio. 01:06:44.650 --> 01:06:48.210 align:middle line:84% And that's the social welfare optimum. 01:06:48.210 --> 01:06:54.330 align:middle line:84% We also know that on the production side, 01:06:54.330 --> 01:06:59.610 align:middle line:84% firms will produce until the marginal cost of producing ice 01:06:59.610 --> 01:07:03.970 align:middle line:84% cream relative to the marginal cost of producing cookies 01:07:03.970 --> 01:07:07.770 align:middle line:84% is equal to the price of ice cream over the price of cookies. 01:07:07.770 --> 01:07:10.270 align:middle line:84% That's the supply side condition. 01:07:10.270 --> 01:07:14.170 align:middle line:84% That's what firms are going to optimize. 01:07:14.170 --> 01:07:20.250 align:middle line:84% So basically, if we are going to assume that the price of cookies 01:07:20.250 --> 01:07:21.810 align:middle line:90% is $1-- 01:07:21.810 --> 01:07:24.890 align:middle line:84% OK, let's assume the marginal cost of cookies is also $1-- 01:07:24.890 --> 01:07:28.490 align:middle line:84% we can safely say the marginal cost of ice cream 01:07:28.490 --> 01:07:30.710 align:middle line:84% is equal to the price of ice cream. 01:07:30.710 --> 01:07:34.570 align:middle line:84% We're just turning ratios into levels by using numeraire goods. 01:07:34.570 --> 01:07:39.310 align:middle line:84% So basically, the marginal cost of ice cream 01:07:39.310 --> 01:07:41.570 align:middle line:84% and the price of ice cream says this is $1. 01:07:41.570 --> 01:07:43.410 align:middle line:84% That means the marginal cost of ice cream 01:07:43.410 --> 01:07:46.570 align:middle line:84% is equal to the marginal rate of substitution between ice cream 01:07:46.570 --> 01:07:47.770 align:middle line:90% and cookies. 01:07:47.770 --> 01:07:49.110 align:middle line:90% And that's the equilibrium. 01:07:49.110 --> 01:07:50.716 align:middle line:84% We know the price ratio is equal to marginal rate 01:07:50.716 --> 01:07:51.570 align:middle line:90% of substitution. 01:07:51.570 --> 01:07:52.730 align:middle line:84% Since we assume the price of a cookie 01:07:52.730 --> 01:07:54.410 align:middle line:84% is $1, that means the price level is 01:07:54.410 --> 01:07:56.410 align:middle line:84% equal to the marginal rate of substitution. 01:07:56.410 --> 01:08:00.090 align:middle line:84% So price ratio is equal to the marginal rate of substitution. 01:08:00.090 --> 01:08:03.075 align:middle line:84% And that is the equilibrium condition in a private market. 01:08:03.075 --> 01:08:03.950 align:middle line:90% Questions about that? 01:08:03.950 --> 01:08:05.075 align:middle line:90% Once again, I'm going fast. 01:08:05.075 --> 01:08:07.730 align:middle line:84% But I'm assuming you guys all know this from chapter 2. 01:08:07.730 --> 01:08:08.330 align:middle line:90% OK? 01:08:08.330 --> 01:08:10.290 align:middle line:90% Questions about that? 01:08:10.290 --> 01:08:12.010 align:middle line:84% Now, imagine we change the problem 01:08:12.010 --> 01:08:15.650 align:middle line:84% in a small way, which is we're now choosing between missiles 01:08:15.650 --> 01:08:17.133 align:middle line:90% and cookies. 01:08:17.133 --> 01:08:18.550 align:middle line:84% OK, no more ice cream and cookies. 01:08:18.550 --> 01:08:21.090 align:middle line:84% Ben and Jerry got militaristic and are now choosing 01:08:21.090 --> 01:08:23.620 align:middle line:90% between missiles and cookies. 01:08:23.620 --> 01:08:27.500 align:middle line:84% So the difference between missiles and ice cream cones 01:08:27.500 --> 01:08:29.660 align:middle line:84% is that missiles are public goods. 01:08:29.660 --> 01:08:32.700 align:middle line:84% They're non-rival and non-excludable. 01:08:32.700 --> 01:08:35.859 align:middle line:84% So now how does that change the problem? 01:08:35.859 --> 01:08:37.580 align:middle line:90% OK, let's look at figure 7-2. 01:08:37.580 --> 01:08:40.100 align:middle line:90% 01:08:40.100 --> 01:08:44.899 align:middle line:84% The difference is now when we choose a quantity of missiles, 01:08:44.899 --> 01:08:46.720 align:middle line:84% we do not choose them individually. 01:08:46.720 --> 01:08:48.740 align:middle line:84% We not choose each person's number of missiles 01:08:48.740 --> 01:08:51.032 align:middle line:84% like we choose each person's number of ice cream cones. 01:08:51.032 --> 01:08:53.220 align:middle line:90% We choose a number of missiles-- 01:08:53.220 --> 01:08:55.660 align:middle line:84% a socially optimal number of missiles. 01:08:55.660 --> 01:09:01.700 align:middle line:84% And so what we now have here is that we have each individual's 01:09:01.700 --> 01:09:03.580 align:middle line:84% demand for missiles, which is essentially-- 01:09:03.580 --> 01:09:04.997 align:middle line:84% remember, what's the demand curve? 01:09:04.997 --> 01:09:06.979 align:middle line:90% it's a willingness to pay curve. 01:09:06.979 --> 01:09:09.300 align:middle line:84% So we can graph each individual's willingness 01:09:09.300 --> 01:09:10.899 align:middle line:90% to pay for missiles. 01:09:10.899 --> 01:09:13.260 align:middle line:90% So take Ben. 01:09:13.260 --> 01:09:18.180 align:middle line:84% Ben is willing to pay $2 for one missile 01:09:18.180 --> 01:09:20.800 align:middle line:90% and $1 for five missiles. 01:09:20.800 --> 01:09:21.300 align:middle line:90% OK? 01:09:21.300 --> 01:09:22.960 align:middle line:84% It's a downward-sloping demand curve. 01:09:22.960 --> 01:09:25.240 align:middle line:84% So for the first missile, he's going to pay $2. 01:09:25.240 --> 01:09:26.979 align:middle line:84% But then demand is diminishing, because one missile is going 01:09:26.979 --> 01:09:28.123 align:middle line:90% to protect him pretty well. 01:09:28.123 --> 01:09:29.540 align:middle line:84% For a second missile-- if you want 01:09:29.540 --> 01:09:33.140 align:middle line:84% to buy-- then its demand is downwards, is diminishing. 01:09:33.140 --> 01:09:34.979 align:middle line:90% OK? 01:09:34.979 --> 01:09:38.660 align:middle line:84% Jerry, on the other hand, is a real militaristic guy. 01:09:38.660 --> 01:09:41.080 align:middle line:84% Jerry is willing to pay $4 for the first missile. 01:09:41.080 --> 01:09:45.939 align:middle line:84% Even by the fifth missile, he's still willing to pay $2. 01:09:45.939 --> 01:09:48.080 align:middle line:84% So that is their underlying willingness to pay, 01:09:48.080 --> 01:09:49.580 align:middle line:90% which is a virtual demand curve. 01:09:49.580 --> 01:09:52.859 align:middle line:84% The willingness to pay and demand are symmetric. 01:09:52.859 --> 01:09:53.920 align:middle line:90% What does that mean? 01:09:53.920 --> 01:09:57.420 align:middle line:84% That means that society's willingness to pay 01:09:57.420 --> 01:09:59.820 align:middle line:90% is the vertical sum-- 01:09:59.820 --> 01:10:02.540 align:middle line:84% not the horizontal sum, as in the previous case-- 01:10:02.540 --> 01:10:07.420 align:middle line:84% but the vertical sum of the individual's willingness to pay. 01:10:07.420 --> 01:10:13.260 align:middle line:84% So society-- remember, Ben was willing to pay 01:10:13.260 --> 01:10:15.620 align:middle line:90% $2 for the first missile. 01:10:15.620 --> 01:10:17.760 align:middle line:84% Jerry's willing to pay $4 for the first missile. 01:10:17.760 --> 01:10:20.810 align:middle line:84% So society is willing to pay $6 for the first missile. 01:10:20.810 --> 01:10:21.310 align:middle line:90% Why? 01:10:21.310 --> 01:10:24.190 align:middle line:84% Because the missile benefits them both equally. 01:10:24.190 --> 01:10:27.150 align:middle line:84% Unlike the ice cream cone, ice cream cone Ben gets and Jerry 01:10:27.150 --> 01:10:27.930 align:middle line:90% doesn't get. 01:10:27.930 --> 01:10:30.270 align:middle line:84% Now if Ben gets the missile, Jerry gets the missile. 01:10:30.270 --> 01:10:31.370 align:middle line:90% It's a public good. 01:10:31.370 --> 01:10:34.030 align:middle line:84% So if Ben's willing to pay $3 to get protected 01:10:34.030 --> 01:10:36.010 align:middle line:84% and Jerry's willing to pay $4 to get protected, 01:10:36.010 --> 01:10:37.730 align:middle line:90% society is willing to pay $6. 01:10:37.730 --> 01:10:40.710 align:middle line:84% Remember, society is the sum of these two guys. 01:10:40.710 --> 01:10:45.270 align:middle line:84% OK, so that vertical sum shows the social marginal benefit 01:10:45.270 --> 01:10:50.430 align:middle line:84% for missiles is the vertical sum of these lines. 01:10:50.430 --> 01:10:52.390 align:middle line:84% Then if there's some social marginal cost 01:10:52.390 --> 01:10:56.030 align:middle line:84% curve of producing missiles, you get an equilibrium, 01:10:56.030 --> 01:10:59.230 align:middle line:84% in this case, which is five missiles we 01:10:59.230 --> 01:11:01.910 align:middle line:90% produce at a price of $3. 01:11:01.910 --> 01:11:04.470 align:middle line:90% Questions about that? 01:11:04.470 --> 01:11:06.950 align:middle line:90% Let's do this mathematically. 01:11:06.950 --> 01:11:09.590 align:middle line:84% Mathematically, the difference is 01:11:09.590 --> 01:11:13.750 align:middle line:84% now, what is the marginal-- what is the slope? 01:11:13.750 --> 01:11:16.490 align:middle line:84% What is the marginal rate of substitution for society? 01:11:16.490 --> 01:11:18.310 align:middle line:84% What's the social marginal benefit? 01:11:18.310 --> 01:11:22.280 align:middle line:84% Well, it's the sum of Ben and Jerry's marginal rates 01:11:22.280 --> 01:11:24.160 align:middle line:90% of substitution. 01:11:24.160 --> 01:11:28.920 align:middle line:84% So now the social marginal benefit of missiles 01:11:28.920 --> 01:11:34.040 align:middle line:84% is equal to the marginal rate of substitution 01:11:34.040 --> 01:11:39.160 align:middle line:84% of Ben plus the marginal rate of substitution of Jerry. 01:11:39.160 --> 01:11:42.560 align:middle line:84% That's the social marginal benefit of missiles. 01:11:42.560 --> 01:11:45.320 align:middle line:84% But the social marginal cost is just the same 01:11:45.320 --> 01:11:47.400 align:middle line:90% as it was in the private market. 01:11:47.400 --> 01:11:49.480 align:middle line:84% It's the marginal costs of producing a missile. 01:11:49.480 --> 01:11:52.040 align:middle line:84% Remember, we're seeing marginal cost of cookies is $1. 01:11:52.040 --> 01:11:56.820 align:middle line:84% So that is equal to the marginal cost of producing a missile. 01:11:56.820 --> 01:12:01.480 align:middle line:90% 01:12:01.480 --> 01:12:04.400 align:middle line:84% So look at what the difference is here. 01:12:04.400 --> 01:12:04.960 align:middle line:90% OK? 01:12:04.960 --> 01:12:06.600 align:middle line:90% Look at the difference. 01:12:06.600 --> 01:12:09.360 align:middle line:84% It's equal to the marginal cost of producing a missile. 01:12:09.360 --> 01:12:11.040 align:middle line:84% It's the marginal rate of substitution 01:12:11.040 --> 01:12:12.268 align:middle line:90% of missiles for cookies. 01:12:12.268 --> 01:12:14.560 align:middle line:84% And this is the marginal rate of substitution for Jerry 01:12:14.560 --> 01:12:17.040 align:middle line:90% of missiles for cookies. 01:12:17.040 --> 01:12:19.840 align:middle line:90% Note the key difference. 01:12:19.840 --> 01:12:21.520 align:middle line:90% OK? 01:12:21.520 --> 01:12:25.880 align:middle line:84% Here, we said each individual's MRS gets 01:12:25.880 --> 01:12:28.600 align:middle line:90% set equal to the marginal cost. 01:12:28.600 --> 01:12:31.560 align:middle line:84% Here, it's the sum of all MRSes which 01:12:31.560 --> 01:12:34.320 align:middle line:90% is equal to the marginal cost. 01:12:34.320 --> 01:12:37.120 align:middle line:84% Subtle, but part of why Paul Samuelson won the Nobel 01:12:37.120 --> 01:12:39.560 align:middle line:90% Prize for proving this in 1953. 01:12:39.560 --> 01:12:43.640 align:middle line:84% That basically, the difference with the public good is since 01:12:43.640 --> 01:12:46.080 align:middle line:90% consumption is non-rival-- 01:12:46.080 --> 01:12:48.640 align:middle line:84% since multiple people will benefit for the same good, 01:12:48.640 --> 01:12:51.560 align:middle line:90% we want to produce more of it. 01:12:51.560 --> 01:12:54.240 align:middle line:84% So as a result, we want to account for the fact 01:12:54.240 --> 01:12:56.880 align:middle line:84% that since everyone benefits, we're 01:12:56.880 --> 01:12:59.280 align:middle line:84% going to sum up everybody's marginal benefit 01:12:59.280 --> 01:13:01.120 align:middle line:84% to get the social benefit and set 01:13:01.120 --> 01:13:03.320 align:middle line:90% that equal the marginal cost. 01:13:03.320 --> 01:13:04.640 align:middle line:90% That's the difference. 01:13:04.640 --> 01:13:07.840 align:middle line:84% With ice cream, I only want to do it 01:13:07.840 --> 01:13:09.790 align:middle line:84% till my benefit is equal to the cost, 01:13:09.790 --> 01:13:11.540 align:middle line:84% because only I benefit from the ice cream. 01:13:11.540 --> 01:13:12.290 align:middle line:90% You don't get any. 01:13:12.290 --> 01:13:13.480 align:middle line:90% I ate it. 01:13:13.480 --> 01:13:15.142 align:middle line:90% With a missile, we both benefit. 01:13:15.142 --> 01:13:17.350 align:middle line:84% So we want to wait until the sum of both our benefits 01:13:17.350 --> 01:13:18.837 align:middle line:90% are equal to marginal cost. 01:13:18.837 --> 01:13:21.170 align:middle line:84% And that's the key difference between public and private 01:13:21.170 --> 01:13:25.010 align:middle line:84% goods, is the private good you produce until an individual's 01:13:25.010 --> 01:13:28.330 align:middle line:90% benefit equals society's cost. 01:13:28.330 --> 01:13:29.890 align:middle line:90% And every individual's the same. 01:13:29.890 --> 01:13:32.210 align:middle line:84% Here, produce till the sum of society's benefits 01:13:32.210 --> 01:13:34.290 align:middle line:90% equal society's costs. 01:13:34.290 --> 01:13:36.370 align:middle line:90% OK, questions about that? 01:13:36.370 --> 01:13:38.990 align:middle line:84% A little mind blowing, so take your time on that one. 01:13:38.990 --> 01:13:39.672 align:middle line:90% Yeah. 01:13:39.672 --> 01:13:41.130 align:middle line:84% AUDIENCE: Are there any assumptions 01:13:41.130 --> 01:13:44.318 align:middle line:84% that we're making in this, or is this just like-- 01:13:44.318 --> 01:13:46.110 align:middle line:84% JON GRUBER: This is a pretty general proof. 01:13:46.110 --> 01:13:48.290 align:middle line:84% Like, this is a general case where 01:13:48.290 --> 01:13:52.305 align:middle line:84% if you have a good that's benefiting many people, 01:13:52.305 --> 01:13:54.930 align:middle line:84% you're going to want to account for all those people's benefits 01:13:54.930 --> 01:13:56.670 align:middle line:84% in deciding how much of it to produce. 01:13:56.670 --> 01:13:58.090 align:middle line:90% That's really the insight. 01:13:58.090 --> 01:14:00.050 align:middle line:90% Yeah. 01:14:00.050 --> 01:14:02.250 align:middle line:90% Another question? 01:14:02.250 --> 01:14:02.830 align:middle line:90% Steven. 01:14:02.830 --> 01:14:04.955 align:middle line:84% AUDIENCE: Yeah, in what way is the second different 01:14:04.955 --> 01:14:10.010 align:middle line:84% from saying, OK, if Ben's optimal amount of ice cream 01:14:10.010 --> 01:14:13.250 align:middle line:84% is x and Ben Jerry's is y, the sum of those 01:14:13.250 --> 01:14:15.130 align:middle line:84% is all the ice cream that's produced? 01:14:15.130 --> 01:14:18.690 align:middle line:84% Why is that quantity different from the quantity that you would 01:14:18.690 --> 01:14:20.555 align:middle line:84% get when you're just like, we know 01:14:20.555 --> 01:14:21.930 align:middle line:84% they both want a lot of ice cream 01:14:21.930 --> 01:14:24.290 align:middle line:84% and it should add up to something. 01:14:24.290 --> 01:14:26.610 align:middle line:84% JON GRUBER: But the difference is if they both-- that's 01:14:26.610 --> 01:14:27.430 align:middle line:90% absolutely right. 01:14:27.430 --> 01:14:30.930 align:middle line:84% Their total demand will depend on how much they want ice cream. 01:14:30.930 --> 01:14:37.250 align:middle line:84% The difference is that if you raise the cost of ice cream, 01:14:37.250 --> 01:14:39.915 align:middle line:84% that's going to affect each of them by a certain amount. 01:14:39.915 --> 01:14:42.290 align:middle line:84% And you only care about how much it affects each of them. 01:14:42.290 --> 01:14:45.290 align:middle line:84% Here it's going to matter how it affects their total value 01:14:45.290 --> 01:14:46.068 align:middle line:90% of the missiles. 01:14:46.068 --> 01:14:47.110 align:middle line:90% It's a little bit subtle. 01:14:47.110 --> 01:14:48.902 align:middle line:84% We'll actually come to this more next time. 01:14:48.902 --> 01:14:51.890 align:middle line:84% May become clearer when we talk about optimality. 01:14:51.890 --> 01:14:54.168 align:middle line:90% But other questions about that? 01:14:54.168 --> 01:14:56.210 align:middle line:84% OK, why don't I stop there rather than giving you 01:14:56.210 --> 01:14:57.270 align:middle line:90% the next section? 01:14:57.270 --> 01:15:00.490 align:middle line:84% So we'll come back next time and talk about, 01:15:00.490 --> 01:15:03.910 align:middle line:84% can the private sector provide public goods efficiently? 01:15:03.910 --> 01:15:05.890 align:middle line:84% And if not, what can the public sector do? 01:15:05.890 --> 01:15:08.220 align:middle line:90% And I think you know the answer. 01:15:08.220 --> 01:15:15.000 align:middle line:90%