WEBVTT 00:00:00.000 --> 00:00:00.500 align:middle line:90% 00:00:00.500 --> 00:00:02.444 align:middle line:90% [SQUEAKING] 00:00:02.444 --> 00:00:04.388 align:middle line:90% [RUSTLING] 00:00:04.388 --> 00:00:08.276 align:middle line:90% [CLICKING] 00:00:08.276 --> 00:00:11.795 align:middle line:90% 00:00:11.795 --> 00:00:12.420 align:middle line:90% JON GRUBER: OK. 00:00:12.420 --> 00:00:15.880 align:middle line:84% So we are talking about health care. 00:00:15.880 --> 00:00:18.833 align:middle line:84% We, just to remember the context, 00:00:18.833 --> 00:00:21.000 align:middle line:84% we were in some [INAUDIBLE] general social insurance 00:00:21.000 --> 00:00:22.680 align:middle line:84% framework, thinking about the costs 00:00:22.680 --> 00:00:25.960 align:middle line:84% and benefits of our social insurance programs. 00:00:25.960 --> 00:00:28.760 align:middle line:84% And we talked last time-- we ended last time a discussion 00:00:28.760 --> 00:00:31.882 align:middle line:84% of how can we apply the principles of chapter 12 00:00:31.882 --> 00:00:33.840 align:middle line:84% to think about what an optimal health insurance 00:00:33.840 --> 00:00:35.200 align:middle line:90% plan should look like? 00:00:35.200 --> 00:00:37.487 align:middle line:84% And we talked about, essentially, 00:00:37.487 --> 00:00:39.320 align:middle line:84% this trade-off between consumption smoothing 00:00:39.320 --> 00:00:41.760 align:middle line:84% and moral hazard, and talked about how 00:00:41.760 --> 00:00:44.280 align:middle line:84% it's hard to really justify a health insurance plan that 00:00:44.280 --> 00:00:46.680 align:middle line:84% doesn't impose some patient co-payments, 00:00:46.680 --> 00:00:49.320 align:middle line:84% because for small co-payments, there's 00:00:49.320 --> 00:00:52.400 align:middle line:84% not a lot of consumption-smoothing loss, 00:00:52.400 --> 00:00:54.320 align:middle line:90% as long as you're not too poor. 00:00:54.320 --> 00:00:58.360 align:middle line:84% And there could be a large moral hazard gain. 00:00:58.360 --> 00:01:00.010 align:middle line:90% However, there are exceptions. 00:01:00.010 --> 00:01:03.990 align:middle line:84% In particular, really if people are very resource-constrained, 00:01:03.990 --> 00:01:06.405 align:middle line:84% this could lead them to not use care that's beneficial. 00:01:06.405 --> 00:01:08.030 align:middle line:84% And even studies I reviewed in the book 00:01:08.030 --> 00:01:10.447 align:middle line:84% show that even people who aren't that resource-constrained 00:01:10.447 --> 00:01:13.570 align:middle line:84% do cut back in a lot of care that might be beneficial. 00:01:13.570 --> 00:01:16.330 align:middle line:84% And so that suggested a more subtle approach 00:01:16.330 --> 00:01:19.690 align:middle line:84% of a system where the cost sharing imposed on patients 00:01:19.690 --> 00:01:21.370 align:middle line:84% should relate both on their resources 00:01:21.370 --> 00:01:25.610 align:middle line:84% and on the value of the care, which is fine in theory. 00:01:25.610 --> 00:01:27.410 align:middle line:90% In practice, it's hard. 00:01:27.410 --> 00:01:31.530 align:middle line:84% But that's what an optimal health insurance 00:01:31.530 --> 00:01:33.490 align:middle line:90% program might look like. 00:01:33.490 --> 00:01:37.410 align:middle line:84% Now what's interesting is health insurance 00:01:37.410 --> 00:01:40.130 align:middle line:90% doesn't look like that. 00:01:40.130 --> 00:01:41.930 align:middle line:84% 15 years ago, health insurance actually 00:01:41.930 --> 00:01:44.210 align:middle line:84% was pretty much first-dollar coverage. 00:01:44.210 --> 00:01:46.770 align:middle line:84% There's been a growth in deductibles. 00:01:46.770 --> 00:01:50.410 align:middle line:84% If you look at the figure from chapter 15, 00:01:50.410 --> 00:01:53.730 align:middle line:84% you see the growth in the average deductibles 00:01:53.730 --> 00:01:56.090 align:middle line:84% for employer-sponsored insurance over time. 00:01:56.090 --> 00:01:58.780 align:middle line:84% There's been a large growth in the last 20 years 00:01:58.780 --> 00:02:00.480 align:middle line:84% in the deductibles people have to pay, 00:02:00.480 --> 00:02:03.940 align:middle line:84% which is the most common form of coinsurance. 00:02:03.940 --> 00:02:07.420 align:middle line:84% So we have moved towards more patient cost sharing. 00:02:07.420 --> 00:02:10.460 align:middle line:84% But we haven't really done so in a discriminatory way of the kind 00:02:10.460 --> 00:02:12.340 align:middle line:90% that theory might suggest. 00:02:12.340 --> 00:02:16.300 align:middle line:84% And we're still in a world where these deductibles are still 00:02:16.300 --> 00:02:19.560 align:middle line:84% relatively small for a higher-income person. 00:02:19.560 --> 00:02:23.040 align:middle line:84% Even a $2,000 deductible, which is a lot for a lot of people, 00:02:23.040 --> 00:02:27.260 align:middle line:84% is still pretty trivial for higher-income people. 00:02:27.260 --> 00:02:30.380 align:middle line:84% And the question is, why is insurance not even 00:02:30.380 --> 00:02:31.160 align:middle line:90% less generous? 00:02:31.160 --> 00:02:32.860 align:middle line:90% Why is it so-- 00:02:32.860 --> 00:02:34.883 align:middle line:84% why is it really, in some sense, a system, 00:02:34.883 --> 00:02:37.300 align:middle line:84% which is so, in many ways, overly generous to a large part 00:02:37.300 --> 00:02:39.020 align:middle line:90% of the population? 00:02:39.020 --> 00:02:42.692 align:middle line:84% And the answer traditionally has been-- one answer 00:02:42.692 --> 00:02:44.900 align:middle line:84% has been the tax subsidy to employer-sponsored health 00:02:44.900 --> 00:02:46.020 align:middle line:90% insurance. 00:02:46.020 --> 00:02:48.540 align:middle line:84% Because remember, the way that health insurance 00:02:48.540 --> 00:02:54.580 align:middle line:84% works is if you're paid in wages, you pay taxes. 00:02:54.580 --> 00:02:56.700 align:middle line:84% If you're paying health insurance, you don't. 00:02:56.700 --> 00:02:58.960 align:middle line:84% We talked last time about how that 00:02:58.960 --> 00:03:00.335 align:middle line:84% would lead people to be distorted 00:03:00.335 --> 00:03:01.918 align:middle line:84% towards getting their health insurance 00:03:01.918 --> 00:03:03.120 align:middle line:90% through their employer. 00:03:03.120 --> 00:03:05.080 align:middle line:84% But there's another implication, which 00:03:05.080 --> 00:03:07.560 align:middle line:84% is it makes health insurance relatively cheap compared 00:03:07.560 --> 00:03:09.200 align:middle line:90% to other goods. 00:03:09.200 --> 00:03:14.000 align:middle line:84% So if I want to buy apples, I have to pay taxes first, then 00:03:14.000 --> 00:03:15.280 align:middle line:90% buy apples. 00:03:15.280 --> 00:03:17.687 align:middle line:84% If I want to buy health insurance through my employer, 00:03:17.687 --> 00:03:19.020 align:middle line:90% I don't have to pay taxes first. 00:03:19.020 --> 00:03:22.160 align:middle line:84% I just buy health insurance with pre-tax dollars. 00:03:22.160 --> 00:03:25.280 align:middle line:84% So it makes health insurance relatively cheap. 00:03:25.280 --> 00:03:29.120 align:middle line:84% As a result, people overconsume health insurance. 00:03:29.120 --> 00:03:32.840 align:middle line:84% The story I always tell is I had a choice when 00:03:32.840 --> 00:03:35.760 align:middle line:84% my kids were little of buying the basic or comprehensive MIT 00:03:35.760 --> 00:03:37.560 align:middle line:90% dental plan. 00:03:37.560 --> 00:03:40.520 align:middle line:84% And basically, you could think of it as I could say, 00:03:40.520 --> 00:03:42.540 align:middle line:84% imagine MIT coming to me and say, look, 00:03:42.540 --> 00:03:45.600 align:middle line:84% would you rather earn $1,000 in salaries or $1,000 in dental 00:03:45.600 --> 00:03:46.760 align:middle line:90% benefits? 00:03:46.760 --> 00:03:51.960 align:middle line:84% Well, if the $1,000 in salary, that would basically mean, 00:03:51.960 --> 00:03:55.200 align:middle line:84% if I had another $1,000 salary, I'd take about $500 home. 00:03:55.200 --> 00:03:57.350 align:middle line:84% If I get another $1,000 in dental benefits, 00:03:57.350 --> 00:03:58.550 align:middle line:90% I get the whole $1,000. 00:03:58.550 --> 00:03:59.070 align:middle line:90% So I did it. 00:03:59.070 --> 00:04:01.790 align:middle line:84% And my daughter got these cool braces that change colors. 00:04:01.790 --> 00:04:03.110 align:middle line:90% And they were awesome. 00:04:03.110 --> 00:04:05.130 align:middle line:90% Because why the hell not? 00:04:05.130 --> 00:04:08.890 align:middle line:84% So basically, we think part of the use of the over-generosity 00:04:08.890 --> 00:04:12.130 align:middle line:84% of health care, of health insurance in America, 00:04:12.130 --> 00:04:15.464 align:middle line:90% is because of this tax subsidy. 00:04:15.464 --> 00:04:18.089 align:middle line:84% So the tax subsidy-- lets focus on the tax subsidy for a minute 00:04:18.089 --> 00:04:20.450 align:middle line:84% because it's an important policy tool. 00:04:20.450 --> 00:04:23.410 align:middle line:84% Basically, there's three fundamental things 00:04:23.410 --> 00:04:29.130 align:middle line:84% wrong with this unequal treatment of compensation 00:04:29.130 --> 00:04:31.357 align:middle line:84% in the form of wages and compensation 00:04:31.357 --> 00:04:32.690 align:middle line:90% in the form of health insurance. 00:04:32.690 --> 00:04:35.490 align:middle line:84% The first thing wrong is it costs a lot of money. 00:04:35.490 --> 00:04:38.690 align:middle line:84% So if we think that, look, either one's compensation, 00:04:38.690 --> 00:04:42.210 align:middle line:84% by treating one tax-free and taxing the other one, 00:04:42.210 --> 00:04:46.170 align:middle line:84% we lose $350 billion a year in revenues. 00:04:46.170 --> 00:04:48.950 align:middle line:84% The second is what I just described, 00:04:48.950 --> 00:04:51.290 align:middle line:84% which is it's potentially inefficient because it induces 00:04:51.290 --> 00:04:53.290 align:middle line:84% excessively generous health insurance. 00:04:53.290 --> 00:04:57.343 align:middle line:84% The third is that it's also inequitable. 00:04:57.343 --> 00:04:59.760 align:middle line:84% We haven't really talked a lot about equity in this class. 00:04:59.760 --> 00:05:01.580 align:middle line:90% We'll get there more soon. 00:05:01.580 --> 00:05:03.760 align:middle line:84% But the bottom line is, if you think about it, 00:05:03.760 --> 00:05:05.940 align:middle line:84% the way this works is you can deduct-- 00:05:05.940 --> 00:05:09.260 align:middle line:84% you don't get taxed on the cost of your health insurance. 00:05:09.260 --> 00:05:13.218 align:middle line:84% That means the higher your tax rate, the bigger break you get. 00:05:13.218 --> 00:05:15.260 align:middle line:84% Well, what causes people to have a high tax rate? 00:05:15.260 --> 00:05:16.860 align:middle line:90% Being high income. 00:05:16.860 --> 00:05:19.140 align:middle line:84% So basically, higher-income people 00:05:19.140 --> 00:05:22.980 align:middle line:84% get the larger tax break in this system. 00:05:22.980 --> 00:05:25.240 align:middle line:84% So if you think about it, what the tax subsidy is, 00:05:25.240 --> 00:05:27.700 align:middle line:84% it's basically a bribe for employers 00:05:27.700 --> 00:05:29.140 align:middle line:90% to offer health insurance. 00:05:29.140 --> 00:05:32.380 align:middle line:84% And that bribe gets larger the richer the employee is. 00:05:32.380 --> 00:05:35.500 align:middle line:84% That's not really-- doesn't sound very good as a policy 00:05:35.500 --> 00:05:37.140 align:middle line:90% design principle. 00:05:37.140 --> 00:05:39.480 align:middle line:84% So many people have said we should get rid of it. 00:05:39.480 --> 00:05:45.460 align:middle line:84% Indeed, one fun family story is I was coming back from the beach 00:05:45.460 --> 00:05:47.908 align:middle line:84% one day with my kids, and I got a call from a friend 00:05:47.908 --> 00:05:48.700 align:middle line:90% in the White House. 00:05:48.700 --> 00:05:51.340 align:middle line:90% This is in 2009 or 2008. 00:05:51.340 --> 00:05:53.428 align:middle line:84% 2009-- a friend from the White House saying, 00:05:53.428 --> 00:05:54.970 align:middle line:84% the president wants you to come down. 00:05:54.970 --> 00:05:57.137 align:middle line:84% The President wants you to come down and talk to him 00:05:57.137 --> 00:05:58.490 align:middle line:90% about the Affordable Care Act. 00:05:58.490 --> 00:05:59.250 align:middle line:90% And I was all excited. 00:05:59.250 --> 00:05:59.650 align:middle line:90% I hung up. 00:05:59.650 --> 00:06:00.730 align:middle line:84% I said to my kids, this is great. 00:06:00.730 --> 00:06:02.188 align:middle line:84% I'm going to talk to the President. 00:06:02.188 --> 00:06:05.670 align:middle line:84% And my daughter, who was, I guess, 10, said, that's so cool. 00:06:05.670 --> 00:06:07.170 align:middle line:84% You're going to meet with him alone. 00:06:07.170 --> 00:06:08.430 align:middle line:84% I said, no, I'll be in a room with a bunch 00:06:08.430 --> 00:06:09.263 align:middle line:90% of other economists. 00:06:09.263 --> 00:06:11.950 align:middle line:84% And she said, oh, that's not that impressive, then. 00:06:11.950 --> 00:06:14.872 align:middle line:90% So kids keep you humble. 00:06:14.872 --> 00:06:17.330 align:middle line:84% But I did get to meet with him with these other economists. 00:06:17.330 --> 00:06:18.830 align:middle line:84% And he said, well, how should we control health care 00:06:18.830 --> 00:06:19.890 align:middle line:90% costs in this bill? 00:06:19.890 --> 00:06:21.390 align:middle line:84% And the only thing we all agreed on for sure 00:06:21.390 --> 00:06:23.098 align:middle line:84% is get rid of the goddamn employer health 00:06:23.098 --> 00:06:24.050 align:middle line:90% insurance subsidy. 00:06:24.050 --> 00:06:25.350 align:middle line:90% That makes no sense. 00:06:25.350 --> 00:06:32.630 align:middle line:84% The problem is that-- there's a political economy problem, which 00:06:32.630 --> 00:06:35.310 align:middle line:84% is that while I can talk till I'm blue in the face about how 00:06:35.310 --> 00:06:39.570 align:middle line:84% this is removing a tax subsidy, people say it's a new tax. 00:06:39.570 --> 00:06:41.670 align:middle line:90% So that's hard to do. 00:06:41.670 --> 00:06:45.370 align:middle line:84% The other problem is that there's a political problem. 00:06:45.370 --> 00:06:47.590 align:middle line:84% There's also potentially an economic problem, which 00:06:47.590 --> 00:06:51.800 align:middle line:84% is this part of the reason we have so much health 00:06:51.800 --> 00:06:54.600 align:middle line:84% insurance through employers, remember, is this tax break. 00:06:54.600 --> 00:06:56.000 align:middle line:84% If you take it away and employers 00:06:56.000 --> 00:06:59.600 align:middle line:84% stop offering health insurance, will people become uninsured? 00:06:59.600 --> 00:07:01.367 align:middle line:90% And that's a concern as well. 00:07:01.367 --> 00:07:03.700 align:middle line:84% So that's one reason people didn't want to take it away. 00:07:03.700 --> 00:07:06.483 align:middle line:84% Now the answer to that is to have systems outside employers 00:07:06.483 --> 00:07:07.900 align:middle line:84% where we can get health insurance. 00:07:07.900 --> 00:07:09.720 align:middle line:84% And I'll talk about that next lecture with the Affordable Care 00:07:09.720 --> 00:07:10.440 align:middle line:90% Act. 00:07:10.440 --> 00:07:12.120 align:middle line:84% But nonetheless, that was a concern, 00:07:12.120 --> 00:07:13.700 align:middle line:84% that basically if we got rid of this tax break, 00:07:13.700 --> 00:07:15.908 align:middle line:84% people would lose their employer-sponsored insurance. 00:07:15.908 --> 00:07:17.040 align:middle line:90% They'd become uninsured. 00:07:17.040 --> 00:07:20.560 align:middle line:84% For that reason, a much more popular idea 00:07:20.560 --> 00:07:23.780 align:middle line:84% is the idea of capping the tax exclusion. 00:07:23.780 --> 00:07:26.760 align:middle line:84% So let's say, instead of saying you're never 00:07:26.760 --> 00:07:28.840 align:middle line:84% taxed on your health insurance, let's 00:07:28.840 --> 00:07:31.320 align:middle line:84% say you're only taxed on health insurance that 00:07:31.320 --> 00:07:35.600 align:middle line:84% costs above the average level of health insurance. 00:07:35.600 --> 00:07:38.360 align:middle line:84% So if you buy a typical plan, it's tax-free. 00:07:38.360 --> 00:07:39.860 align:middle line:84% But if you buy a more generous plan, 00:07:39.860 --> 00:07:41.800 align:middle line:90% we're going to start to tax you. 00:07:41.800 --> 00:07:46.440 align:middle line:84% The idea there is A, we raise money. 00:07:46.440 --> 00:07:50.090 align:middle line:84% B, then the tax break isn't encouraging excessively generous 00:07:50.090 --> 00:07:50.830 align:middle line:90% insurance. 00:07:50.830 --> 00:07:53.455 align:middle line:84% We're basically saying, look, we can't get rid of the tax break 00:07:53.455 --> 00:07:54.110 align:middle line:90% altogether. 00:07:54.110 --> 00:07:55.370 align:middle line:84% We want to make sure people get insurance. 00:07:55.370 --> 00:07:56.590 align:middle line:84% We're worried about them losing insurance. 00:07:56.590 --> 00:07:58.890 align:middle line:84% But we don't want to subsidize generous insurance. 00:07:58.890 --> 00:08:00.610 align:middle line:84% So let's have a tax break that's tied 00:08:00.610 --> 00:08:02.930 align:middle line:84% to a typical cost of an insurance plan. 00:08:02.930 --> 00:08:05.530 align:middle line:84% Anything above that, you lose the tax break. 00:08:05.530 --> 00:08:09.210 align:middle line:84% It's called capping the employer tax exclusion. 00:08:09.210 --> 00:08:12.468 align:middle line:84% So that became a very popular idea. 00:08:12.468 --> 00:08:14.010 align:middle line:84% And I'll come back next time and talk 00:08:14.010 --> 00:08:15.290 align:middle line:84% about what happened when we tried to do that 00:08:15.290 --> 00:08:16.547 align:middle line:90% in the Affordable Care Act. 00:08:16.547 --> 00:08:18.130 align:middle line:84% But that's one way people have thought 00:08:18.130 --> 00:08:21.290 align:middle line:84% about dealing with the economic fallout, the potential fallout 00:08:21.290 --> 00:08:22.707 align:middle line:84% of people losing health insurance. 00:08:22.707 --> 00:08:24.540 align:middle line:84% We'd say, we'll let you keep this tax break, 00:08:24.540 --> 00:08:26.010 align:middle line:90% but only up to a certain level. 00:08:26.010 --> 00:08:30.010 align:middle line:90% And that's one idea people had. 00:08:30.010 --> 00:08:33.010 align:middle line:90% Questions about that? 00:08:33.010 --> 00:08:35.530 align:middle line:84% So that's all I want-- yeah, Enoch? 00:08:35.530 --> 00:08:37.530 align:middle line:84% AUDIENCE: Higher income people will basically 00:08:37.530 --> 00:08:38.470 align:middle line:90% get more out of it? 00:08:38.470 --> 00:08:39.872 align:middle line:90% JON GRUBER: Yeah. 00:08:39.872 --> 00:08:42.330 align:middle line:84% AUDIENCE: They get more money back because otherwise they'd 00:08:42.330 --> 00:08:42.913 align:middle line:90% be taxed more. 00:08:42.913 --> 00:08:43.622 align:middle line:90% JON GRUBER: Yeah. 00:08:43.622 --> 00:08:45.850 align:middle line:84% AUDIENCE: More money means less to them than like-- 00:08:45.850 --> 00:08:46.930 align:middle line:90% JON GRUBER: Less what? 00:08:46.930 --> 00:08:48.638 align:middle line:84% AUDIENCE: More money doesn't mean as much 00:08:48.638 --> 00:08:49.990 align:middle line:90% to high-income people. 00:08:49.990 --> 00:08:52.316 align:middle line:90% JON GRUBER: Oh, well, no. 00:08:52.316 --> 00:08:53.810 align:middle line:90% But that's a great point. 00:08:53.810 --> 00:08:55.750 align:middle line:84% So you're saying, as a percent of income, 00:08:55.750 --> 00:08:57.790 align:middle line:90% are they getting more back? 00:08:57.790 --> 00:09:00.710 align:middle line:84% The answer as a percent of income, they-- 00:09:00.710 --> 00:09:01.910 align:middle line:90% so the very richest aren't. 00:09:01.910 --> 00:09:04.065 align:middle line:90% The near-rich are getting more. 00:09:04.065 --> 00:09:06.190 align:middle line:84% They're also more likely to have insurance coverage 00:09:06.190 --> 00:09:07.350 align:middle line:90% in the first place. 00:09:07.350 --> 00:09:08.935 align:middle line:90% But you're right. 00:09:08.935 --> 00:09:10.310 align:middle line:84% This becomes something we discuss 00:09:10.310 --> 00:09:13.390 align:middle line:84% a lot in chapter 18, which is how do you measure fairness? 00:09:13.390 --> 00:09:16.230 align:middle line:90% It's very tricky. 00:09:16.230 --> 00:09:17.110 align:middle line:90% Yeah, Evan. 00:09:17.110 --> 00:09:20.550 align:middle line:84% AUDIENCE: We'll still capturing information about some people 00:09:20.550 --> 00:09:23.557 align:middle line:84% might just prefer to have more generous insurance, 00:09:23.557 --> 00:09:25.390 align:middle line:84% but that's fine because they can pay for it. 00:09:25.390 --> 00:09:26.130 align:middle line:90% JON GRUBER: Yeah, exactly. 00:09:26.130 --> 00:09:28.380 align:middle line:84% The question is, why is the government subsidizing it? 00:09:28.380 --> 00:09:29.890 align:middle line:84% We're not saying you can't have it. 00:09:29.890 --> 00:09:31.270 align:middle line:84% We're just saying, why is the government paying for you 00:09:31.270 --> 00:09:32.370 align:middle line:90% to have it on the margin? 00:09:32.370 --> 00:09:35.670 align:middle line:84% The idea would be, if we go back to our flat of the curve, 00:09:35.670 --> 00:09:37.790 align:middle line:84% that we think giving people some insurance might 00:09:37.790 --> 00:09:39.110 align:middle line:90% be good for their health. 00:09:39.110 --> 00:09:41.193 align:middle line:84% But there's no reason why we should be subsidizing 00:09:41.193 --> 00:09:44.350 align:middle line:90% very generous insurance. 00:09:44.350 --> 00:09:45.650 align:middle line:90% Yeah? 00:09:45.650 --> 00:09:47.590 align:middle line:84% AUDIENCE: Last class, we talked about some 00:09:47.590 --> 00:09:50.400 align:middle line:84% of the benefits of the employer-based insurance, which 00:09:50.400 --> 00:09:53.520 align:middle line:84% was that it mitigates the adverse selection of fixed 00:09:53.520 --> 00:09:54.480 align:middle line:90% costs. 00:09:54.480 --> 00:09:56.760 align:middle line:84% Is there a reason why it might be good to incentivize 00:09:56.760 --> 00:09:57.940 align:middle line:90% employer-based insurance? 00:09:57.940 --> 00:09:59.500 align:middle line:84% JON GRUBER: That's one reason why. 00:09:59.500 --> 00:10:04.013 align:middle line:84% So in some sense, you could think of this as-- 00:10:04.013 --> 00:10:05.680 align:middle line:84% in economics, we talk a lot about what's 00:10:05.680 --> 00:10:07.600 align:middle line:90% called second-best problems. 00:10:07.600 --> 00:10:10.720 align:middle line:84% The first best is 14.01, like everything works. 00:10:10.720 --> 00:10:13.080 align:middle line:84% The second best is, well, if something goes wrong, 00:10:13.080 --> 00:10:15.840 align:middle line:84% what's the thing you can do under those conditions, 00:10:15.840 --> 00:10:18.240 align:middle line:90% the constrained best optimum? 00:10:18.240 --> 00:10:20.600 align:middle line:84% You might make what's called a second-best argument 00:10:20.600 --> 00:10:22.953 align:middle line:84% for the employer tax subsidy saying, well, look. 00:10:22.953 --> 00:10:25.120 align:middle line:84% There's reasons why we want people to have insurance 00:10:25.120 --> 00:10:26.320 align:middle line:90% through their employer. 00:10:26.320 --> 00:10:27.820 align:middle line:84% The only way we can make that happen 00:10:27.820 --> 00:10:29.660 align:middle line:84% is through this bribe to employers. 00:10:29.660 --> 00:10:30.740 align:middle line:90% So let's do it. 00:10:30.740 --> 00:10:33.080 align:middle line:84% It's like a subsidy to insurance. 00:10:33.080 --> 00:10:34.538 align:middle line:84% The problem-- and you could say-- 00:10:34.538 --> 00:10:36.080 align:middle line:84% the counter would be, well, if you're 00:10:36.080 --> 00:10:38.180 align:middle line:84% going to subsidize insurance, why do it through employers? 00:10:38.180 --> 00:10:39.847 align:middle line:84% And the argument would be, well, they're 00:10:39.847 --> 00:10:41.320 align:middle line:90% a natural pooling mechanism. 00:10:41.320 --> 00:10:43.320 align:middle line:84% And we'll come back next time to talk 00:10:43.320 --> 00:10:45.120 align:middle line:84% about why it might make more sense to use 00:10:45.120 --> 00:10:46.340 align:middle line:90% other pooling mechanisms. 00:10:46.340 --> 00:10:48.970 align:middle line:84% But that would be the kind of-- this important kind of argument 00:10:48.970 --> 00:10:51.650 align:middle line:84% we need to get good at developing, which is recognizing 00:10:51.650 --> 00:10:53.270 align:middle line:90% that we're not in 14.01. 00:10:53.270 --> 00:10:55.350 align:middle line:84% Given failures, what's the best thing to do 00:10:55.350 --> 00:10:57.050 align:middle line:90% and the trade-offs in those. 00:10:57.050 --> 00:10:57.670 align:middle line:90% Yeah? 00:10:57.670 --> 00:10:59.090 align:middle line:90% AUDIENCE: The curve [INAUDIBLE] 00:10:59.090 --> 00:11:01.430 align:middle line:84% JON GRUBER: The medical effectiveness curve. 00:11:01.430 --> 00:11:04.330 align:middle line:90% 00:11:04.330 --> 00:11:06.930 align:middle line:84% So that's what I say about the first-- 00:11:06.930 --> 00:11:09.350 align:middle line:84% like I said, we talked about the benefits of insurance, 00:11:09.350 --> 00:11:10.425 align:middle line:90% the cost of insurance. 00:11:10.425 --> 00:11:12.050 align:middle line:84% But there's a whole other dimension now 00:11:12.050 --> 00:11:15.330 align:middle line:84% that we skipped in chapter 12 and 13 and 14, which 00:11:15.330 --> 00:11:19.597 align:middle line:84% is the role of the medical provider. 00:11:19.597 --> 00:11:21.430 align:middle line:84% When we talked about unemployment insurance, 00:11:21.430 --> 00:11:24.730 align:middle line:84% for example, the key thing that determined 00:11:24.730 --> 00:11:28.130 align:middle line:84% how long you stayed unemployed was your trade-off 00:11:28.130 --> 00:11:30.730 align:middle line:90% between leisure and UI benefits. 00:11:30.730 --> 00:11:33.450 align:middle line:84% But what determines how much medical care you use 00:11:33.450 --> 00:11:34.600 align:middle line:90% is mostly your doctor. 00:11:34.600 --> 00:11:36.850 align:middle line:84% I mean, you might decide whether you go to the doctor. 00:11:36.850 --> 00:11:39.490 align:middle line:84% When you get there, you're largely not going 00:11:39.490 --> 00:11:41.410 align:middle line:90% to not do what the doctor says. 00:11:41.410 --> 00:11:43.150 align:middle line:84% The doctor largely determines it. 00:11:43.150 --> 00:11:44.670 align:middle line:84% So now we have to ask ourselves-- 00:11:44.670 --> 00:11:46.980 align:middle line:84% there's another component of moral hazard we have 00:11:46.980 --> 00:11:49.380 align:middle line:84% to worry about, which is moral hazard on the provider 00:11:49.380 --> 00:11:54.337 align:middle line:84% side, which is, how do we think about not just what patients 00:11:54.337 --> 00:11:55.920 align:middle line:84% have to pay when they get health care, 00:11:55.920 --> 00:11:58.340 align:middle line:84% but how we should pay the providers to deliver that health 00:11:58.340 --> 00:11:59.780 align:middle line:90% care. 00:11:59.780 --> 00:12:03.620 align:middle line:84% Now traditionally in America, and the whole world, 00:12:03.620 --> 00:12:09.700 align:middle line:84% we had what was called a fee-for-service, fee-for-service 00:12:09.700 --> 00:12:10.732 align:middle line:90% medical system. 00:12:10.732 --> 00:12:12.440 align:middle line:84% What is a fee-for-service medical system? 00:12:12.440 --> 00:12:14.940 align:middle line:84% What it sounds like-- the doctor does a service and they get 00:12:14.940 --> 00:12:16.900 align:middle line:90% paid. 00:12:16.900 --> 00:12:19.040 align:middle line:90% Dr. Ahmet fixes my knee. 00:12:19.040 --> 00:12:20.940 align:middle line:90% Dr. Ahmet bills me for a knee. 00:12:20.940 --> 00:12:22.600 align:middle line:84% As long as the bill isn't insane, 00:12:22.600 --> 00:12:24.600 align:middle line:90% I pay it, as the insurer. 00:12:24.600 --> 00:12:27.120 align:middle line:84% If Dr. Ahmet bills me $1 trillion fixing my knee, 00:12:27.120 --> 00:12:29.320 align:middle line:84% I know Dr. Ahmet's not on the level. 00:12:29.320 --> 00:12:32.860 align:middle line:84% But basically, they had a thing they called usual and customary 00:12:32.860 --> 00:12:33.620 align:middle line:90% reimbursement. 00:12:33.620 --> 00:12:35.780 align:middle line:90% Sort of sounds sensible. 00:12:35.780 --> 00:12:42.620 align:middle line:84% The problem with this is that this led to a concern of doctors 00:12:42.620 --> 00:12:45.550 align:middle line:90% over overusing health care. 00:12:45.550 --> 00:12:49.870 align:middle line:84% So think about a doctor as having a utility function. 00:12:49.870 --> 00:12:52.290 align:middle line:84% I'm not about malicious doctors, scumbag doctors. 00:12:52.290 --> 00:12:54.870 align:middle line:84% No, a regular Joe doctor, good guy. 00:12:54.870 --> 00:12:58.150 align:middle line:84% Regular Joe doctor's utility function includes two things-- 00:12:58.150 --> 00:13:00.610 align:middle line:84% your health as a patient and their income. 00:13:00.610 --> 00:13:01.850 align:middle line:90% They care about their income. 00:13:01.850 --> 00:13:02.770 align:middle line:90% They're not selfless. 00:13:02.770 --> 00:13:03.778 align:middle line:90% I said they're regular. 00:13:03.778 --> 00:13:06.070 align:middle line:84% They're not criminal, but they're not selfless, either. 00:13:06.070 --> 00:13:10.470 align:middle line:84% As a result, think about the incentives facing that doctor 00:13:10.470 --> 00:13:13.910 align:middle line:84% to do any service which has an even an infinitesimal 00:13:13.910 --> 00:13:16.990 align:middle line:84% chance of benefiting you, but might make them money. 00:13:16.990 --> 00:13:19.310 align:middle line:90% They'll do it. 00:13:19.310 --> 00:13:20.990 align:middle line:84% Indeed, you could go further and say 00:13:20.990 --> 00:13:23.270 align:middle line:84% if it doesn't hurt you too much, they might still do. 00:13:23.270 --> 00:13:26.890 align:middle line:84% Measures which infinitesimally hurts you but makes them money, 00:13:26.890 --> 00:13:29.790 align:middle line:84% they might still do it because they care about both things. 00:13:29.790 --> 00:13:32.630 align:middle line:84% As a result, as you think about that flat of the curve, 00:13:32.630 --> 00:13:34.270 align:middle line:84% doctor incentives are pushing way out 00:13:34.270 --> 00:13:36.505 align:middle line:90% towards the end of the curve. 00:13:36.505 --> 00:13:38.130 align:middle line:84% Because as long as it doesn't hurt you, 00:13:38.130 --> 00:13:39.510 align:middle line:84% and as long as there's even a chance 00:13:39.510 --> 00:13:41.870 align:middle line:84% it might benefit you-- so let's say the doctor sees you 00:13:41.870 --> 00:13:44.660 align:middle line:84% and you have something which, once in the last 50 years, 00:13:44.660 --> 00:13:48.820 align:middle line:84% someone found someone with those conditions had brain cancer. 00:13:48.820 --> 00:13:52.080 align:middle line:84% So they ordered a brain scan for you and it finds nothing 00:13:52.080 --> 00:13:54.225 align:middle line:84% because it's one-out-of-a-gazillion cases 00:13:54.225 --> 00:13:55.100 align:middle line:90% that found something. 00:13:55.100 --> 00:13:56.040 align:middle line:90% But why not? 00:13:56.040 --> 00:13:57.340 align:middle line:90% It doesn't hurt you that much. 00:13:57.340 --> 00:13:58.020 align:middle line:90% Doesn't hurt you at all. 00:13:58.020 --> 00:13:59.040 align:middle line:90% It's just a brain scan. 00:13:59.040 --> 00:14:00.363 align:middle line:90% Makes them some money. 00:14:00.363 --> 00:14:02.280 align:middle line:84% So the concern with the fee-for-service system 00:14:02.280 --> 00:14:05.100 align:middle line:84% was it would lead to overuse of health care, 00:14:05.100 --> 00:14:07.520 align:middle line:84% overprescribing health by doctors. 00:14:07.520 --> 00:14:12.440 align:middle line:84% And that's why, in the 1980s, we started moving away from 00:14:12.440 --> 00:14:20.800 align:middle line:84% the fee-for-service system towards managed care, 00:14:20.800 --> 00:14:23.840 align:middle line:84% a term you're all familiar with by now. 00:14:23.840 --> 00:14:26.300 align:middle line:84% Managed care is the notion of not just a purely 00:14:26.300 --> 00:14:28.800 align:middle line:84% system where doctors do what the hell they want and bill you 00:14:28.800 --> 00:14:30.560 align:middle line:90% for it. 00:14:30.560 --> 00:14:35.240 align:middle line:84% Managed care began in a very simple way 00:14:35.240 --> 00:14:37.680 align:middle line:84% with what was called preferred provider 00:14:37.680 --> 00:14:43.180 align:middle line:84% organizations, preferred provider organizations, or PPOs. 00:14:43.180 --> 00:14:44.880 align:middle line:90% This is what started early on. 00:14:44.880 --> 00:14:47.742 align:middle line:84% The way these worked is they simply said, look, 00:14:47.742 --> 00:14:49.700 align:middle line:84% this wasn't really dealing with a moral hazard. 00:14:49.700 --> 00:14:52.700 align:middle line:84% This was saying, look, there's all these extra rents 00:14:52.700 --> 00:14:54.230 align:middle line:90% being made by providers. 00:14:54.230 --> 00:14:55.980 align:middle line:84% And could we redistribute those rents back 00:14:55.980 --> 00:14:56.938 align:middle line:90% to insurers and people? 00:14:56.938 --> 00:14:58.605 align:middle line:84% And the way it worked was the following. 00:14:58.605 --> 00:15:00.740 align:middle line:84% It started out in California, Northern California. 00:15:00.740 --> 00:15:04.140 align:middle line:84% And an insurer went to Hewlett-Packard, a or a company 00:15:04.140 --> 00:15:08.100 align:middle line:84% like Hewlett-Packard, and said, if I can save you 00:15:08.100 --> 00:15:10.023 align:middle line:84% 25% on your health care bill, would you 00:15:10.023 --> 00:15:11.940 align:middle line:84% be willing to tell your patients they can only 00:15:11.940 --> 00:15:13.700 align:middle line:90% use hospital x, y, or z? 00:15:13.700 --> 00:15:15.198 align:middle line:90% And Hewlett-Packard said, sure. 00:15:15.198 --> 00:15:17.240 align:middle line:84% Then they went to hospitals in the area and said, 00:15:17.240 --> 00:15:20.218 align:middle line:84% whoever will give us 40% off, you 00:15:20.218 --> 00:15:21.760 align:middle line:84% can get the Hewlett-Packard patients. 00:15:21.760 --> 00:15:23.100 align:middle line:84% But if you don't, you won't get them. 00:15:23.100 --> 00:15:24.280 align:middle line:84% The hospitals are like, well, geez, 00:15:24.280 --> 00:15:25.480 align:middle line:84% we're making these zillions of profits. 00:15:25.480 --> 00:15:27.820 align:middle line:84% We'll happily give you 40% back to get the patients. 00:15:27.820 --> 00:15:29.717 align:middle line:84% The PPO pocketed the 15% difference, 00:15:29.717 --> 00:15:31.800 align:middle line:84% went back to Hewlett-Packard, and said, hey, look. 00:15:31.800 --> 00:15:32.925 align:middle line:90% Here's a list of hospitals. 00:15:32.925 --> 00:15:35.160 align:middle line:84% Restrict your employees to go into these hospitals, 00:15:35.160 --> 00:15:36.820 align:middle line:90% and you can save 25%. 00:15:36.820 --> 00:15:40.190 align:middle line:84% Essentially, they became a middleman scooping up 00:15:40.190 --> 00:15:43.230 align:middle line:84% the rents that were being excessively earned 00:15:43.230 --> 00:15:44.870 align:middle line:90% by these health care providers. 00:15:44.870 --> 00:15:46.905 align:middle line:84% And PPOs lowered health care costs a lot, 00:15:46.905 --> 00:15:49.030 align:middle line:84% but not necessarily through changing the efficiency 00:15:49.030 --> 00:15:49.630 align:middle line:90% with health care. 00:15:49.630 --> 00:15:51.570 align:middle line:84% They weren't really addressing the moral hazard problem. 00:15:51.570 --> 00:15:54.030 align:middle line:84% They were just scooping up the rents that providers at that 00:15:54.030 --> 00:15:57.830 align:middle line:84% point-- essentially, imperfect competition created rents. 00:15:57.830 --> 00:16:00.190 align:middle line:84% PPOs became a competitive mechanism 00:16:00.190 --> 00:16:03.870 align:middle line:90% to recapture those rents. 00:16:03.870 --> 00:16:05.330 align:middle line:90% So that was the first wave. 00:16:05.330 --> 00:16:07.870 align:middle line:84% But the really exciting wave was the wave 00:16:07.870 --> 00:16:10.910 align:middle line:84% of the HMO wave, health maintenance organizations, 00:16:10.910 --> 00:16:14.230 align:middle line:84% because what HMOs were is they were essentially 00:16:14.230 --> 00:16:20.150 align:middle line:84% the integration of the delivery and the insurance of health. 00:16:20.150 --> 00:16:23.110 align:middle line:84% HMOs said, let's get rid of the separation 00:16:23.110 --> 00:16:25.510 align:middle line:84% of the company that insures you and the company 00:16:25.510 --> 00:16:26.290 align:middle line:90% delivers the care. 00:16:26.290 --> 00:16:29.950 align:middle line:84% Let's integrate them, the most famous of which, 00:16:29.950 --> 00:16:35.470 align:middle line:84% initially, was Kaiser Permanente in California. 00:16:35.470 --> 00:16:37.310 align:middle line:84% What they did is they said, we are not 00:16:37.310 --> 00:16:39.440 align:middle line:84% going to have a system where your insurer-- instead 00:16:39.440 --> 00:16:41.273 align:middle line:84% of signing up for Blue Cross and Blue Cross, 00:16:41.273 --> 00:16:44.760 align:middle line:84% pays whatever doctors want, you sign up for Kaiser. 00:16:44.760 --> 00:16:46.780 align:middle line:84% You have to go to a Kaiser Hospital or doctor, 00:16:46.780 --> 00:16:48.920 align:middle line:90% so you get the PPO effect. 00:16:48.920 --> 00:16:52.200 align:middle line:84% But moreover, we're going to change the way our providers are 00:16:52.200 --> 00:16:53.600 align:middle line:90% reimbursed. 00:16:53.600 --> 00:16:56.120 align:middle line:84% We're going to move from a fee-for-service system 00:16:56.120 --> 00:17:01.868 align:middle line:84% to a prospective reimbursement system. 00:17:01.868 --> 00:17:03.660 align:middle line:84% What is a prospective reimbursement system? 00:17:03.660 --> 00:17:05.077 align:middle line:84% A prospective reimbursement system 00:17:05.077 --> 00:17:09.760 align:middle line:84% is where doctors are paid not based on what they do to you, 00:17:09.760 --> 00:17:12.319 align:middle line:90% but based on something else. 00:17:12.319 --> 00:17:14.440 align:middle line:84% So for example, one type of prospective system 00:17:14.440 --> 00:17:17.119 align:middle line:90% is doctors are paid a salary. 00:17:17.119 --> 00:17:19.140 align:middle line:84% Kaiser just paid their doctors a salary. 00:17:19.140 --> 00:17:21.692 align:middle line:84% It didn't matter how many people you saw or what you did. 00:17:21.692 --> 00:17:23.859 align:middle line:84% So there's no more incentive to give the extra test, 00:17:23.859 --> 00:17:25.192 align:middle line:90% didn't make you much more money. 00:17:25.192 --> 00:17:27.839 align:middle line:84% You made your x $100,000 a year, regardless of what you did 00:17:27.839 --> 00:17:29.360 align:middle line:90% to people. 00:17:29.360 --> 00:17:34.520 align:middle line:84% So we got rid of the incentive to overtreat people 00:17:34.520 --> 00:17:37.040 align:middle line:84% because your salary didn't depend on it. 00:17:37.040 --> 00:17:38.470 align:middle line:90% You could go further than that. 00:17:38.470 --> 00:17:39.730 align:middle line:90% So that's one approach. 00:17:39.730 --> 00:17:44.530 align:middle line:84% Further approach is the evil-sounding capitation, 00:17:44.530 --> 00:17:48.770 align:middle line:84% which is not decapitation, but capitation approach, 00:17:48.770 --> 00:17:52.330 align:middle line:84% which is to say to doctors, look, 00:17:52.330 --> 00:17:54.458 align:middle line:84% we're going to pay you per-- you're 00:17:54.458 --> 00:17:56.250 align:middle line:84% going to have a panel of-- it's the easiest 00:17:56.250 --> 00:17:59.083 align:middle line:84% thing with a primary care doc, a pediatrician. 00:17:59.083 --> 00:18:00.750 align:middle line:84% We're going to say to the pediatricians, 00:18:00.750 --> 00:18:02.410 align:middle line:84% you're going to have a panel of kids. 00:18:02.410 --> 00:18:04.630 align:middle line:84% For every kid who's enrolled with you as their doc, 00:18:04.630 --> 00:18:08.070 align:middle line:84% you get x dollars, stop, full stop. 00:18:08.070 --> 00:18:09.850 align:middle line:84% It doesn't matter what you do to them. 00:18:09.850 --> 00:18:11.250 align:middle line:90% You get that. 00:18:11.250 --> 00:18:14.210 align:middle line:84% The thing about this system-- it flips all the incentives 00:18:14.210 --> 00:18:15.450 align:middle line:90% on their head. 00:18:15.450 --> 00:18:18.090 align:middle line:84% It used to be when that kid came in to my system, every test 00:18:18.090 --> 00:18:19.570 align:middle line:90% I did, it made more money. 00:18:19.570 --> 00:18:22.550 align:middle line:84% Now, every test I do, I lose money. 00:18:22.550 --> 00:18:23.050 align:middle line:90% Why? 00:18:23.050 --> 00:18:24.967 align:middle line:84% Because I get paid the same amount for the kid 00:18:24.967 --> 00:18:27.050 align:middle line:90% if I ignore him or treat him. 00:18:27.050 --> 00:18:28.750 align:middle line:84% So it flips the incentives on the head. 00:18:28.750 --> 00:18:34.410 align:middle line:84% And now, instead [INAUDIBLE] to undertreat because, basically-- 00:18:34.410 --> 00:18:38.580 align:middle line:84% what it does is it changes the risk, who bears the risk. 00:18:38.580 --> 00:18:42.940 align:middle line:84% Under a fee-for-service system, the payer bears the risk. 00:18:42.940 --> 00:18:45.660 align:middle line:84% Under capitation system, the provider 00:18:45.660 --> 00:18:48.155 align:middle line:84% bears the risk because they get a fixed payment 00:18:48.155 --> 00:18:49.280 align:middle line:90% regardless of what they do. 00:18:49.280 --> 00:18:51.080 align:middle line:84% So they do more-- they're going to lose money. 00:18:51.080 --> 00:18:52.320 align:middle line:84% If they do less, they're make money. 00:18:52.320 --> 00:18:52.820 align:middle line:90% Yeah. 00:18:52.820 --> 00:18:55.020 align:middle line:84% AUDIENCE: And that money loss is for the doctor 00:18:55.020 --> 00:18:57.260 align:middle line:84% because it's them spending their time doing that-- 00:18:57.260 --> 00:18:59.020 align:middle line:84% JON GRUBER: No, because they literally-- 00:18:59.020 --> 00:19:00.620 align:middle line:90% AUDIENCE: [INAUDIBLE] 00:19:00.620 --> 00:19:02.120 align:middle line:84% JON GRUBER: But their time is money. 00:19:02.120 --> 00:19:03.520 align:middle line:90% That's why I use a pediatrician. 00:19:03.520 --> 00:19:04.760 align:middle line:84% I mean, I'm not talking about surgery. 00:19:04.760 --> 00:19:06.060 align:middle line:90% I'm talking about pediatrician. 00:19:06.060 --> 00:19:07.640 align:middle line:90% They get paid for their time. 00:19:07.640 --> 00:19:09.180 align:middle line:84% So now they're spending their time. 00:19:09.180 --> 00:19:14.800 align:middle line:84% So the point is if Ahmet's a pediatrician 00:19:14.800 --> 00:19:16.860 align:middle line:84% and Val's a pediatrician, and Ahmet's 00:19:16.860 --> 00:19:19.333 align:middle line:84% going to spend three minutes-- like my workers' comp doc, 00:19:19.333 --> 00:19:21.000 align:middle line:84% spend three minutes seeing kids all day, 00:19:21.000 --> 00:19:23.000 align:middle line:84% he can see a ton more kids than Valerie can, 00:19:23.000 --> 00:19:25.460 align:middle line:90% so he'll make more money. 00:19:25.460 --> 00:19:28.220 align:middle line:84% So basically, that was the switch, 00:19:28.220 --> 00:19:32.620 align:middle line:84% was basically the idea of moving towards a prospective 00:19:32.620 --> 00:19:34.180 align:middle line:90% reimbursement system. 00:19:34.180 --> 00:19:36.310 align:middle line:90% This was a very radical change. 00:19:36.310 --> 00:19:39.710 align:middle line:84% Now, HMOs eventually evolved into two types. 00:19:39.710 --> 00:19:42.030 align:middle line:84% The original Kaiser type was what 00:19:42.030 --> 00:19:45.710 align:middle line:84% was called IPAs, independent practice organizations, 00:19:45.710 --> 00:19:50.510 align:middle line:84% literally, the full integration of an insurer and a doctor. 00:19:50.510 --> 00:19:52.990 align:middle line:84% But many people found that too restrictive. 00:19:52.990 --> 00:19:58.630 align:middle line:84% So we actually went to panel models of HMOs. 00:19:58.630 --> 00:20:01.430 align:middle line:84% The way the panel model worked is the HMO 00:20:01.430 --> 00:20:02.540 align:middle line:90% set up a panel of doctors. 00:20:02.540 --> 00:20:04.290 align:middle line:84% They didn't have to work for that company. 00:20:04.290 --> 00:20:05.190 align:middle line:90% So there's now a-- 00:20:05.190 --> 00:20:08.110 align:middle line:90% I belong to the Blue Cross HMO. 00:20:08.110 --> 00:20:10.010 align:middle line:84% The doctors don't work for Blue Cross. 00:20:10.010 --> 00:20:13.630 align:middle line:84% But if you want a Blue Cross HMO, it's like a PPO. 00:20:13.630 --> 00:20:15.550 align:middle line:90% You have to sign up for it. 00:20:15.550 --> 00:20:18.410 align:middle line:84% And you agree, but it's a PPO with prospective system, 00:20:18.410 --> 00:20:20.030 align:middle line:90% with prospective reimbursement. 00:20:20.030 --> 00:20:22.030 align:middle line:84% So my doctor gets paid a fixed amount 00:20:22.030 --> 00:20:25.130 align:middle line:84% for having enrolled me by Blue Cross Blue Shield, 00:20:25.130 --> 00:20:28.110 align:middle line:84% but doesn't get paid any more based on what he does to me. 00:20:28.110 --> 00:20:31.638 align:middle line:84% So IPAs are the most extreme form, 00:20:31.638 --> 00:20:33.930 align:middle line:84% where, basically, literally, the insurer and the doctor 00:20:33.930 --> 00:20:35.570 align:middle line:90% are one and the same. 00:20:35.570 --> 00:20:37.090 align:middle line:90% The typical HMO is not that. 00:20:37.090 --> 00:20:40.990 align:middle line:84% Typical HMO is like Blue Cross, where the doctor's independent, 00:20:40.990 --> 00:20:43.990 align:middle line:84% but the doctor, to be signed up by Blue Cross, has to accept, 00:20:43.990 --> 00:20:46.830 align:middle line:84% I'll take this payment per patient, or Blue Cross says, 00:20:46.830 --> 00:20:47.790 align:middle line:90% I won't sign you up. 00:20:47.790 --> 00:20:50.730 align:middle line:90% 00:20:50.730 --> 00:20:53.610 align:middle line:84% Because of this-- and we'll come back to this next time. 00:20:53.610 --> 00:20:56.050 align:middle line:84% This is why the key differentiator 00:20:56.050 --> 00:20:58.760 align:middle line:84% across insurance companies is becoming their panel of doctors. 00:20:58.760 --> 00:21:01.010 align:middle line:84% Insurance companies don't differ that much, especially 00:21:01.010 --> 00:21:03.570 align:middle line:84% since the ACA, in what they cover. 00:21:03.570 --> 00:21:06.850 align:middle line:84% What they differ is in which doctors they've signed up. 00:21:06.850 --> 00:21:09.030 align:middle line:84% And obviously, the more money you pay doctors, 00:21:09.030 --> 00:21:11.370 align:middle line:90% the more you can sign up. 00:21:11.370 --> 00:21:15.030 align:middle line:84% So we've moved towards this prospective system. 00:21:15.030 --> 00:21:20.290 align:middle line:84% Now, what effect has this had, of moving to this? 00:21:20.290 --> 00:21:24.250 align:middle line:84% Well, on the one hand, the strong prediction 00:21:24.250 --> 00:21:28.110 align:middle line:84% would be it would save money because you move from one where 00:21:28.110 --> 00:21:29.610 align:middle line:84% the doctors make more money by doing 00:21:29.610 --> 00:21:32.050 align:middle line:84% more on the margin to the doctors 00:21:32.050 --> 00:21:35.100 align:middle line:84% make more money by doing less on the margin. 00:21:35.100 --> 00:21:38.340 align:middle line:84% The problem is you can't just compare 00:21:38.340 --> 00:21:39.780 align:middle line:84% those who are in HMOs to those who 00:21:39.780 --> 00:21:42.020 align:middle line:90% aren't because it's endogenous. 00:21:42.020 --> 00:21:44.260 align:middle line:90% People get to choose. 00:21:44.260 --> 00:21:46.480 align:middle line:84% A sick person is not going to want to be in an HMO. 00:21:46.480 --> 00:21:49.500 align:middle line:84% They're going to want be able to go to whatever doctor they want. 00:21:49.500 --> 00:21:51.142 align:middle line:84% HMOs limit your choice of doctor. 00:21:51.142 --> 00:21:53.100 align:middle line:84% It's only the doctor who's signed up with them. 00:21:53.100 --> 00:21:55.540 align:middle line:84% So the problem is you have a natural bias 00:21:55.540 --> 00:21:57.663 align:middle line:84% problem in comparing those HMOs and outside. 00:21:57.663 --> 00:21:59.580 align:middle line:84% There's been a bunch of papers that have dealt 00:21:59.580 --> 00:22:01.940 align:middle line:90% with this in various ways. 00:22:01.940 --> 00:22:05.760 align:middle line:84% They all come to a very consistent conclusion, which is, 00:22:05.760 --> 00:22:08.180 align:middle line:90% A, there is enormous selection. 00:22:08.180 --> 00:22:10.700 align:middle line:84% Sicker people are much less likely to sign up for HMOs 00:22:10.700 --> 00:22:12.940 align:middle line:90% than healthier people are. 00:22:12.940 --> 00:22:15.580 align:middle line:84% B, when you control for that selection, 00:22:15.580 --> 00:22:22.740 align:middle line:84% HMOs save a lot of money, like, maybe 20%, 25% A lot of money, 00:22:22.740 --> 00:22:25.260 align:middle line:90% through this alternative model. 00:22:25.260 --> 00:22:28.080 align:middle line:84% So the first factor's the selection. 00:22:28.080 --> 00:22:30.020 align:middle line:84% The second factor's they're lowering spending. 00:22:30.020 --> 00:22:31.070 align:middle line:90% Yeah. 00:22:31.070 --> 00:22:32.670 align:middle line:84% AUDIENCE: That saved money is accrued 00:22:32.670 --> 00:22:35.670 align:middle line:84% just generally or to the actual person that saved it? 00:22:35.670 --> 00:22:37.613 align:middle line:90% JON GRUBER: Ah, great question. 00:22:37.613 --> 00:22:39.030 align:middle line:84% As we'll talk about in tax, what's 00:22:39.030 --> 00:22:42.310 align:middle line:90% the incidence of that savings? 00:22:42.310 --> 00:22:46.310 align:middle line:84% And basically, we think that a lot of it 00:22:46.310 --> 00:22:49.653 align:middle line:84% has gone to the insurers, and some has gone to the people. 00:22:49.653 --> 00:22:51.070 align:middle line:84% In a perfectly competitive market, 00:22:51.070 --> 00:22:52.550 align:middle line:90% it would all go to the people. 00:22:52.550 --> 00:22:55.470 align:middle line:84% In fact, one test of lack of perfect competition insurance 00:22:55.470 --> 00:22:58.230 align:middle line:84% markets is showing that when insurers save money, 00:22:58.230 --> 00:22:59.512 align:middle line:90% their profits go up. 00:22:59.512 --> 00:23:01.470 align:middle line:84% In a perfect competitive market, they wouldn't. 00:23:01.470 --> 00:23:03.730 align:middle line:84% So they have an imperfect insurance market, 00:23:03.730 --> 00:23:06.070 align:middle line:84% and so they've been keeping a lot of it. 00:23:06.070 --> 00:23:09.230 align:middle line:84% Now, that raises the other question, which is, 00:23:09.230 --> 00:23:11.830 align:middle line:84% what about health care-- what about quality? 00:23:11.830 --> 00:23:14.390 align:middle line:84% Have they saved money by going too far? 00:23:14.390 --> 00:23:16.670 align:middle line:90% And the answer is, no. 00:23:16.670 --> 00:23:20.030 align:middle line:84% The answer is that the quality of care, overall, delivered 00:23:20.030 --> 00:23:23.240 align:middle line:84% by these HMOs was no lower than it 00:23:23.240 --> 00:23:25.490 align:middle line:84% was delivered by traditional fee-for-service medicine. 00:23:25.490 --> 00:23:26.448 align:middle line:90% Now, there was a range. 00:23:26.448 --> 00:23:27.782 align:middle line:90% Some studies found it was lower. 00:23:27.782 --> 00:23:29.250 align:middle line:84% Some studies found it was higher. 00:23:29.250 --> 00:23:31.990 align:middle line:84% But the consensus of the literature is basically zero. 00:23:31.990 --> 00:23:34.240 align:middle line:84% The consensus of the literature is these organizations 00:23:34.240 --> 00:23:39.480 align:middle line:84% are delivering comparable quality care, at least 00:23:39.480 --> 00:23:42.560 align:middle line:84% in terms of patient health, comparable quality of care 00:23:42.560 --> 00:23:45.560 align:middle line:84% in terms of patient health, at lower cost. 00:23:45.560 --> 00:23:46.160 align:middle line:90% Why? 00:23:46.160 --> 00:23:47.785 align:middle line:84% Because we're on the flat of the curve. 00:23:47.785 --> 00:23:49.400 align:middle line:84% Because doctors under fee-for-service 00:23:49.400 --> 00:23:52.760 align:middle line:84% were delivering excessive medical care. 00:23:52.760 --> 00:23:55.720 align:middle line:84% These HMOs managed to get rid of the excess 00:23:55.720 --> 00:23:57.232 align:middle line:90% without going too far. 00:23:57.232 --> 00:23:59.440 align:middle line:84% Now, there's a question-- why didn't they go too far? 00:23:59.440 --> 00:24:02.440 align:middle line:84% Why doesn't Ahmet see a patient every 15 seconds 00:24:02.440 --> 00:24:04.178 align:middle line:84% and just make an oodles of money? 00:24:04.178 --> 00:24:05.720 align:middle line:84% And the answer is because there seems 00:24:05.720 --> 00:24:07.595 align:middle line:84% to be enough incentives in our system pushing 00:24:07.595 --> 00:24:11.680 align:middle line:84% the other way that doctors still don't undertreat people. 00:24:11.680 --> 00:24:14.580 align:middle line:84% In particular, medical malpractice pressure-- 00:24:14.580 --> 00:24:15.517 align:middle line:90% they might get sued. 00:24:15.517 --> 00:24:17.600 align:middle line:84% Or just norms-- doctors still have patient utility 00:24:17.600 --> 00:24:20.137 align:middle line:84% in their-- patient health in their utility function. 00:24:20.137 --> 00:24:22.720 align:middle line:84% So the bottom line is, it feels like this move to managed care 00:24:22.720 --> 00:24:25.137 align:middle line:84% was a move which managed to move us in the right direction 00:24:25.137 --> 00:24:30.010 align:middle line:84% and not too far, if you measure by patient health outcomes. 00:24:30.010 --> 00:24:31.570 align:middle line:84% But there is another reason people 00:24:31.570 --> 00:24:34.885 align:middle line:84% like managed care, which is it's just a less pleasant experience. 00:24:34.885 --> 00:24:37.010 align:middle line:84% It used to be you'd go to whatever doctor you want, 00:24:37.010 --> 00:24:38.490 align:middle line:90% and they just treated you. 00:24:38.490 --> 00:24:41.110 align:middle line:84% Now they have to review what procedure you're going to get. 00:24:41.110 --> 00:24:42.710 align:middle line:84% There's a bunch of extra paperwork. 00:24:42.710 --> 00:24:43.550 align:middle line:90% It's extra time. 00:24:43.550 --> 00:24:45.010 align:middle line:90% You might have to wait longer. 00:24:45.010 --> 00:24:48.810 align:middle line:84% So they've made the incidence in terms of who's won-- 00:24:48.810 --> 00:24:51.770 align:middle line:84% part of it is patients have lost because it's 00:24:51.770 --> 00:24:54.895 align:middle line:84% a less pleasant experience, but it's a lot cheaper. 00:24:54.895 --> 00:24:56.770 align:middle line:84% This is a little bit like-- if you remember-- 00:24:56.770 --> 00:24:58.353 align:middle line:84% those who took 14.01 with me, remember 00:24:58.353 --> 00:25:00.130 align:middle line:84% I taught you about airline deregulation, 00:25:00.130 --> 00:25:01.810 align:middle line:84% where everyone bitches about airline service, 00:25:01.810 --> 00:25:03.185 align:middle line:84% but no one bitches about the fact 00:25:03.185 --> 00:25:05.690 align:middle line:84% that airline-- that flights are a third, lower, third less 00:25:05.690 --> 00:25:07.857 align:middle line:84% expensive than they used to be before we deregulated 00:25:07.857 --> 00:25:10.273 align:middle line:84% because, basically, they were providing excessive quality. 00:25:10.273 --> 00:25:12.397 align:middle line:84% Doctors were providing excessive quality before. 00:25:12.397 --> 00:25:14.730 align:middle line:84% They're providing less, which is a little less pleasant, 00:25:14.730 --> 00:25:17.130 align:middle line:84% but we all like having the savings. 00:25:17.130 --> 00:25:20.850 align:middle line:90% So that is the bottom line. 00:25:20.850 --> 00:25:23.057 align:middle line:84% Where does this leave us on optimal-- 00:25:23.057 --> 00:25:24.890 align:middle line:84% so we talk about optimal design of insurance 00:25:24.890 --> 00:25:27.515 align:middle line:84% from the patient's perspective, which is about high-design cost 00:25:27.515 --> 00:25:28.180 align:middle line:90% sharing? 00:25:28.180 --> 00:25:30.412 align:middle line:84% What an optimal design of reimbursement. 00:25:30.412 --> 00:25:32.620 align:middle line:84% Where this leaves the optimal design of reimbursement 00:25:32.620 --> 00:25:35.580 align:middle line:84% is that, basically, a pure fee-for-service system almost 00:25:35.580 --> 00:25:38.980 align:middle line:90% certainly makes no sense. 00:25:38.980 --> 00:25:40.820 align:middle line:84% The evidence is good that a managed care 00:25:40.820 --> 00:25:44.700 align:middle line:84% system may be a better approach, but there 00:25:44.700 --> 00:25:47.100 align:middle line:84% remains concerns in certain contexts 00:25:47.100 --> 00:25:48.685 align:middle line:90% of providers going too far. 00:25:48.685 --> 00:25:51.060 align:middle line:84% So you really want a managed care system with some checks 00:25:51.060 --> 00:25:54.320 align:middle line:84% on it to make sure, so for example, 00:25:54.320 --> 00:25:57.100 align:middle line:84% something which may allow some escape valves so that providers 00:25:57.100 --> 00:25:59.860 align:middle line:84% don't-- or just monitoring to make sure providers don't go 00:25:59.860 --> 00:26:01.740 align:middle line:90% excessively too far. 00:26:01.740 --> 00:26:02.620 align:middle line:90% Yeah. 00:26:02.620 --> 00:26:04.900 align:middle line:84% AUDIENCE: The airline example [INAUDIBLE] 00:26:04.900 --> 00:26:07.700 align:middle line:84% an important part of that is that people are experiencing 00:26:07.700 --> 00:26:09.400 align:middle line:90% much lower flight rates now. 00:26:09.400 --> 00:26:11.200 align:middle line:90% Is that the case for insurance? 00:26:11.200 --> 00:26:12.415 align:middle line:90% JON GRUBER: Much lower what? 00:26:12.415 --> 00:26:15.040 align:middle line:84% AUDIENCE: Are people paying much less for insurance these days? 00:26:15.040 --> 00:26:16.060 align:middle line:90% JON GRUBER: Yes. 00:26:16.060 --> 00:26:17.720 align:middle line:84% Well, it's just a time series, right? 00:26:17.720 --> 00:26:19.300 align:middle line:84% But certainly, insurance is cheaper 00:26:19.300 --> 00:26:21.340 align:middle line:84% than it would be under the old system. 00:26:21.340 --> 00:26:22.260 align:middle line:90% Yeah. 00:26:22.260 --> 00:26:25.680 align:middle line:84% AUDIENCE: This sounds like vertically integrating. 00:26:25.680 --> 00:26:29.160 align:middle line:84% JON GRUBER: Yeah, this model is very much vertical integration. 00:26:29.160 --> 00:26:34.640 align:middle line:84% AUDIENCE: So in the [INAUDIBLE] this would also 00:26:34.640 --> 00:26:36.360 align:middle line:90% have some effect [INAUDIBLE]. 00:26:36.360 --> 00:26:38.520 align:middle line:84% JON GRUBER: Well that's a great question. 00:26:38.520 --> 00:26:40.180 align:middle line:84% We don't do antitrust in this class. 00:26:40.180 --> 00:26:43.080 align:middle line:84% You can take 1420, which is a terrific class to learn more 00:26:43.080 --> 00:26:44.462 align:middle line:90% about antitrust issues. 00:26:44.462 --> 00:26:45.920 align:middle line:84% But basically, that's the trade-off 00:26:45.920 --> 00:26:49.320 align:middle line:84% with antitrust in a health care setting 00:26:49.320 --> 00:26:52.360 align:middle line:84% is that vertical integration can deliver efficiencies 00:26:52.360 --> 00:26:54.980 align:middle line:84% of this type, but it also lowers competition. 00:26:54.980 --> 00:26:59.560 align:middle line:90% 00:26:59.560 --> 00:27:03.403 align:middle line:84% Horizontal mergers are very bad for competition, obviously. 00:27:03.403 --> 00:27:05.320 align:middle line:84% Vertical are not obviously bad for competition 00:27:05.320 --> 00:27:06.320 align:middle line:84% because the insurer and the doctor 00:27:06.320 --> 00:27:07.900 align:middle line:84% weren't competing with each other. 00:27:07.900 --> 00:27:10.040 align:middle line:84% The question is whether you end up 00:27:10.040 --> 00:27:12.720 align:middle line:84% with these vertical institutions that self-refer 00:27:12.720 --> 00:27:14.600 align:middle line:90% and end up limiting competition. 00:27:14.600 --> 00:27:19.600 align:middle line:84% And so there's a rigorous debate about that. 00:27:19.600 --> 00:27:20.100 align:middle line:90% Yeah. 00:27:20.100 --> 00:27:20.680 align:middle line:90% Enoch? 00:27:20.680 --> 00:27:22.347 align:middle line:84% AUDIENCE: So we think the quality hasn't 00:27:22.347 --> 00:27:24.250 align:middle line:84% changed that much, but the experience 00:27:24.250 --> 00:27:25.412 align:middle line:90% is much less enjoyable. 00:27:25.412 --> 00:27:27.370 align:middle line:84% And there's a lot more procedures and paperwork 00:27:27.370 --> 00:27:28.950 align:middle line:90% that people have to go through. 00:27:28.950 --> 00:27:32.810 align:middle line:84% Does that point towards-- not the purpose, because that's-- 00:27:32.810 --> 00:27:36.210 align:middle line:84% I get that [INAUDIBLE], but is the point of health insurance 00:27:36.210 --> 00:27:39.890 align:middle line:84% that's if you don't die or just that your utility function-- 00:27:39.890 --> 00:27:42.110 align:middle line:84% JON GRUBER: Well, that is a great question. 00:27:42.110 --> 00:27:46.810 align:middle line:84% I mean, in some sense, this comes to the very fact 00:27:46.810 --> 00:27:49.850 align:middle line:84% that economists think about health insurance 00:27:49.850 --> 00:27:51.610 align:middle line:90% differently than people do. 00:27:51.610 --> 00:27:54.290 align:middle line:84% Not that we're not people, but you know what I mean. 00:27:54.290 --> 00:27:56.502 align:middle line:84% Basically, economists talk about health insurance 00:27:56.502 --> 00:27:58.210 align:middle line:84% the way I taught you in this class, which 00:27:58.210 --> 00:28:01.450 align:middle line:84% is as a means of protecting you against risk. 00:28:01.450 --> 00:28:04.810 align:middle line:84% That is something which says that, for example, rich people 00:28:04.810 --> 00:28:06.730 align:middle line:90% should have high deductibles. 00:28:06.730 --> 00:28:10.005 align:middle line:84% People think about health insurance as medical prepayment. 00:28:10.005 --> 00:28:10.630 align:middle line:90% Think about it. 00:28:10.630 --> 00:28:11.880 align:middle line:90% I've already paid my premiums. 00:28:11.880 --> 00:28:13.970 align:middle line:84% I don't want to have to deal with anything else. 00:28:13.970 --> 00:28:16.330 align:middle line:84% Now, we can tell them till we're blue in the face they're wrong. 00:28:16.330 --> 00:28:17.747 align:middle line:84% That's not the point of insurance. 00:28:17.747 --> 00:28:19.830 align:middle line:84% But that's what people-- part of what people like. 00:28:19.830 --> 00:28:21.370 align:middle line:84% So for example, people heat the idea 00:28:21.370 --> 00:28:23.203 align:middle line:84% of having to pay when they go to the doctor. 00:28:23.203 --> 00:28:24.840 align:middle line:84% And you can explain, well, actually, 00:28:24.840 --> 00:28:26.165 align:middle line:90% it's a small price to pay-- 00:28:26.165 --> 00:28:27.540 align:middle line:84% it lowers your premium, whatever. 00:28:27.540 --> 00:28:28.060 align:middle line:90% No. 00:28:28.060 --> 00:28:28.840 align:middle line:90% They hate it. 00:28:28.840 --> 00:28:29.900 align:middle line:84% So if you really think of insurance 00:28:29.900 --> 00:28:31.940 align:middle line:84% as just something which makes your-- you pay once every month 00:28:31.940 --> 00:28:33.980 align:middle line:84% and forget about it, then many people 00:28:33.980 --> 00:28:35.540 align:middle line:84% would rather pay more every month 00:28:35.540 --> 00:28:37.860 align:middle line:84% than have a hassle-free life later on. 00:28:37.860 --> 00:28:39.060 align:middle line:90% And the trick is-- 00:28:39.060 --> 00:28:41.780 align:middle line:84% I'll come back to this later in this lecture-- do they really 00:28:41.780 --> 00:28:42.913 align:middle line:90% have that choice or not? 00:28:42.913 --> 00:28:44.080 align:middle line:90% And we'll come back to that. 00:28:44.080 --> 00:28:44.580 align:middle line:90% Yeah. 00:28:44.580 --> 00:28:46.820 align:middle line:84% AUDIENCE: I understand the different ways 00:28:46.820 --> 00:28:49.340 align:middle line:84% of paying providers offer different incentives 00:28:49.340 --> 00:28:50.340 align:middle line:90% for providers. 00:28:50.340 --> 00:28:53.223 align:middle line:84% What happens when one provider is paid by multiple payers? 00:28:53.223 --> 00:28:54.640 align:middle line:84% JON GRUBER: It's a great question. 00:28:54.640 --> 00:28:59.298 align:middle line:84% Very complicated and interesting question is, does it spill over? 00:28:59.298 --> 00:29:01.340 align:middle line:84% If half my patients are managed care and half are 00:29:01.340 --> 00:29:02.900 align:middle line:84% fee-for-service, does that affect 00:29:02.900 --> 00:29:04.608 align:middle line:84% how I treat the fee-for-service patients, 00:29:04.608 --> 00:29:08.060 align:middle line:84% compared to all-- and the evidence is not great on that. 00:29:08.060 --> 00:29:08.820 align:middle line:90% OK, yeah. 00:29:08.820 --> 00:29:10.420 align:middle line:84% AUDIENCE: I just have a question. 00:29:10.420 --> 00:29:13.300 align:middle line:84% So for fee-for-service-- is incentive for the provider to be 00:29:13.300 --> 00:29:15.040 align:middle line:84% efficient and be more productive-- 00:29:15.040 --> 00:29:15.440 align:middle line:90% JON GRUBER: Right. 00:29:15.440 --> 00:29:16.260 align:middle line:84% AUDIENCE: --even though he might be 00:29:16.260 --> 00:29:18.540 align:middle line:84% on the other side of the curve where the benefit to the patient 00:29:18.540 --> 00:29:19.540 align:middle line:90% may not be high. 00:29:19.540 --> 00:29:22.670 align:middle line:84% But I would imagine that with the managed care 00:29:22.670 --> 00:29:24.090 align:middle line:90% system, the incentive for-- 00:29:24.090 --> 00:29:25.590 align:middle line:84% JON GRUBER: No, there's no incentive 00:29:25.590 --> 00:29:27.430 align:middle line:84% to be more productive with fee-for-service. 00:29:27.430 --> 00:29:29.990 align:middle line:84% There's incentive to do more stuff, 00:29:29.990 --> 00:29:31.390 align:middle line:90% but that's not productivity. 00:29:31.390 --> 00:29:33.150 align:middle line:84% AUDIENCE: In the sense that if you look at [INAUDIBLE]. 00:29:33.150 --> 00:29:34.610 align:middle line:84% JON GRUBER: The incentive to work harder. 00:29:34.610 --> 00:29:35.777 align:middle line:90% AUDIENCE: Yeah, work harder. 00:29:35.777 --> 00:29:39.080 align:middle line:84% JON GRUBER: Yeah, but working harder-- 00:29:39.080 --> 00:29:40.632 align:middle line:90% it's an interesting point. 00:29:40.632 --> 00:29:42.590 align:middle line:84% It flips on its head what we learned about when 00:29:42.590 --> 00:29:43.870 align:middle line:90% we thought about labor supply. 00:29:43.870 --> 00:29:44.930 align:middle line:84% Talking about labor supply, we're like, 00:29:44.930 --> 00:29:46.890 align:middle line:84% look don't want to get in the way of people working hard. 00:29:46.890 --> 00:29:49.223 align:middle line:84% Well, the problem is people working hard on labor supply 00:29:49.223 --> 00:29:50.665 align:middle line:90% is increasing social welfare. 00:29:50.665 --> 00:29:52.790 align:middle line:84% People working hard here is lowering social welfare 00:29:52.790 --> 00:29:54.457 align:middle line:84% because they're limiting a bunch of care 00:29:54.457 --> 00:29:57.550 align:middle line:84% that, as Feldstein showed, has deadweight loss. 00:29:57.550 --> 00:29:59.710 align:middle line:90% So it's not productivity. 00:29:59.710 --> 00:30:03.270 align:middle line:90% It's a good distinction. 00:30:03.270 --> 00:30:08.470 align:middle line:84% So now, that's a lecture and a half on basic health ace 00:30:08.470 --> 00:30:08.977 align:middle line:90% economics. 00:30:08.977 --> 00:30:11.310 align:middle line:84% Now we're going to come for the next lecture and a half, 00:30:11.310 --> 00:30:14.710 align:middle line:84% to the government programs and how we can take the tools 00:30:14.710 --> 00:30:16.785 align:middle line:84% we just learned and use them to think 00:30:16.785 --> 00:30:18.910 align:middle line:84% about optimal design of government health insurance 00:30:18.910 --> 00:30:19.830 align:middle line:90% programs. 00:30:19.830 --> 00:30:22.040 align:middle line:90% So let's start with Medicaid. 00:30:22.040 --> 00:30:24.840 align:middle line:90% Medicaid is insurance for the? 00:30:24.840 --> 00:30:25.680 align:middle line:90% Poor. 00:30:25.680 --> 00:30:27.560 align:middle line:84% Medicare insurance for the elderly. 00:30:27.560 --> 00:30:30.027 align:middle line:84% So let's start with Medicaid, insurance for the poor. 00:30:30.027 --> 00:30:32.360 align:middle line:84% Let me-- now, like I did with the other social insurance 00:30:32.360 --> 00:30:36.240 align:middle line:84% programs, let's do five minutes on institutions, and let's dive 00:30:36.240 --> 00:30:38.160 align:middle line:90% into the economics. 00:30:38.160 --> 00:30:39.540 align:middle line:90% So what is Medicaid? 00:30:39.540 --> 00:30:45.320 align:middle line:84% Medicaid is a federally subsidized but state-run program 00:30:45.320 --> 00:30:49.520 align:middle line:84% that provides health insurance to low-income individuals who 00:30:49.520 --> 00:30:53.140 align:middle line:90% reside in that state. 00:30:53.140 --> 00:30:55.520 align:middle line:90% Now, who is eligible? 00:30:55.520 --> 00:30:57.060 align:middle line:84% That has changed a lot over time. 00:30:57.060 --> 00:31:00.320 align:middle line:84% Traditionally, Medicaid began as a program 00:31:00.320 --> 00:31:02.480 align:middle line:84% for just those who were very poor, 00:31:02.480 --> 00:31:05.560 align:middle line:84% who were getting cash welfare checks from the state. 00:31:05.560 --> 00:31:08.240 align:middle line:84% So think people whose earnings were, like, 00:31:08.240 --> 00:31:11.040 align:middle line:90% $5,000 a year or less, tiny. 00:31:11.040 --> 00:31:18.980 align:middle line:84% Then during the 1980s and 1990s, it expanded for two groups, 00:31:18.980 --> 00:31:22.980 align:middle line:84% for children, especially children under 5, 00:31:22.980 --> 00:31:24.320 align:middle line:90% and for pregnant women. 00:31:24.320 --> 00:31:28.060 align:middle line:84% It expanded quite a lot, to the point where, 00:31:28.060 --> 00:31:29.820 align:middle line:84% in many states people were eligible-- 00:31:29.820 --> 00:31:32.200 align:middle line:84% instead of-- if you weren't pregnant as a woman, 00:31:32.200 --> 00:31:35.480 align:middle line:84% you couldn't get Medicaid if your income was above $5,000. 00:31:35.480 --> 00:31:37.980 align:middle line:84% But if you were pregnant, you'd get it if your income was up 00:31:37.980 --> 00:31:38.740 align:middle line:90% to $50,000. 00:31:38.740 --> 00:31:42.080 align:middle line:84% So there was a huge expansion up into the middle class, 00:31:42.080 --> 00:31:44.560 align:middle line:84% from the very poor into the middle class for those groups. 00:31:44.560 --> 00:31:45.060 align:middle line:90% Yeah. 00:31:45.060 --> 00:31:46.220 align:middle line:90% AUDIENCE: 5 or 5-0? 00:31:46.220 --> 00:31:47.380 align:middle line:90% JON GRUBER: 5-0. 00:31:47.380 --> 00:31:49.900 align:middle line:84% it was it's a huge expansion up into the middle class 00:31:49.900 --> 00:31:52.860 align:middle line:90% for those groups. 00:31:52.860 --> 00:31:55.980 align:middle line:84% And basically, that happened-- that kept going through 00:31:55.980 --> 00:31:58.100 align:middle line:90% the 1990s. 00:31:58.100 --> 00:32:03.180 align:middle line:84% And basically, you had a situation, by the mid-- 00:32:03.180 --> 00:32:10.260 align:middle line:84% by the year-- into the 2000s where, basically, essentially, 00:32:10.260 --> 00:32:13.220 align:middle line:84% basically, any kid, anyone from pregnancy was eligible up 00:32:13.220 --> 00:32:15.160 align:middle line:84% to about the typical income in America. 00:32:15.160 --> 00:32:17.670 align:middle line:84% So basically, if you were below typical income in America, 00:32:17.670 --> 00:32:20.630 align:middle line:84% in most states, you're eligible for health insurance as a kid 00:32:20.630 --> 00:32:22.510 align:middle line:90% or for expense of pregnancy. 00:32:22.510 --> 00:32:25.730 align:middle line:84% But except for those groups, it was very hard to get. 00:32:25.730 --> 00:32:27.790 align:middle line:90% You had to be very, very poor. 00:32:27.790 --> 00:32:30.030 align:middle line:84% That changed with the Affordable Care Act. 00:32:30.030 --> 00:32:33.050 align:middle line:84% The Affordable Care Act made it a universal, 00:32:33.050 --> 00:32:35.070 align:middle line:84% what we call means-tested program, 00:32:35.070 --> 00:32:38.370 align:middle line:84% which is for everyone below 133% of the poverty line, which, 00:32:38.370 --> 00:32:40.790 align:middle line:84% as I said, for a family, is probably about $30,000 00:32:40.790 --> 00:32:44.830 align:middle line:84% or $32,000 a year, you now get Medicaid-- 00:32:44.830 --> 00:32:46.690 align:middle line:90% kid, adult, whatever. 00:32:46.690 --> 00:32:48.690 align:middle line:84% Now, the kids still get it up to higher incomes. 00:32:48.690 --> 00:32:50.330 align:middle line:90% Kids still get it up to higher. 00:32:50.330 --> 00:32:54.910 align:middle line:84% But for everyone now, they get it, 133% of poverty line. 00:32:54.910 --> 00:32:59.150 align:middle line:84% But because of a Supreme Court decision in 2012, 00:32:59.150 --> 00:33:01.210 align:middle line:84% states can decide if they want to offer this. 00:33:01.210 --> 00:33:03.110 align:middle line:84% And I'll come back to this next time. 00:33:03.110 --> 00:33:06.710 align:middle line:84% But for reasons that I will charitably 00:33:06.710 --> 00:33:09.830 align:middle line:84% characterize as political malpractice, a number of states 00:33:09.830 --> 00:33:11.250 align:middle line:90% have not taken up this option. 00:33:11.250 --> 00:33:14.190 align:middle line:84% So in about 10 to 12 states now, you still 00:33:14.190 --> 00:33:18.160 align:middle line:84% can't get Medicaid if you're poor, even if you're 00:33:18.160 --> 00:33:20.500 align:middle line:90% at 100% of the poverty line. 00:33:20.500 --> 00:33:21.000 align:middle line:90% Yeah. 00:33:21.000 --> 00:33:22.520 align:middle line:84% AUDIENCE: The money comes the state? 00:33:22.520 --> 00:33:23.160 align:middle line:90% JON GRUBER: From the feds. 00:33:23.160 --> 00:33:23.880 align:middle line:90% No. 00:33:23.880 --> 00:33:29.380 align:middle line:84% Medicaid, in general, is basically 67% fed, 33% state. 00:33:29.380 --> 00:33:32.280 align:middle line:84% And the formula is an inverse function of state income. 00:33:32.280 --> 00:33:35.400 align:middle line:84% So basically, Massachusetts, 50/50. 00:33:35.400 --> 00:33:39.180 align:middle line:90% Arkansas, it's, like, 70/30. 00:33:39.180 --> 00:33:39.680 align:middle line:90% I'm sorry. 00:33:39.680 --> 00:33:41.638 align:middle line:84% On average, the states-- the federal government 00:33:41.638 --> 00:33:44.800 align:middle line:84% pays about 56% of the cost of Medicaid. 00:33:44.800 --> 00:33:48.540 align:middle line:84% But these expansions under the ACA were 100% federal. 00:33:48.540 --> 00:33:50.163 align:middle line:84% So they went to the states and said, 00:33:50.163 --> 00:33:51.580 align:middle line:84% we'll pay all the costs to expand. 00:33:51.580 --> 00:33:53.600 align:middle line:84% A number of states still said, no. 00:33:53.600 --> 00:33:56.420 align:middle line:90% So that's who's eligible. 00:33:56.420 --> 00:33:57.400 align:middle line:90% What's covered? 00:33:57.400 --> 00:34:00.160 align:middle line:84% Basically, Medicaid is really good insurance. 00:34:00.160 --> 00:34:03.280 align:middle line:84% I like to call it the best insurance money can't buy. 00:34:03.280 --> 00:34:07.160 align:middle line:84% It covers pretty much everything, pretty much free. 00:34:07.160 --> 00:34:09.280 align:middle line:84% So basically, any medical service 00:34:09.280 --> 00:34:11.440 align:middle line:90% is pretty much covered for free. 00:34:11.440 --> 00:34:14.040 align:middle line:84% In many states, that includes dental, vision, et cetera, 00:34:14.040 --> 00:34:18.510 align:middle line:84% things that many health insurance plans don't cover. 00:34:18.510 --> 00:34:20.250 align:middle line:90% That's the good news. 00:34:20.250 --> 00:34:23.730 align:middle line:84% The bad news is that state Medicaid programs-- 00:34:23.730 --> 00:34:26.070 align:middle line:84% they don't have a lot of choice over what they cover. 00:34:26.070 --> 00:34:28.310 align:middle line:84% They have a choice over two things, who's covered-- 00:34:28.310 --> 00:34:30.790 align:middle line:84% and as I said, the ACA's taken a lot of that choice away. 00:34:30.790 --> 00:34:33.090 align:middle line:84% So pretty much, you have to cover poor people. 00:34:33.090 --> 00:34:35.402 align:middle line:90% And how much providers get paid. 00:34:35.402 --> 00:34:37.610 align:middle line:84% And states vary a lot in how much providers get paid. 00:34:37.610 --> 00:34:39.193 align:middle line:84% And in a particular, number of states, 00:34:39.193 --> 00:34:41.170 align:middle line:84% providers get paid very, very poorly 00:34:41.170 --> 00:34:45.429 align:middle line:84% for seeing Medicaid patients, typically on the order of-- it 00:34:45.429 --> 00:34:47.770 align:middle line:84% can be as low as 10% of what they get for a privately 00:34:47.770 --> 00:34:49.130 align:middle line:90% insured patient. 00:34:49.130 --> 00:34:55.730 align:middle line:84% So for many, many years, doctors in New York-- 00:34:55.730 --> 00:34:57.150 align:middle line:90% if they saw a Medicaid patient-- 00:34:57.150 --> 00:34:58.317 align:middle line:90% I don't know what it is now. 00:34:58.317 --> 00:35:00.850 align:middle line:84% You got about $12, where you're getting $200 00:35:00.850 --> 00:35:04.230 align:middle line:90% for seeing a private patient. 00:35:04.230 --> 00:35:06.490 align:middle line:84% And this has been shown in survey after survey 00:35:06.490 --> 00:35:09.770 align:middle line:84% to cause doctors to not want to see Medicaid patients. 00:35:09.770 --> 00:35:10.797 align:middle line:90% And that's legal. 00:35:10.797 --> 00:35:13.130 align:middle line:84% It's illegal to turn someone away at the emergency room, 00:35:13.130 --> 00:35:16.540 align:middle line:84% but it's legal to turn someone away at the doctor's office. 00:35:16.540 --> 00:35:19.600 align:middle line:84% So as a result, there was-- on the one hand, Medicaid 00:35:19.600 --> 00:35:22.980 align:middle line:84% this great insurance, but you can't find doctors to take it, 00:35:22.980 --> 00:35:23.740 align:middle line:90% is the challenge. 00:35:23.740 --> 00:35:25.448 align:middle line:84% In particular, the problem's particularly 00:35:25.448 --> 00:35:28.058 align:middle line:84% acute in rural areas and inner city areas. 00:35:28.058 --> 00:35:30.100 align:middle line:84% It's particularly acute finding doctors that will 00:35:30.100 --> 00:35:31.600 align:middle line:90% take the Medicaid coverage. 00:35:31.600 --> 00:35:35.340 align:middle line:84% So it's a mixed-- it's a mixed bag. 00:35:35.340 --> 00:35:36.820 align:middle line:90% That's Medicaid. 00:35:36.820 --> 00:35:41.403 align:middle line:84% Now, let's go back to the chapter 12 framework Moral 00:35:41.403 --> 00:35:43.320 align:middle line:84% hazard-- we don't need to cover with Medicaid. 00:35:43.320 --> 00:35:44.500 align:middle line:84% Basically, it's got all the moral hazard 00:35:44.500 --> 00:35:45.900 align:middle line:84% free health insurance would have, 00:35:45.900 --> 00:35:48.108 align:middle line:84% subject to the fact that if you can't see the doctor, 00:35:48.108 --> 00:35:49.820 align:middle line:84% there's no scope for moral hazard. 00:35:49.820 --> 00:35:52.703 align:middle line:84% Let's talk about, though, the other side, the benefits. 00:35:52.703 --> 00:35:55.120 align:middle line:84% So how do we think about the benefits of health insurance? 00:35:55.120 --> 00:35:58.020 align:middle line:84% Well, one benefit is the consumption-smoothing benefit. 00:35:58.020 --> 00:36:01.180 align:middle line:84% Does having Medicaid help you smooth your consumption, 00:36:01.180 --> 00:36:03.740 align:middle line:90% in particular when you get ill? 00:36:03.740 --> 00:36:06.645 align:middle line:90% And the answer is strongly, yes. 00:36:06.645 --> 00:36:08.020 align:middle line:84% There's a number of studies which 00:36:08.020 --> 00:36:12.180 align:middle line:84% show that when Medicaid expands and people get Medicaid, 00:36:12.180 --> 00:36:14.610 align:middle line:84% they are much less likely to go bankrupt. 00:36:14.610 --> 00:36:16.970 align:middle line:84% They're much more-- the bankruptcy's mixed. 00:36:16.970 --> 00:36:20.030 align:middle line:84% But much less likely to have bad credit scores, much more likely 00:36:20.030 --> 00:36:22.910 align:middle line:84% to be able to have more to eat, much more likely to be 00:36:22.910 --> 00:36:25.790 align:middle line:90% financially stable and secure. 00:36:25.790 --> 00:36:31.283 align:middle line:84% So it looks like it is smoothing consumption of individuals. 00:36:31.283 --> 00:36:32.450 align:middle line:90% So it does lower bankruptcy. 00:36:32.450 --> 00:36:33.590 align:middle line:90% That's right. 00:36:33.590 --> 00:36:36.050 align:middle line:90% So that's the one piece. 00:36:36.050 --> 00:36:39.110 align:middle line:84% And in some sense, that's the main piece. 00:36:39.110 --> 00:36:40.050 align:middle line:90% It is health insurers. 00:36:40.050 --> 00:36:41.830 align:middle line:84% But remember, the role of health insurance 00:36:41.830 --> 00:36:43.950 align:middle line:90% is to protect you financially. 00:36:43.950 --> 00:36:45.450 align:middle line:84% That's the goal of health insurance, 00:36:45.450 --> 00:36:47.110 align:middle line:90% is to protect you financially. 00:36:47.110 --> 00:36:48.910 align:middle line:84% And the answer is Medicaid does that job. 00:36:48.910 --> 00:36:52.030 align:middle line:84% People are not well self-insured against these medical risks. 00:36:52.030 --> 00:36:55.590 align:middle line:84% And when they get Medicaid, they're protected financially. 00:36:55.590 --> 00:36:57.840 align:middle line:84% However, really, truly, what most people seem to think 00:36:57.840 --> 00:36:59.632 align:middle line:84% is they want to know if health insurance is 00:36:59.632 --> 00:37:00.910 align:middle line:90% good for your health or not. 00:37:00.910 --> 00:37:03.470 align:middle line:84% That's what we want to know with health care. 00:37:03.470 --> 00:37:06.390 align:middle line:84% And so the question then is, does having Medicaid actually 00:37:06.390 --> 00:37:07.630 align:middle line:90% improve your health? 00:37:07.630 --> 00:37:11.290 align:middle line:84% And it turns out this is a complicated question. 00:37:11.290 --> 00:37:14.290 align:middle line:84% And so to do that, I developed this highly technical chart, 00:37:14.290 --> 00:37:17.650 align:middle line:84% which is the first chart we have, figure 16-1, 00:37:17.650 --> 00:37:20.930 align:middle line:84% which talks about the steps through which making someone 00:37:20.930 --> 00:37:23.730 align:middle line:84% eligible for Medicaid might actually lead to improved 00:37:23.730 --> 00:37:26.050 align:middle line:90% health. 00:37:26.050 --> 00:37:27.830 align:middle line:90% Let's take these ACA expansions. 00:37:27.830 --> 00:37:30.198 align:middle line:84% We increase eligibility for Medicaid. 00:37:30.198 --> 00:37:31.990 align:middle line:84% How does that translate to improved health? 00:37:31.990 --> 00:37:33.810 align:middle line:84% Well, it does so through a number of steps. 00:37:33.810 --> 00:37:36.330 align:middle line:84% The first step is, well, you have to actually take 00:37:36.330 --> 00:37:38.170 align:middle line:90% that health insurance. 00:37:38.170 --> 00:37:40.130 align:middle line:84% So for someone who was previously 00:37:40.130 --> 00:37:42.570 align:middle line:84% uninsured to the right-- previously uninsured 00:37:42.570 --> 00:37:44.970 align:middle line:84% to the right, you have to take up. 00:37:44.970 --> 00:37:48.210 align:middle line:84% Remember, we said take-up is imperfect. 00:37:48.210 --> 00:37:52.690 align:middle line:84% Moreover, it may actually impact enrollment and insurance 00:37:52.690 --> 00:37:54.650 align:middle line:90% if you're already insured. 00:37:54.650 --> 00:37:59.450 align:middle line:84% Like I said, MIT makes me pay $6,000 a year for health 00:37:59.450 --> 00:38:02.570 align:middle line:84% insurance where I have a $3,000 deductible. 00:38:02.570 --> 00:38:06.050 align:middle line:84% Medicaid would pay nothing for a plan where everything's covered. 00:38:06.050 --> 00:38:08.798 align:middle line:84% I might think, gee, that's a pretty good deal. 00:38:08.798 --> 00:38:10.340 align:middle line:84% If I'm eligible for Medicaid, maybe I 00:38:10.340 --> 00:38:12.580 align:middle line:84% want to drop my private insurance 00:38:12.580 --> 00:38:15.140 align:middle line:90% and move over to Medicaid. 00:38:15.140 --> 00:38:17.240 align:middle line:84% That's a typical example of crowd-out, 00:38:17.240 --> 00:38:19.540 align:middle line:84% where you had a private insurance that's 00:38:19.540 --> 00:38:22.513 align:middle line:84% crowded out by this public insurance, 00:38:22.513 --> 00:38:23.680 align:middle line:90% and that can happen as well. 00:38:23.680 --> 00:38:25.860 align:middle line:84% So there's two ways that increased 00:38:25.860 --> 00:38:28.260 align:middle line:84% eligibility could lead to increased Medicaid coverage 00:38:28.260 --> 00:38:32.380 align:middle line:84% through take-up of the uninsured and crowd-out of the insured. 00:38:32.380 --> 00:38:35.980 align:middle line:84% But coverage then-- the question is, does that 00:38:35.980 --> 00:38:38.340 align:middle line:90% lead to increased utilization? 00:38:38.340 --> 00:38:41.300 align:middle line:84% And the answer is, it's not clear because of the doctor 00:38:41.300 --> 00:38:43.100 align:middle line:90% access problem. 00:38:43.100 --> 00:38:46.608 align:middle line:84% Coverage clearly will increase the demand for health care, 00:38:46.608 --> 00:38:48.900 align:middle line:84% but will it increase the actual delivery of health care 00:38:48.900 --> 00:38:51.395 align:middle line:84% depends on how supply-constrained you are. 00:38:51.395 --> 00:38:53.520 align:middle line:84% So the next question is, well, if you get coverage, 00:38:53.520 --> 00:38:56.405 align:middle line:84% does it actually increase their utilization? 00:38:56.405 --> 00:38:58.780 align:middle line:84% Then the next question is, does that utilization actually 00:38:58.780 --> 00:39:00.442 align:middle line:84% translate to better health outcomes? 00:39:00.442 --> 00:39:01.900 align:middle line:84% We've talked about a lot of studies 00:39:01.900 --> 00:39:04.233 align:middle line:84% which are showing that changing health insurance doesn't 00:39:04.233 --> 00:39:05.440 align:middle line:90% seem to change health. 00:39:05.440 --> 00:39:07.760 align:middle line:84% We saw that with the RAND Health Insurance Experiment. 00:39:07.760 --> 00:39:10.530 align:middle line:84% We saw that with these managed care example. 00:39:10.530 --> 00:39:12.930 align:middle line:84% So does it actually improve your health? 00:39:12.930 --> 00:39:14.430 align:middle line:84% And then the last step we care about 00:39:14.430 --> 00:39:17.950 align:middle line:84% as public finance economists is, does it do so cost-effectively, 00:39:17.950 --> 00:39:20.590 align:middle line:84% or should we just give people cash 00:39:20.590 --> 00:39:23.870 align:middle line:84% because it's a cost effective way to improve health? 00:39:23.870 --> 00:39:27.530 align:middle line:84% So to study this, there's been a large literature-- 00:39:27.530 --> 00:39:29.590 align:middle line:84% I've been a contributor to this literature-- 00:39:29.590 --> 00:39:33.870 align:middle line:84% which has looked at the effect of Medicaid on various outcomes. 00:39:33.870 --> 00:39:37.230 align:middle line:84% And what we have is we have the huge advantage that 00:39:37.230 --> 00:39:41.030 align:middle line:84% the expansions in the '80s and '90s in Medicaid happened 00:39:41.030 --> 00:39:45.670 align:middle line:84% in a very differential way across states and across 00:39:45.670 --> 00:39:47.590 align:middle line:90% demographic groups. 00:39:47.590 --> 00:39:52.470 align:middle line:84% So for example, if you look at table 16-1, 00:39:52.470 --> 00:39:57.390 align:middle line:84% this shows what happened to eligibility for Medicaid among 00:39:57.390 --> 00:40:00.390 align:middle line:90% kids in two states. 00:40:00.390 --> 00:40:04.430 align:middle line:84% So in 1982, 12% of kids in Missouri 00:40:04.430 --> 00:40:05.770 align:middle line:90% were eligible for Medicaid. 00:40:05.770 --> 00:40:08.160 align:middle line:90% 20% of kids in Michigan were. 00:40:08.160 --> 00:40:11.020 align:middle line:84% Michigan is higher income and had higher eligibility. 00:40:11.020 --> 00:40:11.520 align:middle line:90% Why? 00:40:11.520 --> 00:40:15.080 align:middle line:84% Because their program was more generous in 1982. 00:40:15.080 --> 00:40:17.820 align:middle line:84% By 2000, when the states had expanded, 00:40:17.820 --> 00:40:21.187 align:middle line:84% now 76% of all kids in Missouri are eligible for Medicaid. 00:40:21.187 --> 00:40:22.020 align:middle line:90% That's what I meant. 00:40:22.020 --> 00:40:23.680 align:middle line:84% It really went up to the middle class. 00:40:23.680 --> 00:40:27.840 align:middle line:84% It's 34% in Michigan because Michigan's a richer state. 00:40:27.840 --> 00:40:29.880 align:middle line:84% So you have a difference in difference 00:40:29.880 --> 00:40:32.960 align:middle line:84% here, where because of a national expansion that 00:40:32.960 --> 00:40:36.080 align:middle line:84% increased it to a certain poverty, certain income level, 00:40:36.080 --> 00:40:39.580 align:middle line:84% that had a much bigger effect in Missouri, 00:40:39.580 --> 00:40:42.000 align:middle line:84% where people are poor, than it did in Michigan, where 00:40:42.000 --> 00:40:42.800 align:middle line:90% people are rich. 00:40:42.800 --> 00:40:43.300 align:middle line:90% Yeah. 00:40:43.300 --> 00:40:45.120 align:middle line:84% AUDIENCE: The CHIP program [INAUDIBLE]? 00:40:45.120 --> 00:40:46.370 align:middle line:90% JON GRUBER: The CHIP program-- 00:40:46.370 --> 00:40:49.120 align:middle line:90% I'm folding it into Medicaid. 00:40:49.120 --> 00:40:52.880 align:middle line:84% So basically, that is a classic diff-in-diff 00:40:52.880 --> 00:40:54.600 align:middle line:90% we can use to study. 00:40:54.600 --> 00:40:57.160 align:middle line:84% But it's better than that because states 00:40:57.160 --> 00:41:00.240 align:middle line:84% expanded differentially for different ages of kids. 00:41:00.240 --> 00:41:10.270 align:middle line:84% So if you look at Missouri, in 1982, 18% of kids at age 13 00:41:10.270 --> 00:41:10.770 align:middle line:90% were-- 00:41:10.770 --> 00:41:12.853 align:middle line:84% I'm sorry, this is Washington DC, another example. 00:41:12.853 --> 00:41:16.330 align:middle line:84% In 1982, 18% of kids at age 13 were eligible, 00:41:16.330 --> 00:41:18.930 align:middle line:90% and 48% were eligible at birth. 00:41:18.930 --> 00:41:21.670 align:middle line:84% By 2000, the share at birth had barely budged, 00:41:21.670 --> 00:41:23.420 align:middle line:84% but the share at age 18 had gone up a lot. 00:41:23.420 --> 00:41:25.295 align:middle line:84% So that's another difference you can exploit, 00:41:25.295 --> 00:41:26.910 align:middle line:84% is different age kids within a state. 00:41:26.910 --> 00:41:29.153 align:middle line:84% So it's got all the fun chapter 3 stuff going on. 00:41:29.153 --> 00:41:30.570 align:middle line:84% Lots of differences in differences 00:41:30.570 --> 00:41:31.890 align:middle line:90% floating around here. 00:41:31.890 --> 00:41:33.833 align:middle line:84% And so lots of people have used this now. 00:41:33.833 --> 00:41:35.750 align:middle line:84% There's probably hundreds of articles written. 00:41:35.750 --> 00:41:37.970 align:middle line:84% And now, of course, we have a whole new literature 00:41:37.970 --> 00:41:41.210 align:middle line:84% because some states expanded under the ACA, and some didn't. 00:41:41.210 --> 00:41:42.553 align:middle line:90% So we had an old literature. 00:41:42.553 --> 00:41:43.970 align:middle line:84% Now we have a whole new literature 00:41:43.970 --> 00:41:46.370 align:middle line:90% that's using those experiments. 00:41:46.370 --> 00:41:48.250 align:middle line:90% So what have we learned? 00:41:48.250 --> 00:41:49.570 align:middle line:90% OK. 00:41:49.570 --> 00:41:54.117 align:middle line:84% First of all, we've learned that take-up is very impartial. 00:41:54.117 --> 00:41:56.450 align:middle line:84% If you look at a straight line from Medicare eligibility 00:41:56.450 --> 00:42:01.210 align:middle line:84% to Medicaid coverage and you ask for every person made eligible, 00:42:01.210 --> 00:42:03.530 align:middle line:84% what share actually gained coverage? 00:42:03.530 --> 00:42:08.020 align:middle line:84% The answer is about 20% to 25%, really low. 00:42:08.020 --> 00:42:13.340 align:middle line:84% Of every person made eligible, only 20% to 25% get covered. 00:42:13.340 --> 00:42:16.860 align:middle line:84% But that ignores the fact that there's two channels, 00:42:16.860 --> 00:42:20.460 align:middle line:84% because it turns out most of the people made eligible 00:42:20.460 --> 00:42:23.260 align:middle line:84% by the expansion of the '80s and '90s already had health 00:42:23.260 --> 00:42:26.020 align:middle line:90% insurance. 00:42:26.020 --> 00:42:28.900 align:middle line:84% So what you really want to know is the two separate channels. 00:42:28.900 --> 00:42:30.733 align:middle line:84% How much is covered [INAUDIBLE] among people 00:42:30.733 --> 00:42:34.020 align:middle line:84% who were previously uninsured, and how much is crowd-out? 00:42:34.020 --> 00:42:36.680 align:middle line:84% And what you find is that among people previously uninsured, 00:42:36.680 --> 00:42:39.140 align:middle line:84% the take-up is probably more like 2/3, 00:42:39.140 --> 00:42:41.860 align:middle line:84% which is typical for a social insurance program. 00:42:41.860 --> 00:42:44.540 align:middle line:84% 2/3 is about how often people seem to take up these 00:42:44.540 --> 00:42:45.240 align:middle line:90% entitlements. 00:42:45.240 --> 00:42:46.300 align:middle line:90% By the way, it's crazy. 00:42:46.300 --> 00:42:49.037 align:middle line:84% This is a free entitlement that you're 00:42:49.037 --> 00:42:51.120 align:middle line:84% getting, worth thousands and thousands of dollars, 00:42:51.120 --> 00:42:52.780 align:middle line:84% and a third of the people turn it down. 00:42:52.780 --> 00:42:54.420 align:middle line:84% We've talked about reasons why that be. 00:42:54.420 --> 00:42:57.760 align:middle line:84% That might be information failures, hassle, cost, stigma, 00:42:57.760 --> 00:42:58.580 align:middle line:90% et cetera. 00:42:58.580 --> 00:43:01.280 align:middle line:84% But nonetheless, it looks like it's about 2/3 take-up. 00:43:01.280 --> 00:43:02.100 align:middle line:90% Yeah. 00:43:02.100 --> 00:43:04.720 align:middle line:84% AUDIENCE: Is there a reason why it can't just 00:43:04.720 --> 00:43:06.780 align:middle line:84% be given automatically to those [INAUDIBLE]? 00:43:06.780 --> 00:43:07.760 align:middle line:84% JON GRUBER: Well, the question is, how would you 00:43:07.760 --> 00:43:09.120 align:middle line:90% give it automatically? 00:43:09.120 --> 00:43:10.340 align:middle line:90% It is given automatically-- 00:43:10.340 --> 00:43:12.560 align:middle line:84% I mean, literally, what you could do-- 00:43:12.560 --> 00:43:15.680 align:middle line:84% I mean, you could send people letters. 00:43:15.680 --> 00:43:17.048 align:middle line:90% But the problem is-- 00:43:17.048 --> 00:43:18.840 align:middle line:84% the reason it can't be given automatically, 00:43:18.840 --> 00:43:22.200 align:middle line:84% necessarily, easily, is because it's income-targeted. 00:43:22.200 --> 00:43:24.417 align:middle line:84% So you have to first confirm my income. 00:43:24.417 --> 00:43:26.500 align:middle line:84% Now, you can go back and confirm it with tax data, 00:43:26.500 --> 00:43:28.020 align:middle line:84% but that's old and takes a while to get. 00:43:28.020 --> 00:43:29.437 align:middle line:84% And the federal government doesn't 00:43:29.437 --> 00:43:31.240 align:middle line:84% like sharing tax data across branches. 00:43:31.240 --> 00:43:35.080 align:middle line:84% So this leads to a major issue we're having now in Medicaid, 00:43:35.080 --> 00:43:38.837 align:middle line:84% which is that during the crisis, during COVID, 00:43:38.837 --> 00:43:40.920 align:middle line:84% we essentially made everybody who applied eligible 00:43:40.920 --> 00:43:41.500 align:middle line:90% for Medicaid. 00:43:41.500 --> 00:43:43.643 align:middle line:84% We didn't check your income, basically. 00:43:43.643 --> 00:43:46.060 align:middle line:84% As long as you self-attested you were poor, we let you on. 00:43:46.060 --> 00:43:48.120 align:middle line:90% We didn't check if it was true. 00:43:48.120 --> 00:43:51.160 align:middle line:84% So the problem is now, we're checking, 00:43:51.160 --> 00:43:54.200 align:middle line:84% and it turns out a huge, huge number of kids on Medicaid 00:43:54.200 --> 00:43:56.343 align:middle line:84% weren't actually eligible or aren't eligible now. 00:43:56.343 --> 00:43:58.760 align:middle line:84% And there's millions of kids losing their health insurance 00:43:58.760 --> 00:44:01.100 align:middle line:84% as a result, which is an issue we're dealing with now. 00:44:01.100 --> 00:44:03.230 align:middle line:84% AUDIENCE: [INAUDIBLE] Is that fraud? 00:44:03.230 --> 00:44:05.250 align:middle line:84% JON GRUBER: Well, no, it's a great question. 00:44:05.250 --> 00:44:07.090 align:middle line:84% It's fraud if they lied about their income 00:44:07.090 --> 00:44:07.670 align:middle line:90% when they signed up. 00:44:07.670 --> 00:44:09.450 align:middle line:84% What about if their income has changed? 00:44:09.450 --> 00:44:10.770 align:middle line:90% Then it's not fraud. 00:44:10.770 --> 00:44:12.490 align:middle line:90% They ask about your income now. 00:44:12.490 --> 00:44:15.210 align:middle line:84% Now, technically-- so the way Medicaid works is they 00:44:15.210 --> 00:44:17.255 align:middle line:84% do a reevaluation every-- periodically, 00:44:17.255 --> 00:44:19.130 align:middle line:84% every six months or every year, where they're 00:44:19.130 --> 00:44:20.530 align:middle line:90% re-ask about your income. 00:44:20.530 --> 00:44:22.270 align:middle line:84% So the idea is when you're made eligible, 00:44:22.270 --> 00:44:23.770 align:middle line:84% you are guaranteed that eligibility, 00:44:23.770 --> 00:44:26.770 align:middle line:84% even if your income goes up, until the next recertification. 00:44:26.770 --> 00:44:28.630 align:middle line:84% At the next recertification, they ask you. 00:44:28.630 --> 00:44:30.170 align:middle line:90% The problem is the following. 00:44:30.170 --> 00:44:31.830 align:middle line:90% They do these recertifications. 00:44:31.830 --> 00:44:33.390 align:middle line:90% Tons of kids lose Medicaid. 00:44:33.390 --> 00:44:38.170 align:middle line:84% It turns out many of them-- the best estimate is about 2/3 , 00:44:38.170 --> 00:44:40.670 align:middle line:84% or 1/2 to 2/3-- are actually still eligible. 00:44:40.670 --> 00:44:42.610 align:middle line:84% They just missed the letter, or they 00:44:42.610 --> 00:44:44.770 align:middle line:84% couldn't read the letter, or something like that. 00:44:44.770 --> 00:44:46.770 align:middle line:84% So this fascinating trade-off that no one's ever 00:44:46.770 --> 00:44:49.870 align:middle line:84% really figured out how to deal with, which is, on the one hand, 00:44:49.870 --> 00:44:51.537 align:middle line:84% I don't want to give rich kids Medicaid. 00:44:51.537 --> 00:44:53.578 align:middle line:84% On the other hand, when I try to kick people off, 00:44:53.578 --> 00:44:56.050 align:middle line:84% I kick a lot of people off who shouldn't be kicked off. 00:44:56.050 --> 00:44:57.390 align:middle line:84% And how do you resolve that trade-off? 00:44:57.390 --> 00:44:58.410 align:middle line:84% That's a really interesting problem 00:44:58.410 --> 00:44:59.410 align:middle line:84% we're dealing with right now, that many 00:44:59.410 --> 00:45:01.160 align:middle line:84% of the kids who are losing eligibility now 00:45:01.160 --> 00:45:02.135 align:middle line:90% may still be eligible. 00:45:02.135 --> 00:45:03.760 align:middle line:84% The parents just didn't get the letter, 00:45:03.760 --> 00:45:06.740 align:middle line:84% or they just can't read it, or they're not paying attention. 00:45:06.740 --> 00:45:07.780 align:middle line:90% Yeah. 00:45:07.780 --> 00:45:10.960 align:middle line:84% AUDIENCE: So just at the time of service, 00:45:10.960 --> 00:45:13.980 align:middle line:84% can you not just have the equivalent of that letter thing 00:45:13.980 --> 00:45:15.260 align:middle line:90% happening? 00:45:15.260 --> 00:45:17.218 align:middle line:84% JON GRUBER: No the problem is, how's the doctor 00:45:17.218 --> 00:45:19.860 align:middle line:90% going to know your income? 00:45:19.860 --> 00:45:22.060 align:middle line:84% I mean, the problem is-- once again, universal 00:45:22.060 --> 00:45:24.740 align:middle line:84% would solve this, but without universal, it's hard to solve. 00:45:24.740 --> 00:45:26.175 align:middle line:84% AUDIENCE: Is it fair to say that-- 00:45:26.175 --> 00:45:27.800 align:middle line:84% I don't know-- if society is worse off, 00:45:27.800 --> 00:45:30.100 align:middle line:84% if people have health insurance for, let's say, during COVID 00:45:30.100 --> 00:45:32.350 align:middle line:84% and then they get kicked off versus them not having it 00:45:32.350 --> 00:45:32.940 align:middle line:90% at all? 00:45:32.940 --> 00:45:35.320 align:middle line:90% JON GRUBER: I don't think so. 00:45:35.320 --> 00:45:39.180 align:middle line:90% 00:45:39.180 --> 00:45:41.380 align:middle line:84% Certainly, you could talk about a model where-- 00:45:41.380 --> 00:45:44.053 align:middle line:90% 00:45:44.053 --> 00:45:46.220 align:middle line:84% you'd have to have a situation where-- yeah, I mean, 00:45:46.220 --> 00:45:47.720 align:middle line:84% you could imagine you got health insurance. 00:45:47.720 --> 00:45:49.300 align:middle line:84% They found out they had cancer, and they kicked them off 00:45:49.300 --> 00:45:51.260 align:middle line:84% before they could deal with it, and they're sadder 00:45:51.260 --> 00:45:52.360 align:middle line:90% knowing they have cancer. 00:45:52.360 --> 00:45:53.840 align:middle line:84% Either way, their cancer would have been dealt with, 00:45:53.840 --> 00:45:55.200 align:middle line:84% but they're sadder knowing they have cancer. 00:45:55.200 --> 00:45:56.100 align:middle line:84% So it'd have to be something which 00:45:56.100 --> 00:45:58.660 align:middle line:84% involved people's well-being of knowing whether they're sick 00:45:58.660 --> 00:45:59.160 align:middle line:90% or not. 00:45:59.160 --> 00:46:02.590 align:middle line:84% That'd be the story you'd have to tell. 00:46:02.590 --> 00:46:03.730 align:middle line:90% What about crowd-out? 00:46:03.730 --> 00:46:04.773 align:middle line:90% Well, crowd-out is real. 00:46:04.773 --> 00:46:07.190 align:middle line:84% By the way, it's ambiguous whether crowd-out would be real 00:46:07.190 --> 00:46:08.490 align:middle line:84% or not, because remember, on the one hand, 00:46:08.490 --> 00:46:10.950 align:middle line:84% I described why I might want to drop my MIT insurance. 00:46:10.950 --> 00:46:14.693 align:middle line:84% On the other hand, if I drop my MIT insurance and my income 00:46:14.693 --> 00:46:16.610 align:middle line:84% goes back up and I can't get Medicaid anymore, 00:46:16.610 --> 00:46:18.723 align:middle line:84% maybe I can't get my insurance back. 00:46:18.723 --> 00:46:20.140 align:middle line:84% And maybe I like my MIT insurance. 00:46:20.140 --> 00:46:22.140 align:middle line:84% I don't want to be in the government, et cetera. 00:46:22.140 --> 00:46:25.330 align:middle line:84% So it's ambiguous how big crowd-out out would be. 00:46:25.330 --> 00:46:27.670 align:middle line:84% The estimates here are varied, but the best estimates 00:46:27.670 --> 00:46:30.810 align:middle line:84% are crowd-out and even how you define it. 00:46:30.810 --> 00:46:37.870 align:middle line:84% The best estimates are that of the people who get covered, 00:46:37.870 --> 00:46:40.230 align:middle line:84% about half we're covering-- about half 00:46:40.230 --> 00:46:42.350 align:middle line:84% came from previously uninsured, and half 00:46:42.350 --> 00:46:43.970 align:middle line:90% came from previously insured. 00:46:43.970 --> 00:46:45.930 align:middle line:84% We call that about 50% crowd-out, 00:46:45.930 --> 00:46:47.430 align:middle line:84% that of every person who's gaining-- 00:46:47.430 --> 00:46:50.370 align:middle line:84% of every 1,000 people gaining coverage through Medicaid, 00:46:50.370 --> 00:46:53.270 align:middle line:84% about 500 used to have health insurance, 00:46:53.270 --> 00:46:55.790 align:middle line:84% and about 500 didn't have health insurance. 00:46:55.790 --> 00:46:57.070 align:middle line:90% Now, why do we care? 00:46:57.070 --> 00:46:58.720 align:middle line:84% Well, we care because the people who 00:46:58.720 --> 00:46:59.840 align:middle line:84% used to have health insurance now 00:46:59.840 --> 00:47:00.940 align:middle line:90% are-- it's on the public bill. 00:47:00.940 --> 00:47:02.360 align:middle line:84% It used to be on the private bill. 00:47:02.360 --> 00:47:04.220 align:middle line:84% Distribution-- we might be OK with that, 00:47:04.220 --> 00:47:06.700 align:middle line:84% but efficiency-wise we're not really changing anything. 00:47:06.700 --> 00:47:07.960 align:middle line:84% We're just shifting costs from the private 00:47:07.960 --> 00:47:08.835 align:middle line:90% to the public sector. 00:47:08.835 --> 00:47:11.120 align:middle line:90% Classic crowd-out issue. 00:47:11.120 --> 00:47:14.080 align:middle line:84% So that's the first node of this chart. 00:47:14.080 --> 00:47:17.600 align:middle line:84% The second node is, well, given that those people already 00:47:17.600 --> 00:47:19.288 align:middle line:84% had health insurance anyway and given 00:47:19.288 --> 00:47:21.080 align:middle line:84% it's hard to find a doctor, do you actually 00:47:21.080 --> 00:47:22.840 align:middle line:90% get more health care? 00:47:22.840 --> 00:47:25.360 align:middle line:84% And the answer is unambiguously, yes. 00:47:25.360 --> 00:47:28.400 align:middle line:84% Study after study shows large, large increase 00:47:28.400 --> 00:47:33.920 align:middle line:84% in Medicaid eligibility-- in using health care when 00:47:33.920 --> 00:47:35.200 align:middle line:90% you get Medicaid. 00:47:35.200 --> 00:47:37.640 align:middle line:84% Despite the barriers, despite the crowd-out, 00:47:37.640 --> 00:47:42.020 align:middle line:84% huge increase in health utilization. 00:47:42.020 --> 00:47:46.360 align:middle line:84% Then the question is, well, does it improve health? 00:47:46.360 --> 00:47:48.345 align:middle line:90% Now, here, it's a fascinating-- 00:47:48.345 --> 00:47:49.720 align:middle line:84% I'm going to diverge for a minute 00:47:49.720 --> 00:47:52.090 align:middle line:84% to a fascinating lecture on methods, if you will, 00:47:52.090 --> 00:47:54.340 align:middle line:84% in economics because it's interesting, at least to me, 00:47:54.340 --> 00:47:55.898 align:middle line:90% and you got to listen. 00:47:55.898 --> 00:47:57.190 align:middle line:90% Well, you don't have to listen. 00:47:57.190 --> 00:47:58.150 align:middle line:90% I'd like you to listen. 00:47:58.150 --> 00:48:05.330 align:middle line:84% Basically, there was a literature up until about 2010, 00:48:05.330 --> 00:48:06.890 align:middle line:90% about 2008. 00:48:06.890 --> 00:48:10.730 align:middle line:84% And that literature was fairly consistent, 00:48:10.730 --> 00:48:14.370 align:middle line:84% that giving people Medicaid improved their health, 00:48:14.370 --> 00:48:19.390 align:middle line:84% that they use more health care, and health outcomes were better. 00:48:19.390 --> 00:48:22.050 align:middle line:84% And there were all these cool quasi natural experiment 00:48:22.050 --> 00:48:24.170 align:middle line:84% studies, quasi experimental studies, showing-- 00:48:24.170 --> 00:48:26.910 align:middle line:84% some by myself, showing that, for example, 00:48:26.910 --> 00:48:30.290 align:middle line:84% when mothers got Medicaid, they were much-- the babies were much 00:48:30.290 --> 00:48:32.308 align:middle line:84% less likely to die, when kids got Medicaid, 00:48:32.308 --> 00:48:34.850 align:middle line:84% they were much less likely to be sick and go to the hospital. 00:48:34.850 --> 00:48:37.570 align:middle line:84% Lots of evidence that Medicaid was improving health, 00:48:37.570 --> 00:48:40.798 align:middle line:84% and actually, that it was doing so cost-effectively, 00:48:40.798 --> 00:48:43.090 align:middle line:84% that if you looked at the cost of increasing what we're 00:48:43.090 --> 00:48:46.217 align:middle line:84% spending on Medicaid versus what we're getting for health, 00:48:46.217 --> 00:48:48.050 align:middle line:84% we're doing things like saving a baby's life 00:48:48.050 --> 00:48:51.450 align:middle line:84% for a million bucks, which, if you think back to chapter 8, 00:48:51.450 --> 00:48:51.950 align:middle line:90% is not bad. 00:48:51.950 --> 00:48:54.050 align:middle line:84% We think lives are worth $10.5 million. 00:48:54.050 --> 00:48:56.690 align:middle line:84% Saving a baby's life for a million bucks is not bad. 00:48:56.690 --> 00:48:59.190 align:middle line:84% So we were doing things-- we were improving health and doing 00:48:59.190 --> 00:49:01.470 align:middle line:90% so fairly cost-effectively. 00:49:01.470 --> 00:49:03.030 align:middle line:90% That was round 1. 00:49:03.030 --> 00:49:05.830 align:middle line:84% Round 2 was this very famous Oregon Health Insurance 00:49:05.830 --> 00:49:06.850 align:middle line:90% Experiment. 00:49:06.850 --> 00:49:08.670 align:middle line:84% The Oregon Health Insurance Experiment 00:49:08.670 --> 00:49:12.070 align:middle line:84% was a study where the state of Oregon 00:49:12.070 --> 00:49:14.767 align:middle line:84% decided they wanted to expand Medicaid-- this is pre-ACA. 00:49:14.767 --> 00:49:16.850 align:middle line:84% Decided they wanted to expand Medicaid for adults. 00:49:16.850 --> 00:49:19.142 align:middle line:84% They wanted to add 10,000 more adults to their Medicaid 00:49:19.142 --> 00:49:20.110 align:middle line:90% program. 00:49:20.110 --> 00:49:22.110 align:middle line:84% So they took applications from uninsured people, 00:49:22.110 --> 00:49:24.590 align:middle line:90% and 100,000 people applied. 00:49:24.590 --> 00:49:26.270 align:middle line:84% So they decided a fair way to allocate 00:49:26.270 --> 00:49:29.105 align:middle line:84% it would be to have a lottery, so they ran an experiment. 00:49:29.105 --> 00:49:30.980 align:middle line:84% They ran a lottery, and 10,000 people got it, 00:49:30.980 --> 00:49:32.550 align:middle line:90% and 90,000 people didn't. 00:49:32.550 --> 00:49:34.430 align:middle line:84% And a team that I was part of but which 00:49:34.430 --> 00:49:37.090 align:middle line:84% was led by my colleague Amy Finkelstein and Kate Baker, 00:49:37.090 --> 00:49:40.750 align:middle line:84% formerly of Harvard, now the provost at Chicago-- 00:49:40.750 --> 00:49:45.030 align:middle line:84% they looked at what effect this had 00:49:45.030 --> 00:49:47.430 align:middle line:90% using-- it was an experiment. 00:49:47.430 --> 00:49:50.470 align:middle line:84% And the findings were people used a lot more health care. 00:49:50.470 --> 00:49:52.190 align:middle line:90% Like I said, that's clear. 00:49:52.190 --> 00:49:56.080 align:middle line:84% But their health wasn't measurably much better. 00:49:56.080 --> 00:50:01.880 align:middle line:84% Now, one exception-- their measures of physical health 00:50:01.880 --> 00:50:02.460 align:middle line:90% were better. 00:50:02.460 --> 00:50:04.760 align:middle line:84% Their blood pressure didn't change, things like that. 00:50:04.760 --> 00:50:07.680 align:middle line:90% One exception was mental health. 00:50:07.680 --> 00:50:12.440 align:middle line:84% Depression went down by 30%, almost instantly. 00:50:12.440 --> 00:50:16.640 align:middle line:84% So that's weird because even Prozac doesn't work that 00:50:16.640 --> 00:50:20.490 align:middle line:84% quickly, and it wasn't like they were putting it in the water. 00:50:20.490 --> 00:50:22.240 align:middle line:84% What happened was-- it comes back to where 00:50:22.240 --> 00:50:23.920 align:middle line:90% discussing a few minutes ago. 00:50:23.920 --> 00:50:26.240 align:middle line:84% Our theory-- the co-authors' theory was it 00:50:26.240 --> 00:50:28.960 align:middle line:84% was just the lack of stress of being uninsured, 00:50:28.960 --> 00:50:33.040 align:middle line:84% that facing constant financial risk is depressing, 00:50:33.040 --> 00:50:35.898 align:middle line:84% and that basically, by relaxing that risk on people, 00:50:35.898 --> 00:50:37.440 align:middle line:84% we improve their mental health, which 00:50:37.440 --> 00:50:40.400 align:middle line:84% comes to these other questions about, what is insurance for? 00:50:40.400 --> 00:50:43.267 align:middle line:84% Contradicts, a bit, my earlier perspective-- it doesn't really. 00:50:43.267 --> 00:50:44.600 align:middle line:90% Actually, no, it's a contradict. 00:50:44.600 --> 00:50:45.642 align:middle line:90% I would argue it doesn't. 00:50:45.642 --> 00:50:48.110 align:middle line:84% What it's saying is, that's a measure of the benefit 00:50:48.110 --> 00:50:50.360 align:middle line:84% of insuring-- the idea of insurance to protect against 00:50:50.360 --> 00:50:51.473 align:middle line:90% financial risk-- 00:50:51.473 --> 00:50:53.890 align:middle line:84% a measure of the value of that is how much better people's 00:50:53.890 --> 00:50:54.790 align:middle line:90% mental health was. 00:50:54.790 --> 00:50:55.290 align:middle line:90% Yeah. 00:50:55.290 --> 00:50:58.290 align:middle line:90% AUDIENCE: [INAUDIBLE] 00:50:58.290 --> 00:50:59.830 align:middle line:90% JON GRUBER: Adults. 00:50:59.830 --> 00:51:00.950 align:middle line:90% It was only expansion. 00:51:00.950 --> 00:51:04.610 align:middle line:84% For adults, it was only expansion for adults. 00:51:04.610 --> 00:51:07.610 align:middle line:84% Side note on that, side note of a side note, 00:51:07.610 --> 00:51:10.730 align:middle line:84% talking about the way health care's viewed in America. 00:51:10.730 --> 00:51:13.283 align:middle line:84% This study was very famous, got coverage in The New York 00:51:13.283 --> 00:51:13.950 align:middle line:90% Times and stuff. 00:51:13.950 --> 00:51:15.370 align:middle line:84% And they wrote the articles, and they said, 00:51:15.370 --> 00:51:17.350 align:middle line:84% study finds health insurance doesn't improve health. 00:51:17.350 --> 00:51:19.558 align:middle line:84% In the last sentence, they said, oh, yeah, depression 00:51:19.558 --> 00:51:22.690 align:middle line:84% went down 30%, which talks about the weak role mental health 00:51:22.690 --> 00:51:24.750 align:middle line:84% has in our society, relative to physical health. 00:51:24.750 --> 00:51:26.550 align:middle line:84% And I think that's changing with your generation. 00:51:26.550 --> 00:51:28.425 align:middle line:84% There's a lot more attention to mental health 00:51:28.425 --> 00:51:30.930 align:middle line:84% than there was even 15 years ago when we did this study. 00:51:30.930 --> 00:51:33.047 align:middle line:84% But an interesting side note about the value 00:51:33.047 --> 00:51:35.630 align:middle line:84% of-- people's view of the value of mental and physical health. 00:51:35.630 --> 00:51:37.795 align:middle line:84% Anyway, this was a real shot across the bow 00:51:37.795 --> 00:51:39.170 align:middle line:84% because people said, wow, we have 00:51:39.170 --> 00:51:41.462 align:middle line:84% this whole literature saying, Medicaid improved health. 00:51:41.462 --> 00:51:44.290 align:middle line:90% What if it doesn't? 00:51:44.290 --> 00:51:47.530 align:middle line:84% Then we ran another experiment called the Affordable Care 00:51:47.530 --> 00:51:50.050 align:middle line:84% Act, where we had some people getting 00:51:50.050 --> 00:51:53.500 align:middle line:84% health insurance because their states expanded, others not. 00:51:53.500 --> 00:51:54.660 align:middle line:90% OK. 00:51:54.660 --> 00:51:58.260 align:middle line:84% And what that study-- what they did in those studies is they 00:51:58.260 --> 00:52:00.080 align:middle line:84% took massive samples of individuals 00:52:00.080 --> 00:52:02.960 align:middle line:84% so they could find really precisely what effect that had. 00:52:02.960 --> 00:52:05.900 align:middle line:84% And they found huge positive effects on health, 00:52:05.900 --> 00:52:09.012 align:middle line:84% much like the previous literature, 00:52:09.012 --> 00:52:10.220 align:middle line:90% from very convincing methods. 00:52:10.220 --> 00:52:12.040 align:middle line:84% Indeed, there was another randomized trial, 00:52:12.040 --> 00:52:13.615 align:middle line:90% which is really cool. 00:52:13.615 --> 00:52:16.240 align:middle line:84% Under the Affordable Care Act, which I'll talk about next time, 00:52:16.240 --> 00:52:17.900 align:middle line:90% there was an individual mandate. 00:52:17.900 --> 00:52:20.400 align:middle line:84% That individual mandate said you had to report on your taxes 00:52:20.400 --> 00:52:24.460 align:middle line:84% whether you were insured, and if you weren't, you paid a penalty. 00:52:24.460 --> 00:52:27.220 align:middle line:84% Well, the IRS, after the first year-- 00:52:27.220 --> 00:52:29.740 align:middle line:90% the IRS is our tax system. 00:52:29.740 --> 00:52:32.260 align:middle line:84% After the first year of this program, 00:52:32.260 --> 00:52:35.380 align:middle line:84% decided people who pay the mandate penalty should remind 00:52:35.380 --> 00:52:37.380 align:middle line:84% them, hey, get health insurance, and you 00:52:37.380 --> 00:52:38.705 align:middle line:90% won't have to pay the penalty. 00:52:38.705 --> 00:52:40.580 align:middle line:84% Someone convinced them, some brilliant person 00:52:40.580 --> 00:52:43.620 align:middle line:84% convinced them to randomly remind people. 00:52:43.620 --> 00:52:46.740 align:middle line:84% So they randomly randomized, and some people got a reminder 00:52:46.740 --> 00:52:47.940 align:middle line:90% letter, and some didn't. 00:52:47.940 --> 00:52:51.070 align:middle line:84% I don't know how they got away with that. 00:52:51.070 --> 00:52:52.830 align:middle line:84% What they found was the people who 00:52:52.830 --> 00:52:54.630 align:middle line:84% got the reminder letter were much more 00:52:54.630 --> 00:52:56.950 align:middle line:84% likely to sign up for health insurance, 00:52:56.950 --> 00:52:59.710 align:middle line:90% and their health improved. 00:52:59.710 --> 00:53:02.550 align:middle line:84% So here, we have a situation where 00:53:02.550 --> 00:53:05.730 align:middle line:84% we have this contradiction in the literature. 00:53:05.730 --> 00:53:07.293 align:middle line:84% I think the answer is pretty clear 00:53:07.293 --> 00:53:09.710 align:middle line:84% that the problem of the Oregon Health Insurance Experiment 00:53:09.710 --> 00:53:11.543 align:middle line:84% is it only followed people for about a year, 00:53:11.543 --> 00:53:12.987 align:middle line:90% and the samples were too small. 00:53:12.987 --> 00:53:15.070 align:middle line:84% Now, as a co-author, I'm not sure the lead authors 00:53:15.070 --> 00:53:15.490 align:middle line:90% would agree. 00:53:15.490 --> 00:53:17.032 align:middle line:84% I'm a junior co-author on that paper. 00:53:17.032 --> 00:53:19.670 align:middle line:84% I'm not sure they'd agree with my assessment of that. 00:53:19.670 --> 00:53:23.270 align:middle line:84% But that would be my assessment of this debate, 00:53:23.270 --> 00:53:26.310 align:middle line:84% that, basically, we've pretty much sealed the deal, 00:53:26.310 --> 00:53:30.430 align:middle line:84% that Medicaid coverage is good for your health 00:53:30.430 --> 00:53:33.030 align:middle line:84% and that it probably improves your health cost-effectively. 00:53:33.030 --> 00:53:36.430 align:middle line:84% But it is an interesting study in how 00:53:36.430 --> 00:53:39.790 align:middle line:90% scientific consensus evolves. 00:53:39.790 --> 00:53:41.230 align:middle line:90% Questions about that. 00:53:41.230 --> 00:53:41.802 align:middle line:90% Yeah. 00:53:41.802 --> 00:53:43.510 align:middle line:84% AUDIENCE: I remember, when I was younger, 00:53:43.510 --> 00:53:45.093 align:middle line:84% hearing about the Affordable Care Act, 00:53:45.093 --> 00:53:46.550 align:middle line:84% and some of the politics around it 00:53:46.550 --> 00:53:49.250 align:middle line:84% were such that what people hated the [INAUDIBLE]. 00:53:49.250 --> 00:53:51.310 align:middle line:84% JON GRUBER: Oh, come to the next lecture. 00:53:51.310 --> 00:53:52.310 align:middle line:90% We'll talk all about it. 00:53:52.310 --> 00:53:52.810 align:middle line:90% Believe me. 00:53:52.810 --> 00:53:53.990 align:middle line:90% I'll show you my scars. 00:53:53.990 --> 00:53:56.910 align:middle line:90% [LAUGHTER] 00:53:56.910 --> 00:53:59.220 align:middle line:84% That's what I want to say about Medicaid. 00:53:59.220 --> 00:54:01.470 align:middle line:84% Now I want to turn to the other big program, Medicare, 00:54:01.470 --> 00:54:02.690 align:middle line:90% which is insurance for the? 00:54:02.690 --> 00:54:03.677 align:middle line:90% AUDIENCE: Elderly. 00:54:03.677 --> 00:54:04.510 align:middle line:90% JON GRUBER: Elderly. 00:54:04.510 --> 00:54:05.210 align:middle line:90% All right, good. 00:54:05.210 --> 00:54:08.210 align:middle line:84% Remember, any time I can sneak up on you and ask you 00:54:08.210 --> 00:54:11.330 align:middle line:84% this question, I have the right, any time, to ask this question. 00:54:11.330 --> 00:54:14.170 align:middle line:90% OK, so how does Medicare work? 00:54:14.170 --> 00:54:17.290 align:middle line:84% Well, Medicare, unlike Medicaid, is a federal program, not 00:54:17.290 --> 00:54:18.890 align:middle line:90% a state program. 00:54:18.890 --> 00:54:21.770 align:middle line:90% It's a federal program. 00:54:21.770 --> 00:54:23.730 align:middle line:84% As I said, it's a program where if you've 00:54:23.730 --> 00:54:27.010 align:middle line:84% worked for at least 10 years paying in your Medicare taxes, 00:54:27.010 --> 00:54:29.930 align:middle line:84% you then, when you turn 65, become eligible 00:54:29.930 --> 00:54:31.450 align:middle line:90% for the Medicare program. 00:54:31.450 --> 00:54:34.870 align:middle line:84% Or if you become disabled, after two years, 00:54:34.870 --> 00:54:36.690 align:middle line:84% you also become eligible for Medicare. 00:54:36.690 --> 00:54:39.348 align:middle line:84% So it's insurance for the elderly and the disabled. 00:54:39.348 --> 00:54:40.390 align:middle line:90% Elderly's a fine summary. 00:54:40.390 --> 00:54:42.010 align:middle line:84% I won't hit you with a disabled part. 00:54:42.010 --> 00:54:43.730 align:middle line:84% But for the elderly and disabled, 00:54:43.730 --> 00:54:46.890 align:middle line:90% people on disability insurance. 00:54:46.890 --> 00:54:51.420 align:middle line:90% Now, how does Medicare look? 00:54:51.420 --> 00:54:55.220 align:middle line:84% Traditional Medicare looks very much like old-school 00:54:55.220 --> 00:54:57.140 align:middle line:90% fee-for-service-medicine. 00:54:57.140 --> 00:54:58.720 align:middle line:84% Basically, you get the insurance. 00:54:58.720 --> 00:54:59.680 align:middle line:90% The government's your insurer. 00:54:59.680 --> 00:55:01.722 align:middle line:84% You go to any doctor you want, and the government 00:55:01.722 --> 00:55:02.740 align:middle line:90% pays the Bills. 00:55:02.740 --> 00:55:06.100 align:middle line:84% And until the 1980s, that's basically how it worked. 00:55:06.100 --> 00:55:08.580 align:middle line:84% And we'll come back to what changed in a few minutes. 00:55:08.580 --> 00:55:12.700 align:middle line:84% Now, Medicare, however, is really three different programs. 00:55:12.700 --> 00:55:15.078 align:middle line:84% There's part A, part B, and Part D. 00:55:15.078 --> 00:55:17.120 align:middle line:84% I'll come back later to part C. Why do we skip C? 00:55:17.120 --> 00:55:18.340 align:middle line:90% I'll come back to that. 00:55:18.340 --> 00:55:21.980 align:middle line:90% Part A is hospital care. 00:55:21.980 --> 00:55:24.900 align:middle line:84% That is what we pay for with our payroll tax. 00:55:24.900 --> 00:55:28.060 align:middle line:84% Our payroll tax goes into Medicare Part A trust fund. 00:55:28.060 --> 00:55:31.020 align:middle line:84% That pays for your hospital care. 00:55:31.020 --> 00:55:33.180 align:middle line:90% Part B is Doctor care. 00:55:33.180 --> 00:55:36.900 align:middle line:84% Doctor care you actually have to sign up for. 00:55:36.900 --> 00:55:38.802 align:middle line:90% And, like, 97% of elders do. 00:55:38.802 --> 00:55:41.260 align:middle line:84% But basically, it's a program where the government provides 00:55:41.260 --> 00:55:42.600 align:middle line:90% and pays for your doctor care. 00:55:42.600 --> 00:55:44.680 align:middle line:84% That one, you don't pay any payroll tax, 00:55:44.680 --> 00:55:48.670 align:middle line:84% but you pay a premium that is 25% of the program costs. 00:55:48.670 --> 00:55:51.470 align:middle line:84% So that's more like the kind of insurance you have an employer. 00:55:51.470 --> 00:55:55.230 align:middle line:84% You pay a premium that's 25% of program costs. 00:55:55.230 --> 00:56:00.430 align:middle line:84% And that's part B. Part D is coverage for drugs, 00:56:00.430 --> 00:56:04.270 align:middle line:84% and I'll talk about that separately in a little bit. 00:56:04.270 --> 00:56:07.870 align:middle line:84% Now, what this means is while Medicaid 00:56:07.870 --> 00:56:09.850 align:middle line:84% may be the best insurance money can't buy, 00:56:09.850 --> 00:56:12.095 align:middle line:84% Medicare's actually not that generous. 00:56:12.095 --> 00:56:13.720 align:middle line:84% 20 years ago, it was much less generous 00:56:13.720 --> 00:56:15.210 align:middle line:90% than the typical employer plan. 00:56:15.210 --> 00:56:17.730 align:middle line:84% Now it's probably equivalent to a typical employer plan. 00:56:17.730 --> 00:56:20.190 align:middle line:90% This is shown in table 16-2. 00:56:20.190 --> 00:56:22.910 align:middle line:84% This compares Medicare and Medicaid 00:56:22.910 --> 00:56:24.950 align:middle line:84% in terms of what you have to pay. 00:56:24.950 --> 00:56:26.590 align:middle line:90% First, I have the eligibles. 00:56:26.590 --> 00:56:28.870 align:middle line:90% Talk about premiums. 00:56:28.870 --> 00:56:31.070 align:middle line:84% Well, for Medicaid, there's none. 00:56:31.070 --> 00:56:33.717 align:middle line:84% For Medicare, you have to pay for physician coverage, 00:56:33.717 --> 00:56:36.050 align:middle line:84% and you also have to pay for prescription drug coverage. 00:56:36.050 --> 00:56:36.717 align:middle line:90% That's variable. 00:56:36.717 --> 00:56:38.430 align:middle line:90% We'll come back to that. 00:56:38.430 --> 00:56:41.790 align:middle line:84% Deductibles and co-payments-- nothing in Medicaid. 00:56:41.790 --> 00:56:46.000 align:middle line:84% In Medicare, you have to pay a $1,484 deductible. 00:56:46.000 --> 00:56:48.420 align:middle line:84% And then for hospitals-- for physicians, 00:56:48.420 --> 00:56:49.740 align:middle line:90% there's a $200 deductible. 00:56:49.740 --> 00:56:50.900 align:middle line:90% And this is important. 00:56:50.900 --> 00:56:54.120 align:middle line:90% A 20% uncapped coinsurance. 00:56:54.120 --> 00:56:56.580 align:middle line:84% This is worse than any other insurance program. 00:56:56.580 --> 00:56:59.300 align:middle line:84% Every other insurance program you sign up for has a maximum. 00:56:59.300 --> 00:57:00.920 align:middle line:84% They say, you have to pay out of pocket, but at some maximum, 00:57:00.920 --> 00:57:01.820 align:middle line:90% we'll stop you. 00:57:01.820 --> 00:57:02.900 align:middle line:90% Medicare does not. 00:57:02.900 --> 00:57:06.600 align:middle line:84% If you spend $100,000 on doctors, you owe $20,000. 00:57:06.600 --> 00:57:08.250 align:middle line:84% Uncapped-- and we'll come back to this, 00:57:08.250 --> 00:57:10.360 align:middle line:84% but it's a real weak feature of Medicare. 00:57:10.360 --> 00:57:12.400 align:middle line:90% Uncapped coinsurance system. 00:57:12.400 --> 00:57:16.080 align:middle line:84% And Medicaid doesn't exclude anything. 00:57:16.080 --> 00:57:19.040 align:middle line:84% Medicare used to exclude prescription drugs. 00:57:19.040 --> 00:57:20.740 align:middle line:84% It used to exclude routine checkups. 00:57:20.740 --> 00:57:22.168 align:middle line:84% At least it wouldn't make it free. 00:57:22.168 --> 00:57:24.460 align:middle line:84% It excludes dental care, nursing home care, eyeglasses, 00:57:24.460 --> 00:57:26.400 align:middle line:90% hearing aids, et cetera. 00:57:26.400 --> 00:57:29.840 align:middle line:84% And then, finally, reimbursement under Medicare 00:57:29.840 --> 00:57:33.060 align:middle line:84% is between private insurance and Medicaid. 00:57:33.060 --> 00:57:37.500 align:middle line:84% If Medicaid is, on average, less than half of private insurance, 00:57:37.500 --> 00:57:39.920 align:middle line:84% Medicare reimbursement's on the order of about 70% 00:57:39.920 --> 00:57:42.280 align:middle line:90% of private insurance. 00:57:42.280 --> 00:57:46.850 align:middle line:84% So that's basically how the program works. 00:57:46.850 --> 00:57:48.390 align:middle line:84% Let's talk about prescription drugs. 00:57:48.390 --> 00:57:50.970 align:middle line:84% This is a fascinating debate, real-time-- at least 00:57:50.970 --> 00:57:53.650 align:middle line:84% in my life, real-time debate When 00:57:53.650 --> 00:57:58.890 align:middle line:84% Medicare was set up in 1965, it did not include drugs. 00:57:58.890 --> 00:58:02.155 align:middle line:84% That's because out-of-pocket drugs wasn't really a thing. 00:58:02.155 --> 00:58:04.530 align:middle line:84% It's like if you got drugs, you got them in the hospital. 00:58:04.530 --> 00:58:06.697 align:middle line:84% You took two aspirin and called them in the morning. 00:58:06.697 --> 00:58:08.750 align:middle line:90% It wasn't a thing. 00:58:08.750 --> 00:58:14.370 align:middle line:84% But by the late 1990s, the typical elder was spending more 00:58:14.370 --> 00:58:18.362 align:middle line:84% on prescription drugs than the whole program cost in 1965. 00:58:18.362 --> 00:58:20.070 align:middle line:84% Prescription drugs had become a big deal, 00:58:20.070 --> 00:58:22.190 align:middle line:84% and there was an increasing awareness 00:58:22.190 --> 00:58:25.970 align:middle line:84% this is a hole in the nature of the Medicare program. 00:58:25.970 --> 00:58:31.758 align:middle line:84% So basically, in George W Bush's first term, 00:58:31.758 --> 00:58:34.050 align:middle line:84% there was consensus that something should be done about 00:58:34.050 --> 00:58:36.950 align:middle line:84% this, but there was a vigorous disagreement-- 00:58:36.950 --> 00:58:38.730 align:middle line:90% it goes back to chapter 7-- 00:58:38.730 --> 00:58:41.530 align:middle line:84% between the two parties and how it should be done. 00:58:41.530 --> 00:58:44.660 align:middle line:84% Democrats want to do this new drug program 00:58:44.660 --> 00:58:47.500 align:middle line:84% like we do the rest of Medicare, a single-payer fee-for-service 00:58:47.500 --> 00:58:48.820 align:middle line:90% system. 00:58:48.820 --> 00:58:53.060 align:middle line:84% Republicans said, no, we should have a privately provided 00:58:53.060 --> 00:58:54.398 align:middle line:90% but federally paid system. 00:58:54.398 --> 00:58:56.940 align:middle line:84% We should have private insurers provide the prescription drug 00:58:56.940 --> 00:58:58.455 align:middle line:84% benefit, competing with each other, 00:58:58.455 --> 00:59:00.580 align:middle line:84% and the federal government will pay in the back end 00:59:00.580 --> 00:59:01.948 align:middle line:90% to those insurers. 00:59:01.948 --> 00:59:03.740 align:middle line:84% So they said it'll be like other insurance, 00:59:03.740 --> 00:59:06.140 align:middle line:84% where the contractor's the federal government, 00:59:06.140 --> 00:59:08.320 align:middle line:84% but it'll be private insurers delivering it. 00:59:08.320 --> 00:59:10.220 align:middle line:84% And you see arguments on both sides. 00:59:10.220 --> 00:59:13.480 align:middle line:84% On the one side, the Republican argument makes a lot of sense. 00:59:13.480 --> 00:59:15.580 align:middle line:84% Competition works in many contexts. 00:59:15.580 --> 00:59:18.300 align:middle line:84% The idea is to have these competitive insurers, 00:59:18.300 --> 00:59:21.260 align:middle line:84% and they can innovate in things like that. 00:59:21.260 --> 00:59:26.780 align:middle line:84% On the other hand, what is innovation in insurance? 00:59:26.780 --> 00:59:29.587 align:middle line:84% Largely, it's innovation about how to keep out sick people. 00:59:29.587 --> 00:59:31.420 align:middle line:84% I mean, that's how insurers really innovate, 00:59:31.420 --> 00:59:33.860 align:middle line:84% is by trying to decide how to keep sick people out. 00:59:33.860 --> 00:59:37.020 align:middle line:84% So it's not quite clear why you want innovation. 00:59:37.020 --> 00:59:39.760 align:middle line:84% Now, I think the counterargument to that, which is valid-- 00:59:39.760 --> 00:59:42.760 align:middle line:84% which is that-- so the Republican counterargument is, 00:59:42.760 --> 00:59:45.920 align:middle line:84% actually, a lot of innovation is in negotiations with the drug 00:59:45.920 --> 00:59:49.120 align:middle line:84% companies, that, basically, is innovation in how you decide 00:59:49.120 --> 00:59:53.160 align:middle line:84% to get the prices down on those drugs, to which the Democrats 00:59:53.160 --> 00:59:55.160 align:middle line:84% responded, well, the best way to get prices down 00:59:55.160 --> 00:59:56.813 align:middle line:84% is have one giant payer who just screws 00:59:56.813 --> 00:59:58.480 align:middle line:84% the pharmaceutical companies, and that's 00:59:58.480 --> 00:59:59.800 align:middle line:90% what Medicare would do. 00:59:59.800 --> 01:00:04.640 align:middle line:84% Vigorous debate between these two things, and Republicans won. 01:00:04.640 --> 01:00:08.280 align:middle line:84% And we set up this program where basically people get 01:00:08.280 --> 01:00:10.040 align:middle line:84% their Medicare prescription drug benefit 01:00:10.040 --> 01:00:12.800 align:middle line:84% from a private insurer that's largely 01:00:12.800 --> 01:00:14.500 align:middle line:90% paid by the federal government. 01:00:14.500 --> 01:00:18.220 align:middle line:84% People pay a premium, which is a small fraction of the cost. 01:00:18.220 --> 01:00:20.680 align:middle line:84% The federal government pays the rest. 01:00:20.680 --> 01:00:23.440 align:middle line:90% So that's how it ended up. 01:00:23.440 --> 01:00:25.200 align:middle line:90% How did it do? 01:00:25.200 --> 01:00:28.733 align:middle line:84% Well, basically, it was much cheaper than we expected. 01:00:28.733 --> 01:00:31.400 align:middle line:84% The early evidence on part D was quite good in terms of relative 01:00:31.400 --> 01:00:33.020 align:middle line:84% to what we thought the program would cost. 01:00:33.020 --> 01:00:34.080 align:middle line:84% These private insurance companies 01:00:34.080 --> 01:00:35.420 align:middle line:90% really got the cost down. 01:00:35.420 --> 01:00:37.160 align:middle line:84% Did they get them down more than the government could have, 01:00:37.160 --> 01:00:38.060 align:middle line:90% negotiating? 01:00:38.060 --> 01:00:39.318 align:middle line:90% Not clear. 01:00:39.318 --> 01:00:40.860 align:middle line:84% And we'll come back to that next time 01:00:40.860 --> 01:00:43.130 align:middle line:84% when we talk about the Inflation Reduction Act. 01:00:43.130 --> 01:00:45.190 align:middle line:84% But certainly, it was less than we expected. 01:00:45.190 --> 01:00:48.210 align:middle line:84% They certainly did a good job of negotiating these prices down. 01:00:48.210 --> 01:00:51.410 align:middle line:84% At the same time, we had the problem 01:00:51.410 --> 01:00:54.850 align:middle line:84% that people were confused by this array of choice 01:00:54.850 --> 01:00:57.110 align:middle line:84% and did not do a very good job choosing their plans. 01:00:57.110 --> 01:00:59.450 align:middle line:84% That's something I've done a lot of research on. 01:00:59.450 --> 01:01:02.165 align:middle line:84% We've shown that the typical senior-- 01:01:02.165 --> 01:01:04.290 align:middle line:84% we looked at-- all seniors-- we looked at the plans 01:01:04.290 --> 01:01:06.270 align:middle line:84% they had available and what they chose, 01:01:06.270 --> 01:01:09.110 align:middle line:84% and we estimated what their cost would be under the alternative. 01:01:09.110 --> 01:01:10.970 align:middle line:84% And we found about 12% of seniors 01:01:10.970 --> 01:01:12.650 align:middle line:90% chose the lowest-cost plan. 01:01:12.650 --> 01:01:15.530 align:middle line:84% And the typical senior could have saved on the order of $600 01:01:15.530 --> 01:01:17.050 align:middle line:90% if they chose better. 01:01:17.050 --> 01:01:21.275 align:middle line:84% And basically, it's a very confusing choice. 01:01:21.275 --> 01:01:22.650 align:middle line:84% So the downside of the choice was 01:01:22.650 --> 01:01:25.190 align:middle line:84% seniors were not always choosing the best plan for them. 01:01:25.190 --> 01:01:25.890 align:middle line:90% Yeah. 01:01:25.890 --> 01:01:28.250 align:middle line:84% AUDIENCE: Would there be a benefit 01:01:28.250 --> 01:01:30.490 align:middle line:84% to having the government, right on the end of that, 01:01:30.490 --> 01:01:33.550 align:middle line:84% and say, OK, you competed to create these different plans, 01:01:33.550 --> 01:01:36.170 align:middle line:84% and now, here's a smaller menu of plans 01:01:36.170 --> 01:01:37.650 align:middle line:90% for senior citizens [INAUDIBLE]? 01:01:37.650 --> 01:01:39.150 align:middle line:84% JON GRUBER: We'll come back to that. 01:01:39.150 --> 01:01:39.660 align:middle line:90% Absolutely. 01:01:39.660 --> 01:01:41.160 align:middle line:84% Basically, could the government play 01:01:41.160 --> 01:01:42.740 align:middle line:84% with the choice architecture in a way 01:01:42.740 --> 01:01:44.323 align:middle line:84% that you have the best of both worlds? 01:01:44.323 --> 01:01:45.860 align:middle line:90% And we'll come back to that. 01:01:45.860 --> 01:01:46.820 align:middle line:90% OK. 01:01:46.820 --> 01:01:51.140 align:middle line:84% What was most funky about Part D was the way 01:01:51.140 --> 01:01:52.680 align:middle line:90% they set up the cost sharing. 01:01:52.680 --> 01:01:54.340 align:middle line:90% Let me explain how it worked. 01:01:54.340 --> 01:01:56.620 align:middle line:84% The numbers aren't quite precise, 01:01:56.620 --> 01:02:00.780 align:middle line:84% and they're out of date, but here's how it worked. 01:02:00.780 --> 01:02:05.420 align:middle line:84% On this axis, we have how much you spend on drugs. 01:02:05.420 --> 01:02:12.060 align:middle line:84% And on this axis, we have your co-insurance rate, what percent 01:02:12.060 --> 01:02:14.340 align:middle line:90% of the cost of drugs you pay. 01:02:14.340 --> 01:02:16.700 align:middle line:84% Well, there was a $250 deductible. 01:02:16.700 --> 01:02:26.220 align:middle line:84% That meant for the first $250, you paid 100% Then after $250, 01:02:26.220 --> 01:02:30.820 align:middle line:90% it fell to 25%. 01:02:30.820 --> 01:02:33.760 align:middle line:84% So at that point, you paid 25% of the cost of the drug. 01:02:33.760 --> 01:02:38.790 align:middle line:84% The insurer paid 75%, until you got to-- 01:02:38.790 --> 01:02:46.150 align:middle line:84% let me get the number-- until you got to $2,500 of spending. 01:02:46.150 --> 01:02:51.910 align:middle line:84% At that point, it went back to you paying 100% until you got 01:02:51.910 --> 01:02:56.970 align:middle line:84% to $5,100, when it went down to 5%. 01:02:56.970 --> 01:03:00.910 align:middle line:90% 01:03:00.910 --> 01:03:08.590 align:middle line:84% Now, I hope you can see this does not make sense. 01:03:08.590 --> 01:03:14.470 align:middle line:84% It is hard to think of a chapter 12 justification for a structure 01:03:14.470 --> 01:03:16.375 align:middle line:84% like this because remember, the whole idea 01:03:16.375 --> 01:03:18.750 align:middle line:84% is we're trying to combat moral hazard versus consumption 01:03:18.750 --> 01:03:19.430 align:middle line:90% smoothing. 01:03:19.430 --> 01:03:22.772 align:middle line:84% Well, these people-- they're spending a lot on drugs. 01:03:22.772 --> 01:03:25.230 align:middle line:84% They need the consumption-- why do these people need better 01:03:25.230 --> 01:03:26.690 align:middle line:90% consumption than these people? 01:03:26.690 --> 01:03:28.648 align:middle line:84% The moral hazard's probably less because you're 01:03:28.648 --> 01:03:31.150 align:middle line:84% taking expensive drugs, and you probably really need them. 01:03:31.150 --> 01:03:33.790 align:middle line:84% And the consumption smoothing is probably higher 01:03:33.790 --> 01:03:36.040 align:middle line:84% because this is a lot of money to spend. 01:03:36.040 --> 01:03:38.640 align:middle line:84% So this is just clearly wrong-headed. 01:03:38.640 --> 01:03:39.620 align:middle line:90% Why did they do it? 01:03:39.620 --> 01:03:43.440 align:middle line:90% 01:03:43.440 --> 01:03:46.160 align:middle line:84% Why did they set up the structure this way. 01:03:46.160 --> 01:03:47.473 align:middle line:90% Chapter 9 answer. 01:03:47.473 --> 01:03:48.140 align:middle line:90% Give you a hint. 01:03:48.140 --> 01:03:49.932 align:middle line:84% It's a chapter 9 answer for why they set up 01:03:49.932 --> 01:03:51.700 align:middle line:90% the structure this way. 01:03:51.700 --> 01:03:52.200 align:middle line:90% Yeah. 01:03:52.200 --> 01:03:54.765 align:middle line:84% AUDIENCE: [INAUDIBLE] drop in that range, 01:03:54.765 --> 01:03:56.640 align:middle line:84% where you have the highest co-pay [INAUDIBLE] 01:03:56.640 --> 01:04:00.400 align:middle line:84% and the pharmaceutical companies are lobbying for that. 01:04:00.400 --> 01:04:02.140 align:middle line:84% JON GRUBER: Most drugs are here or here? 01:04:02.140 --> 01:04:03.080 align:middle line:90% AUDIENCE: Yeah, the one-- 01:04:03.080 --> 01:04:03.800 align:middle line:90% JON GRUBER: Here. 01:04:03.800 --> 01:04:04.425 align:middle line:90% AUDIENCE: Yeah. 01:04:04.425 --> 01:04:07.077 align:middle line:90% 01:04:07.077 --> 01:04:09.160 align:middle line:84% JON GRUBER: Yeah, but the pharmaceutical companies 01:04:09.160 --> 01:04:10.260 align:middle line:84% get the same amount of money either way. 01:04:10.260 --> 01:04:11.440 align:middle line:90% It's just what people pay. 01:04:11.440 --> 01:04:12.500 align:middle line:90% So that's not quite it. 01:04:12.500 --> 01:04:13.220 align:middle line:90% What else? 01:04:13.220 --> 01:04:14.840 align:middle line:84% AUDIENCE: Most voters said the drugs [INAUDIBLE]. 01:04:14.840 --> 01:04:15.720 align:middle line:90% JON GRUBER: Yeah. 01:04:15.720 --> 01:04:17.400 align:middle line:90% The risk is here. 01:04:17.400 --> 01:04:18.660 align:middle line:90% The voters are here. 01:04:18.660 --> 01:04:21.640 align:middle line:90% 01:04:21.640 --> 01:04:26.200 align:middle line:84% Or the voters are really here, in this flat part. 01:04:26.200 --> 01:04:29.520 align:middle line:84% So the idea was instead of designing insurance to minimize 01:04:29.520 --> 01:04:32.698 align:middle line:84% risk, they designed insurance to maximize votes, which is-- 01:04:32.698 --> 01:04:34.740 align:middle line:84% if they'd done the right thing, which said, look, 01:04:34.740 --> 01:04:38.430 align:middle line:84% we're going to have a 25% coinsurance to say $4,000, 01:04:38.430 --> 01:04:41.013 align:middle line:84% and then we'll get rid of it, all these others would be like, 01:04:41.013 --> 01:04:42.930 align:middle line:84% no, no, no, no, I don't-- or if they had said, 01:04:42.930 --> 01:04:45.550 align:middle line:84% we'll have a 30% coinsurance or something, then get rid of it, 01:04:45.550 --> 01:04:47.050 align:middle line:84% people would be like, oh no, I'm so upset. 01:04:47.050 --> 01:04:48.800 align:middle line:84% This way, we're like, OK, for most voters, 01:04:48.800 --> 01:04:49.912 align:middle line:90% we'll keep them protected. 01:04:49.912 --> 01:04:51.370 align:middle line:84% These guys get screwed, but there's 01:04:51.370 --> 01:04:55.290 align:middle line:84% where the risk is, but not many voters. 01:04:55.290 --> 01:04:58.610 align:middle line:84% So it's a classic political economy trade off. 01:04:58.610 --> 01:05:01.090 align:middle line:84% Fortunately, the Affordable Care Act fixed this, 01:05:01.090 --> 01:05:08.450 align:middle line:84% changed this to 25% And now there's now a cap at $2,000, 01:05:08.450 --> 01:05:09.800 align:middle line:90% where it goes to 0. 01:05:09.800 --> 01:05:11.550 align:middle line:84% So the new structure, starting next year-- 01:05:11.550 --> 01:05:15.223 align:middle line:84% the Affordable Care Act made this 25% to here. 01:05:15.223 --> 01:05:16.390 align:middle line:90% You don't need to know this. 01:05:16.390 --> 01:05:18.310 align:middle line:90% It's just for edification. 01:05:18.310 --> 01:05:22.250 align:middle line:84% Now the new structure is 25% till $2,000 and then 01:05:22.250 --> 01:05:25.290 align:middle line:90% you're capped. 01:05:25.290 --> 01:05:26.010 align:middle line:90% Yeah. 01:05:26.010 --> 01:05:28.210 align:middle line:84% AUDIENCE: That latter change happened when? 01:05:28.210 --> 01:05:30.530 align:middle line:84% JON GRUBER: That's part of the Inflation Reduction Act. 01:05:30.530 --> 01:05:32.810 align:middle line:90% Takes place in January. 01:05:32.810 --> 01:05:36.990 align:middle line:84% So that's an interesting-- just interesting political economy 01:05:36.990 --> 01:05:38.170 align:middle line:90% of how these things happen. 01:05:38.170 --> 01:05:41.433 align:middle line:84% Now, with Medicare-- with Medicaid, 01:05:41.433 --> 01:05:43.350 align:middle line:84% I talked about the benefits and not the costs. 01:05:43.350 --> 01:05:45.183 align:middle line:84% With Medicare, I'm going to flip the tables. 01:05:45.183 --> 01:05:45.750 align:middle line:90% Yeah. 01:05:45.750 --> 01:05:47.625 align:middle line:84% AUDIENCE: You said it takes place in January. 01:05:47.625 --> 01:05:50.350 align:middle line:84% Inflation Reduction Act was passed, what, 2023? 01:05:50.350 --> 01:05:52.050 align:middle line:90% JON GRUBER: 2022, I think. 01:05:52.050 --> 01:05:52.830 align:middle line:90% Yeah. 01:05:52.830 --> 01:05:54.610 align:middle line:90% AUDIENCE: Why wait two years? 01:05:54.610 --> 01:05:58.232 align:middle line:90% 01:05:58.232 --> 01:05:59.690 align:middle line:84% JON GRUBER: That's a good question. 01:05:59.690 --> 01:06:02.232 align:middle line:84% I mean you need to wait a little while because companies need 01:06:02.232 --> 01:06:04.222 align:middle line:84% to-- that's a big change for insurers 01:06:04.222 --> 01:06:06.430 align:middle line:84% and how they structure the nature of their insurance. 01:06:06.430 --> 01:06:08.350 align:middle line:84% [INAUDIBLE] they have things like call formularies, 01:06:08.350 --> 01:06:09.610 align:middle line:90% about which drugs are where. 01:06:09.610 --> 01:06:11.277 align:middle line:84% So you need to give them a little while. 01:06:11.277 --> 01:06:11.985 align:middle line:90% It's interesting. 01:06:11.985 --> 01:06:13.610 align:middle line:84% I'll talk about the Affordable Care Act 01:06:13.610 --> 01:06:15.830 align:middle line:84% and why the fundamental mistake was waiting so long. 01:06:15.830 --> 01:06:18.050 align:middle line:84% More time allows for better design, 01:06:18.050 --> 01:06:20.130 align:middle line:90% but it has costs as well. 01:06:20.130 --> 01:06:21.130 align:middle line:90% We'll come back to that. 01:06:21.130 --> 01:06:24.618 align:middle line:90% 01:06:24.618 --> 01:06:27.270 align:middle line:84% So with Medicaid, I talked about trying 01:06:27.270 --> 01:06:29.190 align:middle line:84% to measure the benefits that ignore the costs. 01:06:29.190 --> 01:06:30.610 align:middle line:84% With Medicare, let's ignore the benefits. 01:06:30.610 --> 01:06:32.260 align:middle line:84% It's good people have health insurance. 01:06:32.260 --> 01:06:34.200 align:middle line:90% Let's focus on the costs. 01:06:34.200 --> 01:06:39.000 align:middle line:84% The costs here are interesting because this-- remember 01:06:39.000 --> 01:06:41.680 align:middle line:84% in the first lecture, I talked about CBO, the Congressional 01:06:41.680 --> 01:06:43.440 align:middle line:84% Budget Office, and how they're the only thing keeping 01:06:43.440 --> 01:06:45.648 align:middle line:84% our democracy safe because they're the people who are 01:06:45.648 --> 01:06:47.220 align:middle line:90% truth tellers in Washington. 01:06:47.220 --> 01:06:50.480 align:middle line:84% Well, it's a fascinating example that I like to use CBO. 01:06:50.480 --> 01:06:53.440 align:middle line:84% Medicare was developed by Lyndon Johnson before CBO. 01:06:53.440 --> 01:06:55.760 align:middle line:90% And he basically lied. 01:06:55.760 --> 01:06:58.820 align:middle line:84% He basically had his advisors say what Medicare would cost, 01:06:58.820 --> 01:07:00.760 align:middle line:90% and they were just way off. 01:07:00.760 --> 01:07:02.385 align:middle line:84% Now, were advisors just wrong-headed? 01:07:02.385 --> 01:07:03.260 align:middle line:90% I think probably not. 01:07:03.260 --> 01:07:05.280 align:middle line:84% I think they just understated it. 01:07:05.280 --> 01:07:08.040 align:middle line:90% But Medicare grew exponentially. 01:07:08.040 --> 01:07:12.240 align:middle line:84% So basically, Medicare started out-- 01:07:12.240 --> 01:07:17.040 align:middle line:84% it was a $64-million program in 1966. 01:07:17.040 --> 01:07:21.960 align:middle line:84% 14 years later, it was a $32-billion program. 01:07:21.960 --> 01:07:25.560 align:middle line:84% It had gone from $64 million to $32 billion in 14 years 01:07:25.560 --> 01:07:27.160 align:middle line:90% and was growing exponentially. 01:07:27.160 --> 01:07:29.840 align:middle line:84% And Congress was like, wait a second. 01:07:29.840 --> 01:07:32.330 align:middle line:84% That didn't quite work out like we thought. 01:07:32.330 --> 01:07:33.970 align:middle line:90% We need to control costs. 01:07:33.970 --> 01:07:36.210 align:middle line:84% So Congress came up with two different approaches 01:07:36.210 --> 01:07:39.210 align:middle line:84% to controlling costs, which are fascinating to study. 01:07:39.210 --> 01:07:40.972 align:middle line:84% The first approach-- once again, always, I 01:07:40.972 --> 01:07:42.430 align:middle line:84% use the term, fascinating, loosely. 01:07:42.430 --> 01:07:43.888 align:middle line:84% My family makes fun of me for this. 01:07:43.888 --> 01:07:47.650 align:middle line:84% But once again, you have to listen. 01:07:47.650 --> 01:07:50.570 align:middle line:84% The first approach was to say, hey, 01:07:50.570 --> 01:07:53.570 align:middle line:84% let's move away from fee-for-service towards 01:07:53.570 --> 01:07:55.990 align:middle line:90% a prospective managed care-- 01:07:55.990 --> 01:07:58.170 align:middle line:84% not managed care, but a prospective reimbursement 01:07:58.170 --> 01:07:58.670 align:middle line:90% system. 01:07:58.670 --> 01:08:00.250 align:middle line:84% In particular, they set up what was 01:08:00.250 --> 01:08:04.670 align:middle line:84% called the prospective payment system for hospitals in 1980. 01:08:04.670 --> 01:08:08.010 align:middle line:84% This was very early, very innovative. 01:08:08.010 --> 01:08:09.172 align:middle line:90% How did the system work? 01:08:09.172 --> 01:08:10.630 align:middle line:84% The way it works-- they said, look, 01:08:10.630 --> 01:08:12.713 align:middle line:84% we're not just going to pay you whatever you bill. 01:08:12.713 --> 01:08:15.530 align:middle line:84% We're going to do the following thing. 01:08:15.530 --> 01:08:17.930 align:middle line:84% We're going to pay you based on how sick people are 01:08:17.930 --> 01:08:19.430 align:middle line:90% when they come to the hospital. 01:08:19.430 --> 01:08:21.450 align:middle line:90% We'll pay you a fixed amount. 01:08:21.450 --> 01:08:24.310 align:middle line:84% So someone comes in with this level of sickness, 01:08:24.310 --> 01:08:25.550 align:middle line:90% we'll pay you this much. 01:08:25.550 --> 01:08:27.550 align:middle line:84% This level of sickness, we'll pay you this much. 01:08:27.550 --> 01:08:28.830 align:middle line:84% And then you do whatever the hell you want. 01:08:28.830 --> 01:08:30.420 align:middle line:90% We're not paying you anymore. 01:08:30.420 --> 01:08:32.819 align:middle line:84% So it's prospective reimbursement in hospitals. 01:08:32.819 --> 01:08:35.420 align:middle line:84% Instead of hospitals billing for every test, 01:08:35.420 --> 01:08:40.180 align:middle line:84% hospitals billed based on how sick the people are coming in. 01:08:40.180 --> 01:08:42.939 align:middle line:84% Now, there are roughly 10,000 different diagnoses, 01:08:42.939 --> 01:08:45.064 align:middle line:84% and they didn't want to deal with different prices, 01:08:45.064 --> 01:08:47.640 align:middle line:84% so they set up what was called diagnosis-related groups, 01:08:47.640 --> 01:08:49.939 align:middle line:90% 457 DRGs. 01:08:49.939 --> 01:08:54.740 align:middle line:84% These are basically like hysterectomy, not hysterectomy, 01:08:54.740 --> 01:08:58.439 align:middle line:84% depending on a woman's age or various things, or broken arm, 01:08:58.439 --> 01:09:01.217 align:middle line:84% not where-- which part of the arm is broken, et cetera. 01:09:01.217 --> 01:09:03.300 align:middle line:84% Heart attack-- there's four or five different ones 01:09:03.300 --> 01:09:05.925 align:middle line:84% for heart attack, where there's hundreds of different diagnoses 01:09:05.925 --> 01:09:06.859 align:middle line:90% for heart attack. 01:09:06.859 --> 01:09:09.317 align:middle line:84% So basically, they said we're going to have a fixed amount, 01:09:09.317 --> 01:09:10.620 align:middle line:90% we pay per DRG. 01:09:10.620 --> 01:09:13.800 align:middle line:84% So it's a true prospective reimbursement system. 01:09:13.800 --> 01:09:18.220 align:middle line:84% We are going to pay a fixed amount per DRG. 01:09:18.220 --> 01:09:22.220 align:middle line:90% Now, this is a great experiment. 01:09:22.220 --> 01:09:23.220 align:middle line:90% What effect did it have? 01:09:23.220 --> 01:09:24.678 align:middle line:84% You had, basically, something where 01:09:24.678 --> 01:09:27.740 align:middle line:84% we're moving from retrospective to prospective reimbursement, 01:09:27.740 --> 01:09:29.770 align:middle line:84% but without all the other managed care features. 01:09:29.770 --> 01:09:31.870 align:middle line:84% This is just a pure reimbursement change. 01:09:31.870 --> 01:09:33.670 align:middle line:84% So it's not tangled up in all the things 01:09:33.670 --> 01:09:36.010 align:middle line:84% like independent practice organization stuff. 01:09:36.010 --> 01:09:37.430 align:middle line:90% It's the same doctors. 01:09:37.430 --> 01:09:38.925 align:middle line:84% You're just paying them different. 01:09:38.925 --> 01:09:40.050 align:middle line:90% So it's a great experiment. 01:09:40.050 --> 01:09:41.430 align:middle line:90% What did we find? 01:09:41.430 --> 01:09:46.229 align:middle line:84% What we found was more evidence for the flat of the curve, 01:09:46.229 --> 01:09:48.350 align:middle line:84% that basically, what we found was 01:09:48.350 --> 01:09:51.470 align:middle line:84% there was an enormous reduction overnight 01:09:51.470 --> 01:09:53.510 align:middle line:90% in how the elderly were treated. 01:09:53.510 --> 01:09:56.670 align:middle line:84% So the average hospital stay for an elderly person 01:09:56.670 --> 01:09:59.910 align:middle line:84% fell from 10 days to eight days within, like, two years. 01:09:59.910 --> 01:10:01.190 align:middle line:90% That's amazing. 01:10:01.190 --> 01:10:08.430 align:middle line:84% Basically, we saw a huge drop in the intensity of hospital care 01:10:08.430 --> 01:10:12.950 align:middle line:84% with no evidence of adverse effects on elder outcomes. 01:10:12.950 --> 01:10:15.250 align:middle line:84% Classic flat of the curve, that these hospitals 01:10:15.250 --> 01:10:18.033 align:middle line:84% were delivering excess care into the fee-for-service system. 01:10:18.033 --> 01:10:19.450 align:middle line:84% You move to a prospective payment. 01:10:19.450 --> 01:10:21.410 align:middle line:84% They deliver less, but not so much less 01:10:21.410 --> 01:10:23.370 align:middle line:84% the patients are worse off, just enough less. 01:10:23.370 --> 01:10:24.870 align:middle line:84% They don't quite make as much money. 01:10:24.870 --> 01:10:28.040 align:middle line:90% 01:10:28.040 --> 01:10:30.680 align:middle line:84% I did skip one thing that I want to come back 01:10:30.680 --> 01:10:34.520 align:middle line:84% to for one second, which is what I didn't mention 01:10:34.520 --> 01:10:36.827 align:middle line:84% is you could have asked me-- you could have said, 01:10:36.827 --> 01:10:37.660 align:middle line:90% wait a second, John. 01:10:37.660 --> 01:10:40.040 align:middle line:90% I'm confused by something. 01:10:40.040 --> 01:10:42.800 align:middle line:84% The RAND Health Insurance Experiment 01:10:42.800 --> 01:10:44.520 align:middle line:84% showed that the generosity of insurance 01:10:44.520 --> 01:10:46.920 align:middle line:84% didn't seem to matter for people's health. 01:10:46.920 --> 01:10:49.320 align:middle line:84% But you've just argued passionately that 01:10:49.320 --> 01:10:51.800 align:middle line:84% getting Medicaid matters for people's health. 01:10:51.800 --> 01:10:53.560 align:middle line:90% How is that consistent? 01:10:53.560 --> 01:10:57.240 align:middle line:84% And I would argue those two put together exactly illustrate 01:10:57.240 --> 01:11:00.160 align:middle line:84% the curve, which is getting people from having 01:11:00.160 --> 01:11:02.400 align:middle line:90% no coverage to some coverage-- 01:11:02.400 --> 01:11:05.080 align:middle line:84% it matters for the health, but how generous that coverage is 01:11:05.080 --> 01:11:07.120 align:middle line:90% doesn't matter nearly as much. 01:11:07.120 --> 01:11:09.040 align:middle line:84% So those two pieces put together are really 01:11:09.040 --> 01:11:11.640 align:middle line:84% what traces out this curve I keep coming back to, 01:11:11.640 --> 01:11:14.360 align:middle line:84% that it matters that you get some health care. 01:11:14.360 --> 01:11:17.590 align:middle line:84% It might not matter as much how generous that health care is. 01:11:17.590 --> 01:11:21.400 align:middle line:84% Just coming back to that before I forget. 01:11:21.400 --> 01:11:24.380 align:middle line:84% Now we're back on Medicare prospective payment. 01:11:24.380 --> 01:11:26.060 align:middle line:90% They put in prospective payment. 01:11:26.060 --> 01:11:26.997 align:middle line:90% It worked. 01:11:26.997 --> 01:11:28.080 align:middle line:90% Lowered health care costs. 01:11:28.080 --> 01:11:31.380 align:middle line:84% Health care outcomes weren't any worse for a while. 01:11:31.380 --> 01:11:32.757 align:middle line:90% Then it stopped working. 01:11:32.757 --> 01:11:34.340 align:middle line:84% Hospital costs grew slowly for a while 01:11:34.340 --> 01:11:38.380 align:middle line:84% and then started growing again, very rapidly, because basically, 01:11:38.380 --> 01:11:41.620 align:middle line:84% hospitals figured out multiple different ways 01:11:41.620 --> 01:11:43.660 align:middle line:90% around this system. 01:11:43.660 --> 01:11:45.580 align:middle line:90% You guys are smart. 01:11:45.580 --> 01:11:47.980 align:middle line:84% Let's say you're running a profit-maximizing hospital. 01:11:47.980 --> 01:11:49.980 align:middle line:90% I put in the DRG system. 01:11:49.980 --> 01:11:52.940 align:middle line:84% How would you get around it to get back to making more money? 01:11:52.940 --> 01:11:54.860 align:middle line:90% What could you do? 01:11:54.860 --> 01:11:57.300 align:middle line:84% Doctors aren't allowed to answer. 01:11:57.300 --> 01:11:59.540 align:middle line:90% What would you do? 01:11:59.540 --> 01:12:00.040 align:middle line:90% Come on. 01:12:00.040 --> 01:12:01.080 align:middle line:90% You guys are smart. 01:12:01.080 --> 01:12:02.640 align:middle line:84% I know you're not evil, but imagine you were. 01:12:02.640 --> 01:12:02.980 align:middle line:90% Yeah. 01:12:02.980 --> 01:12:04.220 align:middle line:90% AUDIENCE: Change the diagnosis. 01:12:04.220 --> 01:12:06.500 align:middle line:84% JON GRUBER: You could change how people are diagnosed. 01:12:06.500 --> 01:12:10.100 align:middle line:84% If you look at the data, America's elders 01:12:10.100 --> 01:12:14.360 align:middle line:90% got incredibly sick overnight. 01:12:14.360 --> 01:12:17.780 align:middle line:84% There was a big jump in the diagnosis. 01:12:17.780 --> 01:12:19.720 align:middle line:84% Because you're paid based on diagnosis, 01:12:19.720 --> 01:12:21.480 align:middle line:84% they started diagnosing people sicker. 01:12:21.480 --> 01:12:24.790 align:middle line:84% Now, you can't diagnose a broken arm as a heart attack. 01:12:24.790 --> 01:12:27.223 align:middle line:84% But you can diagnose someone who came up 01:12:27.223 --> 01:12:28.890 align:middle line:84% with the broken arm and a heart attack-- 01:12:28.890 --> 01:12:30.807 align:middle line:84% you can focus on the heart attack part and not 01:12:30.807 --> 01:12:32.710 align:middle line:84% the broken arm part, because the heart attack 01:12:32.710 --> 01:12:33.730 align:middle line:90% part gets paid more. 01:12:33.730 --> 01:12:41.010 align:middle line:90% 01:12:41.010 --> 01:12:43.310 align:middle line:84% There's the general, gee, we'll sort of on the margin 01:12:43.310 --> 01:12:44.977 align:middle line:84% code a little more complicated, and then 01:12:44.977 --> 01:12:49.938 align:middle line:84% there's the extreme examples, like in 1995, 01:12:49.938 --> 01:12:52.230 align:middle line:84% Cedars-Sinai Medical Center in Miami-- once again, it's 01:12:52.230 --> 01:12:53.010 align:middle line:90% always Florida. 01:12:53.010 --> 01:12:55.670 align:middle line:90% 01:12:55.670 --> 01:13:00.310 align:middle line:84% There were two DRGs, one for respiratory illness 01:13:00.310 --> 01:13:03.450 align:middle line:84% with a complex respiratory infection 01:13:03.450 --> 01:13:06.055 align:middle line:84% and one with pneumonia with complications. 01:13:06.055 --> 01:13:07.930 align:middle line:84% So I have a respiratory illness-- pneumonia's 01:13:07.930 --> 01:13:08.850 align:middle line:90% a respiratory illness. 01:13:08.850 --> 01:13:11.058 align:middle line:84% I could either call it, pneumonia with complications, 01:13:11.058 --> 01:13:14.910 align:middle line:84% where I got paid $1,700, or respiratory illness with 01:13:14.910 --> 01:13:16.760 align:middle line:90% complex-- 01:13:16.760 --> 01:13:20.590 align:middle line:84% complex respiratory illness, for which I got $5,700. 01:13:20.590 --> 01:13:27.240 align:middle line:84% Well, at Jackson Memorial, which was a nonprofit hospital nearby, 01:13:27.240 --> 01:13:30.280 align:middle line:84% 28% of the cases were coded as the complex. 01:13:30.280 --> 01:13:33.020 align:middle line:84% In the for-profit Cedars Medical Center, 01:13:33.020 --> 01:13:35.500 align:middle line:90% 93% were coded as complex. 01:13:35.500 --> 01:13:37.500 align:middle line:84% Now, maybe they got a little more complex cases, 01:13:37.500 --> 01:13:40.440 align:middle line:84% but it's hard to imagine it was that different. 01:13:40.440 --> 01:13:45.760 align:middle line:84% And indeed, a number of hospitals, 01:13:45.760 --> 01:13:48.480 align:middle line:84% actually, were subject to huge lawsuits 01:13:48.480 --> 01:13:51.080 align:middle line:90% for clearly abusing the system. 01:13:51.080 --> 01:13:55.400 align:middle line:84% But to be fair, the system was designed to be abused. 01:13:55.400 --> 01:13:56.460 align:middle line:90% What do I mean by that? 01:13:56.460 --> 01:14:00.615 align:middle line:84% What I mean by that is, literally, any-- 01:14:00.615 --> 01:14:02.240 align:middle line:84% it's designed to be abused in the sense 01:14:02.240 --> 01:14:05.062 align:middle line:84% that it was partially retrospective. 01:14:05.062 --> 01:14:06.020 align:middle line:90% What do I mean by that? 01:14:06.020 --> 01:14:08.080 align:middle line:84% What I mean by that is, for example, 01:14:08.080 --> 01:14:11.440 align:middle line:84% if someone comes in with a heart attack, 01:14:11.440 --> 01:14:15.640 align:middle line:84% you could code them as cardiac arrest unexplained and get 01:14:15.640 --> 01:14:17.680 align:middle line:90% $5,000. 01:14:17.680 --> 01:14:21.970 align:middle line:84% Or you could do a coronary bypass, 01:14:21.970 --> 01:14:23.570 align:middle line:90% which is a different DRG-- 01:14:23.570 --> 01:14:27.410 align:middle line:84% heart attack with a coronary bypass got $33,000. 01:14:27.410 --> 01:14:30.003 align:middle line:84% Well, that sounds a lot like fee-for-service medicine to me. 01:14:30.003 --> 01:14:31.670 align:middle line:84% You do something, and you get paid more. 01:14:31.670 --> 01:14:33.810 align:middle line:84% That's not perspective reimbursement anymore. 01:14:33.810 --> 01:14:35.685 align:middle line:84% So part of the problem was they didn't really 01:14:35.685 --> 01:14:38.030 align:middle line:84% commit to true prospective reimbursement. 01:14:38.030 --> 01:14:40.210 align:middle line:84% They said, well, gee, we'll actually pay some more 01:14:40.210 --> 01:14:42.010 align:middle line:90% if you do some of this stuff. 01:14:42.010 --> 01:14:44.530 align:middle line:84% So that was part of the problem, that it was still 01:14:44.530 --> 01:14:47.970 align:middle line:90% a fee-for-service system. 01:14:47.970 --> 01:14:53.690 align:middle line:84% Another problem was it paid hospitals per admission. 01:14:53.690 --> 01:14:55.250 align:middle line:90% So what do you do? 01:14:55.250 --> 01:14:58.997 align:middle line:90% Well, Paul comes in. 01:14:58.997 --> 01:15:00.830 align:middle line:84% Don't worry, Paul, you're not actually sick. 01:15:00.830 --> 01:15:01.872 align:middle line:90% I'm just using your name. 01:15:01.872 --> 01:15:03.310 align:middle line:84% Paul comes in-- he looks shocked. 01:15:03.310 --> 01:15:06.507 align:middle line:90% Paul comes in, and we treat him. 01:15:06.507 --> 01:15:08.590 align:middle line:84% And we're like, Paul could use a couple more days. 01:15:08.590 --> 01:15:12.970 align:middle line:84% We're going to discharge him and readmit him and get paid twice. 01:15:12.970 --> 01:15:16.090 align:middle line:84% So hospitals started doing a lot of excessive discharge 01:15:16.090 --> 01:15:18.290 align:middle line:90% and readmission. 01:15:18.290 --> 01:15:19.930 align:middle line:84% Indeed, part of the Affordable Care 01:15:19.930 --> 01:15:21.680 align:middle line:84% Act I'll talk about next time was actually 01:15:21.680 --> 01:15:26.580 align:middle line:84% a policy to penalize hospitals for the success of readmission. 01:15:26.580 --> 01:15:29.620 align:middle line:84% So that's another thing you could do. 01:15:29.620 --> 01:15:32.540 align:middle line:84% Another thing you could do is you could recognize 01:15:32.540 --> 01:15:35.940 align:middle line:84% that patients get treated in many locations, only some 01:15:35.940 --> 01:15:38.780 align:middle line:84% of which were reimbursed by PPS and some of which 01:15:38.780 --> 01:15:40.440 align:middle line:90% were reimbursed fee-for-service. 01:15:40.440 --> 01:15:44.740 align:middle line:84% So for example, if you break your hip, 01:15:44.740 --> 01:15:47.500 align:middle line:84% you go from-- you go to an acute-care hospital 01:15:47.500 --> 01:15:50.440 align:middle line:90% to a rehab hospital. 01:15:50.440 --> 01:15:52.440 align:middle line:84% You get your surgery at the acute-care hospital. 01:15:52.440 --> 01:15:54.080 align:middle line:84% They keep you for some amount of time. 01:15:54.080 --> 01:15:55.955 align:middle line:84% Then they ship you off to the rehab hospital, 01:15:55.955 --> 01:15:58.300 align:middle line:90% where you get rehabbed. 01:15:58.300 --> 01:16:01.180 align:middle line:84% The acute-care hospital is reimbursed by PPS. 01:16:01.180 --> 01:16:03.768 align:middle line:84% The rehab hospital is still fee-for-service. 01:16:03.768 --> 01:16:04.560 align:middle line:90% So what do they do? 01:16:04.560 --> 01:16:07.102 align:middle line:84% They kick people, real quick, out of the acute-care hospital, 01:16:07.102 --> 01:16:09.460 align:middle line:84% put them in rehab, and they paid-- so they save money. 01:16:09.460 --> 01:16:11.000 align:middle line:84% Our acute-care hospitals went down, 01:16:11.000 --> 01:16:13.780 align:middle line:84% but our rehab hospital bills went up. 01:16:13.780 --> 01:16:15.540 align:middle line:84% So basically, hospitals became very good 01:16:15.540 --> 01:16:19.812 align:middle line:84% at figuring out ways to game the system. 01:16:19.812 --> 01:16:22.270 align:middle line:84% What I talk about in the book is, basically, the difficulty 01:16:22.270 --> 01:16:23.410 align:middle line:90% of partial reform. 01:16:23.410 --> 01:16:26.010 align:middle line:84% It's like squeezing a-- squeezing a pillow. 01:16:26.010 --> 01:16:28.310 align:middle line:84% You squeeze the middle, and it comes out the sides, 01:16:28.310 --> 01:16:32.010 align:middle line:84% that basically, it's very hard-- when you have underpaid, 01:16:32.010 --> 01:16:34.930 align:middle line:84% overworked government regulators and overpaid, 01:16:34.930 --> 01:16:37.190 align:middle line:84% under-worked private executives-- 01:16:37.190 --> 01:16:38.430 align:middle line:90% [LAUGHTER] 01:16:38.430 --> 01:16:40.310 align:middle line:84% --the private executives are going to win. 01:16:40.310 --> 01:16:41.910 align:middle line:84% And basically, the trick is whenever 01:16:41.910 --> 01:16:44.270 align:middle line:84% you try to play whack-a-mole, you're 01:16:44.270 --> 01:16:45.670 align:middle line:90% always going to be behind. 01:16:45.670 --> 01:16:48.150 align:middle line:84% It doesn't mean it's not worth doing. 01:16:48.150 --> 01:16:51.030 align:middle line:84% Even with all these changes, hospital costs 01:16:51.030 --> 01:16:55.210 align:middle line:84% still grew more slowly after PPS than they did before. 01:16:55.210 --> 01:16:58.030 align:middle line:84% It's just they slowed a lot, then sped back up, but never as 01:16:58.030 --> 01:17:00.203 align:middle line:90% fast as they were before. 01:17:00.203 --> 01:17:01.870 align:middle line:84% The problem is people take these lessons 01:17:01.870 --> 01:17:04.537 align:middle line:84% as nihilistic like, oh, my god, we shouldn't have done anything. 01:17:04.537 --> 01:17:05.678 align:middle line:90% That's not the answer. 01:17:05.678 --> 01:17:07.470 align:middle line:84% We were still controlling health care costs 01:17:07.470 --> 01:17:10.370 align:middle line:84% better with this system than we were before we had the system. 01:17:10.370 --> 01:17:13.518 align:middle line:84% It just had a number of flaws because of its-- 01:17:13.518 --> 01:17:14.810 align:middle line:90% so we've tried to deal with it. 01:17:14.810 --> 01:17:15.518 align:middle line:90% So what do we do? 01:17:15.518 --> 01:17:16.370 align:middle line:90% We've whacked moles. 01:17:16.370 --> 01:17:18.030 align:middle line:84% We've put in a readmission penalty. 01:17:18.030 --> 01:17:20.430 align:middle line:84% I'll talk about that next time, to try to lower costs. 01:17:20.430 --> 01:17:22.770 align:middle line:84% We've moved the whole system to PPS, 01:17:22.770 --> 01:17:24.930 align:middle line:84% so now you can't game the system by moving someone 01:17:24.930 --> 01:17:27.330 align:middle line:84% to rehab because the rehab's on PPS, too. 01:17:27.330 --> 01:17:30.410 align:middle line:84% So we've taken steps to try to address some of these holes, 01:17:30.410 --> 01:17:32.090 align:middle line:84% but it's never going to be perfect. 01:17:32.090 --> 01:17:32.590 align:middle line:90% Yeah. 01:17:32.590 --> 01:17:34.132 align:middle line:84% AUDIENCE: So the heart-type example-- 01:17:34.132 --> 01:17:36.050 align:middle line:84% does that create an incentive for the hospital 01:17:36.050 --> 01:17:37.410 align:middle line:90% to give a coronary bypass? 01:17:37.410 --> 01:17:38.868 align:middle line:84% JON GRUBER: Well, certainly no more 01:17:38.868 --> 01:17:41.230 align:middle line:84% than they had before because they always got paid more. 01:17:41.230 --> 01:17:44.090 align:middle line:84% It just didn't decrease the incentive. 01:17:44.090 --> 01:17:46.670 align:middle line:84% Basically, it used to be you'd bill more for it. 01:17:46.670 --> 01:17:48.070 align:middle line:90% Now you get a higher DRG for it. 01:17:48.070 --> 01:17:48.970 align:middle line:90% It's the same thing. 01:17:48.970 --> 01:17:50.270 align:middle line:84% So it didn't decrease the incentive. 01:17:50.270 --> 01:17:50.770 align:middle line:90% Yeah. 01:17:50.770 --> 01:17:54.050 align:middle line:84% AUDIENCE: Going back to what you argued how [INAUDIBLE] 01:17:54.050 --> 01:17:56.610 align:middle line:84% before or after, I guess that's much harder than just 01:17:56.610 --> 01:17:58.067 align:middle line:90% looking at general health. 01:17:58.067 --> 01:17:59.150 align:middle line:90% JON GRUBER: Oh, excellent. 01:17:59.150 --> 01:17:59.870 align:middle line:90% Excellent question. 01:17:59.870 --> 01:18:01.578 align:middle line:84% So how do you do the empirical work here? 01:18:01.578 --> 01:18:02.970 align:middle line:90% Great question. 01:18:02.970 --> 01:18:04.050 align:middle line:90% Two ways. 01:18:04.050 --> 01:18:07.710 align:middle line:84% One is that PPS was phased in different parts. 01:18:07.710 --> 01:18:09.072 align:middle line:90% It was phased in over time. 01:18:09.072 --> 01:18:10.530 align:middle line:84% And the rate at which hospitals got 01:18:10.530 --> 01:18:13.290 align:middle line:84% phased in differed based on what their costs were initially. 01:18:13.290 --> 01:18:15.010 align:middle line:84% So you have a diff-in-diff based on 01:18:15.010 --> 01:18:17.360 align:middle line:84% at what point hospitals were subject to PPS. 01:18:17.360 --> 01:18:18.543 align:middle line:90% It was over several years. 01:18:18.543 --> 01:18:20.460 align:middle line:84% You also had the fact that some states already 01:18:20.460 --> 01:18:23.520 align:middle line:84% had a system like PPS, so they were exempt from the PPS reform, 01:18:23.520 --> 01:18:25.980 align:middle line:84% so you have a state difference-in-difference. 01:18:25.980 --> 01:18:28.940 align:middle line:84% You can also do things like-- if you look at-- 01:18:28.940 --> 01:18:32.980 align:middle line:84% last thing-- if you look at figure 16-2, what happened was-- 01:18:32.980 --> 01:18:35.100 align:middle line:84% and this is just a national change. 01:18:35.100 --> 01:18:40.380 align:middle line:84% It used to be that at long-term care hospitals, basically, 01:18:40.380 --> 01:18:43.262 align:middle line:84% they would keep patients a long time. 01:18:43.262 --> 01:18:45.720 align:middle line:84% And they were worried they were overpaying these hospitals. 01:18:45.720 --> 01:18:47.020 align:middle line:84% They put in a policy, which said-- 01:18:47.020 --> 01:18:48.940 align:middle line:84% they were worried that people were abusing long-term care 01:18:48.940 --> 01:18:50.913 align:middle line:84% hospitals when they didn't need to be there. 01:18:50.913 --> 01:18:53.080 align:middle line:84% Said, we're going to pay less for the first 29 days. 01:18:53.080 --> 01:18:55.460 align:middle line:84% Only once you've been there 30 days will we pay more. 01:18:55.460 --> 01:18:56.718 align:middle line:90% Look what happened. 01:18:56.718 --> 01:18:58.260 align:middle line:84% Afterwards, there was a huge increase 01:18:58.260 --> 01:18:59.927 align:middle line:84% in people staying 30 days because that's 01:18:59.927 --> 01:19:01.300 align:middle line:90% when the reimbursement went up. 01:19:01.300 --> 01:19:03.300 align:middle line:84% It's a national change, but it's pretty obvious something 01:19:03.300 --> 01:19:04.175 align:middle line:90% like that's going on. 01:19:04.175 --> 01:19:06.700 align:middle line:84% So you either look for sharp event studies like that, 01:19:06.700 --> 01:19:08.160 align:middle line:84% or you try to find control groups. 01:19:08.160 --> 01:19:10.080 align:middle line:90% Great question. 01:19:10.080 --> 01:19:12.050 align:middle line:90% Let's stop there. 01:19:12.050 --> 01:19:16.000 align:middle line:90%