1 00:00:00,000 --> 00:00:04,480 The following is part of Cornell  Contemporary China Initiative lecture series   2 00:00:04,480 --> 00:00:08,240 under the Cornell East Asia Program.  The arguments and viewpoints of this   3 00:00:08,240 --> 00:00:11,200 talk belong solely to the  speaker. We hope you enjoy. 4 00:00:11,840 --> 00:00:16,640 And uh for our last talk today we have  our very own Professor Shanjun Li,   5 00:00:17,440 --> 00:00:26,160 who joined the Dyson school in 2011 as an  economist, and he had been working prior to that   6 00:00:27,120 --> 00:00:35,200 with a think tank in D.C., yes, had been there  several years since finishing his PhD in economics   7 00:00:35,200 --> 00:00:43,840 at Duke. So we're very happy to have him here.  Since arriving here, he has of course published uh   8 00:00:43,840 --> 00:00:51,040 many interesting things but also has recently  uh co-founded, together with Panle Jia,   9 00:00:51,040 --> 00:00:58,000 the CICER, the Cornell Institute for China  Economic Research, which is often partnered   10 00:00:58,000 --> 00:01:04,320 with us, CCCI, with this lecture series. Uh so,  without uh further ado, let us welcome Shanjun. 11 00:01:05,440 --> 00:01:11,920 Alright, and thanks to you all for coming,  alright, um so I have to warn you today I'm   12 00:01:11,920 --> 00:01:17,120 going to show a lot of things that that are quite  depressing and you know in this kind of weather,   13 00:01:17,120 --> 00:01:21,600 in this type of political environment, uh but in  the end there will be something more positive,   14 00:01:21,600 --> 00:01:28,160 so stick with me till the end if you can. Okay,  um, so I'm going to talk about environmental and   15 00:01:28,160 --> 00:01:34,240 energy challenges and policy options in China.  Okay, in 2005, China was the largest energy   16 00:01:34,240 --> 00:01:40,640 consumer accounting for 24 percent of world total  energy consumption, okay? 24 percent. I want you   17 00:01:40,640 --> 00:01:46,240 to think about these numbers really for a moment.  By far the largest coal consumer: 50 percent of   18 00:01:46,240 --> 00:01:52,480 world total coal consumption. The largest CO2  emitter: 30 percent of world total. The largest   19 00:01:52,480 --> 00:02:00,000 automobile market, 26 percent of world total  automobile sales, okay? So these shares are only   20 00:02:00,000 --> 00:02:05,520 getting bigger, that is, these things are really  have very big implications if you think about it.   21 00:02:05,520 --> 00:02:12,400 Think about energy demand, right? We know energy demand, for example, oil is inherently now   22 00:02:12,400 --> 00:02:17,600 actually world market, so anything happens in  China, in terms of automobile demand, in terms   23 00:02:17,600 --> 00:02:24,720 of gasoline demand, is going to put pressure or  affect oil market, affects the price, affect the   24 00:02:24,720 --> 00:02:31,040 price [vp, unintelligible] here in the United  States. Okay. Think about CO2 emission, right?   25 00:02:31,040 --> 00:02:38,400 CO2 we know is a global pollutant. A ton of CO2  emitted from China is no different from one-time   26 00:02:38,400 --> 00:02:43,600 CO2 emitted in the U.S. in terms of its impact  on climate, in terms of impact for the world,   27 00:02:44,320 --> 00:02:50,800 okay? So we care about not only CO2 emissions  from here but also in China and elsewhere, right?   28 00:02:52,720 --> 00:02:58,000 So domestic policy in China really have global  impacts and that's why we really need to,   29 00:02:58,000 --> 00:03:01,840 we really should understand what's  happening in China, and really,   30 00:03:01,840 --> 00:03:08,640 that can inform us in terms of policy making in  the U.S. Because of these numbers that I mentioned,   31 00:03:09,360 --> 00:03:16,160 okay, air pollution and traffic congestion really  uh have become the most pressing challenges in   32 00:03:16,160 --> 00:03:21,120 major urban areas in China. I'm going to talk  about these two issues and actually going   33 00:03:21,120 --> 00:03:27,840 to focus on traffic congestion in the later  half of of the talk. Um in turn these were,   34 00:03:27,840 --> 00:03:34,800 air pollution traffic we know actually affect  quality of lives in urban areas, it really affects   35 00:03:35,520 --> 00:03:42,960 firm location, affects labor productivity,  affects human capital investment, and affects,   36 00:03:42,960 --> 00:03:50,240 in the end, economic growth in the long run.  Okay, so this shows a GDP growth on the right   37 00:03:50,240 --> 00:03:59,840 axis and energy consumption on the left. For  the past um 50, actually more than 50 years,   38 00:04:01,040 --> 00:04:07,920 can look at the increase in GDP that, that is  this black line here, so we all know annual   39 00:04:07,920 --> 00:04:15,920 average is about 10 percent during the last 35  years. If you look at uh coal consumption this   40 00:04:15,920 --> 00:04:22,240 this uh gray area it's increasing roughly the same  speed. If you look at oil consumption increase,   41 00:04:22,240 --> 00:04:30,240 look at gasoline, sorry this is natural  gas, and hydro and nuclear is the blue,   42 00:04:30,240 --> 00:04:37,280 okay? so you can look at this this, the  majority of, of the um fuel source is coal. 43 00:04:39,600 --> 00:04:50,800 So in 2014, 89 percent of energy uh was  from fossil fuel including: coal 66 percent,   44 00:04:52,160 --> 00:04:58,240 oil 18 percent, this is really mainly for  automobiles, okay, very little uh gas, uh   45 00:04:58,240 --> 00:05:04,240 natural gas, if you look at in look at U.S., this  will be for coal, would be roughly 30 percent,   46 00:05:04,800 --> 00:05:12,080 uh natural gas would be would be larger share,  oil would be I think roughly the same here,   47 00:05:12,080 --> 00:05:19,200 okay? So fossil fuel, especially coal, is the  dominant source, and if you look at all the uh   48 00:05:19,200 --> 00:05:24,400 resources here coal coal  is the dirtiest one, okay? By far the dirtiest   49 00:05:24,400 --> 00:05:28,880 one. That that was the consumption. What about  energy balance? So this is the difference between   50 00:05:29,440 --> 00:05:34,480 what China produced, what China consumed, so  this, these are negative. We look at this,   51 00:05:34,480 --> 00:05:39,440 for the majority of this, actually, these are  negative, meaning China has been importing oil,   52 00:05:40,320 --> 00:05:48,640 okay? And also increasingly uh coal, natural gas,  from other countries, okay? That's why I said in   53 00:05:48,640 --> 00:05:54,000 the beginning that domestic demand in China  actually has important implications for the   54 00:05:54,000 --> 00:06:01,680 world market as a whole. So this is new automobile  sales in China. This is uh annual sales uh in   55 00:06:01,680 --> 00:06:09,440 China in 2001, was less than 2.5 million, now um  it's here, actually 2015 the total new vehicle   56 00:06:09,440 --> 00:06:16,800 sales was over 21 million units in China. This  is for the U.S., China surpassed the U.S. in 2009   57 00:06:16,800 --> 00:06:25,920 to become the largest automobile market, okay? So  the dramatic increase here in uh in coal, sorry in   58 00:06:26,720 --> 00:06:33,840 oil consumption was largely driven by the  increase in vehicle ownership in China. 59 00:06:36,000 --> 00:06:45,200 Alright. So this is crude oil consumption uh  in China in orange, and the world blue. This   60 00:06:45,200 --> 00:06:50,960 is relative to 2001 level so everything  is relative to 2001. If you look at, if you   61 00:06:50,960 --> 00:06:57,600 look at the increase in oil consumption in  China and then the blue bar are the world total   62 00:06:57,600 --> 00:07:03,840 consumption relative to 2001 again, if you look  at this uh orange bar relative to the blue bar,   63 00:07:03,840 --> 00:07:09,360 China accounted for 46 percent increasing  oil consumption during this data period.   64 00:07:10,400 --> 00:07:18,160 So this is CO2 emission again relative to 2001  level, orange is for China, blue is for the world,   65 00:07:18,160 --> 00:07:25,840 during this period 2001-2013, the increase in  CO2 emission in China accounted for 60 percent   66 00:07:25,840 --> 00:07:34,240 of growth in global total. So the next, I'm going  to show you um a few slides about environmental   67 00:07:34,240 --> 00:07:40,400 quality or environmental issues in China. And  before I show them, and we all understand there's   68 00:07:41,920 --> 00:07:47,280 big issues, environmental degradation is a big  problem there, and if you think about why we   69 00:07:47,280 --> 00:07:53,360 had these issues, uh I call this here, I call  them, really, this is a perfect storm for China,   70 00:07:53,360 --> 00:08:02,160 so several really important factors combined  together that contributed to the really dramatic   71 00:08:02,160 --> 00:08:06,800 degradation in terms of environmental  quality in China during the past 15 years,   72 00:08:06,800 --> 00:08:12,080 okay? So the first one, we all know, is the in  unprecedented growth in the industry sector and   73 00:08:12,080 --> 00:08:18,880 vehicle ownership that I showed you and the, we,  re, China heavily rely uh relies on fossil fuels   74 00:08:18,880 --> 00:08:25,120 especially coal for energy, as I showed you as  well. We also know the political environment,   75 00:08:25,120 --> 00:08:32,160 the political personnel system, is a top-down  approach, that is the government officials   76 00:08:32,160 --> 00:08:39,040 at the federal level appoint provincial  government officials then they in turn appoint uh   77 00:08:39,040 --> 00:08:46,080 government officials in the municipal and in  county levels. And the most important criteria,   78 00:08:46,080 --> 00:08:55,280 used to be really the sole criteria for  promotion was GDP growth, okay? Um, as a result,   79 00:08:55,280 --> 00:08:59,120 government officials really have a very  strong incentive to do all kinds of things   80 00:08:59,760 --> 00:09:07,840 to increase their GDP growth, including attract  a lot of very dirty industries: petrochemical, um   81 00:09:10,480 --> 00:09:14,080 plastic production, petroleum  refinery, those type of stuff,   82 00:09:14,960 --> 00:09:21,840 to increase GDP without giving really a lot of  thoughts about the environmental consequences. 83 00:09:24,080 --> 00:09:29,920 So environmental protection really was after  uh was an afterthought for a lot of government   84 00:09:29,920 --> 00:09:35,200 officials in making decisions, how to actually  balance economic growth and environmental   85 00:09:35,200 --> 00:09:40,960 protection, okay? It has changed during the last  uh, during the last five years. I'm going to talk   86 00:09:40,960 --> 00:09:51,440 about them toward the end. So as, at the same time  we have, we have, China had really very relaxed   87 00:09:51,440 --> 00:09:58,560 environmental regulations and enforcement and  uh to give you a sense, China's, we know in the   88 00:09:58,560 --> 00:10:04,640 United States the environmental regulation agency  is Environmental Protection Agency, the EPA.   89 00:10:05,680 --> 00:10:10,800 China's counterpart is Ministry of Environmental  Protection, was established only in 2008   90 00:10:12,000 --> 00:10:19,440 as a cabinet level ministry, okay? In the  U.S., we have about 17,000 employees under EPA.   91 00:10:20,640 --> 00:10:27,840 China's MEP employs about 2,600 people.  Think about, compare the size with uh   92 00:10:27,840 --> 00:10:35,440 the uh the population, the country um and this is,  this is much much smaller number of course, right?   93 00:10:37,440 --> 00:10:47,520 Another important issue is that local government  bureaus, local environmental regulators   94 00:10:48,720 --> 00:10:55,680 or think about these are local MBP offices, are  actually part of local government. In the U.S.,   95 00:10:56,400 --> 00:11:01,600 EPA has regional offices, but those  region offices are directly managed   96 00:11:01,600 --> 00:11:10,320 from EPA in D.C., okay? So regional directors  are actually appointed by EPA director.   97 00:11:11,760 --> 00:11:17,520 But in China, the local environmental bureau  chiefs are actually appointed by local government   98 00:11:18,400 --> 00:11:23,040 officials, right? And think about, these  are really environmental uh enforcers,   99 00:11:23,040 --> 00:11:30,480 right? Regulation and also monitors. And they have  to monitor people who will appoint them later on,   100 00:11:30,480 --> 00:11:36,000 right, or promote them later on. So this  creates, of course, perverse incentive   101 00:11:37,520 --> 00:11:40,480 in terms of environmental  regulation in China, okay?   102 00:11:41,680 --> 00:11:48,240 Alight, so I'm gonna, you know, as I mentioned,  show you some quite uh nice pictures. Uh so this   103 00:11:48,240 --> 00:11:54,800 is one of them, this air pollution. Uh this  is from uh from Hebei, uh which is the largest   104 00:11:55,360 --> 00:12:02,240 uh steel producer in the world, if you don't  uh consider China as a whole as a country,   105 00:12:02,240 --> 00:12:09,840 okay? Um so there is a saying that if you look at  steel production in the world, China is the first,   106 00:12:10,480 --> 00:12:16,960 Hebei province is actually second, okay?  Now let's look at graphically, compare   107 00:12:17,600 --> 00:12:24,240 Beijing's PM2.5 and L.A. L.A. is considered an  area in the U.S. with the worst air pollution,   108 00:12:24,240 --> 00:12:29,680 right? For those those of you who don't know  PM2.5, this is called particular matter,   109 00:12:29,680 --> 00:12:35,600 these are fine particular matter,  particular matter with diameter that is   110 00:12:35,600 --> 00:12:43,440 less than 2.5 millimeter, okay? 2.5 millimeter  is about 1-30th of a human hair, so tiny tiny um   111 00:12:45,040 --> 00:12:49,040 particular matter, and these are  really damaging to human health.   112 00:12:49,040 --> 00:12:54,160 During the past 20 years, actually 25  years, there were a lot of medical research,   113 00:12:54,160 --> 00:12:59,760 and really a lot of research, that looked  at health impacts PM2.5 and found them to be   114 00:13:00,960 --> 00:13:06,000 very damaging to human health, because they are  really small, you can breathe them, they can   115 00:13:06,000 --> 00:13:14,720 then go deep into your lung and blood streams,  cause lung cancer, and cause uh heart problems,   116 00:13:14,720 --> 00:13:21,920 okay? So if you look at the daily concentration  here in China, in orange, from 2012 to 2013,   117 00:13:22,480 --> 00:13:29,680 and the black is for, sorry, this is for Beijing,  and this is for L.A., okay? I also want you look   118 00:13:29,680 --> 00:13:37,440 at the EPA categories here. So this level is good  to moderate. This this range is unhealthy. This   119 00:13:37,440 --> 00:13:45,360 is very unhealthy. This level above is what EPA  calls header, hazardous, meaning EPA actually   120 00:13:45,360 --> 00:13:51,840 will advise people not to go outside, okay?  But that happens quite often in Beijing, for   121 00:13:51,840 --> 00:13:57,360 those of us who lived in Beijing, we all know  this, but this is the record, uh before 2000,   122 00:13:57,360 --> 00:14:04,960 before the state was near, nearly 600 micrograms  per cubic meter, this is the measurement we use.   123 00:14:04,960 --> 00:14:11,280 I mean if you look at L.A. again, this is the most  polluting city in the U.S. There's no comparison.   124 00:14:11,280 --> 00:14:18,480 So another thing mentioned that U.S. has a daily  standard that's 35 micrograms per cubic meter,   125 00:14:18,480 --> 00:14:25,120 right, and for L.A., actually, most of the  days were below this 35 line there, okay?   126 00:14:26,320 --> 00:14:33,280 This is for whole China PM2.5 concentration on an  annual basis, okay? So if you look at this dark   127 00:14:33,280 --> 00:14:41,440 red um and this, this red provinces, their annual  level is about, it's over 50 or even over 70, okay   128 00:14:41,440 --> 00:14:47,120 micrograms per cubic meter, this is the unit we  use, if you look at the U.S. annual standard 12,   129 00:14:48,320 --> 00:14:54,400 okay, WHO, World Health Organization guidelines is  only 10. So for most actually we all we know that   130 00:14:55,040 --> 00:15:02,480 this part of China has a lot of population, so  for most of these areas the concentration level is   131 00:15:02,480 --> 00:15:12,240 four, five, six, seven, times the WHO guideline,  okay? Okay we know air pollution is bad, okay,   132 00:15:12,240 --> 00:15:17,520 what is the cause of air pollution? So there  are a lot of research on that, uh, well it's not   133 00:15:17,520 --> 00:15:23,200 quite recent anymore, because this old slide,  there was article in Nature September 2015,   134 00:15:23,200 --> 00:15:29,040 they estimate that air pollution economy for  1.3 million premature deaths in China in 2010,   135 00:15:29,840 --> 00:15:35,120 and this number accounted for 40  percent of world total premature deaths. 136 00:15:38,080 --> 00:15:41,520 2007 world report, they estimate  the cost of pollution in China,   137 00:15:42,080 --> 00:15:50,720 um and the health cost of air pollution health  costs alone is, uh was 1.2 to 3.8 of GDP in 2013.   138 00:15:50,720 --> 00:15:57,280 I can give you a quick idea how the estimate this  cost. What they use is so called dose response   139 00:15:57,280 --> 00:16:05,120 functions, right? They have data,  for example, in terms, air pollution and mortality   140 00:16:05,120 --> 00:16:10,960 and morbidity rates in different locations in  the U.S. or elsewhere, they try to find the   141 00:16:10,960 --> 00:16:16,320 relationship between pollution and mortality rate,  and they try to estimate that relationship, that's   142 00:16:16,320 --> 00:16:21,040 called dose response relation,  is well established in the medical literature and   143 00:16:21,040 --> 00:16:27,520 then they can use those functional relationships  to estimate, well if China's air pollution was   144 00:16:27,520 --> 00:16:35,440 was lower than what we had before, for example,  as low as the World Health Organization guideline,   145 00:16:35,440 --> 00:16:40,640 what would happen to their premature deaths,  right? And these are what happened to all kind   146 00:16:40,640 --> 00:16:47,360 of diseases, and health cost is only one part of  the cost, of course, okay, there are other costs.   147 00:16:48,880 --> 00:16:55,920 Okay a 2012 MIT study estimates that health cost  air pollution uh to be 112 billion U.S. dollar   148 00:16:55,920 --> 00:17:03,840 in 2005, five percent of GDP, okay? So these are  really big numbers if you think about it, okay?   149 00:17:05,280 --> 00:17:12,720 Alright, soil pollution: about 20 percent China's  farmland are polluted, from industry wastewater   150 00:17:12,720 --> 00:17:17,840 and levels of organic pollutants  are often 20 times the standard. 151 00:17:20,080 --> 00:17:23,600 Water pollution: 60 percent of  groundwater are contaminated,   152 00:17:24,320 --> 00:17:29,840 90 percent for cities, and two-thirds of  China's rural public population use water   153 00:17:29,840 --> 00:17:38,080 contaminated by human and industry waste, okay? As  a result, especially for rural areas, um Ministry   154 00:17:38,080 --> 00:17:45,360 of Environmental Protection um classified  459 villages as cancer villages. There are,   155 00:17:45,360 --> 00:17:50,240 if you look at location, it's not surprising,  these are located in heavily industry areas,   156 00:17:50,240 --> 00:17:56,480 uh with actually a lot of so-called industry  clusters, okay? Cluster of textile firms,   157 00:17:56,480 --> 00:18:03,840 cluster of steel uh mills, etc okay? And if  uh research has found that water pollution   158 00:18:04,480 --> 00:18:10,080 is the main cause for cancer villages, okay? These  are really, um as mentioned in the beginning,   159 00:18:10,080 --> 00:18:17,360 kind of dire [unintelligible] pictures. You  might ask, what, where do we go from here, okay?   160 00:18:18,800 --> 00:18:24,320 So some of you actually saw these pictures  before, so I put these two pictures,   161 00:18:24,320 --> 00:18:29,600 these are two pictures in different uh  continent, and different time period,   162 00:18:30,240 --> 00:18:38,640 okay? I want you to guess what is this about, what  was this, and then what what was this? Exactly,   163 00:18:38,640 --> 00:18:45,040 this is actually Dust Bowl from the from from the  U.S. in the 30s that happened in the Great Plains,   164 00:18:45,040 --> 00:18:51,680 Texas, Oklahoma, those areas, and the main cause  of the dust bowl, as well, a combination of weather   165 00:18:51,680 --> 00:18:58,960 but mostly erosion of topsoil due to  intensive uh agriculture um and this,   166 00:19:00,880 --> 00:19:05,680 and this is this is Ningxia in  uh in 2000 in 2010 in China,   167 00:19:05,680 --> 00:19:12,240 okay? This is one of the sandstorm um you  know episodes, okay? Okay, so these two,   168 00:19:14,400 --> 00:19:20,320 uh well, this is New York City, George Washington  Bridge, okay? This is, uh was from 1970   169 00:19:21,920 --> 00:19:25,280 okay? And this you can't really  see this much of this, right?   170 00:19:25,280 --> 00:19:32,320 And then this, is this is Beijing, okay? Um  sometime, actually very often, this is Beijing,   171 00:19:32,320 --> 00:19:39,200 okay? So you think about, well, you know, U.S.  had, you know, very bad environmental issues   172 00:19:39,200 --> 00:19:44,640 um in the 60s, 70s or 50s, 40s even earlier,  you know, we all know if you study environmental   173 00:19:44,640 --> 00:19:50,160 history um there were a lot of environmental  disasters, right? One of them was land and smog,   174 00:19:51,840 --> 00:19:57,520 1952, 4,000 people died during that uh a  few days, right? So there were incidents   175 00:19:57,520 --> 00:20:01,840 like this in the U.S. in Pittsburgh,  in a place called Donora, there were,   176 00:20:03,200 --> 00:20:09,600 many people died in a short span, few days,  because combinations of air pollution and   177 00:20:09,600 --> 00:20:20,160 bad weather conditions. So how did U.S. do it,  right? How did U.K. clean up their air? Okay,   178 00:20:21,040 --> 00:20:26,240 environmental regulations in the U.S.  So 1970 was a very important year,   179 00:20:27,520 --> 00:20:32,400 and before 1917 many things happened, including,  I don't know whether you know this book called the   180 00:20:32,400 --> 00:20:39,040 Silent Spring by Richard Carson, and many also  bad environmental incidents, and that lead to   181 00:20:39,040 --> 00:20:45,920 environmental movement, and Richard Nixon created  EPA in 1970. They also passed Clean Air Act.   182 00:20:47,040 --> 00:20:52,880 This act is very important, uh, the goal  of that act was to regulate air pollutions,   183 00:20:53,520 --> 00:21:00,160 so they set up what they call National  Ambient Air Quality Standards um,   184 00:21:00,160 --> 00:21:07,840 and um for six pollutants including carbon  monoxide, uh sulfur dioxide, particular matter,   185 00:21:07,840 --> 00:21:16,720 it was PM10, uh NOG, NO2, and  ozone, and lead. They said, well, for each county,   186 00:21:17,440 --> 00:21:22,480 right, you have to actually comply with the  the national standard. If you don't comply,   187 00:21:23,120 --> 00:21:27,840 then we are going to pull out federal fundings,  you are going to suffer from consequences,   188 00:21:27,840 --> 00:21:32,720 so you have to do anything you need to do  to make sure you comply with the air quality   189 00:21:32,720 --> 00:21:39,120 standards. So different states use different  strategy to do that. This was a very important   190 00:21:39,840 --> 00:21:46,160 regulation in the U.S. in terms of environmental  regulation. Numerous laws passed since 1970,   191 00:21:46,160 --> 00:21:51,600 they addressed a lot of things: clean air, clean  water, energy conservation hazardous waste,   192 00:21:51,600 --> 00:21:58,960 pesticides, and more recently toxics from  power plants, okay? Okay, let's look at what   193 00:21:58,960 --> 00:22:06,240 what happened from 1970, okay, actually to large  extent due to environmental regulation. Look at   194 00:22:06,240 --> 00:22:15,280 SO2 air quality, right, so from 1980 to 2010 there  was a 23 decrease in national average. In terms   195 00:22:15,280 --> 00:22:21,200 of SO2, in terms of NO2 there was a 52  percent decrease in national average, okay,   196 00:22:22,640 --> 00:22:28,720 carbon monoxide, 82 percent  drop, and 28 decrease in ozone,   197 00:22:29,680 --> 00:22:36,960 so it's a remarkable uh uh improvement during  this 20 years, actually 20, 30 years, right? 198 00:22:39,840 --> 00:22:46,400 Okay, um, what's the result of environmental,  like, these effects of environmental legislations?   199 00:22:46,400 --> 00:22:54,480 Well, we all know the air now is is very good, but  if you look at the uh, the health consequences, or   200 00:22:54,480 --> 00:23:01,440 health benefits, right? If you look at 1990 alone,  there are studies show that clean air programs   201 00:23:01,440 --> 00:23:08,880 prevented this year alone over 20,000 premature  deaths, this many cases of chronic bronchitis,   202 00:23:08,880 --> 00:23:14,800 and you know heart disease, asthma attacks, uh  all these things, and also 10 point million loss   203 00:23:14,800 --> 00:23:20,720 IQ point in children from lead reductions, right,  these are these are numbers that are produced by   204 00:23:20,720 --> 00:23:27,040 researchers, okay, and you know these are these  are the health benefits in 1990 alone, okay?   205 00:23:27,040 --> 00:23:32,560 And there are recent studies that actually show,  that try to quantify the benefit of clean air   206 00:23:32,560 --> 00:23:41,600 act uh in risk, in more recent years. For example,  they found that uh that you know premature deaths   207 00:23:41,600 --> 00:23:51,440 was reduced by over 160,000 uh in 2010 due to  uh due to Clean Air Act uh regulations, okay?   208 00:23:52,240 --> 00:23:59,280 So I showed you uh the progress that  U.S. made during the last 30, 40 years,   209 00:23:59,280 --> 00:24:05,360 okay, and I mentioned briefly the environmental  regulations, the big really regulations,   210 00:24:07,120 --> 00:24:09,360 and now I'm going to go through very quickly   211 00:24:10,320 --> 00:24:15,280 approaches to environmental regulation, okay?  So think about China, there are many options on   212 00:24:15,280 --> 00:24:22,320 the table, right? Um some have tried by other  countries, some have not, on a large scale,   213 00:24:22,880 --> 00:24:25,440 okay? So Chinese government are  thinking about these policies   214 00:24:26,000 --> 00:24:32,160 and we need to decide: what are the policies that  that we should use uh to address the issues? So   215 00:24:33,520 --> 00:24:41,040 economists tend to categorize different  approaches into three categories. The first,   216 00:24:41,040 --> 00:24:47,280 we call that command-and-control approach.  To give you example, for example we ban DDT,   217 00:24:48,000 --> 00:24:56,560 this is pesticide, um we ban CFCs, these are  ozone depleting CFCs, and we have these type   218 00:24:56,560 --> 00:25:01,760 of regulations, complete prohibition,  okay, of use of some of the chemicals,   219 00:25:02,320 --> 00:25:08,640 we can do that. Second type of command-and-control  approach, is is less forceful but   220 00:25:10,400 --> 00:25:17,120 the government essentially mandate firms,  okay, coal fire plants or chemical companies or   221 00:25:18,720 --> 00:25:24,720 petroleum refinement companies to use the  best available technology on the market in   222 00:25:24,720 --> 00:25:29,840 terms of environmental performance, right, you  have to do that, so government can mandate um   223 00:25:29,840 --> 00:25:36,640 that, um so these are called uh  these are kind of uh are within   224 00:25:36,640 --> 00:25:43,120 um this command-and-control approach, okay? Think  about the pros and cons, right? Well because you   225 00:25:43,120 --> 00:25:49,120 have, the government has a lot of control, right,  so you can expect some results will be achieved,   226 00:25:49,120 --> 00:25:52,880 right, in terms of the environment  performance you can expect that   227 00:25:52,880 --> 00:25:59,120 certain improvement will be achieved, okay,  but you have to think about at what cost, okay?   228 00:26:01,040 --> 00:26:07,680 The second approach called market-based  approach okay, this approach basically consider   229 00:26:09,360 --> 00:26:16,160 clean air, clean water as resources, as natural  resources, okay, and we need to think about how   230 00:26:16,160 --> 00:26:22,160 to allocate these resources just as we think  about how we allocate goods on the markets,   231 00:26:22,160 --> 00:26:26,640 computer, uh you know clothes, or anything  that we purchase on the market, right?   232 00:26:27,440 --> 00:26:33,440 And we know for things that we purchase  on market, price is the mechanism which   233 00:26:34,160 --> 00:26:38,000 guides us where these goods should  go to or who the goods should go to,   234 00:26:38,000 --> 00:26:44,720 right? The goods should go to those who value  these um this um those who value them the most,   235 00:26:44,720 --> 00:26:51,440 right, um so um so this approach basically says  well let's think about really treating air and   236 00:26:51,440 --> 00:26:57,200 water as natural resources and let's think  about use market mechanism to re-allocate   237 00:26:58,160 --> 00:27:02,960 resources, right? That is, if  you uh power plants, you pollute,   238 00:27:04,880 --> 00:27:08,880 your generate pollutions, that is, you  are using clean air for your production,   239 00:27:09,600 --> 00:27:17,600 right? And you need to pay for the input, that  is clean air. Under this approach, for example,   240 00:27:17,600 --> 00:27:23,680 as I mentioned, if you need to pay for clean air,  that's we call that tax, example is carbon tax,   241 00:27:23,680 --> 00:27:28,160 if power plants you generate carbon emissions,  for each time carbon emission, you need to pay   242 00:27:28,160 --> 00:27:33,840 a price, okay? How much you need to pay depends  on the damage carbon does to to the society,   243 00:27:34,560 --> 00:27:41,520 okay, so we need to quantify the damage.  Gasoline tax is another uh kind of uh tax   244 00:27:43,040 --> 00:27:51,600 that try to actually control for this damages.  Subsidy on electric cars, on other fuel saving   245 00:27:51,600 --> 00:28:00,640 technologies, okay? This is essentially putting a  price on using these resources, okay, for subsidy   246 00:28:00,640 --> 00:28:07,120 basically you give you a subsidy to reduce your  pollution, okay. That's the opposite to the tax.   247 00:28:08,560 --> 00:28:14,640 Cap and trade, this is, you hear a lot about this  cap and trade, that's another type of market-based   248 00:28:14,640 --> 00:28:22,560 approach. The idea is that, if you want to  use tax, for example, use carbon tax, okay,   249 00:28:24,960 --> 00:28:33,360 a lot of times people, consumers, or general  public do not like the idea of tax, okay,   250 00:28:33,360 --> 00:28:37,520 especially in the U.S. and in other countries  as well. One of the reasons is they don't trust   251 00:28:38,080 --> 00:28:42,000 the government in terms of how they are  going to use those tax revenues later on,   252 00:28:42,560 --> 00:28:48,720 right, so some economists suggest well let's  think about alternative way to do this,   253 00:28:48,720 --> 00:28:54,720 right? We call that cap and trade. Um the idea  is that if we want to reduce air pollution,   254 00:28:54,720 --> 00:29:00,160 for example, if we want to reduce  carbon emissions, okay, in order to, um   255 00:29:01,360 --> 00:29:08,400 a word, uh in order to prevent, from uh,  prevent the, you know, disastrous scenarios   256 00:29:08,400 --> 00:29:14,720 to happen, okay, so let's think about well carbon  emissions should be below this certain threshold   257 00:29:15,360 --> 00:29:21,200 okay for many many years onward from, this year  onward, and so we need to set a cap; that is,   258 00:29:21,920 --> 00:29:29,440 each year we cannot emit more than this,  okay, then we are going to have licenses   259 00:29:32,080 --> 00:29:37,680 allocated to firms so that when  firms are made, each time carbon,   260 00:29:37,680 --> 00:29:45,360 they need to surrender one unit of license  to the government, okay? So in the end,   261 00:29:49,360 --> 00:29:59,760 the cap is uh is um should not be, um should  not be violated, okay? Um once you allocate   262 00:29:59,760 --> 00:30:03,920 the permits to firms, so there are many different  ways to allocate permits, but one of them is   263 00:30:03,920 --> 00:30:11,920 based on history uh emissions. Once you allocate  these permits, and actually firms then can trade   264 00:30:11,920 --> 00:30:18,960 among themselves. For example, if I am the  for, as a firm get 200 tons of permits,   265 00:30:19,680 --> 00:30:28,560 okay, but during my production process, I emitted  250 tons, then my own permit is not enough,   266 00:30:29,520 --> 00:30:35,840 but what I can do, I can actually go to the  market and buy permits from other firms, right?   267 00:30:36,800 --> 00:30:43,120 If I produce only 150 tons of CO2 that year,  but I have 200 permits, then I can sell the extra   268 00:30:43,120 --> 00:30:49,920 50 tons permits on the market to other firms,  okay, so you allow firms to trade. In the end,   269 00:30:49,920 --> 00:30:55,040 the permit, the right to pollute, you can  think about, will have a market price,   270 00:30:56,080 --> 00:31:01,600 okay? The market price was determined by  the cap, by actually the cost of the firms,   271 00:31:01,600 --> 00:31:06,960 in reducing their pollution. What  happens in this system is that   272 00:31:07,680 --> 00:31:14,240 the permit price actually is going to guide,  as actually signal or mechanism, to guide firms   273 00:31:15,760 --> 00:31:22,880 optimally choose their abatement behavior, or  behavior of reducing emissions, okay? For example,   274 00:31:22,880 --> 00:31:28,720 if I am a very very efficient firm, I can easily  reduce my air pollution without incurring much   275 00:31:28,720 --> 00:31:35,520 of a cost, okay? Another firm, that is a firm  that is not very good in terms of technology,   276 00:31:35,520 --> 00:31:41,520 so it's going to take a lot of efforts or cost  for that firm to reduce emissions, so think   277 00:31:41,520 --> 00:31:47,360 about what would happen to this type of firms,  well, if I'm very good at reducing pollution,   278 00:31:47,360 --> 00:31:52,400 in the end, I'm actually in in that I'm going  to reduce a lot of pollution, and I'm going to   279 00:31:52,400 --> 00:31:58,560 sell a lot of premise to the other firm, who is  not as good reducing pollution, so he will not   280 00:31:58,560 --> 00:32:03,280 reduce a lot of pollution, but he, that firm is  going to use a lot of permits from the market,   281 00:32:03,280 --> 00:32:09,120 he's going to buy a lot of permits from the  market uh to to do that, because for that firm   282 00:32:09,120 --> 00:32:13,840 reducing pollution by himself actually is  more costly than buying from the market,   283 00:32:14,800 --> 00:32:21,280 okay? So in the end, you, we we actually showed  through theory, through empirical studies,   284 00:32:21,280 --> 00:32:26,720 based on past experience, this type of system  cap and trade, can be very cost effective   285 00:32:26,720 --> 00:32:31,760 in reducing emission. That is, if you want  achieve that cap, or certain level of emission,   286 00:32:32,320 --> 00:32:36,400 and this actually is very efficient in  doing that, so we are going to use the   287 00:32:37,360 --> 00:32:43,520 lower cost in achieving the same amount reduction.  Than, for example, this type of approach,   288 00:32:44,320 --> 00:32:48,640 because this type of approach, you  do not give firms a lot of freedom   289 00:32:48,640 --> 00:32:56,000 in choosing their best strategy, okay, so this  type of strategy for example technology mandates   290 00:32:56,000 --> 00:33:00,640 basically assumes the government knows a lot  about the technology, and we will choose the   291 00:33:00,640 --> 00:33:08,880 best technology which actually is not often  the case, okay? So there are a lot of examples   292 00:33:08,880 --> 00:33:13,760 uh in the U.S., for example, in the  80s, the U.S., the way that the U.S.   293 00:33:15,600 --> 00:33:21,680 phase out leaded gasoline was through this  type of idea, and that program was actually   294 00:33:21,680 --> 00:33:27,600 under Ronald Reagan, okay? SO2 permit trading,  which is very big, very successful program in   295 00:33:27,600 --> 00:33:35,520 reducing SO2 was authorized by first Bush in 1989.  So cap and trade actually was a Republican idea,   296 00:33:36,400 --> 00:33:43,680 okay, but we all know in recent years and a  lot of lawmakers call that cap and tax, right,   297 00:33:43,680 --> 00:33:50,000 they didn't really like this idea. They blocked a  lot of actually uh proposals in the Congress. Um   298 00:33:51,840 --> 00:33:57,840 emission trading for CO2, so this is European  Union emission treaty for CO2, this is the largest   299 00:33:58,400 --> 00:34:03,040 emission trading program or cap and trade  program in the world, started from 2005.   300 00:34:04,240 --> 00:34:08,240 And we're going to talk about  China's CO2 cap and trade programs.   301 00:34:08,240 --> 00:34:12,720 So this is one of the policies that are  really being used or being considered   302 00:34:13,280 --> 00:34:18,160 in many parts of the world to deal with  not only local pollutants such as SO2,   303 00:34:18,160 --> 00:34:26,000 but also global pollutant such as CO2, okay?  So the third category, call them information,   304 00:34:26,000 --> 00:34:33,200 you know, let consumers know that the the cause  of climate change, the damage of climate change,   305 00:34:33,200 --> 00:34:38,800 therefore, think about more carefully, how much  electricity you use, how much heat do you use,   306 00:34:38,800 --> 00:34:45,280 and then think about the actions you take, right,  moral suasion, norms, social pressures, this is   307 00:34:45,280 --> 00:34:50,960 uh, these are really in one this category, right?  This could be uh useful it could be helpful,   308 00:34:50,960 --> 00:34:58,360 uh many many very often they will be helpful  when you combine this and this together, okay?   309 00:34:59,360 --> 00:35:04,560 As I mentioned, traffic congestion is really  one of the, to me, really one of the two most   310 00:35:04,560 --> 00:35:10,880 pressing issues in urban China, and there are  many options to deal with traffic congestion,   311 00:35:11,440 --> 00:35:17,920 and I'm going to talk a specific approach that  we propose. Uh it's a market-based approach,   312 00:35:19,440 --> 00:35:28,400 and look at how we actually set up uh that  poli- that policy, and how to implement it,   313 00:35:28,400 --> 00:35:34,240 what would be the consequences of that policy.  In this paper, this really a combination of   314 00:35:35,200 --> 00:35:39,440 economic insights, okay very basic  fundamental economic insights,   315 00:35:40,240 --> 00:35:47,280 and big data on a very important social  issue, okay? So if you rank world cities   316 00:35:47,280 --> 00:35:53,440 in terms of traffic congestion, okay, the top one  is the worst, okay, so Mexico City was the worst,   317 00:35:53,440 --> 00:36:00,880 it's top 15. If you look at them, seven of those  cities are actually from China, okay? A lot of   318 00:36:00,880 --> 00:36:06,400 them, if you look at them, actually, are not are  actually from mid or or mid-income countries,   319 00:36:07,120 --> 00:36:14,320 okay? L.A. has the worst traffic congestion  in the United States, it's number six, okay?   320 00:36:14,320 --> 00:36:21,600 This is based on 2015 data. So this is Beijing  15 years ago, actually, yeah, 15-16 years ago,   321 00:36:23,280 --> 00:36:28,960 and this is Beijing now. I showed you the the  dramatic increase in vehicle ownership before.   322 00:36:29,760 --> 00:36:34,400 From economics perspective, traffic  congestion is very simple, actually,   323 00:36:34,400 --> 00:36:40,840 okay, it's fundamentally, it is  excess demand of road capacity, okay,   324 00:36:44,560 --> 00:36:50,480 and why there is excess demand? Okay, we  think, well that's because mispricing.   325 00:36:51,360 --> 00:36:56,640 If a good resource is free of charge, and  if this is something that is, that is good   326 00:36:56,640 --> 00:37:01,920 thing for people, and it is free, then the  demand for that will be will be high, right?   327 00:37:02,640 --> 00:37:11,840 That's true for road capacity. So there are two  broad strategies to deal with traffic congestion,   328 00:37:11,840 --> 00:37:18,000 or to deal with exit demand issue, okay? One type  of strategy, we call that supply side strategy.   329 00:37:18,560 --> 00:37:22,240 For example, building more new roads,  improving public transportation.   330 00:37:23,520 --> 00:37:28,320 Think about what what will that do if you build  more roads, right? Well it's going to reduce   331 00:37:28,320 --> 00:37:35,440 congestion for sure, but that will also reduce  the cost of travel for people, okay? That is,   332 00:37:35,440 --> 00:37:39,840 reduce the price that people have to pay for  travel, the time, the price that I'm talking   333 00:37:39,840 --> 00:37:45,680 about, think about that's the, uh the cost  of time, right? If something becomes cheaper,   334 00:37:46,320 --> 00:37:51,600 then the demand for that will be higher. So what  I'm saying is when you build more road, you're   335 00:37:51,600 --> 00:37:56,720 going to reduce travel cost, therefore you're  going to lead to more driving, higher demand for   336 00:37:56,720 --> 00:38:03,040 driving, that will in turn will lead to congestion  again, right? So in the short term, supply side   337 00:38:03,040 --> 00:38:09,600 policies can work, and tend to work, but in  the long term, they cannot fundamentally deal   338 00:38:09,600 --> 00:38:15,360 with or address congestion issue. Another set of  strategies, we call them demand side strategies,   339 00:38:16,400 --> 00:38:21,040 for example the command-and-control approach  that I mentioned driving restrictions that   340 00:38:21,040 --> 00:38:27,600 are being used in many uh cities in the world,  including Mexico City, Bogota, some other South   341 00:38:27,600 --> 00:38:35,280 American cities, and Beijing, and uh probably New  Delhi soon, right? The idea here is that, well,   342 00:38:36,160 --> 00:38:41,680 if there are too many cars on the road, why  can't we restrict some of them from driving,   343 00:38:42,560 --> 00:38:47,440 right? So in Beijing, what they do is, depending  on the last digits of your license, well you can   344 00:38:47,440 --> 00:38:53,200 only drive four days out of five days, right?  Um so this is kind of very straightforward kind   345 00:38:53,200 --> 00:38:58,240 of linear thinking, right, if there are too many  cars let's take some of the cars off the road by   346 00:38:58,240 --> 00:39:03,920 forcing people or prohibiting people from driving,  right, by not allowing people to buy cars,   347 00:39:04,480 --> 00:39:11,200 okay? This we call them command-and-control  approach. They might work, sometimes actually   348 00:39:11,200 --> 00:39:17,440 might not work. The other approach, as I  mentioned, we call that market-based or   349 00:39:17,440 --> 00:39:23,840 price-based approach. So here we are going to  use actually price signals as a mechanism to   350 00:39:23,840 --> 00:39:29,680 allocate resources. The resource here is road  capacity. The same thing as clean air, right,   351 00:39:29,680 --> 00:39:37,120 we can use price signal to thinking about allocate  resources but if you compare command approach and,   352 00:39:39,600 --> 00:39:44,480 with the price-based, we call that congestion  pricing, congestion pricing is going to be   353 00:39:44,480 --> 00:39:51,600 much more effective or it's going to achieve the  same goal with the lease cost. The reason, as I   354 00:39:51,600 --> 00:39:56,880 mentioned a moment ago, is that when you use this  market-based approach, or congestion price in this   355 00:39:57,520 --> 00:40:03,520 instance, you have a lot of margins. You have,  actually, you give travelers a lot of margins   356 00:40:03,520 --> 00:40:09,440 to adjust their behavior. For example, when you  have a congestion pricing, well, that can induce   357 00:40:09,440 --> 00:40:16,720 adjustment uh for travelers in many margins, such  as trip frequency, how many times they travel,   358 00:40:16,720 --> 00:40:20,960 what kind of mode they use, what's the time  of travel, and what's the route they use,   359 00:40:20,960 --> 00:40:28,080 so they are going to adjust their behavior or  travel behavior accordingly um in an optimal   360 00:40:28,080 --> 00:40:34,320 way under this pricing strategy, okay? So this  is actually what we call first-best policy   361 00:40:35,280 --> 00:40:43,040 in addressing traffic congestion, okay? And this  first-best policy addressing congestion was first   362 00:40:43,040 --> 00:40:49,760 proposed uh by William Vickrey in 1955. William  Vickrey is a Nobel prize winner in economics,   363 00:40:49,760 --> 00:40:56,480 1992. He had he made a lot of very important  contributions in many areas in economics   364 00:40:57,200 --> 00:41:03,680 including auction, and  in 1955 he proposed congestion pricing to   365 00:41:03,680 --> 00:41:09,360 deal with urban congestion in New York City, and  then he, he talked about congestion pricing for   366 00:41:09,920 --> 00:41:17,520 urban transportation in general in the U.S., okay?  In 1963, in his American Economic Review article,   367 00:41:17,520 --> 00:41:23,360 he he mentioned this, he said, well, no other  areas are pricing practices so irrational,   368 00:41:23,360 --> 00:41:30,160 so out of date, and so conducive to waste as in  urban transportation, okay? So basically arguing,   369 00:41:30,160 --> 00:41:37,440 well, pricing practices in the U.S. was really  out of date. Essentially there was no pricing,   370 00:41:37,440 --> 00:41:43,760 so we had very serious mispricing of road  capacity, as a result we have congestion issues   371 00:41:43,760 --> 00:41:50,960 everywhere, right, in urban areas.  Okay, if you think more about it, okay,   372 00:41:51,760 --> 00:41:58,000 congestion is what we call, economists call,  a classic a classical classic externality.   373 00:41:59,600 --> 00:42:06,480 Externality arises when one's action, when  my action inadvertently affects other people,   374 00:42:07,200 --> 00:42:13,760 okay? For example, when I drive on the road, I  will actually slow down other people on the road.   375 00:42:14,720 --> 00:42:18,800 But when I make my decision in terms,  well, when I'm going to travel,   376 00:42:18,800 --> 00:42:24,400 how far I'm going to travel, I actually don't  consider my impact on others into my decision,   377 00:42:25,040 --> 00:42:33,760 right? So my action is causing impacts or impacts  or affecting others that I'm not considering   378 00:42:33,760 --> 00:42:41,840 or taking into account, and that really my uh, the  the impact of my trouble on others is externality,   379 00:42:42,400 --> 00:42:49,600 okay? The same thing is pollution, right, when  firms make, uh produce electricity, for example,   380 00:42:50,160 --> 00:42:55,040 they will make their decision based on the cost  of the input, based on the price of the output,   381 00:42:55,840 --> 00:42:59,200 but they don't actually take into account  the amount of pollution they will produce,   382 00:43:00,160 --> 00:43:06,080 right, and therefore they are going to lead to  health consequences. We call them externalities as   383 00:43:06,080 --> 00:43:13,520 well. Pigou, author Pigou in 1920 actually talked  about externalities, and he proposed use tax   384 00:43:14,160 --> 00:43:19,600 to correct for this externality generating  activities. We call that Pigouvian tax. So   385 00:43:19,600 --> 00:43:26,880 congestion pricing is essentially one type of  congestion cost, that is, if you are, your action   386 00:43:26,880 --> 00:43:31,920 is hurting other people, you have to pay for  your actions, for for the damage that you cause,   387 00:43:31,920 --> 00:43:38,800 right? So in in terms of congestion, you actually  need to pay the cost that you impose on other   388 00:43:38,800 --> 00:43:45,520 road travelers, okay? How much do, should you  pay? Well, that's called congestion pricing,   389 00:43:45,520 --> 00:43:49,520 right? What's the level of congestion  pricing? But that's an empirical question.   390 00:43:50,160 --> 00:43:56,560 Let's look at, so this is the cost of travel, this  is the, think about this as a demand of travel,   391 00:43:56,560 --> 00:44:04,560 this is a density, number of vehicles on the road  per kilometer, okay? And at, a certain level,   392 00:44:04,560 --> 00:44:13,360 below this level of density, then, this is the  cost, the travel cost is flat, travel cost,   393 00:44:13,360 --> 00:44:17,920 including time cost, including operating  cost, and road maintenance cost, okay?   394 00:44:20,240 --> 00:44:27,440 This is flat, but when the number of vehicles is  above this threshold, is above this threshold,   395 00:44:27,440 --> 00:44:33,440 then there will be difference between, this is  what we call a cost average, sorry we call that   396 00:44:34,880 --> 00:44:43,600 marginal um private cost, think about the cost  that travelers pay themselves, and the cost to   397 00:44:43,600 --> 00:44:48,720 the society, okay, this we we call marginal  social cost. So there will be a difference   398 00:44:48,720 --> 00:44:54,800 between the cost to the society, and the cost  to individual travelers. The reason there is   399 00:44:54,800 --> 00:45:01,280 difference is because congestion, right? At this  point, when we add one more traveler on the road,   400 00:45:02,400 --> 00:45:07,280 it's going to slow down other people on the road,  therefore, they are gonna, it's gonna take longer   401 00:45:07,280 --> 00:45:13,760 for them to travel, and there will be a cost,  because um time is money. So the difference   402 00:45:13,760 --> 00:45:23,360 that we, between these two lines, is externality,  from congest, from uh congestion, okay, um   403 00:45:24,720 --> 00:45:30,320 and uh and this is, this is what we call a  demand curve for travel, when price is high,   404 00:45:30,320 --> 00:45:37,600 you travel less, price is low you travel more.  So Pigou in 1920 said, in this type of situation,   405 00:45:37,600 --> 00:45:43,760 right, he looked at other type of externalities.  Well we can actually impose a tax, okay,   406 00:45:43,760 --> 00:45:49,840 so that, as a society of a whole, we achieve  the best outcome. If you don't do anything, if   407 00:45:49,840 --> 00:45:55,360 the government does not do anything, then in the  the society actually will achieve this outcome,   408 00:45:56,160 --> 00:46:01,840 this level of density, and there will be big  wedge between the social costs and private costs,   409 00:46:01,840 --> 00:46:05,600 there'll be a lot of congestion. This is  not ideal from a society point of view.   410 00:46:06,880 --> 00:46:09,920 From a society point of view,  this point is the best, okay?   411 00:46:11,120 --> 00:46:18,560 So in order to achieve this point, this level  of density, or travel demand, we actually need   412 00:46:18,560 --> 00:46:24,720 to impose a tax. So the tax, the amount of tax  should be this much, okay, so this this is the   413 00:46:24,720 --> 00:46:30,240 [unintelligible], we call that congestion  charge, okay? So in this paper, what we are going   414 00:46:30,240 --> 00:46:35,120 to do, we are going to look at Beijing, we are  going to try to actually estimate these curves,   415 00:46:36,000 --> 00:46:41,080 this curve, this curve, this curve, and try  to estimate the congestion pricing, okay?   416 00:46:41,840 --> 00:46:46,000 So that is how much you should charge  road users so that in the end, if we   417 00:46:46,000 --> 00:46:51,600 achieve a level congestion, that it actually is  optimal for the, for the society point of view.   418 00:46:51,600 --> 00:46:59,680 Okay, alright, um so there are many cities around  the world using congestion pricing, starting from   419 00:46:59,680 --> 00:47:05,280 Singapore in 1975 um, you know there are  several European cities adopt in early 2000s,   420 00:47:06,240 --> 00:47:10,880 the studies show that this congestion pricing  practice is actually reduce congestion by   421 00:47:10,880 --> 00:47:17,280 10 to 30 percent in different cities of the  world, okay? So Beijing Municipal Government   422 00:47:17,280 --> 00:47:22,800 uh in two thousand, December 2015, announced  that they will introduce congestion pricing,   423 00:47:22,800 --> 00:47:26,960 so now they are actually studying congestion  pricing think about how to implement it think   424 00:47:26,960 --> 00:47:34,560 about the potential impacts of the policy,  okay? So what we are going to do is, actually,   425 00:47:36,000 --> 00:47:42,000 really provide suggestions or information  for them to use to think about how they   426 00:47:42,000 --> 00:47:48,400 should implement the practice, okay, so this is  Singapore to just give you an idea, what they do,   427 00:47:48,400 --> 00:47:55,200 this is called electric road pricing, so  every car has a uh, what do you call that, um   428 00:47:56,160 --> 00:48:03,200 receptor or um and there's also a credit card  here, uh when this car passed this this point,   429 00:48:03,200 --> 00:48:11,200 this will this will you know have a received  signal, and um there are different um things on   430 00:48:11,200 --> 00:48:16,640 the different, you know, uh at different points on  the road, there are many uh structures like this,   431 00:48:16,640 --> 00:48:24,080 um and uh at the end of a month, essentially your  charges will be deducted from the credit card, um.   432 00:48:26,240 --> 00:48:31,200 This is uh, this is, we call that second  generation elect- um congestion pricing.   433 00:48:31,200 --> 00:48:35,200 The third generation of, which they are thinking  about, they are actually planning to adopt in   434 00:48:35,200 --> 00:48:41,600 Singapore is based on gps. In that case, you don't  need this anymore, okay, what you need actually is   435 00:48:41,600 --> 00:48:53,440 a transponder on your car, um and um that will  that will basically tell the central operator   436 00:48:53,440 --> 00:49:00,880 where you are at a given point in time, and  if you drive very often during the rush hour,   437 00:49:00,880 --> 00:49:07,200 and on congestion road, you will receive a larger  bill during the end of the month than if you drive   438 00:49:07,760 --> 00:49:13,360 during more during non-rush hours or less  congested road, okay? So this type of policy will   439 00:49:13,360 --> 00:49:20,400 really guide people, make their optimal decisions  in terms when to travel, where to travel. Our   440 00:49:20,400 --> 00:49:27,360 goal, as I said, is to estimate that congestion  pricing, okay? The optimal congestion charge,   441 00:49:29,120 --> 00:49:33,440 the optimal congestion charge is going to  be effect- it's going to be essentially,   442 00:49:33,440 --> 00:49:38,080 this is externality or or the cause  that you impose to other people,   443 00:49:38,080 --> 00:49:43,200 okay? When you drive on the road. And that,  to estimate this, we actually need to know   444 00:49:43,200 --> 00:49:50,400 the relationship between the density and speed,  okay? So essentially what we need to know is that,   445 00:49:50,400 --> 00:49:55,360 given point in time, if there are already 10  cars on the road, okay, if you now decide enter   446 00:49:55,360 --> 00:50:00,400 the road, you become the 11th car, how much you  are going to red-, how much uh you are going to   447 00:50:01,280 --> 00:50:06,720 reduce the speed, therefore how much cost you're  going to impose on other road users. Therefore,   448 00:50:06,720 --> 00:50:12,560 we need to know this relationship between how many  cars on the road, and the speed of those cars,   449 00:50:12,560 --> 00:50:17,440 okay, and this is going to tell us, allow  us to actually estimate the cost you cause   450 00:50:17,440 --> 00:50:23,840 or impose on other road users, okay?  So the data we we have is from Beijing,   451 00:50:23,840 --> 00:50:30,320 um so these are uh this uh orange dot  is location for uh monitoring stations,   452 00:50:30,320 --> 00:50:35,520 traffic monitoring stations in Beijing, there  are more than fifteen hundred of them here,   453 00:50:35,520 --> 00:50:43,600 and uh we have data for every two minutes, we know  the traffic speed and flow in each point here,   454 00:50:44,240 --> 00:50:50,320 and for a whole year this is about half billion  observations, okay, so we aggregate the data   455 00:50:51,360 --> 00:50:55,840 into our interval so we know for  each point, we know how many cars   456 00:50:58,240 --> 00:51:03,120 on the road segment during given point in  time, and we know the speed, so based on,   457 00:51:04,000 --> 00:51:09,440 essentially the observations, on density or number  of cars on the road, and also observed speed,   458 00:51:09,440 --> 00:51:16,800 we can try to infer or estimate that relationship  between density and speed, okay? So we aggregate   459 00:51:16,800 --> 00:51:22,800 the data, so we have over 12 million observations,  and we have other variables including weather,   460 00:51:23,360 --> 00:51:29,520 you know wind speed, visibility, temperature,  all these things could affect speed, okay, so we,   461 00:51:29,520 --> 00:51:34,800 in the regression, we we control those variables.  This is the raw data, it shows you the average   462 00:51:34,800 --> 00:51:40,720 density and also speed. Of course, when there  are fewer number of cars on the road, the speed   463 00:51:40,720 --> 00:51:49,600 is higher and there are more cars on the road, you  know, per kilometer the speed is lower, okay, so   464 00:51:49,600 --> 00:51:54,320 you see this relationship. So what we are going to  do actually, we are going to estimate a function,   465 00:51:54,320 --> 00:51:59,920 speed as a function of density, okay, and there  are some uh technical or econometric issues that   466 00:51:59,920 --> 00:52:05,920 we have to deal with uh that I will escape, but  when we do the estimation, so you regress speed   467 00:52:05,920 --> 00:52:11,360 as a function of density, so this is the last  column that that you can focus on different   468 00:52:11,360 --> 00:52:18,240 specifications, try to control weather conditions,  baseline traffic speed during different railroad,   469 00:52:19,040 --> 00:52:25,040 etc. Okay, so the last column basically says, if  you increase the number of cars per kilometer by   470 00:52:25,040 --> 00:52:35,520 one, okay, going from 10 to 11, 11 to 12, or etc.  then the speed is going to reduce by 0.3 kilometer   471 00:52:35,520 --> 00:52:41,120 per hour, okay? That's the relationship. It's a  linear relationship that'll be estimated because   472 00:52:41,120 --> 00:52:46,800 if you look at the curve here, it's it's quite  linear, especially this part of the curve, okay?   473 00:52:48,240 --> 00:52:56,720 Okay, now there are some issues uh with technical  issues um if if you do that and then in this table   474 00:52:56,720 --> 00:53:00,320 we actually try to address the issue, we call  that endogeneity in economics,   475 00:53:02,000 --> 00:53:06,800 and we believe this is actually a better  specification, or better estimation results,   476 00:53:06,800 --> 00:53:12,080 so the last column basically says, well if you  increase number of cards by one, going from 10   477 00:53:12,080 --> 00:53:18,640 to 11 per kilometer, then the speed is going  to reduce by one kilometer per hour, okay,   478 00:53:18,640 --> 00:53:24,240 so the effect of density on speed is actually  much larger when you try to deal with some of   479 00:53:24,240 --> 00:53:30,720 the technical issues in the relationship, okay?  So in the end, we are going to use this number,   480 00:53:31,680 --> 00:53:36,720 and try to estimate the congestion  cost. We call that marginal external   481 00:53:37,280 --> 00:53:43,280 congestion cost, so this is the cost you impose  when you drive on the road to other road users,   482 00:53:43,280 --> 00:53:48,160 and this cost essentially comes from the time  loss, because you slow down other people,   483 00:53:48,880 --> 00:53:56,480 and time worth money, we translate really lost  time to money uh term, that's the external   484 00:53:56,480 --> 00:54:00,880 congestion cost, right, there are a lot of  assumptions that we have to use including,   485 00:54:00,880 --> 00:54:08,480 for example, what's the value of time? How much  does one hour worth to people? And we estimate   486 00:54:09,120 --> 00:54:15,920 um this marginal, this congestion cost curve.  So that congestion cost curve of course varies   487 00:54:15,920 --> 00:54:21,760 depending on the traffic density. If there are  already a lot of cars, you enter into the road,   488 00:54:21,760 --> 00:54:27,520 that's going to cause a high, much higher cost to  other road users than when there are fewer cars   489 00:54:27,520 --> 00:54:33,040 on the road, right? You see this relationship when  there are a lot already a lot of cars on the road,   490 00:54:33,040 --> 00:54:37,840 the congestion cost will be higher, okay,  because you slow down other people much more,   491 00:54:38,800 --> 00:54:45,680 right? So for Beijing, focus on this, we do this  for different days, for of different hours of the   492 00:54:45,680 --> 00:54:51,920 day, for different location, for example, this is  within the second ring road, second to the third,   493 00:54:51,920 --> 00:54:57,280 third to the fourth, fourth to the fifth, outside  the fifth ring road. So for example this point   494 00:54:57,280 --> 00:55:03,840 basically says well, if you travel during this  time of the day, roughly uh five or six pm in   495 00:55:03,840 --> 00:55:10,000 the afternoon, within the six, sorry, within the  second ring road, for every kilometer you travel,   496 00:55:10,000 --> 00:55:17,200 you are going to impose 92 cents on other  road users, so this is the cost to the society   497 00:55:17,200 --> 00:55:24,400 when you travel one kilometer during five or  six pm within the sixth ring road, right? This   498 00:55:24,400 --> 00:55:29,120 is the cost to the society but you do not take  that into a consideration in your own decision,   499 00:55:30,000 --> 00:55:36,640 right? You only consider your time cost,  your fuel cost, etc. but you don't consider   500 00:55:36,640 --> 00:55:43,120 the cost imposed on other people, okay? So this is  the external cost that we estimated for different   501 00:55:43,120 --> 00:55:47,360 hour of the day, as I said, different location,  if you look at outside fifth ring road, there is   502 00:55:47,360 --> 00:55:52,720 really not much of a congestion cost because uh,  congestion is not not bad outside, I mean, there   503 00:55:52,720 --> 00:56:00,000 are not many cars simply there, okay? Okay, so now  come back to this graph, so we essentially we can,   504 00:56:00,000 --> 00:56:06,400 we estimate these two lines here, and then we  can also estimate this curve, uh once we do that,   505 00:56:06,400 --> 00:56:12,400 we can then estimate the congestion charge, okay?  Think about these are the curves we estimated,   506 00:56:12,400 --> 00:56:18,800 right, this is the the private cost that  road users need to pay, they need to pay,   507 00:56:18,800 --> 00:56:24,480 incur, these are the social cost right, so the  difference is, as I said, is the congestion cost.   508 00:56:26,000 --> 00:56:32,800 So now let's think about, if we want to impose  the first scheme, we call that uniform pricing,   509 00:56:32,800 --> 00:56:39,440 okay? That is, you are going to charge a same  price per kilometer, no matter when people drive,   510 00:56:39,440 --> 00:56:44,720 where people drive, so that is, if you drive in  Beijing, any time of the day, anywhere in Beijing,   511 00:56:44,720 --> 00:56:50,240 you are going to need to pay a certain  price, okay? If we want to find that price,   512 00:56:50,240 --> 00:56:55,920 well that's going to be this, okay? So this  will be the price, that's one strategy,   513 00:56:56,640 --> 00:57:01,760 okay? But this is not the best strategy, because  we know during different time of the day, during   514 00:57:01,760 --> 00:57:06,560 different places, the city, at different place  of the city, the congestion cost is different,   515 00:57:07,360 --> 00:57:13,680 right? Um, as I showed you, so here, what we  do is, we are going to look at a time-varying   516 00:57:13,680 --> 00:57:19,280 pricing, that is, if you think about, well if you  travel during peak hours versus non-peak hours,   517 00:57:20,560 --> 00:57:26,000 this will be the demand curve for peak travel,  peak travel demand, this is non- peak hours,   518 00:57:26,000 --> 00:57:31,520 and this is gonna be the optimal congestion  charge during peak hours, this is optimal   519 00:57:31,520 --> 00:57:38,160 congestion charge during non-peak hours, okay?  This graph says, well, during peak hours,   520 00:57:38,160 --> 00:57:45,040 from seven to nine in the morning, five to seven  uh in the afternoon, you have to pay a higher   521 00:57:45,040 --> 00:57:52,000 price when you travel during non-peak hours. The  last one is more complicated. This is time-varying   522 00:57:52,000 --> 00:57:59,680 and location specific so we separate time into  peak and non-peak location, into more condensed   523 00:57:59,680 --> 00:58:06,080 areas within the third ring road, less congested  third to fifth and outside fifth ring road, okay,   524 00:58:07,120 --> 00:58:12,720 so this basically, is the congestion charge  during peak hour within the third ring road,   525 00:58:13,360 --> 00:58:19,760 this is during non-peak hours within the third  ring road, peak hour between third and fifth   526 00:58:19,760 --> 00:58:27,280 ring road, non peak hour between third and fifth,  right? Um so this is the uniform pricing that I   527 00:58:27,280 --> 00:58:31,360 told you. If you travel in Beijing, if we want  to use this structure, that basically means   528 00:58:31,360 --> 00:58:35,840 no matter when you drive, no matter where you  drive, you need to pay 10 cents per kilometer,   529 00:58:36,560 --> 00:58:42,640 okay? So think about, if you enter Beijing  proper, you need to pay 10 cents every   530 00:58:42,640 --> 00:58:48,000 kilometer you drive. This is one strategy, and  this is the time-varying strategy that I said,   531 00:58:48,000 --> 00:58:53,360 that I mentioned. During peak hour, you pay 13  cents per kilometer during 9 peak hour you pay   532 00:58:53,360 --> 00:59:00,640 9 cents, okay? But we can also look at more  elaborate strategy, that is time-varying and   533 00:59:00,640 --> 00:59:06,080 location-specific. So location, we have three  kind of categories, within third ring road,   534 00:59:06,080 --> 00:59:12,880 between third and fifth, outside fifth, right?  During the peak hour, you need to pay 33 cents   535 00:59:13,920 --> 00:59:22,480 within third ring road, 20 cents within third ring  road off peak hour, and um you'll pay less between   536 00:59:22,480 --> 00:59:28,880 third, fifth, and uh even, even more, even less  outside fifth ring road. Let's let's look at just   537 00:59:28,880 --> 00:59:35,360 the impacts on speed, um so that basically says,  if we use this uniform pricing, the reduction in   538 00:59:35,360 --> 00:59:41,440 speed is going to be 1.4 percent, um and if you  use time varying price, it's going to be this,   539 00:59:41,440 --> 00:59:46,560 if you use the other more elaborated strategy,  as I said, time-varying and location-specific,   540 00:59:46,560 --> 00:59:52,320 we should expect four to six percent reduction  within the third ring road in terms of congestion,   541 00:59:52,880 --> 00:59:58,480 reduction, oh sorry, increase in speed, uh  reduction in congestion, yeah. This uh this uh   542 00:59:58,480 --> 01:00:06,720 exercise shows we can actually use uh some  principle, economic principles to thinking about   543 01:00:06,720 --> 01:00:12,560 how to address a very important social problems  in China, right? What we need to do here actually   544 01:00:12,560 --> 01:00:21,120 is to apply that economic principle to the problem  using using data, using empirical analysis, right?   545 01:00:21,120 --> 01:00:26,240 So in the end, the policy we're going to make  actually need to be based on empirical evidence,   546 01:00:26,240 --> 01:00:30,880 right? Think about, as I said, if we want to adopt  congestion pricing, what should be the right level   547 01:00:30,880 --> 01:00:35,520 of congestion pricing? For different time of the  day? For different locations of the day? And what   548 01:00:35,520 --> 01:00:41,200 kind of impacts we should expect from these type  of policies, so our our framework actually allow   549 01:00:41,200 --> 01:00:46,480 us to do those things? So policymakers in China  actually increasingly aware of economic social   550 01:00:46,480 --> 01:00:50,480 all kinds of air pollution  and also congestion. I talked about, in 2015,   551 01:00:50,480 --> 01:00:54,880 Beijing Municipal Government actually announced  they were, they plan to use congestion pricing,   552 01:00:54,880 --> 01:00:59,680 use market-based pricing mechanism, to deal  with the congestion they're actually doing a   553 01:00:59,680 --> 01:01:03,680 lot of things also, or thinking about doing  a lot of things, in terms of air pollution,   554 01:01:03,680 --> 01:01:11,680 as well. Thirteenth 5-year plan for China,  we know that that is for 2015, 2016 to 2020,   555 01:01:12,480 --> 01:01:17,680 okay? In that five-year plan, if you read it,  they actually set a goal for PM2.5 reduction.   556 01:01:18,240 --> 01:01:23,280 The goal is to reduce non-compliance  days by 18 percent for major cities,   557 01:01:23,280 --> 01:01:29,600 and this is the first time ever in a five-year  plan they actually mentioned PM2.5, and Beijing   558 01:01:29,600 --> 01:01:36,480 Municipal Government has also its own five-year  plan, and the, their goal is reduce PM2.5 by   559 01:01:36,480 --> 01:01:41,973 30 percent by the end of the five-year plan, okay?  Now there was a 12-year plan of course, twelfth   560 01:01:41,973 --> 01:01:48,240 5 year plan, these were the targets for all kind  of things, not PM2.5 as I mentioned, but   561 01:01:48,240 --> 01:01:56,800 energy intensity reduction, carbon intensity, uh  sulfur dioxide, NOx, you know, chemical, you know,   562 01:01:56,800 --> 01:02:02,080 COD, these are water pollutants, and forest  coverage, okay and these are the achievements,   563 01:02:02,080 --> 01:02:08,560 uh at the end of the five-year plan, that  is in 2015, they actually achieve all these   564 01:02:08,560 --> 01:02:15,920 targets. These are the targets for thirteenth  5-year plan, so this is for 2016 to 2020,   565 01:02:15,920 --> 01:02:21,440 these are the plan, these are the goals, right? So  again, there are quite ambitious goals in terms of   566 01:02:21,440 --> 01:02:29,840 energy intensity, carbon intensity, all the other  things. As well as PM2.5 reduction, okay? Alright,   567 01:02:29,840 --> 01:02:36,560 we all know in 2014 China U.S. announced, when  President Obama visited Beijing, they announced   568 01:02:36,560 --> 01:02:43,720 this U.S.-China agreement on climate change, they  said um bilaterally we are going to actually uh   569 01:02:45,360 --> 01:02:52,560 agree on these things, as part of the Paris  Agreement that was later on made in 2015   570 01:02:53,120 --> 01:03:01,280 but U.S. pledged to do, is, the U.S. aims to  reduce emission by 26 to 28 percent below its   571 01:03:01,280 --> 01:03:10,320 2005 level in 2012, or 2025, China's goal was  to peak its emission, carbon emission by 2030.   572 01:03:10,320 --> 01:03:16,320 Okay, so that is, you should allow us to increase  carbon emissions over time because we are still   573 01:03:16,320 --> 01:03:23,920 developing, but we promise, by 2030, our uh carbon  emission will be peaked. We're gonna, not gonna   574 01:03:23,920 --> 01:03:31,120 increase beyond 2030. We're going to do a lot  of things uh to to achieve that including   575 01:03:32,320 --> 01:03:37,840 increased dense fossil fuels in  energy consumption by 20 percent by 2030. Okay,   576 01:03:38,960 --> 01:03:44,160 we all know that our president-elect  want to abandon the Paris Agreement,   577 01:03:44,160 --> 01:03:49,600 okay, but Chinese government said we are going  to fulfill and honor its commitment to agreement,   578 01:03:49,600 --> 01:03:58,800 so this is very welcome, right, news.  In fact, actually, China has been doing   579 01:04:00,080 --> 01:04:05,920 carbon reduction programs, such as cap-and-trade  programs in seven regions starting from 2013.   580 01:04:05,920 --> 01:04:14,160 So in seven uh locations, including Beijing uh  Guangzhou and Chongqing and Hubei uh et cetera,   581 01:04:16,640 --> 01:04:23,680 they set up cap-and-trade programs for for carbon  emissions. The goal was to actually, to have a   582 01:04:23,680 --> 01:04:29,440 national program start starting 2017, okay, so  these are pilot programs. The idea is that let   583 01:04:29,440 --> 01:04:35,120 government agencies, let firms understand how this  program work, and then in 2017, we are going to   584 01:04:35,120 --> 01:04:41,200 have a national program, uh that's gonna, this  program is gonna be larger, will be larger than   585 01:04:41,200 --> 01:04:49,280 the current E.U. emission trading program, okay,  and there will be significant benefit to reducing   586 01:04:49,280 --> 01:04:55,520 local air pollution. Although this program, the  goal of these programs is to peak carbon emission,   587 01:04:55,520 --> 01:05:01,840 but as we know, when you reduce carbon emission,  for example, from using natural gas rather than   588 01:05:01,840 --> 01:05:07,840 coal, actually the majority of the benefit will  be local pollutants, will be reduction in PM2.5   589 01:05:08,480 --> 01:05:14,080 and and associated health benefits from the  reduction of local pollutants rather than   590 01:05:14,080 --> 01:05:19,360 actually climate change benefit itself, right?  There are many studies, for example, look at   591 01:05:21,440 --> 01:05:26,560 U.S. SO2 program, where they show the  majority of the benefit actually not come from   592 01:05:26,560 --> 01:05:31,520 reduction in SO2 but the co-benefit, reduction in reducing PM2.5,   593 01:05:33,520 --> 01:05:40,560 okay? So I believe that in the process, in the  process of cleaning up China's own air, China   594 01:05:40,560 --> 01:05:46,560 actually is uniquely positioned, really, to be a  leader in climate change, and more importantly in   595 01:05:46,560 --> 01:05:52,640 actually producing clean energy for China, for the  U.S., for the world as a whole, okay? Which China   596 01:05:52,640 --> 01:05:59,760 has been doing, but right now, uh I think China is  in an even better position to do that, uh moving   597 01:05:59,760 --> 01:06:10,320 forward. Okay, so my overall thought, uh so this  is my overall thought, um so Chairman Mao said,   598 01:06:10,320 --> 01:06:15,840 in 1945, that the future is bright but the roads  are you know, twists and turns, so there are a   599 01:06:15,840 --> 01:06:20,880 lot of things we need to do, but in the end,  I think China, I I actually I have confidence   600 01:06:21,520 --> 01:06:26,400 that Chinese government will be able to  actually address the environmental issues   601 01:06:26,400 --> 01:06:31,840 uh although not in the immediate short term,  but in a 20, 30 years period or even shorter.