WEBVTT 00:00:00.000 --> 00:00:00.997 align:middle line:90% 00:00:00.997 --> 00:00:02.488 align:middle line:90% [SQUEAKING] 00:00:02.488 --> 00:00:03.482 align:middle line:90% [RUSTLING] 00:00:03.482 --> 00:00:09.960 align:middle line:90% [CLICKING] 00:00:09.960 --> 00:00:13.160 align:middle line:84% JON GRUBER: We are now going to shift gears 00:00:13.160 --> 00:00:15.190 align:middle line:84% to the next section of the course. 00:00:15.190 --> 00:00:17.440 align:middle line:84% So really, as I said, the course divided fundamentally 00:00:17.440 --> 00:00:18.800 align:middle line:90% into three sections. 00:00:18.800 --> 00:00:21.040 align:middle line:84% We have externalities and public goods. 00:00:21.040 --> 00:00:23.080 align:middle line:84% That's what we've covered so far. 00:00:23.080 --> 00:00:26.120 align:middle line:84% We then talk about social insurance for the next 00:00:26.120 --> 00:00:28.600 align:middle line:90% probably eight lectures. 00:00:28.600 --> 00:00:30.840 align:middle line:90% And then we turn to taxation. 00:00:30.840 --> 00:00:32.700 align:middle line:90% Now, what is social insurance? 00:00:32.700 --> 00:00:36.240 align:middle line:84% Well, let's start by just looking at figure 12-1, 00:00:36.240 --> 00:00:38.380 align:middle line:84% and reminding-- it's the figure we saw earlier, 00:00:38.380 --> 00:00:39.513 align:middle line:90% just recast a little bit. 00:00:39.513 --> 00:00:41.680 align:middle line:84% Reminding ourselves of how the nature of what the US 00:00:41.680 --> 00:00:43.200 align:middle line:90% government does has changed. 00:00:43.200 --> 00:00:48.040 align:middle line:84% In 1953, the government was mostly defense. 00:00:48.040 --> 00:00:50.880 align:middle line:84% There was a very small Social Security program 00:00:50.880 --> 00:00:55.120 align:middle line:84% that took up 3.5%, 3.6% of our government spending, 00:00:55.120 --> 00:00:56.880 align:middle line:84% and the government virtually did nothing 00:00:56.880 --> 00:01:00.540 align:middle line:84% in the realm of health, 0.5% of government spending. 00:01:00.540 --> 00:01:06.260 align:middle line:84% By 2019, defense had shrunk to 15%, Social Security to 18%, 00:01:06.260 --> 00:01:08.180 align:middle line:90% and health to 24%. 00:01:08.180 --> 00:01:11.520 align:middle line:84% This has been a massive shift in what the US government does. 00:01:11.520 --> 00:01:14.780 align:middle line:84% As Paul Krugman wrote, "The US government is essentially now 00:01:14.780 --> 00:01:18.037 align:middle line:84% a giant pension fund with a small army attached." 00:01:18.037 --> 00:01:20.620 align:middle line:84% Basically, we have gone from a government that's fundamentally 00:01:20.620 --> 00:01:23.900 align:middle line:84% responsibilities defending us to fundamental responsibility 00:01:23.900 --> 00:01:25.700 align:middle line:90% of insuring us. 00:01:25.700 --> 00:01:26.260 align:middle line:90% Why? 00:01:26.260 --> 00:01:27.620 align:middle line:90% What does that mean? 00:01:27.620 --> 00:01:31.420 align:middle line:84% And what are all these big programs that are so expensive? 00:01:31.420 --> 00:01:33.700 align:middle line:84% Well, these are a class of what we're 00:01:33.700 --> 00:01:37.220 align:middle line:84% going to call social insurance programs. 00:01:37.220 --> 00:01:40.520 align:middle line:84% Social insurance programs are going to be programs-- 00:01:40.520 --> 00:01:47.860 align:middle line:84% and that's what we're going to cover in the next few lectures, 00:01:47.860 --> 00:01:49.360 align:middle line:90% are social insurance programs. 00:01:49.360 --> 00:01:56.780 align:middle line:84% These are a class of programs that share several key features. 00:01:56.780 --> 00:01:59.600 align:middle line:90% The first feature is workers-- 00:01:59.600 --> 00:02:02.280 align:middle line:84% they're restricted to workers who 00:02:02.280 --> 00:02:04.600 align:middle line:84% participate by essentially making 00:02:04.600 --> 00:02:07.560 align:middle line:90% involuntary contributions. 00:02:07.560 --> 00:02:09.660 align:middle line:84% They can be taxes, they can be insurance premiums, 00:02:09.660 --> 00:02:11.327 align:middle line:84% but the point is, they're not voluntary. 00:02:11.327 --> 00:02:22.720 align:middle line:84% So basically, involuntary worker contributions is one feature. 00:02:22.720 --> 00:02:28.960 align:middle line:84% The second feature is these programs are not means-tested. 00:02:28.960 --> 00:02:30.040 align:middle line:90% What does that mean? 00:02:30.040 --> 00:02:34.680 align:middle line:84% That means that they're universal. 00:02:34.680 --> 00:02:37.280 align:middle line:90% Everyone gets them. 00:02:37.280 --> 00:02:40.540 align:middle line:84% This is not something where only poor people get them. 00:02:40.540 --> 00:02:43.040 align:middle line:84% Now the benefits-- the nature of the program 00:02:43.040 --> 00:02:45.760 align:middle line:84% can differ by income, but everyone's entitled. 00:02:45.760 --> 00:02:48.200 align:middle line:90% These are universal programs. 00:02:48.200 --> 00:02:56.590 align:middle line:84% And the third point is that benefit's payments are 00:02:56.590 --> 00:02:58.650 align:middle line:90% triggered by an adverse event. 00:02:58.650 --> 00:03:03.870 align:middle line:90% 00:03:03.870 --> 00:03:07.890 align:middle line:84% So for example, a classic example 00:03:07.890 --> 00:03:10.630 align:middle line:84% of a social insurance program is unemployment insurance. 00:03:10.630 --> 00:03:12.050 align:middle line:90% We'll talk about that a lot. 00:03:12.050 --> 00:03:14.612 align:middle line:84% Some today, a lot in a couple lectures. 00:03:14.612 --> 00:03:16.070 align:middle line:84% Unemployment insurance is a program 00:03:16.070 --> 00:03:19.790 align:middle line:84% where everybody-- actually, their employers on their behalf, 00:03:19.790 --> 00:03:22.910 align:middle line:84% pays a tax to the government, everyone 00:03:22.910 --> 00:03:26.230 align:middle line:84% who loses their job qualifies, and you 00:03:26.230 --> 00:03:27.650 align:middle line:90% qualify by losing your job. 00:03:27.650 --> 00:03:29.390 align:middle line:90% That's the adverse event. 00:03:29.390 --> 00:03:32.590 align:middle line:84% So everyone's entitled to qualify by losing your job. 00:03:32.590 --> 00:03:36.190 align:middle line:84% Another example is Social Security, the nation's largest 00:03:36.190 --> 00:03:38.870 align:middle line:90% single social insurance program. 00:03:38.870 --> 00:03:41.430 align:middle line:84% That's a program where we pay-- you know that line 00:03:41.430 --> 00:03:43.750 align:middle line:90% FICA on your pay stub? 00:03:43.750 --> 00:03:47.430 align:middle line:84% That's the Social Security contribution. 00:03:47.430 --> 00:03:50.470 align:middle line:84% Everyone gets Social Security, and it's triggered 00:03:50.470 --> 00:03:53.410 align:middle line:90% by retirement or disability. 00:03:53.410 --> 00:03:55.265 align:middle line:84% These are the kinds of programs that 00:03:55.265 --> 00:03:56.890 align:middle line:84% are under the social insurance umbrella 00:03:56.890 --> 00:04:00.490 align:middle line:84% that we'll cover over the next several lectures, 00:04:00.490 --> 00:04:02.750 align:middle line:84% but that's the general idea of social insurance. 00:04:02.750 --> 00:04:05.610 align:middle line:84% This would be distinguished from things 00:04:05.610 --> 00:04:08.450 align:middle line:84% which are redistribution programs, which explicitly are 00:04:08.450 --> 00:04:10.730 align:middle line:84% restricted to some people and not others, 00:04:10.730 --> 00:04:13.130 align:middle line:84% and which are triggered by what your income is, 00:04:13.130 --> 00:04:15.410 align:middle line:90% not what happens to you. 00:04:15.410 --> 00:04:19.730 align:middle line:84% These are really like insurance programs. 00:04:19.730 --> 00:04:22.970 align:middle line:84% Now, to understand why the government spends so much 00:04:22.970 --> 00:04:24.850 align:middle line:84% of its revenue on these programs, 00:04:24.850 --> 00:04:28.610 align:middle line:84% we have to step back and ask, what is insurance, 00:04:28.610 --> 00:04:30.667 align:middle line:84% and why do we want it in the first place? 00:04:30.667 --> 00:04:32.250 align:middle line:84% So let's step back and do a little bit 00:04:32.250 --> 00:04:35.970 align:middle line:84% of basic insurance theory, this is some review of 14.01. 00:04:35.970 --> 00:04:38.730 align:middle line:84% Let's step back and do some basic insurance theory. 00:04:38.730 --> 00:04:41.490 align:middle line:84% Now, first of all, let's stipulate, insurance 00:04:41.490 --> 00:04:43.290 align:middle line:90% is big business in America. 00:04:43.290 --> 00:04:46.330 align:middle line:84% We have a private health insurance industry, 00:04:46.330 --> 00:04:50.710 align:middle line:84% which is currently collects about $1.2 trillion in premiums. 00:04:50.710 --> 00:04:54.235 align:middle line:90% 00:04:54.235 --> 00:04:56.110 align:middle line:84% We have an insurance industry, which collects 00:04:56.110 --> 00:04:59.270 align:middle line:90% about $250 billion in premiums. 00:04:59.270 --> 00:05:03.430 align:middle line:84% We have a life insurance program which collects 00:05:03.430 --> 00:05:05.930 align:middle line:90% about $160 billion in premiums. 00:05:05.930 --> 00:05:08.010 align:middle line:84% And then we have property and casualty insurance, 00:05:08.010 --> 00:05:10.110 align:middle line:84% like if something catches on fire or gets 00:05:10.110 --> 00:05:13.670 align:middle line:84% stolen, that's about $600 billion a year. 00:05:13.670 --> 00:05:16.550 align:middle line:84% So we're, as a nation, spending on the order of $3 trillion-ish 00:05:16.550 --> 00:05:18.970 align:middle line:90% a year to insure ourselves. 00:05:18.970 --> 00:05:21.550 align:middle line:84% The nature of insurance is essentially 00:05:21.550 --> 00:05:24.110 align:middle line:84% you pay a month-- the nature of private insurance 00:05:24.110 --> 00:05:26.630 align:middle line:84% is you pay a regular payment, called 00:05:26.630 --> 00:05:30.190 align:middle line:84% a premium, usually monthly; in return, 00:05:30.190 --> 00:05:33.030 align:middle line:84% that insurance covers you against some adverse event. 00:05:33.030 --> 00:05:35.318 align:middle line:84% So health insurance that covers if we're sick. 00:05:35.318 --> 00:05:37.110 align:middle line:84% Property and casualty insurance to cover us 00:05:37.110 --> 00:05:38.830 align:middle line:90% if our house burns down. 00:05:38.830 --> 00:05:40.530 align:middle line:90% And that's big dollars. 00:05:40.530 --> 00:05:43.030 align:middle line:84% It's trillions of dollars in the US. 00:05:43.030 --> 00:05:47.652 align:middle line:84% Now why do people want insurance that badly? 00:05:47.652 --> 00:05:49.610 align:middle line:84% Why are we, as the economy, is spending so much 00:05:49.610 --> 00:05:51.990 align:middle line:90% of our GDP on insurance? 00:05:51.990 --> 00:05:54.010 align:middle line:90% Why do individuals value it? 00:05:54.010 --> 00:05:56.882 align:middle line:84% And I'll start with the general intuition. 00:05:56.882 --> 00:05:58.590 align:middle line:84% In the book, there's a numerical example, 00:05:58.590 --> 00:06:00.090 align:middle line:90% but I'll cover the math. 00:06:00.090 --> 00:06:03.470 align:middle line:84% The general intuition is quite simple. 00:06:03.470 --> 00:06:07.130 align:middle line:84% Think about-- remember our typical intertemporal choice 00:06:07.130 --> 00:06:10.090 align:middle line:84% problem, which is you can eat today or tomorrow. 00:06:10.090 --> 00:06:13.530 align:middle line:84% Instead, we're going to do what we call an uncertainty 00:06:13.530 --> 00:06:17.110 align:middle line:84% problem, which is there's multiple states of the world. 00:06:17.110 --> 00:06:18.810 align:middle line:84% So it's like a science fiction thing 00:06:18.810 --> 00:06:21.710 align:middle line:84% where basically in the future, one of two things can happen. 00:06:21.710 --> 00:06:24.170 align:middle line:84% You can think of yourself in either state of the world. 00:06:24.170 --> 00:06:26.530 align:middle line:84% One state of the world is the bad thing doesn't happen, 00:06:26.530 --> 00:06:29.210 align:middle line:84% one state of the world is a bad thing does happen. 00:06:29.210 --> 00:06:31.770 align:middle line:84% And from today's perspective, you 00:06:31.770 --> 00:06:33.910 align:middle line:84% can't know with certainty if it's going to happen. 00:06:33.910 --> 00:06:35.730 align:middle line:90% You can know the probability-- 00:06:35.730 --> 00:06:38.310 align:middle line:84% or you may know the probability of the bad things can happen, 00:06:38.310 --> 00:06:40.393 align:middle line:84% but you don't know, from your current perspective, 00:06:40.393 --> 00:06:42.370 align:middle line:90% if it's going to happen or not. 00:06:42.370 --> 00:06:46.290 align:middle line:84% So you have this uncertain risk you face with some probability. 00:06:46.290 --> 00:06:51.410 align:middle line:84% And then if that risk happens, if something bad happens to you, 00:06:51.410 --> 00:06:53.550 align:middle line:84% you end up in a worse state than if something good 00:06:53.550 --> 00:06:54.830 align:middle line:90% happens to you. 00:06:54.830 --> 00:06:57.930 align:middle line:84% And the basic principle that we learned in 14.01, 00:06:57.930 --> 00:07:00.630 align:middle line:84% we talked about uncertainty, is that individuals 00:07:00.630 --> 00:07:03.590 align:middle line:90% want consumption smoothing. 00:07:03.590 --> 00:07:08.070 align:middle line:84% That because our utility functions 00:07:08.070 --> 00:07:11.190 align:middle line:84% feature diminishing marginal utility, 00:07:11.190 --> 00:07:13.220 align:middle line:84% we're always happier having the same consumption 00:07:13.220 --> 00:07:15.470 align:middle line:84% in different states in the world rather than different 00:07:15.470 --> 00:07:16.870 align:middle line:90% consumption. 00:07:16.870 --> 00:07:19.350 align:middle line:84% Because remember, with diminishing marginal utility, 00:07:19.350 --> 00:07:22.510 align:middle line:84% the extra dollar makes us less happy than one 00:07:22.510 --> 00:07:27.107 align:middle line:84% less dollar makes us sad because it's diminishing. 00:07:27.107 --> 00:07:29.690 align:middle line:84% So if you're at a certain point and your income goes up a lot, 00:07:29.690 --> 00:07:31.750 align:middle line:84% you're happier, but you're not nearly as happy 00:07:31.750 --> 00:07:33.375 align:middle line:84% as you are sad if your income goes down 00:07:33.375 --> 00:07:37.250 align:middle line:84% a lot because of diminishing marginal utility. 00:07:37.250 --> 00:07:39.590 align:middle line:84% Think of the limit, if everyone goes down all the way 00:07:39.590 --> 00:07:42.630 align:middle line:84% to 0, that's much worse, whereas how happy 00:07:42.630 --> 00:07:43.630 align:middle line:90% are the billion dollars? 00:07:43.630 --> 00:07:46.570 align:middle line:90% It's still limited. 00:07:46.570 --> 00:07:50.650 align:middle line:84% So basically, essentially the idea is-- 00:07:50.650 --> 00:07:52.330 align:middle line:84% that's about it, forget that intuition, 00:07:52.330 --> 00:07:53.538 align:middle line:90% I'll get to math in a second. 00:07:53.538 --> 00:07:56.530 align:middle line:84% The bottom line is, just focus on the curvature, which 00:07:56.530 --> 00:07:59.250 align:middle line:84% is, remember, that because of the curvature of the utility 00:07:59.250 --> 00:08:01.970 align:middle line:84% function, because diminishing marginal utility, 00:08:01.970 --> 00:08:04.890 align:middle line:84% you would want to make sure consumption's smooth 00:08:04.890 --> 00:08:07.130 align:middle line:90% across states of the world. 00:08:07.130 --> 00:08:11.207 align:middle line:84% That's the basic principle that drives our demand for insurance. 00:08:11.207 --> 00:08:13.290 align:middle line:84% So to understand that, there's a numerical example 00:08:13.290 --> 00:08:15.123 align:middle line:84% in the chapter, but let's go to the appendix 00:08:15.123 --> 00:08:16.650 align:middle line:90% and actually do the math. 00:08:16.650 --> 00:08:20.630 align:middle line:84% So, let's imagine the following model. 00:08:20.630 --> 00:08:25.930 align:middle line:84% Let's imagine you are hit by a car with some probability-- 00:08:25.930 --> 00:08:28.570 align:middle line:90% here, let's set it up here. 00:08:28.570 --> 00:08:33.610 align:middle line:84% There's some probability P that an adverse event happens. 00:08:33.610 --> 00:08:34.690 align:middle line:90% Say you're hit by a car. 00:08:34.690 --> 00:08:37.357 align:middle line:90% 00:08:37.357 --> 00:08:39.690 align:middle line:84% Regardless of whether you get hit by or not, hit or not, 00:08:39.690 --> 00:08:44.080 align:middle line:84% your income is W. So P is the risk getting hit by a car, 00:08:44.080 --> 00:08:48.400 align:middle line:84% W is your income regardless, you still go to work either way. 00:08:48.400 --> 00:08:52.280 align:middle line:84% But if you get hit by the car, you incur medical costs delta. 00:08:52.280 --> 00:08:56.760 align:middle line:90% 00:08:56.760 --> 00:09:03.840 align:middle line:84% You can buy insurance for a premium of M per dollar. 00:09:03.840 --> 00:09:07.240 align:middle line:84% So M is the number of cents it costs you to buy 00:09:07.240 --> 00:09:10.960 align:middle line:90% a dollar's worth of insurance. 00:09:10.960 --> 00:09:16.280 align:middle line:84% That insurance you buy pays out a benefit b. 00:09:16.280 --> 00:09:20.640 align:middle line:84% Pays out a benefit b if you get hit. 00:09:20.640 --> 00:09:23.280 align:middle line:84% b is the amount of benefit it pays out. 00:09:23.280 --> 00:09:26.523 align:middle line:84% So once again, in this world, you only risk hit by a car. 00:09:26.523 --> 00:09:28.440 align:middle line:84% You're a young, healthy man-- and I choose man 00:09:28.440 --> 00:09:30.240 align:middle line:84% because there's no risk of pregnancy. 00:09:30.240 --> 00:09:32.000 align:middle line:90% You're a young, healthy man. 00:09:32.000 --> 00:09:35.920 align:middle line:84% The only thing that can go wrong is you getting hit by a car. 00:09:35.920 --> 00:09:37.420 align:middle line:90% This is pre-COVID. 00:09:37.420 --> 00:09:40.280 align:middle line:84% Things go wrong, you get hit by a car. 00:09:40.280 --> 00:09:42.340 align:middle line:84% There's a P probability hit by a car. 00:09:42.340 --> 00:09:44.480 align:middle line:84% Your income's W regardless of whether you get hit. 00:09:44.480 --> 00:09:46.003 align:middle line:90% There's a d chance you get hit. 00:09:46.003 --> 00:09:47.420 align:middle line:84% You can buy insurance at a premium 00:09:47.420 --> 00:09:51.900 align:middle line:84% of M per dollar of insurance and it pays you a benefit b. 00:09:51.900 --> 00:09:57.260 align:middle line:84% Now, we write down your expected utility. 00:09:57.260 --> 00:09:59.380 align:middle line:90% What is your expected utility? 00:09:59.380 --> 00:10:05.660 align:middle line:84% Well, your expected utility is 1 minus P, the chance that you 00:10:05.660 --> 00:10:12.420 align:middle line:84% don't hit by a car, times the utility of W 00:10:12.420 --> 00:10:16.180 align:middle line:84% minus the insurance-- what you pay for insurance, Mb. 00:10:16.180 --> 00:10:21.160 align:middle line:90% 00:10:21.160 --> 00:10:24.380 align:middle line:84% That's what you pay for insurance, is Mb. 00:10:24.380 --> 00:10:25.860 align:middle line:84% So if you want to benefit $1,000, 00:10:25.860 --> 00:10:31.860 align:middle line:84% you pay 1,000 in because M is per dollar. 00:10:31.860 --> 00:10:35.900 align:middle line:84% If you get hit by a car, there's a probability 00:10:35.900 --> 00:10:42.600 align:middle line:84% P times the utility of-- you still have income W. You still 00:10:42.600 --> 00:10:44.280 align:middle line:90% pay the insurance premium. 00:10:44.280 --> 00:10:45.980 align:middle line:84% You pay insurance no matter what. 00:10:45.980 --> 00:10:51.520 align:middle line:84% So you still pay Mb, but you also 00:10:51.520 --> 00:10:58.540 align:middle line:84% lose delta because you have to pay that medical cost. 00:10:58.540 --> 00:11:02.600 align:middle line:84% But you get b, which is the benefit because they 00:11:02.600 --> 00:11:04.360 align:middle line:90% pay for your medical costs. 00:11:04.360 --> 00:11:08.080 align:middle line:90% They pay some amount b. 00:11:08.080 --> 00:11:12.960 align:middle line:84% So, basically that is your that's your setup. 00:11:12.960 --> 00:11:14.600 align:middle line:90% Questions about that? 00:11:14.600 --> 00:11:15.540 align:middle line:90% Yeah? 00:11:15.540 --> 00:11:16.960 align:middle line:90% AUDIENCE: [INAUDIBLE] 00:11:16.960 --> 00:11:17.880 align:middle line:90% JON GRUBER: Yes. 00:11:17.880 --> 00:11:19.640 align:middle line:84% Because M per dollar of insurance, 00:11:19.640 --> 00:11:22.240 align:middle line:90% you buy b dollars of insurance. 00:11:22.240 --> 00:11:23.980 align:middle line:84% And b is how much you pay for insurance. 00:11:23.980 --> 00:11:26.640 align:middle line:90% 00:11:26.640 --> 00:11:29.160 align:middle line:84% Now, we are going to want to optimize this. 00:11:29.160 --> 00:11:32.800 align:middle line:84% But if you'll notice, there is two unknowns here. 00:11:32.800 --> 00:11:34.640 align:middle line:84% We're going to stipulate W-- we'll tell you 00:11:34.640 --> 00:11:36.800 align:middle line:84% what w is, we'll tell you what P is, 00:11:36.800 --> 00:11:43.060 align:middle line:84% we'll tell you what delta is, but you have to decide M and b. 00:11:43.060 --> 00:11:46.300 align:middle line:84% So we need a second equation to close this system. 00:11:46.300 --> 00:11:49.100 align:middle line:84% And the second equation we're going to get to close the system 00:11:49.100 --> 00:11:53.260 align:middle line:90% is the supply side of insurance. 00:11:53.260 --> 00:11:56.180 align:middle line:84% And we're going to get the price of insurance 00:11:56.180 --> 00:11:58.960 align:middle line:84% by assuming a zero-profit insurer. 00:11:58.960 --> 00:12:01.798 align:middle line:84% We're going to assume a zero-profit in expectation. 00:12:01.798 --> 00:12:03.340 align:middle line:84% We're going to assume the insurer has 00:12:03.340 --> 00:12:05.460 align:middle line:90% zero expected profits. 00:12:05.460 --> 00:12:06.680 align:middle line:90% What does that mean? 00:12:06.680 --> 00:12:09.780 align:middle line:84% Well, the insurer's expected profits 00:12:09.780 --> 00:12:15.660 align:middle line:84% are what it collects in premiums, which is Mb. 00:12:15.660 --> 00:12:17.940 align:middle line:84% That's what it collects in premiums. 00:12:17.940 --> 00:12:23.940 align:middle line:84% Minus what it expects to pay out, which is Pb. 00:12:23.940 --> 00:12:28.180 align:middle line:84% P percent of the time, they'll pay out b. 00:12:28.180 --> 00:12:31.100 align:middle line:84% Well, what you can see from this is that zero profits 00:12:31.100 --> 00:12:35.760 align:middle line:90% implies M equals b. 00:12:35.760 --> 00:12:39.880 align:middle line:84% That's the implication of the zero-profit condition. 00:12:39.880 --> 00:12:42.740 align:middle line:84% That basically-- I'm sorry, M equals P, my bad. 00:12:42.740 --> 00:12:49.080 align:middle line:84% Sorry, M equals P. M equals P. That the zero-profit condition 00:12:49.080 --> 00:12:50.680 align:middle line:84% is the rate per dollar charged to you 00:12:50.680 --> 00:12:53.440 align:middle line:84% is equal to the probability event of the event. 00:12:53.440 --> 00:12:55.020 align:middle line:90% And it makes sense. 00:12:55.020 --> 00:12:59.320 align:middle line:84% If it's a 1 in 100 event, then by charging a cent per dollar 00:12:59.320 --> 00:13:02.730 align:middle line:90% of insurance, they break even. 00:13:02.730 --> 00:13:03.980 align:middle line:90% So that's how they break even. 00:13:03.980 --> 00:13:06.920 align:middle line:84% They basically say a rare event, you pay less; a lesser event, 00:13:06.920 --> 00:13:08.600 align:middle line:90% you pay more. 00:13:08.600 --> 00:13:12.840 align:middle line:84% So that's what delivers zero profits for the insurer. 00:13:12.840 --> 00:13:17.420 align:middle line:84% Now, with that in mind, let's take a typical assumption. 00:13:17.420 --> 00:13:19.640 align:middle line:84% You assume utility equals square root 00:13:19.640 --> 00:13:24.160 align:middle line:84% of C, which is our typical risk-averse diminishing 00:13:24.160 --> 00:13:26.960 align:middle line:84% marginal utility utility function. 00:13:26.960 --> 00:13:28.920 align:middle line:84% And let's plug in and write this down. 00:13:28.920 --> 00:13:32.320 align:middle line:84% So we say expected utility now equals 00:13:32.320 --> 00:13:40.140 align:middle line:84% 1 minus P times the square root of W minus Mb, which 00:13:40.140 --> 00:13:45.660 align:middle line:84% we can rewrite as W minus bP, because M equals-- 00:13:45.660 --> 00:13:46.280 align:middle line:90% I'm sorry. 00:13:46.280 --> 00:13:49.510 align:middle line:84% We can rewrite as W minus bP, right. 00:13:49.510 --> 00:13:53.770 align:middle line:84% So we can rewrite that as W minus bP because Mb, 00:13:53.770 --> 00:13:58.540 align:middle line:84% b is b and M is P. So just rewrote that. 00:13:58.540 --> 00:14:00.700 align:middle line:84% And then for the second part of the equation, 00:14:00.700 --> 00:14:09.940 align:middle line:84% we have P plus P times the square root of W minus 00:14:09.940 --> 00:14:18.200 align:middle line:90% minus bP minus delta plus b. 00:14:18.200 --> 00:14:19.680 align:middle line:84% So all I did here I did two things. 00:14:19.680 --> 00:14:21.940 align:middle line:84% I plugged in from the zero-profit condition, 00:14:21.940 --> 00:14:25.900 align:middle line:84% and I assumed a functional form for utility. 00:14:25.900 --> 00:14:30.980 align:middle line:84% We can maximize this with respect to b. 00:14:30.980 --> 00:14:33.330 align:middle line:84% So let's maximize this equation with respect 00:14:33.330 --> 00:14:35.370 align:middle line:84% to b, you want to choose the optimal benefits. 00:14:35.370 --> 00:14:36.745 align:middle line:84% By choosing the optimal benefits, 00:14:36.745 --> 00:14:38.810 align:middle line:84% you choose how much you pay in premiums. 00:14:38.810 --> 00:14:41.250 align:middle line:84% So basically, you just want to choose 00:14:41.250 --> 00:14:44.090 align:middle line:84% the optimal-- you just want to choose the optimal benefits. 00:14:44.090 --> 00:14:48.670 align:middle line:90% And that condition is if you-- 00:14:48.670 --> 00:14:51.370 align:middle line:84% I want to ask, what's the optimal benefits? 00:14:51.370 --> 00:14:54.050 align:middle line:84% Well basically, you're going to get-- 00:14:54.050 --> 00:14:56.930 align:middle line:84% if you take the derivative of this with respect to b, 00:14:56.930 --> 00:15:08.410 align:middle line:84% you're going to get minus 1 minus P times P over-- 00:15:08.410 --> 00:15:11.170 align:middle line:90% 1 minus b of P, 1 minus-- 00:15:11.170 --> 00:15:16.150 align:middle line:84% 1 minus P times P over square root of W minus bP. 00:15:16.150 --> 00:15:19.650 align:middle line:90% 00:15:19.650 --> 00:15:36.070 align:middle line:84% Plus P times 1 minus P over the square root of W minus bP 00:15:36.070 --> 00:15:38.030 align:middle line:90% minus delta plus b. 00:15:38.030 --> 00:15:40.030 align:middle line:84% That's what you get when you take the derivative 00:15:40.030 --> 00:15:41.870 align:middle line:90% of this equation. 00:15:41.870 --> 00:15:46.750 align:middle line:84% If you set that equal to 0 and solve, 00:15:46.750 --> 00:15:54.730 align:middle line:84% you get that b star equals delta. 00:15:54.730 --> 00:15:57.990 align:middle line:84% That the optimal level of benefits 00:15:57.990 --> 00:16:04.870 align:middle line:84% is equal to the medical cost of the adverse event. 00:16:04.870 --> 00:16:07.310 align:middle line:90% You can do the math to see that. 00:16:07.310 --> 00:16:10.590 align:middle line:84% You can also cheat and say, look, 00:16:10.590 --> 00:16:14.950 align:middle line:84% we know that optimal insurance involves consumption smoothing. 00:16:14.950 --> 00:16:18.110 align:middle line:84% That involves these two terms being the same. 00:16:18.110 --> 00:16:20.330 align:middle line:84% If these two terms are different, it's not optimal. 00:16:20.330 --> 00:16:20.530 align:middle line:90% Why? 00:16:20.530 --> 00:16:21.710 align:middle line:84% Because then they have a different utility 00:16:21.710 --> 00:16:23.085 align:middle line:84% in different states of the world. 00:16:23.085 --> 00:16:25.950 align:middle line:84% They want the same utility in both states of the world. 00:16:25.950 --> 00:16:31.312 align:middle line:84% What sets these two equations equal? b star equals delta. 00:16:31.312 --> 00:16:33.270 align:middle line:84% So you can take the derivative and do the math, 00:16:33.270 --> 00:16:36.033 align:middle line:84% or you can just remember the intuition that the alpha comes 00:16:36.033 --> 00:16:36.950 align:middle line:90% consumption smoothing. 00:16:36.950 --> 00:16:40.570 align:middle line:84% Consumption smoothing implies the term in this-- 00:16:40.570 --> 00:16:42.930 align:middle line:84% here in this term equals this term. 00:16:42.930 --> 00:16:46.890 align:middle line:84% And that is only true when b star equals delta. 00:16:46.890 --> 00:16:51.570 align:middle line:84% So consumption smoothing requires full insurance. 00:16:51.570 --> 00:16:54.730 align:middle line:84% That's the key outcome of insurance theory. 00:16:54.730 --> 00:17:00.810 align:middle line:84% That the first best, the optimum, is full insurance. 00:17:00.810 --> 00:17:02.890 align:middle line:84% Because if you don't have full insurance, 00:17:02.890 --> 00:17:06.050 align:middle line:90% then you're consuming less. 00:17:06.050 --> 00:17:07.500 align:middle line:84% If you're less than fully insured, 00:17:07.500 --> 00:17:09.750 align:middle line:84% you consume less in the bad state than the good state, 00:17:09.750 --> 00:17:10.690 align:middle line:90% you don't like that. 00:17:10.690 --> 00:17:12.630 align:middle line:84% You want to consume the same in every state. 00:17:12.630 --> 00:17:15.609 align:middle line:84% That's the whole principle of consumption smoothing. 00:17:15.609 --> 00:17:16.893 align:middle line:90% Yeah, Alec? 00:17:16.893 --> 00:17:18.810 align:middle line:84% AUDIENCE: This assumes that the insurance firm 00:17:18.810 --> 00:17:20.342 align:middle line:90% has no cost other than payouts? 00:17:20.342 --> 00:17:21.550 align:middle line:90% JON GRUBER: I'll get to that. 00:17:21.550 --> 00:17:26.109 align:middle line:84% Right now, it's a perfectly zero-profit, perfectly efficient 00:17:26.109 --> 00:17:27.750 align:middle line:90% insurer. 00:17:27.750 --> 00:17:28.830 align:middle line:90% Yeah, Stephen? 00:17:28.830 --> 00:17:32.710 align:middle line:84% AUDIENCE: So can you just explain specifically what b is 00:17:32.710 --> 00:17:33.458 align:middle line:90% and what M-- 00:17:33.458 --> 00:17:34.750 align:middle line:90% JON GRUBER: Sure, I understand. 00:17:34.750 --> 00:17:38.830 align:middle line:84% M is what you pay per dollar of insurance. 00:17:38.830 --> 00:17:43.132 align:middle line:84% b is how many dollars of insurance you get. 00:17:43.132 --> 00:17:44.090 align:middle line:90% AUDIENCE: So it's the-- 00:17:44.090 --> 00:17:46.230 align:middle line:84% JON GRUBER: So in other words, it-- 00:17:46.230 --> 00:17:49.150 align:middle line:84% it's the cost-- so, the total amount I pay for insurance 00:17:49.150 --> 00:17:51.110 align:middle line:90% is Mb. 00:17:51.110 --> 00:17:56.948 align:middle line:84% So if I want to buy $1,000 of insurance, I pay M times 1,000. 00:17:56.948 --> 00:17:57.490 align:middle line:90% AUDIENCE: OK. 00:17:57.490 --> 00:17:59.230 align:middle line:84% So b is the amount of coverage that you're receiving? 00:17:59.230 --> 00:18:00.063 align:middle line:90% JON GRUBER: Exactly. 00:18:00.063 --> 00:18:02.030 align:middle line:84% The amount of coverage you're receiving. 00:18:02.030 --> 00:18:03.017 align:middle line:90% AUDIENCE: And M is-- 00:18:03.017 --> 00:18:04.850 align:middle line:84% JON GRUBER: The cost per dollar of coverage. 00:18:04.850 --> 00:18:08.110 align:middle line:90% 00:18:08.110 --> 00:18:11.190 align:middle line:90% All right, other questions? 00:18:11.190 --> 00:18:12.090 align:middle line:90% Yeah? 00:18:12.090 --> 00:18:14.470 align:middle line:84% AUDIENCE: I'm just a little bit lost 00:18:14.470 --> 00:18:17.590 align:middle line:84% how we got the second equation for expected utility 00:18:17.590 --> 00:18:19.320 align:middle line:90% where we just plug in the-- 00:18:19.320 --> 00:18:20.570 align:middle line:90% JON GRUBER: Where we get this? 00:18:20.570 --> 00:18:21.150 align:middle line:90% AUDIENCE: Yeah. 00:18:21.150 --> 00:18:22.817 align:middle line:84% JON GRUBER: So what I did is two things. 00:18:22.817 --> 00:18:24.590 align:middle line:84% One is I replace u with square root. 00:18:24.590 --> 00:18:25.890 align:middle line:90% That's the easy part. 00:18:25.890 --> 00:18:30.250 align:middle line:84% The second is, wherever I had P-- 00:18:30.250 --> 00:18:32.650 align:middle line:84% wherever I had M, I'm sorry, I plugged in P. 00:18:32.650 --> 00:18:34.070 align:middle line:90% So I place that M with a P-- 00:18:34.070 --> 00:18:36.010 align:middle line:84% I flipped the order just to fuck you up. 00:18:36.010 --> 00:18:38.930 align:middle line:84% And then I replaced that M with a P, 00:18:38.930 --> 00:18:40.430 align:middle line:84% and once again, I flipped the order. 00:18:40.430 --> 00:18:44.310 align:middle line:90% 00:18:44.310 --> 00:18:45.890 align:middle line:90% Other questions? 00:18:45.890 --> 00:18:46.710 align:middle line:90% Good questions. 00:18:46.710 --> 00:18:47.330 align:middle line:90% OK. 00:18:47.330 --> 00:18:52.150 align:middle line:84% So, the bottom line is, people want full insurance. 00:18:52.150 --> 00:18:54.490 align:middle line:84% And once again, it's useful to go to the book 00:18:54.490 --> 00:18:55.730 align:middle line:90% and look through the-- 00:18:55.730 --> 00:18:57.473 align:middle line:84% look through the mathematical example-- 00:18:57.473 --> 00:18:58.890 align:middle line:84% look through the numerical example 00:18:58.890 --> 00:19:01.890 align:middle line:84% just so you can confirm the intuition. 00:19:01.890 --> 00:19:09.370 align:middle line:84% Now, one important feature is that by assuming this utility 00:19:09.370 --> 00:19:16.050 align:middle line:84% function, we've assumed a given level of risk aversion. 00:19:16.050 --> 00:19:20.330 align:middle line:84% And remember the key lesson from 14.01. 00:19:20.330 --> 00:19:21.950 align:middle line:84% Risk aversion is about the curvature 00:19:21.950 --> 00:19:24.310 align:middle line:90% of the utility function. 00:19:24.310 --> 00:19:28.470 align:middle line:84% The more rapidly your marginal utility diminishes, 00:19:28.470 --> 00:19:31.470 align:middle line:84% the more risk-averse you are, because the more it rapidly 00:19:31.470 --> 00:19:34.510 align:middle line:84% diminishes, that's how quickly you get sad 00:19:34.510 --> 00:19:36.590 align:middle line:90% when things go south. 00:19:36.590 --> 00:19:41.490 align:middle line:84% So no matter your level of risk aversion, 00:19:41.490 --> 00:19:43.230 align:middle line:84% you'll always want full insurance. 00:19:43.230 --> 00:19:45.950 align:middle line:84% That's true as long as you're at all risk-averse. 00:19:45.950 --> 00:19:49.870 align:middle line:84% In fact, as long as you're not risk-loving. 00:19:49.870 --> 00:19:53.270 align:middle line:84% But how much you care about full insurance 00:19:53.270 --> 00:19:55.430 align:middle line:84% will depend on how risk-averse you are. 00:19:55.430 --> 00:19:59.310 align:middle line:84% If you're risk-neutral, you won't care a lot. 00:19:59.310 --> 00:20:02.350 align:middle line:84% And remember our intuition from 14.01, 00:20:02.350 --> 00:20:05.670 align:middle line:84% that the value of insurance depends 00:20:05.670 --> 00:20:10.870 align:middle line:84% on the size of the risk relative to your income. 00:20:10.870 --> 00:20:13.070 align:middle line:90% Remember, if your income is-- 00:20:13.070 --> 00:20:16.230 align:middle line:84% if it's a $1 risk and you're very rich, 00:20:16.230 --> 00:20:17.830 align:middle line:90% then you're risk-neutral. 00:20:17.830 --> 00:20:19.380 align:middle line:84% The utility function is approximately 00:20:19.380 --> 00:20:22.257 align:middle line:84% flat at that point, so you don't care-- 00:20:22.257 --> 00:20:23.840 align:middle line:84% I mean, you still want full insurance, 00:20:23.840 --> 00:20:25.380 align:middle line:90% but you don't really care. 00:20:25.380 --> 00:20:30.620 align:middle line:84% But if it's a $1,000 risk, it becomes $2,000, you care a lot. 00:20:30.620 --> 00:20:35.020 align:middle line:84% And so you really are willing to pay a lot for insurance. 00:20:35.020 --> 00:20:39.060 align:middle line:84% So that's the role of risk aversion. 00:20:39.060 --> 00:20:39.940 align:middle line:90% Yeah? 00:20:39.940 --> 00:20:42.540 align:middle line:84% AUDIENCE: So which of these statements 00:20:42.540 --> 00:20:45.760 align:middle line:84% depend upon the particular form of utility there? 00:20:45.760 --> 00:20:46.260 align:middle line:90% Or-- 00:20:46.260 --> 00:20:47.560 align:middle line:84% JON GRUBER: None of this is dependent on that. 00:20:47.560 --> 00:20:49.352 align:middle line:84% You always want full insurance, unless this 00:20:49.352 --> 00:20:51.340 align:middle line:84% is a risk-loving utility function, which 00:20:51.340 --> 00:20:53.820 align:middle line:84% violates other norms we have in the-- as long as we have 00:20:53.820 --> 00:20:56.958 align:middle line:84% a diminishing marginal utility consumption, this is all true. 00:20:56.958 --> 00:20:57.500 align:middle line:90% AUDIENCE: OK. 00:20:57.500 --> 00:20:58.092 align:middle line:90% [INAUDIBLE] 00:20:58.092 --> 00:20:58.800 align:middle line:90% JON GRUBER: Yeah. 00:20:58.800 --> 00:21:00.740 align:middle line:84% It's just how much you care depends 00:21:00.740 --> 00:21:03.140 align:middle line:90% on how risk-averse you are. 00:21:03.140 --> 00:21:04.360 align:middle line:90% We'll come back to that. 00:21:04.360 --> 00:21:06.440 align:middle line:90% Now, that's all a setup. 00:21:06.440 --> 00:21:08.882 align:middle line:84% That's a review of 14.01, which sets up, well, 00:21:08.882 --> 00:21:10.840 align:middle line:84% why do we need the government in this business? 00:21:10.840 --> 00:21:12.993 align:middle line:84% We've got a $3 trillion of private insurance 00:21:12.993 --> 00:21:14.660 align:middle line:84% floating around, why do we need, really, 00:21:14.660 --> 00:21:17.220 align:middle line:84% almost another $3 trillion of public insurance 00:21:17.220 --> 00:21:18.520 align:middle line:90% added to the mix? 00:21:18.520 --> 00:21:22.720 align:middle line:84% Literally, if you look at total insurance expenditures 00:21:22.720 --> 00:21:24.860 align:middle line:84% in America, it's about half-public, half-private-- 00:21:24.860 --> 00:21:27.640 align:middle line:84% maybe a little more private, but close to half and half. 00:21:27.640 --> 00:21:28.660 align:middle line:90% Why do we need that? 00:21:28.660 --> 00:21:31.760 align:middle line:84% And the main reason we feel like we need a public sector 00:21:31.760 --> 00:21:34.340 align:middle line:84% intervention is you have to ask yourself, 00:21:34.340 --> 00:21:35.430 align:middle line:90% what's the market failure? 00:21:35.430 --> 00:21:37.680 align:middle line:84% And the main market failure here is the market failure 00:21:37.680 --> 00:21:39.260 align:middle line:90% of asymmetric information. 00:21:39.260 --> 00:21:43.148 align:middle line:90% 00:21:43.148 --> 00:21:45.440 align:middle line:84% That's the main market failure we're dealing with here, 00:21:45.440 --> 00:21:47.800 align:middle line:90% is asymmetric information. 00:21:47.800 --> 00:21:50.840 align:middle line:84% The classic example that goes back to the Akerlof lemons 00:21:50.840 --> 00:21:51.460 align:middle line:90% problem. 00:21:51.460 --> 00:21:54.550 align:middle line:84% George Akerlof won the Nobel Prize for-- 00:21:54.550 --> 00:21:56.800 align:middle line:84% George Akerlof, Nobel Prize-winning economist and half 00:21:56.800 --> 00:21:58.040 align:middle line:84% of the power couple because his wife 00:21:58.040 --> 00:22:00.060 align:middle line:84% is Janet Yellen, who's Secretary of Treasury, 00:22:00.060 --> 00:22:03.720 align:middle line:90% so that's quite a power couple. 00:22:03.720 --> 00:22:06.635 align:middle line:84% George Akerlof made the following talk in-- 00:22:06.635 --> 00:22:08.260 align:middle line:84% and once again, it's a review in 14.01, 00:22:08.260 --> 00:22:09.880 align:middle line:84% but he talked in his 1970 article 00:22:09.880 --> 00:22:13.280 align:middle line:84% about the used car market, and talked about the fact 00:22:13.280 --> 00:22:16.440 align:middle line:84% that in the used car market, there's asymmetric information. 00:22:16.440 --> 00:22:19.420 align:middle line:84% The person selling the car knows more 00:22:19.420 --> 00:22:22.540 align:middle line:84% about the car than the person buying the car. 00:22:22.540 --> 00:22:25.460 align:middle line:84% As a result, you can't necessarily 00:22:25.460 --> 00:22:28.580 align:middle line:84% come to an agreement on what the car is worth. 00:22:28.580 --> 00:22:31.020 align:middle line:84% And if you can't agree with what the car is worth, 00:22:31.020 --> 00:22:35.700 align:middle line:84% you might not be able to make a mutually beneficial trade. 00:22:35.700 --> 00:22:38.940 align:middle line:84% Buyers are perfectly happy to pay a lot of money 00:22:38.940 --> 00:22:40.820 align:middle line:90% for a good car. 00:22:40.820 --> 00:22:43.300 align:middle line:84% Sellers may be perfectly happy to accept even less money 00:22:43.300 --> 00:22:46.720 align:middle line:84% than that for their car, but as they can't prove it's good, 00:22:46.720 --> 00:22:49.380 align:middle line:90% they can't make the deal. 00:22:49.380 --> 00:22:51.780 align:middle line:84% Health insurance is the opposite. 00:22:51.780 --> 00:22:55.260 align:middle line:84% In health insurance, it's the buyers who know more 00:22:55.260 --> 00:22:58.220 align:middle line:90% and the sellers who know less. 00:22:58.220 --> 00:23:01.220 align:middle line:84% So basically, the problem is, the insurance sellers 00:23:01.220 --> 00:23:05.440 align:middle line:84% don't know exactly what's wrong with you, 00:23:05.440 --> 00:23:07.340 align:middle line:84% so they don't know how to price it. 00:23:07.340 --> 00:23:10.420 align:middle line:84% And that can, but not necessarily will, 00:23:10.420 --> 00:23:12.642 align:middle line:90% lead to market failure. 00:23:12.642 --> 00:23:15.100 align:middle line:84% And you want to remember, it can, but not necessarily will. 00:23:15.100 --> 00:23:16.400 align:middle line:84% So to do that, once again, we'll skip over 00:23:16.400 --> 00:23:19.120 align:middle line:84% the easy stuff in the chapter, let's go to the appendix. 00:23:19.120 --> 00:23:20.200 align:middle line:90% OK. 00:23:20.200 --> 00:23:24.160 align:middle line:84% So basically, imagine we're in the same world, 00:23:24.160 --> 00:23:30.720 align:middle line:84% but now there's two types, careful and careless. 00:23:30.720 --> 00:23:32.520 align:middle line:90% Careful and careless. 00:23:32.520 --> 00:23:37.920 align:middle line:84% So now we're going to say the probability of an accident 00:23:37.920 --> 00:23:40.798 align:middle line:90% is, a, we're going to say Pa-- 00:23:40.798 --> 00:23:42.340 align:middle line:84% I shouldn't use careful and careless, 00:23:42.340 --> 00:23:45.240 align:middle line:90% that's really not helpful. 00:23:45.240 --> 00:23:53.260 align:middle line:84% So Pa is the careless and Pc is the careful. 00:23:53.260 --> 00:23:56.643 align:middle line:90% 00:23:56.643 --> 00:23:58.060 align:middle line:84% Valerie, can you send me an email, 00:23:58.060 --> 00:23:59.980 align:middle line:84% remind me to change that in the next section of the textbook? 00:23:59.980 --> 00:24:01.240 align:middle line:90% That's not very helpful. 00:24:01.240 --> 00:24:01.900 align:middle line:90% OK. 00:24:01.900 --> 00:24:05.960 align:middle line:90% So, we're going to assume P-- 00:24:05.960 --> 00:24:13.420 align:middle line:84% the probability of an accident, Pa, is greater than Pc. 00:24:13.420 --> 00:24:17.420 align:middle line:84% So now we have the same model as before with two types, 00:24:17.420 --> 00:24:20.340 align:middle line:84% with different odds of an accident. 00:24:20.340 --> 00:24:23.898 align:middle line:84% Now, imagine that insurers know your type. 00:24:23.898 --> 00:24:26.440 align:middle line:84% When you walk into the office, it's stamped on your forehead, 00:24:26.440 --> 00:24:30.540 align:middle line:84% it's irremovable, and it says I'm careful, I'm careless. 00:24:30.540 --> 00:24:32.340 align:middle line:84% Then this whole model works, they just 00:24:32.340 --> 00:24:34.380 align:middle line:90% set two separate prices. 00:24:34.380 --> 00:24:37.140 align:middle line:84% They just say, well, I've got a Pc and a Pa, 00:24:37.140 --> 00:24:39.220 align:middle line:90% and everything goes through. 00:24:39.220 --> 00:24:40.080 align:middle line:90% Nothing changes. 00:24:40.080 --> 00:24:42.560 align:middle line:84% It's just-- they say, oh, I see you're careful, 00:24:42.560 --> 00:24:44.360 align:middle line:84% so I'm going to do this whole math, 00:24:44.360 --> 00:24:48.965 align:middle line:84% and I'm going to set up this math according to P sub c. 00:24:48.965 --> 00:24:50.340 align:middle line:84% If I see you're reckless, I'll do 00:24:50.340 --> 00:24:53.007 align:middle line:84% the whole math, set to P sub a, and I'll get different premiums. 00:24:53.007 --> 00:24:56.420 align:middle line:90% Which premium will be higher? 00:24:56.420 --> 00:24:58.420 align:middle line:90% Which premium will be higher? 00:24:58.420 --> 00:24:59.020 align:middle line:90% Pa. 00:24:59.020 --> 00:25:04.860 align:middle line:84% Because remember, the premium that you set here, 00:25:04.860 --> 00:25:06.780 align:middle line:90% M, is equal to the probability. 00:25:06.780 --> 00:25:09.340 align:middle line:84% So you get a higher M because P is higher. 00:25:09.340 --> 00:25:11.320 align:middle line:90% Full stop, done. 00:25:11.320 --> 00:25:13.140 align:middle line:84% But now imagine, as in the real world, 00:25:13.140 --> 00:25:16.360 align:middle line:84% we don't know who's careful and careless. 00:25:16.360 --> 00:25:19.000 align:middle line:84% Now imagine we don't know who's careful and careless. 00:25:19.000 --> 00:25:21.480 align:middle line:84% And imagine the insurance company only 00:25:21.480 --> 00:25:26.440 align:middle line:84% knows the proportions of each in society. 00:25:26.440 --> 00:25:29.040 align:middle line:84% They know the proportions of each in society, 00:25:29.040 --> 00:25:32.962 align:middle line:84% but they don't know who's careful and who's careless. 00:25:32.962 --> 00:25:34.420 align:middle line:84% So what's the insurance company do? 00:25:34.420 --> 00:25:37.560 align:middle line:84% Well, one strategy is they could say, look, 00:25:37.560 --> 00:25:40.368 align:middle line:84% I'm going to assume people are honest. 00:25:40.368 --> 00:25:42.160 align:middle line:84% And I'm just going to charge them according 00:25:42.160 --> 00:25:44.640 align:middle line:90% to what they tell me they are. 00:25:44.640 --> 00:25:49.520 align:middle line:84% So what happens is, the careful people come in 00:25:49.520 --> 00:25:54.360 align:middle line:84% and say, I'm careful, so they say, fine, 00:25:54.360 --> 00:25:57.320 align:middle line:90% I am going to charge you-- 00:25:57.320 --> 00:26:01.160 align:middle line:84% I'm going to charge you a premium-- 00:26:01.160 --> 00:26:06.910 align:middle line:84% I'm going to-- so for the careful person, my profit 00:26:06.910 --> 00:26:08.910 align:middle line:84% condition-- so go back to expression for profit, 00:26:08.910 --> 00:26:10.490 align:middle line:90% would be M times b. 00:26:10.490 --> 00:26:18.717 align:middle line:84% M-- so my profit condition is M, which is going to be b. 00:26:18.717 --> 00:26:20.550 align:middle line:84% Remember, people are still going to choose-- 00:26:20.550 --> 00:26:24.090 align:middle line:84% they're still going to choose M equals-- 00:26:24.090 --> 00:26:27.810 align:middle line:90% 00:26:27.810 --> 00:26:28.310 align:middle line:90% right. 00:26:28.310 --> 00:26:31.950 align:middle line:84% So I'm going to choose M careful-- 00:26:31.950 --> 00:26:33.830 align:middle line:84% let's write this out, let's not skip steps. 00:26:33.830 --> 00:26:42.190 align:middle line:90% M careful minus b times Pc. 00:26:42.190 --> 00:26:46.950 align:middle line:84% I'm going to set that equal to 0, just like here. 00:26:46.950 --> 00:26:53.590 align:middle line:84% Mcb minus P times bc I'm going to set to 0. 00:26:53.590 --> 00:26:58.270 align:middle line:84% That is-- and remember, M equals P, 00:26:58.270 --> 00:27:05.450 align:middle line:84% so it's just going to be Pb Pcb minus b times Pc equals 00:27:05.450 --> 00:27:07.230 align:middle line:84% 0-- wait, hold on one second, one second. 00:27:07.230 --> 00:27:08.150 align:middle line:90% Man, I'm getting the math wrong. 00:27:08.150 --> 00:27:09.610 align:middle line:90% That, by definition, equals 0. 00:27:09.610 --> 00:27:11.290 align:middle line:90% M minus b times P-- 00:27:11.290 --> 00:27:13.960 align:middle line:90% Mc, b times Pc-- right. 00:27:13.960 --> 00:27:14.598 align:middle line:90% Exactly. 00:27:14.598 --> 00:27:15.390 align:middle line:90% So they break even. 00:27:15.390 --> 00:27:16.770 align:middle line:90% They make zero profit. 00:27:16.770 --> 00:27:21.310 align:middle line:84% So if they charge you the right premium, which is M equals P, 00:27:21.310 --> 00:27:22.770 align:middle line:90% they make zero profit. 00:27:22.770 --> 00:27:26.650 align:middle line:84% Contrast that with what happens with the careless. 00:27:26.650 --> 00:27:32.530 align:middle line:84% The careless walks in and they say-- what do they say? 00:27:32.530 --> 00:27:33.470 align:middle line:90% I'm a careful guy. 00:27:33.470 --> 00:27:35.750 align:middle line:84% I walk in carefully, I'm a careful guy, trust me. 00:27:35.750 --> 00:27:36.690 align:middle line:90% I'm a careful guy. 00:27:36.690 --> 00:27:38.330 align:middle line:90% So what happens there? 00:27:38.330 --> 00:27:40.130 align:middle line:90% What happens with the careless? 00:27:40.130 --> 00:27:46.730 align:middle line:84% Well, the careless, they charge them Mc, 00:27:46.730 --> 00:27:52.610 align:middle line:90% they charge them Mc minus-- 00:27:52.610 --> 00:27:55.810 align:middle line:84% they charge them Mc, they charge them the careful premium. 00:27:55.810 --> 00:28:00.645 align:middle line:90% But their payout is b times Pa. 00:28:00.645 --> 00:28:02.770 align:middle line:84% That's what they pay out, because actually, they're 00:28:02.770 --> 00:28:05.350 align:middle line:90% careless. 00:28:05.350 --> 00:28:10.150 align:middle line:84% Which we can rewrite as b times Pc-- 00:28:10.150 --> 00:28:16.230 align:middle line:84% Mc is b times Pc minus b times Pa. 00:28:16.230 --> 00:28:20.070 align:middle line:84% And we know Pa is greater than Pc, 00:28:20.070 --> 00:28:22.470 align:middle line:90% so we know this is less than 0. 00:28:22.470 --> 00:28:23.830 align:middle line:90% So profits fall. 00:28:23.830 --> 00:28:29.390 align:middle line:84% So once again, in a world with careful, I charge them Mc-- 00:28:29.390 --> 00:28:35.570 align:middle line:84% I charge them Mc minus b times Pc, I rewrite this, 00:28:35.570 --> 00:28:38.070 align:middle line:90% and I see that it equals 0. 00:28:38.070 --> 00:28:40.370 align:middle line:84% In a world with careless, I charge them Mc, 00:28:40.370 --> 00:28:43.550 align:middle line:90% but I pay out b times Pa. 00:28:43.550 --> 00:28:45.030 align:middle line:90% So I earn less than 0. 00:28:45.030 --> 00:28:47.490 align:middle line:84% So I lose money, I go out of business. 00:28:47.490 --> 00:28:48.990 align:middle line:90% The market fails. 00:28:48.990 --> 00:28:51.490 align:middle line:84% The market fails when you don't have supply equaling demand. 00:28:51.490 --> 00:28:52.450 align:middle line:84% Well, here, if I'm going to lose money, 00:28:52.450 --> 00:28:54.158 align:middle line:84% I'm going to be in the market, therefore, 00:28:54.158 --> 00:28:57.070 align:middle line:84% there's no supply to meet the demand. 00:28:57.070 --> 00:28:59.690 align:middle line:84% So you can see in this world where 00:28:59.690 --> 00:29:03.150 align:middle line:84% we trust people will be honest, you lose money. 00:29:03.150 --> 00:29:05.150 align:middle line:84% If we trust people to be honest, you lose money. 00:29:05.150 --> 00:29:08.490 align:middle line:90% Questions about that? 00:29:08.490 --> 00:29:09.170 align:middle line:90% OK. 00:29:09.170 --> 00:29:12.570 align:middle line:84% Now, let's say I don't trust people to be honest. 00:29:12.570 --> 00:29:16.330 align:middle line:84% Let's say instead, let's say instead of trusting people 00:29:16.330 --> 00:29:19.490 align:middle line:84% to be honest, I just say, look, I know the population is 00:29:19.490 --> 00:29:22.570 align:middle line:90% half-careful and half-careless. 00:29:22.570 --> 00:29:25.090 align:middle line:84% And so what I'll do is I'll set the price 00:29:25.090 --> 00:29:31.050 align:middle line:84% to be the average, to be Pv, which 00:29:31.050 --> 00:29:35.010 align:middle line:84% is the average of the careful and the careless. 00:29:35.010 --> 00:29:36.190 align:middle line:90% And then I'll get it right. 00:29:36.190 --> 00:29:37.430 align:middle line:84% I was stupid to trust people to be honest. 00:29:37.430 --> 00:29:38.178 align:middle line:90% Nobody's honest. 00:29:38.178 --> 00:29:39.970 align:middle line:84% But I know that half the people are careful 00:29:39.970 --> 00:29:44.130 align:middle line:84% and half the people are careless, so I'll just set Pv. 00:29:44.130 --> 00:29:50.770 align:middle line:84% Now, at that price, what happens? 00:29:50.770 --> 00:29:55.770 align:middle line:84% Well, what might happen, will the careful buy insurance 00:29:55.770 --> 00:29:57.730 align:middle line:90% at a price Pv? 00:29:57.730 --> 00:30:00.270 align:middle line:90% The careful-- 00:30:00.270 --> 00:30:02.690 align:middle line:84% AUDIENCE: Depending on what other insurers are offering. 00:30:02.690 --> 00:30:03.890 align:middle line:84% JON GRUBER: Well, no, forget there's other insurers, 00:30:03.890 --> 00:30:06.223 align:middle line:84% there's one insurers, but it still depends on something. 00:30:06.223 --> 00:30:08.432 align:middle line:84% AUDIENCE: Probably not, but depends on risk aversion? 00:30:08.432 --> 00:30:10.230 align:middle line:84% JON GRUBER: Depends on risk aversion. 00:30:10.230 --> 00:30:14.390 align:middle line:84% But let's-- but will the careless buy insurance? 00:30:14.390 --> 00:30:16.770 align:middle line:84% For sure, because it's a great deal for them, 00:30:16.770 --> 00:30:18.510 align:middle line:90% because Pv is less. 00:30:18.510 --> 00:30:27.470 align:middle line:84% Pv is less than Pa, but more than Pc. 00:30:27.470 --> 00:30:29.250 align:middle line:84% So for sure the careless will buy it, 00:30:29.250 --> 00:30:31.990 align:middle line:90% it's a great deal for them. 00:30:31.990 --> 00:30:33.930 align:middle line:90% The careful may or may not buy. 00:30:33.930 --> 00:30:35.550 align:middle line:84% Remember, whether they buy depends 00:30:35.550 --> 00:30:38.750 align:middle line:84% on something we derived in 14.01 called the risk premium. 00:30:38.750 --> 00:30:41.310 align:middle line:84% Are they risk-averse enough that they're 00:30:41.310 --> 00:30:47.390 align:middle line:90% willing to pay the extra amount? 00:30:47.390 --> 00:30:52.590 align:middle line:90% So in that case, if only-- 00:30:52.590 --> 00:30:57.470 align:middle line:84% if, in fact, the rich-- the careful don't buy, 00:30:57.470 --> 00:31:01.020 align:middle line:84% then you'll see, you lose money by the same math here. 00:31:01.020 --> 00:31:04.260 align:middle line:90% Replace this with Pv. 00:31:04.260 --> 00:31:08.620 align:middle line:84% And since Pv is still less than Pa, you still lose money. 00:31:08.620 --> 00:31:09.233 align:middle line:90% Yeah? 00:31:09.233 --> 00:31:11.900 align:middle line:84% AUDIENCE: Would you expect to be careful to buy some amount of-- 00:31:11.900 --> 00:31:13.280 align:middle line:84% JON GRUBER: Hold on, you're getting ahead, 00:31:13.280 --> 00:31:14.113 align:middle line:90% we're getting ahead. 00:31:14.113 --> 00:31:16.020 align:middle line:84% Let's imagine you buy it or you don't. 00:31:16.020 --> 00:31:20.700 align:middle line:90% Now-- so that's kind of the-- 00:31:20.700 --> 00:31:21.600 align:middle line:90% let me back up. 00:31:21.600 --> 00:31:23.517 align:middle line:84% Let's say that we're selling a certain-- yeah, 00:31:23.517 --> 00:31:26.740 align:middle line:84% we're selling a certain product that's for X dollars. 00:31:26.740 --> 00:31:29.725 align:middle line:84% So you'll see that basically, you're 00:31:29.725 --> 00:31:31.100 align:middle line:84% going to end up losing money even 00:31:31.100 --> 00:31:33.420 align:middle line:90% if you use the average price. 00:31:33.420 --> 00:31:38.940 align:middle line:84% But if the careful do buy, then Pv is calculated correctly 00:31:38.940 --> 00:31:42.820 align:middle line:84% and you'll be back to making zero profits. 00:31:42.820 --> 00:31:45.100 align:middle line:84% So the bottom line is, the market 00:31:45.100 --> 00:31:49.783 align:middle line:84% can fail if you don't have perfect information. 00:31:49.783 --> 00:31:52.200 align:middle line:84% If you don't have perfect information, you assume honesty, 00:31:52.200 --> 00:31:53.497 align:middle line:90% the market will fail. 00:31:53.497 --> 00:31:55.080 align:middle line:84% If you don't have perfect information, 00:31:55.080 --> 00:31:59.120 align:middle line:84% you use an average price, the market may fail. 00:31:59.120 --> 00:32:04.280 align:middle line:84% So basically, that is kind of-- and whether it fails or not 00:32:04.280 --> 00:32:05.800 align:middle line:84% depends on whether we end up what 00:32:05.800 --> 00:32:07.820 align:middle line:90% we call a pooling equilibrium. 00:32:07.820 --> 00:32:10.800 align:middle line:90% 00:32:10.800 --> 00:32:14.720 align:middle line:84% A pooling equilibrium is an equilibrium 00:32:14.720 --> 00:32:20.040 align:middle line:84% where everyone does buy even though you 00:32:20.040 --> 00:32:21.340 align:middle line:90% can't have full information. 00:32:21.340 --> 00:32:23.780 align:middle line:84% As long as everyone buys, you're all in one big pool, 00:32:23.780 --> 00:32:25.540 align:middle line:84% so the average premium gets it right. 00:32:25.540 --> 00:32:29.160 align:middle line:90% 00:32:29.160 --> 00:32:31.840 align:middle line:84% But if you don't have that pooling equilibrium, 00:32:31.840 --> 00:32:35.160 align:middle line:90% the market will fail. 00:32:35.160 --> 00:32:39.380 align:middle line:84% Now, that is not the only thing that can happen. 00:32:39.380 --> 00:32:41.340 align:middle line:84% Stephen's question comes another point. 00:32:41.340 --> 00:32:45.680 align:middle line:84% There's another way the market cannot collapse here. 00:32:45.680 --> 00:32:48.840 align:middle line:84% There's another way where the market might exist. 00:32:48.840 --> 00:32:51.380 align:middle line:84% And that's if you say, well, you know what? 00:32:51.380 --> 00:32:54.940 align:middle line:84% I'm going to offer two separate products. 00:32:54.940 --> 00:32:57.980 align:middle line:84% I'm going to offer a crappy product for cheap money 00:32:57.980 --> 00:33:00.580 align:middle line:84% and a good product for expensive money. 00:33:00.580 --> 00:33:03.340 align:middle line:84% It is possible-- there are certain conditions, 00:33:03.340 --> 00:33:06.500 align:middle line:84% you may or may not have to solve in a problem set or exam. 00:33:06.500 --> 00:33:11.100 align:middle line:84% It's possible that careful people 00:33:11.100 --> 00:33:13.500 align:middle line:84% are OK with the crappy product at low money 00:33:13.500 --> 00:33:15.420 align:middle line:90% because the risk is low. 00:33:15.420 --> 00:33:18.340 align:middle line:84% Careless people only want the good product at high money 00:33:18.340 --> 00:33:20.220 align:middle line:90% and they're happy to buy it. 00:33:20.220 --> 00:33:22.520 align:middle line:84% So what you do is you offer two products. 00:33:22.520 --> 00:33:24.960 align:middle line:84% One is the actual careless product. 00:33:24.960 --> 00:33:27.580 align:middle line:84% So remember, we said up here, you can literally just 00:33:27.580 --> 00:33:28.480 align:middle line:90% offer two products. 00:33:28.480 --> 00:33:30.900 align:middle line:84% If you had full information-- well, if you had information 00:33:30.900 --> 00:33:32.420 align:middle line:90% you offer two products. 00:33:32.420 --> 00:33:36.820 align:middle line:84% OK, you redo this math, once with Pa and once with Pc. 00:33:36.820 --> 00:33:42.040 align:middle line:84% In that world, you do it once with Pa, you offer that product, 00:33:42.040 --> 00:33:45.020 align:middle line:84% and remember, the careless people will buy it because they 00:33:45.020 --> 00:33:46.280 align:middle line:90% want full insurance. 00:33:46.280 --> 00:33:51.540 align:middle line:84% Then you do it once at Pc, but at a much-- 00:33:51.540 --> 00:33:54.400 align:middle line:84% you do it once at Pc, but at a smaller amount. 00:33:54.400 --> 00:33:57.000 align:middle line:84% A small enough amount that the careless people 00:33:57.000 --> 00:33:58.980 align:middle line:84% aren't interested anymore buying it, 00:33:58.980 --> 00:34:01.800 align:middle line:84% but the careful people might want to buy it. 00:34:01.800 --> 00:34:05.140 align:middle line:84% And you can get what's called a separating equilibrium. 00:34:05.140 --> 00:34:08.810 align:middle line:90% 00:34:08.810 --> 00:34:10.060 align:middle line:90% We have another word for this. 00:34:10.060 --> 00:34:12.199 align:middle line:90% We call it HMOs. 00:34:12.199 --> 00:34:15.480 align:middle line:84% When I was young, the way health insurance 00:34:15.480 --> 00:34:19.639 align:middle line:84% worked is they covered you for everything very generously. 00:34:19.639 --> 00:34:24.120 align:middle line:84% And basically what happened was they charged a high price. 00:34:24.120 --> 00:34:27.219 align:middle line:84% And since it was cheap, a lot of people bought it, 00:34:27.219 --> 00:34:30.040 align:middle line:84% but as the price of medical care started going up, 00:34:30.040 --> 00:34:32.280 align:middle line:84% the careful-- i.g., the healthy, started 00:34:32.280 --> 00:34:35.127 align:middle line:84% dropping out because they didn't want to pay that price. 00:34:35.127 --> 00:34:36.960 align:middle line:84% So companies came up with this, they came up 00:34:36.960 --> 00:34:38.793 align:middle line:84% with HMOs, Health Maintenance Organizations, 00:34:38.793 --> 00:34:41.199 align:middle line:84% we'll talk about these a lot more in a few lectures. 00:34:41.199 --> 00:34:41.860 align:middle line:90% What are they? 00:34:41.860 --> 00:34:46.420 align:middle line:84% They're kind of like mediocre insurance at low prices. 00:34:46.420 --> 00:34:48.780 align:middle line:84% And what that did was it offered a cheaper option 00:34:48.780 --> 00:34:51.219 align:middle line:84% so the healthy people could still buy something. 00:34:51.219 --> 00:34:53.780 align:middle line:84% They created a separating equilibrium. 00:34:53.780 --> 00:34:56.300 align:middle line:84% And in chapter 16, we'll talk about the welfare effects 00:34:56.300 --> 00:34:58.820 align:middle line:90% of separating equilibrium. 00:34:58.820 --> 00:35:01.580 align:middle line:84% But they created a separating equilibrium 00:35:01.580 --> 00:35:03.220 align:middle line:84% where you had really good insurance 00:35:03.220 --> 00:35:05.680 align:middle line:84% for the sick and mediocre insurance for the healthy, 00:35:05.680 --> 00:35:08.500 align:middle line:90% and both bought. 00:35:08.500 --> 00:35:12.340 align:middle line:84% Now, here's the real tricky insight, 00:35:12.340 --> 00:35:18.220 align:middle line:84% which is separating market failure from market collapse. 00:35:18.220 --> 00:35:24.860 align:middle line:84% In the world where you counted on people to be honest, 00:35:24.860 --> 00:35:28.100 align:middle line:84% the market collapsed because insurers couldn't make money. 00:35:28.100 --> 00:35:30.780 align:middle line:84% In the world where you charge average prices 00:35:30.780 --> 00:35:32.800 align:middle line:84% and you didn't get a pooling equilibrium, 00:35:32.800 --> 00:35:35.100 align:middle line:90% the market collapsed. 00:35:35.100 --> 00:35:38.440 align:middle line:84% In the separating equilibrium, the market doesn't collapse, 00:35:38.440 --> 00:35:39.800 align:middle line:90% but it still fails. 00:35:39.800 --> 00:35:40.740 align:middle line:90% Why? 00:35:40.740 --> 00:35:44.780 align:middle line:84% Why does the market still exist but is no longer efficient? 00:35:44.780 --> 00:35:46.962 align:middle line:84% In a separating equilibrium-- yeah? 00:35:46.962 --> 00:35:48.920 align:middle line:84% AUDIENCE: Because the optimal is fully insured. 00:35:48.920 --> 00:35:50.712 align:middle line:84% JON GRUBER: Yes, because the healthy people 00:35:50.712 --> 00:35:53.400 align:middle line:84% are getting less than full insurance, so it's not optimal. 00:35:53.400 --> 00:35:57.240 align:middle line:84% So market collapse equals market failure. 00:35:57.240 --> 00:36:01.520 align:middle line:84% Market existing doesn't equal market efficiency. 00:36:01.520 --> 00:36:03.808 align:middle line:84% It's an inefficient outcome because healthy people 00:36:03.808 --> 00:36:05.600 align:middle line:84% are getting less than full insurance, which 00:36:05.600 --> 00:36:09.740 align:middle line:84% is the first best, which is the efficient outcome. 00:36:09.740 --> 00:36:11.600 align:middle line:90% Important distinction. 00:36:11.600 --> 00:36:14.600 align:middle line:84% So the only case in which the market actually doesn't fail 00:36:14.600 --> 00:36:16.960 align:middle line:84% is the pooling equilibrium, where everyone still 00:36:16.960 --> 00:36:19.120 align:middle line:84% buys full insurance because the rich are 00:36:19.120 --> 00:36:21.680 align:middle line:84% just risk-averse-- because the careful are just 00:36:21.680 --> 00:36:23.477 align:middle line:90% risk-averse enough. 00:36:23.477 --> 00:36:25.560 align:middle line:84% Then there's markets-- the market doesn't collapse 00:36:25.560 --> 00:36:26.977 align:middle line:90% and it doesn't fail. 00:36:26.977 --> 00:36:29.560 align:middle line:84% But in every other example, it either collapses, in which case 00:36:29.560 --> 00:36:31.960 align:middle line:84% it fails, or it exists, but still fails. 00:36:31.960 --> 00:36:34.140 align:middle line:84% It still fails to achieve-- by market failure, 00:36:34.140 --> 00:36:37.160 align:middle line:84% we mean not that the market doesn't exist. 00:36:37.160 --> 00:36:39.537 align:middle line:84% Plenty of markets that exist have still failed. 00:36:39.537 --> 00:36:41.120 align:middle line:84% By market failure, we mean it does not 00:36:41.120 --> 00:36:45.520 align:middle line:84% deliver the efficient outcome, the welfare-maximizing outcome. 00:36:45.520 --> 00:36:47.530 align:middle line:84% And that's true with the separating equilibrium. 00:36:47.530 --> 00:36:50.150 align:middle line:90% 00:36:50.150 --> 00:36:50.950 align:middle line:90% Yeah? 00:36:50.950 --> 00:36:53.910 align:middle line:84% AUDIENCE: So just to clarify, complete insurance 00:36:53.910 --> 00:36:56.570 align:middle line:84% means that in both the good and bad state of the world, 00:36:56.570 --> 00:36:57.443 align:middle line:90% utilities are equal? 00:36:57.443 --> 00:36:58.110 align:middle line:90% JON GRUBER: Yes. 00:36:58.110 --> 00:36:59.860 align:middle line:84% AUDIENCE: So in this case, it's incomplete 00:36:59.860 --> 00:37:03.148 align:middle line:84% because healthy people are better off in the good-- 00:37:03.148 --> 00:37:05.690 align:middle line:84% JON GRUBER: Yes, because they're not fully insuring the risk. 00:37:05.690 --> 00:37:06.790 align:middle line:90% Well put. 00:37:06.790 --> 00:37:08.150 align:middle line:90% Well put. 00:37:08.150 --> 00:37:14.070 align:middle line:84% So it's not-- we're not getting full consumption smoothing 00:37:14.070 --> 00:37:16.630 align:middle line:84% because the healthy people are ending up worse 00:37:16.630 --> 00:37:20.390 align:middle line:90% off if they actually get sick. 00:37:20.390 --> 00:37:20.958 align:middle line:90% OK. 00:37:20.958 --> 00:37:22.750 align:middle line:84% Let's talk about how this works in practice 00:37:22.750 --> 00:37:25.270 align:middle line:84% with a particularly pernicious example, down the street 00:37:25.270 --> 00:37:26.670 align:middle line:90% at Harvard University. 00:37:26.670 --> 00:37:28.990 align:middle line:84% All the same thing sort of happened at MIT. 00:37:28.990 --> 00:37:33.350 align:middle line:84% And the consequence of what can happen, we call this problem-- 00:37:33.350 --> 00:37:37.350 align:middle line:84% the problem that arises only because people who are healthy 00:37:37.350 --> 00:37:39.430 align:middle line:84% might not want to buy is called the problem 00:37:39.430 --> 00:37:42.570 align:middle line:90% of adverse selection. 00:37:42.570 --> 00:37:46.010 align:middle line:90% 00:37:46.010 --> 00:37:48.490 align:middle line:84% Adverse selection is a problem that 00:37:48.490 --> 00:37:54.290 align:middle line:84% arises because only the sick people want to buy, 00:37:54.290 --> 00:37:59.490 align:middle line:84% and they might skip the market and it might collapse. 00:37:59.490 --> 00:38:01.907 align:middle line:84% Adverse selection can have very pernicious examples, 00:38:01.907 --> 00:38:03.990 align:middle line:84% as we can see-- you don't have to write this down, 00:38:03.990 --> 00:38:04.930 align:middle line:84% it's in the book, but let's talk about what 00:38:04.930 --> 00:38:06.222 align:middle line:90% happened at Harvard University. 00:38:06.222 --> 00:38:09.050 align:middle line:84% Harvard University had a health insurance plan 00:38:09.050 --> 00:38:11.910 align:middle line:84% which was what was called experience-rated. 00:38:11.910 --> 00:38:18.130 align:middle line:90% 00:38:18.130 --> 00:38:19.110 align:middle line:90% Experience-rated. 00:38:19.110 --> 00:38:22.887 align:middle line:84% What that means is Harvard effectively was its own insurer. 00:38:22.887 --> 00:38:25.470 align:middle line:84% The insurers administered the program, at the end of the year, 00:38:25.470 --> 00:38:28.050 align:middle line:84% they charged Harvard whatever people spent in medical care. 00:38:28.050 --> 00:38:30.650 align:middle line:84% They were essentially what's called self-insured. 00:38:30.650 --> 00:38:34.410 align:middle line:84% That basically, the more people spent, the more Harvard paid. 00:38:34.410 --> 00:38:38.850 align:middle line:84% The insurer effectively didn't bear the risk, Harvard did. 00:38:38.850 --> 00:38:41.870 align:middle line:84% Now, Harvard had a set of plans, some which are more generous 00:38:41.870 --> 00:38:43.910 align:middle line:84% and some which are less generous. 00:38:43.910 --> 00:38:46.510 align:middle line:84% And Harvard had a system where basically they 00:38:46.510 --> 00:38:49.090 align:middle line:84% subsidized the more generous plans, 00:38:49.090 --> 00:38:51.743 align:middle line:84% so the gap between the two wasn't very large. 00:38:51.743 --> 00:38:53.910 align:middle line:84% The more generous plans didn't really cost much more 00:38:53.910 --> 00:38:55.932 align:middle line:90% than the less generous plans. 00:38:55.932 --> 00:38:58.390 align:middle line:84% So everyone just was in the more generous plans because why 00:38:58.390 --> 00:38:59.030 align:middle line:90% the hell not? 00:38:59.030 --> 00:39:01.790 align:middle line:84% Everyone likes more generous insurance. 00:39:01.790 --> 00:39:05.470 align:middle line:84% Then Harvard changed its system where now they 00:39:05.470 --> 00:39:07.310 align:middle line:84% allowed there to be-- they allowed 00:39:07.310 --> 00:39:10.190 align:middle line:84% there to be a price gap between the two plans 00:39:10.190 --> 00:39:13.270 align:middle line:84% that reflected their own experience. 00:39:13.270 --> 00:39:16.070 align:middle line:84% So each plan now reflected the actual experience of people 00:39:16.070 --> 00:39:16.750 align:middle line:90% in it. 00:39:16.750 --> 00:39:19.270 align:middle line:84% What that meant was, the really generous plans 00:39:19.270 --> 00:39:23.185 align:middle line:84% that attracted the very sickest people was expensive. 00:39:23.185 --> 00:39:25.310 align:middle line:84% The other plan, which didn't attract as sick people 00:39:25.310 --> 00:39:28.390 align:middle line:84% because it wasn't very generous, was less expensive. 00:39:28.390 --> 00:39:31.878 align:middle line:84% So the healthy people went to the less expensive plan. 00:39:31.878 --> 00:39:32.670 align:middle line:90% What happens, then? 00:39:32.670 --> 00:39:34.910 align:middle line:84% Let's just think about the math for a second. 00:39:34.910 --> 00:39:38.310 align:middle line:84% You've got sick people here, healthy people here. 00:39:38.310 --> 00:39:40.910 align:middle line:84% Healthy people leave here and move to here. 00:39:40.910 --> 00:39:43.650 align:middle line:84% What happens to the price of this plan? 00:39:43.650 --> 00:39:44.385 align:middle line:90% Goes up. 00:39:44.385 --> 00:39:46.010 align:middle line:84% What happens to the price of this plan? 00:39:46.010 --> 00:39:47.290 align:middle line:90% Goes down. 00:39:47.290 --> 00:39:48.850 align:middle line:84% Well, now what's going to happen? 00:39:48.850 --> 00:39:50.110 align:middle line:84% Healthy people are going to leave here. 00:39:50.110 --> 00:39:52.050 align:middle line:84% Now it's like, well shit, none of the healthy people 00:39:52.050 --> 00:39:53.370 align:middle line:84% want to be here, so they keep leaving, 00:39:53.370 --> 00:39:55.030 align:middle line:84% and they keep leaving and leaving. 00:39:55.030 --> 00:39:57.090 align:middle line:90% And we call that a death spiral. 00:39:57.090 --> 00:39:59.970 align:middle line:84% It got to the point where just like three super sick guys left 00:39:59.970 --> 00:40:01.930 align:middle line:84% and Harvard just closed down the plan. 00:40:01.930 --> 00:40:04.610 align:middle line:90% Same thing happened at MIT. 00:40:04.610 --> 00:40:07.108 align:middle line:84% So adverse selection can actually destroy markets. 00:40:07.108 --> 00:40:08.650 align:middle line:84% That's how adverse selection destroys 00:40:08.650 --> 00:40:13.010 align:middle line:84% a market, is through that kind of mechanism, where basically-- 00:40:13.010 --> 00:40:14.850 align:middle line:84% because each-- because basically, 00:40:14.850 --> 00:40:16.902 align:middle line:84% healthy people want to be with healthy people 00:40:16.902 --> 00:40:18.110 align:middle line:90% because the premiums are low. 00:40:18.110 --> 00:40:19.610 align:middle line:84% Sick people don't want to be with healthy people 00:40:19.610 --> 00:40:21.150 align:middle line:84% because they don't want cheap insurance, 00:40:21.150 --> 00:40:22.290 align:middle line:84% and the thing drifts further and further 00:40:22.290 --> 00:40:23.630 align:middle line:90% apart until it collapses. 00:40:23.630 --> 00:40:24.450 align:middle line:90% Yeah? 00:40:24.450 --> 00:40:28.290 align:middle line:84% AUDIENCE: So once the more expensive plan collapses, 00:40:28.290 --> 00:40:30.785 align:middle line:84% and those people who were covered under it, I guess, 00:40:30.785 --> 00:40:31.410 align:middle line:90% move into the-- 00:40:31.410 --> 00:40:33.570 align:middle line:90% JON GRUBER: They end up-- yeah. 00:40:33.570 --> 00:40:36.037 align:middle line:84% AUDIENCE: It just ends up always returning to one-- 00:40:36.037 --> 00:40:37.870 align:middle line:84% JON GRUBER: It ends up with one shitty plan. 00:40:37.870 --> 00:40:40.310 align:middle line:90% It's a race to the bottom. 00:40:40.310 --> 00:40:43.030 align:middle line:84% That basically you end up in one mediocre plan 00:40:43.030 --> 00:40:46.850 align:middle line:84% because that plan is the one that everyone can-- 00:40:46.850 --> 00:40:48.330 align:middle line:90% that can exist in equilibrium. 00:40:48.330 --> 00:40:50.990 align:middle line:84% That's where the new pooling equilibrium is. 00:40:50.990 --> 00:40:53.870 align:middle line:84% The old pooling equilibrium worked because the price gap 00:40:53.870 --> 00:40:55.310 align:middle line:90% wasn't very large. 00:40:55.310 --> 00:40:58.607 align:middle line:84% Now you get this new pooling equilibrium in a mediocre plan. 00:40:58.607 --> 00:41:00.190 align:middle line:84% And reacting to this, Harvard actually 00:41:00.190 --> 00:41:02.065 align:middle line:84% then reintroduced the subsidies to bring them 00:41:02.065 --> 00:41:03.630 align:middle line:90% back closer together. 00:41:03.630 --> 00:41:05.630 align:middle line:84% AUDIENCE: And in that final state where everyone 00:41:05.630 --> 00:41:07.670 align:middle line:84% is in a mediocre plan, presumably people 00:41:07.670 --> 00:41:09.930 align:middle line:84% would all be better off if they were fully insured, 00:41:09.930 --> 00:41:11.390 align:middle line:90% but there's no way of getting-- 00:41:11.390 --> 00:41:13.348 align:middle line:84% JON GRUBER: Well, OK, so let's talk about that. 00:41:13.348 --> 00:41:16.130 align:middle line:84% What can the government do about adverse selection? 00:41:16.130 --> 00:41:16.630 align:middle line:90% Yeah? 00:41:16.630 --> 00:41:18.130 align:middle line:90% AUDIENCE: [INAUDIBLE] 00:41:18.130 --> 00:41:19.110 align:middle line:90% JON GRUBER: Yeah. 00:41:19.110 --> 00:41:21.270 align:middle line:84% The government could enforce a pooling equilibrium. 00:41:21.270 --> 00:41:22.290 align:middle line:84% The government could just say, look, 00:41:22.290 --> 00:41:24.310 align:middle line:84% everyone has to buy insurance at the average price. 00:41:24.310 --> 00:41:25.768 align:middle line:84% Let's go back to our simple example 00:41:25.768 --> 00:41:27.110 align:middle line:90% of one plan, one product. 00:41:27.110 --> 00:41:29.510 align:middle line:84% The government could enforce a pooling equilibrium 00:41:29.510 --> 00:41:32.830 align:middle line:84% so everyone has to buy insurance at the pooling equilibrium 00:41:32.830 --> 00:41:35.110 align:middle line:90% price. 00:41:35.110 --> 00:41:37.970 align:middle line:84% So, the government could say, everybody has to pay Pv, 00:41:37.970 --> 00:41:41.250 align:middle line:84% and you have to be an individual mandate. 00:41:41.250 --> 00:41:44.570 align:middle line:84% That would be efficient in the sense 00:41:44.570 --> 00:41:47.490 align:middle line:84% that everyone will be fully insured. 00:41:47.490 --> 00:41:50.330 align:middle line:84% But now let's think about it politically. 00:41:50.330 --> 00:41:52.223 align:middle line:90% Who does not like that plan? 00:41:52.223 --> 00:41:53.890 align:middle line:84% AUDIENCE: People who don't want to pay-- 00:41:53.890 --> 00:41:55.765 align:middle line:84% JON GRUBER: People who are healthy and aren't 00:41:55.765 --> 00:41:59.170 align:middle line:84% risk-averse are basically forced into something they don't want. 00:41:59.170 --> 00:42:01.650 align:middle line:84% So foreshadowing what happened with the ACA, 00:42:01.650 --> 00:42:04.330 align:middle line:84% individual mandate is economically efficient, 00:42:04.330 --> 00:42:06.098 align:middle line:84% but politically challenging because you're 00:42:06.098 --> 00:42:08.390 align:middle line:84% forced into by people to buy something they don't want. 00:42:08.390 --> 00:42:08.890 align:middle line:90% Yeah? 00:42:08.890 --> 00:42:11.370 align:middle line:84% AUDIENCE: Is forcing a minimum standard of coverage, 00:42:11.370 --> 00:42:12.965 align:middle line:90% does that also have the same-- 00:42:12.965 --> 00:42:13.590 align:middle line:90% JON GRUBER: No. 00:42:13.590 --> 00:42:14.650 align:middle line:84% Forcing a minimum standard of coverage 00:42:14.650 --> 00:42:15.970 align:middle line:84% would be different because you don't force people to 00:42:15.970 --> 00:42:18.387 align:middle line:84% by a minimum standard, you just make the adverse selection 00:42:18.387 --> 00:42:21.855 align:middle line:84% spiral worse because people just drop out. 00:42:21.855 --> 00:42:22.730 align:middle line:90% AUDIENCE: [INAUDIBLE] 00:42:22.730 --> 00:42:23.550 align:middle line:90% JON GRUBER: Yeah, exactly. 00:42:23.550 --> 00:42:25.550 align:middle line:84% As long as you have an outside operative with no insurance, 00:42:25.550 --> 00:42:27.430 align:middle line:84% the more expensive you make the insurance-- 00:42:27.430 --> 00:42:28.550 align:middle line:84% this is what I struggle with, I had 00:42:28.550 --> 00:42:30.330 align:middle line:84% to set insurance for the state of Massachusetts 00:42:30.330 --> 00:42:33.090 align:middle line:84% under our health insurance plan, and the thing we struggled with 00:42:33.090 --> 00:42:35.382 align:middle line:84% is, well, we want people to have really good insurance, 00:42:35.382 --> 00:42:38.110 align:middle line:84% but our mandate wasn't that strong. 00:42:38.110 --> 00:42:39.860 align:middle line:84% So unless you have a super strong mandate, 00:42:39.860 --> 00:42:41.485 align:middle line:84% if you force a minimum standard, you're 00:42:41.485 --> 00:42:44.268 align:middle line:84% just going to chase people out of the insurance market. 00:42:44.268 --> 00:42:46.060 align:middle line:84% AUDIENCE: Can you force a full equilibrium, 00:42:46.060 --> 00:42:48.260 align:middle line:84% but only if you mandate people have to have insurance, 00:42:48.260 --> 00:42:49.510 align:middle line:90% insurance has to be this good? 00:42:49.510 --> 00:42:50.640 align:middle line:90% JON GRUBER: Yeah, exactly. 00:42:50.640 --> 00:42:52.920 align:middle line:84% OK, so that's one thing you could do. 00:42:52.920 --> 00:42:55.200 align:middle line:84% What's the other thing you could do? 00:42:55.200 --> 00:43:00.180 align:middle line:84% How else could you get to a pooling equilibrium? 00:43:00.180 --> 00:43:03.360 align:middle line:90% 00:43:03.360 --> 00:43:04.960 align:middle line:84% What's one way you could absolutely 00:43:04.960 --> 00:43:06.280 align:middle line:84% get to pooling equilibrium if you didn't care 00:43:06.280 --> 00:43:07.447 align:middle line:90% about the government budget? 00:43:07.447 --> 00:43:12.040 align:middle line:90% 00:43:12.040 --> 00:43:12.880 align:middle line:90% Yeah? 00:43:12.880 --> 00:43:13.520 align:middle line:90% What? 00:43:13.520 --> 00:43:14.600 align:middle line:90% Make insurance free. 00:43:14.600 --> 00:43:17.720 align:middle line:84% At a price of zero, you get a pooling equilibrium. 00:43:17.720 --> 00:43:24.560 align:middle line:84% By that same math, subsidizing insurance 00:43:24.560 --> 00:43:27.120 align:middle line:84% offsets the adverse selection problem. 00:43:27.120 --> 00:43:30.440 align:middle line:84% The more you subsidize insurance with the limit being zero, 00:43:30.440 --> 00:43:32.420 align:middle line:84% the less adverse selection there is. 00:43:32.420 --> 00:43:34.737 align:middle line:84% Now by the way, it turns out, in practice-- 00:43:34.737 --> 00:43:37.320 align:middle line:84% this is the theory chapter, in practice with health insurance, 00:43:37.320 --> 00:43:38.340 align:middle line:90% that's still not good enough. 00:43:38.340 --> 00:43:39.940 align:middle line:84% Because even at free, a lot of healthy people 00:43:39.940 --> 00:43:41.148 align:middle line:90% just don't bother to sign up. 00:43:41.148 --> 00:43:42.160 align:middle line:90% It's never really free. 00:43:42.160 --> 00:43:44.120 align:middle line:84% You've got to go sign up, stuff like that. 00:43:44.120 --> 00:43:45.745 align:middle line:84% So you still, even with free insurance, 00:43:45.745 --> 00:43:47.440 align:middle line:84% get adverse selection, but it's not-- 00:43:47.440 --> 00:43:49.760 align:middle line:84% but it's minimized by the amount of subsidy. 00:43:49.760 --> 00:43:52.200 align:middle line:84% So the more-- so there's an interesting trade-off here. 00:43:52.200 --> 00:43:54.492 align:middle line:84% The more you make insurance-- subsidize insurance, make 00:43:54.492 --> 00:43:57.140 align:middle line:84% it cheaper, the less adverse selection problem you have. 00:43:57.140 --> 00:44:01.020 align:middle line:84% So you can mandate or you can subsidize, 00:44:01.020 --> 00:44:05.140 align:middle line:84% and that's how you can fight the adverse selection problem. 00:44:05.140 --> 00:44:07.860 align:middle line:90% Questions about that? 00:44:07.860 --> 00:44:09.878 align:middle line:84% Now, I started this whole thing by saying, 00:44:09.878 --> 00:44:11.920 align:middle line:84% why does the government provide social insurance? 00:44:11.920 --> 00:44:14.280 align:middle line:84% Well, adverse selection is one a set of reasons. 00:44:14.280 --> 00:44:16.560 align:middle line:84% There's a number of other reasons that matter as well, 00:44:16.560 --> 00:44:18.227 align:middle line:84% we just don't spend as much time on them 00:44:18.227 --> 00:44:19.420 align:middle line:90% because they're not as fun. 00:44:19.420 --> 00:44:21.975 align:middle line:84% One is externalities, what we started the very lecture with. 00:44:21.975 --> 00:44:24.600 align:middle line:84% Remember, I was sick, I had the flu, I coughed, I got you sick. 00:44:24.600 --> 00:44:25.892 align:middle line:90% So we started the course with-- 00:44:25.892 --> 00:44:27.700 align:middle line:84% I wasn't actually sick, it was the example. 00:44:27.700 --> 00:44:30.600 align:middle line:84% In that case, there's another argument 00:44:30.600 --> 00:44:32.960 align:middle line:84% for government intervening insurance markets, 00:44:32.960 --> 00:44:34.120 align:middle line:90% externalities. 00:44:34.120 --> 00:44:38.040 align:middle line:84% Another argument was mentioned in a question 00:44:38.040 --> 00:44:40.220 align:middle line:84% by Stephen, which is administrative costs. 00:44:40.220 --> 00:44:43.920 align:middle line:84% This whole model only works, only 00:44:43.920 --> 00:44:48.260 align:middle line:84% delivers this efficient outcome under the M equals P assumption. 00:44:48.260 --> 00:44:51.640 align:middle line:84% So only if insurers-- if we have zero profit for insurers, 00:44:51.640 --> 00:44:53.440 align:middle line:84% then we get the efficient outcome. 00:44:53.440 --> 00:44:55.560 align:middle line:90% What if insurers make profit? 00:44:55.560 --> 00:45:00.640 align:middle line:84% Well, if insurers make profit, then M will be higher than P. 00:45:00.640 --> 00:45:02.240 align:middle line:84% And if M is higher than P, people 00:45:02.240 --> 00:45:04.360 align:middle line:90% will not choose full insurance. 00:45:04.360 --> 00:45:06.960 align:middle line:84% You can solve that in your own free time, 00:45:06.960 --> 00:45:10.040 align:middle line:84% but you can see that if M is greater than P, 00:45:10.040 --> 00:45:12.820 align:middle line:84% full insurance is no longer optimal for individuals, 00:45:12.820 --> 00:45:16.048 align:middle line:84% which means you will not be the efficient outcome. 00:45:16.048 --> 00:45:17.840 align:middle line:84% So-- and it turns out, administrative costs 00:45:17.840 --> 00:45:20.340 align:middle line:84% matter a lot, and they're much higher for private insurance. 00:45:20.340 --> 00:45:22.257 align:middle line:84% The administrative costs for private insurance 00:45:22.257 --> 00:45:24.320 align:middle line:90% are on the order of 12% to 15%. 00:45:24.320 --> 00:45:27.000 align:middle line:84% So every dollar you spend on private insurance, 00:45:27.000 --> 00:45:29.900 align:middle line:84% $0.12 to $0.15 goes to paying bills and profit 00:45:29.900 --> 00:45:31.380 align:middle line:90% and things like that. 00:45:31.380 --> 00:45:33.380 align:middle line:84% The administrative cost for government insurance 00:45:33.380 --> 00:45:36.020 align:middle line:90% is on the order of 2%. 00:45:36.020 --> 00:45:37.080 align:middle line:90% Why is it so much lower? 00:45:37.080 --> 00:45:38.163 align:middle line:90% Well, there's two reasons. 00:45:38.163 --> 00:45:40.740 align:middle line:84% First of all, there's no profits being made. 00:45:40.740 --> 00:45:43.380 align:middle line:84% But second of all, many administrative costs 00:45:43.380 --> 00:45:46.388 align:middle line:84% are fixed, so the larger the pool, 00:45:46.388 --> 00:45:48.180 align:middle line:84% the smaller they are as a percentage, which 00:45:48.180 --> 00:45:49.780 align:middle line:90% is what we care about. 00:45:49.780 --> 00:45:51.280 align:middle line:84% So an argument for social insurance, 00:45:51.280 --> 00:45:53.880 align:middle line:84% you can create giant pools by having the government do it, 00:45:53.880 --> 00:45:56.020 align:middle line:84% and that can lower administrative costs. 00:45:56.020 --> 00:46:00.300 align:middle line:84% Not in dollars, but as a percent of the premium. 00:46:00.300 --> 00:46:03.083 align:middle line:84% So that's another reason why we might want social insurance. 00:46:03.083 --> 00:46:04.500 align:middle line:84% We've got adverse selection, we've 00:46:04.500 --> 00:46:07.700 align:middle line:84% got externalities, administrative costs. 00:46:07.700 --> 00:46:10.660 align:middle line:84% Redistribution, we might just want social insurance 00:46:10.660 --> 00:46:13.352 align:middle line:84% because the private market might not deliver a fair outcome. 00:46:13.352 --> 00:46:15.560 align:middle line:84% It might be the rich can afford really good insurance 00:46:15.560 --> 00:46:17.863 align:middle line:90% and we don't like that. 00:46:17.863 --> 00:46:20.280 align:middle line:84% And then finally, we have the reason that economists hate, 00:46:20.280 --> 00:46:22.322 align:middle line:84% but it's probably the major reason governments do 00:46:22.322 --> 00:46:23.780 align:middle line:90% things, which is paternalism. 00:46:23.780 --> 00:46:25.580 align:middle line:84% The government may just think, look, 00:46:25.580 --> 00:46:27.020 align:middle line:84% the private market is not getting it right, 00:46:27.020 --> 00:46:28.562 align:middle line:84% people just aren't figuring this out, 00:46:28.562 --> 00:46:31.560 align:middle line:84% people aren't buying insurance when they should. 00:46:31.560 --> 00:46:34.160 align:middle line:84% There's all sorts of behavioral problems, psychological issues 00:46:34.160 --> 00:46:35.940 align:middle line:84% that people have in buying insurance, so as a result, 00:46:35.940 --> 00:46:38.800 align:middle line:84% we need to force them to buy it or subsidize it because people 00:46:38.800 --> 00:46:41.240 align:middle line:90% aren't doing correctly. 00:46:41.240 --> 00:46:45.163 align:middle line:84% Now, a version of all this is something 00:46:45.163 --> 00:46:47.580 align:middle line:84% I discussed in the book, we don't have time to go through. 00:46:47.580 --> 00:46:50.040 align:middle line:84% There's an application about what we 00:46:50.040 --> 00:46:52.800 align:middle line:90% call the Samaritan's dilemma. 00:46:52.800 --> 00:46:54.440 align:middle line:84% The Samaritan's dilemma is the problem 00:46:54.440 --> 00:47:00.440 align:middle line:84% the government faces, which is that because I know, I 00:47:00.440 --> 00:47:04.642 align:middle line:84% as a government, can't commit to let you die in the street, 00:47:04.642 --> 00:47:07.100 align:middle line:84% I'm going to force you to buy insurance in the first place. 00:47:07.100 --> 00:47:08.200 align:middle line:90% How does this work? 00:47:08.200 --> 00:47:09.300 align:middle line:90% That is the following. 00:47:09.300 --> 00:47:12.800 align:middle line:84% If I don't force you to buy insurance, you won't buy it. 00:47:12.800 --> 00:47:14.620 align:middle line:84% If you don't buy insurance, you get sick. 00:47:14.620 --> 00:47:17.640 align:middle line:84% You stumble into the hospital and I treat you. 00:47:17.640 --> 00:47:20.140 align:middle line:84% So the Samaritan's dilemma is, because I'm a good Samaritan, 00:47:20.140 --> 00:47:24.860 align:middle line:84% because I can't not treat you, I end up spending money on you 00:47:24.860 --> 00:47:27.980 align:middle line:84% anyway, so I might as well force you to insure. 00:47:27.980 --> 00:47:32.060 align:middle line:84% The classic example of this is disaster insurance. 00:47:32.060 --> 00:47:33.980 align:middle line:84% If disaster happens, FEMA is going 00:47:33.980 --> 00:47:37.550 align:middle line:84% to come and rescue you and fix your house and stuff. 00:47:37.550 --> 00:47:39.300 align:middle line:84% And as a result, the government has set up 00:47:39.300 --> 00:47:43.540 align:middle line:84% what's called the National Flood Insurance Program to make sure 00:47:43.540 --> 00:47:46.580 align:middle line:84% that-- to basically subsidize and try 00:47:46.580 --> 00:47:49.060 align:middle line:84% to mandate, although it's not very effective, 00:47:49.060 --> 00:47:53.580 align:middle line:84% try to mandate people to have flood insurance so that when 00:47:53.580 --> 00:47:56.287 align:middle line:84% the time comes, the government doesn't end up picking the cost. 00:47:56.287 --> 00:47:58.620 align:middle line:84% The idea is the government get paid for at the beginning 00:47:58.620 --> 00:47:59.900 align:middle line:90% or the end. 00:47:59.900 --> 00:48:01.738 align:middle line:84% And by mandate and by mandating insurance, 00:48:01.738 --> 00:48:03.780 align:middle line:84% we make you pay for it rather than the government 00:48:03.780 --> 00:48:05.940 align:middle line:90% picking it up at the end. 00:48:05.940 --> 00:48:09.060 align:middle line:84% So you can actually argue that it's 00:48:09.060 --> 00:48:13.560 align:middle line:84% efficient to subsidize insurance upfront to avoid the spending 00:48:13.560 --> 00:48:15.940 align:middle line:90% you get in the back. 00:48:15.940 --> 00:48:20.340 align:middle line:84% And that is an argument for subsidized flood insurance. 00:48:20.340 --> 00:48:22.130 align:middle line:84% The problem, once again, we run into 00:48:22.130 --> 00:48:26.970 align:middle line:84% is the politics, which is that flood insurance, if properly 00:48:26.970 --> 00:48:29.330 align:middle line:84% priced, would make houses unaffordable in a number 00:48:29.330 --> 00:48:31.410 align:middle line:90% of places where houses exist. 00:48:31.410 --> 00:48:34.043 align:middle line:84% And so flood insurance is still not properly priced. 00:48:34.043 --> 00:48:36.210 align:middle line:84% People don't-- people pay too little for their flood 00:48:36.210 --> 00:48:37.030 align:middle line:90% insurance. 00:48:37.030 --> 00:48:40.150 align:middle line:84% As a result, even though we have a system, and it functions, 00:48:40.150 --> 00:48:42.890 align:middle line:84% it constantly runs out of money because it basically 00:48:42.890 --> 00:48:45.415 align:middle line:84% aren't charging people the fair price for insurance 00:48:45.415 --> 00:48:47.290 align:middle line:84% because that would mean houses on Long Island 00:48:47.290 --> 00:48:50.050 align:middle line:84% and on parts of Florida become incredibly expensive, 00:48:50.050 --> 00:48:52.810 align:middle line:84% and politicians in those areas don't want that to happen, 00:48:52.810 --> 00:48:54.490 align:middle line:84% so the flood insurance prices actually 00:48:54.490 --> 00:48:56.170 align:middle line:90% end up where they should be. 00:48:56.170 --> 00:48:59.000 align:middle line:84% So as a result, we end up-- we don't end up in a perfect system 00:48:59.000 --> 00:49:01.250 align:middle line:84% where basically people pay x instead of the government 00:49:01.250 --> 00:49:02.002 align:middle line:90% paying x post. 00:49:02.002 --> 00:49:04.210 align:middle line:84% We have the government subsidize x and the government 00:49:04.210 --> 00:49:06.290 align:middle line:90% pays some x post. 00:49:06.290 --> 00:49:10.290 align:middle line:84% So because we can't politically set flood insurance prices 00:49:10.290 --> 00:49:13.290 align:middle line:84% properly, we end up with this mixed system where we people do 00:49:13.290 --> 00:49:16.610 align:middle line:84% buy flood insurance, but FEMA still comes and bails them out 00:49:16.610 --> 00:49:19.870 align:middle line:90% at the end. 00:49:19.870 --> 00:49:22.390 align:middle line:90% Questions about that? 00:49:22.390 --> 00:49:24.390 align:middle line:84% So, those are all the reasons-- that's 00:49:24.390 --> 00:49:27.710 align:middle line:84% the first hour of the class, is all the reasons why 00:49:27.710 --> 00:49:30.430 align:middle line:84% we need social insurance, from the detailed 00:49:30.430 --> 00:49:32.710 align:middle line:90% math to the intuitive. 00:49:32.710 --> 00:49:39.270 align:middle line:84% Now, remember, the next question of our four questions 00:49:39.270 --> 00:49:41.550 align:middle line:84% is, what should the government actually do? 00:49:41.550 --> 00:49:43.850 align:middle line:84% And we're going to rephrase that here as, well, 00:49:43.850 --> 00:49:47.350 align:middle line:84% how much social insurance should the government provide? 00:49:47.350 --> 00:49:52.510 align:middle line:84% Should the government-- how much-- remember, essentially we 00:49:52.510 --> 00:49:56.030 align:middle line:84% want to ask, if you think about a social insurance program, 00:49:56.030 --> 00:49:59.370 align:middle line:84% and the government can insure you fully or less than fully, 00:49:59.370 --> 00:50:00.365 align:middle line:90% what should it do? 00:50:00.365 --> 00:50:01.990 align:middle line:84% Now you might say, well, you've already 00:50:01.990 --> 00:50:04.870 align:middle line:84% given me that answer, Jon, the government should insure fully, 00:50:04.870 --> 00:50:07.830 align:middle line:84% but there's two reasons why that's wrong. 00:50:07.830 --> 00:50:10.190 align:middle line:84% There's two reasons why, in fact, optimal government 00:50:10.190 --> 00:50:13.710 align:middle line:84% insurance is rarely 100% government insurance. 00:50:13.710 --> 00:50:16.990 align:middle line:90% The first reason is because-- 00:50:16.990 --> 00:50:19.850 align:middle line:84% the first reason why that's not right 00:50:19.850 --> 00:50:23.510 align:middle line:84% is because individuals have self-insurance. 00:50:23.510 --> 00:50:28.050 align:middle line:90% 00:50:28.050 --> 00:50:30.810 align:middle line:84% Remember, the goal of social insurance 00:50:30.810 --> 00:50:34.250 align:middle line:84% is not to provide you insurance to smooth your consumption. 00:50:34.250 --> 00:50:35.930 align:middle line:84% But if you already have the ability 00:50:35.930 --> 00:50:37.770 align:middle line:84% to smooth your consumption, the government 00:50:37.770 --> 00:50:40.490 align:middle line:84% is not helping you out that much. 00:50:40.490 --> 00:50:42.230 align:middle line:90% Now self-insurance is real. 00:50:42.230 --> 00:50:44.690 align:middle line:84% Think about someone who loses their job. 00:50:44.690 --> 00:50:46.690 align:middle line:90% How can they self-insure? 00:50:46.690 --> 00:50:48.850 align:middle line:90% What could they do? 00:50:48.850 --> 00:50:50.650 align:middle line:84% They've lost their wage income, what could 00:50:50.650 --> 00:50:53.650 align:middle line:90% they do so they don't starve? 00:50:53.650 --> 00:50:55.010 align:middle line:90% What's that? 00:50:55.010 --> 00:50:56.770 align:middle line:90% They could have savings. 00:50:56.770 --> 00:50:57.830 align:middle line:90% What else could they do? 00:50:57.830 --> 00:51:01.350 align:middle line:90% 00:51:01.350 --> 00:51:02.350 align:middle line:90% What else could they do? 00:51:02.350 --> 00:51:03.350 align:middle line:90% They could have savings. 00:51:03.350 --> 00:51:05.303 align:middle line:90% What else could they do? 00:51:05.303 --> 00:51:07.470 align:middle line:84% They could find a new-- well, they can find new job. 00:51:07.470 --> 00:51:08.910 align:middle line:84% Let's imagine they can't get a job. 00:51:08.910 --> 00:51:09.550 align:middle line:90% What else could they do? 00:51:09.550 --> 00:51:09.810 align:middle line:90% Yeah? 00:51:09.810 --> 00:51:11.210 align:middle line:84% AUDIENCE: They cd sell their house. 00:51:11.210 --> 00:51:11.710 align:middle line:90% JON GRUBER: They could. 00:51:11.710 --> 00:51:13.030 align:middle line:84% They have assets they could sell. 00:51:13.030 --> 00:51:14.270 align:middle line:84% They could borrow on their credit card. 00:51:14.270 --> 00:51:15.310 align:middle line:90% They might get help from family. 00:51:15.310 --> 00:51:16.770 align:middle line:84% They might get help from friends. 00:51:16.770 --> 00:51:18.710 align:middle line:90% Their spouse might go to work. 00:51:18.710 --> 00:51:20.070 align:middle line:90% There's lots of ways-- 00:51:20.070 --> 00:51:22.930 align:middle line:84% people lose their job, but not suddenly, by definition, 00:51:22.930 --> 00:51:24.190 align:middle line:90% going to starve. 00:51:24.190 --> 00:51:27.270 align:middle line:84% There's lots of other avenues they have that they can turn to. 00:51:27.270 --> 00:51:29.502 align:middle line:90% We call those self-insurance. 00:51:29.502 --> 00:51:31.210 align:middle line:84% In other words, they are private market-- 00:51:31.210 --> 00:51:33.310 align:middle line:84% it's almost a Coasian point, that there 00:51:33.310 --> 00:51:38.710 align:middle line:84% are private market responses to this adverse event. 00:51:38.710 --> 00:51:45.030 align:middle line:84% And if we provide social insurance-- 00:51:45.030 --> 00:51:47.630 align:middle line:90% so let's take a simple example. 00:51:47.630 --> 00:51:53.310 align:middle line:84% Imagine someone is set up such that if they lose their job, 00:51:53.310 --> 00:51:57.790 align:middle line:84% they can easily replace their full income. 00:51:57.790 --> 00:52:00.850 align:middle line:84% They could easily-- if someone lose their job, 00:52:00.850 --> 00:52:03.510 align:middle line:84% they can easily replace their full income. 00:52:03.510 --> 00:52:05.890 align:middle line:84% OK, let's take here's a simple example of that. 00:52:05.890 --> 00:52:07.110 align:middle line:90% My wife's aunt. 00:52:07.110 --> 00:52:09.190 align:middle line:84% My wife's aunt worked for a manufacturing plant 00:52:09.190 --> 00:52:10.870 align:middle line:90% in the Midwest. 00:52:10.870 --> 00:52:12.590 align:middle line:84% The manufacturing plant announced, 00:52:12.590 --> 00:52:18.530 align:middle line:84% every summer we are going to lay you off with perfect certainty. 00:52:18.530 --> 00:52:19.850 align:middle line:90% Well, what did she do? 00:52:19.850 --> 00:52:23.070 align:middle line:84% She saved for nine months of the year 00:52:23.070 --> 00:52:25.870 align:middle line:84% so that she could consume the same amount in the summer. 00:52:25.870 --> 00:52:28.370 align:middle line:84% She could self-smooth her consumption. 00:52:28.370 --> 00:52:29.830 align:middle line:90% I'm a professor. 00:52:29.830 --> 00:52:31.703 align:middle line:84% I only have-- I get paid annually. 00:52:31.703 --> 00:52:32.870 align:middle line:90% Think about school teachers. 00:52:32.870 --> 00:52:35.870 align:middle line:84% My kid's a school teacher, you get paid for 10 months a year. 00:52:35.870 --> 00:52:37.430 align:middle line:84% Does that mean you're going to starve the other two months? 00:52:37.430 --> 00:52:37.730 align:middle line:90% No. 00:52:37.730 --> 00:52:38.850 align:middle line:90% You know with certainty. 00:52:38.850 --> 00:52:42.170 align:middle line:84% You just save so you have the same consumption every month. 00:52:42.170 --> 00:52:43.630 align:middle line:90% You can self-insure. 00:52:43.630 --> 00:52:46.290 align:middle line:84% And indeed, my aunt was a statewide softball pitching 00:52:46.290 --> 00:52:48.332 align:middle line:84% champion during the summer, she had a great time, 00:52:48.332 --> 00:52:50.498 align:middle line:84% and her consumption didn't suffer at all because she 00:52:50.498 --> 00:52:51.970 align:middle line:90% saved all year and knew it. 00:52:51.970 --> 00:52:55.970 align:middle line:84% In that case, what happens when she collects UI? 00:52:55.970 --> 00:52:57.190 align:middle line:90% Let's say UI is available. 00:52:57.190 --> 00:52:59.810 align:middle line:84% She got laid off, UI is available. 00:52:59.810 --> 00:53:01.470 align:middle line:90% What does UI do in that case? 00:53:01.470 --> 00:53:03.830 align:middle line:90% Does it smooth consumption? 00:53:03.830 --> 00:53:04.330 align:middle line:90% No. 00:53:04.330 --> 00:53:06.210 align:middle line:90% What does it do? 00:53:06.210 --> 00:53:07.150 align:middle line:90% Yeah? 00:53:07.150 --> 00:53:08.695 align:middle line:90% AUDIENCE: [INAUDIBLE] 00:53:08.695 --> 00:53:10.570 align:middle line:84% JON GRUBER: Well, no, no, no, but you already 00:53:10.570 --> 00:53:12.408 align:middle line:84% have-- you're already start consuming less. 00:53:12.408 --> 00:53:13.700 align:middle line:90% What does it do to your budget? 00:53:13.700 --> 00:53:15.440 align:middle line:84% What does it do to your decisions? 00:53:15.440 --> 00:53:17.320 align:middle line:84% How does it affect your behavior? 00:53:17.320 --> 00:53:18.400 align:middle line:90% Yeah? 00:53:18.400 --> 00:53:21.480 align:middle line:84% AUDIENCE: It increases your budget in a way that you-- 00:53:21.480 --> 00:53:22.800 align:middle line:90% JON GRUBER: Well, it could. 00:53:22.800 --> 00:53:26.300 align:middle line:84% But remember the first rule of government intervention. 00:53:26.300 --> 00:53:29.480 align:middle line:84% If you can undo a government intervention, you always will. 00:53:29.480 --> 00:53:32.560 align:middle line:84% How can you perfectly undo the government intervention here? 00:53:32.560 --> 00:53:35.340 align:middle line:90% Thinking back to fireworks. 00:53:35.340 --> 00:53:36.200 align:middle line:90% AUDIENCE: Save less. 00:53:36.200 --> 00:53:37.640 align:middle line:90% JON GRUBER: Save less. 00:53:37.640 --> 00:53:44.000 align:middle line:84% So indeed, full self-insurance is the same as full crowdout. 00:53:44.000 --> 00:53:46.720 align:middle line:90% 00:53:46.720 --> 00:53:52.560 align:middle line:84% That all UI is doing here is crowding out her savings. 00:53:52.560 --> 00:53:55.860 align:middle line:84% All UI is doing is crowding out her savings. 00:53:55.860 --> 00:53:58.680 align:middle line:90% 00:53:58.680 --> 00:54:01.960 align:middle line:84% It is doing no consumption smoothing. 00:54:01.960 --> 00:54:04.760 align:middle line:84% Either way, she consume the same amount all year. 00:54:04.760 --> 00:54:08.960 align:middle line:84% The difference has to come from her savings or the government. 00:54:08.960 --> 00:54:14.020 align:middle line:84% But there's no consumption being done by the UI program. 00:54:14.020 --> 00:54:16.740 align:middle line:90% Contrast that to another case. 00:54:16.740 --> 00:54:21.740 align:middle line:84% Imagine I'm someone who's just like, I have no-- 00:54:21.740 --> 00:54:23.480 align:middle line:84% I literally cannot afford to save. 00:54:23.480 --> 00:54:25.188 align:middle line:84% Literally, I'm at a job where literally I 00:54:25.188 --> 00:54:26.780 align:middle line:84% pay my rent and my bills, I literally 00:54:26.780 --> 00:54:30.060 align:middle line:84% have to spend my salary every week. 00:54:30.060 --> 00:54:31.600 align:middle line:84% I have no friends, I have no family, 00:54:31.600 --> 00:54:33.780 align:middle line:84% I have no credit card because I have no good credit, 00:54:33.780 --> 00:54:35.700 align:middle line:90% and I have no savings. 00:54:35.700 --> 00:54:37.860 align:middle line:90% Now I lose my job. 00:54:37.860 --> 00:54:40.900 align:middle line:90% In that case, I will die. 00:54:40.900 --> 00:54:44.020 align:middle line:84% I am unable to smooth my consumption. 00:54:44.020 --> 00:54:48.960 align:middle line:84% In that case, the government-- in that case, 00:54:48.960 --> 00:54:56.080 align:middle line:84% there is no self-insurance, so as a result, I get no crowdout. 00:54:56.080 --> 00:54:57.360 align:middle line:90% There's nothing to crowd out. 00:54:57.360 --> 00:55:00.740 align:middle line:90% 00:55:00.740 --> 00:55:03.640 align:middle line:84% But I get full consumption smoothing. 00:55:03.640 --> 00:55:07.178 align:middle line:90% 00:55:07.178 --> 00:55:08.720 align:middle line:84% Whereas when I have full self-insure, 00:55:08.720 --> 00:55:11.300 align:middle line:84% I get full crowdout and no consumption smoothing. 00:55:11.300 --> 00:55:13.840 align:middle line:90% 00:55:13.840 --> 00:55:16.520 align:middle line:84% So look what we have here, we have a trade-off. 00:55:16.520 --> 00:55:20.040 align:middle line:84% The more self-insurance you have, 00:55:20.040 --> 00:55:24.180 align:middle line:84% the more government insurance is not smoothing your consumption, 00:55:24.180 --> 00:55:27.720 align:middle line:84% it's just crowding out your self-insurance. 00:55:27.720 --> 00:55:31.000 align:middle line:84% And since the goal of social insurance to smooth consumption, 00:55:31.000 --> 00:55:34.800 align:middle line:84% it's less valuable the more self-insurance you have. 00:55:34.800 --> 00:55:41.200 align:middle line:84% UI is worth nothing to my aunt in terms 00:55:41.200 --> 00:55:42.300 align:middle line:90% of smoothing consumption. 00:55:42.300 --> 00:55:43.560 align:middle line:90% It makes her wealthier. 00:55:43.560 --> 00:55:46.700 align:middle line:84% That's just a transfer from the government to her. 00:55:46.700 --> 00:55:48.920 align:middle line:84% But there's nothing in terms of her desired 00:55:48.920 --> 00:55:51.800 align:middle line:84% consumption-- in terms of her consumption smoothing. 00:55:51.800 --> 00:55:53.440 align:middle line:84% All it does is crowded her savings. 00:55:53.440 --> 00:55:56.760 align:middle line:84% For the poor sap I just described a few minutes ago, 00:55:56.760 --> 00:55:59.220 align:middle line:90% this person, UI is vital. 00:55:59.220 --> 00:56:01.253 align:middle line:84% It's fully smoothing their consumption, 00:56:01.253 --> 00:56:02.920 align:middle line:84% but it's not crowding any self-interests 00:56:02.920 --> 00:56:05.640 align:middle line:84% because no self-insurance to crowd out. 00:56:05.640 --> 00:56:08.380 align:middle line:84% So it's the same concept as fireworks. 00:56:08.380 --> 00:56:11.000 align:middle line:84% But instead of crowding out Ben & Jerry's contributions, 00:56:11.000 --> 00:56:15.620 align:middle line:84% it's crowding out what you're doing to protect yourself. 00:56:15.620 --> 00:56:17.380 align:middle line:90% What does that mean? 00:56:17.380 --> 00:56:20.340 align:middle line:84% That means that the value of social insurance 00:56:20.340 --> 00:56:24.820 align:middle line:84% is determined by how much self-insurance you have. 00:56:24.820 --> 00:56:27.455 align:middle line:84% We assumed, in this example, that each dollar 00:56:27.455 --> 00:56:29.580 align:middle line:84% of social insurance was providing $1 of consumption 00:56:29.580 --> 00:56:31.915 align:middle line:90% smoothing, but that's not true. 00:56:31.915 --> 00:56:34.540 align:middle line:84% Each dollar of social insurance is providing less than a dollar 00:56:34.540 --> 00:56:36.915 align:middle line:84% of consumption smoothing because part of what it's doing, 00:56:36.915 --> 00:56:39.700 align:middle line:84% it's just crowding out your self-insurance. 00:56:39.700 --> 00:56:42.860 align:middle line:84% So one reason why social insurance is less valuable 00:56:42.860 --> 00:56:46.660 align:middle line:84% than you might think is that all it's doing 00:56:46.660 --> 00:56:50.500 align:middle line:84% is crowding out your self-insurance. 00:56:50.500 --> 00:56:54.300 align:middle line:84% So it's less-- so that's the first point to remember. 00:56:54.300 --> 00:56:58.203 align:middle line:84% And what is going to determine how much crowdout 00:56:58.203 --> 00:57:00.620 align:middle line:84% there is, or conversely, how much consumption smoothing is 00:57:00.620 --> 00:57:01.380 align:middle line:90% done? 00:57:01.380 --> 00:57:04.580 align:middle line:84% What's going to determine that is two things. 00:57:04.580 --> 00:57:10.320 align:middle line:84% The first of all is how predictable is the event? 00:57:10.320 --> 00:57:14.320 align:middle line:84% And the second is, how severe is the event? 00:57:14.320 --> 00:57:20.160 align:middle line:84% If an event-- if I said to you, for sure, 00:57:20.160 --> 00:57:25.760 align:middle line:84% every year I'm going to take $1 from you, 00:57:25.760 --> 00:57:28.080 align:middle line:90% you will need to insure that. 00:57:28.080 --> 00:57:30.920 align:middle line:84% You just know that I just basically 00:57:30.920 --> 00:57:34.897 align:middle line:84% have to save a penny less than-- $0.08 a month. 00:57:34.897 --> 00:57:36.980 align:middle line:84% At the end of the year, when they take the dollar, 00:57:36.980 --> 00:57:39.520 align:middle line:84% I've got a dollar of savings to replace it. 00:57:39.520 --> 00:57:42.040 align:middle line:84% Social insurance would be useless. 00:57:42.040 --> 00:57:45.960 align:middle line:84% But if I said to you, randomly, you're going to be hit by a car 00:57:45.960 --> 00:57:48.345 align:middle line:84% and it's going to cost you $50,000, well, 00:57:48.345 --> 00:57:50.720 align:middle line:84% that's going to be hard to self-insure unless you're very 00:57:50.720 --> 00:57:54.480 align:middle line:84% rich because you need to have an extra 50K of savings sitting 00:57:54.480 --> 00:57:58.480 align:middle line:90% around at all times. 00:57:58.480 --> 00:58:00.340 align:middle line:84% That's going to make it hard to self-insure. 00:58:00.340 --> 00:58:02.420 align:middle line:84% So basically, the more predictable 00:58:02.420 --> 00:58:04.923 align:middle line:84% is the event, the more you can self-insure like my aunt. 00:58:04.923 --> 00:58:05.840 align:middle line:90% That was a major loss. 00:58:05.840 --> 00:58:08.640 align:middle line:84% She was losing three months of income every year, 00:58:08.640 --> 00:58:11.020 align:middle line:90% but it was predictable. 00:58:11.020 --> 00:58:15.970 align:middle line:84% And how big it is, because even it's unpredictable, 00:58:15.970 --> 00:58:17.720 align:middle line:84% if it's not big, it doesn't really matter, 00:58:17.720 --> 00:58:18.887 align:middle line:90% you can just self-insure it. 00:58:18.887 --> 00:58:20.420 align:middle line:84% So my dollar example, imagine I said 00:58:20.420 --> 00:58:22.170 align:middle line:84% I'm going to take $1 for you, but randomly 00:58:22.170 --> 00:58:24.020 align:middle line:90% throughout the year. 00:58:24.020 --> 00:58:26.120 align:middle line:84% Or I could say it could be $1 or $2. 00:58:26.120 --> 00:58:29.980 align:middle line:84% You're like, I don't care, I can just make up for that. 00:58:29.980 --> 00:58:33.780 align:middle line:84% So self-insurance will work best when events 00:58:33.780 --> 00:58:35.780 align:middle line:90% are predictable and not severe. 00:58:35.780 --> 00:58:38.460 align:middle line:84% But the less predictable, the more severe, the less effective 00:58:38.460 --> 00:58:39.720 align:middle line:90% will self-insurance be. 00:58:39.720 --> 00:58:42.700 align:middle line:90% 00:58:42.700 --> 00:58:49.300 align:middle line:84% Now, that is one reason why our model isn't quite so simple. 00:58:49.300 --> 00:58:52.900 align:middle line:84% The other reason is more important, 00:58:52.900 --> 00:58:58.620 align:middle line:84% which is that insurance also has a cost. 00:58:58.620 --> 00:59:00.890 align:middle line:84% And the cost of insurance is what 00:59:00.890 --> 00:59:06.930 align:middle line:84% we call moral hazard, a term you guys might have heard bandied 00:59:06.930 --> 00:59:10.090 align:middle line:90% about in various contexts. 00:59:10.090 --> 00:59:15.690 align:middle line:84% Moral hazard, to be specific, is the adverse behavior 00:59:15.690 --> 00:59:19.730 align:middle line:90% encouraged by insurance. 00:59:19.730 --> 00:59:22.410 align:middle line:84% The adverse behavior encouraged by insurance 00:59:22.410 --> 00:59:25.410 align:middle line:90% is what we call moral hazard. 00:59:25.410 --> 00:59:29.290 align:middle line:84% Basically, the idea is if you're insured 00:59:29.290 --> 00:59:31.650 align:middle line:84% against your house burning down, you 00:59:31.650 --> 00:59:34.560 align:middle line:84% don't worry about keeping your fire extinguisher up to date. 00:59:34.560 --> 00:59:36.310 align:middle line:84% If you're insured against a bike accident, 00:59:36.310 --> 00:59:39.050 align:middle line:84% you might bike a little more recklessly. 00:59:39.050 --> 00:59:43.010 align:middle line:84% Basically, any action you take, anytime insurance 00:59:43.010 --> 00:59:48.490 align:middle line:84% changes your behavior, that's the moral hazard response. 00:59:48.490 --> 00:59:51.250 align:middle line:84% Anytime it changes your risk-taking behavior 00:59:51.250 --> 00:59:53.990 align:middle line:84% or your protective behavior, that's a moral hazard response. 00:59:53.990 --> 00:59:55.670 align:middle line:84% In fact, Shakespeare wrote about this. 00:59:55.670 --> 00:59:59.550 align:middle line:84% He said nothing emboldens sin so much as is mercy. 00:59:59.550 --> 01:00:01.470 align:middle line:84% That basically, the more you're protected, 01:00:01.470 --> 01:00:04.430 align:middle line:90% the more you might fuck about. 01:00:04.430 --> 01:00:09.750 align:middle line:84% Now, what that means is that if there is moral hazard, 01:00:09.750 --> 01:00:14.310 align:middle line:84% full social insurance is no longer optimal. 01:00:14.310 --> 01:00:18.350 align:middle line:84% It's no longer optimal because now there's a trade-off. 01:00:18.350 --> 01:00:20.910 align:middle line:84% On the one hand, insurance is smoothing consumption 01:00:20.910 --> 01:00:23.550 align:middle line:84% depending on how much self-insurance you have. 01:00:23.550 --> 01:00:26.710 align:middle line:84% On the other hand, insurance is inducing 01:00:26.710 --> 01:00:29.975 align:middle line:90% moral hazard, which is costly. 01:00:29.975 --> 01:00:32.350 align:middle line:84% And there's no better way to see this than to think about 01:00:32.350 --> 01:00:34.070 align:middle line:84% the workers' compensation program, 01:00:34.070 --> 01:00:35.490 align:middle line:84% a program we'll talk about in a couple of lectures, 01:00:35.490 --> 01:00:37.230 align:middle line:84% probably one of the social insurance programs you're least 01:00:37.230 --> 01:00:39.730 align:middle line:84% familiar with, but it's actually the oldest social insurance 01:00:39.730 --> 01:00:42.150 align:middle line:84% program in America, started in the beginning of the 20th 01:00:42.150 --> 01:00:44.567 align:middle line:84% century, before Social Security and other programs we know 01:00:44.567 --> 01:00:45.110 align:middle line:90% about. 01:00:45.110 --> 01:00:48.790 align:middle line:84% This is a program that, if you get hurt on the job, 01:00:48.790 --> 01:00:49.930 align:middle line:90% pays for cash benefits. 01:00:49.930 --> 01:00:51.930 align:middle line:90% You Know your Aflac ads? 01:00:51.930 --> 01:00:53.750 align:middle line:90% Aflac is like a top of this. 01:00:53.750 --> 01:00:56.410 align:middle line:84% This is the stuff, which is a basic government program, 01:00:56.410 --> 01:00:58.610 align:middle line:84% that if you get hurt on the job, they 01:00:58.610 --> 01:01:00.525 align:middle line:90% replace some of your income. 01:01:00.525 --> 01:01:02.150 align:middle line:84% They also provide you medical benefits, 01:01:02.150 --> 01:01:03.308 align:middle line:90% we'll come back to that. 01:01:03.308 --> 01:01:05.350 align:middle line:84% But fundamentally, for ease of thinking about it, 01:01:05.350 --> 01:01:06.470 align:middle line:90% forget the medical piece. 01:01:06.470 --> 01:01:07.830 align:middle line:90% Focus on the cash piece. 01:01:07.830 --> 01:01:09.470 align:middle line:84% The idea is if you get hurt on the job, 01:01:09.470 --> 01:01:12.010 align:middle line:84% they'll replace some of your income. 01:01:12.010 --> 01:01:14.137 align:middle line:84% Now the Aflac ads will be like, well, 01:01:14.137 --> 01:01:16.470 align:middle line:84% they'll replace of your income, so you should buy Aflac, 01:01:16.470 --> 01:01:18.770 align:middle line:84% and here's why you should never buy Aflac. 01:01:18.770 --> 01:01:21.330 align:middle line:84% Because the truth is, they replace almost all 01:01:21.330 --> 01:01:23.210 align:middle line:90% of your income. 01:01:23.210 --> 01:01:26.330 align:middle line:84% Worker's compensation insurance replaces 01:01:26.330 --> 01:01:28.370 align:middle line:84% about three-quarters of your income 01:01:28.370 --> 01:01:30.570 align:middle line:84% that you lose because you're hurt. 01:01:30.570 --> 01:01:34.110 align:middle line:84% And it's not taxed, while your income is. 01:01:34.110 --> 01:01:37.010 align:middle line:84% So if you do the math, which we'll do in a few lectures, 01:01:37.010 --> 01:01:40.410 align:middle line:84% turns out, you make about almost $0.90 on the dollar when 01:01:40.410 --> 01:01:43.370 align:middle line:84% you're hurt compared to when you're working. 01:01:43.370 --> 01:01:45.890 align:middle line:84% Why do you need to buy Aflac at some high premium 01:01:45.890 --> 01:01:47.330 align:middle line:90% to top that up? 01:01:47.330 --> 01:01:52.070 align:middle line:84% The truth is, workers' comp is already pretty damn good. 01:01:52.070 --> 01:01:56.310 align:middle line:84% Now, the problem with that is, it's also 01:01:56.310 --> 01:01:58.383 align:middle line:84% true with workers' comp that it's actually 01:01:58.383 --> 01:02:00.550 align:middle line:84% kind of hard sometimes to tell if someone was really 01:02:00.550 --> 01:02:02.450 align:middle line:90% hurt on the job. 01:02:02.450 --> 01:02:02.950 align:middle line:90% Why? 01:02:02.950 --> 01:02:04.830 align:middle line:84% Because most workers' comp is not 01:02:04.830 --> 01:02:06.670 align:middle line:84% lacerations-- your arm cut off, it's 01:02:06.670 --> 01:02:10.750 align:middle line:84% mostly back pain and mental health issues. 01:02:10.750 --> 01:02:13.710 align:middle line:84% Those are very hard to adjudicate, 01:02:13.710 --> 01:02:15.310 align:middle line:84% A, if they're even happening; and B, 01:02:15.310 --> 01:02:18.270 align:middle line:90% If they were caused at work. 01:02:18.270 --> 01:02:21.110 align:middle line:84% As a result, you have something where 01:02:21.110 --> 01:02:25.110 align:middle line:84% it's pretty easy to make it up, and it's very rewarding 01:02:25.110 --> 01:02:27.350 align:middle line:90% if you do. 01:02:27.350 --> 01:02:29.690 align:middle line:90% And that can cause moral hazard. 01:02:29.690 --> 01:02:32.470 align:middle line:84% That can cause people to claim benefits 01:02:32.470 --> 01:02:36.230 align:middle line:84% and stop working even when they're not really hurt. 01:02:36.230 --> 01:02:40.052 align:middle line:84% In the next chapter, we're going to talk about Social Security. 01:02:40.052 --> 01:02:42.510 align:middle line:84% I'll highlight the end of the lecture that in some European 01:02:42.510 --> 01:02:47.570 align:middle line:84% countries, if you work, you make a $1, and if you don't work, 01:02:47.570 --> 01:02:50.070 align:middle line:90% you make $1.40. 01:02:50.070 --> 01:02:51.210 align:middle line:90% What do you think happens? 01:02:51.210 --> 01:02:53.650 align:middle line:90% No one works. 01:02:53.650 --> 01:02:55.330 align:middle line:84% Moral hazard is a real consequence 01:02:55.330 --> 01:02:56.630 align:middle line:90% of these types of programs. 01:02:56.630 --> 01:03:00.290 align:middle line:84% Worker's comp is fun because all sorts of fun examples. 01:03:00.290 --> 01:03:02.588 align:middle line:84% You guys ever hear about a show called Wicked Tuna? 01:03:02.588 --> 01:03:04.130 align:middle line:84% It's a National Geographic about guys 01:03:04.130 --> 01:03:05.510 align:middle line:90% who fish for these giant tuna. 01:03:05.510 --> 01:03:07.210 align:middle line:84% Turns out, the guy in the show was actually collecting 01:03:07.210 --> 01:03:09.543 align:middle line:84% workers' comp at the same time as he was wrestling these 01:03:09.543 --> 01:03:13.450 align:middle line:84% 3,000-pound tuna fish and he got busted. 01:03:13.450 --> 01:03:17.630 align:middle line:84% There was another guy who was collecting workers' comp, 01:03:17.630 --> 01:03:20.410 align:middle line:84% was also a tandem skydiving instructor at the same time. 01:03:20.410 --> 01:03:22.510 align:middle line:84% But my favorite example, which comes from, 01:03:22.510 --> 01:03:28.130 align:middle line:84% of course, Florida, where all bad behavior in America 01:03:28.130 --> 01:03:30.610 align:middle line:90% happens-- 01:03:30.610 --> 01:03:34.290 align:middle line:84% not all bad behavior, but a high percentage. 01:03:34.290 --> 01:03:36.690 align:middle line:84% A woman was working at her desk in Fort Lauderdale 01:03:36.690 --> 01:03:40.450 align:middle line:84% when she noticed a ceiling fixture had fallen on her desk. 01:03:40.450 --> 01:03:42.808 align:middle line:84% So she looks around, picks it up, bonks herself 01:03:42.808 --> 01:03:45.350 align:middle line:84% in the head with it, and says, oh my God, it fell on my head, 01:03:45.350 --> 01:03:47.230 align:middle line:84% I need workers' comp, not realizing, 01:03:47.230 --> 01:03:48.730 align:middle line:84% of course, the reason I can tell you 01:03:48.730 --> 01:03:50.870 align:middle line:84% this story is it was all on film. 01:03:50.870 --> 01:03:52.732 align:middle line:84% So, there is-- now, if it wasn't on film, 01:03:52.732 --> 01:03:54.190 align:middle line:84% she would have gotten workers' comp 01:03:54.190 --> 01:03:55.590 align:middle line:84% because how are you going to verify the thing bonked her 01:03:55.590 --> 01:03:57.830 align:middle line:84% on the head rather than falling on her desk? 01:03:57.830 --> 01:04:00.570 align:middle line:84% The bottom line is, moral hazard is real. 01:04:00.570 --> 01:04:02.950 align:middle line:84% Now what determines moral hazard? 01:04:02.950 --> 01:04:06.510 align:middle line:84% Moral hazard will be determined by two things. 01:04:06.510 --> 01:04:11.950 align:middle line:90% The first is the observability. 01:04:11.950 --> 01:04:13.650 align:middle line:84% So what will determine moral hazard? 01:04:13.650 --> 01:04:16.570 align:middle line:90% The first is observability. 01:04:16.570 --> 01:04:19.390 align:middle line:90% 01:04:19.390 --> 01:04:21.950 align:middle line:84% How easy it is to see whether you really need the thing 01:04:21.950 --> 01:04:25.350 align:middle line:90% you're qualifying for. 01:04:25.350 --> 01:04:27.930 align:middle line:84% Like, if it's your arm cut off, I can see it pretty easily. 01:04:27.930 --> 01:04:31.630 align:middle line:84% If it's a back pain, it's hard to verify. 01:04:31.630 --> 01:04:36.390 align:middle line:84% And the second is the financial incentive. 01:04:36.390 --> 01:04:38.070 align:middle line:84% Because basically, fundamentally, people 01:04:38.070 --> 01:04:40.350 align:middle line:90% are pretty honest. 01:04:40.350 --> 01:04:42.410 align:middle line:84% So unless there's a-- makes it worth it, 01:04:42.410 --> 01:04:46.190 align:middle line:84% they're not going to try to pretend they're hurt. 01:04:46.190 --> 01:04:49.620 align:middle line:84% Just to emphasize that the second can be important, 01:04:49.620 --> 01:04:52.300 align:middle line:84% let's talk about the most disgusting case of moral hazard 01:04:52.300 --> 01:05:00.580 align:middle line:84% in American history, which is that in the 1950s and 1960s, 01:05:00.580 --> 01:05:04.700 align:middle line:84% the investigators determined that two-thirds of all claims 01:05:04.700 --> 01:05:09.060 align:middle line:84% for losing limbs came from one area of Florida. 01:05:09.060 --> 01:05:11.540 align:middle line:84% And it turned out, people were cutting off their limbs 01:05:11.540 --> 01:05:12.740 align:middle line:90% to get benefits. 01:05:12.740 --> 01:05:15.140 align:middle line:90% They even called it Nub City. 01:05:15.140 --> 01:05:18.968 align:middle line:84% So it turns out, even though it was observable, 01:05:18.968 --> 01:05:21.260 align:middle line:84% there was some moral hazard because the benefit was big 01:05:21.260 --> 01:05:23.500 align:middle line:84% enough relative to people's income. 01:05:23.500 --> 01:05:27.580 align:middle line:84% So basically, the point is, moral hazard is real, 01:05:27.580 --> 01:05:30.020 align:middle line:84% it can be pernicious, and it's going 01:05:30.020 --> 01:05:33.420 align:middle line:90% to depend on these two things. 01:05:33.420 --> 01:05:38.380 align:middle line:84% So that's the first thing about moral hazard, 01:05:38.380 --> 01:05:40.900 align:middle line:84% is what causes it is going to be observability 01:05:40.900 --> 01:05:42.900 align:middle line:90% and financial incentives. 01:05:42.900 --> 01:05:44.940 align:middle line:84% The second thing about moral hazard 01:05:44.940 --> 01:05:47.960 align:middle line:90% is that it multidimensional. 01:05:47.960 --> 01:05:52.080 align:middle line:90% That's what causes it. 01:05:52.080 --> 01:05:54.100 align:middle line:84% The second thing it has is multidimensional. 01:05:54.100 --> 01:05:57.040 align:middle line:84% There could be multiple types of moral hazard. 01:05:57.040 --> 01:05:59.520 align:middle line:84% The types we typically think about 01:05:59.520 --> 01:06:06.160 align:middle line:84% are reduced precaution, or what we might 01:06:06.160 --> 01:06:09.440 align:middle line:90% call ex-ante moral hazard. 01:06:09.440 --> 01:06:12.540 align:middle line:84% This is the, I don't have a fire extinguisher in my house, 01:06:12.540 --> 01:06:16.400 align:middle line:84% or I don't ride my bike as carefully. 01:06:16.400 --> 01:06:18.180 align:middle line:84% Or when I'm climbing the ladder at work, 01:06:18.180 --> 01:06:21.200 align:middle line:84% I rush up instead of taking my time. 01:06:21.200 --> 01:06:25.160 align:middle line:84% That's the classic example of moral hazard we teach, 01:06:25.160 --> 01:06:28.578 align:middle line:84% but it's actually probably by far the least important. 01:06:28.578 --> 01:06:30.620 align:middle line:84% Most people don't want to get hit on their bikes. 01:06:30.620 --> 01:06:32.640 align:middle line:84% Most people don't want their house to burn down. 01:06:32.640 --> 01:06:35.040 align:middle line:84% Most people don't want to get hurt at work. 01:06:35.040 --> 01:06:39.080 align:middle line:84% Except maybe in Nub City, So the bottom line is-- 01:06:39.080 --> 01:06:40.980 align:middle line:84% the bottom line is that basically, 01:06:40.980 --> 01:06:44.140 align:middle line:84% while this is the easiest case to think about as an example, 01:06:44.140 --> 01:06:46.260 align:middle line:84% it's probably the least important case. 01:06:46.260 --> 01:06:49.300 align:middle line:84% The second case is increased odds 01:06:49.300 --> 01:06:50.960 align:middle line:90% of entering the adverse state. 01:06:50.960 --> 01:06:54.380 align:middle line:90% 01:06:54.380 --> 01:06:57.800 align:middle line:90% Odds of adverse event. 01:06:57.800 --> 01:07:01.300 align:middle line:90% 01:07:01.300 --> 01:07:03.960 align:middle line:90% Increased odds of adverse event. 01:07:03.960 --> 01:07:07.060 align:middle line:90% 01:07:07.060 --> 01:07:12.400 align:middle line:84% This is people not trying harder in their job. 01:07:12.400 --> 01:07:15.235 align:middle line:84% So it's like, this is reduced precaution-- let me separate-- 01:07:15.235 --> 01:07:16.360 align:middle line:90% these are hard to separate. 01:07:16.360 --> 01:07:17.440 align:middle line:90% Let me be more precise. 01:07:17.440 --> 01:07:21.740 align:middle line:90% Reduce precaution is I don't-- 01:07:21.740 --> 01:07:25.740 align:middle line:84% because I have fire insurance, I don't buy a fire extinguisher. 01:07:25.740 --> 01:07:29.580 align:middle line:84% Increased odds of any adverse state is, because this is-- 01:07:29.580 --> 01:07:31.000 align:middle line:90% I actually misspoke here. 01:07:31.000 --> 01:07:34.620 align:middle line:84% This is the, I'm not careful-- this is the, 01:07:34.620 --> 01:07:37.460 align:middle line:84% I don't worry about keeping my job because I 01:07:37.460 --> 01:07:39.100 align:middle line:90% have unemployment insurance. 01:07:39.100 --> 01:07:43.023 align:middle line:84% So clumsy on the ladder is actually in this bucket. 01:07:43.023 --> 01:07:44.440 align:middle line:84% So this is literally, I don't take 01:07:44.440 --> 01:07:45.740 align:middle line:90% protective things I could do. 01:07:45.740 --> 01:07:49.760 align:middle line:84% So this would be, I don't buy a really good ladder 01:07:49.760 --> 01:07:51.148 align:middle line:90% to make sure I don't fall. 01:07:51.148 --> 01:07:52.940 align:middle line:84% This would be when I'm climbing the ladder, 01:07:52.940 --> 01:07:55.440 align:middle line:84% I'm just figure, what the hell, I'll fall off, I don't care. 01:07:55.440 --> 01:07:58.040 align:middle line:84% Or I don't really care about who's in my job. 01:07:58.040 --> 01:08:00.820 align:middle line:84% The difference is really in the timing. 01:08:00.820 --> 01:08:02.520 align:middle line:84% Do you take precautionary activities 01:08:02.520 --> 01:08:07.280 align:middle line:84% or while you're doing the thing, are you careful enough? 01:08:07.280 --> 01:08:13.320 align:middle line:84% The third example that may actually be the most costly 01:08:13.320 --> 01:08:26.720 align:middle line:84% is increased spending while under social insurance. 01:08:26.720 --> 01:08:30.160 align:middle line:84% This is the idea that, OK, I lost my job. 01:08:30.160 --> 01:08:31.075 align:middle line:90% I got hurt. 01:08:31.075 --> 01:08:32.700 align:middle line:84% Now I'm not going to try to get better. 01:08:32.700 --> 01:08:34.460 align:middle line:84% I'm not going to try to find a new job. 01:08:34.460 --> 01:08:35.760 align:middle line:90% I'm just going to hang out. 01:08:35.760 --> 01:08:39.660 align:middle line:84% I lost my job, so basically, I didn't try hard, 01:08:39.660 --> 01:08:42.060 align:middle line:84% I entered the state, and now that I'm in the state, 01:08:42.060 --> 01:08:43.380 align:middle line:84% I'm getting checks every week, I'm just 01:08:43.380 --> 01:08:44.755 align:middle line:84% going to sit around and watch TV, 01:08:44.755 --> 01:08:46.460 align:middle line:84% I'm not going to look for a new job. 01:08:46.460 --> 01:08:48.800 align:middle line:84% Or I fell off the ladder because I wasn't being careful, 01:08:48.800 --> 01:08:50.600 align:middle line:84% now that I'm home collecting 90% of my wages, 01:08:50.600 --> 01:08:52.058 align:middle line:84% I'm not going to try to get better. 01:08:52.058 --> 01:08:53.300 align:middle line:90% I'm not going to go to PT. 01:08:53.300 --> 01:08:54.675 align:middle line:84% I'm not going to do things to get 01:08:54.675 --> 01:08:57.500 align:middle line:90% me better because why bother? 01:08:57.500 --> 01:09:00.340 align:middle line:84% That can end up being very expensive as well because that 01:09:00.340 --> 01:09:03.640 align:middle line:84% increases the duration of your social insurance expenditures. 01:09:03.640 --> 01:09:06.500 align:middle line:84% If everyone fell off the ladder was back at work three days 01:09:06.500 --> 01:09:08.819 align:middle line:90% later, we wouldn't really care. 01:09:08.819 --> 01:09:11.433 align:middle line:84% We care because they're out of work for six to 10 weeks. 01:09:11.433 --> 01:09:13.100 align:middle line:84% And the question is, could they actually 01:09:13.100 --> 01:09:16.080 align:middle line:84% be back to work three days later, but they don't bother? 01:09:16.080 --> 01:09:17.580 align:middle line:84% That's another type of moral hazard, 01:09:17.580 --> 01:09:22.460 align:middle line:84% is extending your stay in the social insurance state. 01:09:22.460 --> 01:09:27.520 align:middle line:84% And then finally, we have the other interesting case, 01:09:27.520 --> 01:09:29.560 align:middle line:90% which is supplier responses. 01:09:29.560 --> 01:09:33.020 align:middle line:90% 01:09:33.020 --> 01:09:37.380 align:middle line:84% This is the idea that basically, because-- 01:09:37.380 --> 01:09:41.240 align:middle line:84% this is the idea that because someone has insurance, 01:09:41.240 --> 01:09:42.140 align:middle line:90% I spend more-- 01:09:42.140 --> 01:09:43.920 align:middle line:90% I make more money on them. 01:09:43.920 --> 01:09:45.435 align:middle line:90% So a physician-- 01:09:45.435 --> 01:09:47.560 align:middle line:84% I don't-- you guys probably haven't taken your cars 01:09:47.560 --> 01:09:48.760 align:middle line:90% for repairs much. 01:09:48.760 --> 01:09:50.340 align:middle line:84% When you take your cars for repairs, 01:09:50.340 --> 01:09:52.632 align:middle line:84% the mechanic will ask you, is this for insurance or not 01:09:52.632 --> 01:09:53.600 align:middle line:90% for insurance? 01:09:53.600 --> 01:09:55.880 align:middle line:84% So insurance, they'll bill you more than 01:09:55.880 --> 01:09:59.080 align:middle line:84% if it's not for insurance because you don't care. 01:09:59.080 --> 01:10:03.600 align:middle line:84% So you're not price-sensitive, so they'll just charge more. 01:10:03.600 --> 01:10:05.480 align:middle line:84% Doctors will do more if you're insured 01:10:05.480 --> 01:10:08.320 align:middle line:84% than you're uninsured because you don't care. 01:10:08.320 --> 01:10:12.800 align:middle line:84% And so that is an example of another way 01:10:12.800 --> 01:10:17.600 align:middle line:84% that moral hazard can increase government costs, 01:10:17.600 --> 01:10:20.820 align:middle line:84% is through these type of supplier responses. 01:10:20.820 --> 01:10:23.400 align:middle line:90% 01:10:23.400 --> 01:10:27.720 align:middle line:84% My job after freshman year of MIT, 01:10:27.720 --> 01:10:30.920 align:middle line:84% I worked as a filing clerk in-- actually 01:10:30.920 --> 01:10:33.520 align:middle line:84% started IAP of my freshman year, in the summer of my freshman 01:10:33.520 --> 01:10:33.860 align:middle line:90% year. 01:10:33.860 --> 01:10:35.610 align:middle line:84% I was a filing clerk at a doctor's office, 01:10:35.610 --> 01:10:36.798 align:middle line:90% I did workers' comp. 01:10:36.798 --> 01:10:38.590 align:middle line:84% And eventually I was good enough at filing, 01:10:38.590 --> 01:10:40.790 align:middle line:84% they promoted me to assistant bookkeeper. 01:10:40.790 --> 01:10:42.890 align:middle line:84% Assistant bookkeeper, and then I got fired. 01:10:42.890 --> 01:10:44.510 align:middle line:90% Only time I've ever been fired. 01:10:44.510 --> 01:10:47.790 align:middle line:84% It turned out, much to my justification, 01:10:47.790 --> 01:10:50.123 align:middle line:84% the doctor then went to jail, along with his accountant. 01:10:50.123 --> 01:10:51.540 align:middle line:84% The reason they fired me, I think, 01:10:51.540 --> 01:10:53.790 align:middle line:84% is because they thought I might catch them-- maybe. 01:10:53.790 --> 01:10:56.790 align:middle line:84% Because the doctor was running this incredible scam where 01:10:56.790 --> 01:11:00.350 align:middle line:84% basically, people would come in for workers' comp 01:11:00.350 --> 01:11:03.470 align:middle line:84% claims, and he would just say yes. 01:11:03.470 --> 01:11:04.970 align:middle line:84% He wouldn't even really look at him. 01:11:04.970 --> 01:11:07.530 align:middle line:84% He would see one person every three minutes all day-- 01:11:07.530 --> 01:11:10.070 align:middle line:84% it was one doctor with five full-time assistants-- 01:11:10.070 --> 01:11:12.710 align:middle line:84% full-time secretaries, that should have given some hint, 01:11:12.710 --> 01:11:15.252 align:middle line:84% that literally, all he did was just people come in, he'd say, 01:11:15.252 --> 01:11:17.030 align:middle line:90% yes, you're hurt, give me-- 01:11:17.030 --> 01:11:18.030 align:middle line:90% I'll bill you now. 01:11:18.030 --> 01:11:20.190 align:middle line:84% And just do that over and over and over again. 01:11:20.190 --> 01:11:22.950 align:middle line:84% So that was a supplier response that increased the moral hazard, 01:11:22.950 --> 01:11:25.470 align:middle line:84% affects this program, and I lost my job. 01:11:25.470 --> 01:11:28.750 align:middle line:90% So that's another example. 01:11:28.750 --> 01:11:33.790 align:middle line:84% Now, why do we care about all this? 01:11:33.790 --> 01:11:37.007 align:middle line:84% Why do we care about any of this stuff? 01:11:37.007 --> 01:11:38.590 align:middle line:84% Why do I care if there's moral hazard? 01:11:38.590 --> 01:11:40.632 align:middle line:84% Why do I care if the guy the guy takes a few more 01:11:40.632 --> 01:11:43.730 align:middle line:84% weeks at home getting better rather than working hard with PT 01:11:43.730 --> 01:11:44.417 align:middle line:90% to get better? 01:11:44.417 --> 01:11:46.750 align:middle line:84% Or he stays in relax a bit before he starts his new job? 01:11:46.750 --> 01:11:47.970 align:middle line:90% Why do I care? 01:11:47.970 --> 01:11:49.122 align:middle line:90% AUDIENCE: [INAUDIBLE] 01:11:49.122 --> 01:11:49.830 align:middle line:90% JON GRUBER: Yeah. 01:11:49.830 --> 01:11:51.210 align:middle line:90% I care for two reasons. 01:11:51.210 --> 01:11:55.290 align:middle line:90% First of all, why do I care? 01:11:55.290 --> 01:11:56.510 align:middle line:90% I care for two reasons. 01:11:56.510 --> 01:11:59.370 align:middle line:90% One is government cost. 01:11:59.370 --> 01:12:01.530 align:middle line:90% Cost to the government. 01:12:01.530 --> 01:12:04.330 align:middle line:84% As we'll talk about in the tax chapters, 01:12:04.330 --> 01:12:08.170 align:middle line:90% taxation causes inefficiency. 01:12:08.170 --> 01:12:10.210 align:middle line:84% So the more the government has to raise in taxes 01:12:10.210 --> 01:12:12.430 align:middle line:84% to pay its costs, the more inefficiency 01:12:12.430 --> 01:12:15.930 align:middle line:90% it induces in society. 01:12:15.930 --> 01:12:16.848 align:middle line:90% Did you do chapter 4? 01:12:16.848 --> 01:12:17.890 align:middle line:90% You did that this Friday? 01:12:17.890 --> 01:12:19.370 align:middle line:90% Valerie, was that this Friday? 01:12:19.370 --> 01:12:19.530 align:middle line:90% AUDIENCE: That was last-- 01:12:19.530 --> 01:12:20.250 align:middle line:90% JON GRUBER: Last Friday. 01:12:20.250 --> 01:12:22.542 align:middle line:84% If you did chapter 4, you talked about the consequences 01:12:22.542 --> 01:12:23.370 align:middle line:90% of deficits. 01:12:23.370 --> 01:12:24.890 align:middle line:84% We don't like to have deficits, we 01:12:24.890 --> 01:12:27.330 align:middle line:84% don't want to have excess government spending. 01:12:27.330 --> 01:12:31.410 align:middle line:84% And as a result, it's costly to have these expenditures. 01:12:31.410 --> 01:12:33.510 align:middle line:84% But that may not be the biggest reason 01:12:33.510 --> 01:12:36.310 align:middle line:90% why we don't like moral hazard. 01:12:36.310 --> 01:12:41.172 align:middle line:84% This is-- we can call this the deadweight loss of taxation. 01:12:41.172 --> 01:12:42.630 align:middle line:84% That we care about government costs 01:12:42.630 --> 01:12:44.170 align:middle line:84% rising because we've got to pay for it with taxes 01:12:44.170 --> 01:12:45.550 align:middle line:90% and that causes deadweight loss. 01:12:45.550 --> 01:12:48.670 align:middle line:84% But there's an even bigger deadweight loss problem here, 01:12:48.670 --> 01:12:49.530 align:middle line:90% which is what? 01:12:49.530 --> 01:12:53.150 align:middle line:90% 01:12:53.150 --> 01:12:55.270 align:middle line:84% What's even bigger deadweight loss problem 01:12:55.270 --> 01:12:56.650 align:middle line:90% than moral hazard causes? 01:12:56.650 --> 01:13:01.270 align:middle line:90% 01:13:01.270 --> 01:13:05.077 align:middle line:84% Why do I care if you're sitting at home not looking for a job? 01:13:05.077 --> 01:13:06.410 align:middle line:90% Why would a social planner care? 01:13:06.410 --> 01:13:07.170 align:middle line:90% Yeah? 01:13:07.170 --> 01:13:09.910 align:middle line:90% AUDIENCE: [INAUDIBLE] 01:13:09.910 --> 01:13:12.510 align:middle line:84% JON GRUBER: Well basically, more fundamentally, 01:13:12.510 --> 01:13:15.030 align:middle line:84% you're not producing your marginal product. 01:13:15.030 --> 01:13:17.790 align:middle line:84% And as a result, the social pie shrinks. 01:13:17.790 --> 01:13:21.550 align:middle line:84% Remember, labor supply is optimized 01:13:21.550 --> 01:13:24.710 align:middle line:84% at the point where your wage equals 01:13:24.710 --> 01:13:27.910 align:middle line:90% your marginal value of leisure. 01:13:27.910 --> 01:13:30.170 align:middle line:84% That is the first best for the labor market. 01:13:30.170 --> 01:13:33.235 align:middle line:84% You should work until you're indifferent 01:13:33.235 --> 01:13:35.610 align:middle line:84% between the next step work and the next hour of producing 01:13:35.610 --> 01:13:37.480 align:middle line:90% stuff. 01:13:37.480 --> 01:13:39.230 align:middle line:84% Your wage, which I've assumed, by the way, 01:13:39.230 --> 01:13:43.007 align:middle line:84% I've assumed that equals your value marginal product of labor. 01:13:43.007 --> 01:13:44.590 align:middle line:84% That's really the technical condition. 01:13:44.590 --> 01:13:46.630 align:middle line:84% The wage, I'm just assuming a perfect labor market, 01:13:46.630 --> 01:13:48.470 align:middle line:84% but this is really the technical condition. 01:13:48.470 --> 01:13:50.350 align:middle line:84% When your marginal product of labor, 01:13:50.350 --> 01:13:52.370 align:middle line:84% what you make with the next hour equals 01:13:52.370 --> 01:13:56.170 align:middle line:84% how happy you'd be not making something the next hour. 01:13:56.170 --> 01:13:59.130 align:middle line:84% Well, what happens when you get a workers' comp benefit? 01:13:59.130 --> 01:14:01.660 align:middle line:84% Well now, if you work, you get W. What 01:14:01.660 --> 01:14:02.910 align:middle line:90% do you get if you sit at home? 01:14:02.910 --> 01:14:05.330 align:middle line:84% You get the marginal value of leisure 01:14:05.330 --> 01:14:10.490 align:middle line:84% plus a benefit, which means that by definition, you 01:14:10.490 --> 01:14:12.870 align:middle line:90% will have a wage that's less. 01:14:12.870 --> 01:14:16.210 align:middle line:84% You will now choose-- you will now only work 01:14:16.210 --> 01:14:20.090 align:middle line:84% at a wage that is greater than your marginal value of leisure 01:14:20.090 --> 01:14:21.868 align:middle line:84% because you've got to make these equal. 01:14:21.868 --> 01:14:23.410 align:middle line:84% Here, you will work until wage equals 01:14:23.410 --> 01:14:24.510 align:middle line:90% marginal value of leisure. 01:14:24.510 --> 01:14:27.052 align:middle line:84% Here, you'll work until wage equals marginal value of leisure 01:14:27.052 --> 01:14:30.190 align:middle line:84% plus b, which means you'll need a higher wage. 01:14:30.190 --> 01:14:33.230 align:middle line:84% Which means that wages that you should work, 01:14:33.230 --> 01:14:35.790 align:middle line:84% that would be socially efficient to work, you won't work. 01:14:35.790 --> 01:14:39.270 align:middle line:84% So if I come in and say to you, here's $15 an hour, 01:14:39.270 --> 01:14:42.270 align:middle line:84% and $15 an hour is your marginal value of leisure, 01:14:42.270 --> 01:14:44.390 align:middle line:90% you should take that job. 01:14:44.390 --> 01:14:46.490 align:middle line:84% But if you get a $5 an hour workers' comp benefit, 01:14:46.490 --> 01:14:47.430 align:middle line:90% you won't. 01:14:47.430 --> 01:14:49.470 align:middle line:84% You'll say, forget it, I'm sitting at home. 01:14:49.470 --> 01:14:51.370 align:middle line:90% Society has lost value. 01:14:51.370 --> 01:14:54.030 align:middle line:84% The pie has shrunk because people are not 01:14:54.030 --> 01:14:55.750 align:middle line:90% working optimally. 01:14:55.750 --> 01:14:58.670 align:middle line:90% Remember, work is not a virtue. 01:14:58.670 --> 01:15:00.790 align:middle line:84% Work is something that we do to try to get 01:15:00.790 --> 01:15:02.270 align:middle line:90% the socially efficient outcome. 01:15:02.270 --> 01:15:04.190 align:middle line:84% Work should only occur to this point-- 01:15:04.190 --> 01:15:06.410 align:middle line:84% if someone has a high marginal value of leisure, 01:15:06.410 --> 01:15:07.570 align:middle line:90% they maybe shouldn't work. 01:15:07.570 --> 01:15:09.470 align:middle line:84% I'm not saying everyone should work. 01:15:09.470 --> 01:15:11.470 align:middle line:84% What I'm saying is, if you are working 01:15:11.470 --> 01:15:15.190 align:middle line:84% less than your marginal value of leisure, if the wage you require 01:15:15.190 --> 01:15:17.268 align:middle line:84% is higher than your marginal value of leisure, 01:15:17.268 --> 01:15:19.310 align:middle line:84% than you're turning down productive opportunities 01:15:19.310 --> 01:15:22.110 align:middle line:84% for society, and the pie is shrinking. 01:15:22.110 --> 01:15:25.070 align:middle line:84% And that's an even bigger reason we care. 01:15:25.070 --> 01:15:28.770 align:middle line:84% There's an inefficiency for moral hazard. 01:15:28.770 --> 01:15:30.730 align:middle line:90% Where does this leave us? 01:15:30.730 --> 01:15:32.330 align:middle line:90% Well, this leaves us-- 01:15:32.330 --> 01:15:37.970 align:middle line:84% why I love this topic is that we finally, in this class, 01:15:37.970 --> 01:15:42.170 align:middle line:84% actually get to write down a condition for optimal government 01:15:42.170 --> 01:15:42.670 align:middle line:90% policy. 01:15:42.670 --> 01:15:44.962 align:middle line:84% We don't say, well, on the one hand, on the other hand. 01:15:44.962 --> 01:15:48.050 align:middle line:84% We actually say, now we know what the government should do. 01:15:48.050 --> 01:15:50.290 align:middle line:84% And what the government should do is quite simple. 01:15:50.290 --> 01:15:53.210 align:middle line:84% It should provide you social insurance such 01:15:53.210 --> 01:15:55.750 align:middle line:84% that the benefits equal the cost. 01:15:55.750 --> 01:15:58.050 align:middle line:84% What's the cost of social insurance? 01:15:58.050 --> 01:15:59.730 align:middle line:90% The moral hazard. 01:15:59.730 --> 01:16:01.890 align:middle line:90% What's the benefit? 01:16:01.890 --> 01:16:04.290 align:middle line:90% The consumption smoothing. 01:16:04.290 --> 01:16:08.930 align:middle line:84% So the government should provide you insurance until-- 01:16:08.930 --> 01:16:11.490 align:middle line:90% where's our-- over here. 01:16:11.490 --> 01:16:13.650 align:middle line:90% Is this-- yeah. 01:16:13.650 --> 01:16:16.850 align:middle line:84% It should provide you insurance until the amount of consumption 01:16:16.850 --> 01:16:22.970 align:middle line:84% smoothing you get equals the moral hazard it's causing. 01:16:22.970 --> 01:16:25.640 align:middle line:84% So in other words, in cases where 01:16:25.640 --> 01:16:28.840 align:middle line:84% there's very little self-insurance, 01:16:28.840 --> 01:16:31.280 align:middle line:84% then social insurance is valuable. 01:16:31.280 --> 01:16:33.380 align:middle line:84% In case where there's very little moral hazard, 01:16:33.380 --> 01:16:34.780 align:middle line:90% social insurance is valuable. 01:16:34.780 --> 01:16:36.280 align:middle line:84% Social insurance should be generous. 01:16:36.280 --> 01:16:38.780 align:middle line:84% But if you have a case where there's a lot of self-insurance 01:16:38.780 --> 01:16:41.580 align:middle line:84% and a lot of moral hazard, social should not be generous. 01:16:41.580 --> 01:16:44.320 align:middle line:84% Imagine my wife's aunt, imagine that having unemployment 01:16:44.320 --> 01:16:46.412 align:middle line:84% insurance causes you to stay out of work longer 01:16:46.412 --> 01:16:48.620 align:middle line:84% and it was doing nothing to smooth their consumption, 01:16:48.620 --> 01:16:52.960 align:middle line:84% then there shouldn't be unemployment insurance for her 01:16:52.960 --> 01:16:55.400 align:middle line:84% because it was doing nothing to smooth the consumption 01:16:55.400 --> 01:16:59.380 align:middle line:84% and it was shrinking and it was distorting social efficiency. 01:16:59.380 --> 01:17:01.280 align:middle line:84% It was causing us to have less than the fully 01:17:01.280 --> 01:17:03.440 align:middle line:90% efficient equilibrium. 01:17:03.440 --> 01:17:07.120 align:middle line:84% So the optimal social insurance program will almost never 01:17:07.120 --> 01:17:10.400 align:middle line:90% be 100% full insurance. 01:17:10.400 --> 01:17:14.360 align:middle line:84% Almost never, despite where we started this lecture. 01:17:14.360 --> 01:17:14.980 align:middle line:90% Almost never. 01:17:14.980 --> 01:17:17.397 align:middle line:84% And the reason is because it will be, to some extent, also 01:17:17.397 --> 01:17:19.800 align:middle line:84% it's crowding out, and to some extent, 01:17:19.800 --> 01:17:21.360 align:middle line:90% it's causing moral hazard. 01:17:21.360 --> 01:17:23.113 align:middle line:84% So think about starting from 100. 01:17:23.113 --> 01:17:24.780 align:middle line:84% You shrink it because it's crowding out, 01:17:24.780 --> 01:17:27.155 align:middle line:84% and you shrink it more because it's causing moral hazard, 01:17:27.155 --> 01:17:29.660 align:middle line:90% you get a number well below 100. 01:17:29.660 --> 01:17:32.260 align:middle line:84% And that's how we think about optimal social insurance sides. 01:17:32.260 --> 01:17:34.180 align:middle line:84% And what we'll do in the next set of lectures 01:17:34.180 --> 01:17:36.197 align:middle line:84% is talk about those two sides of that equation. 01:17:36.197 --> 01:17:38.780 align:middle line:84% The one side of the equation is how much consumption smoothing 01:17:38.780 --> 01:17:39.752 align:middle line:90% is a program doing. 01:17:39.752 --> 01:17:41.960 align:middle line:84% The other side is how much moral hazard it's causing. 01:17:41.960 --> 01:17:43.220 align:middle line:84% And then we can put them together, 01:17:43.220 --> 01:17:45.553 align:middle line:84% they actually tell you how big social insurance programs 01:17:45.553 --> 01:17:47.540 align:middle line:90% should be. 01:17:47.540 --> 01:17:49.420 align:middle line:90% which is really cool. 01:17:49.420 --> 01:17:51.810 align:middle line:84% It'll be able to tell you what's the benefit, that's 01:17:51.810 --> 01:17:53.560 align:middle line:84% how much consumption smoothing it's doing, 01:17:53.560 --> 01:17:57.100 align:middle line:84% or alternatively, how much crowdout it's not doing. 01:17:57.100 --> 01:17:58.182 align:middle line:90% What's the cost? 01:17:58.182 --> 01:17:59.640 align:middle line:84% How much moral hazard it's causing, 01:17:59.640 --> 01:18:01.540 align:middle line:84% we'll put together to tell you what optimal social insurance 01:18:01.540 --> 01:18:02.340 align:middle line:90% looks like. 01:18:02.340 --> 01:18:05.820 align:middle line:84% And that's what we're doing for the next set of lectures. 01:18:05.820 --> 01:18:08.660 align:middle line:84% So I'll stop there with your mind freshly blown. 01:18:08.660 --> 01:18:10.480 align:middle line:84% Go away because this is hard stuff, 01:18:10.480 --> 01:18:12.740 align:middle line:84% spend some time with chapter 12 and you'll get it, 01:18:12.740 --> 01:18:17.250 align:middle line:84% and we'll come back on Wednesday and talk about Chapter 13. 01:18:17.250 --> 01:18:22.000 align:middle line:90%