1 00:00:00,040 --> 00:00:02,460 The following content is provided under a Creative 2 00:00:02,460 --> 00:00:03,960 Commons license. 3 00:00:03,960 --> 00:00:06,910 Your support will help MIT OpenCourseWare continue to 4 00:00:06,910 --> 00:00:10,660 offer high-quality educational resources for free. 5 00:00:10,660 --> 00:00:13,460 To make a donation or view additional materials from 6 00:00:13,460 --> 00:00:17,390 hundreds of MIT courses, visit MIT OpenCourseWare at 7 00:00:17,390 --> 00:00:18,656 ocw.mit.edu. 8 00:00:26,030 --> 00:00:29,340 PROFESSOR: Might as well get started. 9 00:00:29,340 --> 00:00:31,460 So this is all carrying on the conversation 10 00:00:31,460 --> 00:00:32,670 that we had last time. 11 00:00:32,670 --> 00:00:45,920 We talked about, possibly for very long-term historical 12 00:00:45,920 --> 00:00:49,560 reasons, some countries might have very different kinds of 13 00:00:49,560 --> 00:00:50,820 institutions from others. 14 00:00:50,820 --> 00:00:55,890 In particular, some might be much worse than others, at 15 00:00:55,890 --> 00:00:59,990 least using the standard measures. 16 00:00:59,990 --> 00:01:02,830 And then what do you do about it? 17 00:01:02,830 --> 00:01:07,846 So the first reaction is, sometimes, well, tough. 18 00:01:07,846 --> 00:01:12,670 If the story is that you had to have some lucky break 400 19 00:01:12,670 --> 00:01:16,700 years ago to get the institutions right, then it's 20 00:01:16,700 --> 00:01:18,820 not entirely clear that there's much you 21 00:01:18,820 --> 00:01:20,410 can do about it. 22 00:01:20,410 --> 00:01:23,500 That's the pessimistic view. 23 00:01:23,500 --> 00:01:32,060 Now, for the people who really worry about these issues, 24 00:01:32,060 --> 00:01:34,220 there's an interesting debate going on. 25 00:01:37,380 --> 00:01:43,500 It's really between these five people listed above, are on 26 00:01:43,500 --> 00:01:45,820 the two sides of the debate. 27 00:01:45,820 --> 00:01:51,780 One side of it is a Stanford professor called Paul Romer, 28 00:01:51,780 --> 00:01:56,770 who has come up with this idea of charter cities, like 29 00:01:56,770 --> 00:01:57,870 charter schools. 30 00:01:57,870 --> 00:02:03,170 So he wants, if you'd like, to import institutions. 31 00:02:03,170 --> 00:02:08,850 So his basic idea is, the US has good institutions, so a US 32 00:02:08,850 --> 00:02:14,900 firm, or a US NGO, would start-- 33 00:02:14,900 --> 00:02:19,610 if South Sudan wants to have a city, then they should just 34 00:02:19,610 --> 00:02:25,030 hire a US firm to set up the city for them, have the same 35 00:02:25,030 --> 00:02:27,020 guarantees that the US has. 36 00:02:27,020 --> 00:02:35,160 So it'll be a part of the US institutional frame. 37 00:02:40,390 --> 00:02:43,850 So it's not clear exactly how that works. 38 00:02:43,850 --> 00:02:50,960 Because for example, suppose you set up a court in that 39 00:02:50,960 --> 00:02:55,790 country, and it has the US laws, but the judge in the 40 00:02:55,790 --> 00:03:00,640 court decides not to implement those laws. 41 00:03:00,640 --> 00:03:02,720 Does the US invade? 42 00:03:02,720 --> 00:03:04,080 Who's making the commitment? 43 00:03:04,080 --> 00:03:07,810 So this is a set of real issues about how 44 00:03:07,810 --> 00:03:08,540 you get that to work. 45 00:03:08,540 --> 00:03:09,039 Yeah. 46 00:03:09,039 --> 00:03:11,534 AUDIENCE: Isn't there even more fundamental issues of 47 00:03:11,534 --> 00:03:14,029 what foreign country would be willing to 48 00:03:14,029 --> 00:03:16,025 sponsor a charter city? 49 00:03:16,025 --> 00:03:18,520 Because seemingly, all the advantages, even if you could 50 00:03:18,520 --> 00:03:21,181 find a country to host it, like a country that would be 51 00:03:21,181 --> 00:03:26,337 willing to give up the land, the outside sponsor, for lack 52 00:03:26,337 --> 00:03:29,696 of a better term-- at least the way you guys talk about it 53 00:03:29,696 --> 00:03:32,028 in the book-- they basically give it up as soon as the 54 00:03:32,028 --> 00:03:34,184 city's kind of successful. 55 00:03:34,184 --> 00:03:36,579 He alludes to Hong Kong being an example. 56 00:03:36,579 --> 00:03:40,370 But Hong Kong had very different circumstances, 57 00:03:40,370 --> 00:03:44,161 because Britain originally colonized it in the hopes of 58 00:03:44,161 --> 00:03:46,616 having it as its own, having it as a commercial center. 59 00:03:46,616 --> 00:03:49,562 And then they got the benefits for 100 years before they 60 00:03:49,562 --> 00:03:51,040 handed it over. 61 00:03:51,040 --> 00:03:53,960 PROFESSOR: And equally importantly, they handed it 62 00:03:53,960 --> 00:03:58,940 over under circumstances when they really 63 00:03:58,940 --> 00:04:01,400 didn't have any choice. 64 00:04:01,400 --> 00:04:12,770 So Britain is certainly in no position to, let's say, 65 00:04:12,770 --> 00:04:16,029 disagree with China about a city that's 66 00:04:16,029 --> 00:04:17,170 next doors to China. 67 00:04:17,170 --> 00:04:21,740 So there was no leverage is there whatsoever. 68 00:04:21,740 --> 00:04:22,019 Yeah. 69 00:04:22,019 --> 00:04:23,260 So there's several issues there. 70 00:04:23,260 --> 00:04:30,250 You've just brought up one, which is, how do I make the, 71 00:04:30,250 --> 00:04:32,110 to use your word, the sponsoring 72 00:04:32,110 --> 00:04:33,830 country, give up the city. 73 00:04:33,830 --> 00:04:35,570 Then there's the opposite problem. 74 00:04:35,570 --> 00:04:41,450 How do I make sure the city actually runs, and who's 75 00:04:41,450 --> 00:04:42,700 guaranteeing it? 76 00:04:49,580 --> 00:04:55,230 Suppose the judge who's supposed to enforce US laws 77 00:04:55,230 --> 00:05:06,020 in, I don't know, Benin, doesn't do it, who's liable to 78 00:05:06,020 --> 00:05:07,704 make it happen? 79 00:05:07,704 --> 00:05:11,113 AUDIENCE: But there's no one preventing those countries 80 00:05:11,113 --> 00:05:12,087 [INAUDIBLE] 81 00:05:12,087 --> 00:05:15,983 extracting institutions, they have no incentive to be 82 00:05:15,983 --> 00:05:17,931 [INAUDIBLE] either. 83 00:05:17,931 --> 00:05:19,181 PROFESSOR: Right. 84 00:05:22,290 --> 00:05:28,218 AUDIENCE: So in the same was as colonies where Europeans 85 00:05:28,218 --> 00:05:32,170 did not settle permanently [INAUDIBLE]. 86 00:05:32,170 --> 00:05:36,122 They set up [INAUDIBLE] institutions that would be 87 00:05:36,122 --> 00:05:37,590 [INAUDIBLE]. 88 00:05:37,590 --> 00:05:39,470 PROFESSOR: So you're right. 89 00:05:39,470 --> 00:05:41,740 There's the question of the demand side for it. 90 00:05:41,740 --> 00:05:45,600 Maybe there's no country that wants it. 91 00:05:45,600 --> 00:05:49,180 I think Paul's presumption is that there are countries that 92 00:05:49,180 --> 00:05:51,370 would like to have better institutions. 93 00:05:51,370 --> 00:05:53,930 They cannot domestically manufacture it. 94 00:05:53,930 --> 00:05:58,130 But if they just somehow imported institutions, and 95 00:05:58,130 --> 00:05:59,620 there was a [? guarantor ?] 96 00:05:59,620 --> 00:06:01,720 behind it, then it would work. 97 00:06:01,720 --> 00:06:07,940 So what you're saying is completely right. 98 00:06:07,940 --> 00:06:09,360 There needs to be a demand for it. 99 00:06:09,360 --> 00:06:10,720 If the country doesn't want it, it's 100 00:06:10,720 --> 00:06:11,580 never going to happen. 101 00:06:11,580 --> 00:06:15,310 But let's say the country even wants these institutions. 102 00:06:15,310 --> 00:06:17,880 Even in that case, there are several problems. 103 00:06:17,880 --> 00:06:20,230 One is, how do you actually run a charter city? 104 00:06:20,230 --> 00:06:22,490 I mean, who's in charge? 105 00:06:30,290 --> 00:06:36,420 Is the US prepared to back up its commitment 106 00:06:36,420 --> 00:06:37,890 to the charter city? 107 00:06:37,890 --> 00:06:40,530 If it says the US laws, does that mean the US is going to 108 00:06:40,530 --> 00:06:45,034 send in an army whenever the laws are violated? 109 00:06:45,034 --> 00:06:47,026 AUDIENCE: Likewise, does it mean that if the country gets 110 00:06:47,026 --> 00:06:50,263 overrun by an outside invader and the charter city gets 111 00:06:50,263 --> 00:06:53,430 overrun, will the US extend its nuclear arms-- 112 00:06:53,430 --> 00:06:54,170 like, let's say-- 113 00:06:54,170 --> 00:06:54,330 PROFESSOR: Right. 114 00:06:54,330 --> 00:06:55,060 Exactly. 115 00:06:55,060 --> 00:06:56,880 So [? what is ?] commitment to it. 116 00:06:56,880 --> 00:06:57,706 AUDIENCE: Foreign policy. 117 00:06:57,706 --> 00:06:58,120 PROFESSOR: Absolutely. 118 00:06:58,120 --> 00:07:03,100 So all forms of commitment are an issue, obviously. 119 00:07:03,100 --> 00:07:07,420 It's not enough to say that we have US-style laws. 120 00:07:07,420 --> 00:07:09,680 In fact, if you look across the world, laws 121 00:07:09,680 --> 00:07:13,190 don't vary a lot. 122 00:07:13,190 --> 00:07:16,640 Legal systems vary relatively little. 123 00:07:16,640 --> 00:07:18,790 What varies is the enforcement of the law. 124 00:07:18,790 --> 00:07:21,345 So you have to make-- who's watching that the laws are 125 00:07:21,345 --> 00:07:21,820 getting enforced? 126 00:07:21,820 --> 00:07:22,302 Yeah. 127 00:07:22,302 --> 00:07:22,784 Melissa. 128 00:07:22,784 --> 00:07:25,194 AUDIENCE: I guess as an example of this on a much 129 00:07:25,194 --> 00:07:29,050 smaller scale is you have these cruise lines set up 130 00:07:29,050 --> 00:07:31,460 these enclaves in Caribbean islands. 131 00:07:31,460 --> 00:07:34,518 And it'd be a town in Haiti, and it functions much better 132 00:07:34,518 --> 00:07:36,116 than the rest of the country. 133 00:07:36,116 --> 00:07:38,032 And part of that's because they have economic demand. 134 00:07:38,032 --> 00:07:40,906 But the cruise line has basically bought the town and 135 00:07:40,906 --> 00:07:42,343 runs things. 136 00:07:42,343 --> 00:07:44,766 So like I said, it's all on a much smaller scale. 137 00:07:44,766 --> 00:07:46,758 But it seems like that that's the only thing. 138 00:07:46,758 --> 00:07:49,995 AUDIENCE: In those instances, generally speaking, the cruise 139 00:07:49,995 --> 00:07:51,156 line owns all of the land. 140 00:07:51,156 --> 00:07:52,570 And it's private property. 141 00:07:52,570 --> 00:07:53,920 And they setup the town. 142 00:07:53,920 --> 00:07:56,130 So it's really no-- 143 00:07:56,130 --> 00:07:56,780 PROFESSOR: It's like a hotel. 144 00:07:56,780 --> 00:07:58,410 AUDIENCE: It's like a privatized government. 145 00:07:58,410 --> 00:07:58,570 AUDIENCE: Yeah. 146 00:07:58,570 --> 00:07:59,360 It's like a large hotel. 147 00:07:59,360 --> 00:08:01,130 AUDIENCE: But it's larger than a hotel. 148 00:08:01,130 --> 00:08:03,430 PROFESSOR: But a large hotel. 149 00:08:03,430 --> 00:08:06,410 Some hotels have their internal currency. 150 00:08:06,410 --> 00:08:09,520 I was once trapped into this completely horrible hotel 151 00:08:09,520 --> 00:08:12,300 which had this internal currency with high inflation. 152 00:08:12,300 --> 00:08:13,780 This was a hotel. 153 00:08:13,780 --> 00:08:15,750 Because whenever I wanted to buy something-- 154 00:08:15,750 --> 00:08:17,940 so basically, there you go into the hotel. 155 00:08:17,940 --> 00:08:18,770 It's like a resort. 156 00:08:18,770 --> 00:08:23,210 And they give you a certain amount of currency, and you 157 00:08:23,210 --> 00:08:26,170 spend it anywhere you want in the resort. 158 00:08:26,170 --> 00:08:28,150 But it turns out that the prices are 159 00:08:28,150 --> 00:08:29,400 extraordinarily high. 160 00:08:32,340 --> 00:08:34,689 So that's sort of the same idea. 161 00:08:34,689 --> 00:08:35,647 AUDIENCE: It's quite different. 162 00:08:35,647 --> 00:08:39,805 But I guess it's just that we don't really have anything on 163 00:08:39,805 --> 00:08:43,796 a smaller scale to this that we could even say, does it 164 00:08:43,796 --> 00:08:45,150 work on a smaller scale? 165 00:08:45,150 --> 00:08:48,090 PROFESSOR: So why is it different from a resort where 166 00:08:48,090 --> 00:08:51,270 you can go from one shop to another? 167 00:08:51,270 --> 00:08:52,730 The shops are competing. 168 00:08:52,730 --> 00:08:55,155 Why is it different? 169 00:08:55,155 --> 00:08:58,792 AUDIENCE: At least one thing that's different is there's a 170 00:08:58,792 --> 00:09:01,362 major industry there that creates demand. 171 00:09:01,362 --> 00:09:05,790 Whereas if you set some city up in the middle of Africa, 172 00:09:05,790 --> 00:09:08,988 where there's not already an economic industry, and 173 00:09:08,988 --> 00:09:09,740 everybody's-- 174 00:09:09,740 --> 00:09:12,530 PROFESSOR: So part of it is that there is an economic 175 00:09:12,530 --> 00:09:14,685 engine that's already there. 176 00:09:17,460 --> 00:09:19,920 Then there's a chicken-and-egg problem, which is that the 177 00:09:19,920 --> 00:09:24,150 city will only generate value if it comes with an economic 178 00:09:24,150 --> 00:09:27,120 model, and the economic model isn't there already. 179 00:09:27,120 --> 00:09:30,640 If you just guarantee the institutions, how good does 180 00:09:30,640 --> 00:09:32,830 the guarantee need to be before the 181 00:09:32,830 --> 00:09:35,140 economic engine arrives? 182 00:09:35,140 --> 00:09:36,790 That's a challenge. 183 00:09:36,790 --> 00:09:39,870 As you might imagine, the problem might be that if I 184 00:09:39,870 --> 00:09:43,590 tell you that I'm in Cote d'Ivoire in the middle of the 185 00:09:43,590 --> 00:09:47,260 civil war, but we have this enclave that's protected, you 186 00:09:47,260 --> 00:09:50,430 really need very good assurances of protection 187 00:09:50,430 --> 00:09:54,400 before some industry's going to be willing to move to this 188 00:09:54,400 --> 00:09:57,760 enclave in Cote d'Ivoire in the middle of the civil war. 189 00:09:57,760 --> 00:10:01,560 So you're always this question, who guarantees? 190 00:10:01,560 --> 00:10:05,070 But also, I think there is a third question, which is, 191 00:10:05,070 --> 00:10:12,180 what's to stop the country from withdrawing its 192 00:10:12,180 --> 00:10:12,890 concessions? 193 00:10:12,890 --> 00:10:15,480 So let's say the country gave you some land. 194 00:10:15,480 --> 00:10:17,110 You build a lot of buildings. 195 00:10:17,110 --> 00:10:19,760 You've got a few factories running there. 196 00:10:19,760 --> 00:10:22,620 What's to start a country with bad institutions with coming 197 00:10:22,620 --> 00:10:25,280 back the next year and saying, well, now that we have all of 198 00:10:25,280 --> 00:10:27,420 these factories, I want them. 199 00:10:27,420 --> 00:10:28,920 And who's going to stop that? 200 00:10:28,920 --> 00:10:31,150 Will the US send in an army to stop that? 201 00:10:31,150 --> 00:10:35,650 It seems like there are myriad issues of implementation. 202 00:10:35,650 --> 00:10:39,342 It would make this a rather wacky scheme. 203 00:10:39,342 --> 00:10:40,818 AUDIENCE: [INAUDIBLE] 204 00:10:40,818 --> 00:10:42,786 also the fact that [INAUDIBLE] 205 00:10:42,786 --> 00:10:43,770 affect the rest of the country. 206 00:10:43,770 --> 00:10:49,182 [? They're ?] going to build a whole new city [INAUDIBLE]. 207 00:10:49,182 --> 00:10:52,626 And all the existing cities and things surrounding that 208 00:10:52,626 --> 00:10:55,086 will not necessarily be receptive to it. 209 00:10:55,086 --> 00:10:57,580 [INAUDIBLE]. 210 00:10:57,580 --> 00:10:58,120 PROFESSOR: Yeah. 211 00:10:58,120 --> 00:10:59,170 That too. 212 00:10:59,170 --> 00:11:05,070 Just how do you deal with the inequality and all that within 213 00:11:05,070 --> 00:11:05,630 the country? 214 00:11:05,630 --> 00:11:06,598 Yeah. 215 00:11:06,598 --> 00:11:09,502 AUDIENCE: So when Paul Romer proposed this, it sounds very 216 00:11:09,502 --> 00:11:12,406 theoretical, and I don't think he actually proposed how to 217 00:11:12,406 --> 00:11:13,374 implement it. 218 00:11:13,374 --> 00:11:16,278 But did he mention the timeline for these cities? 219 00:11:16,278 --> 00:11:19,575 Because although you might have the framework for good 220 00:11:19,575 --> 00:11:21,725 institutions, you might have all the laws, you still need 221 00:11:21,725 --> 00:11:24,000 the people to implement it, the people to respect it. 222 00:11:24,000 --> 00:11:28,220 And that's somewhat a cultural issue, that if you grew up in 223 00:11:28,220 --> 00:11:31,670 a corrupt society, you have it ingrained in you how to go 224 00:11:31,670 --> 00:11:31,950 about things. 225 00:11:31,950 --> 00:11:36,209 So it would take at least one generation, or several 226 00:11:36,209 --> 00:11:39,480 generations, to get people to change their ways and actually 227 00:11:39,480 --> 00:11:44,320 follow the law and help make these institutions work. 228 00:11:44,320 --> 00:11:48,617 For example, Russia transiting from the Communist society to 229 00:11:48,617 --> 00:11:51,192 democracy, they have all the laws and all the rules and all 230 00:11:51,192 --> 00:11:51,770 the institutions. 231 00:11:51,770 --> 00:11:59,330 But people still have this old mindset that's not working. 232 00:11:59,330 --> 00:12:02,580 PROFESSOR: That goes back to the first point we were 233 00:12:02,580 --> 00:12:09,360 saying, which is that the evidence on institutions says 234 00:12:09,360 --> 00:12:12,390 that if you got the right accident which made your 235 00:12:12,390 --> 00:12:17,480 institutions the right ones 200 or 400 years ago, then 236 00:12:17,480 --> 00:12:20,140 that persists. 237 00:12:20,140 --> 00:12:22,850 That doesn't say anything about transplanting 238 00:12:22,850 --> 00:12:23,560 institutions. 239 00:12:23,560 --> 00:12:27,910 But I think Romer is asking the right question, which is, 240 00:12:27,910 --> 00:12:31,210 imagine that the US has good institutions. 241 00:12:31,210 --> 00:12:34,980 Can a US company take advantage of the fact that it 242 00:12:34,980 --> 00:12:38,220 is a US company, and the US has good institutions, to set 243 00:12:38,220 --> 00:12:43,590 up a kind of a charter city in Cote d'Ivoire which will have 244 00:12:43,590 --> 00:12:44,350 good institutions? 245 00:12:44,350 --> 00:12:48,070 So he's thought about this issue. 246 00:12:48,070 --> 00:12:51,130 He thinks that, well, the way you do it is, you import 247 00:12:51,130 --> 00:12:52,610 institutions. 248 00:12:52,610 --> 00:12:53,980 That's the idea. 249 00:12:53,980 --> 00:12:58,670 And so in some sense, you're totally right. 250 00:12:58,670 --> 00:13:02,800 It's not enough to just write down the rules of the 251 00:13:02,800 --> 00:13:03,710 institutions. 252 00:13:03,710 --> 00:13:04,900 And that's completely right. 253 00:13:04,900 --> 00:13:06,070 But he's not saying that. 254 00:13:06,070 --> 00:13:09,250 He's saying that plus implementation will come from 255 00:13:09,250 --> 00:13:10,310 US company. 256 00:13:10,310 --> 00:13:12,650 The question is whether a US company can implement 257 00:13:12,650 --> 00:13:13,190 anything like that. 258 00:13:13,190 --> 00:13:14,480 Yeah. 259 00:13:14,480 --> 00:13:16,530 AUDIENCE: Couldn't to some degree-- 260 00:13:16,530 --> 00:13:18,544 I think it'll be interesting to see in 20 or 30 years, what 261 00:13:18,544 --> 00:13:20,179 Afghanistan and Iraq look like? 262 00:13:20,179 --> 00:13:22,047 I mean, obviously, they had a pretty distinct population. 263 00:13:22,047 --> 00:13:24,860 But to some degree, the closest thing to what Romer 264 00:13:24,860 --> 00:13:28,540 suggested we're seeing in action in Iraq. 265 00:13:28,540 --> 00:13:31,840 I mean, the level of US monetary involvement has been 266 00:13:31,840 --> 00:13:33,190 huge [INAUDIBLE] 267 00:13:33,190 --> 00:13:34,320 military expenses. 268 00:13:34,320 --> 00:13:36,510 And they're trying to create these new institutions. 269 00:13:36,510 --> 00:13:39,876 So that could be an interesting example case to 270 00:13:39,876 --> 00:13:44,540 build off of in the future, when changed institutions on a 271 00:13:44,540 --> 00:13:45,770 wide scale he was talking about. 272 00:13:45,770 --> 00:13:48,700 I mean, obviously, it's not very sustainable, I imagine. 273 00:13:48,700 --> 00:13:49,960 And there will be a series of issues. 274 00:13:49,960 --> 00:13:51,910 But just as an example. 275 00:13:51,910 --> 00:13:56,606 PROFESSOR: This is an aside, but it's worth reading the 276 00:13:56,606 --> 00:13:59,960 book, I think it's called Imperial Life in the Emerald 277 00:13:59,960 --> 00:14:01,310 City, or something. 278 00:14:01,310 --> 00:14:03,310 Wonderful book. 279 00:14:03,310 --> 00:14:08,710 The Washington Post bureau chief in Iraq, after a year 280 00:14:08,710 --> 00:14:13,750 and a half there, wrote this book about the implementation 281 00:14:13,750 --> 00:14:16,120 of US institutions in Iraq. 282 00:14:16,120 --> 00:14:21,000 And what was wonderful about the story in that book was-- 283 00:14:21,000 --> 00:14:22,970 and this is relevant for what we're going to 284 00:14:22,970 --> 00:14:23,920 talk about in a bit. 285 00:14:23,920 --> 00:14:25,880 So it's not entirely an aside. 286 00:14:25,880 --> 00:14:29,530 So the book was wonderful in describing exactly how the US 287 00:14:29,530 --> 00:14:33,860 went about implementing good institutions in Iraq. 288 00:14:33,860 --> 00:14:40,310 So basically, rule one was that the people who got hired 289 00:14:40,310 --> 00:14:49,880 to do that were either staffers of prominent 290 00:14:49,880 --> 00:14:54,440 Republican congressmen or senators, or prominent 291 00:14:54,440 --> 00:14:57,210 contributors to the Republican cause. 292 00:14:57,210 --> 00:14:59,300 Those were the only people who were selected. 293 00:14:59,300 --> 00:15:01,990 Second, most of the staffers who were given the 294 00:15:01,990 --> 00:15:06,250 responsibility for, for example, designing the 295 00:15:06,250 --> 00:15:12,110 stock-market laws, were 23-year-olds who had an 296 00:15:12,110 --> 00:15:15,110 undergraduate degree in accounting were sent to the 297 00:15:15,110 --> 00:15:17,850 set up these laws. 298 00:15:17,850 --> 00:15:19,560 And there was usually one person. 299 00:15:24,300 --> 00:15:27,310 A 23-year-old with an undergraduate degree was 300 00:15:27,310 --> 00:15:34,700 assigned to be in charge of setting up the stock market or 301 00:15:34,700 --> 00:15:38,270 figuring out how to reform the legal system. 302 00:15:38,270 --> 00:15:44,430 So these where people with strong background in being 303 00:15:44,430 --> 00:15:50,300 Republican operatives, and no background in actually running 304 00:15:50,300 --> 00:15:51,180 an institution. 305 00:15:51,180 --> 00:15:55,400 So the book is wonderful in that it goes through how the 306 00:15:55,400 --> 00:15:58,340 whole process was doomed from the start. 307 00:15:58,340 --> 00:16:01,710 Because the people who were sent to do these institutions 308 00:16:01,710 --> 00:16:04,980 had no competence whatsoever in doing the job, because they 309 00:16:04,980 --> 00:16:06,680 were all political appointees. 310 00:16:06,680 --> 00:16:11,790 And the book is actually worth reading. 311 00:16:11,790 --> 00:16:14,180 And it comes back to point I'll come back to later. 312 00:16:14,180 --> 00:16:17,400 Which is that, imagine you wanted to make 313 00:16:17,400 --> 00:16:18,810 institutions better. 314 00:16:18,810 --> 00:16:20,300 You still need competence. 315 00:16:20,300 --> 00:16:24,640 And it's not just a matter of having some 316 00:16:24,640 --> 00:16:26,160 playbook from the US. 317 00:16:26,160 --> 00:16:29,310 It's a matter of actually having the competence to turn 318 00:16:29,310 --> 00:16:30,940 that playbook into reality. 319 00:16:30,940 --> 00:16:37,060 And I think that was a very major problem in Iraq. 320 00:16:46,970 --> 00:16:54,600 So then there is Paul Collier, who has a different view. 321 00:16:54,600 --> 00:16:59,050 In some ways, this view I find less unrealistic-- 322 00:16:59,050 --> 00:17:01,740 maybe less defensible, but less unrealistic. 323 00:17:01,740 --> 00:17:07,329 Which is, if the country has really horrible institutions, 324 00:17:07,329 --> 00:17:08,579 just invade. 325 00:17:12,630 --> 00:17:15,040 This is, I think, internally coherent. 326 00:17:15,040 --> 00:17:18,099 Much more, I think, internally coherent than the Romer 327 00:17:18,099 --> 00:17:21,260 suggestion, which I don't see how it would be implemented. 328 00:17:21,260 --> 00:17:26,869 But this one, at least the policy prescription's simple. 329 00:17:26,869 --> 00:17:34,200 Invade, and then just set up colonial rule for a while. 330 00:17:34,200 --> 00:17:35,560 This has been tried. 331 00:17:35,560 --> 00:17:37,760 It's been implemented in the past. 332 00:17:37,760 --> 00:17:40,490 We know how it's to be done, so there's no 333 00:17:40,490 --> 00:17:41,970 implementation problem. 334 00:17:41,970 --> 00:17:44,050 You might have various objections to it. 335 00:17:48,080 --> 00:17:51,570 But in many ways, this is a more difficult one to 336 00:17:51,570 --> 00:17:58,160 challenge on purely a priori grounds. 337 00:17:58,160 --> 00:18:02,190 One of the sad facts is that we know very little about the 338 00:18:02,190 --> 00:18:05,590 effect of colonial rule from history. 339 00:18:05,590 --> 00:18:08,550 That is to say, if you look at places that didn't get 340 00:18:08,550 --> 00:18:11,290 colonized, so few of them didn't get colonized, and 341 00:18:11,290 --> 00:18:14,240 they're so unique, like China is one country 342 00:18:14,240 --> 00:18:17,750 that didn't get colonized. 343 00:18:17,750 --> 00:18:27,040 But basically, poor countries, most countries outside Europe 344 00:18:27,040 --> 00:18:30,470 and North America were colonized. 345 00:18:30,470 --> 00:18:32,865 And as a result, there's just very little. 346 00:18:32,865 --> 00:18:35,460 And when you compare the ones that got colonized with the 347 00:18:35,460 --> 00:18:38,400 ones that didn't, the ones that didn't get colonized were 348 00:18:38,400 --> 00:18:39,760 Japan and China. 349 00:18:39,760 --> 00:18:42,540 And those are not random draws from any population. 350 00:18:42,540 --> 00:18:47,990 So it's very hard to see what the effect of 351 00:18:47,990 --> 00:18:49,580 being colonized is. 352 00:18:49,580 --> 00:18:53,500 The one exception to that that I know, which is not quite 353 00:18:53,500 --> 00:18:58,490 perfect, but it does a little bit of that, is, an MIT Ph.D. 354 00:18:58,490 --> 00:19:02,610 student some years ago, Lakshmi Iyer, did this paper 355 00:19:02,610 --> 00:19:09,090 where what she did is, she looked at, in India, there was 356 00:19:09,090 --> 00:19:11,260 what was called the Doctrine of Lapse. 357 00:19:11,260 --> 00:19:13,320 The Doctrine of Lapse was a beautiful idea. 358 00:19:13,320 --> 00:19:16,655 So the way India was colonized was bit by bit. 359 00:19:22,220 --> 00:19:24,030 Some bits got left out. 360 00:19:24,030 --> 00:19:27,400 At any point, they would fight a battle, take over some land, 361 00:19:27,400 --> 00:19:30,030 and then some would be left out. 362 00:19:30,030 --> 00:19:34,910 And then they'd have a treaty with the local potentate at 363 00:19:34,910 --> 00:19:40,020 that time to have a kind of a dependent ruler. 364 00:19:45,080 --> 00:19:50,070 Basically, this process of colonizing India ends in 1857 365 00:19:50,070 --> 00:19:55,490 with about 2/3 of India under direct British rule, and one 366 00:19:55,490 --> 00:19:59,700 third of India under rule of a whole bunch of rulers-- 367 00:19:59,700 --> 00:20:03,990 500-odd rulers who had a little small territory, and 368 00:20:03,990 --> 00:20:06,030 was the king of that territory, and had a treaty 369 00:20:06,030 --> 00:20:06,760 with the British. 370 00:20:06,760 --> 00:20:11,940 The British basically were overseeing this person. 371 00:20:11,940 --> 00:20:14,260 So that's how the colonization happened. 372 00:20:14,260 --> 00:20:18,320 Now, one of the ways in which the British would take over 373 00:20:18,320 --> 00:20:21,490 one of these little kingdoms was under what was called the 374 00:20:21,490 --> 00:20:22,510 Doctrine of Lapse. 375 00:20:22,510 --> 00:20:28,430 The Doctrine of Lapse was, one might say, a cynical idea. 376 00:20:28,430 --> 00:20:34,800 But it was the idea that if the king didn't have a natural 377 00:20:34,800 --> 00:20:41,080 heir, then the kingdom would lapse to British rule. 378 00:20:41,080 --> 00:20:49,090 So basically, what this paper does, is she compares places 379 00:20:49,090 --> 00:20:50,760 which, during-- 380 00:20:50,760 --> 00:20:56,540 Doctrine of Lapse was enforced very heavily for eight years, 381 00:20:56,540 --> 00:20:59,500 from 1848 to 1856. 382 00:20:59,500 --> 00:21:15,440 During this period, about 20 places had the king die. 383 00:21:15,440 --> 00:21:20,770 And I guess some of those places had a 384 00:21:20,770 --> 00:21:22,490 male natural heir. 385 00:21:22,490 --> 00:21:25,540 Natural heir means a son. 386 00:21:25,540 --> 00:21:27,470 And the other places didn't. 387 00:21:27,470 --> 00:21:29,910 So what she does is, she compares the places which had 388 00:21:29,910 --> 00:21:32,040 a natural heir with the places which didn't 389 00:21:32,040 --> 00:21:34,870 have a natural heir. 390 00:21:34,870 --> 00:21:36,920 So if you look at these places, if you think that the 391 00:21:36,920 --> 00:21:41,330 birth of a male child is kind of a random accident, then you 392 00:21:41,330 --> 00:21:42,970 might say that these are comparable. 393 00:21:42,970 --> 00:21:45,560 And on most things, they look comparable. 394 00:21:45,560 --> 00:21:51,550 If you look at these places 150 years later, the places 395 00:21:51,550 --> 00:21:57,920 that were under British rule do worse in almost every 396 00:21:57,920 --> 00:22:00,690 attribute that can be measured. 397 00:22:00,690 --> 00:22:05,360 Less good infrastructure, less good education, less good 398 00:22:05,360 --> 00:22:06,080 other things. 399 00:22:06,080 --> 00:22:09,260 So that's the only piece of evidence I've seen that has 400 00:22:09,260 --> 00:22:12,060 some bearing on this question. 401 00:22:12,060 --> 00:22:16,670 It's still a very special set of 20 places. 402 00:22:16,670 --> 00:22:20,490 So I would say, we know very little about even the economic 403 00:22:20,490 --> 00:22:22,110 effects of colonialism. 404 00:22:22,110 --> 00:22:24,480 And then there are all the political effects, which are 405 00:22:24,480 --> 00:22:30,560 that colonists often create hierarchies, and those 406 00:22:30,560 --> 00:22:34,570 hierarchies create all kinds of resentment, and a whole 407 00:22:34,570 --> 00:22:42,540 bunch of other stuff that potentially can go wrong. 408 00:22:42,540 --> 00:22:45,310 But I don't think there's a great natural experiment that 409 00:22:45,310 --> 00:22:51,270 tells you what would have happened had 410 00:22:51,270 --> 00:22:53,130 Algeria not been colonized. 411 00:22:53,130 --> 00:22:54,585 I don't think anybody knows the answer. 412 00:22:54,585 --> 00:22:55,060 Yeah. 413 00:22:55,060 --> 00:22:57,435 AUDIENCE: [INAUDIBLE] the fact that everywhere from the Congo 414 00:22:57,435 --> 00:23:01,890 to the US and Canada have been colonized suggests that it 415 00:23:01,890 --> 00:23:07,576 doesn't matter so much on the nature of it. 416 00:23:07,576 --> 00:23:10,890 Why do we want to make a general statement? 417 00:23:10,890 --> 00:23:14,360 PROFESSOR: I guess the question is whether or not 418 00:23:14,360 --> 00:23:20,480 once you interact race with colony, you 419 00:23:20,480 --> 00:23:23,040 get much worse outcomes. 420 00:23:23,040 --> 00:23:24,150 You almost surely do. 421 00:23:24,150 --> 00:23:26,800 The question is whether that's interpretable or not. 422 00:23:26,800 --> 00:23:32,220 If I said, take colonies which are nonwhite colonies, they do 423 00:23:32,220 --> 00:23:34,490 a lot worse than white colonies. 424 00:23:34,490 --> 00:23:35,920 Now, what that means I don't know. 425 00:23:35,920 --> 00:23:42,570 But I do think that you could make a firm hypothesis, which 426 00:23:42,570 --> 00:23:46,210 I don't think is testable given the data, that if you 427 00:23:46,210 --> 00:23:49,420 were nonwhite and you were colonized by white people, you 428 00:23:49,420 --> 00:23:50,780 ended up badly. 429 00:23:50,780 --> 00:23:52,120 You could make the hypothesis. 430 00:23:52,120 --> 00:23:53,160 It wouldn't be rejected. 431 00:23:53,160 --> 00:23:54,330 AUDIENCE: [? Half the ?] people in the US who were not 432 00:23:54,330 --> 00:23:56,787 white who were colonized by white people, it still ended 433 00:23:56,787 --> 00:24:00,040 up badly, economically, relative to white Americans. 434 00:24:00,040 --> 00:24:00,530 PROFESSOR: Exactly. 435 00:24:00,530 --> 00:24:03,760 So I'm not saying that people resolved this question. 436 00:24:03,760 --> 00:24:06,620 I'm saying the hypothesis would not be prima facie 437 00:24:06,620 --> 00:24:09,670 rejected by the data. 438 00:24:09,670 --> 00:24:12,420 But my point is only that we don't know. 439 00:24:12,420 --> 00:24:13,710 I wasn't saying more than that. 440 00:24:13,710 --> 00:24:16,420 I would just say that it's not implausible hypothesis, but 441 00:24:16,420 --> 00:24:21,780 one that I don't think we know enough to 442 00:24:21,780 --> 00:24:23,500 answer Collier's question. 443 00:24:23,500 --> 00:24:26,010 I actually don't think that this is a realistic idea, 444 00:24:26,010 --> 00:24:26,790 because I don't think 445 00:24:26,790 --> 00:24:29,880 anybody's prepared to actually-- 446 00:24:29,880 --> 00:24:33,050 I think if you look at the US, the real problem with getting 447 00:24:33,050 --> 00:24:37,910 the US to be a colonist is that the US actually loves the 448 00:24:37,910 --> 00:24:43,330 idea, maybe, of being influential in the world, but 449 00:24:43,330 --> 00:24:46,490 it doesn't actually like the idea of having young US 450 00:24:46,490 --> 00:24:50,750 soldiers fighting to protect some foreign land from its own 451 00:24:50,750 --> 00:24:51,850 internal conflicts. 452 00:24:51,850 --> 00:24:54,900 That is not a politically popular idea. 453 00:24:54,900 --> 00:24:58,570 So I don't actually think any country is volunteering to be 454 00:24:58,570 --> 00:24:59,750 a colonist. 455 00:24:59,750 --> 00:25:02,430 And that might be a good thing, but it is certainly 456 00:25:02,430 --> 00:25:05,190 true that, I think, after the Vietnam war, the US has had 457 00:25:05,190 --> 00:25:07,860 enough of long-term colonialism. 458 00:25:07,860 --> 00:25:11,130 I think that's the given. 459 00:25:11,130 --> 00:25:13,340 So I don't think there is any country that's volunteering. 460 00:25:13,340 --> 00:25:15,860 So again, I don't think Collier's solution, whether 461 00:25:15,860 --> 00:25:17,120 it's right or wrong-- 462 00:25:17,120 --> 00:25:19,170 I'm inclined to believe it wouldn't work, even if you 463 00:25:19,170 --> 00:25:22,390 tried, but that's a [INAUDIBLE]. 464 00:25:22,390 --> 00:25:24,740 I don't think there is any evidence that anybody's 465 00:25:24,740 --> 00:25:25,560 willing to do it. 466 00:25:25,560 --> 00:25:26,810 So it's maybe moot. 467 00:25:29,610 --> 00:25:33,370 And then there is Bill Easterly, who's been very 468 00:25:33,370 --> 00:25:37,780 critical of both Romer and Collier. 469 00:25:37,780 --> 00:25:41,820 And his position is very internally consistent. 470 00:25:41,820 --> 00:25:46,070 His view is that there's nothing you 471 00:25:46,070 --> 00:25:47,300 can do to help countries. 472 00:25:47,300 --> 00:25:49,040 Countries have to help themselves. 473 00:25:49,040 --> 00:25:55,380 There is no sense in which there is really any outside 474 00:25:55,380 --> 00:25:58,510 wisdom that's useful, with an important caveat I'll come 475 00:25:58,510 --> 00:25:59,620 back to in a minute. 476 00:25:59,620 --> 00:26:03,850 But therefore, colonialism won't help, because countries 477 00:26:03,850 --> 00:26:04,820 help themselves. 478 00:26:04,820 --> 00:26:07,200 They solve their own problems. 479 00:26:07,200 --> 00:26:10,670 You can't really help any country. 480 00:26:10,670 --> 00:26:15,180 He doesn't quite stop there, which would be really 481 00:26:15,180 --> 00:26:17,350 internally consistent-- 482 00:26:17,350 --> 00:26:19,900 whether right or wrong is a different question. 483 00:26:19,900 --> 00:26:23,910 He says, except you should have free trade, democracy, 484 00:26:23,910 --> 00:26:26,000 and capitalism. 485 00:26:26,000 --> 00:26:28,360 Those are the three things that he's in favor of. 486 00:26:28,360 --> 00:26:31,370 And then he's not really in favor of anything else. 487 00:26:31,370 --> 00:26:40,740 So he does believe that the country should not have choice 488 00:26:40,740 --> 00:26:43,650 on those dimensions, but once you have implemented those, 489 00:26:43,650 --> 00:26:47,130 you can leave them alone. 490 00:26:47,130 --> 00:26:50,390 Now, again, the problem is that 491 00:26:50,390 --> 00:26:52,870 there is not much evidence-- 492 00:26:52,870 --> 00:26:57,100 I mean, certainly some countries have figured it out. 493 00:26:57,100 --> 00:27:01,030 It's not clear that there's any evidence that the process 494 00:27:01,030 --> 00:27:02,340 of figuring out is a 495 00:27:02,340 --> 00:27:04,790 particularly efficient process. 496 00:27:04,790 --> 00:27:08,460 We've talked a lot through this semester about policies 497 00:27:08,460 --> 00:27:13,630 that were tried and didn't work, and they didn't work for 498 00:27:13,630 --> 00:27:16,000 very good reasons. 499 00:27:16,000 --> 00:27:20,010 So in other words, there is a lot of knowledge out there 500 00:27:20,010 --> 00:27:25,030 that's useful for countries to use. 501 00:27:25,030 --> 00:27:27,460 So it's not clear that you want to take the position that 502 00:27:27,460 --> 00:27:32,360 countries need to make every mistake and figure it out. 503 00:27:32,360 --> 00:27:35,560 And it's not clear what a country means. 504 00:27:35,560 --> 00:27:42,930 Is it really the case that, if I think of Tunisia under Ben 505 00:27:42,930 --> 00:27:47,500 Ali, there was a collective experiment in dictatorship? 506 00:27:47,500 --> 00:27:51,270 It doesn't seem like that's the right description. 507 00:27:51,270 --> 00:27:54,260 So a country is just a set of people. 508 00:27:54,260 --> 00:27:59,920 But it's not clear that there is any organizing body which 509 00:27:59,920 --> 00:28:02,810 is implementing a national experiment. 510 00:28:02,810 --> 00:28:07,960 I mean, maybe some countries are, but most countries, when 511 00:28:07,960 --> 00:28:10,480 they fail, they seem to be not doing anything particularly 512 00:28:10,480 --> 00:28:10,980 interesting. 513 00:28:10,980 --> 00:28:13,870 Some guy is stealing a lot of money, and the rest of the 514 00:28:13,870 --> 00:28:15,120 people are suffering. 515 00:28:21,320 --> 00:28:24,860 It's not clear that's a symptom of the process of 516 00:28:24,860 --> 00:28:27,370 figuring out where you're going, rather than just a 517 00:28:27,370 --> 00:28:30,430 symptom of failure. 518 00:28:30,430 --> 00:28:33,890 And then, I think the most internally consistent are 519 00:28:33,890 --> 00:28:39,280 definitely our colleagues Acemoglu and Robinson, whose 520 00:28:39,280 --> 00:28:43,680 view is very much that there's nothing you can do to help, 521 00:28:43,680 --> 00:28:47,560 but also that mostly nothing good will happen. 522 00:28:47,560 --> 00:28:51,370 So their view is, I think, most solidly internally 523 00:28:51,370 --> 00:28:52,490 consistent. 524 00:28:52,490 --> 00:28:58,290 Basically, countries are probably screwed. 525 00:28:58,290 --> 00:28:58,760 Yeah. 526 00:28:58,760 --> 00:29:01,195 AUDIENCE: I don't think that's a fair assessment. 527 00:29:01,195 --> 00:29:03,873 Jim Robinson's doing all this stuff in Sierra Leone now with 528 00:29:03,873 --> 00:29:06,065 the World Bank and [INAUDIBLE]. 529 00:29:06,065 --> 00:29:08,500 PROFESSOR: But I know Jim very well. 530 00:29:08,500 --> 00:29:09,830 I know Jim very well. 531 00:29:09,830 --> 00:29:11,080 I think he would say he's a pessimist. 532 00:29:14,560 --> 00:29:19,040 He might be trying to do something, but I think he's-- 533 00:29:19,040 --> 00:29:23,220 I would say that that's the most internally consistent of 534 00:29:23,220 --> 00:29:26,830 all of these positions that you see, is the one of saying 535 00:29:26,830 --> 00:29:29,710 that, look, if I really believe that institutional 536 00:29:29,710 --> 00:29:32,820 changes are very difficult, and they only happen because 537 00:29:32,820 --> 00:29:37,260 of major historical events, then it's internally 538 00:29:37,260 --> 00:29:38,970 consistent be pessimistic. 539 00:29:38,970 --> 00:29:43,500 Because why would we have much hope? 540 00:29:43,500 --> 00:29:49,380 And I think the example they give of changes are the French 541 00:29:49,380 --> 00:29:51,580 Revolution and the Glorious Revolution. 542 00:29:51,580 --> 00:29:54,395 That's the two examples that they mostly spend a lot of 543 00:29:54,395 --> 00:29:55,645 time talking about. 544 00:30:05,070 --> 00:30:09,040 So their view is that change is possible, but change 545 00:30:09,040 --> 00:30:14,450 happens when internal forces within a country just happen 546 00:30:14,450 --> 00:30:16,920 to be aligned in the perfect way. 547 00:30:16,920 --> 00:30:20,190 And they give these two examples of the Glorious 548 00:30:20,190 --> 00:30:24,950 Revolution in Britain in 1688 and the French Revolution in 549 00:30:24,950 --> 00:30:29,960 1789 as the two examples of where you just had the right 550 00:30:29,960 --> 00:30:31,410 alignment of forces. 551 00:30:31,410 --> 00:30:34,260 And then the internal compulsions of the country 552 00:30:34,260 --> 00:30:37,190 just drove the country towards good outcomes. 553 00:30:37,190 --> 00:30:41,000 That in each case, these were just moments of liberation, 554 00:30:41,000 --> 00:30:43,485 and they did wonderful things for the country. 555 00:30:46,600 --> 00:30:50,210 There are many, many other such revolutions, and they 556 00:30:50,210 --> 00:30:51,900 have been less successful. 557 00:30:55,700 --> 00:30:59,920 The Russian Revolution, the Mexican Revolution, the 558 00:30:59,920 --> 00:31:01,670 Chinese Revolution. 559 00:31:01,670 --> 00:31:04,270 There are many others you could go through, and they 560 00:31:04,270 --> 00:31:05,850 have been less successful, clearly. 561 00:31:05,850 --> 00:31:09,740 So their view is that it's not enough to have revolutions. 562 00:31:09,740 --> 00:31:13,370 You need to change societies massively. 563 00:31:13,370 --> 00:31:15,130 You need to have revolutions. 564 00:31:15,130 --> 00:31:17,150 But revolutions are not easy. 565 00:31:17,150 --> 00:31:18,590 Mostly they go wrong. 566 00:31:18,590 --> 00:31:23,190 It's only when you have the lucky matching of different 567 00:31:23,190 --> 00:31:25,800 forces that you get the right outcome. 568 00:31:25,800 --> 00:31:29,250 And therefore, mostly there's not much you can do. 569 00:31:29,250 --> 00:31:30,230 You can get lucky. 570 00:31:30,230 --> 00:31:31,850 If you get lucky, things will work. 571 00:31:31,850 --> 00:31:34,690 If they do not, there's not much you can do about it. 572 00:31:34,690 --> 00:31:37,970 And that's very consistent with the view that they have 573 00:31:37,970 --> 00:31:43,890 from their evidence that all we really know is that, if 574 00:31:43,890 --> 00:31:48,080 countries got it right several hundred years ago, then that 575 00:31:48,080 --> 00:31:48,760 seems to persist. 576 00:31:48,760 --> 00:31:51,580 So when good things happen, good things build on each 577 00:31:51,580 --> 00:31:55,280 other, but we have no recipe for creating good things. 578 00:31:55,280 --> 00:32:00,180 So that's a very, very pessimistic and tough-minded, 579 00:32:00,180 --> 00:32:04,140 but, I think in some ways, admirably internally 580 00:32:04,140 --> 00:32:07,685 consistent point of view. 581 00:32:07,685 --> 00:32:10,150 AUDIENCE: I think [INAUDIBLE] 582 00:32:10,150 --> 00:32:13,108 has even tried to see [INAUDIBLE] 583 00:32:13,108 --> 00:32:17,052 institutions, changing institutions, is that US 584 00:32:17,052 --> 00:32:21,489 conventions, which are in the treaties, even on things like 585 00:32:21,489 --> 00:32:24,447 child labor or [INAUDIBLE]. 586 00:32:24,447 --> 00:32:27,898 So in your assessment, or your generally economic 587 00:32:27,898 --> 00:32:30,856 [INAUDIBLE], is it part of the reason, in many case, not 588 00:32:30,856 --> 00:32:32,828 effective enough, or [INAUDIBLE] 589 00:32:32,828 --> 00:32:35,300 will not be effective? 590 00:32:35,300 --> 00:32:38,960 PROFESSOR: So I'm going to actually argue that lots of 591 00:32:38,960 --> 00:32:41,960 institutional changes are possible. 592 00:32:41,960 --> 00:32:46,980 Just, they're possible at a different level from the ones 593 00:32:46,980 --> 00:32:48,680 that we are talking about. 594 00:32:48,680 --> 00:32:52,560 And so I would agree that many things 595 00:32:52,560 --> 00:32:53,730 change within countries. 596 00:32:53,730 --> 00:32:59,830 Indeed, I would say that if you look at status of women, 597 00:32:59,830 --> 00:33:02,300 the standing fact about the world is how much has changed 598 00:33:02,300 --> 00:33:04,570 over the last 150 years, not that there's 599 00:33:04,570 --> 00:33:05,860 not a lot of change. 600 00:33:05,860 --> 00:33:10,400 And I think the change has happened in China as much as 601 00:33:10,400 --> 00:33:13,300 it's happened in the US. 602 00:33:13,300 --> 00:33:17,010 And China was an extraordinarily, I would say, 603 00:33:17,010 --> 00:33:18,880 anti-woman country. 604 00:33:18,880 --> 00:33:21,980 And you see massive changes going on. 605 00:33:21,980 --> 00:33:25,770 So I'm going to argue mostly that there is actually lots 606 00:33:25,770 --> 00:33:29,550 and lots of possibilities for changing institutions, just 607 00:33:29,550 --> 00:33:31,200 that you don't-- 608 00:33:31,200 --> 00:33:35,080 this is asking a different question. 609 00:33:35,080 --> 00:33:49,280 So what we want to argue in the book is, 610 00:33:49,280 --> 00:33:51,870 lots of things change. 611 00:33:51,870 --> 00:33:55,380 They don't necessarily change wholesale, and you don't 612 00:33:55,380 --> 00:34:02,350 necessarily get better institutions 613 00:34:02,350 --> 00:34:07,140 in any macro sense. 614 00:34:07,140 --> 00:34:10,500 But that doesn't mean that things are not changing all 615 00:34:10,500 --> 00:34:12,850 the time, even within political 616 00:34:12,850 --> 00:34:14,330 systems which are bad. 617 00:34:14,330 --> 00:34:17,639 So it isn't the case that-- 618 00:34:17,639 --> 00:34:24,179 for example, countries which have had very large 619 00:34:24,179 --> 00:34:28,870 improvements in education and health. 620 00:34:28,870 --> 00:34:30,600 Indonesia under Suharto-- 621 00:34:30,600 --> 00:34:34,710 which was extraordinarily corrupt, was often rated the 622 00:34:34,710 --> 00:34:39,179 most corrupt country in the world, and was under an 623 00:34:39,179 --> 00:34:44,100 extraordinarily repressive dictator who at some point, 624 00:34:44,100 --> 00:34:47,969 it's claimed, killed a million people by labeling them 625 00:34:47,969 --> 00:34:51,389 Communists in the mid-'60s. 626 00:34:51,389 --> 00:34:55,000 Nevertheless, if you look at what's happening in Indonesia, 627 00:34:55,000 --> 00:35:01,600 between the period of 1972 and 1990, is one of the world's 628 00:35:01,600 --> 00:35:09,050 biggest, most successful investors in human 629 00:35:09,050 --> 00:35:10,290 capital in the world. 630 00:35:10,290 --> 00:35:15,070 You see big changes in education levels, you see big 631 00:35:15,070 --> 00:35:20,670 changes in health Suharto was very, very concerned about 632 00:35:20,670 --> 00:35:29,040 Indonesian health, nutrition, and he created a nationwide 633 00:35:29,040 --> 00:35:34,390 campaign of young people going back to villages and bringing 634 00:35:34,390 --> 00:35:36,270 message of better nutrition. 635 00:35:36,270 --> 00:35:38,710 And it's claimed to have improved nutrition massively 636 00:35:38,710 --> 00:35:41,560 in Indonesia. 637 00:35:41,560 --> 00:35:45,540 So in other words, that's not to say that Indonesia became a 638 00:35:45,540 --> 00:35:51,030 better country immediately, or that corruption went away. 639 00:35:51,030 --> 00:35:55,370 In fact, during the same period that education is 640 00:35:55,370 --> 00:36:00,590 improving in Indonesia, corruption is also going up. 641 00:36:00,590 --> 00:36:02,630 Suharto is becoming more and more corrupt. 642 00:36:02,630 --> 00:36:05,890 As his children get older, they're becoming more and more 643 00:36:05,890 --> 00:36:09,060 richer and richer, and buying up larger and larger shares of 644 00:36:09,060 --> 00:36:09,540 the country. 645 00:36:09,540 --> 00:36:12,620 So there was not that there was a one-way of movement in 646 00:36:12,620 --> 00:36:13,680 institutions. 647 00:36:13,680 --> 00:36:14,990 And indeed, that's my point. 648 00:36:18,940 --> 00:36:26,020 There is no sense in which there's one institution in 649 00:36:26,020 --> 00:36:29,100 place, one set of institutions in place. 650 00:36:29,100 --> 00:36:32,050 There are many, many things going on in any country. 651 00:36:32,050 --> 00:36:36,230 And many of those are getting better. 652 00:36:36,230 --> 00:36:39,220 Others are getting worse. 653 00:36:39,220 --> 00:36:42,910 I think we're always looking for leverage points. 654 00:36:42,910 --> 00:36:51,110 What are the best places to make a small push and see 655 00:36:51,110 --> 00:36:53,230 where we can make a change? 656 00:36:53,230 --> 00:37:00,090 So once we start thinking about institutions in small 657 00:37:00,090 --> 00:37:00,780 letters, rather than 658 00:37:00,780 --> 00:37:02,820 institutions in capital letters-- 659 00:37:02,820 --> 00:37:08,080 so not capitalism, but how a specific market is run-- 660 00:37:08,080 --> 00:37:09,990 we do see lots of changes. 661 00:37:09,990 --> 00:37:13,850 And that's the level at which we should think of 662 00:37:13,850 --> 00:37:14,830 institutional change. 663 00:37:14,830 --> 00:37:16,273 Yeah. 664 00:37:16,273 --> 00:37:20,010 AUDIENCE: I get the sense that, looking at it from one 665 00:37:20,010 --> 00:37:23,250 perspective, you should almost always prefer gradual reform. 666 00:37:23,250 --> 00:37:25,590 Because all the things that you showed us on the last 667 00:37:25,590 --> 00:37:28,940 slide are so out of sample that the chances that they'll 668 00:37:28,940 --> 00:37:33,213 actually work, just because they're so far beyond what we 669 00:37:33,213 --> 00:37:40,218 can observe or test and day-to-day things, that doing 670 00:37:40,218 --> 00:37:43,950 gradual reform would be the path of lesser resistance. 671 00:37:43,950 --> 00:37:48,345 But then Jim and [? Daron ?] and [? Pavel ?] have this 672 00:37:48,345 --> 00:37:50,080 paper where they talk about this thing called the seesaw 673 00:37:50,080 --> 00:37:52,152 effect, where you make one dimension of institutions 674 00:37:52,152 --> 00:37:54,450 better, another gets worse to compensate. 675 00:37:54,450 --> 00:37:57,140 Or Jeff Sachs would say that we need to change 676 00:37:57,140 --> 00:37:59,806 everything at once. 677 00:37:59,806 --> 00:38:00,798 Do we actually have-- 678 00:38:00,798 --> 00:38:02,782 I mean, I guess that Jim and [? Daron and ?] 679 00:38:02,782 --> 00:38:04,766 [? Pavel ?] have a little bit of evidence in their paper. 680 00:38:04,766 --> 00:38:10,010 But do we really have much in the way of systematic 681 00:38:10,010 --> 00:38:12,690 evidence for that? 682 00:38:12,690 --> 00:38:15,270 PROFESSOR: So I think the answer's no. 683 00:38:19,231 --> 00:38:29,230 It could be that they're right, that you can never 684 00:38:29,230 --> 00:38:33,440 improve anything without making something else worse. 685 00:38:33,440 --> 00:38:36,340 I find that implausible, let's say. 686 00:38:36,340 --> 00:38:40,340 But I'll leave that as a-- 687 00:38:40,340 --> 00:38:44,450 My main point I'm trying to make here is a simpler one, 688 00:38:44,450 --> 00:38:52,420 which is that there is no grand theorem which says that 689 00:38:52,420 --> 00:38:58,750 everything that you can change is determined by some 690 00:38:58,750 --> 00:39:01,070 overarching institutional frame. 691 00:39:01,070 --> 00:39:04,060 That even within bad institutions, there's lots of 692 00:39:04,060 --> 00:39:06,020 things that do change-- some for the better, 693 00:39:06,020 --> 00:39:07,340 some for the worse-- 694 00:39:07,340 --> 00:39:11,740 and that there is enough slack within the system that you can 695 00:39:11,740 --> 00:39:16,030 often change things even when things look pretty dire. 696 00:39:16,030 --> 00:39:19,310 So here's an example. 697 00:39:19,310 --> 00:39:21,120 In Chinese democracy-- 698 00:39:21,120 --> 00:39:24,730 this is not well-known, because China's not supposed 699 00:39:24,730 --> 00:39:25,910 to be a democracy. 700 00:39:25,910 --> 00:39:29,770 But in fact, China introduced village elections quite early. 701 00:39:29,770 --> 00:39:34,900 And the elections were phased in, so you could compare 702 00:39:34,900 --> 00:39:39,160 places which had elections with places which had a 703 00:39:39,160 --> 00:39:42,880 centrally-appointed village head. 704 00:39:42,880 --> 00:39:47,840 So this was just a comparison of what happens to a place 705 00:39:47,840 --> 00:39:51,340 when it becomes elected. 706 00:39:51,340 --> 00:40:00,750 So you been basically seem to see that when these 707 00:40:00,750 --> 00:40:03,720 institutional changes happen in a place like China, which 708 00:40:03,720 --> 00:40:06,560 is, after all, not a democracy. 709 00:40:06,560 --> 00:40:10,430 Huge control of the Communist party. 710 00:40:10,430 --> 00:40:16,410 Certainly, whenever there is any real attempt to change 711 00:40:16,410 --> 00:40:20,360 institutions at the high level, it's been stamped out 712 00:40:20,360 --> 00:40:23,160 with a great deal of clarity. 713 00:40:23,160 --> 00:40:27,475 There's never been any debate on how-- 714 00:40:30,400 --> 00:40:33,710 challenges is to the power of the Party have not been 715 00:40:33,710 --> 00:40:34,960 entertained. 716 00:40:37,060 --> 00:40:39,670 So you might imagine that when China introduces village 717 00:40:39,670 --> 00:40:41,210 elections, this would do nothing. 718 00:40:41,210 --> 00:40:44,070 Because power has not shifted. 719 00:40:44,070 --> 00:40:45,385 The same people have power. 720 00:40:48,480 --> 00:40:50,540 Turns out that when you introduce elections, you get 721 00:40:50,540 --> 00:40:51,790 very large changes. 722 00:40:54,240 --> 00:40:55,330 And this is not now. 723 00:40:55,330 --> 00:40:57,440 This started, I think, in the '80s. 724 00:40:57,440 --> 00:40:59,760 So when you start introducing elections, you 725 00:40:59,760 --> 00:41:03,570 get very large changes. 726 00:41:03,570 --> 00:41:04,980 Basically, two things happen. 727 00:41:04,980 --> 00:41:07,650 One is that the central policies that are unpopular 728 00:41:07,650 --> 00:41:09,040 are less enforced. 729 00:41:09,040 --> 00:41:14,150 So the one-child policy, which is the policy of not allowing 730 00:41:14,150 --> 00:41:17,960 people to have more than one child, is relaxed much faster 731 00:41:17,960 --> 00:41:21,920 than places which have elections. 732 00:41:21,920 --> 00:41:24,890 So somehow, despite the fact that the central government is 733 00:41:24,890 --> 00:41:29,240 all-powerful, when the village has an election, it does 734 00:41:29,240 --> 00:41:30,900 enforce different things. 735 00:41:30,900 --> 00:41:34,370 Second, one other very important part of Chinese 736 00:41:34,370 --> 00:41:36,470 policy is reallocation of farmland. 737 00:41:36,470 --> 00:41:40,970 Basically, what they were doing was, they would, for 738 00:41:40,970 --> 00:41:48,000 example, take farmland from farmers and build cities, or 739 00:41:48,000 --> 00:41:53,240 roads, or industrial estates, or special economic zones, and 740 00:41:53,240 --> 00:41:54,440 things like that. 741 00:41:54,440 --> 00:41:57,920 And basically, when the village had elected 742 00:41:57,920 --> 00:42:00,520 government, that stopped, or went down a lot. 743 00:42:00,520 --> 00:42:04,030 So you couldn't take land away from the farmers that easily. 744 00:42:04,030 --> 00:42:07,890 And this is where the all-powerful state which was 745 00:42:07,890 --> 00:42:10,050 supposed to have impunity. 746 00:42:10,050 --> 00:42:14,140 It was still true that you could make many changes within 747 00:42:14,140 --> 00:42:15,190 that system. 748 00:42:15,190 --> 00:42:22,330 And it's so striking, even in a system like China, how much 749 00:42:22,330 --> 00:42:27,530 responsiveness there is to local demands 750 00:42:27,530 --> 00:42:29,140 once you start democracy. 751 00:42:29,140 --> 00:42:32,080 So you might have thought that, well, the big 752 00:42:32,080 --> 00:42:36,610 institution is, there is no democracy in China. 753 00:42:36,610 --> 00:42:40,150 China is classified as a non-democratic state. 754 00:42:40,150 --> 00:42:42,820 Yet at some level, China is actually quite democratic at 755 00:42:42,820 --> 00:42:43,970 the local level. 756 00:42:43,970 --> 00:42:47,360 And that's a way to understand, also, why China 757 00:42:47,360 --> 00:42:52,540 has managed its conflicts relatively effectively. 758 00:42:52,540 --> 00:42:56,960 It's because there is actually pressure involved at the local 759 00:42:56,960 --> 00:42:59,640 level, which allow it to let off pressure. 760 00:42:59,640 --> 00:43:07,180 So there's a sense in which this is an example of why we 761 00:43:07,180 --> 00:43:13,750 shouldn't look at institutions from a thousand miles up, but 762 00:43:13,750 --> 00:43:16,216 we should look at how they function on the ground a 763 00:43:16,216 --> 00:43:17,466 little bit. 764 00:43:19,920 --> 00:43:23,490 And, I think, equally importantly-- 765 00:43:23,490 --> 00:43:30,580 so I just talked about why bad institutions in capitals, like 766 00:43:30,580 --> 00:43:33,450 lack of democracy, doesn't really necessarily always mean 767 00:43:33,450 --> 00:43:35,830 that is not democracy on the ground-- 768 00:43:35,830 --> 00:43:37,630 the opposite is also true. 769 00:43:37,630 --> 00:43:40,470 That you could have good institutions, meaning good 770 00:43:40,470 --> 00:43:46,070 electoral laws, nominally fair elections, but those elections 771 00:43:46,070 --> 00:43:48,810 may actually not be at all fair. 772 00:43:48,810 --> 00:43:50,470 And here's a nice example. 773 00:43:50,470 --> 00:43:58,280 In Brazil, there was a complicated 774 00:43:58,280 --> 00:44:01,890 paper ballot system. 775 00:44:01,890 --> 00:44:05,450 Brazil has, for the last many years, had free and fair 776 00:44:05,450 --> 00:44:07,690 elections, since about the mid-'80s. 777 00:44:10,680 --> 00:44:17,800 And yet, in the Brazilian electoral system, for a long 778 00:44:17,800 --> 00:44:20,140 time, people just didn't think about it. 779 00:44:20,140 --> 00:44:24,200 So you would have to write in the name of the person who you 780 00:44:24,200 --> 00:44:25,930 were electing, or the number. 781 00:44:25,930 --> 00:44:27,140 Either name or number. 782 00:44:27,140 --> 00:44:29,400 You have to say, Mr. seven, or something. 783 00:44:29,400 --> 00:44:33,800 And since many people were not very literate, when they write 784 00:44:33,800 --> 00:44:36,490 the names of somebody, it is illegible. 785 00:44:36,490 --> 00:44:39,130 So as a result, 11% of the names were 786 00:44:39,130 --> 00:44:43,100 rejected in any election. 787 00:44:43,100 --> 00:44:44,080 That's a lot, right? 788 00:44:44,080 --> 00:44:46,460 11% of the votes were rejected. 789 00:44:46,460 --> 00:44:48,760 That's a huge amount. 790 00:44:48,760 --> 00:44:53,830 So this was taking too much effort. 791 00:44:53,830 --> 00:44:56,300 Counting votes was too complicated. 792 00:44:56,300 --> 00:44:59,500 So they replaced it with electronic voting. 793 00:44:59,500 --> 00:45:03,230 So you to press a button. 794 00:45:03,230 --> 00:45:07,940 So that meant that the invalid votes went away. 795 00:45:07,940 --> 00:45:16,790 And as a result, the fraction of the elected representatives 796 00:45:16,790 --> 00:45:25,900 who were uneducated or poor went up. 797 00:45:25,900 --> 00:45:28,205 The poor were voting for people like themselves. 798 00:45:30,950 --> 00:45:40,600 As a result, health expenditures went up, and a 799 00:45:40,600 --> 00:45:44,170 bunch of pro-poor policies were implemented. 800 00:45:44,170 --> 00:45:49,690 So this is an example of where Brazil had good institutions 801 00:45:49,690 --> 00:45:54,280 in capitals to start with, but somehow those institutions 802 00:45:54,280 --> 00:45:57,310 weren't doing what they were supposed to do. 803 00:45:57,310 --> 00:46:00,640 So if you looked at it from a thousand feet up, you will see 804 00:46:00,640 --> 00:46:01,810 good institutions. 805 00:46:01,810 --> 00:46:04,810 But that doesn't mean that the institutions were actually 806 00:46:04,810 --> 00:46:08,260 doing what the were supposed to be doing. 807 00:46:08,260 --> 00:46:12,980 And so about 10% of the population, and a huge 808 00:46:12,980 --> 00:46:16,830 proportion of the poor, were disenfranchised. 809 00:46:16,830 --> 00:46:19,710 11% percent of the votes were eliminated. 810 00:46:19,710 --> 00:46:23,950 That means that, if you take the poor, probably 30, 40% of 811 00:46:23,950 --> 00:46:25,360 the poor were disenfranchised. 812 00:46:25,360 --> 00:46:26,970 They couldn't vote. 813 00:46:26,970 --> 00:46:30,120 That changed the nature of voting and changed the nature 814 00:46:30,120 --> 00:46:30,730 of outcomes. 815 00:46:30,730 --> 00:46:36,210 So it's a beautiful example of how something-- 816 00:46:36,210 --> 00:46:38,000 nobody thought about this. 817 00:46:38,000 --> 00:46:42,170 It's not that anybody had this as a conspiracy. 818 00:46:42,170 --> 00:46:43,990 This was not somebody's plan. 819 00:46:43,990 --> 00:46:45,970 It just was something that was just there. 820 00:46:45,970 --> 00:46:48,010 People thought paper ballots were fine. 821 00:46:48,010 --> 00:46:49,380 Nobody ever thought about it. 822 00:46:49,380 --> 00:46:52,520 Once you change the rules, and you mostly changed the rules 823 00:46:52,520 --> 00:46:55,850 not to fix it, but to just make it easier to count votes, 824 00:46:55,850 --> 00:46:57,850 it turned out that that had massive consequences. 825 00:47:00,890 --> 00:47:01,590 This is the picture. 826 00:47:01,590 --> 00:47:02,950 I'll just skip this. 827 00:47:08,300 --> 00:47:12,525 So here's another couple of nice experiments. 828 00:47:17,080 --> 00:47:27,860 So one argument that Paul Collier makes a lot of is, he 829 00:47:27,860 --> 00:47:32,030 says the reason why you have to invade this country is that 830 00:47:32,030 --> 00:47:36,100 these countries, have too much internal conflict, that they 831 00:47:36,100 --> 00:47:39,310 can't govern themselves because they're fighting each 832 00:47:39,310 --> 00:47:41,080 other all the time. 833 00:47:41,080 --> 00:47:43,730 Paul is very sympathetic to these people, you have to 834 00:47:43,730 --> 00:47:44,650 understand. 835 00:47:44,650 --> 00:47:50,930 He's a very, very soft-hearted colonist-- 836 00:47:50,930 --> 00:47:52,180 imperialist. 837 00:47:56,010 --> 00:47:57,340 I've known him for many, many years. 838 00:47:57,340 --> 00:48:01,070 And he's certainly not someone who's cynical about this. 839 00:48:01,070 --> 00:48:05,110 He certainly does really believe that these people are 840 00:48:05,110 --> 00:48:09,760 just caught in this eternal fight against each other. 841 00:48:09,760 --> 00:48:11,350 They hate each other's guts. 842 00:48:11,350 --> 00:48:13,890 And so, therefore, they can never govern. 843 00:48:13,890 --> 00:48:16,870 Whenever one group takes over, it starts 844 00:48:16,870 --> 00:48:18,340 beating up on the others. 845 00:48:18,340 --> 00:48:22,550 And that's why you always have a permanent state of civil war 846 00:48:22,550 --> 00:48:24,000 and conflict. 847 00:48:24,000 --> 00:48:26,700 They're just not capable of governing each other because 848 00:48:26,700 --> 00:48:30,300 there's so much fundamental hatred of each other. 849 00:48:30,300 --> 00:48:38,000 So to look at this, Leonard Wantchekon, who is a political 850 00:48:38,000 --> 00:48:41,690 scientist from Benin, did the following. 851 00:48:41,690 --> 00:48:49,630 Leonard was in the student movement in Benin in the late 852 00:48:49,630 --> 00:48:51,330 '70s, early '80s. 853 00:48:51,330 --> 00:48:56,290 And that's when, I think, Benin had a dictatorship. 854 00:48:56,290 --> 00:49:00,910 And so he went to jail with all the other people who were 855 00:49:00,910 --> 00:49:02,720 leading the student movement. 856 00:49:02,720 --> 00:49:06,270 And so he's very good friends with all the leading 857 00:49:06,270 --> 00:49:08,370 politicians in Benin now. 858 00:49:08,370 --> 00:49:10,310 So Benin is now a democracy. 859 00:49:10,310 --> 00:49:14,590 And in the democracy, he's good friends 860 00:49:14,590 --> 00:49:17,110 with all the leaders. 861 00:49:17,110 --> 00:49:19,660 So he basically talked them into doing something quite 862 00:49:19,660 --> 00:49:20,700 remarkable. 863 00:49:20,700 --> 00:49:26,610 His idea was, every leader, he basically got 864 00:49:26,610 --> 00:49:27,855 the following deal. 865 00:49:27,855 --> 00:49:33,240 He said, let's go to the area where you are really dominant. 866 00:49:33,240 --> 00:49:36,080 Go to the area where you are really powerful. 867 00:49:36,080 --> 00:49:37,860 So you're never going to lose that area. 868 00:49:37,860 --> 00:49:39,690 You're always going to win that area. 869 00:49:39,690 --> 00:49:42,910 So it doesn't really matter if you get a little less votes. 870 00:49:42,910 --> 00:49:46,210 So let's do an experiment where you send different 871 00:49:46,210 --> 00:49:52,250 messages in different areas. 872 00:49:52,250 --> 00:49:59,490 So within in that area, let's say there's this region like 873 00:49:59,490 --> 00:50:02,700 that, where this is where you're totally dominant. 874 00:50:02,700 --> 00:50:03,980 You're always going to win. 875 00:50:03,980 --> 00:50:07,930 So now let's chop this up into little pieces and randomly 876 00:50:07,930 --> 00:50:10,380 choose some of them. 877 00:50:10,380 --> 00:50:14,720 This is all within the area where you're sure to win. 878 00:50:14,720 --> 00:50:18,850 We're going to randomize the message you send. 879 00:50:18,850 --> 00:50:20,610 So there were two messages they were sending. 880 00:50:31,930 --> 00:50:44,600 So this is the message that primes people on their narrow 881 00:50:44,600 --> 00:50:46,140 tribal interests, if you like. 882 00:50:50,640 --> 00:50:53,965 This says, the Bariba people will get jobs. 883 00:50:56,750 --> 00:50:59,270 I am the representative of the Bariba people. 884 00:50:59,270 --> 00:51:03,470 This is what Paul Collier is particularly concerned with. 885 00:51:03,470 --> 00:51:07,290 So this is the same guy going to different in villages and 886 00:51:07,290 --> 00:51:11,190 giving different messages. 887 00:51:11,190 --> 00:51:14,080 The villages are randomly chosen, so you can compare 888 00:51:14,080 --> 00:51:16,690 what happens. 889 00:51:16,690 --> 00:51:18,380 And they were willing to do it because they 890 00:51:18,380 --> 00:51:19,850 were sure of winning. 891 00:51:19,850 --> 00:51:22,250 So they didn't have to worry about winning there. 892 00:51:22,250 --> 00:51:25,400 The question was many votes do they get. 893 00:51:25,400 --> 00:51:29,390 This was the Clientelist message, the message that I'll 894 00:51:29,390 --> 00:51:33,080 serve you guys, my people. 895 00:51:33,080 --> 00:51:34,880 That's what people are worried about. 896 00:51:34,880 --> 00:51:37,530 Here's the public policy message. 897 00:51:37,530 --> 00:51:40,050 Same person, different village, goes and delivers 898 00:51:40,050 --> 00:51:41,300 this message. 899 00:51:45,570 --> 00:51:49,410 We will fight corruption and promote peace among all ethnic 900 00:51:49,410 --> 00:51:52,600 groups in all regions of Benin. 901 00:51:52,600 --> 00:51:54,940 This was completely rigged. 902 00:51:54,940 --> 00:51:57,300 They were given these messages. 903 00:51:57,300 --> 00:51:59,210 They agreed to deliver it because they 904 00:51:59,210 --> 00:52:03,020 were friends of his. 905 00:52:03,020 --> 00:52:08,890 So this is what makes Paul collier worried. 906 00:52:08,890 --> 00:52:11,830 So this was the first experiment that Leonard did. 907 00:52:11,830 --> 00:52:15,520 So if you look at what happens to vote shares, if you go to 908 00:52:15,520 --> 00:52:19,240 the villages where you did the public policy message, you got 909 00:52:19,240 --> 00:52:20,830 only 59% of the vote. 910 00:52:20,830 --> 00:52:23,320 When you did Clientelist message, you 911 00:52:23,320 --> 00:52:26,460 got 79% of the vote. 912 00:52:26,460 --> 00:52:28,790 So this is what gets Paul Collier worried. 913 00:52:28,790 --> 00:52:32,540 He says it's not possible to have democracy in these places 914 00:52:32,540 --> 00:52:35,780 because people are so narrow-minded. 915 00:52:35,780 --> 00:52:37,750 Now what happened-- 916 00:52:37,750 --> 00:52:39,630 let me just go back-- 917 00:52:39,630 --> 00:52:40,880 is-- 918 00:52:48,300 --> 00:52:50,000 OK. 919 00:52:50,000 --> 00:52:50,670 Let me see. 920 00:52:50,670 --> 00:52:51,920 Where did it go? 921 00:53:02,390 --> 00:53:03,640 OK. 922 00:53:07,905 --> 00:53:11,100 Let me mention one more, and I'll come back to-- 923 00:53:13,726 --> 00:53:16,640 no, actually-- 924 00:53:16,640 --> 00:53:18,470 so let me show you one other experiment 925 00:53:18,470 --> 00:53:19,620 before I come to this. 926 00:53:19,620 --> 00:53:22,302 Because this is sort of interesting. 927 00:53:22,302 --> 00:53:25,280 I'll show you one more. 928 00:53:25,280 --> 00:53:31,370 So here's an experiment that we did in India, which is kind 929 00:53:31,370 --> 00:53:40,890 of the opposite of this Wantchekon experiment, and has 930 00:53:40,890 --> 00:53:42,300 the opposite result. 931 00:53:42,300 --> 00:53:47,330 And I'll try and interpret why. 932 00:53:52,510 --> 00:53:54,930 This is probably the most corrupt place in India. 933 00:54:00,540 --> 00:54:06,740 About 40% of the people who are elected from there have 934 00:54:06,740 --> 00:54:09,600 criminal charges against them. 935 00:54:09,600 --> 00:54:13,550 In this area, we did an experiment where we randomly 936 00:54:13,550 --> 00:54:14,770 chose villages. 937 00:54:14,770 --> 00:54:23,620 And in some of the villages, we basically showed up with a 938 00:54:23,620 --> 00:54:24,540 puppet show. 939 00:54:24,540 --> 00:54:28,410 Basically, we showed up with a puppet show. 940 00:54:28,410 --> 00:54:32,860 The point of the public show was to say, you should not 941 00:54:32,860 --> 00:54:35,540 vote based on your caste, which is the equivalent of 942 00:54:35,540 --> 00:54:36,750 ethnicity in India. 943 00:54:36,750 --> 00:54:40,740 You should vote based on who's going to do more 944 00:54:40,740 --> 00:54:42,450 development for you. 945 00:54:42,450 --> 00:54:43,730 So that was the message. 946 00:54:43,730 --> 00:54:44,980 It was a completely neutral message. 947 00:54:48,280 --> 00:54:57,965 And the neutral message was delivered by an NGO. 948 00:55:03,230 --> 00:55:09,170 When you look at what happened, the people who voted 949 00:55:09,170 --> 00:55:16,700 for their ethnic party went down by a quarter. 950 00:55:16,700 --> 00:55:19,040 If you just went and told people, don't do it, they 951 00:55:19,040 --> 00:55:20,300 don't do it. 952 00:55:20,300 --> 00:55:23,000 So this is sort of the opposite story, which is, in 953 00:55:23,000 --> 00:55:28,650 Benin, the experiment was that you went and told people, 954 00:55:28,650 --> 00:55:32,150 please vote for me based on my ethnicity. 955 00:55:32,150 --> 00:55:34,190 Then they seem to vote on ethnicity. 956 00:55:34,190 --> 00:55:35,410 If you tell people, don't vote on 957 00:55:35,410 --> 00:55:37,800 ethnicity, they do the opposite. 958 00:55:37,800 --> 00:55:39,360 Now, is Benin different? 959 00:55:39,360 --> 00:55:41,040 I'll come back to that later. 960 00:55:41,040 --> 00:55:44,560 But what this made us think about was, maybe the reason 961 00:55:44,560 --> 00:55:48,750 why you observe this ethnicity-based voting is not 962 00:55:48,750 --> 00:55:53,440 because people are passionate about their ethnic roots. 963 00:55:53,440 --> 00:55:54,300 Sometimes they are. 964 00:55:54,300 --> 00:55:56,460 But maybe a lot of people are not. 965 00:55:56,460 --> 00:55:59,400 But because they have no other information, maybe they don't 966 00:55:59,400 --> 00:56:01,170 know anything about the candidate. 967 00:56:01,170 --> 00:56:06,445 And therefore, if I have no idea who to vote for, there 968 00:56:06,445 --> 00:56:09,370 are three people running, I just vote for the one who has 969 00:56:09,370 --> 00:56:11,310 the same name as me. 970 00:56:11,310 --> 00:56:11,520 Why? 971 00:56:11,520 --> 00:56:14,370 Because I have no idea who they are. 972 00:56:14,370 --> 00:56:15,140 I've never met them. 973 00:56:15,140 --> 00:56:17,030 I have no idea what they believe in. 974 00:56:17,030 --> 00:56:21,390 I don't trust them to have told me the truth in any case. 975 00:56:21,390 --> 00:56:34,850 So do you actually get the right outcome? 976 00:56:34,850 --> 00:56:39,050 So maybe the reason why people vote based on ethnicity, 977 00:56:39,050 --> 00:56:47,620 whereas we think that somehow people to just have some 978 00:56:47,620 --> 00:56:52,330 biological or sociological hatred of those other people 979 00:56:52,330 --> 00:56:54,950 and they can't resist voting, in fact, maybe the reason why 980 00:56:54,950 --> 00:56:58,290 they vote for their own ethnicity is only because they 981 00:56:58,290 --> 00:57:00,570 have no better choices. 982 00:57:00,570 --> 00:57:02,830 They don't know anything about any of the candidates, so they 983 00:57:02,830 --> 00:57:05,340 might as well vote for their ethnicity. 984 00:57:05,340 --> 00:57:07,230 So this is what we were investigating. 985 00:57:07,230 --> 00:57:07,510 Keep going. 986 00:57:07,510 --> 00:57:08,283 Yeah. 987 00:57:08,283 --> 00:57:10,583 AUDIENCE: In the Benin experiment as well, it seemed 988 00:57:10,583 --> 00:57:12,227 like it was a way to target things. 989 00:57:12,227 --> 00:57:14,445 Like, I'm this ethnicity, and I'm going to get things for 990 00:57:14,445 --> 00:57:16,664 people of this ethnicity, versus, I'm going to do good 991 00:57:16,664 --> 00:57:20,115 things for everybody. 992 00:57:20,115 --> 00:57:22,590 I don't know if that would be so different than if somebody 993 00:57:22,590 --> 00:57:26,150 was running for Senate in the US and was like, I'm going to 994 00:57:26,150 --> 00:57:28,406 do good things for the people of Massachusetts, versus, I'm 995 00:57:28,406 --> 00:57:30,080 going to do good things for everyone in the US. 996 00:57:33,580 --> 00:57:35,230 PROFESSOR: So you may well be right. 997 00:57:35,230 --> 00:57:37,936 So that's another point. 998 00:57:37,936 --> 00:57:41,490 I'll come back to the Benin in a minute. 999 00:57:41,490 --> 00:57:45,100 I just wanted to show this experiment. 1000 00:57:45,100 --> 00:57:48,220 And I'll come back to the Benin one in a minute. 1001 00:57:48,220 --> 00:57:52,980 So based on this idea that maybe voters don't know who 1002 00:57:52,980 --> 00:57:55,650 they are voting for-- 1003 00:57:55,650 --> 00:57:58,500 and therefore, they just vote at random, and maybe they just 1004 00:57:58,500 --> 00:58:02,050 vote for the guy who has the right name or something-- 1005 00:58:02,050 --> 00:58:05,850 we did another experiment where we gave information 1006 00:58:05,850 --> 00:58:09,520 about the candidates to the voters. 1007 00:58:09,520 --> 00:58:17,460 We told them, this candidate works 1008 00:58:17,460 --> 00:58:18,720 harder than that candidate. 1009 00:58:18,720 --> 00:58:23,530 We showed newspapers printed people's performance. 1010 00:58:23,530 --> 00:58:27,780 And when you do that, you see that people do change their 1011 00:58:27,780 --> 00:58:30,190 voting patterns, suggesting that people don't really know 1012 00:58:30,190 --> 00:58:31,540 a lot about the candidates. 1013 00:58:31,540 --> 00:58:34,730 If you give them information, they react, which suggests 1014 00:58:34,730 --> 00:58:37,300 that information is valuable to them. 1015 00:58:37,300 --> 00:58:41,010 They don't really know what's going on. 1016 00:58:41,010 --> 00:58:41,360 OK. 1017 00:58:41,360 --> 00:58:46,660 Let me now go back to Benin, and I'll come back to this. 1018 00:58:46,660 --> 00:58:50,420 So here's what happened in Benin. 1019 00:58:50,420 --> 00:58:53,290 Wantchekon did a second experiment. 1020 00:58:53,290 --> 00:58:59,660 Second experiment was, instead of people just saying, I'm 1021 00:58:59,660 --> 00:59:05,750 going to do good for everybody, he actually had 1022 00:59:05,750 --> 00:59:09,790 them run a conference where, in the conference, they 1023 00:59:09,790 --> 00:59:12,780 discussed real policies. 1024 00:59:12,780 --> 00:59:16,450 So they had a long discussion of real, substantive policies. 1025 00:59:16,450 --> 00:59:19,520 Experts were invited. 1026 00:59:19,520 --> 00:59:25,500 And so parties took up the policies that came out of that 1027 00:59:25,500 --> 00:59:27,970 conference and put it on the platform. 1028 00:59:27,970 --> 00:59:31,120 These were well-worked-out policies, not claims about, 1029 00:59:31,120 --> 00:59:32,370 I'll do good for everybody. 1030 00:59:35,810 --> 00:59:37,430 So the only difference-- 1031 00:59:37,430 --> 00:59:41,080 then he does the same experiment as before in a 1032 00:59:41,080 --> 00:59:43,080 different election, later election. 1033 00:59:43,080 --> 00:59:50,790 He does the same experiment, which is telling some people, 1034 00:59:50,790 --> 00:59:54,280 either send an ethnic message or a non-ethnic message. 1035 01:00:01,870 --> 01:00:05,830 And now, when you go and tell I'm going to do good for 1036 01:00:05,830 --> 01:00:10,040 everybody, that actually goes the other way. 1037 01:00:10,040 --> 01:00:12,250 So when you say, I'm going to do good for everybody, but 1038 01:00:12,250 --> 01:00:15,510 it's backed with a real policy proposal, you get much more 1039 01:00:15,510 --> 01:00:18,410 positive results. 1040 01:00:18,410 --> 01:00:21,510 So it looks like the voters were actually rightly cynical 1041 01:00:21,510 --> 01:00:22,870 about these politicians. 1042 01:00:22,870 --> 01:00:26,310 They were claiming they're going to do stuff for 1043 01:00:26,310 --> 01:00:29,760 everybody, but nobody believed them, because they had not put 1044 01:00:29,760 --> 01:00:31,210 any content into it. 1045 01:00:31,210 --> 01:00:34,770 When you actually put content into what you're claiming, 1046 01:00:34,770 --> 01:00:37,030 voters react very differently. 1047 01:00:37,030 --> 01:00:39,660 They feel like there's something that's going on 1048 01:00:39,660 --> 01:00:41,870 that's maybe worth supporting. 1049 01:00:41,870 --> 01:00:47,310 So you see, instead of this strong reaction for the ethnic 1050 01:00:47,310 --> 01:00:49,750 point of view, now you get the opposite, which is a strong 1051 01:00:49,750 --> 01:00:53,408 reaction for the non-ethnic point of view. 1052 01:01:08,660 --> 01:01:10,350 Another example. 1053 01:01:10,350 --> 01:01:20,180 So the point I'm trying to make here is that in some 1054 01:01:20,180 --> 01:01:24,130 sense, what looks like structural 1055 01:01:24,130 --> 01:01:27,892 resistance to any change-- 1056 01:01:27,892 --> 01:01:30,820 countries are just hopeless, they have such ethnic conflict 1057 01:01:30,820 --> 01:01:32,760 that they can't do anything about it-- that might be true 1058 01:01:32,760 --> 01:01:34,210 under some situations. 1059 01:01:34,210 --> 01:01:44,090 But, for example, if you want to name countries which have 1060 01:01:44,090 --> 01:01:55,090 had some of the most serious ethnic violence in the last 20 1061 01:01:55,090 --> 01:01:59,440 years, they include many countries which are actually 1062 01:01:59,440 --> 01:02:03,460 very economically successful countries. 1063 01:02:03,460 --> 01:02:12,040 So it's not clear that that's something that the effects are 1064 01:02:12,040 --> 01:02:12,900 that strong. 1065 01:02:12,900 --> 01:02:15,960 But even if you believe it, it's not clear whether ethnic 1066 01:02:15,960 --> 01:02:21,690 conflicts are a result of other failures or the cause. 1067 01:02:21,690 --> 01:02:24,350 So in other words, is it the case that people are basically 1068 01:02:24,350 --> 01:02:27,390 cynical about the political system, and therefore they'll 1069 01:02:27,390 --> 01:02:33,450 vote for the person whose name is the same as theirs, or is 1070 01:02:33,450 --> 01:02:40,220 it really the case that they vote for people who have the 1071 01:02:40,220 --> 01:02:41,900 same name as theirs, and therefore the 1072 01:02:41,900 --> 01:02:44,600 political system fails? 1073 01:02:44,600 --> 01:02:47,150 That's the challenge in figuring out. 1074 01:02:47,150 --> 01:02:50,000 And at least there seems to be some data 1075 01:02:50,000 --> 01:02:51,940 which suggests that-- 1076 01:02:51,940 --> 01:02:54,170 this evidence that I've been talking about-- 1077 01:02:54,170 --> 01:02:58,240 suggests that it may well be that, to a certain extent, if 1078 01:02:58,240 --> 01:03:01,400 people are cynical about the political system, makes them 1079 01:03:01,400 --> 01:03:04,360 then vote more unreliably. 1080 01:03:04,360 --> 01:03:06,650 And if you could actually get the political system to be 1081 01:03:06,650 --> 01:03:10,760 more effective, then people would actually be less 1082 01:03:10,760 --> 01:03:13,420 inclined to vote on the ethnic basis. 1083 01:03:13,420 --> 01:03:17,380 So in other words, I'm trying to make the case that what 1084 01:03:17,380 --> 01:03:20,270 looks like-- 1085 01:03:20,270 --> 01:03:22,350 in Collier's view, the fundamental 1086 01:03:22,350 --> 01:03:26,040 problem is ethnic hatred. 1087 01:03:26,040 --> 01:03:30,400 Whereas it may well be that the fundamental problem is 1088 01:03:30,400 --> 01:03:34,610 just economic failure, and ethnic hatred or ethnic 1089 01:03:34,610 --> 01:03:36,900 conflicts are just a result of that. 1090 01:03:36,900 --> 01:03:40,970 And it's not clear which way that goes. 1091 01:03:40,970 --> 01:03:44,060 And in particular, it could be very innocent. 1092 01:03:44,060 --> 01:03:45,760 People just are completely-- 1093 01:03:45,760 --> 01:03:47,480 it's not that they hate anybody. 1094 01:03:47,480 --> 01:03:51,340 They just think, OK, well if I don't have any information, 1095 01:03:51,340 --> 01:03:53,900 I'm going to vote for the guy who has the same name as me. 1096 01:03:53,900 --> 01:03:56,030 That might be a significant part of what is 1097 01:03:56,030 --> 01:03:59,110 called ethnic voting. 1098 01:03:59,110 --> 01:04:02,650 Here's another instance that's interesting. 1099 01:04:02,650 --> 01:04:16,930 So going down this same path of making the point that it is 1100 01:04:16,930 --> 01:04:18,240 possible to-- 1101 01:04:20,920 --> 01:04:24,820 So in other words, I made two points. 1102 01:04:24,820 --> 01:04:30,210 One is that having good institutions is no guarantee 1103 01:04:30,210 --> 01:04:33,970 that you get good outcomes. 1104 01:04:33,970 --> 01:04:39,740 And the good institution at the 1,000-feet level is not a 1105 01:04:39,740 --> 01:04:42,840 guarantee that you get good outcomes on the ground level. 1106 01:04:42,840 --> 01:04:50,150 And conversely, if you look at it from 1,000 feet, what looks 1107 01:04:50,150 --> 01:04:55,970 like bad outcomes may not be necessarily some things that 1108 01:04:55,970 --> 01:04:58,610 you should just take as, well, it's not possible to do 1109 01:04:58,610 --> 01:04:59,630 anything in this country. 1110 01:04:59,630 --> 01:05:01,500 There may be lots of slack in the country. 1111 01:05:01,500 --> 01:05:04,890 Because, for example, even ethnic preferences, which is 1112 01:05:04,890 --> 01:05:08,150 seen as a fundamental constraint, may not be that 1113 01:05:08,150 --> 01:05:09,120 much of a constraint. 1114 01:05:09,120 --> 01:05:12,950 They may just be an outcome rather than a constraint. 1115 01:05:12,950 --> 01:05:20,560 To pursue that point, that it's often possible to have 1116 01:05:20,560 --> 01:05:26,670 substantial changes even when the economic incentives have 1117 01:05:26,670 --> 01:05:30,630 not been changed very much-- so the structures of power 1118 01:05:30,630 --> 01:05:33,550 have not changed, the institutions are the same-- 1119 01:05:33,550 --> 01:05:36,530 here's an example. 1120 01:05:36,530 --> 01:05:38,660 This is not someone saying whether the institutions are 1121 01:05:38,660 --> 01:05:42,480 good or bad to start with, but to say that it's possible to 1122 01:05:42,480 --> 01:05:44,770 change institutions at the margin, a 1123 01:05:44,770 --> 01:05:46,900 little, and get big effects. 1124 01:05:46,900 --> 01:05:48,150 So here's an example from Brazil. 1125 01:05:51,040 --> 01:05:51,990 Brazil is interesting. 1126 01:05:51,990 --> 01:05:54,190 The political system in Brazil is among the most interesting 1127 01:05:54,190 --> 01:05:55,580 in the world. 1128 01:05:55,580 --> 01:05:59,550 One thing they do is every month, 60 municipalities are 1129 01:05:59,550 --> 01:06:02,425 chosen at random, and their accounts are audited. 1130 01:06:08,120 --> 01:06:14,840 So you audit the accounts, and then the audits are 1131 01:06:14,840 --> 01:06:16,610 basically-- 1132 01:06:16,610 --> 01:06:17,860 there's a lottery. 1133 01:06:20,240 --> 01:06:21,070 Literally. 1134 01:06:21,070 --> 01:06:23,240 That's on TV. 1135 01:06:23,240 --> 01:06:25,860 They're doing a lottery. 1136 01:06:25,860 --> 01:06:29,020 The winner of the lottery doesn't get a lot of money. 1137 01:06:29,020 --> 01:06:31,160 The winner of the lottery gets audited. 1138 01:06:31,160 --> 01:06:33,156 So you don't want to do be the winner. 1139 01:06:36,430 --> 01:06:39,030 It's on TV, so it's very, very transparent. 1140 01:06:39,030 --> 01:06:42,130 Everybody can see that it is really a lottery. 1141 01:06:42,130 --> 01:06:44,570 That's a big advantage of lottery. 1142 01:06:44,570 --> 01:06:46,840 The audit is also given to the government and 1143 01:06:46,840 --> 01:06:50,090 disclosed to the media. 1144 01:06:50,090 --> 01:06:53,300 So there's a nice paper studying the 1145 01:06:53,300 --> 01:06:55,510 effect of these audits. 1146 01:06:55,510 --> 01:07:00,510 And basically what they do is, they compare places that were 1147 01:07:00,510 --> 01:07:03,650 audited just before the election and places that were 1148 01:07:03,650 --> 01:07:05,030 audited just after the elections. 1149 01:07:12,060 --> 01:07:25,250 And in particular, it turns out that if you look at the 1150 01:07:25,250 --> 01:07:31,950 average effect of being audited before or after 1151 01:07:31,950 --> 01:07:33,620 elections, there's not a big difference. 1152 01:07:33,620 --> 01:07:35,010 Why isn't there a big difference? 1153 01:07:35,010 --> 01:07:38,510 Well, turns out most people don't find out the results. 1154 01:07:38,510 --> 01:07:42,250 But if you look at the places which have a which have a 1155 01:07:42,250 --> 01:07:48,080 radio station, and compare the effect of being audited on 1156 01:07:48,080 --> 01:07:51,220 places which have a radio station with the effect of not 1157 01:07:51,220 --> 01:07:53,890 being audited on places which have a radio station, you see 1158 01:07:53,890 --> 01:07:55,140 a massive effect. 1159 01:07:57,620 --> 01:08:03,080 If you're corrupt, and you were a mayor from a place that 1160 01:08:03,080 --> 01:08:06,040 had a radio station, and you were audited before the 1161 01:08:06,040 --> 01:08:13,100 election, you are 25 percentage points less likely 1162 01:08:13,100 --> 01:08:22,030 to get elected than if you were a non-corrupt mayor from 1163 01:08:22,030 --> 01:08:26,560 a place which has a radio station and which was audited 1164 01:08:26,560 --> 01:08:27,800 before the election. 1165 01:08:27,800 --> 01:08:30,940 So the difference in election probability-- 1166 01:08:35,490 --> 01:08:45,819 So what we're doing here is comparing places that were 1167 01:08:45,819 --> 01:08:48,460 audited before the election and after the election. 1168 01:08:48,460 --> 01:08:51,510 And we're comparing good mayors with bad mayors. 1169 01:08:51,510 --> 01:08:55,069 So good mayors are much more likely to get elected in 1170 01:08:55,069 --> 01:09:00,500 places where the audit was before the elections. 1171 01:09:00,500 --> 01:09:03,060 Whereas there's little difference between good mayors 1172 01:09:03,060 --> 01:09:06,800 and bad mayors if the audit was to 1173 01:09:06,800 --> 01:09:08,080 happen after the elections. 1174 01:09:08,080 --> 01:09:12,990 So just this act of publishing the audits seems to have 1175 01:09:12,990 --> 01:09:15,040 substantially changed the incentive for 1176 01:09:15,040 --> 01:09:16,390 corruption in Brazil. 1177 01:09:16,390 --> 01:09:20,670 So this is an example of something where the system has 1178 01:09:20,670 --> 01:09:21,670 not changed. 1179 01:09:21,670 --> 01:09:23,470 The whole system is the same. 1180 01:09:23,470 --> 01:09:26,279 Brazilian democracy has not changed a lot. 1181 01:09:26,279 --> 01:09:29,560 But just the fact that they instituted these audits means 1182 01:09:29,560 --> 01:09:31,769 that being corrupt is much more costly. 1183 01:09:36,240 --> 01:09:41,380 This is a picture which makes that point. 1184 01:09:41,380 --> 01:09:47,649 The blue is those who the audit was after the election. 1185 01:09:47,649 --> 01:09:50,370 And being corrupt has no effect on 1186 01:09:50,370 --> 01:09:51,620 your reelection rate. 1187 01:09:54,860 --> 01:10:06,320 Whereas, if you had zero corrupt violations and you had 1188 01:10:06,320 --> 01:10:13,630 a pre-election audit, you were 35 percentage points-- 1189 01:10:13,630 --> 01:10:14,330 I said 25-- 1190 01:10:14,330 --> 01:10:19,060 35 percentage points less likely to be elected then if 1191 01:10:19,060 --> 01:10:23,700 you had three or more, if there was a 1192 01:10:23,700 --> 01:10:25,370 pre-election audit. 1193 01:10:25,370 --> 01:10:27,300 If you have a post-election audit, it doesn't matter, 1194 01:10:27,300 --> 01:10:30,510 because the election is done. 1195 01:10:30,510 --> 01:10:33,550 So getting information has huge effects here. 1196 01:10:33,550 --> 01:10:37,010 These effects are on a magnitude which-- 1197 01:10:37,010 --> 01:10:44,140 clearly, you go from having 50% chance of being elected to 1198 01:10:44,140 --> 01:10:47,780 being a 20% chance of being elected, just when you publish 1199 01:10:47,780 --> 01:10:49,030 the results. 1200 01:10:51,780 --> 01:10:55,320 So again, making the point that small changes in the 1201 01:10:55,320 --> 01:10:58,940 institutional frame, the little institutions, changing 1202 01:10:58,940 --> 01:11:01,915 them a little might actually give you big effects. 1203 01:11:14,600 --> 01:11:18,190 So, so far, what I've been trying to say is a bit that, 1204 01:11:18,190 --> 01:11:22,300 even within bad institutional frames, or without changing 1205 01:11:22,300 --> 01:11:25,510 the institutional framework generally, you can make good 1206 01:11:25,510 --> 01:11:28,010 things happen. 1207 01:11:28,010 --> 01:11:30,240 That doesn't mean that good things are guaranteed to 1208 01:11:30,240 --> 01:11:32,760 happen-- whenever you try things, good things happen. 1209 01:11:32,760 --> 01:11:34,320 That's not at all what I'm saying. 1210 01:11:34,320 --> 01:11:39,930 So I'm saying that what's true about economy at the 1211 01:11:39,930 --> 01:11:47,340 1,000-feet level is entirely one thing. 1212 01:11:47,340 --> 01:11:52,660 And then there is possibly what's irrelevant at the 1213 01:11:52,660 --> 01:11:55,430 10-feet level, and those things are only weakly 1214 01:11:55,430 --> 01:11:56,680 correlated. 1215 01:11:58,330 --> 01:12:03,020 So in other words, things fail at the ground level whether or 1216 01:12:03,020 --> 01:12:06,470 not you have good institutional structures. 1217 01:12:06,470 --> 01:12:11,330 So let me give you an example of where the institutional 1218 01:12:11,330 --> 01:12:17,520 structures were all good, and the system completely failed. 1219 01:12:17,520 --> 01:12:23,110 So we did an experiment where there was a district 1220 01:12:23,110 --> 01:12:28,350 administration in India worked with an NGO to set up 1221 01:12:28,350 --> 01:12:34,050 monitoring system to make sure that nurses come to work. 1222 01:12:34,050 --> 01:12:36,290 Nurses often don't come to work. 1223 01:12:36,290 --> 01:12:38,460 You saw that, I think, at the beginning of the 1224 01:12:38,460 --> 01:12:40,090 semester with Esther. 1225 01:12:40,090 --> 01:12:44,150 And this is a big concern for the health system. 1226 01:12:44,150 --> 01:12:48,330 Because you show up, nobody's there, that's not so good. 1227 01:12:48,330 --> 01:12:52,210 So the way it was implemented was that there was a date and 1228 01:12:52,210 --> 01:12:55,770 time stamp given to every nurse. 1229 01:12:55,770 --> 01:12:58,450 This was just on Mondays when she was required 1230 01:12:58,450 --> 01:12:59,920 to be in the center. 1231 01:12:59,920 --> 01:13:04,730 She was required to stamp her presence on something that was 1232 01:13:04,730 --> 01:13:05,850 stuck to the wall. 1233 01:13:05,850 --> 01:13:11,430 So she would have stamped it there so that people could 1234 01:13:11,430 --> 01:13:13,000 check whether she was there or not. 1235 01:13:13,000 --> 01:13:15,000 So she had to stamp it on the wall. 1236 01:13:15,000 --> 01:13:18,420 So she had to be there. 1237 01:13:18,420 --> 01:13:24,060 And this NGO was in charge of collecting the stamps and 1238 01:13:24,060 --> 01:13:25,930 counting them up and giving them to the government. 1239 01:13:31,250 --> 01:13:33,530 So the district administration was very 1240 01:13:33,530 --> 01:13:35,180 taking this very seriously. 1241 01:13:35,180 --> 01:13:40,120 They announced that anybody who doesn't come to work at 1242 01:13:40,120 --> 01:13:44,630 least 50% of the time will get fired. 1243 01:13:44,630 --> 01:13:50,240 So there was a quite major top-down attempt to reform. 1244 01:13:50,240 --> 01:13:54,510 So the institutional frame was doing everything 1245 01:13:54,510 --> 01:13:57,340 it needed to do. 1246 01:13:57,340 --> 01:13:59,650 So when this was announced, immediately-- 1247 01:14:03,640 --> 01:14:07,240 just compare the dashed lines to the dashed lines. 1248 01:14:07,240 --> 01:14:09,200 I didn't tell you the story of the solid lines. 1249 01:14:09,200 --> 01:14:12,690 So just compare the dashed line to the dashed line. 1250 01:14:12,690 --> 01:14:17,520 You see there's a very major effect. 1251 01:14:17,520 --> 01:14:24,600 At the beginning, the presence of the nurse goes from 15%-- 1252 01:14:24,600 --> 01:14:26,500 15%. 1253 01:14:26,500 --> 01:14:28,050 They don't show up at all-- 1254 01:14:28,050 --> 01:14:30,200 to about 55%. 1255 01:14:30,200 --> 01:14:31,800 It goes up massively. 1256 01:14:31,800 --> 01:14:35,750 When you announce this program, it goes up massively. 1257 01:14:35,750 --> 01:14:38,270 They start to come to work. 1258 01:14:38,270 --> 01:14:39,330 And that's not surprising. 1259 01:14:39,330 --> 01:14:40,880 You told them. 1260 01:14:40,880 --> 01:14:44,740 And then what happens, you see those curves getting closer 1261 01:14:44,740 --> 01:14:48,620 and closer, and they finally cross. 1262 01:14:48,620 --> 01:14:54,290 So after a while, it turns out that in the places where the 1263 01:14:54,290 --> 01:14:56,710 nurses were threatened with losing their jobs, they're 1264 01:14:56,710 --> 01:14:59,450 coming to work less than where they were not threatened. 1265 01:15:02,110 --> 01:15:03,250 So what happened? 1266 01:15:03,250 --> 01:15:04,500 Why did it get reversed? 1267 01:15:12,110 --> 01:15:14,330 So what happened was very interesting. 1268 01:15:14,330 --> 01:15:16,845 So two things happened. 1269 01:15:16,845 --> 01:15:22,680 The nurse has an immediate boss whose job is to decide 1270 01:15:22,680 --> 01:15:25,120 whether the nurse needs to come to work or not. 1271 01:15:25,120 --> 01:15:25,550 Why? 1272 01:15:25,550 --> 01:15:28,140 Because the nurses have things like training. 1273 01:15:28,140 --> 01:15:30,700 So some days, they're not required to come to work. 1274 01:15:30,700 --> 01:15:32,160 Those are called exempt days. 1275 01:15:35,020 --> 01:15:40,570 And then the date and time stamp had to work in order to 1276 01:15:40,570 --> 01:15:41,610 monitor the nurses. 1277 01:15:41,610 --> 01:15:45,880 So the date and time stamp got broken, you 1278 01:15:45,880 --> 01:15:47,150 couldn't monitor them. 1279 01:15:47,150 --> 01:15:50,190 Now there was a rule which said that, if it's broken, 1280 01:15:50,190 --> 01:15:51,490 they would have bring it in to the 1281 01:15:51,490 --> 01:15:54,230 center and get it replaced. 1282 01:15:54,230 --> 01:15:58,310 But still, once it gets broken, it's less easy to 1283 01:15:58,310 --> 01:15:59,320 monitor them. 1284 01:15:59,320 --> 01:16:03,880 So what happened was very simple. 1285 01:16:03,880 --> 01:16:07,580 Absence did not go up. 1286 01:16:07,580 --> 01:16:09,300 It's not that the absence went up. 1287 01:16:09,300 --> 01:16:11,110 The exemptions went up. 1288 01:16:11,110 --> 01:16:14,810 Meaning that their immediate boss basically declared, fine, 1289 01:16:14,810 --> 01:16:16,190 you don't have to come to work. 1290 01:16:16,190 --> 01:16:17,610 From now on, you're not absent. 1291 01:16:17,610 --> 01:16:19,620 You're exempt. 1292 01:16:19,620 --> 01:16:22,810 Meaning the boss said, I don't care if you 1293 01:16:22,810 --> 01:16:24,190 don't come to work. 1294 01:16:24,190 --> 01:16:28,480 So I'll give you an official excuse to not come to work. 1295 01:16:28,480 --> 01:16:29,660 That was part of what happened. 1296 01:16:29,660 --> 01:16:31,210 Part of what happened was that they started 1297 01:16:31,210 --> 01:16:32,480 breaking the machines. 1298 01:16:32,480 --> 01:16:35,620 The machines started breaking in increasing frequency. 1299 01:16:35,620 --> 01:16:38,750 And the bosses who should have complained 1300 01:16:38,750 --> 01:16:40,500 about that didn't complain. 1301 01:16:40,500 --> 01:16:47,770 So the machines kept breaking and nobody was objecting, so 1302 01:16:47,770 --> 01:16:49,750 the whole program collapsed. 1303 01:16:49,750 --> 01:16:53,152 Now, why did this happen? 1304 01:17:10,540 --> 01:17:19,060 So clearly, this was a result of the fact that the nurse and 1305 01:17:19,060 --> 01:17:21,130 her immediate boss were colluding. 1306 01:17:21,130 --> 01:17:24,180 That's not surprising. 1307 01:17:24,180 --> 01:17:28,370 But the reason why it actually happened was not because the 1308 01:17:28,370 --> 01:17:30,400 institutional frame was bad. 1309 01:17:30,400 --> 01:17:33,200 In some sense, there is democracy. 1310 01:17:33,200 --> 01:17:36,000 There was supervisor who wanted it to happen. 1311 01:17:36,000 --> 01:17:38,640 There was an NGO that was implementing it. 1312 01:17:38,640 --> 01:17:44,300 The reason why didn't work was mostly because there was no 1313 01:17:44,300 --> 01:17:45,910 demand for it to work. 1314 01:17:45,910 --> 01:17:48,280 It wasn't so much that the institutional 1315 01:17:48,280 --> 01:17:49,275 frame wasn't there. 1316 01:17:49,275 --> 01:17:53,380 It was just that, basically, people figured out that if it 1317 01:17:53,380 --> 01:17:57,680 didn't work, nobody would complain. 1318 01:17:57,680 --> 01:17:59,590 And why would nobody complain? 1319 01:17:59,590 --> 01:18:01,350 Well, you saw this before. 1320 01:18:01,350 --> 01:18:05,500 Nobody ever showed up in these health centers in any case. 1321 01:18:05,500 --> 01:18:07,230 So there was no one to complain. 1322 01:18:07,230 --> 01:18:11,300 So there's no political compulsion to fix it. 1323 01:18:11,300 --> 01:18:16,380 And that meant that at some level, there was no pressure 1324 01:18:16,380 --> 01:18:18,110 to make it work. 1325 01:18:18,110 --> 01:18:24,640 This is what sometimes we call triple-i. 1326 01:18:28,600 --> 01:18:31,690 In other words, the reason why this intervention didn't work 1327 01:18:31,690 --> 01:18:35,080 is not because the institutional frame was not 1328 01:18:35,080 --> 01:18:38,240 working, but because there was no one on the ground who 1329 01:18:38,240 --> 01:18:43,820 actually had any stake in making it work. 1330 01:18:43,820 --> 01:18:45,965 Nobody went to the health center, so nobody complained. 1331 01:18:50,490 --> 01:18:54,580 The system was designed with the idea that these nurses 1332 01:18:54,580 --> 01:18:59,450 have some internal desire to serve of the people. 1333 01:18:59,450 --> 01:19:06,160 So the system has no effective culture of making sure that 1334 01:19:06,160 --> 01:19:10,130 the nurses one want to come to work. 1335 01:19:10,130 --> 01:19:12,020 They never invested in it. 1336 01:19:12,020 --> 01:19:19,450 And finally, there was no one had thought about the fact 1337 01:19:19,450 --> 01:19:32,890 that there were these exempt days, and that these exempt 1338 01:19:32,890 --> 01:19:34,230 days were not regulated. 1339 01:19:34,230 --> 01:19:37,580 So that it was mostly just the fact that, when this 1340 01:19:37,580 --> 01:19:40,740 intervention was introduced, nobody had actually put enough 1341 01:19:40,740 --> 01:19:45,280 thought into how the system would react. 1342 01:19:45,280 --> 01:19:50,780 And so it was not so much that the institutional frame was 1343 01:19:50,780 --> 01:19:53,220 particularly bad here. 1344 01:19:53,220 --> 01:19:55,150 Other interventions do work. 1345 01:19:55,150 --> 01:20:00,380 It's just that this was a system where there was no one 1346 01:20:00,380 --> 01:20:04,170 who actually had thought about how to design an intervention 1347 01:20:04,170 --> 01:20:08,130 that would work, and therefore it didn't work. 1348 01:20:08,130 --> 01:20:13,570 Indeed, when we went back, we figured out why 1349 01:20:13,570 --> 01:20:14,900 you get this crossing. 1350 01:20:14,900 --> 01:20:16,800 Why do you think we get this crossing? 1351 01:20:16,800 --> 01:20:20,420 Why do you think the nurses came less when they were 1352 01:20:20,420 --> 01:20:22,234 incentivized to come? 1353 01:20:34,534 --> 01:20:37,832 AUDIENCE: I think it's a situation where there were new 1354 01:20:37,832 --> 01:20:39,296 measure put in place. 1355 01:20:39,296 --> 01:20:41,614 Once it became clear that those measure weren't going to 1356 01:20:41,614 --> 01:20:45,152 be enforced at all long-term, it was an even clearer 1357 01:20:45,152 --> 01:20:46,970 indication that no one would check up on them. 1358 01:20:46,970 --> 01:20:49,480 It's one thing to have a policy that in theory exists, 1359 01:20:49,480 --> 01:20:51,418 and no one tries to ever enforce it. 1360 01:20:51,418 --> 01:20:58,303 But if then they try to institute this new policy and 1361 01:20:58,303 --> 01:20:59,642 indeed, it was very [INAUDIBLE] that there was no 1362 01:20:59,642 --> 01:21:01,911 enforcement, that sends an even stronger message that you 1363 01:21:01,911 --> 01:21:03,480 can shirk out your responsibilities. 1364 01:21:03,480 --> 01:21:03,940 PROFESSOR: Right. 1365 01:21:03,940 --> 01:21:07,770 So they realized that their bosses were very willing to 1366 01:21:07,770 --> 01:21:10,550 collude with them to not let them work. 1367 01:21:17,690 --> 01:21:21,710 So because they had had a policy which required their 1368 01:21:21,710 --> 01:21:25,790 bosses to enforce, now they could learn that the bosses 1369 01:21:25,790 --> 01:21:27,090 had no interest in enforcing. 1370 01:21:27,090 --> 01:21:28,590 Before this, they didn't know that. 1371 01:21:28,590 --> 01:21:30,840 So in the control villages, they don't yet know that the 1372 01:21:30,840 --> 01:21:32,610 bosses don't care. 1373 01:21:32,610 --> 01:21:35,090 In the treatment villages, they really learned that the 1374 01:21:35,090 --> 01:21:36,730 bosses don't give a damn whether they 1375 01:21:36,730 --> 01:21:37,960 come to work or not. 1376 01:21:37,960 --> 01:21:39,280 So they stopped coming to work. 1377 01:21:49,530 --> 01:21:55,445 So whether these institutions at the 1,000-feet level work 1378 01:21:55,445 --> 01:21:56,860 or don't work-- 1379 01:21:56,860 --> 01:21:59,310 I mean, there's a democracy here. 1380 01:21:59,310 --> 01:22:00,580 Lots of people vote. 1381 01:22:00,580 --> 01:22:02,330 There is political competition. 1382 01:22:02,330 --> 01:22:04,660 All of those things that are supposed to make these things 1383 01:22:04,660 --> 01:22:06,510 work are all present. 1384 01:22:06,510 --> 01:22:12,010 The reason why it doesn't work is simply, at the micro level, 1385 01:22:12,010 --> 01:22:14,380 the incentive to make it work isn't there. 1386 01:22:14,380 --> 01:22:17,780 Because people don't go to the health center, so nobody 1387 01:22:17,780 --> 01:22:21,136 actually demands that the health centers work. 1388 01:22:21,136 --> 01:22:23,122 AUDIENCE: So actually, more people started to go to the 1389 01:22:23,122 --> 01:22:23,430 health centers. 1390 01:22:23,430 --> 01:22:26,260 And yet there were few people [INAUDIBLE]? 1391 01:22:26,260 --> 01:22:27,510 PROFESSOR: No. 1392 01:22:29,130 --> 01:22:32,370 Basically, people have checked out of the system. 1393 01:22:32,370 --> 01:22:36,830 So more people use the health center, but only by accident, 1394 01:22:36,830 --> 01:22:37,750 when it's open. 1395 01:22:37,750 --> 01:22:41,650 So the probability, per day, more people are not present. 1396 01:22:41,650 --> 01:22:45,310 It's open more often, but it doesn't have any effect on who 1397 01:22:45,310 --> 01:22:47,310 gets treated per day. 1398 01:22:47,310 --> 01:22:48,180 Yeah. 1399 01:22:48,180 --> 01:22:48,783 Alyssa? 1400 01:22:48,783 --> 01:22:51,148 AUDIENCE: The fact that nobody is using these can't be that 1401 01:22:51,148 --> 01:22:53,354 there wouldn't be a demand for health care if they 1402 01:22:53,354 --> 01:22:53,513 functioned. 1403 01:22:53,513 --> 01:22:56,840 It's itself an outcome of the fact that-- 1404 01:22:56,840 --> 01:22:58,090 PROFESSOR: Absolutely. 1405 01:22:59,910 --> 01:23:04,220 But this attempt to fix it came from a bureaucratic 1406 01:23:04,220 --> 01:23:07,150 impulse which said, have health center, 1407 01:23:07,150 --> 01:23:08,560 will make it work. 1408 01:23:08,560 --> 01:23:13,530 And they just did not think about what was required to 1409 01:23:13,530 --> 01:23:15,650 make it work, which is to get a demand for it. 1410 01:23:15,650 --> 01:23:17,110 Otherwise, it's never going to work. 1411 01:23:17,110 --> 01:23:20,650 The political system only provides incentives for things 1412 01:23:20,650 --> 01:23:21,380 people want. 1413 01:23:21,380 --> 01:23:23,670 If people don't want it, the political system doesn't 1414 01:23:23,670 --> 01:23:25,610 provide incentives. 1415 01:23:25,610 --> 01:23:31,940 It was a classic example of trying to fix institutions 1416 01:23:31,940 --> 01:23:34,420 without really having understood how the 1417 01:23:34,420 --> 01:23:36,082 institutions work. 1418 01:23:36,082 --> 01:23:39,140 So I'll continue with this next time.