 Donc, dans cette conférence avec Mike Witt-Claude, Yersine et Stéphane, ce que nous essayons de faire cette année, et aussi avec les conférences précédentes, c'est de mettre ensemble deux communités distinctes des économistes, ceux qui travaillent sur les marchés de l'économie et de l'énergie, qui devraient aller plus dans les conférences de l'économie et de l'énergie, et ceux qui sont plus intéressés dans le design des policiers environnementaux et des ressources économiques, comme quelqu'un comme moi. Et ce n'est pas facile, parce que, comme vous le voyez, vous pouvez voir ce qu'ils se diviseront dans les sessions. Mais la bonne nouvelle, c'est que nous avons quelqu'un qui peut faire un merge entre les deux communités, c'est Ron Pablo. So Ron Pablo has made contribution in IO, energy markets, in environmental economy, the choice of environmental policies, contract theory. So he is really much a synthesis of the two type of community approach of the tools, he has graduated from MIT and then he has been a professor at the Pontifical University of Chile in Santiago for many years. And what is impressing is when you see his main contribution to economics that managed to publish in general interest journals and also top feed journals in both in IO, energy economics and environmental economics. And also he is also involved in policy advices. So this is a perfect match for this kind of conferences. So he is going to talk about traffic regulation, which is a new topic I think, one of new topic of research. So now Ron Pablo is yours, thank you. Thank you Stefan, thank you. Well this is the title. First of all, well thank you for the invitation and for the opportunity to give a keynote talk. It's an honor really. I mean the previous keynote speakers really. So thank you very much and also congratulations for the workshop, I think it's great. I mean to you, to Claude, to Stefan and to Missing Agnesine as well. So really, very good. So yeah, thank you. Well I've been working on this for 10 years now. And I think the reason I'm working on this is because I was born and lived in one of these cities. Probably I wouldn't be doing that if I were born in, I don't know, in Little Rock, Arkansas perhaps, or I don't know, any other small city. And one of this is my home city and where you see, we suffer problems of air pollution during winter and the fall and congestion throughout the entire year. Set February, which is the time of the year in which people take vacation. Otherwise, you know, a congestions is getting even worse and there are other cities there, Mexico cities over there, while the US also have a picture there too. So it's happening in many cities around the world, not only in my own city. And what really worries me is this picture here. I don't know if you can see that very well. Which is the rate of motorisation in emerging and developing economies, you know, is growing very fast and they still have a way, you know, large way to go, much higher. In the US it's pretty much stable if you want, with, I don't know, 800 cars per 1,000 people. In Europe, I think it's 581 per 1,000. The same thing in Japan and South Korea. But in the rest of the world, in the developing and emerging, you know, in emerging economies, there's still way below those numbers. And when you look at how much, you know, cars are being sold every single day, it's boom, it's really worrisome. I mean, it's a problem. So the question is, what do you do about that? I mean, these cars create, you know, a lot of externalities. Of course pollution is one of the most important ones, but also traffic congestion and other ones. And I think a good way to start is going back to this paper. It's a nice, very nice paper, a survey paper by Ian Parry and co-authors. They came out in the Journal of Economic Literature more than 10 years ago, talking about these issues. You know, what are the policy instrument that, you know, in different parts of the world, people are using to correct for these externalities. And the externalities is the one you know. I mean, local air pollution, traffic congestion, and also global air pollution, and the traffic accidents, all dependency, particularly for some countries, for some other countries maybe, you know, that much. And other externalities like noise and urban sprawl, which I'm not going to talk much about. I'm mostly going to focus on the first two, which are local problems. So, but indirectly, when you correct these two externalities, local air pollution and traffic congestion, you know, as a byproduct, you also reduce in CO2, automatically. But I'm not going to show you numbers on that, but it's something that is true. So, in the same paper, you know, Parry and co-authors, they say, well, these are the policy instruments that people are using. You know, fuel taxes, it's one. And also, for new cars, you have this fuel economy standard for new vehicles, here in the CAFE standard in the US. And also mission standards for new vehicles, you know, how much they can emit per mile. And Euro 6, for example, I think now in Europe, we're in Euro 6. And then, that's pretty much it. These are the instruments that was mentioned in that paper in 2007. Plus, you know, they explain, well, also we have this policy that is emerging, congestion tolls. By then, we have only, you know, these tolls in four places. In some cities in Norway, Singapore, of course, the first one, and then London, in 2003, which was like a big event. You know, first time, you know, a big city implements this. And not much since then. I mean, Stockholm, Milan, Gothenburg, and New York. This is big news. New York seems to, you know, they already approved to have a congestion toll to enter, you know, the lower, here is London. And you already, you know, agreed to implement a congestion toll. This is like a daily pass. If you enter New York, lower Manhattan, you had to pay this daily pass. And that's supposed to be to start in 2020, not before that. But I mean, the news is that it's going to go through, but it's not 100% sure yet. I mean, there's still maybe some legal issues or, but apparently it's going to go, it's going to go through. But these are only the instruments, you know, considering this paper. There's a lot more going on. This is what I want to talk to you about. A lot more than these instruments. What are these instruments being used elsewhere than in these places? I mean, fuel taxes, well, we're in France. We know what happens when you say, look, we want to increase, you know, fuel taxes to correct for some of these externalities. So the question is, if these instruments are not available, congestion price, you know, what else have, you know, policy makers been using in the, in the world. And we see a lot of things going on. And I want to focus on some of them. In particular, for example, in my home city and in other places. For example, we have this high-currency vehicle lanes in which you allow some people to go in one lane, if they are two or more people, you know, in the car. And if you are only one passenger, then you need to use the other lanes. There's a lot of discussion whether this is efficient or not because you are moving traffic to the other lanes and you may end up with more congestion or even more pollution. There's some, you know, paper showing showing that. In Jakarta, there was a program and you need to have three people in the car and they decided, you know, last year to remove that or two years ago to remove that program, for example. So there's a lot of, you know, controversy about, you know, the efficiency of this instrument. Then we have, and I think in this conference we've seen some people presenting this work on taxes for new cars, just for new cars. Carbon taxes. For example, in France this is a famous, you know, program, malus and bonus in which you have a subsidy or tax depending on the type of the car. If the car has a very low fuel efficiency then you have to pay a tax. This is a high fuel efficiency then you get a subsidy and there's some papers looking at that. Is that it only affects the new cars. So we mean the new cars. And so it's only indirectly can affect the existing fleet and it may take time for this to happen to really have an effect on the existing fleet. And in some cases these subsidies they suffer from, you know, from severe adverse election problems. You're giving subsidy to people that would have bought these cars otherwise. For example, this is a nice NBR working paper I don't know exactly that in the U.S. that you are paying a lot of these subsidies for people that would have bought these cars anyway. And we also see these are in some way better because they are aiming at these very old cars, these scrapping subsidies or these cash for clunkier programs. And we see them in the U.S. also in Europe, in Spain, in France, but the problem with them is that typically they are not, you know, designed as an environmental policy instrument. They are mostly designed to somehow to help the local industry during our recession, for example. So they last only for a short period of time only a few months, three, four months. I think this is how much they last. So, but the good thing is that you are really aiming at the very old cars polluting ones. So that's a good part. The bad part, as I said, is just they don't last too long. So you cannot rely on this as an environmental policy because they are very expensive. I mean, these are, you know, subsidies, they are very, very expensive. You don't see this policy anywhere else than in this few, you know, industrialized countries. I mean, I don't know anybody in the developing world using this kind of cash for clunkier programs. How about pollution charges? And there are some but only two experiences with this. This is a very good instrument. Actually, this is the best instrument that you can think of and local pollution, which is what is being done in London and also in Milan, which is basically had to pay, you know, a congestion toll to enter, you know, some area downtown, say, or even, you know, a larger area. And on top of that, depending on the kind of vehicle that you are driving, whether it's an old or a new, you need to pay like a pollution charge. So this is a very good policy because it's attacking both problems. The problem is that I know, as far as I know, in which you use this. So it's, again, because these are taxes and it's very hard to, you know, to implement. I think the political economy of the kind of instrument that we see around, I think it's very important to understand. And how about smog tax? Basically, you need to, you know, bring your car to an inspection station every year, in some places maybe twice a year to check whether the car, the emissions rate of the car is according to some norm, depending on the age of the car, of course. You know the car ages, you know the norm becomes more relaxed, but you want to make sure that you don't have this, you know, cars emmitting above that, above the norm on the street. What happens? There's some problem. For example, they suffer from some correction problem, some corruption problems in which you just pay, some bribe and then you get your car or you get the certificate. So, it is a good instrument to avoid, for example, those old polluting cars to, you know, to go around. And there's some evidence showing that that's not clear, they're doing that job. And and finally, and this is going to be most on the rest, I mean a large part of the rest of my talk, are these rationing schemes that we see more and more often everywhere in the world. Not only in the developing world, in Athens, I understand was the first, I mean, long time ago in 82 was the first place using this for only for a short period of time. Then in my home city in Santiago, then Mexico City, and now we see them in many, many places. These are rationing schemes of different types. I think sometimes people call them license plate bands in which, you know, depending on the last digit of your license plate, you may not be allowed to use your car, for example, on Monday or maybe twice a week, depending on the programs, sometimes depending on whether it's only in peak hours or during the entire day, so there's different variations of that form. And one extreme form of these rationing schemes is these low emission zones, which are very common now in some cities here in Europe, in Germany, for example, in particular, in which, you know, some old cars just cannot go into the city center, for example. And typically these rationing schemes, the reason why we have them in the first place, if you read, for example, why, what was the motivation for introducing them, was air pollution. That was the reason. In Mexico City was air pollution, they were introduced in 89, where congestion wasn't a big problem yet. In my home city was in 86, congestion wasn't an issue then. I mean, the motorization rate was very low, it was pollution. That was the rational. And I think in Europe where low emission zones, for example, in Germany, it's again, it's pollution. It's not congest. But in some other places, for example, Bogota, in other places, now they're thinking about this rationing scheme for also dealing with congestion. At the end, the two problems are together. But I think it's important to understand the difference between the two, because the kind of, the way you design this rationing scheme, it may change. So it's important to what are you using these schemes for? It's important to keep that in mind. How popular are these restrictions? Increasingly popular. Increasingly popular. Here you have some license plates where we see them in many different places. I could just put a few of them. In Kiel, in Ecuador, there, in São Paulo, in the Philippines, in Manila, actually also La Paz, in Bolivia. So you see them in many, many different places. And so here you have a list. And when you see a star, it means that they were introduced primarily for air pollution reasons. And also here include the low emission zones. Because for me, these are rationing schemes that take a very particular form and also include here Paris, for example, they introduced a very particular restriction which was, I think, two years ago that they restricted all cars 97 and older from entering, you know, the city of Paris. Except during the weekends. But during the weekdays, they cannot enter. So that is also a rationing scheme. These are rationing schemes, there are no prices involved and it's just, so to me, these are all, I call them all driving restrictions of different formats, but these are driving restrictions. Here you have, for example, Madrid. This Madrid central, again, another form of driving restrictions in which you are not allowed to enter the downtown Madrid. And also it depends on the type of cars that you're driving. Some cars are exempted. For example, electric vehicles, all the cars, you know, face larger restrictions. Unfortunately, as we learned, you know, two weeks ago, I think the new local government decided to remove this policy which showed some important reductions of local pollutants. Carbon monoxide and NOx and all that. So I think this is one of the few cases in which a policy of driving restrictions have been removed. Typically, this policy, I like, for example, scappage subsidies that remain only for a few months. In most of those places that I show you, these are policies that remain, you know, for a long time. Here, for example, we have another restriction rationing scheme in Germany in which you can now use diesel cars in Hamburg. And also, another cities in Europe are also implementing the same policy. These are restrictions, again, you're not, but what is interesting is that these restrictions have an important distinction from, for example, the one, the first restrictions that we would so implement it. And when you talk to anybody about driving restrictions, immediately they come to their mind, okay, the policy in Mexico City in 1989 in which you basically impose a restrictions uniform across all cars regardless of emissions rates, regardless of age, so it was a uniform restrictions in basically that once a week you could not use your car. So that's, if you want driving restrictions 1.0. But now we see a very different restrictions which I call vintage-specific restrictions, like the one in Madrid, in Germany, in other places. Which depends on the type of car. If you have a very clean car, you're exempt completely from these restrictions. If your car is old, then you have to be restricted. And I think it's very important to make, you know, the extension between these types, and I'm going to try to explain why is the case. And so the question, I mean to understand why or when these Russian schemes work let me, it's important to differentiate between these two types. Uniform restrictions applied to all cars, regardless of age and these are the ones like in Mexico in 1989 and also in Santiago in 1986. And the answer to the question do they work? The answer is no. They don't work. They work really bad. It's a really bad policy. That's why people tend to think about Russian schemes as really bad policies. When I present this work before, people say ah, this doesn't work. People are going to buy a second car to wait for the restriction and all that. That's true. I mean, you may get some reduction in the show run. I'm going to show you some work on that I did with some co-authors from Mexico City. You may get some reduction but only the first month you get some reduction from these programs. But soon people start buying a second car typically a very old car and highly polluted to bypass the restriction. So now you have two cars in your house the one you have before and a new car and then you end up with more pollution and more congestion. How fast? Very fast. Even before one year you go back to the previous level. So it's a really bad policy. But this is the news that I've been the reason why they're so bad I mean they're already so bad for those two reasons or for this one in particular. But there's another reason. Typical people don't think about it. Is that you have an effect also in how the fleet is going to evolve over time. The new steady state if you want. By taxing all cars uniformly basically what you're doing you are delaying the entry of new cars with a much cleaner. And you end up with an older fleet and you extend the life of the older cars and you end up much worse off. That's the problem. I mean this is a problem it is. But it's not the main problem. The main problem is when you look at the long, very long run is that the fleet that you end up with is really bad. It's a bad composition with too many old cars and with a very delayed entry of new cars. That's the main channel which is operating. All the papers before are just looking at this one month effect or maybe one year effect. The one year effect already show you that this is bad but this is even much worse. So bad idea to have this rational schemes in which are they uniform across all cars. Now this is a work we did in Mexico City with some co-authors, showing why they are so bad and what we use for this is carbon monoxide emissions from all these monitoring stations that are installed in the city. Why carbon monoxide is a very good proxy for car use especially during peak hours. Here you have for example concentrations from these monitoring stations, these dots and here you have the actual emissions. During peak hours here, 7 o'clock in the morning 8 o'clock in the morning, you know they are perfectly correlated. During the rest of the day you don't know why because there is wind and all that but for example carbon monoxide during peak hours is a perfect proxy and that's what we use to test for example for this Mexican for this Mexican policy using these changes in concentrations before and after the policy. And what we found, this is what we found in the short run as I said before 11% decrease. This was a one day a week restriction so you don't get a 20% reduction because there is some substitution but you get 11% reduction. But what happened in the long run? By long run here I mean during the first year well you have more pollution 13% more. Why? Because in part the second car effect was kicking here. I cannot show you anything else because I cannot go beyond with this data I cannot go beyond one year to tell you these very long run effects on the fleet composition that I mentioned before. So the policy was already bad and this is a good exercise because on Sunday there was no restriction so you can look at what happened during Sundays and there was very little change in the short run but in the long run because you bought another car you're also using that on Sunday and then you end up with more pollution not only during the week days in which you have restrictions but also other days of the week. There's also some distributional impacts here for example who are the people that bought the second car in Mexico as many other cities et bien, income is very segregated around the city. So you have for example here low income areas, medium income areas and here you have high income areas where you have the monitoring stations so you can look by looking at each particular monitoring station the impact of this policy and who were the ones that bought the second car this is the famous hypothesis and we found that there were people in the middle income area. Why? Because people with high income already had to handle their restrictions quite well and people you know in the low income areas they didn't have the money to afford a second car. So all the actions was here this was the guy that were and here you have for example summary with this table but I'm going to skip that. So now the question is ok this ration schemes for example in which you impose a uniform restrictions upon all cars are a bad idea, Mexico City is the best example and we are not the only paper showing that, there are other papers but now the question is can they ever work? and this is an important question it's a very long question because we see them more and more around the different places so can deliver anything and the answer is when you allow for this vintage differentiation the answer is yes they can deliver quite a lot actually and this is for example one the first paper looking at this looking at the low emission zones this is a paper by Wolff and then we have a paper just forthcoming with a couple of authors showing why this vintage restrictions can be a very good idea and there are many things here that is important to notice the first one is that now you have there are two ways in which you can bypass a restriction one is by buying a very old car the other one is by buying a cleaner car we have no restriction and what we are serving this program in the program in Santiago in particular is that that's the way people tend to bypass a restriction they anticipate for example the purchase of a cleaner car so they can be accept from the restriction this is exactly what you want you want to accelerate the fleets and over towards cleaner cars with the restriction and this is what we what we want in the data so I think the main message here is that these restrictions when you use them you need to use them to affect the extensive margin not intensive margin you will of course affect the intensive margin because you are imposing a restriction in the usage of the car but mainly you want them to really affect what cars you drive and if you pay attention to that mechanism it's where you can get a lot from this policies so the key to understand when this restriction can work reasonably well is whether correcting for the extensive margin the type of car you drive is very important more important for example than correcting how much you drive and for local pollution maybe not for global pollution local pollution, that is very important why? because most of the local pollution come from old cars 80% come from very old cars and I want you the way to convince you about that is to show you this picture we are not the first one to say this already for example there are papers by Cannes and Neeler and Sandler showing exactly the same thing is the old cars the ones that are contributing mostly with pollution by local pollution I mean carbon monoxide NOx and hydrocarbons and particulates and all that and here you have for example data from Santiago actually from the entire country from Chile showing the clean cars for example there in this vintage could be, this is carbon monoxide could be I don't know 10 times cleaner than cars that are I don't know 15, 20 years older and that's exactly the problem so here the extensive margin is very important because you want to move people towards these cars and here there are two things why? immediately I think you have the question why these guys are much cleaner than these new cars and the old ones there are two reasons if you draw the same picture for fuel economy you wouldn't see much of a change the fuel economy of a car remains pretty much constant throughout the entire life of a car I just learned that I mean there's little deterioration during the lifetime of a car but for local pollutants it's completely different I mean not only because these cars enter without better technologies but also because throughout the lifetime of a car things start getting worse and worse the catalytic converter doesn't work as well as the beginning so you end up with more pollution and the problem is that in this developing and emerging economies the fleet tend to be very old I mean the average if you take a car randomly for example in Mexico City the average life the average age of that car is 14 years old and it's entirely pretty much the same in the US it's 9 years or even less so we have these very old cars you know, running around the city and polluting, you know, a lot so the objective with this policy is to move people faster than otherwise to these cleaner cars and here you have for example evidence from this vintage restriction the first one I think as far as I know that was implemented in Santiago in which the authority decided in 1993 to change the existing restriction program by exempting all cars in 1993 and newer from the restriction why? because those cars were were coming with a converter a catalytic converter cars before that didn't have it so they were much cleaner and the idea was to accelerate that process you know by extending exemption to all those cars so what happened if you look at what happened for example 13 years after the policy in 2006 here you have on the fleet composition in Santiago, you know, the city with the restriction and in the rest of the country so what happened here here the dark bars are old cars, 92 and older the light bars are 93 and newer so you see big difference between the composition of the fleet between the two areas the ones with the restrictions and the ones without the restriction what's happening? what policy is doing its work it's basically creating incentive for people here to buy this cleaner car faster than otherwise and they're exporting a large fraction of these cars to these other areas of the country in which pollution, local pollution of this is not a problem so this is the way the policy is operating and for example here how much people value this restriction here we have some data for example showing what is the price difference between a car with and without the restriction using for example for different models small models like the Fiat Uno small car Toyota Corolla here and we see for example people are ready to pay a premium of something between I don't know, depending on the specification between 10% and 4% people were ready to pay that to get rid of the restriction pay this premium and they did and so now we want to start answering more interesting questions what is the optimal design if you're going to use these restrictions to really accelerate the fleet turnover what is the optimal design for that you need to understand how the market the entire market, how the interaction between the new cars and the existing fleet interact and what happened for example over time and for that you need to develop a dynamic model of the car market that after a shock you want to see how it evolves and what is the new steady state and so what we did we developed this vertical differentiation dynamic model let me give you a little bit of a flavor what it is so there is a continuum of consumers with different willingness to pay for a new car which is cleaner and it's a higher quality in our model we are only one dimension and here are the new cars so this is the consumers this is the willingness to pay for quality or for a new car and so these guys here they are high willingness to pay they are the ones using these new cars then these guys are using cars with 1 year old, 2 year old and so on and at some point you scrap the car and then the rest of the people use public transport so this is the flavor of the model and one shop coming is that we have only one type of car in our model like a representative car that ages over time and start to be traded for people with high willingness to pay for quality to people with lower willingness to pay for quality I want to understand the dynamic of that and how the fleet evolves with that and here you have a let me show you some results of using that model and here is the fleet composition in Santiago in the place in which you have you eventually going to have the policy being implemented and the rest of the country and this applies to any place, you can for example do it in Paris I mean here you have Paris for example in which you are going to place a restriction and then you want to look what happens in the rest of the area in which the secondary market for the car market is operating you can have a restriction in this area but this is going to affect prices because the second hand market is integrated so here you have the composition of the fleet for example in as I said in the potentially restricted area where you are going to impose the policy and here in the rest of the country and the vintages go from up to 24 years old and so you see for example without any intervention without nothing cars in Santiago I mean there is a larger proportion of 2 cars than in the rest of the country and the reason for that is because of income higher income, you need to control for that but these cars here are polluting quite a bit and you are only imposing a restriction here in this region of course you are going to have an impact over here because some of these cars now maybe they are going to be exported over here where they have more value for example we know that for cars for local pollutants and for CO2 we have piguvian tax but for local pollutants there is no piguvian tax we cannot monitor emissions very well although for example the pollution charges in London and in Milan what they do it's like a proxy look this is the type of car you are driving I assume you are emitting this much piguvian damage but if you could implement a piguvian tax in principle this is what the fleet would look like so you get rid of all these very old polluting cars in Santiago and all those guys are here in the rest of the country and because of secondary market operating now, I mentioned before the introducing and uniform now with this model we can think about different policy interventions anything you want capital subsidies for all cars in which you are going to affect the timing in which you scrap the car this capital T is the timing in which you scrap the car you can think about gasoline taxes you can think of for example where we call circulation fees every time you register your car whether it's a new car or it's an existing car every year you need to register your car get a permit to use your car then maybe you can charge a tax we also have that here you can include any instrument that you can think of and start comparing all these different instruments and how much you get from instrument that you already see in practice with this perhaps better instrument or sometimes even worse so the model allows you to do all that and more importantly allows you to understand what is the dynamic of the entire market how the entire fleet evolves with this instrument this is exactly what we do in the paper for example what happens if you introduce a policy like the one in Mexico City uniform restrictions one day a week and this is what I told you before this policy is really bad because look what happened here these are the new cars in the restricted area you really reduce the value of a car of a new car because you're restricting that car 20% of the time so now there's less incentives to buy those new cars and that has an impact for the entire fleet you're really making it older when you're running under the first best, you want to get rid of those guys but you cannot do it with these uniform restrictions so that's why it's so bad and here in these models we're not even considering the second car that you're purchasing a second car, not even that that's not here in this model because every household here has only one car we simplify the model a little bit even without effect this is much more important now what happens for example if you introduce a vintage specific restriction like the one in Paris for example that at a certain age you cannot use your car, period that's it, nothing else and this, think about what is a Russian scheme it's a proportional rationing basically what you're saying is you're saying look, I'm basically restricting on average the use of this car so given that given the form of these policies do you want a car to be used all the time or not to be used at all that is going to depend on the average value of that car if the average value of that car is going to be driven by one particular individual it's about the social value you want that or provide social benefit social net benefits you want that car to be driven always no restriction whatsoever if that is negative you know what car to be driven whatsoever with this proportional rationing rules this is what our driving restriction is it's a proportional rationing rule, that's it you either want the car to be driven 100% of the time or nothing and because of that you have this, the optimal restriction is a threshold from here on you want cars to be completely restricted and from here on no restriction whatsoever so in our model for example you want cars 16 years old to be completely restricted and above that they are not facing no restrictions on the optimal policy it's like the one in Paris 79, 1997 you cannot just go into the city center that of course is going to change for some cars maybe year 2000, for other cars maybe year for diesel could be one year for petrol could be another different year of course in a more complete model but the logic is that either 100% restrictions or no restrictions at all because of the form of the this proportional rationing rule it's not that you are restricting the less valuable trips a driving restriction is a proportional rationing rule it's not an efficient rationing rule if you want that you are destroying the less valuable trips that's what attacks us but not a driving restriction that's an important difference so here you have the optimal driving restrictions as I said is banning all those cars and and what you get you are not in the first best but you are close you got rid of all these guys these very dirty polluting cars and those guys at the beginning this is a very nice mechanism that we didn't anticipate at the beginning those cars were exported to the rest of the country of course but eventually those cars start dropping the price of these very old cars and they were scrapped sooner than otherwise so you end up also with a cleaner fleet in the rest of the country as well at the beginning no but in the steady state at the very end is because you are exporting all these old cars over here and you are reducing the price of these guys here as well and so there is no leakage in the long run there is no leakage so here you have a table summarizing for example when you compare this program to other interventions here is how bad a uniform restriction is the one I mentioned before look if you get 100% this is the first best gain so this is a 92% reduction of welfare really bad now if you use a driving restrictions with some exemptions you get 16% of the first best gain the optimal driving restriction give you 51% in our model not bad now whether you can implement that or not perhaps you are going to be halfway between this and that now here you have the scrappage subsidies the scrappage subsidies have a problem first sometimes for example you want to have these restrictions only during winter and fall on a scrappage subsidy you are also aiming at very old cars because you are removing these cars but you remove them permanently in this exercise I am not including that but in other parts of the paper we do that so sometimes you only want to restrict circulation during maybe half of the year a scrappage subsidy either you remove the car completely or not at all so that's why these policies work even better and the subsidies are expensive to implement now circulation fees this is the best policy that you can implement basically every year you go and register your car where it's new or old you have to pay what is expected is the expected Pigoubian payment the expected Pigoubian bill if you want and for all different cars and with that you get much closer than the first best you are not affecting usage in the margin but here because the most important driver of local pollution is the extensive margin with this instrument you can get very close to the first best gasoline tax is really bad 16% why? because in the long run by imposing a tax on all cars again you are somehow delaying the entry of clean cars new cars because you are taxing everyone with this gasoline you are not separating or distinguishing between all cars and new cars everyone has to pay the same tax now if you combine for example a gasoline tax with acts on the intensive margin with a driving restriction then you can get a 61% so here you have an example the importance of the extensive or the intensive margin and so this we had to do in the last revision of the paper we did this exercise because our referee asked about this and he was absolutely right about this it was a very good comment explain me how the restriction works in the very short run first year for example when the fleet is constant compared to a gasoline tax and then what happens in the very long run and actually he was absolutely on the money they work very differently because one act on the intensive margin which is a gasoline tax and the restrictions on the extensive margin and for local pollution as I hope I convince you already the extensive margin is important you keep for example the fleet constant this is exactly the same exercise done by needle and sandler in a recent paper we get with the gasoline tax which is optimal at $2 per gallon we get 50% of the first best gain in the short run keeping the fleet absolutely constant they get 35% in their paper it's a little bit different because I think our pollution damage figures are a little bit higher and the restriction gives you less than 30% so clearly in the short run when only you are trying to to adjust the intensive margin the gasoline tax does better but what happened in the long run for example in the new state state well the gasoline tax does really bad only 16% because you are affecting also these new cars that are coming to the market you are delaying the entry of those guys and the restrictions does much much much better 51% so I think this is the main message trying to understand okay I mean at the intensive margin or the extensive margin and for local pollution again what matters is the extensive margin so congestion pricing is it feasible I mean we've seen it in some places but clearly we don't see it in much where else I mean nowhere in the emerging or developing world there's nowhere, there's some discussion even in my home city there's some discussion but I think it's going to take a long time to see any form of congestion pricing in the near term now what about the same instrument that we've been talking about so far these rational schemes using it for both for congestion and for pollution which you know it's what we observe we have the two problems together it's hard to separate the two so now you need to deal with two types of exemptions right if you think about congestion you want to impose restrictions upon all cars but if you do that people are going to start buying a second car so what would be an vintage so for pollution we have a vintage exception if you have an exception you have a new car, clean car you face no restriction what would be the equivalent for congestion there is nothing because all cars eagles whatever they should be equally restricted because they all congest the same there's one thing you can do well if you want to get exempted you have to pay a congestion toll so this is what we're proposing so the way you combine these two problems with a restriction is that you allow for these exemptions of two types a toll exemption dealing with a congestion and a vintage exception dealing with a pollution because for example in many places pollution is only a problem during some month of the years the vintage exemptions only act during that time of the year in which you have pollution say winter and fall the other one the toll exemption act every time if you don't do that then you go back to the uniform restrictions which are really really bad idea and it's better not to have any policy at all these driving restrictions four congestions can only work if you allow for these toll exemptions and I think to me you can think about it what happen for example if you strict a car every day every single day of the week but you allow the driver of that car to pay at all then you are in London or in New York perhaps it's very hard to start with that but you may start with one day restriction and if in that day you want to use your car you pay the toll and with that you kill any incentive to buy a second car because it's much cheaper to just pay the toll that day so this is what we are proposing I am talking to policy makers in my city about this and they are more inclined to think about something like this that is less radical than a full congestion pricing for every single day and so you can start with this and eventually moving up and so here I have some numbers we have developed this model based on a previous paper by by Basso and here we really pay attention to different income groups because when you start thinking about this congestion pricing scheme distributional issues are crucial if you don't pay attention to them then you can really screw up the whole thing and so we divide the population in five income groups and also we pay close attention to the different cars that these people own because some of them are going to be respected during winter and fall and it's very ingenious the high income groups they tend to own very clean cars and newer cars and groups one and two and three they tend to own these older cars so you pay attention to all that I need to bring one equation this is the only one that I have and so this is a little bit of a calibration of the model I show you that low income people group one are very price elastic if you introduce at all these people are going to leave the car at home right away but the very high income people actually if you introduce at all they're going to use the cars more why? because now they can't go faster so this is what our model is capturing so I think it's so here we have some results we aggregate overall welfare depending on the past the condition told that you use this is the optimal if you have for example 5 days restriction this would be like London or New York every time you use a car you have to pay the toll this would be with 1 day restriction of course you get a lot less surplus with 1 day restriction basically people only once a day a week so it's not much so we're proposing to have a 2 days restriction otherwise you're not really alleviating congestion much so if you can go all the way up here you have a big amount like 0.5 for GDP this is our estimate now this is the important thing that I want to show here distributional implication the poor do really bad regardless of whether you have 1 day restriction or 5 day restriction they really suffer from this policy so on the rich the high income group the more restriction the more the benefit because the faster they can go they really like the policy they love it so you really need to do something about this otherwise this would be very progressive and the way to do it is to take all the toll revenues and put it back into the public transportation system and there are 2 ways you can do it one is producing the ticket the fare the other one is just a combination of the 2 improving quality improving frequency here I'm going to show you just for example what do you do if you reduce all the money to reduce the fare and if you do that now low income people are very happy so you solve this distributional problem and the rich people are still very happy because they know they can drive much faster so distributional implications are crucial when you start thinking about this these schemes and finally smoke checks this is also doing some research on this and some may think that this is a way to prevent these very old cars and highly polluting cars to go around and it's not true it's not true because we have some evidence this is a very nice paper from Paulina Oliva from I think still in Santa Barbara looking at smoke checks in Mexico City we found a strong evidence of corruption basically you pay someone to just use a different car than yours in the station when you go through the station line they just replace your car with a different one and the way that she did it is very nice because she found a very strange correlation of data sometimes you have like 3 data points with exact the same emissions since this is something suspicious and what evidence of fraud so we are trying to answer a different question in this paper which is sometimes we've had too much competition between this certificate providers these smoke check stations that may be bad they may have incentives to relax the standards and that's what we're finding and so to study this problem so this is a case in which too much competition could be bad basically this is what we're doing in this paper trying to show that and could be bad because you provide less quality but not to the consumers consumers don't care, they want the certificate they provide less quality to all of us we end up with more pollution so it's not the typical question whether if you're allowed to competition then companies don't start providing lower quality products you as a consumer you're going to suffer that here is different me, I don't care I need a certificate just to go around but the problem is that we're going to end up with more pollution and some of our numbers show that you end up with with 10% more but I don't want to show you any numbers yet because we're still working on this and what we do for this competition hypothesis whether more competition is bad for the environment is we look at the exploit this entry, these are all the small check stations in Santiago and so what we do is we exploit for example here you have this station plan this alone in 2014 and one year later has competition from that guy at a much lower prices, 50% and what is interesting is that when this guy instead of reacting by lowering their prices they react by relaxing the standards they keep their prices high because maybe there are some non choppers that people that are new cars they know they're going to pass anyway so the competition is not by lowering the price which they can do but it's by relaxing the standard so now they tend to pass more cars than otherwise and here you have another example of an entry this station here and now there's a new operator because it was auction off the site was auction off there was competition for that site and the guy that wins the site offer a much lower price for the service, 50% less so again you have all these guys suffering from that and we see that these guys are reacting to that not by lowering the prices but by relaxing the standards so conclusions first the political economy of what you can do and what you cannot do is essential and that might change from country to country in some places perhaps in other countries you may I've seen some papers that maybe you are able to use gasoline taxes or all that type of taxes good, fine I mean we economists we like that I think this is but in many other places you cannot and so what I wanted to show you here is that these rational schemes are becoming increasingly popular in many places so the question is how you design them properly in order to get some benefit maybe not close to the first best but some benefit, some positive benefit out of them and subsidies are very expensive, I mean for electric vehicles whether for new cars or to remove all cars scrappage subsidies are very expensive so we hardly see them I mean in some places yes in Norway you know we've been talking about that subsidies to electric vehicles are huge but they want to change that and in the rest of the world we don't see them and I always tell the policy makers in my home country never use that if you can use these other policies to discourage people to own these old polluting cars it's better not to use those subsidies for that because first you have this huge other selection problem you wanna be paying subsidies for things that are gonna be done anyway so it's always better to work with taxes than with subsidies I think now what is key for this well for congestion of course but also for local pollution is to target the existing fleet not just the new cars I mean just by changing for example fuel efficiency standard or emission standards it may take too long now we have a large number of very old cars on the road and you need to take care of all those guys and they are gonna be there for a long long time given you know the motorization rates these guys are gonna stay there for a long time so you really need to target these old cars you need policies for that not to wait for the you know for the new cars to remove them and an important you know related to that it's important to understand the interaction between you know policies aimed at new cars and existing fleet how did you interact this is just one single market the new cars and you know the existing fleet is just one single market you cannot separate the two you cannot just do an analysis for the new market because most of the action especially when it comes to local pollution is in the existing fleet and whether for example you extend the life of these old cars or whether those guys move from one place to another whether it's more local pollution than not so this is what important to understand and I give you two very important examples on that gasoline taxes they may work very nice in the short run when you assume that the fleet is constant but they work really bad even for local pollution and I assume that the fleet is going to you know it's going to evolve and distributional considerations of course are crucial not only in this type of policies in any policy and I show you one example about you know congestion pricing of these congestion tolls combined with these driving restrictions and the way to solve this I think it's very simple keep all the money and give it back to people in the public transport New York is actually doing that London is doing that you need to earmark the money because there's some constitutional for example restrictions this is my case for example my country you cannot just take the money and give it back to the public transport you need to have special law it's very hard to circumvent that restriction but if you don't do that if you don't get this recycling in this way there is nothing you can do I mean it's very hard to pass this kind of legislation and of course there's a lot more research needed here there's a lot of policies you know around we need to understand how they work and sometimes how they don't work and also I encourage you to get more involved in the policy making process talking to policy makers about your findings and all that I think it's important sometimes they listen how much they listen my experience is that they do some not enough but at least you get some so with that thank you