share interactive transcript request transcript/captions live captions
|
MyPlaylist
SPEAKER 1: The following is part of Cornell Contemporary China Initiative lecture series under the Cornell East Asia Program. The arguments and viewpoints of this talk belong solely to the speaker. We hope you enjoy.
SPEAKER 2: And for our last talk today, we have our very own Professor Shanjun Li who joined the Dyson school in 2011 as an economist. And he had been working prior to that with a think tank in DC. Yes. I've been there several years since finishing his PhD in economics at Duke.
So we're very happy to have him here. Since arriving here, he has published many interesting things but also has recently co-founded together with [INAUDIBLE], the sizer, the Cornell Institute for China Economic Research, which has often partnered with us, CCCI, with this lecture series. So without further ado, let us welcome Shanjun.
PROFESSOR SHANJUN LI: All right. Thanks to you all for coming. All right, so I have to warn you. Today, I'm going to show a lot of things that are quite depressing, you know, in this kind of weather, in this type of political environment. But in the end, there will be something more positive. So stick with me till the end, if you can. OK?
So I'm gonna talk about environmental and energy challenges and policy options in China. OK, in 2005, China was the largest energy consumer, accounting for 24% of world total energy consumption. OK, 24%. I want you to think about these numbers really for a moment.
By far, the largest coal consumer. 50% of world total coal consumption. The largest CO2 emitter. 30% of world total. The largest automobile market. 26% world total automobile sales.
These shares are only getting bigger. That is, These things really have very big implications, if you think about it. Think about energy demand, right. We know energy demand-- for example, oil-- is inherently now actually world market. So anything happens in China in terms of automobile demand or in terms of gasoline demand, it's going to put pressure or affect oil market. Affects the price, affect the price may be here in the United States.
Think about CO2 emission. CO2 we know is a global pollutant. A ton of CO2 emitted from China is no different from one ton of CO2 emitted in the US in terms of it's impact on climate. In terms of impact for the world. So we care about not only CO2 emissions from here but also in China and elsewhere.
So domestic policy in China really have global impacts. And that's why we really should understand what's happening in China. And really, that can inform us in terms of policy making in the US. Because of these numbers that I mentioned, air pollution and traffic congestion really have become the most pressing challenges in major urban areas in China.
I'm going to talk about these two issues and actually going to focus on traffic congestion in the later half of the talk. In turn, this is what air pollution-- we know actually affect quality of lives in urban areas. It really affect firm location, affects labor productivity, affects human capital investment, and affects, in the end, economic growth in the long run.
OK, so this shows GDP growth on the right axis. And energy consumption on the left. For the past 50-- actually more than 50 years-- can look at increasing GDP. That is this black line here. So we all know it's annual average is about 10% during the last 35 years.
If you look at coal consumption, this gray area is increasing at, roughly, the same speed. If you look at oil consumption increase, look at gasoline-- sorry, this is natural gas. And hydro and nuclear is the blue. So you can look at the majority of the fuel source is coal.
So in 2014, 89% of energy was from fossil fuel, including coal, 66%, oil, 18%. This is really mainly for automobiles. Very little natural gas. If you look at US, this will be-- for coal-- would be roughly 30%. Natural gas will be larger there. Oil would be, I think, roughly the same here.
So fossil fuel, especially coal, is the dominant source. And if you look at all the fuel sources here, coal is the dirtiest one. By far, the dirtiest one. That was the consumption. What about energy balance? So this is the difference between what China produced and what China consumed.
So these are inactive. You look at this for the majority of this. Actually, these are inactive, meaning China has been importing oil. And also, increasingly coal, natural gas from other countries. That's why I said in the beginning that domestic demand in China actually has important implications for the world market as a whole.
So this is new automobile sales in China. This is annual sales in China in 2001 was less than 2.5 million. Now it's here. Actually, 2015. The total new vehicle in sales was over 21 million units in China. This is for the US.
China surpassed the US in 2009 to become the largest automobile market. So the dramatic increase here in oil consumption was largely driven by the increase in vehicle ownership in China. All right, so this is crude oil consumption.
In China, in orange, and the word, blue, this is relative to 2001 level. So everything is relative to 2001. You look at the increasing oil consumption in China. And then, the blue bar is the world total consumption relative to 2001.
Again, look at this orange bar relative to the blue bar. China accounted for 46%, increasing world's oil consumption during this data period. So this is CO2 emission. Again, relative to 2001 level. Orange is for China. Blue is for the world.
During this period of 2001, 2013, the increase in CO2 emissions in China accounted for 60% of growth in global total. So next, I'm going to show you a few slides about environmental quality or environmental issues in China. And before I show them, we all understand there is big issues. Environmental degradation is a big problem there.
And if you think about why we have these issues, Really, this is a perfect storm for China. So several really important factors combine together that contributes to the really dramatic degradation in terms of environmental quality in China during the past 15 years.
So the first one we all know is the unprecedented growth in the industry sector and vehicle ownership that I showed you. And China heavily relies on fossil fuel, especially coal, for energy, as I showed you, as well. We also know the political environment, the political personnel system, is a top down approach. That is the government officials at the federal level appoint provincial government officials. They, in turn, appoint government officials in the municipal and in county levels.
And the most important criteria used to be that really the sole criteria for promotion was GDP growth. As a result, government officials really have a very strong incentive to do all kinds of things to increase their GDP growth, including attract a lot of very dirty industries. Petrochemical, plastic production, petroleum refineries, those types of stuff to increase GDP without giving, really, a lot of thoughts about the environmental consequences.
So environmental protection really was an afterthought for a lot of government officials in making decisions how to actually balance economic growth and environmental protection. It has changed during the last five years. I'm gonna talk about them toward the end.
So at the same time, we are trying to have a really very lax environmental regulations and enforcement. To give you a sense, we know in the United States the environmental regulation agency is Environmental Protection Agency, the EPA. China's counterpart is Ministry Environmental Protection was established only in 2008 as a cabinet level ministry.
In the US, we have about 17,000 employees under EPA. China's MEP employs about 2,600 people. Compare the size of the population, the country, and this is a much, much smaller number of course. Right? Another important issue is that local government bureaus, local environmental regulators-- or think about this-- local MVP offices are actually part of local government.
In the US, EPA has regional offices. But those regional offices are directly managed from EPA in DC, OK. So regional directors are actually appointed by EPA director. But in China, the local environmental bureau chiefs are actually appointed by local government officials. And I think about it, these are really environmental enforcers, regulation enforcers, and also monitors.
And they have to monitor people who will appoint them the later on. Or promote them later. On. So this create, of course, perverts the incentive in terms of environmental regulation in China. All right. So I'm gonna, as I mentioned, show you some quite nice pictures.
So this is one of them, this air pollution. This is from Hubei, which is the largest steel producer in the world if you don't consider China, as a whole, as a country. So there is a saying that if you look at steel production in the world, China is the first. Hubei province is actually second.
Now let's look at graphically compare Beijing's PM 2.5 and L.A. L.A is considered a area in the US with the worst air pollution. For those of you who don't know PM 2.5, this is called particular matter. These are fine particular matter with a diameter that is less than 2.5 millimeter. 2.5 millimeter is about 130th of a human hair. So it's tiny, tiny particular matter. And these are really damaging to human health.
During the past 20 years-- actually, 25 years-- there was a lot of medical research and really a lot of research that look at health impacts PM 2.5 and found them to be very damaging to human health because they are really small. You can breathe them. They can then go deeply into your lungs and bloodstreams. Cause lung cancer, and cause heart problems.
So if you look at the daily concentration here in China, in orange, from 2012 to 2013, and the black is for-- sorry, this is for Beijing. And this is from L.A. I also want you to look at the EPA categories here. So this level is good to moderate. This range is unhealthy. This is very unhealthy. This level above is what EPA calls hazardous, meaning EPA actually will advise people not to go outside.
But that happens quite often in Beijing. For those of us who lived in Beijing, we all know this. But the record before this date was nearly 600 micrograms per cubic meter. This is the measurement we use. I mean, if you look at L.A, again, this is the most polluting city in the US. There's no comparison.
So another thing to mention is that US has a daily standard that is 35 micrograms per cubic meter. And for L.A, actually, most of the days were below this 35 line there. This is for whole China, PM2.5 concentration on an annual basis. So if you look at these dark red and this red, provinces, their annual level is over 50 or even over 70.
Micrograms per cubic meter. This is the unit we use. If you look at the US annual standard, 12. WHO, World Health Organization, guidelines is only 10. So for most, actually, we know that this part of China has a lot of population. So for most of these areas, the concentration level is four, five, six, seven times the WHO guideline.
OK, we know air pollution is bad. OK, what is the cause of the air pollution? So there are a lot of research on that. Well this is not quite recent anymore because this is an old slide. There was an article in Nature of September 2015. They estimated that air pollution accounted for 1.3 million premature deaths in China in 2010. And this number accounted for 40% of world total premature deaths.
It's 2007 World Report. They estimate the cause of pollution in China, and the health cause of air pollution, health cost alone, was 1.2 to 3.8% of GDP in 2013. I can give you a quick idea how the estimate is cost. What they use is so-called dose response functions. They have data, for example, in terms of air pollution, and mortality, and morbidity rates in different locations in the US or elsewhere.
They tried to find the relationship between pollution and the mortality rate. And they tried to estimate that relationship. That's called dose response relation, as well as establishing the medical literature. And then, they can use those functional relationships to estimate, well, if China's air pollution was lower than what we had before-- for example, as low as the World Health Organization guideline-- what would happen to their premature deaths?
And these are what happened to all kinds of disease. And health cause is only one part of the cause, of course. There are other cause. OK, a 2012 MIT study estimated that health cause air pollution to be $112 billion US dollars in 2005. 5% of GDP. So these are really big numbers, if you think about it.
All right. Soil pollution. About 20% of China's farmland are polluted from industry waste water. And the levels of organic pollutants are often 20 times the standard. Water pollution, 60% of groundwater water are contaminated. 90% for cities. And 2/3 of China's rural population use water contaminated by human industry waste.
As a result, especially for rural areas, Ministry Environmental Protection classify 459 villages as cancer villages. If you look at location, it's not surprising. These are located in heavily industry areas with, actually, a lot of so-called industry clusters. Cluster of textile firms, cluster of steel mills, et cetera. And research has found that water pollution is the main cause for cancer villages.
These are really, as I mentioned in the beginning, kind of, [? jarring ?] pictures. You might ask, where do we go from here? So some of you actually saw these pictures before. So I put these two pictures. These are two pictures in different continents and different time periods.
I want you to guess what is this about. What was this? And then, what was this? Exactly. This is actually a dust bowl from the US in the 30s that happened in the Great Plains. Texas, Oklahoma, and those areas. And the main cause of the dust bowl was a combination of weather, but mostly erosion of topsoil due to intensive agriculture.
And this is Ningxia in 2010 in China. This is one of the sand storm episodes. OK, so these two, this is New York City. George Washington bridge. This is was from 1970. And you can't really see much of this, right?
And this is Beijing. Sometimes, or actually very often, this is Beijing. So you think about, well, US had very bad environmental issues in the 60s, 70s, or 50s. 40s earlier. Now we all know if you study environmental history, there were a lot of environmental disasters. One of them was London's smog, 1952. 4,000 people died during that few days. Right?
So there were incidents like this in the US in Pittsburgh in a place called [INAUDIBLE]. There were many people died in a short span of a few days because combinations with air pollution and bad weather conditions [INAUDIBLE]. So how did US do it? How did UK clean up their air?
OK, environmental regulations in the US. So 1970 was a very important year. And before 1970, many things happened, including-- I don't know whether you know this book called The Silent Spring, by Rachel Carson. And many, also, bad environmental incidents, and that lead to environmental movement. And Richard Nixon created EPA in 1970.
They also passed Clean Air Act. This act is very important. The goal of that act was to regulate air pollutions. So they set up what they call National Ambient Air Quality Standards. And for six pollutants, including carbon monoxide, sulfur dioxide, particular matter, it was PM10, NO2, and ozone and lead, they said, well, for each county, you have to actually comply with the national standard.
If you don't comply, then we are going to pull out federal fundings. You are going to suffer from consequences. So you have to do anything you need to do to make sure you comply with the air quality standards. So different states use different strategies to do that. This was a very important regulation in the US in terms of environmental regulation.
Numerous laws passed since 1970. They addressed a lot of things. Clean air, clean water, energy conservation, hazardous waste, pesticides and, more recently, toxics from power plants. Let's look at what happened from 1970. Actually, to a large extent due to environmental regulation. Look at SO2 air quality. So from 1980 to 2010, there was a 23% decrease in national average in terms of SO2. In terms of NO2, there was a 52% decrease in national average. Carbon monoxide, 82% drop. And 28% decreasing ozone. So it's a remarkable improvement during this 20 years. Actually, 30 years, right.
OK. What's the effects of our environmental legislations? Well we all know the air now is very good. But if you look at the health consequences or health benefits, if you look at 1990 alone, our studies show that clean air programs prevented, this year alone, over 20,000 premature deaths, this many case of chronic bronchitis, and heart disease, asthma attacks, a all these things, and also 10 million lost IQ points in children from lead reductions.
These are numbers that are produced by researchers. And you know, these are the health benefits in 1990 alone. And there are recent studies that actually try to quantify the benefit of Clean Air Act in more recent years. For example, they found that premature deaths was reduced by over 160,000 in 2010 due to Clean Air Act regulations.
So I showed you the progress that US made during the last 30, 40 years. And I mentioned briefly the [? ohmmeter ?] regulations, the big regulations. And now I'm going to go through very quickly [INAUDIBLE] [? ohmmeter ?] regulation. So think about China. There are many options on the table, right.
Some have tried by other countries. Some have not, on a large scale. So Chinese government are thinking about these policies. And we need to decide what are the policies that we should use to address the issues. So economists tend to categorize different approaches into three categories.
The first we call that command and control approach. To give you an example, for example, we ban DDT. This is pesticide. We ban CFCs. These are ozone-depleting CFCs. And we have these type of regulations. Complete prohibition of use of some of the chemicals. We can do that.
Second type of command control approach is less forceful, but the government, essentially, mandate firms-- coal fire plants, or chemical companies, or petroleum refinery companies-- to use the best available technology on the market in terms of environmental performance. You have to do that so government can mandate that.
So these are within this command and control approach. Think about the pros and cons, right. Well the government has a lot of control, so you can expect some results will be achieved. In terms of the environmental performance, you can't expect that certain improvement will be achieved. But you have to think about at what cost.
The second approach called market based approach. This approach, basically, consider clean air, clean water as resources, as natural resources. And we need to think about how to allocate these resources, just as we think about how we allocate rules on the markets. Computer, clothes, or anything that we purchase on the market.
And we know for things that we purchase on the market, price is the mechanism, which guides us where these goods should go to or who the goods should go to. The goods should go to those who value them the most, right. So this approach basically says, well, let's think about really treating air and water as natural resources.
And let's think about using market mechanism to allocate resources. That is, if you are a power plant, you generate pollutions. That is, you are using clean air for your production. And you need to pay for the input that is clean air.
Under this approach, for example, as I mentioned, if you need to pay for clean air, we call that tax. Example is carbon tax. If power plants generate carbon emissions, for each time carbon emission, you need to pay a price. How much you need to pay depends on the damage carbon does to the society. So we need to quantify that damage.
Gasoline tax is another kind of tax that try to actually control for these damages. Subsidy or electric cars are other fuel saving technologies. This is essentially putting a price on using these resources. For subsidy, basically, you give you a subsidy to reduce your pollution. It's the opposite to the tax.
Cap and trade, you hear a lot about this. Cap and trade. That's another type of market based approach. The idea is that, if you want to use tax-- for example, use carbon tax-- a lot of times, consumers or general public do not like the idea of tax, especially in the US, and other countries, as well. One of the reasons is they don't trust the government in terms how they are going to use those tax revenues later on.
So some economists suggest, well, let's think about an alternative way to do this. We call that cap and trade. The idea is that if we want to reduce air pollution-- for example, if we want to reduce carbon emissions-- in order to prevent disastrous scenarios that happen, let's think about, well, carbon emissions should be below this certain threshold for many, many years onward, from this year onward.
And so we need to set a cap. That is, each year, we cannot emit more than this. Then, we are going to have licenses allocated to firms so that it-- now, firms emit-- each time a carbon, they need to surrender one unit of license to the garment, OK.
So in the end, the cap should not be violated. Once you allocated the permits to France-- and there are many different ways to allocate permits. One of them is based on history emissions. Once you allocate these permits, actually firms then can trade among themselves.
For example, as a firm, who had 200 transfer permits. But during my production process, I emitted 250 times. Then, my own permit is not enough. But what I can do is I can actually go to the market and buy permits from other friends. If I produce only 150 times the CO2 that year, but I have 200 permits, then I can sell that extra 50 kinds of permits on the market to other firms.
So you allow firms to trade. In the end, the permit, the right to pollute, you can think about, will have a market price. The market price was determined by the cap, by actually the cost of firms in reducing their pollution. What happens in this system is that the permit price actually is gonna guide as an actual signal or mechanism to guide firms to ultimately choose their abatement behavior or behavior of reducing emissions.
For example, if I am a very, very efficient firm, I can easily reduce my air pollution without incurring much of a cost. Another firm, that is a firm that is not very good in terms of technology, so it's going to take a lot of effort or cost for that firm to reduce emissions. So think about what would have happened to this type of firms.
Well, if I'm very good at reducing pollution, in the end, I'm actually going to reduce a lot of pollution. And I'm going to sell a lot of premise to the other firm who is not as good at reducing pollution. So he will not reduce a lot of pollution. But that firm is going to use a lot of permits from the market. It's going to buy a lot of permits from the market to do that because, for that firm, reducing pollution by himself actually is more costly than buying from the market.
So in the end, we actually showed through theory, through empirical studies based on past experience this type of system, cap and trade, can be very cost effective in reducing emission. That is, if you want achieve that cap or a certain level emission, this actually is very efficient in doing that. So we are going to use a lower cost in achieving the same amount of reduction than, for example, this type of approach because this type of approach do not give firms a lot of freedom in choosing their best strategy.
So these type of strategy, for example, technology mandates-- assumes the government knows a lot about the technology, and they will choose the best technology, which actually is not often the case. So there are a lot of examples in the US. For example, in the 80s, the way that the US phase out leaded gasoline was through this type of idea.
And their program was actually under Ronald Reagan. Also, to permit trading, which is a very successful program in reducing SO2, was authorized by first Bush in 1989. So cap and trade actually was a Republican idea. But we all know in recent years, a lot of lawmakers call that cap and tax. They didn't really like this idea. The blocked a lot of proposals in the Congress.
Emission trading for CO2. So this is European Union Emissions trading for CO2. This is the largest emission trading program or cap and trade program in the world started from 2005. I'm going to talk about China's CO2 cap and trade programs.
So this is one of the policy that are really being used or being considered in many parts of the world to deal with not only local pollutants, such as SO2, but also global pollutant, such as CO2. So the third category [INAUDIBLE] information let consumers know the cause of climate change, the damage of climate change. Therefore, think about more carefully how much electricity you use, how much heat you use, and think about the actions you take right. [INAUDIBLE] norms, social pressures. These are really in this category.
This could be helpful. Very often, they will be helpful when you combine this and these together. As I mentioned, traffic congestion is really, to me, really one of the two most pressing issues in urban China. And there are many options to deal with traffic congestion. And I'm going to talk a specific approach that we propose. It's a market based approach.
And look at how we actually set up that policy. And how to implement it, what would be the consequences of that policy. So in this paper, there's really a combination of economic insights. Very basic, fundamental, economic insights. Big data on a very important social issue.
So if you rank worst cities in terms of traffic congestion, the top one is the worst. So Mexico City was the worst. It's top 15. If you look at them, seven of those cities are actually from China.
A lot of them, if you look at them, are actually from need or mid-income countries. LA has the worst traffic congestion in the United States. It's number six. This is based on 2015 data.
So this is Beijing 15 years ago. Actually, 15, 16 years ago. And this is Beijing now. I showed you the dramatic increase in vehicle ownership before. From an economic perspective, traffic congestion is very simple, actually. Fundamentally, excess demand of road capacity.
Why there is as excess demand? We think, well, that's because mispricing. If a good resource is free of charge. If this is something that is a good thing for people and it is free, then the demand for that will be high. That's two for road capacity.
So there are two broad strategies to deal with traffic congestion or to deal with exit demand issue. One type of strategy, we call that supply-side strategy. For example, building more new roads, improving public transportation. Think about what that would do if you built more roads.
Well it's going to reduce congestion for sure. But that will also reduce the cost of travel for people. That is, reduce the price that people have to pay for travel. The price that I'm talking about, think about it's the cost of time. If something becomes cheaper, then the demand for that will be higher.
So what I'm saying is when you build more road, you're going to reduce travel cost. Therefore, you are going to lead to more driving, higher demand for driving. That, in turn, will lead to congestion again. So in the short term supply side, policies can work and tend to work. But in the long term, they cannot fundamentally deal with or address congestion issues.
Another set of strategy, we call them [INAUDIBLE] strategies. For example, the command and control approach that I mentioned. Driving restrictions that are being used in many cities in the world, including Mexico City, Bogota, some other South American cities, and Beijing, and probably New Delhi soon. The idea here is that, well, if there are too many cars on the road, why can't they restrict some of them from driving.
So in Beijing what they do is, depending on the last digit of your license, well, you can only drive four days out of five days. So this is very straightforward, kind of a linear thinking. If there are too many cars, let's take some of the cars off the road by forcing people or prohibiting people from driving. By not allowing people to buy cars.
This we call the command and control approach. It might work. Sometimes actually might not work. The other approach, as I mentioned, we call that market based or price based approach. So here, we are gonna use, actually, price signals as a mechanism to allocate resources. The resource here is road capacity. The same thing as clean air.
We can use price signal if you're thinking about allocating resources. But if you compare command approach via the price based-- we call that congestion pricing-- congestion pricing is going to be much more effective. Or it's going to achieve the same goal with the least cost.
The reason, as I mentioned a moment ago, is that when you use this market based approach of [INAUDIBLE] pricing this is instance, you have a lot of margins. Actually, you gave travelers a lot of margins to adjust their behavior. For example, when you have a congestion pricing, that can induce adjustment for travelers in many markets, such as trip frequency, how many times they travel, what kind of mode they use, what's the time of travel, and what's the route they use.
So they are going to adjust their behavior or travel behavior accordingly in an optimal way under this pricing strategy. So this is actually what we call first best policy in addressing traffic congestion. And this first best policy addressing congestion was first proposed by William Vickrey in 1955.
William Vickrey is a Nobel Prize winner in economics, 1992. He made a lot of very important contributions in many areas of economics, including auction. And in 1955, he proposed congestion pricing to deal with urban congestion in New York City. And then, you talk about congestion pricing for urban transportation in general in the US.
In 1963 in his American Economic Review article, he mentioned this. He said, "No other areas are pricing practices so irrational, so out of date, and so conducive to waste as in urban transportation." So basically arguing, well, pricing practices in the US was really out of date. Essentially, there was no pricing. So we had very serious mispricing of road capacity. As a result, we have congestion issues everywhere in urban areas.
OK, if you think more about it, congestion is what economists call a classic externality. Externality arises when one's action, when my action, inadvertently affects other people. For example, when I drive on the road, I will actually slow down other people on the road. But when I make my decision in terms of when I'm going to travel, how far I'm going to travel, I actually don't consider my impact on others into my decision.
So my action is causing impacts or affecting others that I'm not considering or taking into account. And the impact on my travel on others is externality. The same thing is pollution. When firms produce electricity, for example, they will make their decision based on the cost of the input, based on the price of the output, but they don't actually take into account the amount of pollution they will produce.
And therefore, they are going to lead to health consequences. We call them externalities, as well. Arthur Pigou, in 1920, actually talk about externalities. And he proposed use tax to correct for this externality generating activities. We call that Pigouvian tax.
So congestion pricing is, essentially, one type of congestion cost. That is, if your action is hurting other people, you have to pay for your actions for the damage that you cause. So in terms of congestion, you actually need to pay the cost that you impose on other road travelers.
How much should you pay? Well that's called congestion pricing. What's the level of congestion pricing? That's an empirical question. So this is the cost of travel. Think about this as a demand of travel. This is a density.
Number of vehicles on the road per kilometer. And at a certain level, below this level of density. Then, this is the cost. The travel cost is flat. Travel cost included in time cost, included in operating cost, and road maintenance cost. This is flat. But then the number of vehicles is about this threshold. Then, there will be a difference between-- this is what we call a cost average. Sorry. We call that marginal [? private ?] cost. Think about the cost that travelers pay themselves, and the cost to the society. This we call marginal social cost.
So there will be a difference between the cost to the society and the cost to individual travelers. The reason there is difference is because of congestion. At this point, when we add one more traveler on the road, it's going to slow down other people on the road. Therefore, it's going to take longer for them to travel. And that will be a cost because time is money.
So the difference between these two lines is externality from congestion. And this is what we call a demand curve for travel. When price is high, you travel less. Price is low, you travel more. So Pigou in 1920 said, in this type of situation-- he look at other type of externalities-- well we can actually impose a tax so that, as a society as a whole, we achieve the best outcome.
If you don't do anything, if the government does not do anything, then the society actually will achieve this outcome, this level of density, and there will be a big wedge between the social cause and private cause. There'll be a lot of congestion. This is not ideal from a society point of view.
From a society point of view, this point is the best. So in order to achieve this point, this level of density or travel demand, we actually need to impose a tax. So the amount of tax should be this much. OK, so this is the [INAUDIBLE]. We call that congestion charge.
So in this paper, what we are gonna do, we are gonna look at Beijing. We are gonna try to actually estimate these curves. This curve, this curve, this curve, and try to estimate the congestion pricing. That is how much you should charge road users so that in the end, if you achieve a level of congestion, that it actually is optimal for the society point of view.
All right. So there are many cities around the world using congestion pricing starting from Singapore in 1975. There are several European cities adopting early 2000. Studies show that with this congestion pricing, practices actually reduce congestion by 10% to 30% in different cities of the world.
So Beijing municipal government in December 2015 announced that they will introduce congestion pricing. So now they are actually studying congestion pricing, thinking about how to implement it, and thinking about the potential impacts of the policy. So what we are gonna do is actually really provide suggestions or information for them to use to think about how they should implement the practice.
So this is Singapore. To just give you an idea of what they do, this is electric road pricing. So every car has-- what do you call that? Receptor? And there's also a credit card here. When this car pass this point, this will receive a signal. And there are different things at different points on the road, there are many structures like this.
And at the end of a month, essentially, your charges will be deducted from the credit card. We call that second generation congestion pricing. The third generation, which they are thinking about-- they are actually planning to adopt in Singapore-- is based on GPS.
In that case, you don't need this anymore. What you need, actually, is a transponder on your car. And that will, basically, tell the central operator where you are at a given point in time. And if you drive very often during the rush hour and on congestion roads, you will receive a larger bill during the end of the month than if you drive more during non-rush hours or less congested roads.
So this type of policy will really guide people to make their optimal decisions in terms when to travel and where to travel. Our goal, as I said, is to estimate that congestion pricing. The optimal congestion charge. The optimal congestion charge is going to be, essentially, that this is an externality or the cost that you impose to other people when you drive on the road.
And to estimate this, we actually need to know the relationship between the density and speed. So essentially, what we need to know is that given point on time. If there are already 10 cars on the road, if you now decide to enter the road and become the 11th car, how much are you going to reduce the speed?
Therefore, how much cost are you gonna impose on other road users. Therefore, we need to know this relationship between how many cars are on the road, and the speed of those cars. And this is gonna tell us, allow us, to actually estimate the cause you cause or impose on other road users.
So the data we have is from Beijing. [INAUDIBLE] is a location for traffic monitoring stations in Beijing. There are more than 1,500 of them here. And we have data for every two minutes. We know the traffic speed and the flow in each point here. For a whole year, this is about half billion observations.
So we aggregate the data into our interval. So we know, for each point, we know how many cars are on the road segment during a given point in time. And we know the speed. So based on, essentially, the observations on density or number of cars on the road, and also observed speed, we can try to infer or estimate that relationship between density and speed.
So we aggregate the data. So we have over 12 million observations. And we have other variables, including weather, wind, speed, visibility, temperature. All these things could affect speed. So in the regression, we control those variables.
This is a raw data. It shows you the average density, and also speed. Of course, when there are fewer number of cars on the road, the speed is higher. When there are more cars on the road per kilometer, the speed is lower. So you see this relationship.
So what we are gonna do, actually, we are gonna estimate a function. Speed as a function of density. And there are some technical or econometric issues that we have to deal with that I [INAUDIBLE]. But when we do the estimation-- so you regress speed as a function of density.
So this is the last column that you can focus on. Different specifications. Try to control weather conditions, baseline, traffic speed, different railroad, et cetera. So the last column basically says, if you increase the number of cars per kilometer by one, or get going from 10 to 11, 11 to 12, or et cetera, then the speed is going to reduce by .3 kilometer per hour.
That's the relationship. It's a linear relationship that will be estimated because if you look at the curve here, it's quite linear, especially this part of the curve. OK, now there are some issues with technical issues if you do that. And then, in this table, we actually try to address the issue. We call that [INAUDIBLE] economics. And they believe this is actually a better specification or better estimation results.
So the last column basically says, well, if you increase number of cars by 1 going from 10 to 11 per kilometer, then the speed is going to reduce by one kilometer per hour. So the effect of density on speed is actually much larger when you try to deal with some of the technical issues in the relationship.
So in the end, we are going to use this number, and try to estimate the congestion cost. We call that marginal external congestion cost. So this is the cost you impose when you drive on the road to other road users. And this cost essentially comes from the time loss because you slow down other people, and time is worth money.
We translate really lost time to money term. That's the external congestion cost. There are a lot of assumptions that we have to use, including, for example, what's the value of time? How much is one hour worth to people? And we estimate this congestion cost curve.
So that congestion cost curve, of course, varies, depending on the traffic density. Is there are already a lot of cars, you enter into the road, that's going to cause a much higher cost to other road users than when there are fewer cars on the road. Right? You see this relationship.
When there are already a lot of cars on the road, the congestion cost will be higher because you slow down other people much more. So for Beijing, focus on this. We do this for different hours of the day for different location. For example, this is within the second ring road. Second to the third, third to the fourth, fourth to the fifth, outside [INAUDIBLE].
So for example, this point basically says, well, if you travel during this time of the day-- roughly 5:00 or 6:00 PM in the afternoon within the second ring road, for every kilometer you travel, you are going to impose $0.92 on other road users. So this is the cost to the society when you travel one kilometer during 5:00 or 6:00 PM within the six ring road.
This is the cost to the society, but you do not take that into consideration in your own decision. You only consider your time cost, your fuel cost, et cetera. But you don't consider the cost imposed on other people. So this is the external cost that we estimated for different hour of the day. As I said, different location.
If you look at outside [INAUDIBLE] ring road, there is really not much of a congestion cost because congestion is not bad outside. There are not many cars, simply, there. OK, so now come back to this graph.
So essentially, we estimate these two lines here. And then, we can also estimate this curve. Once we do that, we can then estimate the congestion charge. Think about these are the curves we estimated. This is the private cost that road users need to pay, incur. These other social costs.
So the difference is, as I said, the congestion cost. So now lets think about if we want to impose the first scheme, we call that uniform pricing. That is, you are going to charge a same price per kilometer, no matter when people drive or where people drive. So that is, if you live in Beijing any time of the day, anywhere in Beijing, you are going to need to pay a certain price.
If we want to find that price, well, that's going to be this. So this will be the price. That's one strategy. But this is not the best strategy because we know during different times of the day at a different place of the city, the congestion cost is different, as I showed you.
So here, what we do is we are going to look at a time varying pricing. That is, if you travel during peak hours. Worse is non-peak hours. This will be the demand curve for peak travel demand. This is non-peak hours. And this is going to be the optimal congestion charge during peak hours. This is optimal congestion charge during non-peak hours.
This graph says, well, during peak hours from 7:00 to 9:00 in the morning and 5:00 to 7:00 in the afternoon, you have to pay a higher price than when you travel during non-peak hours. The last one is more complicated. This is time varying and location specific.
So we separated time into peak and non-peak location into more congested areas within the third ring road, less congested At third to fifth, and outside fifth ring road. So this, basically, is the congestion charge during peak hour within the third ring road. This is during non-peak hours within the third ring road. Peak hour between third and fifth ring road. Non-peak hour between third and fifth.
So this is the uniform pricing that I told you. If you travel in Beijing, if we want to use this structure, that basically means, no matter when you drive, no matter where you drive, you need to pay $0.10 per kilometer. So think about if you enter Beijing proper, you need to pay $0.10 every kilometer you drive. This is one strategy. And this is the time variance strategy that I mentioned.
During peak hour, you pay $0.13 per kilometer. During non-peak hour, you pay $0.09. OK? But we can also look at a more elaborate strategy that is time varying and location specific. So location, we have three categories within third ring road between third and fifth, outside fifth.
During the peak hour, you need to pay $0.33 within third ring road, $0.20 within third ring road off peak hour, and you pay less between third, fifth, and even less outside of fifth ring road. Let's look at just the impacts on speed. So that, basically, says if we use this uniform pricing, the reduction in speed is going to be 1.4%.
And if we use time varying price, it's gonna be this. If you use the other more elaborate strategy-- as I said, time varying and location specific-- we should expect 4% to 6% reduction within the third ring road in terms of congestion. Oh, sorry. Increase in speed. Reduction in congestion.
This exercise shows we can actually use some economic principles to think about how to address very important social problems in China. What we need to do here, actually, is to apply that economic principle to the problem using data, using improper analysis. So in the end, the policy [INAUDIBLE] actually need to be based on empirical evidence.
Think about, as I said, if we want to adopt [INAUDIBLE] pricing, what should be the right level of congestion pricing for different time of the day, for different location of the day, and what kind of impacts we should expect from these type of policies? So our framework will actually allow us to do those things.
So policy makers in China are actually increasingly aware of economic social [INAUDIBLE] air pollution and also congestion. I talk about in 2015, Beijing and municipal government announced their plan to use congestion pricing. Use market based pricing mechanism to deal with the congestion. They are actually doing a lot of things also, or thinking about doing a lot of things in terms of air pollution, as well.
35 year plan for China. We know that is for 2015, 2016, to 2020. In that five year plan, if you [INAUDIBLE], they actually set a goal for PM 2.5 reduction. The goal is to reduce noncompliance days by 18% for major cities. And this is the first time ever in a five year plan they actually mention the PM 2.5.
And Beijing municipal government has also its own five year plan. And their goal is to reduce PM 2.5 by 30% by the end of the five year plan. Now there was a 12th five year plan. These were the targets for all kinds of things, not PM 2.5 [INAUDIBLE] but energy intensity reduction, carbon intensity, sulfur dioxide,
NOX chemical, COD, these are water pollutants. And four is courage. And these are the achievements at the end of the five year plan. That is, in 2015, they actually achieve all these targets. These are the targets for 35 year plan. So this is for 2016 to 2020. These are the goals.
So again, there are quite ambitious goals in terms of energy, intensity of carbon, intensity of all the other things, as well as PM 2.5 reduction. All right. We all know in 2014 China, US announced-- when President Obama visit Beijing-- [INAUDIBLE] US-China agreement on climate change. They said, bilaterally, we are going to actually agree on these things as part of the Paris agreement that was later on made in 2015.
What US pledged to do is the US aims to reduce emissions by 26% to 28% below its 2005 level in 2012 or 2025. China's goal was to peak its carbon emission by 2030. So that is, you should allow us to increase carbon emissions over time because we are still developing. But we promise, by 2030, our carbon emission will be peaked.
We're not gonna increase beyond 2030. We are gonna do a lot of things to achieve that, including increase fossil fuels and energy consumption by 20% by 2030. We all know that our President-elect wanted to abandon the Paris agreement. But Chinese government said we are going to fulfill and honor his commitment to the agreement. So this is very welcome right. news.
In fact, actually, China has been doing carbon reduction programs, such as cap and trade programs in seven regions, starting from 2013. So in seven locations, including Beijing, Guangzhou, and Chongqing, and Hubei, et cetera, they set up a cap and trade programs for carbon emissions.
The goal was actually to have a national program starting 2017. So these are pilot programs. The idea is that lead garment agencies, lead firms, understand how this program works. And then, in 2017, we are going to have a national program. This program is going to be larger, will be larger, than the current EU emission trading program.
And that will be a significant coal benefit in reducing local air pollution. Although, the goal for this program is to peak carbon emission. But as we know, when you reduce carbon emission, for example, from using natural gas rather than coal, actually, the majority of the benefit will be local pollutant, It will be a reduction in PM 2.5 and associated health benefits from the reduction of local pollutants rather than actually climate change benefit itself.
There are many studies. For Example, look at US SO2 program where they show the majority of the benefit actually do not come from reduction of SO2, but the coal benefit reduction in reducing PM 2.5. OK, so I believe that in the process of cleaning up China's own air, China is actually uniquely positioned really to be a leader in climate change.
And more importantly, in actually producing clean energy for China, for the US, for the world as a whole. Which China has been doing. But right now, I think China is in an even better position to do that moving forward. OK, so my overall thought. So this is my overall thought.
So Chairman Mao said in 1945 that the future is bright, but the roads are twists and turns. So there are a lot of things we need to do. But in the end, I have confidence that Chinese government will be able to actually address the environmental issues. Although, not in the immediate short term, but in a 20, 30 years period, or even shorter.
[APPLAUSE]
Shanjun Li, Professor in the Dyson School of Applied Economics & Management at Cornell University, presents an overview of environmental challenges in China and suggests an approach to dealing with traffic congestion in Beijing. Recorded November 21, 2016 as part of East Asia Program’s Cornell Contemporary China Initiative.
