WEBVTT 00:00:00.000 --> 00:00:02.520 The following content is provided under a Creative 00:00:02.520 --> 00:00:03.940 Commons license. 00:00:03.940 --> 00:00:06.360 Your support will help MIT OpenCourseWare 00:00:06.360 --> 00:00:10.660 continue to offer high-quality educational resources for free. 00:00:10.660 --> 00:00:13.320 To make a donation or view additional materials 00:00:13.320 --> 00:00:17.160 from hundreds of MIT courses, visit MIT OpenCourseWare 00:00:17.160 --> 00:00:18.370 at ocw.mit.edu. 00:00:26.330 --> 00:00:30.530 RICHARD SCHMALENSEE: OK, let's get underway. 00:00:30.530 --> 00:00:31.610 Yeah, this thing is on. 00:00:31.610 --> 00:00:34.250 Any burning questions from last time? 00:00:34.250 --> 00:00:39.710 Any debate points you really had to make, and didn't get made? 00:00:39.710 --> 00:00:41.420 OK. 00:00:41.420 --> 00:00:45.060 Today we're talking about path dependence. 00:00:45.060 --> 00:00:45.560 Sorry. 00:00:45.560 --> 00:00:48.780 You were raising your hand, or was that a wave? 00:00:48.780 --> 00:00:49.700 AUDIENCE: Yeah. 00:00:49.700 --> 00:00:56.990 I didn't say much yesterday, but on the note of labor and-- 00:00:56.990 --> 00:01:01.220 a lot of times I'm on the labor side of things yesterday, 00:01:01.220 --> 00:01:04.040 or at least a bit of it. 00:01:04.040 --> 00:01:10.820 And people were talking about how providing, let's say, 00:01:10.820 --> 00:01:14.210 workers' rights, union rights, et cetera, things like that, 00:01:14.210 --> 00:01:19.730 would impede the freedom of the working process. 00:01:19.730 --> 00:01:25.520 And I feel that almost every industrialized country nowadays 00:01:25.520 --> 00:01:28.700 has gone through this process of horrible workers' rights 00:01:28.700 --> 00:01:33.480 to a system more like we have now, let's say. 00:01:33.480 --> 00:01:38.780 China's going through what we did 100 years ago currently, 00:01:38.780 --> 00:01:42.800 with some of their whole stuff, with workers working 00:01:42.800 --> 00:01:45.980 way too long shifts, and there being not much safety 00:01:45.980 --> 00:01:48.670 on chemicals, et cetera, et cetera, et cetera. 00:01:48.670 --> 00:01:51.170 RICHARD SCHMALENSEE: You're just arguing that Sutherland was 00:01:51.170 --> 00:01:52.950 on the wrong side of history. 00:01:52.950 --> 00:01:53.450 Yes? 00:01:57.140 --> 00:02:00.628 And so what we've come to is where we should have come? 00:02:00.628 --> 00:02:01.420 AUDIENCE: Kind of-- 00:02:01.420 --> 00:02:02.670 RICHARD SCHMALENSEE: Kind of-- 00:02:02.670 --> 00:02:03.860 OK. 00:02:03.860 --> 00:02:08.810 But you could understand how, circa 1935-- 00:02:08.810 --> 00:02:12.990 wasn't obvious which was the right side of history. 00:02:12.990 --> 00:02:15.885 It is now, of course. 00:02:15.885 --> 00:02:17.870 AUDIENCE: You could say hindsight's 2020, 00:02:17.870 --> 00:02:21.500 but if someone's working 10 hours a day 00:02:21.500 --> 00:02:24.600 in highly dangerous conditions, you 00:02:24.600 --> 00:02:27.820 might say that's not kosher then either. 00:02:29.992 --> 00:02:31.950 RICHARD SCHMALENSEE: The right side of the room 00:02:31.950 --> 00:02:34.890 would have replied, and maybe will reply, that-- 00:02:34.890 --> 00:02:38.130 take another job, if you don't like it. 00:02:38.130 --> 00:02:39.460 Become an investment banker. 00:02:39.460 --> 00:02:40.297 Come on. 00:02:40.297 --> 00:02:42.630 AUDIENCE: There are high impedances to doing [INAUDIBLE] 00:02:42.630 --> 00:02:43.922 RICHARD SCHMALENSEE: There are. 00:02:43.922 --> 00:02:45.808 There are. 00:02:45.808 --> 00:02:47.850 Good set of points-- didn't mean to put you down. 00:02:50.640 --> 00:02:54.090 One the problems with doing that case is we sort of 00:02:54.090 --> 00:02:56.400 know how it worked out after that. 00:02:56.400 --> 00:02:59.570 We know that Cardozo and the whole 00:02:59.570 --> 00:03:02.130 you got to allow more regulation of the market 00:03:02.130 --> 00:03:04.770 to make-- for a variety of reasons, triumphed. 00:03:07.790 --> 00:03:10.140 But I'm not enough of a determinist 00:03:10.140 --> 00:03:11.570 to think that was inevitable. 00:03:11.570 --> 00:03:15.230 It happened, and it happened lots of places. 00:03:15.230 --> 00:03:17.272 AUDIENCE: It took a long time for that to happen. 00:03:17.272 --> 00:03:18.438 RICHARD SCHMALENSEE: It did. 00:03:18.438 --> 00:03:19.160 It did. 00:03:19.160 --> 00:03:22.510 Anything else on last time? 00:03:22.510 --> 00:03:23.200 OK. 00:03:23.200 --> 00:03:28.450 What I want to do today is talk about path dependence. 00:03:28.450 --> 00:03:32.500 And let me just do a little visual. 00:03:32.500 --> 00:03:38.810 We always like to think that choices get made, 00:03:38.810 --> 00:03:41.140 you look at the alternatives, and the choices 00:03:41.140 --> 00:03:43.990 are the obvious rational choices. 00:03:43.990 --> 00:03:45.820 So we use the energy systems that 00:03:45.820 --> 00:03:48.550 make the most sense to use. 00:03:48.550 --> 00:03:54.130 And let me just anecdotally suggest 00:03:54.130 --> 00:03:58.780 that can't be right, just that statement that, at any point 00:03:58.780 --> 00:04:01.510 in time, we're obviously using the systems that 00:04:01.510 --> 00:04:04.780 are the, technically, or even economically, best. 00:04:04.780 --> 00:04:09.910 The reason is, if you look across rich countries, 00:04:09.910 --> 00:04:13.120 they make different choices with the same available technology, 00:04:13.120 --> 00:04:15.590 the same rough standards of living. 00:04:15.590 --> 00:04:21.160 So here's rush hour in Copenhagen. The city's designed 00:04:21.160 --> 00:04:22.330 for bicycling. 00:04:22.330 --> 00:04:24.070 Lots of people use bicycles. 00:04:24.070 --> 00:04:28.090 You got similar photos from Amsterdam. 00:04:28.090 --> 00:04:30.130 I've got friends who have moved to Amsterdam. 00:04:30.130 --> 00:04:31.390 They never drive. 00:04:31.390 --> 00:04:32.590 They bicycle every place. 00:04:32.590 --> 00:04:36.520 It's just a different system of urban design. 00:04:36.520 --> 00:04:38.410 We've got bicycles. 00:04:38.410 --> 00:04:40.600 We could have gone that direction. 00:04:40.600 --> 00:04:42.970 There's a Japanese high-speed train. 00:04:42.970 --> 00:04:46.290 They've had high-speed trains for a long time. 00:04:46.290 --> 00:04:48.130 We don't really have any. 00:04:48.130 --> 00:04:50.100 There are a bunch of them in Europe. 00:04:50.100 --> 00:04:51.530 We've decided not to. 00:04:54.540 --> 00:04:56.850 Oh, that's the thing we saw earlier. 00:04:56.850 --> 00:05:00.570 That's that diagram for France that's shorthand for, 00:05:00.570 --> 00:05:03.000 and they decided to use a lot of nuclear power-- 00:05:03.000 --> 00:05:05.580 more than anybody else decided to do. 00:05:05.580 --> 00:05:07.860 Everybody had those technologies. 00:05:07.860 --> 00:05:09.720 Certainly, everybody had bicycles and people 00:05:09.720 --> 00:05:11.340 can build trains. 00:05:11.340 --> 00:05:14.010 Different countries made different choices. 00:05:14.010 --> 00:05:20.250 And just to gild the lily, that's the Paris subway system. 00:05:20.250 --> 00:05:22.710 I have a sister who lives in Munich, who occasionally 00:05:22.710 --> 00:05:26.460 visits us and talks about our third-world subway system. 00:05:26.460 --> 00:05:28.920 And I point out, that's really an insult 00:05:28.920 --> 00:05:32.670 to places like Mexico City that have pretty decent subway 00:05:32.670 --> 00:05:34.110 systems. 00:05:34.110 --> 00:05:36.270 But Paris and London-- 00:05:36.270 --> 00:05:38.490 I couldn't find a great London picture-- have clearly 00:05:38.490 --> 00:05:42.960 made much greater investments in that energy using technology 00:05:42.960 --> 00:05:45.060 than we have. 00:05:45.060 --> 00:05:48.170 That's a trolley in Munich. 00:05:48.170 --> 00:05:49.170 There are a lot of them. 00:05:49.170 --> 00:05:49.830 They're fast. 00:05:49.830 --> 00:05:51.360 They're efficient. 00:05:51.360 --> 00:05:53.490 They're everywhere. 00:05:53.490 --> 00:05:56.820 That is a Swiss railway schedule. 00:05:56.820 --> 00:06:00.870 That's, how do you get from the Zurich airport 00:06:00.870 --> 00:06:04.530 to Davos, where the World Economic Forum is held? 00:06:04.530 --> 00:06:06.870 And I've actually done this. 00:06:06.870 --> 00:06:10.650 Those trains arrive on the minute. 00:06:10.650 --> 00:06:13.740 There is no, gee, I wonder when the train's coming. 00:06:13.740 --> 00:06:16.020 The train is due at 3:41. 00:06:16.020 --> 00:06:20.280 The train is there at 3:41, unless something terrible 00:06:20.280 --> 00:06:21.210 has happened. 00:06:21.210 --> 00:06:23.640 So that's lots of very convenient ways 00:06:23.640 --> 00:06:26.280 you can get from the Zurich airport 00:06:26.280 --> 00:06:29.070 to this little village in the mountains in a couple 00:06:29.070 --> 00:06:33.160 hours very reliably. 00:06:33.160 --> 00:06:37.260 There's equivalent operations elsewhere in Europe, 00:06:37.260 --> 00:06:39.210 but the Swiss really stand out in terms 00:06:39.210 --> 00:06:46.140 of having invested in a very efficient, very reliable system 00:06:46.140 --> 00:06:49.770 of rail transport from little towns to little towns. 00:06:53.130 --> 00:06:56.100 And that's a solar field in Germany, 00:06:56.100 --> 00:06:58.170 where the Germans-- again, same technology 00:06:58.170 --> 00:07:03.430 we have-- have decided to put a lot of money into solar power. 00:07:03.430 --> 00:07:08.805 So it can't be the case that, once you 00:07:08.805 --> 00:07:10.830 know the level of income and once 00:07:10.830 --> 00:07:16.170 the available technologies, the choices follow. 00:07:16.170 --> 00:07:19.416 The processes must be different from that, 00:07:19.416 --> 00:07:22.540 from just rational calculation. 00:07:22.540 --> 00:07:24.960 At the very least, that story's incomplete, right? 00:07:24.960 --> 00:07:30.270 Because what determines what are the technologies available? 00:07:30.270 --> 00:07:32.640 What determines where the innovation dollars go? 00:07:32.640 --> 00:07:34.740 What determines whether you spend 00:07:34.740 --> 00:07:36.600 money making automobiles more efficient 00:07:36.600 --> 00:07:39.060 or nuclear power plants more efficient? 00:07:39.060 --> 00:07:43.920 So at the very least, the set of technologies available 00:07:43.920 --> 00:07:51.270 isn't given from heaven, and the choices aren't necessarily 00:07:51.270 --> 00:07:53.190 done on a spreadsheet-- because if they 00:07:53.190 --> 00:07:56.550 were done on a spreadsheet, lots of countries 00:07:56.550 --> 00:07:58.410 would make very similar choices. 00:07:58.410 --> 00:08:00.810 And in fact, we see lots of rich countries 00:08:00.810 --> 00:08:04.830 making very different choices. 00:08:04.830 --> 00:08:09.990 We'll talk a lot about this, but people make choices. 00:08:09.990 --> 00:08:13.170 It's very standard to say the market makes choices. 00:08:13.170 --> 00:08:16.770 People make choices in their individual demands. 00:08:16.770 --> 00:08:20.460 People make choices in collective decision making 00:08:20.460 --> 00:08:22.260 in politics. 00:08:22.260 --> 00:08:27.180 Those choices reflect a lot of things, as we'll talk about, 00:08:27.180 --> 00:08:30.870 but countries have different cultures, 00:08:30.870 --> 00:08:32.760 countries have different histories, 00:08:32.760 --> 00:08:35.429 countries make different choices. 00:08:35.429 --> 00:08:41.100 Besides those examples-- it's really a fascinating historical 00:08:41.100 --> 00:08:46.770 exercise-- if you look at Chinese technology circa 1500 00:08:46.770 --> 00:08:50.710 or 1600 AD, there are a number of areas-- 00:08:50.710 --> 00:08:54.180 including, in particular, ships, in which China 00:08:54.180 --> 00:08:56.700 is far ahead of the West-- 00:08:56.700 --> 00:08:59.880 China did not exploit those advantages, 00:08:59.880 --> 00:09:02.750 for a variety of reasons-- 00:09:02.750 --> 00:09:06.600 cultural reasons, primarily. 00:09:06.600 --> 00:09:12.150 If you think about the cultural decision in Denmark 00:09:12.150 --> 00:09:14.430 and in Holland to retain reliance 00:09:14.430 --> 00:09:18.290 on bicycles as a main urban transport, 00:09:18.290 --> 00:09:21.500 as opposed to, say, China where, years ago, you 00:09:21.500 --> 00:09:25.430 would go to a Chinese city-- you would see hordes of bicycles. 00:09:25.430 --> 00:09:29.030 Now you see hordes of cars not moving. 00:09:29.030 --> 00:09:31.850 But China switched. 00:09:31.850 --> 00:09:32.750 Denmark didn't. 00:09:32.750 --> 00:09:35.820 Holland didn't-- chose not to. 00:09:35.820 --> 00:09:37.370 So what I want to talk about is not 00:09:37.370 --> 00:09:42.320 that whole set of complexities affecting decisions, 00:09:42.320 --> 00:09:45.080 because it is interesting-- and we'll spend a lot of time 00:09:45.080 --> 00:09:46.620 on it as we go forward. 00:09:46.620 --> 00:09:48.470 I want to just talk about history 00:09:48.470 --> 00:09:51.710 and the influence of history-- 00:09:51.710 --> 00:09:57.860 three different versions of path dependence-- 00:09:57.860 --> 00:10:01.820 not all of them necessarily deserving the name, 00:10:01.820 --> 00:10:04.580 but all of them having to do with the way history matters. 00:10:07.110 --> 00:10:11.420 So the first one is many energy systems 00:10:11.420 --> 00:10:13.580 involved durable capital. 00:10:13.580 --> 00:10:16.415 In fact, durable capital's important. 00:10:16.415 --> 00:10:19.880 Now, just because I love to pull pictures, 00:10:19.880 --> 00:10:23.630 there's the railroad system. 00:10:23.630 --> 00:10:26.300 There are tracks laid down in the 1860s that 00:10:26.300 --> 00:10:28.600 are still in use. 00:10:28.600 --> 00:10:32.510 There's a really particularly ugly power plant. 00:10:32.510 --> 00:10:35.480 I think that one was built in the 1920s, 00:10:35.480 --> 00:10:39.140 but there is a power plant in Salem, Massachusetts 00:10:39.140 --> 00:10:44.130 that was first built in 1951, that is still operating. 00:10:44.130 --> 00:10:45.140 Half of it's shut down. 00:10:45.140 --> 00:10:47.240 Half of it's still operating. 00:10:47.240 --> 00:10:49.490 There are power lines. 00:10:49.490 --> 00:10:52.890 They last a long time. 00:10:52.890 --> 00:10:54.790 There's an oil refinery. 00:10:54.790 --> 00:10:59.100 They also last a pretty long time. 00:10:59.100 --> 00:11:02.400 There's the highway system. 00:11:02.400 --> 00:11:06.540 If it weren't for heavy trucks, I'm told, 00:11:06.540 --> 00:11:09.810 highways would be essentially eternal. 00:11:09.810 --> 00:11:12.970 Cars don't damage highways. 00:11:12.970 --> 00:11:13.728 Yeah? 00:11:13.728 --> 00:11:16.186 AUDIENCE: [INAUDIBLE] I'm from Chicago, and we did a lot of 00:11:16.186 --> 00:11:17.802 [INAUDIBLE] 00:11:17.802 --> 00:11:19.260 RICHARD SCHMALENSEE: That'll do it. 00:11:19.260 --> 00:11:19.500 Yeah. 00:11:19.500 --> 00:11:20.280 Time will do it. 00:11:20.280 --> 00:11:23.268 AUDIENCE: [INAUDIBLE] I went to California for the first time 00:11:23.268 --> 00:11:26.007 last year, and I was amazed at how their roads didn't 00:11:26.007 --> 00:11:26.840 have cracks in them. 00:11:26.840 --> 00:11:28.715 RICHARD SCHMALENSEE: Well, all you have to do 00:11:28.715 --> 00:11:31.350 is drive north from Illinois into Wisconsin. 00:11:31.350 --> 00:11:32.730 Wisconsin has decent roads. 00:11:32.730 --> 00:11:33.480 AUDIENCE: They do. 00:11:33.480 --> 00:11:34.350 RICHARD SCHMALENSEE: Part of this is money. 00:11:34.350 --> 00:11:35.736 AUDIENCE: Illinois is state-funded roads, 00:11:35.736 --> 00:11:37.590 but Wisconsin's federally funded roads. 00:11:37.590 --> 00:11:39.465 RICHARD SCHMALENSEE: Well, there's a mixture. 00:11:39.465 --> 00:11:41.880 Illinois is also, let us say, not the best-run state 00:11:41.880 --> 00:11:45.060 in the union, historically. 00:11:45.060 --> 00:11:47.340 For those of you who-- 00:11:47.340 --> 00:11:51.630 well, Rod Blagojevich is just the latest in a-- 00:11:51.630 --> 00:11:55.020 I'm from Illinois originally, so I sympathize, but oh 00:11:55.020 --> 00:11:56.370 my gosh-- what a history. 00:11:56.370 --> 00:11:59.010 Anyway, yes, roads get damaged by other-- 00:11:59.010 --> 00:12:00.960 things other than traffic. 00:12:00.960 --> 00:12:03.060 And among things that last a long time, 00:12:03.060 --> 00:12:06.630 that's Los Angeles, a snap of Los Angeles. 00:12:06.630 --> 00:12:12.400 When you lay out a city, that layout lasts a long time. 00:12:12.400 --> 00:12:16.020 And old buildings like that one, that was first built in 1916 00:12:16.020 --> 00:12:20.400 and is still going strong, last a long time. 00:12:20.400 --> 00:12:24.210 Now, I want to be a little clearer here about how 00:12:24.210 --> 00:12:31.410 durability matters, because the natural thing to say is, yeah, 00:12:31.410 --> 00:12:33.420 but those are all sunk costs. 00:12:33.420 --> 00:12:34.980 What's it matter? 00:12:34.980 --> 00:12:36.450 You build a railroad-- 00:12:36.450 --> 00:12:37.920 that money's gone. 00:12:37.920 --> 00:12:40.440 You can walk away, do something else. 00:12:40.440 --> 00:12:45.000 Well, the fact is you're usually replacing something that's 00:12:45.000 --> 00:12:48.277 capital-intensive with something else that's capital-intensive. 00:12:52.140 --> 00:12:56.980 And that fact by itself makes it difficult to change. 00:12:56.980 --> 00:13:02.640 I want to walk through a little bit of discounting, 00:13:02.640 --> 00:13:05.070 some of which you have probably seen. 00:13:05.070 --> 00:13:08.460 But let me just walk through a little bit of analysis 00:13:08.460 --> 00:13:11.070 of a hypothetical plant. 00:13:11.070 --> 00:13:13.650 We're thinking of replacing a coal-fired power plant. 00:13:13.650 --> 00:13:17.100 It produces Q units of output per year. 00:13:17.100 --> 00:13:20.220 The variable cost is v. It costs F to build. 00:13:20.220 --> 00:13:22.920 It'll last T years. 00:13:22.920 --> 00:13:25.290 The interest rate's r. 00:13:25.290 --> 00:13:27.580 We're considering replacing it with a new plant. 00:13:27.580 --> 00:13:29.730 Let's say it's more efficient-- 00:13:29.730 --> 00:13:34.050 has variable cost v star, cost F star to build. 00:13:34.050 --> 00:13:38.160 It'll also last 2 years, just to keep things simple. 00:13:38.160 --> 00:13:41.940 When do you scrap the old plant, replace it with the new plant? 00:13:41.940 --> 00:13:45.870 Well, I'm going to do-- 00:13:45.870 --> 00:13:47.670 clears that up? 00:13:47.670 --> 00:13:51.060 I'm going to use continuous compounding, because I'm lazy. 00:13:51.060 --> 00:13:54.720 How many of you have seen this before, continuous compounding? 00:13:54.720 --> 00:13:57.825 OK, most of you haven't, so I will walk slowly. 00:14:07.140 --> 00:14:12.330 If you say we're going to invest $1 today and we're going 00:14:12.330 --> 00:14:18.330 to ask, what happens in-- what's it worth in T years, 00:14:18.330 --> 00:14:25.590 and the interest's r, then, if you compounded annually, 00:14:25.590 --> 00:14:34.920 that becomes 1 plus r to the T. If you compounded N times 00:14:34.920 --> 00:14:37.830 a year-- well, the interest rate is divided by N-- 00:14:37.830 --> 00:14:41.655 1 plus r/N, now to the NT. 00:14:45.530 --> 00:14:47.220 So I'm compounding it more often. 00:14:47.220 --> 00:14:54.500 So if it's 12% a year, then $1 today is worth $1.12 in a year. 00:14:54.500 --> 00:15:00.410 If it's compounded monthly, then it's 00:15:00.410 --> 00:15:05.810 1.01 to the 12, which is going to be pretty close, but not 00:15:05.810 --> 00:15:07.340 quite the same as-- 00:15:07.340 --> 00:15:16.300 if you take the limit, this becomes e to the rT. 00:15:19.570 --> 00:15:26.350 And the slide-- well, the slide gets you there, 00:15:26.350 --> 00:15:31.700 asking now, what's $1 in T years worth today? 00:15:31.700 --> 00:15:45.640 So if you put $1 here at time T, that would be e to the minus rT 00:15:45.640 --> 00:15:46.220 today. 00:15:46.220 --> 00:15:48.970 I use continuous compounding a lot, 00:15:48.970 --> 00:15:52.900 because keeping track of whether things happen at the beginning 00:15:52.900 --> 00:15:56.530 or end of periods is tiresome. 00:15:56.530 --> 00:16:00.250 And continuous compounding works pretty well 00:16:00.250 --> 00:16:04.570 compared to daily, monthly, weekly 00:16:04.570 --> 00:16:07.150 for interest rates at matter. 00:16:07.150 --> 00:16:09.520 OK, is the problem clear? 00:16:09.520 --> 00:16:13.870 Is this limiting exercise reasonably clear? 00:16:13.870 --> 00:16:18.820 Questions, reactions, stunned silence? 00:16:18.820 --> 00:16:19.570 OK. 00:16:19.570 --> 00:16:20.653 Yeah, David? 00:16:20.653 --> 00:16:22.570 AUDIENCE: We've already built the first plant? 00:16:22.570 --> 00:16:23.440 RICHARD SCHMALENSEE: We built the first plant. 00:16:23.440 --> 00:16:23.860 We have it. 00:16:23.860 --> 00:16:24.580 AUDIENCE: OK, so that's-- 00:16:24.580 --> 00:16:25.000 RICHARD SCHMALENSEE: And you're going 00:16:25.000 --> 00:16:27.640 to tell me that F is irrelevant, and you're right. 00:16:30.250 --> 00:16:33.300 But comes now the second plant. 00:16:33.300 --> 00:16:35.020 So I want to ask-- 00:16:35.020 --> 00:16:41.085 I want to look at the capital cost per unit of output. 00:16:41.085 --> 00:16:42.460 You sometimes do this-- sometimes 00:16:42.460 --> 00:16:43.570 called levelized cost. 00:16:43.570 --> 00:16:48.500 But let me define it like this. 00:16:48.500 --> 00:16:54.220 Suppose the capital cost per unit of output is c star. 00:16:54.220 --> 00:16:55.070 What's that mean? 00:16:55.070 --> 00:16:57.880 Well, if that's the cost per unit of output, 00:16:57.880 --> 00:17:01.780 and I produce Q units per period, 00:17:01.780 --> 00:17:04.750 then, if I discount that capital cost, 00:17:04.750 --> 00:17:08.780 I ought to get the amount of the capital investment. 00:17:08.780 --> 00:17:10.600 So that's my definition of c star. 00:17:13.990 --> 00:17:16.300 There would be a c up here too, if I went through it, 00:17:16.300 --> 00:17:19.869 but that's-- as you say, that's sunk. 00:17:19.869 --> 00:17:20.650 That's sunk. 00:17:20.650 --> 00:17:24.280 But I'm interested now in-- 00:17:24.280 --> 00:17:26.619 because this isn't sunk. 00:17:26.619 --> 00:17:31.420 If I build my coal-fired-- my more efficient coal-fired power 00:17:31.420 --> 00:17:37.520 plant costs me F star, then that gives me-- 00:17:37.520 --> 00:17:41.620 what's the total cost per unit of output? 00:17:41.620 --> 00:17:44.320 I go out to the end of the plant's life. 00:17:44.320 --> 00:17:45.460 I look at output. 00:17:45.460 --> 00:17:46.930 I say, if that's the cost-- 00:17:46.930 --> 00:17:50.680 if c star is the cost per unit of output, 00:17:50.680 --> 00:17:53.140 then the discounted value of all those costs 00:17:53.140 --> 00:17:55.030 must equal the initial investment. 00:17:55.030 --> 00:17:58.300 That's this simple expression over there. 00:17:58.300 --> 00:18:01.030 Good so far? 00:18:01.030 --> 00:18:03.850 Comment, question? 00:18:03.850 --> 00:18:08.650 So people do this, for instance, if you think about how people 00:18:08.650 --> 00:18:11.440 calculate-- this is not the best calculation in the world, 00:18:11.440 --> 00:18:13.690 but if you ask, what's the cost of electricity 00:18:13.690 --> 00:18:17.110 from a solar plant, and you get a number 00:18:17.110 --> 00:18:23.890 in cents per kilowatt hour, this is how you do it, 00:18:23.890 --> 00:18:28.060 because the cost of solar power is almost exclusively 00:18:28.060 --> 00:18:30.380 an upfront fixed cost. 00:18:30.380 --> 00:18:33.190 And so you look at the life of the plant. 00:18:33.190 --> 00:18:35.740 You look at the output over that life. 00:18:35.740 --> 00:18:39.460 You ask, what cost per kilowatt hour discounts back 00:18:39.460 --> 00:18:42.430 to the original cost of the plant? 00:18:42.430 --> 00:18:44.860 And that gives you a cost per kilowatt hour 00:18:44.860 --> 00:18:47.710 from a solar power plant. 00:18:47.710 --> 00:18:49.360 OK? 00:18:49.360 --> 00:18:52.620 That's how you do it. 00:18:52.620 --> 00:18:53.970 You're very quiet. 00:18:53.970 --> 00:18:56.100 I don't think this is hard, but I'm not 00:18:56.100 --> 00:18:57.840 sure it's quite this easy. 00:18:57.840 --> 00:19:00.610 So are you all on board? 00:19:00.610 --> 00:19:01.110 Yeah? 00:19:01.110 --> 00:19:02.652 AUDIENCE: There's a lot of uncertainy 00:19:02.652 --> 00:19:04.180 in all those numbers, right? 00:19:04.180 --> 00:19:04.680 [INAUDIBLE] 00:19:04.680 --> 00:19:06.097 RICHARD SCHMALENSEE: In life, yes. 00:19:06.097 --> 00:19:08.970 In my story, no. 00:19:08.970 --> 00:19:13.650 This is economics, so we assume God-like powers. 00:19:13.650 --> 00:19:14.813 And I know these numbers. 00:19:14.813 --> 00:19:16.230 But in real life, of course, you'd 00:19:16.230 --> 00:19:18.840 want to react-- you'd want to deal with risk issues. 00:19:18.840 --> 00:19:21.510 But this is a straightforward replacement. 00:19:21.510 --> 00:19:25.830 So obviously, these are obvious properties, 00:19:25.830 --> 00:19:28.710 if you think about it for a minute. 00:19:28.710 --> 00:19:32.880 You basically solve this equation for c star, 00:19:32.880 --> 00:19:36.810 and you can see, as e to the minus rT vanishes, you 00:19:36.810 --> 00:19:39.030 get what you'd expect-- fixed cost divided 00:19:39.030 --> 00:19:42.670 by output made into a flow with the interest rate. 00:19:42.670 --> 00:19:49.140 Now, forget F. The decision to replace 00:19:49.140 --> 00:19:51.570 turns on this comparison-- 00:19:51.570 --> 00:19:55.650 turns on a comparison of the total cost of output 00:19:55.650 --> 00:19:58.080 from the new plant with the variable cost 00:19:58.080 --> 00:20:00.240 of output from the old plant. 00:20:00.240 --> 00:20:03.180 That's the point. 00:20:03.180 --> 00:20:05.190 The cost of the old plant sunk-- doesn't 00:20:05.190 --> 00:20:06.750 matter whether I got it as a gift 00:20:06.750 --> 00:20:09.120 or whether it was ridiculously expensive to build. 00:20:09.120 --> 00:20:11.100 It's gone. 00:20:11.100 --> 00:20:15.400 So what I care about is the cost of the old plant. 00:20:19.182 --> 00:20:23.040 If my plant is a coal-fired generating plant, 00:20:23.040 --> 00:20:24.450 an awful lot-- 00:20:24.450 --> 00:20:27.510 and I'm going to replace it with a new, cleaner, more efficient 00:20:27.510 --> 00:20:32.100 coal-fired generating plant, v star 00:20:32.100 --> 00:20:35.280 is going to be small relative to c star. 00:20:35.280 --> 00:20:38.550 v is going to be small relative to the original capital cost. 00:20:38.550 --> 00:20:39.510 That doesn't matter. 00:20:39.510 --> 00:20:43.500 Fuel is cheap for a coal-fired power plant. 00:20:43.500 --> 00:20:46.060 The plant's expensive. 00:20:46.060 --> 00:20:49.890 So to replace it, I got to say-- 00:20:49.890 --> 00:20:52.650 putting everything else to one side, 00:20:52.650 --> 00:20:56.850 I got to say, how does the running cost 00:20:56.850 --> 00:20:59.430 of this old inefficient dog compare 00:20:59.430 --> 00:21:03.510 with the total cost of this bright, shiny, new, high-tech, 00:21:03.510 --> 00:21:08.310 very expensive to build, cheap to operate plant? 00:21:08.310 --> 00:21:11.625 That gets you stuff lasting for a long time. 00:21:14.815 --> 00:21:15.315 Clear? 00:21:20.640 --> 00:21:25.080 If it's capital-intensive, you're comparing variable cost 00:21:25.080 --> 00:21:26.760 with total cost. 00:21:26.760 --> 00:21:30.850 And even if this is a much better plant, 00:21:30.850 --> 00:21:36.772 even if v star is much below v, and c star isn't bad-- 00:21:36.772 --> 00:21:38.730 first of all, it doesn't matter how it compares 00:21:38.730 --> 00:21:39.990 with the old capital cost. 00:21:39.990 --> 00:21:41.190 That's gone. 00:21:41.190 --> 00:21:44.460 What matters is, what is it-- how does total cost compare 00:21:44.460 --> 00:21:48.690 with the operating cost? 00:21:48.690 --> 00:21:52.870 So why exactly do we have vibrating PowerPoint? 00:21:52.870 --> 00:21:55.170 I don't know. 00:21:55.170 --> 00:21:57.270 If you think about other kinds of investments 00:21:57.270 --> 00:22:00.210 that might be made-- 00:22:00.210 --> 00:22:04.860 and we'll come to this when we talk about big systems-- 00:22:04.860 --> 00:22:09.630 institutions, know how, policies tend 00:22:09.630 --> 00:22:11.280 to come from upfront investments. 00:22:11.280 --> 00:22:12.968 They tend to last a long time. 00:22:12.968 --> 00:22:14.760 So if you think about switching technology, 00:22:14.760 --> 00:22:19.230 you will think about retraining a workforce, building a-- 00:22:19.230 --> 00:22:21.450 building technological knowledge, and so forth. 00:22:27.530 --> 00:22:29.580 Those are systems that are very long-lived, 00:22:29.580 --> 00:22:31.560 and are all up front. 00:22:31.560 --> 00:22:35.460 There's Salem Harbor right there, still-- 00:22:35.460 --> 00:22:42.160 it's coal-fired, and in fact, it actually burns a little oil. 00:22:42.160 --> 00:22:45.670 You don't see much oil, but it burns a little bit of oil 00:22:45.670 --> 00:22:46.780 and burns some gas. 00:22:46.780 --> 00:22:48.790 But it began as coal-fired-- 00:22:48.790 --> 00:22:50.980 1951-- still going. 00:22:50.980 --> 00:22:52.390 Part of it shut down in December. 00:22:52.390 --> 00:22:55.720 Part of it shuts down in 2014. 00:22:55.720 --> 00:22:58.520 You remember that graph from early on, 00:22:58.520 --> 00:23:00.880 but the point I make is all these coal 00:23:00.880 --> 00:23:04.490 fired plants are cheap to run. 00:23:04.490 --> 00:23:07.780 They may be very inefficient, but they are cheap to run. 00:23:07.780 --> 00:23:12.560 So they last a long time. 00:23:12.560 --> 00:23:16.060 They have a high ratio of capital cost per unit 00:23:16.060 --> 00:23:16.915 to variable cost. 00:23:21.370 --> 00:23:24.310 One kind of side point-- 00:23:24.310 --> 00:23:26.740 if you follow environmental debates, 00:23:26.740 --> 00:23:31.650 you will hear that lots of coal plants, 00:23:31.650 --> 00:23:34.010 in fact, will be shut down for environmental reasons 00:23:34.010 --> 00:23:37.640 in the next few years, and you will hear numbers 00:23:37.640 --> 00:23:42.740 on the incredible costs of replacing those plants. 00:23:42.740 --> 00:23:46.260 The point is those plants eventually die. 00:23:46.260 --> 00:23:48.725 Even Salem Harbor eventually dies. 00:23:54.860 --> 00:23:58.910 The economic cost of replacement isn't the upfront cost. 00:23:58.910 --> 00:24:01.220 It's the discounted cost of replacing 00:24:01.220 --> 00:24:05.220 now versus replacing in x years, whatever x happens to be. 00:24:05.220 --> 00:24:07.670 So when you hear the entire Midwest is going to go broke 00:24:07.670 --> 00:24:10.040 because they're replacing power plants, 00:24:10.040 --> 00:24:12.960 they're replacing them earlier. 00:24:12.960 --> 00:24:15.095 So you got to know how much earlier. 00:24:17.660 --> 00:24:22.750 So I have to tell you one other aspect of history that 00:24:22.750 --> 00:24:28.580 matters here, and that's the Clean Air Act. 00:24:28.580 --> 00:24:33.230 Those of you with phenomenal memories 00:24:33.230 --> 00:24:40.520 will recall that, when we talked about the American coal, rather 00:24:40.520 --> 00:24:44.480 than Saudi oil discussion, we talked about the Clean Air Act. 00:24:44.480 --> 00:24:47.690 And I mentioned that the architecture 00:24:47.690 --> 00:24:50.390 has states responsible for meeting air quality 00:24:50.390 --> 00:24:55.850 standards and EPA responsible for setting standards for 00:24:55.850 --> 00:24:59.440 new electric generating plants. 00:24:59.440 --> 00:25:04.190 So EPA starts up in here, tightens standards in there, 00:25:04.190 --> 00:25:09.330 tightens them later on. 00:25:09.330 --> 00:25:13.197 But there are no standards for these plants, 00:25:13.197 --> 00:25:14.655 for all those plants built earlier. 00:25:17.610 --> 00:25:20.520 Complying with EPA standards is expensive-- 00:25:20.520 --> 00:25:22.770 depends on the plant, depends on the standard, 00:25:22.770 --> 00:25:25.540 but it adds to cost. 00:25:25.540 --> 00:25:29.580 So now you're replacing an old plant that may be filthy, 00:25:29.580 --> 00:25:34.320 dirty with a new plant that has to meet new source performance 00:25:34.320 --> 00:25:35.341 standards. 00:25:38.050 --> 00:25:40.300 You can see why the law did it-- 00:25:40.300 --> 00:25:44.950 why Congress did that, because it's cheaper 00:25:44.950 --> 00:25:50.470 normally to design a new plant to be clean 00:25:50.470 --> 00:25:54.160 than to clean up an old plant. 00:25:54.160 --> 00:25:58.120 We'll talk about reasons later on, but that sort of makes 00:25:58.120 --> 00:26:00.170 sense. 00:26:00.170 --> 00:26:04.660 But what it means is building a new plant 00:26:04.660 --> 00:26:07.420 is more expensive than maintaining an old-- even 00:26:07.420 --> 00:26:10.000 more expensive than maintaining an old plant. 00:26:10.000 --> 00:26:15.010 So one reason all these old dogs are still alive 00:26:15.010 --> 00:26:17.170 is because they would have to be replaced 00:26:17.170 --> 00:26:21.790 by plants that would be cleaner, more efficient, and more 00:26:21.790 --> 00:26:22.480 expensive. 00:26:25.890 --> 00:26:28.500 There is an interesting question that I won't bore you with-- 00:26:28.500 --> 00:26:33.340 is what can you do to an old plant by way of repowering, 00:26:33.340 --> 00:26:35.710 changing the equipment, and modernizing it before it 00:26:35.710 --> 00:26:40.130 becomes a new plant, for purposes of the law-- 00:26:40.130 --> 00:26:42.910 a question that has no answer, that I once 00:26:42.910 --> 00:26:47.620 spent four months of my life trying to write a rule on. 00:26:47.620 --> 00:26:49.900 OK, is this all clear-- 00:26:49.900 --> 00:26:50.800 semi-clear? 00:26:50.800 --> 00:26:52.630 Questions, comments? 00:26:52.630 --> 00:26:53.130 Yeah? 00:26:53.130 --> 00:26:54.630 AUDIENCE: I just wanted to ask you-- 00:26:54.630 --> 00:27:03.434 so given [INAUDIBLE] forgetting about any environmental 00:27:03.434 --> 00:27:08.636 standards that probably will add to that or whatever EPA rules 00:27:08.636 --> 00:27:12.003 you may have, wouldn't you expect the environemtal cost 00:27:12.003 --> 00:27:14.470 at some point to start increasing, just because-- 00:27:14.470 --> 00:27:15.553 RICHARD SCHMALENSEE: Yeah. 00:27:15.553 --> 00:27:18.210 AUDIENCE: --your machines are so old that you can't even 00:27:18.210 --> 00:27:20.630 find replacements for what you have and stuff like that? 00:27:20.630 --> 00:27:21.340 RICHARD SCHMALENSEE: And that's when, 00:27:21.340 --> 00:27:23.830 putting environmental considerations aside-- 00:27:23.830 --> 00:27:26.680 and you would also expect the performance of new plants, 00:27:26.680 --> 00:27:29.740 in terms of both capital and variable cost, to get better. 00:27:29.740 --> 00:27:32.230 At some point, those curves cross, and you replace. 00:27:32.230 --> 00:27:35.470 Or I gave it a finite lifetime. 00:27:35.470 --> 00:27:38.470 At some point, the thing just can't be maintained, 00:27:38.470 --> 00:27:40.480 one way or another-- either, for some reason, 00:27:40.480 --> 00:27:46.270 it falls apart at your T, or the costs just keep rising 00:27:46.270 --> 00:27:48.862 and you say, this is an old dog. 00:27:48.862 --> 00:27:49.820 There are other issues. 00:27:49.820 --> 00:27:53.140 One reason Salem Harbor is still going is, 00:27:53.140 --> 00:27:56.200 for stability reasons, they need to have power injected 00:27:56.200 --> 00:28:00.580 into the system around there, and there's 00:28:00.580 --> 00:28:02.300 no obvious other way to do it. 00:28:02.300 --> 00:28:04.780 In fact, they're going to, I think, do some repowering 00:28:04.780 --> 00:28:06.040 and do some gas generation. 00:28:06.040 --> 00:28:08.050 But anyway, yeah, there comes a point-- 00:28:08.050 --> 00:28:09.830 I just had the cost constant. 00:28:09.830 --> 00:28:13.430 What happens in life is that variable cost rises. 00:28:13.430 --> 00:28:13.930 Yeah? 00:28:13.930 --> 00:28:16.210 AUDIENCE: How does the licensing work with this? 00:28:16.210 --> 00:28:19.150 Because I know at some power plants 00:28:19.150 --> 00:28:21.460 you have to be licensed to run, so couldn't you 00:28:21.460 --> 00:28:23.535 say they have a 60-year license, and you're 00:28:23.535 --> 00:28:25.577 pretty sure they're going to replace that anyway. 00:28:25.577 --> 00:28:27.460 [INAUDIBLE] 00:28:27.460 --> 00:28:30.670 RICHARD SCHMALENSEE: Nuclear plants in the US 00:28:30.670 --> 00:28:31.630 were originally given-- 00:28:31.630 --> 00:28:33.940 I think it was 40-year licenses. 00:28:33.940 --> 00:28:37.660 And they have the ability to apply for extensions. 00:28:37.660 --> 00:28:42.010 I don't believe any other power plants 00:28:42.010 --> 00:28:45.410 have licenses of that sort. 00:28:45.410 --> 00:28:49.210 The reason for the nuclear license is-- 00:28:49.210 --> 00:28:54.310 I think is that the federal government limits 00:28:54.310 --> 00:28:56.530 the liability of nuclear plant owners, 00:28:56.530 --> 00:29:00.280 and in exchange for that limit of liability, 00:29:00.280 --> 00:29:03.610 wants to make sure the plants are OK to go on. 00:29:03.610 --> 00:29:06.130 But they can apply for extensions 00:29:06.130 --> 00:29:09.550 for another 20 years. 00:29:09.550 --> 00:29:11.570 And at that point, they're done. 00:29:11.570 --> 00:29:14.860 And that's an engineering estimate at the start 00:29:14.860 --> 00:29:16.753 that we're pretty sure it'll go 40. 00:29:16.753 --> 00:29:18.170 We don't know if it can go longer. 00:29:18.170 --> 00:29:20.920 Let's take another look at it at 40. 00:29:20.920 --> 00:29:23.410 Coal-fired plants don't have this sort of buildup 00:29:23.410 --> 00:29:25.000 of radiation problem. 00:29:25.000 --> 00:29:27.010 Other plants don't, so I don't think 00:29:27.010 --> 00:29:31.780 they tend to have time-limited licenses. 00:29:31.780 --> 00:29:35.740 I think some private hydroelectric facilities 00:29:35.740 --> 00:29:40.690 had state limited licenses, but it can happen. 00:29:40.690 --> 00:29:42.840 Yeah. 00:29:42.840 --> 00:29:46.410 Anything else at all? 00:29:46.410 --> 00:29:48.360 OK. 00:29:48.360 --> 00:29:50.980 Let me talk about a second way. 00:29:50.980 --> 00:29:54.030 So here's the first way that history matters. 00:29:54.030 --> 00:29:59.160 History can matter just because of capital intensity. 00:29:59.160 --> 00:30:01.830 Your replacement technology's capital-intensive. 00:30:01.830 --> 00:30:03.510 You're comparing it with a technology 00:30:03.510 --> 00:30:06.330 that has low variable costs. 00:30:06.330 --> 00:30:11.440 A second way is a little more subtle, 00:30:11.440 --> 00:30:15.190 and it has to do with the political process. 00:30:15.190 --> 00:30:19.510 And that is, just as a matter of fact-- 00:30:19.510 --> 00:30:21.870 I don't know there's an elegant theory here, 00:30:21.870 --> 00:30:26.610 but as a matter of fact, political systems 00:30:26.610 --> 00:30:30.570 don't make big policy changes easily. 00:30:30.570 --> 00:30:32.310 They do, but they don't make them easily, 00:30:32.310 --> 00:30:35.460 because they can be very disruptive. 00:30:35.460 --> 00:30:40.980 So once you've gone down a particular policy path, 00:30:40.980 --> 00:30:43.170 it becomes-- 00:30:43.170 --> 00:30:44.820 the system tends to freeze. 00:30:44.820 --> 00:30:48.660 The US system in particular is prone to that, 00:30:48.660 --> 00:30:51.360 because if you think about-- 00:30:51.360 --> 00:30:53.090 it's more true now than it used to be-- 00:30:53.090 --> 00:30:55.410 you think about the need to have 60 votes in the Senate 00:30:55.410 --> 00:31:00.630 to get anything done at all, a big change 00:31:00.630 --> 00:31:04.230 is hard to get 60 votes for in the absence of a crisis. 00:31:04.230 --> 00:31:06.120 Little changes-- this and that-- sure. 00:31:06.120 --> 00:31:07.770 Fine tuning-- sure. 00:31:07.770 --> 00:31:13.170 But big structural changes the system just doesn't want to do. 00:31:13.170 --> 00:31:15.480 Think about the Clean Air Act. 00:31:15.480 --> 00:31:18.310 I've described a couple of issues with it. 00:31:18.310 --> 00:31:21.450 One is the way standards are set-- the one I just described, 00:31:21.450 --> 00:31:23.430 the new source performance standards. 00:31:23.430 --> 00:31:25.740 These problems have been known since the act was 00:31:25.740 --> 00:31:27.250 passed in 1970. 00:31:27.250 --> 00:31:28.710 They've been discussed. 00:31:28.710 --> 00:31:31.140 Remedies have been proposed. 00:31:31.140 --> 00:31:34.560 That architecture hasn't moved. 00:31:34.560 --> 00:31:35.060 Why? 00:31:35.060 --> 00:31:39.080 Because you've got a bureaucracy in place, 00:31:39.080 --> 00:31:41.180 you've got states relating to it, 00:31:41.180 --> 00:31:44.060 you've got a whole machinery going, 00:31:44.060 --> 00:31:48.410 and you have uncertainty as to how alternatives would work. 00:31:48.410 --> 00:31:53.030 So absent a crisis, that's not going to change. 00:31:53.030 --> 00:31:57.600 Agricultural price supports-- why do we still do this? 00:31:57.600 --> 00:32:00.230 Why does the European Union spend so much money 00:32:00.230 --> 00:32:01.880 supporting European farmers? 00:32:01.880 --> 00:32:04.340 Well, a policy's in place. 00:32:04.340 --> 00:32:08.210 Their interest's in place, their votes. 00:32:08.210 --> 00:32:10.700 It's a reinforcing system-- 00:32:10.700 --> 00:32:14.730 and because there's a fear of radical change. 00:32:14.730 --> 00:32:17.750 So every so often we debate our agricultural program, 00:32:17.750 --> 00:32:20.990 which benefits 1%, 2%, 3% of the population-- 00:32:20.990 --> 00:32:24.260 maybe as much as 5%, depending on how you do your arithmetic. 00:32:24.260 --> 00:32:25.740 And we spend a lot of money on it. 00:32:25.740 --> 00:32:29.420 Similarly, we subsidize oil drilling. 00:32:29.420 --> 00:32:34.550 We treat oil companies and oil exploration in a tax-- 00:32:34.550 --> 00:32:38.330 in a favorable way from the point of view of taxes. 00:32:38.330 --> 00:32:39.300 Why do we do that? 00:32:39.300 --> 00:32:42.700 Well, there are a whole set of reasons, 00:32:42.700 --> 00:32:47.200 but the fact is it's not only that we don't abolish it-- 00:32:47.200 --> 00:32:50.220 we also don't make it more generous. 00:32:50.220 --> 00:32:53.500 We all talk about more US oil drilling, 00:32:53.500 --> 00:32:56.080 but the tax system as it affects oil 00:32:56.080 --> 00:32:59.620 has been in place since, I think, the early '50s. 00:32:59.620 --> 00:33:02.457 Big policy changes just don't get made. 00:33:02.457 --> 00:33:04.540 We pass energy bills every few years, but that one 00:33:04.540 --> 00:33:05.590 doesn't get touched much. 00:33:09.250 --> 00:33:13.150 Oh yeah, and you will recall the pre-World War 1 London 00:33:13.150 --> 00:33:16.990 electricity system from last time-- crazy system-- 00:33:16.990 --> 00:33:18.280 crazy, crazy, crazy. 00:33:18.280 --> 00:33:20.270 It really took close to a crisis. 00:33:20.270 --> 00:33:24.760 It took World War 1 to cause them to say, look, we know 00:33:24.760 --> 00:33:26.570 local control is important. 00:33:26.570 --> 00:33:30.190 We know it's dangerous to fundamentally restructure 00:33:30.190 --> 00:33:33.250 the electricity system, but come on. 00:33:33.250 --> 00:33:34.540 And they did it. 00:33:34.540 --> 00:33:38.540 They, in fact, have done that twice. 00:33:38.540 --> 00:33:41.450 Actually, I think they did that in the '20s, but in any case, 00:33:41.450 --> 00:33:44.560 that was just the nutty system that persisted. 00:33:44.560 --> 00:33:46.210 It persisted. 00:33:46.210 --> 00:33:49.330 Big policy changes don't get made. 00:33:49.330 --> 00:33:53.080 We assigned the piece on health care reform 00:33:53.080 --> 00:33:55.127 to illustrate this-- 00:33:55.127 --> 00:33:56.710 the New Yorker piece, which I hope you 00:33:56.710 --> 00:33:58.877 read, because it was very readable, and interesting, 00:33:58.877 --> 00:34:00.850 and relevant to the current health care 00:34:00.850 --> 00:34:04.420 debate-- if not particularly to this course. 00:34:04.420 --> 00:34:06.610 I will refresh your recollection. 00:34:06.610 --> 00:34:10.989 He argues that everybody, when contemplating health 00:34:10.989 --> 00:34:14.830 care reform, has a tendency to take a blank piece of paper 00:34:14.830 --> 00:34:18.949 and write down some sort of ideal system. 00:34:18.949 --> 00:34:23.199 But if you look at health care, just like you look at energy, 00:34:23.199 --> 00:34:26.260 different countries made different choices. 00:34:26.260 --> 00:34:29.199 Yeah, everybody but the US among the rich countries 00:34:29.199 --> 00:34:33.670 has universal health care, but no two systems are the same. 00:34:33.670 --> 00:34:37.420 No two systems are the same. 00:34:37.420 --> 00:34:40.510 I thought the English example, the UK example 00:34:40.510 --> 00:34:41.590 was the most interesting. 00:34:41.590 --> 00:34:45.730 They moved to complete government-- 00:34:45.730 --> 00:34:47.139 government-run, government-owned, 00:34:47.139 --> 00:34:53.139 government-funded system in 1945 without an eye being batted. 00:34:53.139 --> 00:34:55.540 And the reason was they had to take over the health care 00:34:55.540 --> 00:34:56.889 system during World War 2. 00:34:56.889 --> 00:34:58.480 They had massive movements of people. 00:34:58.480 --> 00:35:00.610 They had massive casualties. 00:35:00.610 --> 00:35:02.985 The government took over a lot of things in this country, 00:35:02.985 --> 00:35:04.943 but it didn't take over the health care system. 00:35:04.943 --> 00:35:06.760 We didn't have that set of problems. 00:35:06.760 --> 00:35:09.070 It took over automobile companies, and steel companies, 00:35:09.070 --> 00:35:11.570 and aircraft companies, but it didn't take over health care. 00:35:11.570 --> 00:35:14.120 It did in the UK. 00:35:14.120 --> 00:35:16.030 So after the war, when they decided, 00:35:16.030 --> 00:35:20.230 gee, wouldn't it be good to have universal health care, 00:35:20.230 --> 00:35:23.140 well, they basically already had it. 00:35:23.140 --> 00:35:24.960 So they just needed to write it up 00:35:24.960 --> 00:35:27.460 and say, yeah, the government's been running the health care 00:35:27.460 --> 00:35:30.100 system since 1939. 00:35:30.100 --> 00:35:32.200 It's now 1945. 00:35:32.200 --> 00:35:34.870 I guess we'll just keep this system-- 00:35:34.870 --> 00:35:36.600 sounded like a huge policy change. 00:35:36.600 --> 00:35:38.350 They moved to the National Health Service, 00:35:38.350 --> 00:35:44.590 they moved to universal health care, but they already had it-- 00:35:44.590 --> 00:35:47.830 tiny policy change, because-- 00:35:47.830 --> 00:35:55.270 you going to make a huge change in 1945? 00:35:55.270 --> 00:35:57.370 France was defeated. 00:35:57.370 --> 00:36:02.170 They didn't have this kind of government-run system in 1945. 00:36:02.170 --> 00:36:06.220 They look back to the way they provided health care 00:36:06.220 --> 00:36:08.650 before World War 2. 00:36:08.650 --> 00:36:11.020 That was something they could use as a base 00:36:11.020 --> 00:36:13.458 and make changes from to make it universal, 00:36:13.458 --> 00:36:14.500 and that's what they did. 00:36:14.500 --> 00:36:16.720 They're a set of regional funds. 00:36:16.720 --> 00:36:19.480 The Swiss system is interesting. 00:36:19.480 --> 00:36:21.710 Switzerland had only private insurance. 00:36:21.710 --> 00:36:23.890 You go buy insurer. 00:36:23.890 --> 00:36:25.570 So when the Swiss decided they were 00:36:25.570 --> 00:36:29.890 going to have universal health care, what did they do? 00:36:29.890 --> 00:36:33.160 They said, OK, A, everybody has to buy insurance, and B, 00:36:33.160 --> 00:36:36.670 we'll subsidize insurance for poor people-- 00:36:36.670 --> 00:36:40.150 tiny change-- no big argument about, 00:36:40.150 --> 00:36:42.280 let's make a huge change, let's do a single payer 00:36:42.280 --> 00:36:44.493 system like the British, let's use regional funds 00:36:44.493 --> 00:36:45.160 like the French. 00:36:45.160 --> 00:36:45.660 No. 00:36:45.660 --> 00:36:47.920 Why would you do that? 00:36:47.920 --> 00:36:51.190 The system tends to move incrementally. 00:36:51.190 --> 00:36:54.100 The policy system moves incrementally, 00:36:54.100 --> 00:36:56.620 because big policy changes tend to have 00:36:56.620 --> 00:36:57.645 unforeseen consequences. 00:36:57.645 --> 00:36:59.020 And if you can get where you want 00:36:59.020 --> 00:37:04.740 to go without running that risk, policy systems do-- 00:37:04.740 --> 00:37:06.570 hence, history has consequences. 00:37:06.570 --> 00:37:09.030 The US case, you must-- 00:37:09.030 --> 00:37:10.440 well, maybe you don't know this. 00:37:10.440 --> 00:37:13.730 Why do most Americans get health care through their employers? 00:37:13.730 --> 00:37:15.660 They didn't in Switzerland. 00:37:15.660 --> 00:37:17.760 They certainly didn't in the UK-- 00:37:17.760 --> 00:37:21.180 French case only if you had a large employer. 00:37:21.180 --> 00:37:24.800 This was because, during World War 2, 00:37:24.800 --> 00:37:28.450 among other things, the government fixed wages. 00:37:28.450 --> 00:37:30.880 The economy was running quite hot. 00:37:30.880 --> 00:37:33.430 There was a lot of demand for labor, 00:37:33.430 --> 00:37:36.340 and they were afraid that, if we didn't control wages, 00:37:36.340 --> 00:37:38.630 we'd have inflation. 00:37:38.630 --> 00:37:40.830 We had inflation, but really rapid inflation-- 00:37:40.830 --> 00:37:43.150 so during World War 2, they controlled wages. 00:37:43.150 --> 00:37:45.460 So employers competed for workers 00:37:45.460 --> 00:37:49.960 by subsidizing health insurance, which was a way 00:37:49.960 --> 00:37:52.660 to get around wage controls. 00:37:52.660 --> 00:37:55.493 I don't know the political story about why that subsidy wasn't 00:37:55.493 --> 00:37:57.160 counted as income, but it wasn't counted 00:37:57.160 --> 00:37:59.260 as income for tax purposes. 00:37:59.260 --> 00:38:01.510 Therefore, we have employer-provided health 00:38:01.510 --> 00:38:05.800 insurance in this country, because 00:38:05.800 --> 00:38:08.170 of that accident of history. 00:38:08.170 --> 00:38:12.250 Therefore, when you do health care reform in Massachusetts, 00:38:12.250 --> 00:38:16.360 as in 2009-- and for that matter, in the federal system-- 00:38:16.360 --> 00:38:20.080 we built on employer-provided health care. 00:38:20.080 --> 00:38:21.790 Is that an obviously good thing to do? 00:38:21.790 --> 00:38:21.940 No. 00:38:21.940 --> 00:38:23.380 Most people think it's a mistake. 00:38:23.380 --> 00:38:26.380 Most economists think it's a mistake, that having that tax 00:38:26.380 --> 00:38:29.080 deduction distorts decisions. 00:38:29.080 --> 00:38:30.820 It makes it cheaper for employers 00:38:30.820 --> 00:38:33.670 to give you income in the form of health insurance 00:38:33.670 --> 00:38:35.530 and to give you income in other ways. 00:38:35.530 --> 00:38:37.690 That's distorting the compensation decision. 00:38:37.690 --> 00:38:39.268 But we have it. 00:38:39.268 --> 00:38:40.435 We've had it since the '40s. 00:38:43.110 --> 00:38:47.280 Changing it would be disruptive, so the system 00:38:47.280 --> 00:38:49.120 doesn't change it. 00:38:49.120 --> 00:38:58.050 So this is the second reason why history casts a long shadow. 00:38:58.050 --> 00:38:59.550 The nice illustrations for health 00:38:59.550 --> 00:39:02.250 care, but the point is the same. 00:39:02.250 --> 00:39:09.490 Once you set up a policy regime now, there are costs of change, 00:39:09.490 --> 00:39:11.610 so you will tend-- the system will tend, 00:39:11.610 --> 00:39:14.660 except in the case of crisis, when-- 00:39:14.660 --> 00:39:17.480 God knows-- but except in the case of crisis, 00:39:17.480 --> 00:39:20.060 the system will tend to use a prior regime 00:39:20.060 --> 00:39:21.663 as a point of departure. 00:39:24.500 --> 00:39:29.580 Health care reform in the US was controversial enough. 00:39:29.580 --> 00:39:32.570 Can you imagine what would have happened if the Obama 00:39:32.570 --> 00:39:34.500 administration or anybody else had said, 00:39:34.500 --> 00:39:36.560 OK, we're not going to get health insurance 00:39:36.560 --> 00:39:38.550 through your employer anymore-- 00:39:38.550 --> 00:39:40.517 we're going to move to the UK system, 00:39:40.517 --> 00:39:42.350 where the government owns all the hospitals, 00:39:42.350 --> 00:39:44.270 and hires all the doctors and nurses, 00:39:44.270 --> 00:39:45.578 and manages health care? 00:39:45.578 --> 00:39:46.370 We know it'll work. 00:39:46.370 --> 00:39:48.590 They do it in the UK. 00:39:48.590 --> 00:39:52.020 Could that have gotten four votes? 00:39:52.020 --> 00:39:54.000 Could a move to employer-provided health 00:39:54.000 --> 00:39:57.090 insurance have gotten three votes in parliament in the UK? 00:39:57.090 --> 00:39:58.440 No. 00:39:58.440 --> 00:40:01.660 It would have been disruptive in either case. 00:40:01.660 --> 00:40:03.160 So the system doesn't do it. 00:40:03.160 --> 00:40:07.800 So once you go down a road, it becomes 00:40:07.800 --> 00:40:13.650 hard to change the policy regime that supports it. 00:40:13.650 --> 00:40:17.640 Third way-- history matters. 00:40:17.640 --> 00:40:21.000 And this is the most interesting, in a way. 00:40:21.000 --> 00:40:25.975 Those were fun, but this is the-- 00:40:25.975 --> 00:40:28.770 what's his name-- Unruh article. 00:40:28.770 --> 00:40:32.760 He's a little bit hysterical, and I'll 00:40:32.760 --> 00:40:35.400 try to make some points about that-- 00:40:35.400 --> 00:40:40.930 and Mahoney, if you can possibly read sociological theory. 00:40:40.930 --> 00:40:45.270 The argument is that, in a lot of cases, what really makes 00:40:45.270 --> 00:40:49.680 path dependence is choices made that are somewhat 00:40:49.680 --> 00:40:54.360 arbitrary that then puts you on a path that 00:40:54.360 --> 00:40:59.550 involves positive interactions that keep you on the path. 00:40:59.550 --> 00:41:02.940 Health insurance isn't quite a great illustration of that. 00:41:02.940 --> 00:41:04.260 It's not bad. 00:41:04.260 --> 00:41:09.360 So you have this chance event in the US that puts you 00:41:09.360 --> 00:41:15.030 on a path of employer-provided health insurance, that then 00:41:15.030 --> 00:41:18.750 leads to a set of institutions being created around it, 00:41:18.750 --> 00:41:22.770 tax code created around it, benefits systems created 00:41:22.770 --> 00:41:25.020 around it. 00:41:25.020 --> 00:41:27.210 Even forgetting about the fact that you wouldn't 00:41:27.210 --> 00:41:29.910 want to risk the policy change in Washington, 00:41:29.910 --> 00:41:33.850 it would be difficult for the rest of the system to change. 00:41:33.850 --> 00:41:38.670 So Mahoney talks about not inevitable choices that then 00:41:38.670 --> 00:41:42.900 make later choices inevitable. 00:41:42.900 --> 00:41:46.410 A classic example of this kind of interaction 00:41:46.410 --> 00:41:48.780 is the keyboards I assume most of us use, 00:41:48.780 --> 00:41:51.480 the QWERTY keyboards. 00:41:51.480 --> 00:41:53.310 This design was originally chosen 00:41:53.310 --> 00:41:56.730 because, for mechanical typewriters, 00:41:56.730 --> 00:41:59.520 you wanted to keep keys that were frequently used together 00:41:59.520 --> 00:42:04.050 far apart, because otherwise, they jam as they came forward. 00:42:04.050 --> 00:42:07.437 Has anybody ever typed on a mechanical typewriter here? 00:42:07.437 --> 00:42:09.270 OK, a few-- all right, this is good for you. 00:42:09.270 --> 00:42:09.770 Yeah. 00:42:12.750 --> 00:42:17.910 So that was a perfectly good reason to do it. 00:42:17.910 --> 00:42:25.140 Some say it is inferior to other keyboard designs, 00:42:25.140 --> 00:42:30.000 but the fact is, even if it wasn't-- 00:42:30.000 --> 00:42:32.460 and the evidence, as I see it, on this is mixed. 00:42:32.460 --> 00:42:33.720 There's been a debate. 00:42:33.720 --> 00:42:37.320 Even if it wasn't, think about what happens. 00:42:37.320 --> 00:42:41.640 You begin to have people who learn to type this way. 00:42:41.640 --> 00:42:45.000 You begin to have typing teachers who learn to teach you 00:42:45.000 --> 00:42:47.610 how to type this way. 00:42:47.610 --> 00:42:50.880 When the computer came, when PCs came, 00:42:50.880 --> 00:42:53.670 people said, ah, this is the end of this design. 00:42:53.670 --> 00:42:55.718 People can get more efficient keyboards. 00:42:55.718 --> 00:42:57.510 We don't all have to use the same keyboard. 00:42:57.510 --> 00:42:58.530 You can buy a-- 00:42:58.530 --> 00:43:00.970 used to be able to buy a Dvorak keyboard for those who 00:43:00.970 --> 00:43:02.220 thought it was more efficient. 00:43:02.220 --> 00:43:04.470 You can teach yourself to type this more efficient way 00:43:04.470 --> 00:43:06.360 and plug it in. 00:43:06.360 --> 00:43:08.580 Well, as far as I know, almost nobody 00:43:08.580 --> 00:43:14.320 did that, because it apparently isn't more efficient. 00:43:14.320 --> 00:43:19.260 So we still use this old design, because once you've 00:43:19.260 --> 00:43:21.870 got the educational process, once you've 00:43:21.870 --> 00:43:24.720 got the standardization process, if there isn't 00:43:24.720 --> 00:43:27.760 a compelling reason to change, there 00:43:27.760 --> 00:43:29.530 are reasons not to change. 00:43:29.530 --> 00:43:31.390 You'd have to redo the instruction books, 00:43:31.390 --> 00:43:34.130 you'd have to reteach the teachers, et cetera, et cetera. 00:43:34.130 --> 00:43:37.060 So if you're going to teach people to type, 00:43:37.060 --> 00:43:39.130 it makes sense to do it one way, and this 00:43:39.130 --> 00:43:42.070 is a perfectly good way. 00:43:42.070 --> 00:43:47.600 An example I think is better is Swiss watch making. 00:43:47.600 --> 00:43:49.880 Why is watchmaking a Swiss thing? 00:43:54.320 --> 00:43:58.400 The historical story you get is Swiss farmers 00:43:58.400 --> 00:44:02.090 had craft ability, and in the wintertime in Switzerland, 00:44:02.090 --> 00:44:04.410 you don't get out much in the fields. 00:44:04.410 --> 00:44:06.830 So there was a market for stuff you could do 00:44:06.830 --> 00:44:08.540 in the house in the wintertime. 00:44:11.050 --> 00:44:14.090 And then, over time, you begin to get this-- 00:44:16.790 --> 00:44:20.150 you develop expertise in design. 00:44:20.150 --> 00:44:22.700 You develop expertise in construction. 00:44:22.700 --> 00:44:25.430 You start to build educational institutions 00:44:25.430 --> 00:44:26.850 to train a workforce. 00:44:26.850 --> 00:44:28.400 You start up a new watch company. 00:44:28.400 --> 00:44:30.260 You can draw on skilled labor. 00:44:30.260 --> 00:44:32.450 All of a sudden, you have computation for people. 00:44:32.450 --> 00:44:34.220 You have design computation. 00:44:34.220 --> 00:44:37.190 And the Swiss are tough to dislodge-- 00:44:37.190 --> 00:44:40.910 training, distribution channels, design expertise. 00:44:40.910 --> 00:44:46.260 But the key thing is the initial location was pretty arbitrary. 00:44:46.260 --> 00:44:50.140 Danish farmers are inside in the winter too. 00:44:50.140 --> 00:44:51.660 It's cold. 00:44:51.660 --> 00:44:54.240 Norwegian farmers are inside during the winter-- 00:44:54.240 --> 00:44:56.370 Swedish farmers-- it's cold a lot of places, 00:44:56.370 --> 00:44:57.600 and you don't get out. 00:44:57.600 --> 00:45:00.420 New Hampshire is cold in the winter. 00:45:00.420 --> 00:45:01.530 Why Switzerland? 00:45:01.530 --> 00:45:07.080 Well, no really good story for why Switzerland starts, 00:45:07.080 --> 00:45:10.215 but pretty good-- pretty easy to see why Switzerland persists. 00:45:12.730 --> 00:45:15.280 There all these economies, all these interactions 00:45:15.280 --> 00:45:20.560 that, once you're going down the road, give you an advantage. 00:45:20.560 --> 00:45:22.720 You say, OK, Danish farmers-- that's fabulous. 00:45:22.720 --> 00:45:24.095 They're inside during the winter. 00:45:24.095 --> 00:45:27.950 I'm going to do watchmaking in Denmark. 00:45:27.950 --> 00:45:31.520 Where are the skilled workers that I can get in Switzerland? 00:45:31.520 --> 00:45:34.370 Where the access to distribution channels? 00:45:34.370 --> 00:45:37.250 Where is the foot traffic of watch buyers 00:45:37.250 --> 00:45:38.960 in downtown Zurich? 00:45:38.960 --> 00:45:42.110 Where are the design people I can get in Switzerland? 00:45:42.110 --> 00:45:42.725 It's hard. 00:45:45.290 --> 00:45:48.320 Starting green field without all that expertise, 00:45:48.320 --> 00:45:52.070 and distribution, and everything is difficult. 00:45:52.070 --> 00:45:54.400 At day 1, it wouldn't have been hard-- 00:45:54.400 --> 00:45:55.400 hard for the Swiss. 00:45:55.400 --> 00:45:56.900 Would have been hard for the Danish. 00:45:56.900 --> 00:45:59.930 But Swiss did it, and once they'd done it, 00:45:59.930 --> 00:46:01.550 you get path dependence. 00:46:01.550 --> 00:46:04.670 They become hard to dislodge. 00:46:04.670 --> 00:46:11.120 So there are a lot of examples of this in energy, 00:46:11.120 --> 00:46:14.390 where institutions and physical systems-- 00:46:14.390 --> 00:46:17.660 institutions include things like training, and education, 00:46:17.660 --> 00:46:19.605 and policy, and all of these things-- 00:46:19.605 --> 00:46:20.105 interact. 00:46:23.000 --> 00:46:24.800 Unruh has a good set of examples, 00:46:24.800 --> 00:46:28.220 but I will say he's a little hysterical. 00:46:28.220 --> 00:46:33.370 Once the transformer was invented, it was over for DC. 00:46:33.370 --> 00:46:35.530 Once you could transmit alternating 00:46:35.530 --> 00:46:38.260 current at high voltage without losses-- 00:46:38.260 --> 00:46:40.000 with low losses over long distance, 00:46:40.000 --> 00:46:41.710 DC was not in the game. 00:46:41.710 --> 00:46:43.900 There are also a variety of other reasons. 00:46:43.900 --> 00:46:45.670 And the notion that electric cars 00:46:45.670 --> 00:46:48.010 were viable, rather than gasoline cars-- 00:46:48.010 --> 00:46:49.180 not so much. 00:46:49.180 --> 00:46:52.390 They used lead acid batteries and-- 00:46:52.390 --> 00:46:55.990 anyway, he's got good stories, but if you think 00:46:55.990 --> 00:46:59.150 about gasoline automobiles-- 00:46:59.150 --> 00:47:07.480 this is his picture, and he misses a little bit here. 00:47:07.480 --> 00:47:10.600 He talks about positive feedbacks 00:47:10.600 --> 00:47:16.990 among demand, the gasoline tax, and the road network. 00:47:16.990 --> 00:47:20.170 And that's true, but if you think 00:47:20.170 --> 00:47:25.570 about the automobile in the US, that's only part of the story. 00:47:25.570 --> 00:47:29.880 You build up manufacturing expertise in Detroit. 00:47:29.880 --> 00:47:31.900 Think 1900 and walk forward. 00:47:31.900 --> 00:47:35.320 You build up all the things that made Switzerland a good place 00:47:35.320 --> 00:47:38.080 to make watches made Detroit a good place to make automobiles. 00:47:38.080 --> 00:47:39.050 It could have been St. Louis. 00:47:39.050 --> 00:47:40.175 It could have been Chicago. 00:47:40.175 --> 00:47:42.160 It was Detroit. 00:47:42.160 --> 00:47:45.400 You began to get an accumulation of skilled workers. 00:47:45.400 --> 00:47:46.570 You got design. 00:47:46.570 --> 00:47:47.710 You got capital. 00:47:47.710 --> 00:47:49.820 You got a legal structure. 00:47:49.820 --> 00:47:53.200 So all of a sudden, you had Detroit being very good 00:47:53.200 --> 00:47:55.820 at making automobiles. 00:47:55.820 --> 00:48:00.690 It was a policy decision in a number of states, 00:48:00.690 --> 00:48:03.330 and then at the federal level that a road network 00:48:03.330 --> 00:48:04.930 would be a good thing to do. 00:48:04.930 --> 00:48:09.360 So you began to develop a set of interests there too. 00:48:09.360 --> 00:48:11.910 You have jobs here making cars. 00:48:11.910 --> 00:48:14.310 You've got jobs making roads. 00:48:14.310 --> 00:48:16.380 You've got companies that pave roads. 00:48:16.380 --> 00:48:18.930 You've got companies that build cars. 00:48:18.930 --> 00:48:21.690 You've got institutional structures. 00:48:21.690 --> 00:48:27.630 Then you have a dynamic that says, through the gasoline tax, 00:48:27.630 --> 00:48:28.800 we will-- 00:48:28.800 --> 00:48:31.990 the more we drive, the more gasoline tax we get, 00:48:31.990 --> 00:48:34.110 the more we put into building the road network, 00:48:34.110 --> 00:48:37.980 the better the road network, the better it is to drive. 00:48:37.980 --> 00:48:40.470 And within all of this-- 00:48:40.470 --> 00:48:44.640 notice acculturation and adaptive preferences-- 00:48:44.640 --> 00:48:46.620 we learn that we can't live without a car. 00:48:49.930 --> 00:48:51.700 People in the 19th century, it turned out, 00:48:51.700 --> 00:48:55.120 had perfectly nice, happy lives and never owned a car. 00:48:55.120 --> 00:48:56.830 So it is possible for human beings 00:48:56.830 --> 00:49:01.360 to live in cities, perfectly, happily without owning cars. 00:49:01.360 --> 00:49:05.320 But we don't believe it, unless you're from New York. 00:49:05.320 --> 00:49:07.240 Most Americans just can't imagine it. 00:49:07.240 --> 00:49:10.660 This is acculturation and adaptive preferences. 00:49:10.660 --> 00:49:13.750 We're with the things so long it becomes 00:49:13.750 --> 00:49:17.950 part of how we think the good life is defined, 00:49:17.950 --> 00:49:20.620 what we think you need. 00:49:20.620 --> 00:49:23.920 The institution over here, the government, 00:49:23.920 --> 00:49:26.440 is built-- you've got the Department of Motor Vehicles, 00:49:26.440 --> 00:49:29.020 for heaven's sakes, in every state with lots of jobs. 00:49:29.020 --> 00:49:34.780 You've got a lot of people invested in this system. 00:49:34.780 --> 00:49:37.660 The more you drive, the more you get tax collection, the more 00:49:37.660 --> 00:49:40.180 you build the road network, the more driving, 00:49:40.180 --> 00:49:41.800 the more you need more roads. 00:49:41.800 --> 00:49:42.790 The system feeds. 00:49:47.830 --> 00:49:51.340 You also have people invested in doing 00:49:51.340 --> 00:49:55.610 R&D. You have people invested in education and training. 00:49:55.610 --> 00:49:57.580 So there's a whole network of people 00:49:57.580 --> 00:50:01.120 who train auto repair people. 00:50:01.120 --> 00:50:05.260 Society of Automotive Engineers gets built. 00:50:05.260 --> 00:50:08.365 You have a lot of jobs in this system-- 00:50:10.880 --> 00:50:14.810 not necessarily easy to move, as Detroit has found. 00:50:14.810 --> 00:50:17.030 When things go bad for automobiles, 00:50:17.030 --> 00:50:20.540 there's a lot of political howling. 00:50:20.540 --> 00:50:22.250 You have the road network. 00:50:22.250 --> 00:50:26.060 It lasts a long time, just as a physical object. 00:50:26.060 --> 00:50:27.350 And it's there for cars. 00:50:27.350 --> 00:50:31.280 So once you've built it, you naturally 00:50:31.280 --> 00:50:32.880 want to use automobiles. 00:50:32.880 --> 00:50:35.360 The culture adapts. 00:50:35.360 --> 00:50:36.600 How do we think of ourselves? 00:50:36.600 --> 00:50:38.570 We think of ourselves as people who 00:50:38.570 --> 00:50:40.100 drive in convertibles with the top 00:50:40.100 --> 00:50:42.050 down on beautiful country roads. 00:50:42.050 --> 00:50:44.398 No, but if you watch enough commercials-- 00:50:44.398 --> 00:50:46.190 particularly if you watch sporting events-- 00:50:46.190 --> 00:50:48.560 you see so many car commercials that involve people 00:50:48.560 --> 00:50:50.690 driving happily down the road. 00:50:50.690 --> 00:50:51.710 This is how you live. 00:50:51.710 --> 00:50:55.160 You also have a policy regime that comes 00:50:55.160 --> 00:50:57.800 to support the whole system. 00:50:57.800 --> 00:50:59.820 It's an important part of the economy. 00:50:59.820 --> 00:51:02.420 It's an important part of the country. 00:51:02.420 --> 00:51:06.710 The political system, absent some good reason, 00:51:06.710 --> 00:51:08.960 is not going to suddenly turn against the people who 00:51:08.960 --> 00:51:10.710 build roads, the people who drive cars, 00:51:10.710 --> 00:51:11.960 and the people who build cars. 00:51:14.720 --> 00:51:16.250 And again, the gas tax-- 00:51:20.210 --> 00:51:24.110 you say, well, when new technologies come along, 00:51:24.110 --> 00:51:26.390 of course, companies will switch. 00:51:26.390 --> 00:51:29.060 Well, one of the things-- and we'll come back to this when we 00:51:29.060 --> 00:51:31.340 talk about corporate decision making-- 00:51:31.340 --> 00:51:33.290 companies tend to focus-- 00:51:33.290 --> 00:51:37.130 as this blurb says, tend to focus on getting better 00:51:37.130 --> 00:51:39.620 at what they're good at. 00:51:39.620 --> 00:51:43.490 So rather than taking dangerous leaps-- 00:51:43.490 --> 00:51:46.850 so you remember Palm and the PalmPilot? 00:51:46.850 --> 00:51:48.050 That didn't go so well. 00:51:48.050 --> 00:51:50.750 Blackberry hasn't quite adjusted. 00:51:50.750 --> 00:51:54.500 General Motors took a long time to turn around. 00:51:54.500 --> 00:51:56.750 There are lots of other examples of companies 00:51:56.750 --> 00:51:59.030 that were really good at something 00:51:59.030 --> 00:52:01.650 and stuck with it too long. 00:52:01.650 --> 00:52:05.030 So you tend not to get the companies that are big players 00:52:05.030 --> 00:52:07.700 in a system like this-- any system like this-- 00:52:07.700 --> 00:52:10.190 to be the ones that drive the change. 00:52:10.190 --> 00:52:11.070 And then finally-- 00:52:11.070 --> 00:52:14.240 I love this phrase-- 00:52:14.240 --> 00:52:16.160 I'm not sure I understand it, but I love it-- 00:52:16.160 --> 00:52:20.390 historically derived subjective modeling of the issues. 00:52:20.390 --> 00:52:22.400 The point he makes-- 00:52:22.400 --> 00:52:29.780 Unruh makes is that, when you're embedded long enough 00:52:29.780 --> 00:52:32.420 in any system that has an impact on your life, 00:52:32.420 --> 00:52:35.960 it colors how you think about the world-- 00:52:35.960 --> 00:52:38.390 just does, necessarily does. 00:52:38.390 --> 00:52:42.800 In the 19th century, horses mattered. 00:52:42.800 --> 00:52:45.710 People would think about the world in a different way 00:52:45.710 --> 00:52:47.710 than they do when cars matter. 00:52:51.870 --> 00:52:57.110 A system that's-- in which you are deeply embedded will affect 00:52:57.110 --> 00:53:01.970 how you think, typically in a way that says more of that 00:53:01.970 --> 00:53:04.940 would be good-- not always, but it will affect it. 00:53:09.470 --> 00:53:15.080 Economics textbooks tend to assume, for simplicity, that we 00:53:15.080 --> 00:53:18.210 are born with our preferences. 00:53:18.210 --> 00:53:22.980 It's probably not a great shock to you to know-- 00:53:22.980 --> 00:53:26.720 and probably you've known all along-- that's just wrong. 00:53:26.720 --> 00:53:29.000 That's just wrong. 00:53:29.000 --> 00:53:31.280 Nobody's born with a preference for video games. 00:53:31.280 --> 00:53:32.720 You learn video games. 00:53:32.720 --> 00:53:35.690 Nobody's born with a preference for automobiles. 00:53:35.690 --> 00:53:37.220 You learn about automobiles. 00:53:37.220 --> 00:53:39.740 Most of our preferences are learned. 00:53:39.740 --> 00:53:41.210 And the way we learn-- 00:53:41.210 --> 00:53:42.840 the automobile's a good example-- 00:53:42.840 --> 00:53:44.760 the way we learn depends on our environment, 00:53:44.760 --> 00:53:47.090 and if x is a big part of the environment-- 00:53:47.090 --> 00:53:49.340 trees, cars, horses, dogs-- 00:53:49.340 --> 00:53:53.130 it will affect your preferences. 00:53:53.130 --> 00:53:56.100 If you grew up with a dog in the house, you probably like dogs. 00:53:56.100 --> 00:54:00.690 If you didn't, you may be ambivalent-- 00:54:00.690 --> 00:54:03.420 not that you're born loving dogs are not loving dogs. 00:54:03.420 --> 00:54:05.160 The environment matters. 00:54:05.160 --> 00:54:09.450 So let me give you another example. 00:54:09.450 --> 00:54:10.650 There are a lot of examples. 00:54:10.650 --> 00:54:13.170 Electricity gives you an example. 00:54:13.170 --> 00:54:15.105 This is another one of his lovely pictures. 00:54:17.880 --> 00:54:20.310 You start doing electricity-- 00:54:20.310 --> 00:54:22.920 he says, well, you develop a regulatory system 00:54:22.920 --> 00:54:25.650 that makes investment in the electric power system 00:54:25.650 --> 00:54:27.930 profitable. 00:54:27.930 --> 00:54:31.350 When that's the case, you develop a supplier network 00:54:31.350 --> 00:54:34.350 that's selling equipment into that-- 00:54:34.350 --> 00:54:35.670 to be invested. 00:54:35.670 --> 00:54:38.040 As the grid network gets bigger, it 00:54:38.040 --> 00:54:41.400 becomes cheaper for you to use electricity. 00:54:41.400 --> 00:54:44.130 What's missing over here are the appliance makers 00:54:44.130 --> 00:54:48.090 that, once electricity's cheap, begin to produce refrigerators, 00:54:48.090 --> 00:54:49.680 and air conditioners, and all kinds 00:54:49.680 --> 00:54:54.120 of things that make electricity more valuable-- dot, dot, dot. 00:54:54.120 --> 00:54:55.560 You have equipment. 00:54:55.560 --> 00:54:57.150 You have education. 00:54:57.150 --> 00:54:58.950 You have lots of jobs tied up. 00:54:58.950 --> 00:55:02.490 You have lots of habits built around that system, 00:55:02.490 --> 00:55:07.380 and it becomes hard to change. 00:55:07.380 --> 00:55:12.420 Examples like this are pretty easy to multiply. 00:55:12.420 --> 00:55:17.340 If you think about air travel, the better the airports, 00:55:17.340 --> 00:55:18.900 the more attractive air travel is. 00:55:18.900 --> 00:55:21.940 The better the planes are, the more attractive air travel is. 00:55:21.940 --> 00:55:24.310 The greater the demand for air travel, 00:55:24.310 --> 00:55:26.070 the more you build airports, the more 00:55:26.070 --> 00:55:29.100 you invest in R&D in aircraft. 00:55:29.100 --> 00:55:33.300 We have a whole legislative structure around air travel. 00:55:33.300 --> 00:55:36.570 We have a lot of training and education around air travel. 00:55:36.570 --> 00:55:38.310 There's all of these investments that 00:55:38.310 --> 00:55:42.270 make the system more attractive, that make it grow, et cetera. 00:55:42.270 --> 00:55:46.810 That's positive interactions along the path. 00:55:46.810 --> 00:55:50.353 Now, this is not evil stuff. 00:55:50.353 --> 00:55:52.020 The fact that you build better airports, 00:55:52.020 --> 00:55:54.570 and you have better airplanes, and the system works better-- 00:55:54.570 --> 00:55:59.220 this is not bad, but this means, once you go down that path-- 00:55:59.220 --> 00:56:03.180 once you decide to do air, rather than high-speed rail-- 00:56:03.180 --> 00:56:06.990 the air path is self-reinforcing. 00:56:06.990 --> 00:56:09.660 It'd be a little awkward in this country, with our distances, 00:56:09.660 --> 00:56:11.100 but you could imagine [INAUDIBLE] 00:56:11.100 --> 00:56:12.558 having decided another path-- we're 00:56:12.558 --> 00:56:15.660 going to do high-speed rail. 00:56:15.660 --> 00:56:18.180 We get manufacturers of those trains. 00:56:18.180 --> 00:56:19.855 We've got people laying the road bed. 00:56:19.855 --> 00:56:21.480 We've got people building the stations. 00:56:21.480 --> 00:56:24.330 It becomes more attractive to travel by high-speed train. 00:56:24.330 --> 00:56:26.070 And somebody says, why would you use 00:56:26.070 --> 00:56:28.260 airplanes along the East Coast? 00:56:31.650 --> 00:56:34.260 Natural gas for heating-- you get the same kind of story. 00:56:34.260 --> 00:56:35.070 You drill wells. 00:56:35.070 --> 00:56:36.030 You build pipelines. 00:56:36.030 --> 00:56:36.660 You have laws. 00:56:36.660 --> 00:56:38.940 You have regulations. 00:56:38.940 --> 00:56:44.850 If you look at the US broadly, our paths tend to be more 00:56:44.850 --> 00:56:49.410 energy-intensive than the paths in other rich countries-- 00:56:49.410 --> 00:56:53.340 fact-- not good, not bad-- fact. 00:56:53.340 --> 00:56:57.820 But I will say there is a little problem with Unruh's tone. 00:56:57.820 --> 00:57:00.630 He calls this carbon lock-in. 00:57:00.630 --> 00:57:02.310 It's not carbon lock-in. 00:57:02.310 --> 00:57:04.770 It's path dependence. 00:57:04.770 --> 00:57:09.210 It wouldn't be easy to move away from bicycles in Amsterdam. 00:57:09.210 --> 00:57:11.130 The city's kind of built for them. 00:57:11.130 --> 00:57:12.660 The streets are laid out for them. 00:57:12.660 --> 00:57:14.040 There are lots of repair shops. 00:57:14.040 --> 00:57:14.790 They're available. 00:57:14.790 --> 00:57:16.650 Everybody knows how to use them. 00:57:16.650 --> 00:57:18.690 People can teach you how to ride a bike. 00:57:18.690 --> 00:57:22.050 All of those interactions are true for the bicycle system 00:57:22.050 --> 00:57:23.550 in Amsterdam. 00:57:23.550 --> 00:57:26.930 It would not be easy for them now. 00:57:26.930 --> 00:57:30.440 Where would you part, if you've been in that city? 00:57:30.440 --> 00:57:32.750 Not easy to move away-- 00:57:32.750 --> 00:57:35.030 Beijing was built for public transportation. 00:57:35.030 --> 00:57:38.490 If you visit Beijing, and if you're like me, 00:57:38.490 --> 00:57:40.490 the first thing you think of is, my God, 00:57:40.490 --> 00:57:41.780 there's no place to part. 00:57:41.780 --> 00:57:44.990 They have these huge apartment buildings close 00:57:44.990 --> 00:57:47.210 by each other-- very dense. 00:57:47.210 --> 00:57:50.520 If all those people had cars, where would you put them? 00:57:53.090 --> 00:57:55.160 Now, they are moving to the automobile path, 00:57:55.160 --> 00:57:56.240 with some pain. 00:57:56.240 --> 00:58:01.250 But once you go down that path, which is a low-carbon path-- 00:58:01.250 --> 00:58:03.170 becomes harder to move off. 00:58:03.170 --> 00:58:06.620 Similarly, London or Paris, cities 00:58:06.620 --> 00:58:13.000 that have subway networks, that carry a lot of people-- 00:58:13.000 --> 00:58:14.470 lot of people use the subway. 00:58:14.470 --> 00:58:18.070 It becomes politically important to make it work well. 00:58:18.070 --> 00:58:19.960 You begin to have a workforce invested 00:58:19.960 --> 00:58:21.460 in the subway, et cetera, et cetera. 00:58:21.460 --> 00:58:25.730 But you also have a housing stock that's dense. 00:58:25.730 --> 00:58:27.320 So if you say, well, we're not going 00:58:27.320 --> 00:58:32.390 to do mass transit anymore, you have a problem. 00:58:32.390 --> 00:58:33.740 Where are people going to park? 00:58:33.740 --> 00:58:38.660 The roads aren't designed to handle cars in quantity. 00:58:38.660 --> 00:58:40.640 You can't move off that path easily. 00:58:40.640 --> 00:58:42.290 That's a low-carbon path. 00:58:42.290 --> 00:58:44.570 Mass transit in Beijing is a low-carbon path. 00:58:44.570 --> 00:58:47.190 Bicycles in Amsterdam are a low-carbon path. 00:58:47.190 --> 00:58:50.060 So it's not just carbon lock-in. 00:58:50.060 --> 00:58:51.800 It's path dependence. 00:58:51.800 --> 00:58:54.650 It's, once you start going down a path-- 00:58:54.650 --> 00:58:57.830 in some systems, not every system you can think of-- 00:58:57.830 --> 00:59:02.520 but in some systems, parts of the system 00:59:02.520 --> 00:59:05.470 interact positively with other parts of the system, 00:59:05.470 --> 00:59:07.740 making it more and more attractive and harder 00:59:07.740 --> 00:59:10.390 and harder to leave. 00:59:10.390 --> 00:59:13.240 OK so far? 00:59:13.240 --> 00:59:16.050 So I hope I've convinced you that you can never 00:59:16.050 --> 00:59:22.560 change energy systems, because that's wrong. 00:59:22.560 --> 00:59:25.470 It has been done in lots-- lots of times. 00:59:25.470 --> 00:59:27.990 Clipper ships-- well, think about it. 00:59:27.990 --> 00:59:29.430 You had people making sales. 00:59:29.430 --> 00:59:30.630 You had designers. 00:59:30.630 --> 00:59:32.190 You had shipyards. 00:59:32.190 --> 00:59:35.280 And sailing technology you can trace over the years. 00:59:35.280 --> 00:59:38.460 Sailing technology got better and better and better, 00:59:38.460 --> 00:59:42.100 and carried more and more commerce faster and faster. 00:59:42.100 --> 00:59:43.620 So this was a path. 00:59:43.620 --> 00:59:47.940 The whole world was on that path. 00:59:47.940 --> 00:59:49.770 That's gas light. 00:59:49.770 --> 00:59:53.130 That's one of the last gas lights in London being lit. 00:59:53.130 --> 00:59:56.610 Streets used to be lit with gas. 00:59:56.610 --> 00:59:59.940 Homes were too, but particularly public areas 00:59:59.940 --> 01:00:02.250 were illuminated at night with gas lights. 01:00:02.250 --> 01:00:03.220 That's a path. 01:00:03.220 --> 01:00:05.850 You have people who learn how to learn the technology. 01:00:05.850 --> 01:00:07.188 You've invested in training. 01:00:07.188 --> 01:00:08.730 You've invested in the infrastructure 01:00:08.730 --> 01:00:10.470 to get the gas there-- 01:00:10.470 --> 01:00:11.280 whole thing works. 01:00:11.280 --> 01:00:14.310 People like light at night. 01:00:14.310 --> 01:00:17.940 That is a coal-powered locomotive 01:00:17.940 --> 01:00:20.430 drawing a passenger train. 01:00:20.430 --> 01:00:23.460 19th century-- locomotives got better and better. 01:00:23.460 --> 01:00:25.290 Railroads employed a lot of people-- 01:00:25.290 --> 01:00:27.330 lot of fixed capital there. 01:00:27.330 --> 01:00:29.910 Passenger trains went from city to city. 01:00:29.910 --> 01:00:32.700 I believe the locomotive speed record in the US 01:00:32.700 --> 01:00:34.680 was set in 1910. 01:00:34.680 --> 01:00:38.730 At any rate, you had a whole system built up-- 01:00:41.610 --> 01:00:42.510 coal-fired. 01:00:45.690 --> 01:00:49.350 Coal-fired passenger trains-- very important 01:00:49.350 --> 01:00:51.737 intercity system. 01:00:51.737 --> 01:00:53.070 And here's everybody's favorite. 01:00:53.070 --> 01:00:57.150 This is Los Angeles in 1909. 01:00:57.150 --> 01:01:01.350 Southern California had an extraordinarily extensive set 01:01:01.350 --> 01:01:02.700 of streetcars. 01:01:02.700 --> 01:01:03.780 You could go anywhere. 01:01:03.780 --> 01:01:05.843 If you've seen the movie Roger Rabbit, where 01:01:05.843 --> 01:01:08.010 the character says, who needs a car in Los Angeles-- 01:01:08.010 --> 01:01:13.200 you ride these-- again, all the system 01:01:13.200 --> 01:01:15.780 effects I talked about are there. 01:01:15.780 --> 01:01:17.640 You've got the owners of the system. 01:01:17.640 --> 01:01:19.520 You've got a large labor force. 01:01:19.520 --> 01:01:21.400 It requires technology. 01:01:21.400 --> 01:01:23.410 There's also a nice network effect. 01:01:23.410 --> 01:01:25.560 The more of these you build, the more attractive 01:01:25.560 --> 01:01:27.750 the whole system becomes. 01:01:27.750 --> 01:01:29.310 So one route-- not so great. 01:01:29.310 --> 01:01:30.270 Two routes-- better. 01:01:30.270 --> 01:01:32.190 Three routes-- even better, because you can 01:01:32.190 --> 01:01:33.540 go lots of different places. 01:01:33.540 --> 01:01:35.700 There are system effects. 01:01:35.700 --> 01:01:39.200 All of those systems went away. 01:01:39.200 --> 01:01:42.650 All of those systems, with all of the same things I was just 01:01:42.650 --> 01:01:45.080 talking about, went away. 01:01:45.080 --> 01:01:48.120 AUDIENCE: [INAUDIBLE] Do you happen 01:01:48.120 --> 01:01:52.540 to know how come it happened that LA really 01:01:52.540 --> 01:01:55.830 abandoned the whole care thing? 01:01:55.830 --> 01:01:58.762 Was it something specific or [INAUDIBLE] 01:01:58.762 --> 01:02:00.720 RICHARD SCHMALENSEE: I will tell you the story. 01:02:00.720 --> 01:02:05.890 The story has a couple of dimensions. 01:02:05.890 --> 01:02:09.000 Sometimes it's just something better comes along. 01:02:09.000 --> 01:02:15.300 Steam turned out, after a lot of R&D, to be better than sail. 01:02:15.300 --> 01:02:18.010 The first steam ships were lousy. 01:02:18.010 --> 01:02:19.740 There's a lovely museum in Munich 01:02:19.740 --> 01:02:21.900 that actually lays this out nicely. 01:02:21.900 --> 01:02:24.000 The first steam ships could basically 01:02:24.000 --> 01:02:27.540 carry enough coal to make it across the Atlantic 01:02:27.540 --> 01:02:29.640 without much cargo. 01:02:29.640 --> 01:02:31.950 They were very inefficient. 01:02:31.950 --> 01:02:33.780 And efficiency really matters-- 01:02:33.780 --> 01:02:36.510 how efficient that engine is, because it depends-- 01:02:36.510 --> 01:02:38.490 it determines how much fuel you have to carry. 01:02:38.490 --> 01:02:41.010 And if you have to carry a lot of fuel to go someplace, 01:02:41.010 --> 01:02:43.260 it's a lousy venture, even though it gets there 01:02:43.260 --> 01:02:46.110 much faster and more reliably. 01:02:46.110 --> 01:02:50.157 So it took a lot of R&D. Gas to electric lights-- 01:02:50.157 --> 01:02:51.240 that was a little quicker. 01:02:51.240 --> 01:02:54.690 The arc light turned out to be pretty good, and pretty early. 01:02:54.690 --> 01:02:57.900 Coal to diesel locomotives-- well, that took a while. 01:02:57.900 --> 01:03:00.690 The first diesel locomotives appeared in the '20s, 01:03:00.690 --> 01:03:04.830 and you really didn't have diesel phased out for long haul 01:03:04.830 --> 01:03:07.500 until the late '40s, '50s. 01:03:07.500 --> 01:03:08.970 You had some diesel. 01:03:08.970 --> 01:03:11.490 It took a while, but it happened. 01:03:11.490 --> 01:03:12.180 It happened. 01:03:18.370 --> 01:03:20.470 So sometimes it took changes in policy. 01:03:20.470 --> 01:03:22.330 To really wipe out the railroads, 01:03:22.330 --> 01:03:25.000 you needed the interstate highway system. 01:03:25.000 --> 01:03:27.880 To get nuclear power, you needed federal policy 01:03:27.880 --> 01:03:30.490 to limit the liability of nuclear plant owners. 01:03:30.490 --> 01:03:36.010 You needed environmental policy to really hit coal generation. 01:03:36.010 --> 01:03:39.790 What happened in Los Angeles-- 01:03:39.790 --> 01:03:42.490 there are two stories. 01:03:42.490 --> 01:03:44.770 The Los Angeles streetcar network 01:03:44.770 --> 01:03:49.030 was largely built by real estate developers, 01:03:49.030 --> 01:03:52.390 because it was a way of making outlying properties more 01:03:52.390 --> 01:03:53.990 valuable. 01:03:53.990 --> 01:03:56.290 So the argument I've heard is that Southern California 01:03:56.290 --> 01:03:59.470 got sprawl before it got automobiles, 01:03:59.470 --> 01:04:03.960 because the idea was, we can build streetcars out 01:04:03.960 --> 01:04:06.850 to there, so they can go from there downtown. 01:04:06.850 --> 01:04:08.770 That makes that land valuable. 01:04:08.770 --> 01:04:14.510 And the argument was that's how they built it. 01:04:14.510 --> 01:04:19.990 They overbuilt it, because this was a land speculation game. 01:04:19.990 --> 01:04:22.030 So they overbuilt it. 01:04:22.030 --> 01:04:27.900 And then you had companies that wanted to sell buses. 01:04:27.900 --> 01:04:32.550 GM wanted to sell buses, so GM lobbied to shut it down. 01:04:32.550 --> 01:04:34.320 GM bought pieces of it. 01:04:34.320 --> 01:04:36.570 Buses have advantages. 01:04:36.570 --> 01:04:39.030 So eventually, the trolleys got replaced by buses, 01:04:39.030 --> 01:04:42.090 which got replaced by cars. 01:04:42.090 --> 01:04:43.890 Again, you had ownership. 01:04:43.890 --> 01:04:45.300 You had a system. 01:04:45.300 --> 01:04:47.093 You had interactions of various kinds. 01:04:47.093 --> 01:04:48.510 The electric companies, of course, 01:04:48.510 --> 01:04:53.320 loved trolleys, because they're low during the daytime. 01:04:53.320 --> 01:04:55.320 One of the problems early electric companies had 01:04:55.320 --> 01:05:00.090 was all the demand was at night, so you used your assets only 01:05:00.090 --> 01:05:01.140 when it was dark. 01:05:01.140 --> 01:05:02.280 Trolleys were great. 01:05:02.280 --> 01:05:04.770 You could keep the whole system running all day long. 01:05:04.770 --> 01:05:06.392 So they loved them. 01:05:06.392 --> 01:05:08.100 The people who worked on them loved them. 01:05:08.100 --> 01:05:09.017 The owners loved them. 01:05:09.017 --> 01:05:11.760 They got beat out in part because the systems were 01:05:11.760 --> 01:05:14.910 overbuilt, in part because buses had 01:05:14.910 --> 01:05:21.660 certain operational advantages, in part for political reasons. 01:05:21.660 --> 01:05:23.910 It wasn't just Los Angeles. 01:05:23.910 --> 01:05:25.740 You could travel across a large part 01:05:25.740 --> 01:05:30.690 of the country on trolleys, and they went away everywhere-- 01:05:30.690 --> 01:05:31.550 mostly went away. 01:05:36.060 --> 01:05:38.610 Old-- "old." 01:05:38.610 --> 01:05:41.670 energy systems can be displaced-- 01:05:41.670 --> 01:05:44.370 sometimes R&D, something better comes along. 01:05:44.370 --> 01:05:46.110 Sometimes changes in policy. 01:05:46.110 --> 01:05:47.100 Sometimes both. 01:05:50.280 --> 01:05:52.950 One of the problems that, again, we 01:05:52.950 --> 01:05:56.400 will look at in a number of settings as we go forward 01:05:56.400 --> 01:05:59.800 is the so-called chicken egg problem. 01:05:59.800 --> 01:06:03.840 So suppose you wanted to say, gee, natural gas 01:06:03.840 --> 01:06:07.140 is a better transportation fuel than oil-- 01:06:07.140 --> 01:06:11.490 right now it's a lot cheaper, and we have a lot of it here-- 01:06:11.490 --> 01:06:14.070 well, I'm not going to buy a natural gas 01:06:14.070 --> 01:06:19.010 car unless I can fill it, unless there are places I can refuel. 01:06:19.010 --> 01:06:22.580 And I'm not going to have a natural gas filling station 01:06:22.580 --> 01:06:25.130 unless there are cars. 01:06:25.130 --> 01:06:26.380 That's a chicken egg problem. 01:06:29.470 --> 01:06:33.520 You need them both, and nobody's going to invest in one side 01:06:33.520 --> 01:06:36.430 unless there's some assurance that you're going to get 01:06:36.430 --> 01:06:38.200 investment in the other side. 01:06:38.200 --> 01:06:43.390 You don't build fancy airplanes for passenger 01:06:43.390 --> 01:06:45.520 travel unless there are airports, 01:06:45.520 --> 01:06:47.260 but why would you invest in an airport 01:06:47.260 --> 01:06:50.620 unless people are going to use it to travel on fancy planes? 01:06:50.620 --> 01:06:52.930 So most of these systems had chicken egg problems. 01:06:52.930 --> 01:06:55.953 Automobiles are pretty useless without good roads, 01:06:55.953 --> 01:06:57.370 but why would you build good roads 01:06:57.370 --> 01:06:59.590 unless there's going to be a lot of traffic? 01:06:59.590 --> 01:07:06.640 So most new systems faced this problem. 01:07:06.640 --> 01:07:10.958 Oh, yeah, you-- why did Edison invent the electric light? 01:07:10.958 --> 01:07:13.000 Well, Edison invented the electric light because, 01:07:13.000 --> 01:07:14.500 without a reliable electric light, 01:07:14.500 --> 01:07:17.320 what was the point of generating electricity? 01:07:17.320 --> 01:07:19.460 There'd be no market for it. 01:07:19.460 --> 01:07:22.150 So without a light, there's no market 01:07:22.150 --> 01:07:25.600 for electricity in the 1880s. 01:07:25.600 --> 01:07:27.995 So he needed to do both. 01:07:27.995 --> 01:07:30.370 If he was going to do something, he had to do the system. 01:07:30.370 --> 01:07:34.120 He had to have the generator and the bulb-- 01:07:34.120 --> 01:07:38.300 useless without each other, without the other. 01:07:38.300 --> 01:07:40.780 So a lot of these new systems face chicken egg problems, 01:07:40.780 --> 01:07:43.180 but one of the things we're going 01:07:43.180 --> 01:07:50.170 to talk about on and off is replacing energy systems, 01:07:50.170 --> 01:07:53.050 and how that might happen. 01:07:53.050 --> 01:07:55.000 You see some obstacles. 01:07:55.000 --> 01:07:57.160 This last slide is intended to point out to you 01:07:57.160 --> 01:07:59.470 that they are not forever. 01:07:59.470 --> 01:08:03.430 Energy systems will change. 01:08:03.430 --> 01:08:05.860 Coal-fired power plants will not be 01:08:05.860 --> 01:08:08.540 a major part of this country's energy system in 2100. 01:08:08.540 --> 01:08:12.430 You can probably bet good money on that-- 01:08:12.430 --> 01:08:14.500 nothing like the ones we have. 01:08:14.500 --> 01:08:17.457 But getting there will take a little work. 01:08:17.457 --> 01:08:19.540 And who knows-- maybe gasoline-powered automobiles 01:08:19.540 --> 01:08:20.439 won't either. 01:08:20.439 --> 01:08:24.729 OK, questions, comments, reactions, thoughts? 01:08:24.729 --> 01:08:25.710 You're on today. 01:08:25.710 --> 01:08:26.660 Yes? 01:08:26.660 --> 01:08:27.979 AUDIENCE: [INAUDIBLE] 01:08:27.979 --> 01:08:28.930 RICHARD SCHMALENSEE: You can't get an A just 01:08:28.930 --> 01:08:29.740 from participation. 01:08:29.740 --> 01:08:30.609 I just want you to know that. 01:08:30.609 --> 01:08:31.240 OK. 01:08:31.240 --> 01:08:32.149 It's all good, but-- 01:08:32.149 --> 01:08:32.649 OK. 01:08:32.649 --> 01:08:35.232 AUDIENCE: [INAUDIBLE] question regarding developing countries. 01:08:35.232 --> 01:08:37.053 RICHARD SCHMALENSEE: Yeah? 01:08:37.053 --> 01:08:38.470 AUDIENCE: One thing I was thinking 01:08:38.470 --> 01:08:42.250 while you were speaking is, at some point, you need to-- 01:08:42.250 --> 01:08:44.500 RICHARD SCHMALENSEE: They need to make choices and get 01:08:44.500 --> 01:08:45.250 on paths. 01:08:45.250 --> 01:08:46.840 AUDIENCE: Yeah, exactly. 01:08:46.840 --> 01:08:49.540 Because if you want, you can perpetuate everything back 01:08:49.540 --> 01:08:52.379 to cavemen and everything [INAUDIBLE] some sort 01:08:52.379 --> 01:08:54.670 of [INAUDIBLE] so it's quite hard to-- 01:08:54.670 --> 01:08:57.823 so I was just thinking it's for [INAUDIBLE] developing 01:08:57.823 --> 01:09:01.649 countries, One is when they make a choice-- 01:09:01.649 --> 01:09:04.330 having spoken about everything today, 01:09:04.330 --> 01:09:06.850 it's that very often I feel that there's 01:09:06.850 --> 01:09:10.210 this narrow-minded-- after what we spoke 01:09:10.210 --> 01:09:13.630 about the cost-benefit analysis [INAUDIBLE] in the sense 01:09:13.630 --> 01:09:16.744 that you might say, OK, a natural gas power 01:09:16.744 --> 01:09:19.660 plant is more expensive than a coal one, so let's go ahead 01:09:19.660 --> 01:09:22.180 and build a coal power plant [INAUDIBLE] 01:09:22.180 --> 01:09:23.500 build coal power plants. 01:09:23.500 --> 01:09:26.689 But you go down a path which you may not-- 01:09:26.689 --> 01:09:29.380 that you should definitely take into account-- 01:09:29.380 --> 01:09:31.149 you might build some sort of expertise 01:09:31.149 --> 01:09:33.575 in something and everything that we talk about. 01:09:33.575 --> 01:09:35.319 So that's one thing, and I'm not sure 01:09:35.319 --> 01:09:37.569 if it's done at all in developing countries. 01:09:37.569 --> 01:09:41.590 And the other thing is, in theory, given that you're-- 01:09:41.590 --> 01:09:46.810 if you think [INAUDIBLE] a blank check have no [INAUDIBLE] 01:09:46.810 --> 01:09:50.350 dependance, wouldn't it makes sense 01:09:50.350 --> 01:09:53.740 to see that whatever choice is made by these sort of countries 01:09:53.740 --> 01:09:59.980 is the most optimal, given what they've 01:09:59.980 --> 01:10:01.520 seen with other countries? 01:10:01.520 --> 01:10:04.510 So if I'm a country that has nothing-- 01:10:04.510 --> 01:10:08.230 I have no universal health care, so-- and I see all these 01:10:08.230 --> 01:10:10.850 models, I say, which one's the best, 01:10:10.850 --> 01:10:13.547 of course for my conditions, but wouldn't that-- 01:10:13.547 --> 01:10:14.630 RICHARD SCHMALENSEE: Yeah. 01:10:17.418 --> 01:10:18.460 You're making two points. 01:10:18.460 --> 01:10:20.980 Let me just respond briefly, and then I'll let everybody go. 01:10:24.820 --> 01:10:27.100 Sometimes you do have a clean slate. 01:10:27.100 --> 01:10:28.600 Sometimes you do have a clean slate. 01:10:33.870 --> 01:10:36.494 I remember a visit to Kyoto-- 01:10:36.494 --> 01:10:38.555 no, Nagoya. 01:10:38.555 --> 01:10:39.930 I was actually in the government, 01:10:39.930 --> 01:10:44.100 and I was being walked around by a local official. 01:10:44.100 --> 01:10:47.100 And I was commenting on how attractive 01:10:47.100 --> 01:10:51.150 the broad boulevards and squares they had were. 01:10:51.150 --> 01:10:56.190 And he looked at me and he said, well, there was nothing here, 01:10:56.190 --> 01:10:57.660 so we could design a city. 01:11:00.480 --> 01:11:05.200 You don't feel good when you've just walked into one of those. 01:11:05.200 --> 01:11:06.360 But sometimes that's true. 01:11:06.360 --> 01:11:08.280 The Cambodians, for instance-- 01:11:08.280 --> 01:11:10.980 you got imported oil to generate your electricity. 01:11:10.980 --> 01:11:12.720 You're burning wood. 01:11:12.720 --> 01:11:15.000 You have a relatively clean slate, 01:11:15.000 --> 01:11:20.220 and you can decide what to do, but the other side of the coin 01:11:20.220 --> 01:11:23.890 is you have a lot of really poor people. 01:11:23.890 --> 01:11:27.210 China's not building coal-fired power plants 01:11:27.210 --> 01:11:31.680 because they're not thinking about tomorrow at all. 01:11:31.680 --> 01:11:34.590 They're balancing tomorrow against today. 01:11:34.590 --> 01:11:37.630 They have hundreds of millions of wretchedly poor people 01:11:37.630 --> 01:11:38.130 still. 01:11:41.940 --> 01:11:46.200 Do I want to make my grandchildren better off, 01:11:46.200 --> 01:11:47.190 or do-- 01:11:47.190 --> 01:11:49.230 their grandchildren better off, or do I 01:11:49.230 --> 01:11:51.330 want to make them better off now? 01:11:51.330 --> 01:11:53.880 That's not an easy choice. 01:11:53.880 --> 01:11:58.710 And a responsible government will wrestle with it, 01:11:58.710 --> 01:12:00.360 but you don't always-- 01:12:00.360 --> 01:12:02.370 the answer isn't always to put yourself 01:12:02.370 --> 01:12:04.170 on the path that looks best in a century. 01:12:07.360 --> 01:12:09.360 They're trying to do the right thing, I believe, 01:12:09.360 --> 01:12:13.020 but figuring out what the right thing is is not trivial. 01:12:13.020 --> 01:12:14.550 Anybody else? 01:12:14.550 --> 01:12:16.100 OK.