Market-based approaches to environmental policy

[00:00:00] Speaker 1: In this lecture on topics in environmental economics, Dr. Sheila Olmsted presents a summary of the use of market-based approaches to dealing with environmental externalities. She first notes the difference between prescriptive regulations, which generally take the form of laws and policy standards, and market-based regulations, which can use price or quantity instruments like taxes and allocation permits to create new markets that capture the cost of externalities. She presents an analytical example of cap-and-trade approaches to highlight the difference between market-based policies and prescriptive policies. She notes the cost-effectiveness of market-based policies in many scenarios and also identifies cases in which prescriptive policies are more appropriate. She ends with several examples of market-based approaches to environmental policy and their relative costs and benefits. [00:00:50] Speaker 2: Just to kind of put this in context, so Jim, at the end of what he was talking about yesterday, talked about what are some solutions to externalities, these negative externalities that are pretty common in the environmental realm. I mean, you know, there are other market failures, public goods, commons, problems, and so on, but what are some solutions to the negative externalities? Well, he talked about the liability regime, and that's something that economists often just kind of forget, especially those of us who work on regulation, that that is just floating there in the background. It's an extraordinarily important part of the apparatus for correcting externalities. It may be even a greater incentive than, you know, anything we might be doing in terms of administrative regulation. But sometimes that's not going to be sufficient for a variety of reasons. The thing that I often put first, at least when I'm teaching, and that we talked a little bit about yesterday as well, was kind of private solutions, right? The mere existence of this external impact on a third party provides in some ways an incentive to start thinking about correcting it. And that can either work through kind of social norms, institution building, a la Eleanor Ostrom, as Jim mentioned yesterday, or through this kind of Coasean bargaining, private bargaining between the kind of externality causer and the externality receiver. And I won't talk too much about that except to say that on the private bargaining side, right, the kind of Coase's theory altogether had some pretty restrictive assumptions about when that would be likely to result in a sort of efficient outcome given the negative externality. And so in most kind of real life cases of pollution and other kinds of negative externalities related to the environment, those conditions are not going to be met. So we don't necessarily think that in lots of cases, this private bargaining can arrive at efficient outcomes, although sometimes it does, and we can certainly talk about some examples of that. There are some good fun ones to talk about maybe in the Q&A. So in that sense, right, if private bargaining fails, or sort of social norm, institutional approaches fail, if the liability regime isn't sufficient either because the kind of harmed parties are so diffused that they're sort of not able to kind of come together in some reasonably cost effective way to deal with, right, whoever's causing the externality, which is often the case when we think about diffuse pollution problems like air pollution problems or water pollution problems, then we're in the world of thinking about some kind of government intervention, usually some kind of administrative or legislative regulation. And in economic terms, some types of regulation are better than others, and so you've already kind of heard this from us, right, Jim talking about thinking about cost when we worry about where to put dollars for conservation, and Pete earlier mentioned individual tradable quotas for fishing, and some of you are already very familiar with those kinds of policy instruments, and those ways of thinking fit very much into this kind of broad structure that I'm going to talk about today, but I thought it would be useful to talk particularly about the case of pollution control and why it is that economists are so crazy about this kind of cap-and-trade and taxes or these kind of market-based approaches to environmental policy which have very strong parallels, right, to the instruments that are applied in resource policy, and then we will kind of get to thinking about how these things are more broadly applied right outside of the world of pollution control. Okay, so first I thought I would just mention the way that we typically as economists will categorize pollution control policy approaches, okay. One, we would create categories called either prescriptive or command and control or traditional regulations, right, so there's all different ways of referring to the same thing. And there we're usually talking about two different classes of approach. One would be a technology standard, right, so I say I've got a problem with pollution, all of you municipal sewage treatment facilities or all of you electric generating utilities have to adopt the same technology to deal with that pollution problem, right, in the electric utility case it might be you have to install a scrubber, right, a sort of flue gas desulphurization device to take sulfur dioxide out of your emissions, right, before they go up your stack or somewhere in the middle of your stack and thereby reducing damages to the public. On the water quality side, it might be a particular wastewater treatment facility piece of technology, but the key thing about that when we call these technology standards prescriptive or command and control regulations is that it's the same across all firms, right, or sort of constrained to a set of technologies or even a single technology requiring all firms to use a particular approach to compliance. A second class of these command and control regulations, and it's listed second because it's somewhat more flexible, right, so it's moving toward, you know, closer toward a market based approach is what we'd call a performance standard, okay, that would be something like I'm going to take electric utilities, I'm going to say everybody, you can figure out how to do this however you want, but everybody has to meet the same ceiling on their emissions or their emissions rate, right, their kind of emissions per unit of output, like the unit of electricity generated or something like that, and again, they're the key that makes it kind of a command and control or prescriptive approach is that it's uniform, right, I'm saying to every firm that, you know, everybody's going to sort of do the same thing, and there's some, a lot of good reasons why that's often a starting point for discussion of regulations, we, you know, if I've got to reduce 100 tons of emissions, well, a kind of 50-50 split if I've got two firms seems like a fair place to start, so it's not at all uncommon to see these uniform performance standards out there, so those two classes of regulations we'll call prescriptive or command and control approaches, and then we have the market based instruments, okay, and those are highly variable, so Pete had a much longer list on his slides that can include things like deposit refund systems and removal of environmentally harmful subsidies, but the two that we'll talk about mostly today are kind of price instruments like a tax and quantity instruments like a cap and trade system, right, where I'm sort of setting some overall cap on pollution from some set of firms and then allowing firms to decide amongst themselves who's going to do, who's going to do what kind of, how much abatement, right, so we'll talk about taxing negative externalities, the flip side of that is subsidizing positive externalities, so trying to subsidize good behavior, I won't talk very much about that, but I'm happy to talk about that in the Q&A, there are some important differences in the dynamic setting, for example, between taxes and subsidies, although in theory, in a static setting, they have a lot in common, and then again, the quantity instruments are, right, I'm going to set a cap, I'm going to allocate some permits, and I'm going to let firms trade, all right, there is also a class of approaches called information-based approaches, these are also market-based in nature in some ways, really the idea there is through some kind of a labeling program or an information disclosure program, providing information to consumers or about the sort of environmental characteristics of the good that they're paying for, like I've got a kind of green power option for my electric utility, or I can buy a shade-grown coffee, right, or the coffee with the little bird on it that says, right, that provides the way in which the coffee is grown, provides bird habitat, so there's lots of these kinds of labeling programs, they're very interesting, there's plenty of interesting research on what those do, at the end of the day, they're not going to change the kind of, they're not going to change enough the kind of fundamental incentives that consumers and producers face in the market, right, so they can't, you know, eliminate a negative externality, they can't fully provide a public good at the level of, that we would think efficiency would, would prescribe, and so from an economic perspective, we're often still going to have to do something else if we really want to get to the sort of efficient level with these labeling programs, so I won't talk much about them, but again, I'm happy to take questions in the, in the Q&A about those practices too. Okay, so first graph, so really what I'm doing here is I'm taking Jim's graph from the end of the discussion yesterday, where what he started out talking about was a cost curve for the production of candy bars, that's my blue curve, except now instead of candy bars, I'm going to talk about steel production, okay, so, so it's not the nougat this time, right, but something, air pollution, I might be, if I'm burning coal to produce steel, I'm generating some air pollution, and that's this kind of, kind of creates these third-party impacts, right, these additional social costs that I'm not taking into account when I think about what the private costs of steel production are, so the private costs of production are in the blue curve, the demand curve for steel, right, that downward sloping one is very similar to what Jim drew yesterday for candy bars, right, so that's, that's the demand curve, the green one, and then what I'm doing here, right, is Jim said, well, that, right, that externality shifts the private curve, private cost curve above the, the social cost curve above the private cost curve, and I'm doing the same thing, but I'm just also specifying that that comes from the fact that there are some marginal, we'll call marginal damages from pollution, as I produce more and more steel, I have these increasing marginal damages from a unit of steel production, right, and that could be because I'm producing more units of pollution every time I produce a unit of steel, but it also just could be that what I'm concerned about is health damages, and as I add more and more pollution, those health damages per unit actually get steeper and steeper, right, as it gets, you know, sort of dirtier and dirtier, the air gets dirtier and dirtier, so I've got these increasing marginal damages, and all I'm doing when I make this purple curve of marginal social costs is I'm adding the sort of brownish-orange one to the blue one, so the sum of this one and this one is this one here, the purple one, and so now I've got this divergence, and as Jim said yesterday, you know, if you sort of look at this, and I thought, well, the efficient level of production of candy bars, this time steel, would have been, you know, not the intersection I'd get from the market, the one that the private cost is sort of determining, but the one that the social cost is determining in conjunction with the social benefits or the demand curve, and so what I'd like to do is kind of move back, right, along that horizontal axis, closer towards the zero point, I'd like to have less production of steel than I do, for that matter, I'd like to have a higher price of steel than I do, right, than what the market is going to give me, and one way that we can think about dealing with that is simply what we would call kind of classic Pigouvian tax, right, so Arthur Pigou is an economist who in 1920 writes this very important piece that sort of says, you know, these externalities may be ubiquitous in society, and pollution, you know, is a good example, and one of the things you might think of doing is simply implementing a tax, make it more costly for firms to produce steel, and in doing so you drive them back, right, along their own cost curve toward a lower level of production, and well, what's the right tax? Pigou said, well, the right tax is the difference between social costs and private costs, which is exactly the height here, right, of this marginal damage function, at whatever you think the efficient level of production would be, right, at this Q*, and so I'm just showing T* would be the Pigouvian or the efficient tax, the height of the damage curve, which is equivalent to the distance between these other two curves, at this efficient level of production, right, so you impose this price on pollution, you internalize that externality, you get firms to do the right thing, right, acting in their own best interest in response to this price that you're setting on the damaging good. So this is a nice, sort of an interesting heuristic, it's a nice way of thinking about it, we do see pollution taxes out there, but when I draw it in this way and in the way that Pigou described back in 1920, one thing that you'll notice is I'm sort of saying the way to reduce pollution, right, the only way on this graph is to reduce production of the polluting good, right, I'm just trying to back everybody up, I'm trying to get less steel production, and that's certainly one way in which a firm might decide to respond to a regulation that increases its cost of production, might produce less steel, right, but a firm might also decide if I did something, you know, very flexible like a tax or tradable permit system, a firm might say, you know, if what you care about is sulfur dioxide pollution, then I might switch from burning high sulfur coal to burning low sulfur coal, or from burning coal altogether to burning gas, natural gas to make my steel, and in doing so I could, you know, dramatically reduce emissions and maybe not have to reduce my output, right, so I could switch my inputs, okay. I could also choose to install a pollution control device like a scrubber, right, I could make my steel using my old dirty coal, but I could scrub my emissions before they go out the stack, and in doing so maybe I'll have to reduce my quantity produced, right, if that is very costly for me, but maybe I won't, right, so this model where we think about the right way of treating the pollution externality as reducing production is not necessarily comprehensive enough for us, okay, so we're going to switch to a different way of thinking about the problem, and that way of thinking about the problem is to think about asking firms, instead of to reduce their production, we're asking them to produce now two products, I want you to produce steel, and I want you to produce pollution abatement, right, I want you to invest some money, right, whether it's switching your inputs, installing pollution control technology, right, reducing your output, whatever that happens to be, I want you to invest, right, so sort of experience some opportunity cost that sort of makes this the level of pollution lower than what it is today, and so the way that we're going to think about that from the regulator's perspective is that we, what we really want is we want to think about a sort of demand curve and a supply curve or a cost curve for pollution abatement, okay, so now I've got the quantity of abatement or pollution control that I'm engaging in on the horizontal axis, and the cost of that per unit per ton is a good measure on the vertical axis, and if I'm the regulator, right, I might do things like use some of the methods that Pete talked about yesterday to think about what this demand curve looks like, right, as abatement increases my kind of willingness to pay for an additional unit should decrease, way to think about that heuristically is that as things get cleaner and cleaner and cleaner, right, what's an additional unit of control worth to you, it should be less worth less on the margin, right, and conversely, if things are getting dirtier and dirtier and dirtier, right, then that first unit of control should be worth a whole lot, okay, in terms of health benefits or whatever, right, the recreational, aesthetic benefits, visibility, you know, whatever it happens to be, okay, fish catch or something like that if we're talking about water pollution. Okay, so I could think about a regulator thinking about, right, either estimating a demand curve or coming up with something qualitative that sort of looks like that and then they have to think about the cost of abatement as well, but the general idea would be, you know, I sort of don't care how you firms get there, right, but what I'd like to do is I'd like to get someplace kind of close to this efficient level of pollution abatement, that's going to be the Q*, right, this part of the equilibrium here, and there's going to be some associated price with that, that would be that P*, and you know, not probably terribly surprisingly that that efficient level of pollution control is almost never going to be all of it, right, so the economic prescription is almost never going to be we should eliminate pollution, that's not to say that we don't have statutes that suggest that that's what we should do, right, the Clean Water Act in 1972 was written saying we're going to eliminate, right, emissions to, you know, effluent to water bodies in the United States by like 1986, okay, so we didn't get there, but that is sometimes an important public goal, a public purpose, it's not generally what would be prescribed by the economic analysis, okay, so sometimes those two things would would conflict. All right, so we're going to find some some Q* and then the idea there is how would I reach that, well, I can have these two kind of market-based alternatives in a very general way, one would be I set that tax, okay, the one we saw in the prior diagram and it's the same thing that is what I'm drawing here, but just with this different kind of context in which firms can kind of do anything, right, not just reduce output, and that would be P*, right, so I could set that efficient tax that would get me to that that efficient level of pollution abatement, and two, I could set the quantity, I could just set the cap, right, I'd say I don't want to have any more pollution than Q*, it has to be, right, less than or equal to Q*, so that would be the efficient cap, but the way in which these two things are sort of the flip sides of the same coin is that it turns out that if I set a cap on pollution at Q*, then the permit price, the equilibrium permit price that I get out of that that process would be P*, and if I set P* as the efficient tax, then what I would expect to get out in terms of the amount of pollution control if I really kind of estimated these things pretty well would be Q*, so really this kind of, you know, whether I choose the price versus the quantity doesn't really make any difference for, right, for what's the efficient level of control, right, and we have this sort of nice parallel between the two, there are lots of other differences between those two approaches, and we'll talk briefly about some of them today, and I can, again, I'm happy to take questions about them, but for the most part you want to think of them in a lot of contexts as being fundamentally equivalent, fundamentally equivalent choices, okay, so why do we like those approaches? We like those approaches because the question that we ask ourselves is all else equal, right, you know, if I say that the standard, right, the sort of the command and control approaches versus the market-based approaches, imagine I can get the same level of control, right, everything else is the same between the two outcomes, but the only difference is that one is less costly than the other, well then I would prefer the less costly option, and that's a different criterion than the efficiency criterion. Remember the efficiency we talked about yesterday, I'm comparing benefits and costs. The cost effectiveness criterion says given some level of benefit, right, given that you tell me that we have to achieve a pollution abatement goal of, right, subtracting 100 tons from current emissions, right, or given that you tell me the total allowable catch in a fishery, right, is going to be whatever 100, I don't have a good sense for what that target might be, but imagine it's 100 tons of fish in a year. So given whatever the kind of politically decided, right, negotiated pollution control or resource management goal is, I can still use the tools of economics to try and achieve that standard at least cost, and there again, this is where this is related pretty tightly to what Jim was talking about this morning, right, how can we sort of preserve the greatest amount of habitat for the least cost or, right, these kind of cost effectiveness questions are often where we have real, relatively productive conversations as economists with other folks engaged in resource management or regulation, because it's, it's, right, we sort of take the benefits question in some ways off the table, right, you tell me what benefits you want to achieve, I'll tell you how to get there at the least cost, okay. All right, so why are they cost effective relative to the technology standards and the uniform performance standards? There are two reasons for this. One is a short run reason, and I'll go through the, the logic of that in a few minutes using some graphs. In the short run, they take advantage of the fact that some firms have different costs of abatement than others, okay. If there's no variation, no heterogeneity at all in, in what firms would do to comply with the regulation, then this doesn't matter at all, right. I can choose whatever instrument I want, but as long as there are firms that would like to do different things, some would like to switch inputs, some would like to install a pollution control technology, some would like to reduce output, right. As long as there is that heterogeneity out there and what they'd like to do, what their least cost abatement choices are, then in the short run we're going to take advantage of that when we use a tax or a cap-and-trade system, and we're not going to take advantage of that when we use a technology standard or a uniform performance standard, and that's going to make those traditional approaches more costly relative to the market-based approaches. And then the second reason they're more cost effective, and this in some ways, if you think about a problem like climate change, is even more important, right, because ultimately our ability to achieve big cuts in emissions of something like carbon dioxide over the long run is going to depend on how much it costs to do that, right, how costly that is. That's important politically, it's important economically, and so what we'd really like is we'd like firms over time to try and bring down the cost of pollution control. We'd like them to invest in either innovation or diffusion, right, or adoption of technologies that reduce those compliance costs, that make it cheaper to control pollution, because the more we do that over time, the more pollution control we get for the same amount of money we spend, and it turns out that this advantage, there's a strong advantage of market-based approaches on that measure as well, and again, both in theory and borne out in empirical studies, and we like that because, right, that has this sort of long-run incentives for technological change to give us some confidence that we're going to be able to meet much more stringent goals in the future, okay. Without that, it's awfully hard to, all right, see how we move toward those more stringent goals. Okay, so now I'm going to go through the theory side of why the market-based approaches are more cost-effective than these kind of uniform approaches like the technology standards and the performance standards. So I'm going to take a case of two firms, and what I'll say here generalizes very nicely to end firms, right, so a large number of polluters. This is a small number because it's easy to show on a graph, and the key thing that I want you to keep your eye on is this cost heterogeneity, right, it all comes down to firms having different costs to comply with the regulation. So what I'm going to draw here is abatement again on the horizontal axis, the cost of abatement in dollars per ton on the vertical axis, and I'm going to have two firms, and firm A is going to be engaging in abatement activities that increase as I move from left to right, okay, so it starts out at zero on the left and goes all the way to a hundred tons on the right, and firm B is going to move in the opposite direction, and these guys are going to have costs associated with that abatement, and firm A is going to look like this, okay, so her marginal abatement cost is going to be upward sloping, it could be linear, it could be all kinds of things, a lot of different shapes, we can talk about that if people have questions, but one reasonable assumption is that the cost of abatement on the margin increases as I engage in more and more abatement, and they might even increase at an increasing rate, okay, as I try and get after those last units right out of my production process, and that last little bit of soot out of the stack, it's going to be much more costly than if I tried to get that first unit out of the stack, okay, so that's firm A's marginal cost curve, it's essentially firm A's supply curve for pollution abatement, okay, that's what, that's what she'd like to do, and firm B is going to move the other way, okay, so from zero to 100 moving from right to left, and B is going to have a marginal cost curve too that looks the same, right, because they're kind of mirror images of each other, they're going to cross in the middle like that, and the way I've drawn firm B's cost function, it's a lot steeper than firm A's, so when I think about, right, kind of characterizing these firms, firm A is the kind of low cost firm, meaning it's cheaper kind of on the whole, not for every unit, right, but sort of on average it's cheaper for A to abate pollution than it is for B to do so, and A's and B's rate of increase in their costs is higher than A's rate of increase as well. So a couple other things to notice about this graph before we think about what's the right way to allocate pollution across these two firms. The first is, remember I said kind of all else equal, you'd like to have the more cost effective approach, you'd like to have the cheaper approach, and so the kind of key all else equal is how effective is the pollution control standard, right, so what I'm going to use here is an example in which I want 100 tons of abatement, and no matter what I choose, no matter how I choose to allocate abatement between these two firms, right, I could say firm A doesn't have to do anything, and firm B has to abate 100 tons, I could say firm B, right, doesn't have to do anything, and A has to abate 100 tons, or I could choose any combination of those two spots in between, and at every point on the horizontal axis I'm still going to get 100 tons of abatement, right, that's kind of a key thing about the graph, I could choose any point and I'm still going to meet my standard, okay. So now, let's start with the kind of intuitive place to go, right, I'm a regulator, I got to reduce 100 tons of abatement of pollution, I've got two firms, one very kind of fair place to start would be to say, well, let's just split it 50-50, right, let's say firm A, you have to abate 50 tons, and firm B, you have to abate 50 tons, and so I'm going to draw this line at the 50-50 distribution of pollution control. So, how much did that last ton of pollution control cost firm A? Does anybody want to take a guess? That 50th ton from this picture. Half as much as firm B? Yeah, why are you saying half as much? Well, just the height of. Good, right, so the height of this cost curve is just telling me, right, for that last ton emitted, how much did it cost, and if I had dollar values here on the vertical axis, you could just kind of read the height across here, and you'd say, well, you know, if this was 10 bucks, it kind of looks like this is, you know, 20 or 25 bucks, right, so you're just comparing the heights of those curves between the two firms. So, am I, with this allocation, the 50-50 allocation, am I achieving that 100 tons at the least cost? No, but Pete doesn't count, so Michelle's shaking her head no. Why no? Is it correct to say no? Why no? Simone? Well, because if A were doing the reducing, it would cost less than for B to do the reducing. Yeah, for that last, you know, if I'm at the 50th ton, it cost me more, right, to have B do it than it would have to have A do it, okay, so at 50-50, I could start at 50-50 and I could start reallocating that pollution control responsibility. I could say, well, you know, B, it's pretty costly for B to reduce that 50th ton, so I'm actually going to back up along B's curve, right, I'm going to sort of move away from 50 and go down, and what if I said, you know, B only has to do 45 and A has to do 55, I still get my 100 but I do it cheaper, right, because I saved, right, all these distances between those two cost curves, and the same thing is true, we can just keep asking ourselves that question, well, can I reallocate it, right, if I go this way, I make things worse, right, so then, right, if I, for that last unit of pollution, my cost to B was a lot bigger than my cost to A, but every time I move to the right and I reallocate some pollution control from B, the high cost firm, to A, the low cost firm, I save some money, right, if I'm sort of moving, and that's true until what point, well, not surprisingly, right, the point of intersection, right, where the two marginal costs are exactly equal to each other, right, for that last unit of pollution at that point of intersection, now it costs me the same amount, right, A and B are both paying the same thing on the margin, it's not possible for me to reallocate control from one firm to another and save money, any time, any way I reallocate, once I get to the intersection, I make things worse, right, either B is more costly than A in this direction, or A is more costly than B in that direction, so where I'd like to be, right, if I'm doing this in the most cost effective manner possible, is right at the intersection of the two curves, which is not in this graph a 50/50 split, but it's a 60/40 split, right, so the low cost firm firm A is going to do 60 tons of abatement, and the high cost firm for B is going to do 40 tons of abatement. So, another way of looking at that is to think about the total costs, right, so these are marginal cost curves, that means I'm just saying for each ton of pollution, how much did it cost me to reduce that ton, to abate that ton. So, to get the total cost, right, at a 50/50 split, I'm going to take the area under that curve. Why? Because, well, for the first ton, right, it cost A this much, and the second ton, it cost that much, and I just start shading in, right, every time I control a ton of pollution, I shade in that space, and as I do that, and I think about pollution being this kind of continuous process, and not just, you know, not available just in discrete chunks of one ton at a time, right, then I can fill in the whole space, and I can call this, this is A's total abatement costs, right, if I give her the 50/50 split, right, and then similarly, right, the area under B's curve represents B's total abatement costs at the 50/50 split, and you can, again, see visually, but kind of in a different way, right, than just comparing the curves, that once I kind of passed this cross, right, I have this excess cost, this shading between the two cost curves, right, so this space in here, it's almost like a triangle, if they were lines, right, then it would be, is the space between the two curves, that's that excess cost above the 60/40 split, right, what we would experience with the 60/40 split. So, the cost effective abatement that we've already identified is, would be here 60/40. Okay, so that's the cost effective split, and what you'll notice, right, is it minimizes the total cost, the same thing, you're never going to get that sort of smaller area under the, smaller area under the curve than what you get at the intersection of the two, so that it's the same thing as comparing the marginals, and then you can identify, right, the kind of cost savings of the 60/40 split relative to the 50/50 split on the curve as well, as we said earlier, and so when I say a cost effective allocation of pollution control, I mean something very specific. I mean, I'm allocating that responsibility for control across firms in a way that equates the marginal cost of abatement across every firm that's engaging in some pollution control. Okay, so if I'm asking you to do something, then the cost of that last ton that you control is exactly the same as everybody else, right, who's controlling pollution under the regulation. And if that's not true, then, right, we're not doing this at the least cost. So now what we want to know is how do these different policy instruments, ways of attaining, right, the 100 tons of abatement stack up against that criterion of cost effectiveness. Okay, so we'll start with the traditional approaches, right, so the technology standard, right, so, you know, 1970, the Clean Air Act, there were exceptions to this, and we'll talk about it in a few minutes because it gave birth to an important sort of, kind of pilot trading program in the United States, the sulfur dioxide trading program. In the 1970 Clean Air Act, one of the things that the EPA did is it required the insulation of scrubbers by certain electric utilities, right, to eliminate sulfur dioxide or to reduce sulfur dioxide pollution from power plants in the U.S. So technology standard saying, right, everybody should adopt the same technology is not going to meet our cost effectiveness criterion. It's not going to equalize marginal costs across the regulated firms for two reasons. One is that different firms have different adoption costs. So if you think about the pollution control case and the requirement to adopt a scrubber, an example of that would be, right, I have a kind of rural electric utility. I've got tons of land. I've got mad big coal piles sitting around. I just clear out a little space and install this big piece of capital equipment called a scrubber, which actually has a pretty big footprint, kind of pretty big land footprint, but it's cheap for me to do that because I don't have to acquire new land or anything like that. But maybe I'm a power plant in downtown Atlanta and then I've got a problem, right, where am I going to fit the scrubber? I'd really rather do anything than adopt the scrubber, right, you know, give me some other options, some other things to do because it's very expensive for me to do that. So I'm not going to meet this criterion of, right, a sort of equalizing marginal costs across firms because firms have different adoption costs for different technologies. And even within the same firm, again, right, the Atlanta power plant is going to say, geez, it'd be cheaper for me to switch fields or it'd be cheaper for me to, you know, to, right, go from coal to gas and so on. So I'm not even going to minimize costs for individual polluters. So that's kind of a red flag, right, that I'm not going to get to that cost-effective allocation. And then the performance standard, right, the one that I gave you was uniform. It just said everybody has to reduce by 50 tons, okay, and that was not cost-effective because it didn't take advantage of this difference in costs across firms. On the other hand, right, that's not because it was a performance standard, which is pretty flexible. It says, right, you can attain that however you want, install your technology, right, or reduce, you know, change your inputs. It was only not cost-effective because I set it the same for everybody, right. So as a regulator, I could, if I had information about firms' marginal costs, I could set a firm-specific standard. I just say, firm A, you're going to do 60. Firm B, you're going to do 40. And that would be the cost-effective allocation of pollution control. But from the regulator's perspective, that's pretty information heavy, right. I have to know a lot about, I'm, I'm taking a guess as the regulator about what firms' marginal costs are in a market economy. I don't know. And if I start asking them, you know, you know, are you a high-cost or are you a low-cost firm? I'm setting some regulations. Are you high-cost or low-cost? You're, you're high-cost. You want to be high-cost, right, because you want your allocation to be as low as possible. So there's an incentive, right. There's this conflictual, potentially conflictual relationship between the regulated communities and the regulator. And everybody wants to say, I, it's extraordinarily costly for me to, right. I, it's so costly. You can't even imagine. Please give me the low allocation, right. So everybody wants to be the high-cost firm. Okay. So regulators have to know the cost curve. It's hard to know that. There's a lot of interesting research on how, how good a job we do of guessing those costs before and then looking at data afterward and trying to learn from that. But in general, the regulator's in a difficult position trying to understand how firms behave when they think about pollution abatement. Okay. What about the market-based approaches? Well, let's think about this, just with the sort of simplest case, which is the tax. All right. So I'm going to simplify the diagram from what we saw before. I'm just going to take firm A. Okay. So now I've just got firm A's cost curve on here. It's the same thing you saw before, but everything else is kind of erased. And firm A is going to engage in some pollution control, right. And if we imagine she has some uncontrolled pollution, that's Q max. That's the most abatement she can do because it's the amount of pollution that she's emitting without any regulations. Okay. So she can, she can engage in abatement activity. She can control pollution anywhere from zero, right. She can do nothing all the way up to Q max, right. Her, her uncontrolled emissions level. So she's got some costs of doing that. And if I want to implement a tax and I'm the regulator, I'm going to set the tax at some level, right. Imagine I set it right here. Okay. So this is my new tax. So every time you emit a unit of pollution, a ton of sulfur dioxide, then you have to pay T. Okay. So now the firm has to decide how much to reduce pollution. They got to decide how much abatement to do. How much are they going to do? What's firm A going to do? She's going to look at this calculus. I can either abate a ton or I can pay the tax. So am I going to abate this first unit of pollution, this first ton right here? Would I rather do that than pay the tax? Yeah. Why so? It's lower. Exactly. Right. So here's my, here's my cost of abatement right here. That's what it cost me to reduce that ton. But if I choose to pay the tax, I'm going to have to pay this much. Okay. So I'm going to keep asking myself that question. I'm going to keep right for every ton that I would emit in the uncontrolled setting until I get to this point. Again, it's a point of intersection, not a big surprise. Okay. But why? Well, she's going to, she's going to engage in as much abatement as she can until it becomes more costly for her to abate those units, right? This guy rises above the tax right at this intersection point, right? So it's cheaper to abate, it's cheaper to abate, right? And then all of a sudden I get to this point and all of a sudden it's cheaper to pay the tax. Okay. So for the pollution beyond that intersection point, she's going to just go ahead and pay the tax. And so what's her total cost going to look like? It just looks exactly like it did before. When we think about the abatement cost, it's the area under this curve, right? This is the quantity that she'll abate given the tax when she faces that. And I can sort of fill in that space. And that's the same space we identified earlier. There's something extra that happens under the tax to the firm, and it's why the firm does not like the tax. And that is the tax bill. Okay. So now, right, I have to pay for every unit that I control, I pay my marginal cost. And then I go to the right of Q tax, the amount that I decide to abate, and now for every unit that I emit, I have to pay the tax. Okay. So now I'm going to fill in this space too, and I'm going to say from, right, the amount that I abate all the way to my uncontrolled level of pollution, right, everything that's left over after I engage in my pollution reduction activities, I have to pay T. Okay. So now I'm going to have this extra cost, which is the area of the rectangle here, and that's my tax bill. Okay. So I've got some abatement costs, and I've got a tax bill, and both of those are compliance costs from the firm's perspective. Okay. All right. So in thinking about whether the tax meets our cost effectiveness criteria, there's two things to think about. One is, does it equate those marginal costs of abatement across firms, right? I define that as the criterion that we have to meet to say something is the least cost way of reducing the pollution to, you know, by 100 tons. And for the tax it is. Why? Because I could have lots of firms with lots of different marginal cost curves. One could look like this, and like that, and like this. And they're all making that same choice that firm A is making. They're trying to decide where should I stop controlling pollution. Oh, I should stop as soon as it's more expensive for me to control than it is to pay the tax. So what are they all doing? Well, they're going right to the point where their marginal cost is equal to the tax. And if the tax is the same across all the firms, then so is the marginal cost of control the same across all the firms, right? They're all setting their costs set to the same constant, right? To the same T. And so that's cost effective. Now you could say, but there's this extra cost, right? The firm has to pay the tax, which is also true. And the firm does not like to pay the tax. But when they pay that tax, that money is not just sort of going right away. It's going to the government, right? And the government could presumably do things with that. And that's the difference between taxes and permits. And we can talk about that later in the Q&A. But we're not going to call this a cost to the economy, okay? So this is a cost to the economy. It's a social cost, cost of abatement. This is a transfer. This is money that I move from one pot to another, okay? And it may take away from the real resources that firm has in its pocket. But as far as sort of social costs go, I'm not increasing social costs by the amount of the tax. I'm only transferring money. Okay, so the tax passes our test of being cost effective. It has some downsides. We get this uncertainty over the quantity of abatement that gets delivered. And firms are not going to be happy with their tax bill, okay? Other than that, the tax can be a good thing. All right. What about cap and trade? How does this work? So we're going to go back to the diagram that I drew earlier. But first, let me just tell you kind of intuitively how this works. It's very similar to the uniform performance standard with one important difference, right? I might say, I've got to reduce pollution by 100 tons. I'm going to give you, right, permits in the amount that basically require you to reduce pollution by 50 tons. I'm going to give you permits in the amount that require you to reduce by 50 tons. But instead of, right, just kind of keeping that at 50-50 the way I did with the uniform performance standard, I'm going to let you guys trade. And you, between the two of you can sort of figure out, right, how to allocate that responsibility for control so that we meet the 100 ton standard. So firms are going to, under those conditions, they'll buy and sell these allowances, these kind of permits, right until those marginal costs are equal across firms. And we'll just look at the graph briefly again so that we can see that that's the case. And so just like the tax, right, where everybody's setting their marginal cost equal to the tax and thus the marginal costs are equal across all firms because firms are trading, right, and they've got their, we end up with this equilibrium permit price in the same way the marginal costs end up equal across all firms, right. They're all equal to the market price of a permit, okay. It's a P* instead of a Q*. So the equilibrium allocation, this is again something I mentioned briefly earlier. One of the nice things is it doesn't matter where I start. So let's go back to the graph and we'll, again, we'll kind of think about this, right. If I sort of, if I think of myself as starting at 50-50, my dashed line, the vertical line here, well, where are my incentives? Well, B is the high cost firm, A is the low cost firm, so B would rather not abate that unit. And he would be willing to pay something to A to abate that unit. And A would be willing to accept something, some compensation, right. For A, as long as the compensation was greater than her marginal cost, she'd be happy. And for B, as long as the payment was less than his marginal cost, he'd be happy. So we get these kind of mutually beneficial gains from trade until we get back to that point of intersection, right. And again, if I started out at, you know, 20-80, right, give firm A 20 tons of responsibility, and B 80 tons, the same thing would be true. I would just get more trading, right. These costs are just even higher than A's at that level. And so the incentive is even stronger to trade, again, until we kind of move to the point of intersection, right. Or I could start out at, you know, 100-0, right. Give firm A all the responsibility and firm B nothing, right. And then I could find, right, I'd find these gains from trade and we'd end up back at 60-40. So this nicety, right, was, again, it's nice and it's tough, right, because politically, it matters very much to firms who gets what in the beginning, right. So I would really rather be the firm that gets a bunch of permits and gets to sell them to other firms than be a firm that, right, gets a small number and has to buy, right, in order to, to kind of do my least cost thing. And so it matters a lot to firms distributionally, but it doesn't matter really at all unless we have a story about market permit in the permit market, right. And we haven't seen empirically that kind of situation in very many, very many actual applied settings. So that would be an unusual thing to happen. And not only that, regulators are pretty wise to that possibility. And so, for example, in the actual SO2 trading program that EPA set up in 1990 that operated until 2008, EPA just sort of kept a certain number of permits aside and would auction them each year to prevent, right, any single firm from trying to kind of corner the market, keep out new entrants and that kind of thing, okay. So we might worry about market power, but in general, it hasn't been an issue. And as long as we don't have that, however we allocate these things in the beginning, we're still going to get to 60-40, right, because we're taking advantage of those incentives between firms for those mutually beneficial gains from trade. So, I mean, politically, that can cause some nice situations too, right. So in 1990, when the George H.W. Bush administration was getting ready to implement what it became the sulfur dioxide trading program, there were some standouts, some states that, right, in the Senate in particular, were really resisting, right, these changes in the Clean Air Act that would make it more stringent. And in particular, states that were resisting were states that were in sort of Appalachia, where there either was coal mining, a lot of big coal industry, or a lot of electric utilities that were using high-solfer coal or both. And one particular standout was the state of Ohio. And then Senator Glenn ended up getting a very, very large sort of chunk of the national permits for the program in order to buy his support for the program in the Senate, which was essential to it passing. And again, that was sort of nice for Ohio utilities because they were in the position where they started out from, you know, this sort of nice big permit allocation, and they could sell some permits, right, in the beginning and make some money, but it didn't matter ultimately, right, for achieving the cost-effective allocation across firms, right. So you can like or dislike that story, but in general, it makes something, it's sort of something that's appealing about cap-and-trade programs from politicians' perspective because they have something to give out, right. We're not just, right, it's not just the stick, but there's also some carrots. Okay. So, how does this work? Right, again, the equilibrium allocation is independent of the initial allocation. That could be a nice thing. Unlike the tax, the regulator doesn't have to know firm's marginal cost, right. This is something that's a little bit magical about cap-and-trade programs, right. When I said, you got to, in order to get Q*, I kind of had to know something about the costs in order to get to Q* with a tax. But if I just sort of know, if I set Q, I know I'm going to get Q. That's my cap, right. I'm going to enforce the cap. And I don't have to know anything about individual firms' marginal cost curves in order to know where, right, right, whether I'm going to achieve Q or not. I just, I get it, right, because the firms are trading with each other. So, unlike the tax, I have this certainty over the quantity. I don't have those certainty over, ultimately, the price per ton, right. So the tax, the permit, equilibrium permit price is going to end up wherever it ends up, right, causing some uncertainty for firms in terms of their compliance costs, whereas the tax, right, I pin down the price, but I don't know the quantity, okay. Is everybody clear on that distinction? Okay. So that's the story, the kind of short-run cost-effectiveness story, right. That's why we tend to like these, again, all else equal for the same standard achieved, we tend to like the taxes and the permits over a uniform standard, a technology standard that doesn't take advantage of these cost differences across firms, okay. So, all right, what about the second question, right, what's the incentive to kind of bring down the cost of pollution control over time? Again, all else equal, we'd kind of like to have a standard or some kind of pollution control policy that would help us do this, okay, that would provide incentives to bring, to do this. And I'm not going to go through the theory on this, but I'll tell you what it says, and that is, right, when this, we're in this long-run setting and my choice of abatement technologies is not fixed and I can choose, right, to change my whole factory or change my pollution control process, I can lower my abatement costs by developing or adopting new technologies, but I'm going to have to usually make a pretty big upfront capital cost to do that. So an example would be, right, what if all of a sudden there's on the market now there's a new, more efficient scrubber, right, so its operating costs are lower, but if I want to do that, if I want to get those sort of lower operating costs of that scrubber, I have to buy the new thing, right, I have to retire my old one and I have to buy a new one. And so firms are always in the position where they're trying to weigh those kinds of decisions. Should I invest in the cost reducing technology or should I not, right, and it depends on the cost savings versus the capital cost of the investment plus discounting, right, I think about this thing lasting over a long period of time. Okay, so it turns out that the market-based policies, because they put a price on pollution, right, because every ton that I abate I have to pay some costs for, because of that I'm going to provide these greater incentives for technological change than what I see under either type of command control policy, the technology standard or the uniform performance standard. Again, we can go through the theory on that, I'd rather kind of get along to some examples of market-based approaches and talk about how that's worked out in practice, but I'm happy to answer questions about that too. Some other things to worry about. Okay, so one thing, what are some potential downsides here? Okay, one would be what I would just going to call the non-uniform mixing problem. So the way I've talked about this and even the way I drew that marginal damage function earlier when we talked about negative externalities, I'm sort of saying, right, that every firm has the same damages from pollution, that it doesn't matter where the pollution is emitted, right, I'm not causing sort of differential marginal damages to society. And that's true in some situations, right, so for carbon emissions, it really doesn't matter whether I emit a ton in Beijing or a ton in Boston, right, a ton in Annapolis, it's causing the same essential quantity of damages in the long run, right, from sort of adding to the stock of carbon in the upper atmosphere. But it's not true of a lot of other pollutants, right, it's not true of most water pollutants, it's not true of some air pollutants, sometimes where the pollution is emitted matters very much, right, for how much damage it causes. And in that situation, we can get into these tricky situations with market-based approaches, right, again, if I take a uniform standard and I say everybody has to cut by 50 tons, okay, it's not necessarily true that everybody is then exposed to the same amount of pollution because there was some variation, right, before I had the standard, there was some variation, there's still going to be variation exposure to pollution. But if I implement a market-based approach and I move around, right, from 50-50 to 60-40, then the place that had only 40 tons of emissions abatement has less abatement relative to its baseline than does the place that has 60 tons. And if it turns out that those high-cost firms, right, the ones that would do less abatement under the market-based approach because it's more costly for them to do so, are also high-damage firms, right, where pollution control gives us big benefits, then this can really kind of mess up the efficiency criteria as well. Then now we're talking about changing the benefits of pollution control because we've redistributed the pollution spatially, okay, and so that can then cause us some problems. And so what are some solutions to that? Well, the solutions are conceptually straightforward but have sort of turned out to be much harder to implement in practice than anyone might have imagined when people started writing about this back in the 1960s. And so the solutions are things kind of like exchange rates, okay, and so this is something that Jim has written about really interestingly, especially thinking about taxes, right, you could either directly tie attacks to the marginal damages at your location, okay, that's one thing you could do, or if it was reasonably straightforward to set up sort of a zonal trading system, you know, if I'm talking about, you know, ozone generated by NOx and VOC emissions in Southern California, and I know the prevailing winds are off the coast from west to east, then I sort of wouldn't like to increase emissions, right, in the kind of western areas, right, and decrease them in the eastern areas necessarily, right, because I'm just sort of blowing that pollution over. So I might have some rules of thumb, right, kind of regional spatial rules of thumb that I can use to create a tax system that taxes, you know, in some simple straightforward way these sort of more damaging emissions at a higher rate, okay. So that can either be firm-specific, location-specific, zones, there's all different kinds of ways of working that out. There could be an intertemporal aspect to this, right, if it matters more in the summer than it does in the winter, if it matters more in the middle of the day than it does in the night, right, so you can think about these different sort of types of taxes to accommodate that system. Okay, so let me give you an example of how this would work with cap-and-trade. So this is a paper back from back in 2005. It's a group of scholars at Carnegie Mellon who did this really neat thing, and they were looking at the problem of combined sewer overflows. Do you ever know what CSOs are, combined sewer overflows, right, so if you live in a place that doesn't have separate storm sewers from the sanitary sewage system, then what happens when you have lots of rainfall all at once, like what happens often in Austin, Texas, although we do have separate stormwater systems, the sanitary system receives all that storm runoff and it overwhelms the capacity the wastewater treatment plant to treat that waste, and so essentially the plants in these areas have permits from EPA to just kind of open the gates, right, and then you get a mix of sort of storm water and raw sewage going into the receiving waters until, right, the flow is sufficiently reduced so that the plant can begin to operate again. And so these are called combined sewer overflows or CSOs, and they're common in, you know, like some northeastern cities in parts of the midwest where it just hasn't been sort of technically feasible/cost effective to like dig up sort of, you know, central Manhattan and install a new storm sewer system and so on, okay, so it's not everywhere but there are parts of the country that have this more than others. So what they did is they looked at the upper Ohio River Basin and the CSOs in the upper Ohio River Basin, and in the basin there are 70 municipal sewage systems that receive runoff during rainfall and have these CSO permits, and the damages that they were looking at were from elevated levels of bacteria, biological oxygen demand, total suspended solids during these CSO events, okay, into receiving water bodies. And so the marginal damages from these depend on, and they're sort of simplified model, right, somebody asked earlier what goes into that again, the sort of flow and some other hydrological characteristics of the receiving water, water as well as the exposed population, okay, so are we talking about big cities or rural areas and so on. And so what they do is they set up this system of exchange rates, okay, so this is the, these are the seven, or sorry, eight of the 70 systems, the kind of eight biggest ones, and this matrix is just saying, okay, here's regulated source one through eight, Clareton through Youngstown, okay, and then here's the, so those are the rows, and then the columns are the same thing, so one here corresponds to the city of Clareton, right, a two corresponds to the city of Greensburg, and the ones across the diagonal of the matrix is just saying, right, this is sort of own trading, right, if Clareton were to trade with Clareton, if Greensburg were to trade with Greensburg, if McKeesport were to trade with McKeesport, so that's why those are ones. But everything else in the matrix is giving you this kind of exchange rate that you would want to use, taking into account the marginal damages at the location of each of these sources, all right, that's the exchange rate you would want to use in a trade, so for example, I circled this, right, Morgantown trading with Pittsburgh, because Pittsburgh is number five, so that's column five, and it turns out that the damages of emissions in Morgantown are higher than the damages from emissions in Pittsburgh, so if Morgantown wants to buy, right, it wants to offset one of its ton, right, one ton of its kind of emissions reduction responsibility by purchasing permits from Pittsburgh, it has to buy sort of four and a half tons for every ton that it wants to reduce of its own responsibility, okay, and so you can read each of these numbers like that, right, if they're smaller, right, then it goes the other way, they can buy less than one ton of emissions, right, from another setting than, rather than reduce it themselves, and so they set up these exchange rates, and you could say, well, why don't they just prohibit trades between Morgantown and Pittsburgh, right, if the damages are higher in Morgantown, and one could do that, you know, you can set those kind of constraints in the system, the problem in a system like this would be, well, then, you know, Morgantown could trade with McKeesport, and then McKeesport could trade with Pittsburgh, right, you'd have to really be very clever, right, in thinking about how to block, right, the movement of pollution toward the cheaper area, and so, right, the system of exchange rates kind of takes care of that, and it says, well, how much is it really worth, right, relative to the damages, how much is it really worth, and then it sort of sets those ratios, just like, you know, an exchange rate does, right, if you're trading currency in another country, and so this is, in theory, how it would work, and there are systems that use these, so a couple to think about or look at if you're interested in this would be the Minnesota River Phosphorus Trading Program has a very carefully, right, again, this is an area where, right, economists are working closely with water quality modelers to understand these impacts, what's important, what are the characteristics that are important, and kind of coming up with these matrices of trading ratios, so the Minnesota River Phosphorus Trading Program is one, the Chesapeake Bay TMDL, right, that was issued back when I was at RFF, I don't know, 2010, maybe, I think, there are three states, Virginia, Maryland, and Pennsylvania that have set up trading programs that also, right, there's this, this, this sort of systems of trading ratios that I think are being used in those programs as well, and there are others, okay, so that's how it would work, in theory, turns out to be, you know, somewhat complicated to implement in practice. Another thing to worry about when we think about cap and trade programs or taxes is monitoring enforcement, so a question came up earlier, kind of, how do you, you know, how do you kind of monitor and enforce regulations like this, and it turns out that while abatement costs tend to be the largest share of social costs when we think about environmental regulation, the kind of monitoring enforcement and associated administrative costs to governments are the second largest category, okay, so to give you a sense of scale with the SO2 trading program that we'll talk about in a minute, that was running from 1990 to 2008, if a firm adopted a scrubber to comply with that, it would have cost on the order of tens to even maybe hundreds of millions of dollars to purchase that scrubber and install it, and probably several million dollars per year to operate it for an individual firm, and in order to enforce that regulation, EPA had to install continuous emissions monitoring in all of the regulated electric utilities stacks, right, so they could, they could know in real time exactly how much pollution was coming out, so they could check and see, right, were firms complying with their permit allocations, and that cost, you know, several hundred thousand dollars to install and about fifty thousand dollars per year to operate, so it's not nothing, right, it's not negligible, it can be a big cost, but on the scale, so relative to abatement costs, it's relatively small, okay, so they tend to be small relative to abatement costs, but it's still, and again, if, you know, you can fit 2500 electric, you know, generating units in the United States on a spreadsheet, but if we're talking about, you know, 140 million cars, or however, you know, hundreds of millions of households are heating homes with home heating oil, but if I wanted to try and do a market-based approach with a much more diffuse, you know, sort of, you know, very large, sort of numerous population of regulated entities, then these monitoring enforcement costs of a cap-and-trade or even a tax-based approach might be very high, okay, so you have to sort of think about the scale of the regulation if I want to compare these things, so, you know, I guess the, the, the point I want to make is that on the cost effectiveness criterion, the market-based approaches are just, in, in most cases, going to be better, okay, and we'll talk about some exceptions to that in a second, but on the monitoring enforcement dimension, there's sort of no clear winner, probably the command and control approaches are cheaper on average, and so you'd want to weigh that against, right, the savings and abatement costs, which are often going to swamp those savings and monitoring enforcement costs, but not always, okay. So are the traditional approaches ever preferable? That's sort of leading into this question, so I would say under three different conditions. One is when I really have these severe kind of non-uniform mixing problems, so I remember in, at the, again, there was this debate we talked briefly about yesterday about over a new arsenic standard under the Safe Drinking Water Act going back to 1999-2000, and at the time an economist, who I know well, but I can't imagine what he was thinking when he wrote this, wrote an article proposing, well, we should have trading, right, and, you know, some communities should have high arsenic levels and others could have low and we could just sort of meet regional standards, and those, you know, highly non-uniformly mixed problems, right, where the the damages are very, very specific, right, to the location of emissions or exposure are really just not well suited, right, for a market-based approach, right. You have, you really run this risk of having the sort of high damage firms doing less abatement or the high, right, or vice versa, and that's not at all going to be what we want for efficiency. So another example of that would be toxic waste, right, I would never say, well, let's just trade toxic waste dumps, right, because all that matters is the total amount of toxic waste produced by the United States. Well, we know that that's not the case, right, the spatial location of that thing matters very much, so we want to be real careful in those situations. A second situation in which we'd prefer the traditional approaches is, you know, a single control technology is just highly effective and everybody would kind of do the same thing anyway, right, abatement costs are very similar across firms. Again, if you think about those marginal cost curves, the only reason we had that space between the curves when we had some trading versus the 50-50 split was because the cost curves were different. I drew one steeper than the other, right, or one higher or one lower, and if that's not the case and everybody has essentially the same cost, then I don't really get any bang for the buck of using a market-based approach, and I might incur some extra costs to do that. So then, you know, in that case, it's not a great idea. So a good example of that historically has been double-hulled oil tankers, right, so what do we do to kind of prevent another Exxon Valdez? Well, one thing that we did was under the Oil Pollution Act of 1990 is that we required the sort of phasing in of these double-hulled oil tankers. Well, if you asked firms to meet some standard, right, that was, had to do with reducing the probability of another spill, they would pretty much universally have chosen double-hulled oil tankers, right, so why go through the trouble of doing some kind of trading, right, or some kind of tax system when that works just fine and is cost effective for most of the regulated entities. And then finally, if the number of regulated entities is really high, again, that's the difference between 2,500 electric generating units regulated by a cap-and-trade system versus, you know, 230 million cars or 140 million households, that kind of thing, right, then we have this problem of monitoring and enforcement. It becomes very expensive to think about implementing, and so as the problem gets more and more diffuse like that, right, then we sort of start to move away from market-based approaches in our kind of mental toolkit of what's available to us. The only thing I would say there is that, you know, I often say, well, kind of, for now, right, so we see, I mean, especially in the aftermath of the Volkswagen, right, this regulatory incident with the German firm Volkswagen, people talk about, well, gee, maybe we should be testing auto emissions on the road, and we just combine license plate recognition technology with, right, sort of field emissions testing, which some states already do, right, to either complement or substitute for some of their kind of in-house emissions testing regulations. Maybe, right, technology, that time, you know, if those technologies combine in a way in which it's cost-effective to regulate emissions in some sort of very diffuse way through a tax or something like that, well, then maybe it would make sense, but in general, the sort of more diffuse the pollution problem, the greater the number of regulated entities, the less it is likely that you would want to go with the market-based. Okay, so just to, I'll just kind of, this is my list of some examples of market-based environmental policies that have been tried. British Columbia has a carbon tax, right, I think it's, does anybody know, $25 a ton, something like that. It's, it's, you know, set some, somewhere reasonably close to what we would think would be, at least from U.S. estimates, the social cost of carbon. It's pretty new, like it's, I don't know of any good empirical analyses yet. I know there are some estimates, B.C. has its own estimates of how it's reduced carbon emissions, but it's hard for me to say how, how good those estimates are. Lots of experiments with cap-and-trade programs, so again, carbon dioxide, right, there's a market in the European Union, there's one in California, there's REGI, the Regional Greenhouse Gas Initiative on the East Coast, there's the SO2 trading program, there's been some NOx trading in the United States, so a fair number now of examples in air pollution control, especially in industrialized countries. As I mentioned earlier, a good example of the tax approach, or the kind of Pigouvian tax approach, are these unit charges for municipal solid waste, or pay-as-you-throw systems. A lot of communities have adopted them, there's some good studies showing that they reduce the sort of volume of the solid waste stream. Interestingly, but not surprisingly, if you don't provide curbside recycling when you do this, right, then you may just not be doing yourself a favor, right, you increase incentives for illegal disposal, and so on. So there are things to think, about other than just, right, what to tax, and how much, you got to sort of think about the whole context. Probably the most successful experiments with these have been individual tradable fishing quotas. Pete, I'm sure, knows much more about those than I do, but New Zealand manages something like 90% or 85% or 90% of its sort of the stocks within its exclusive economic zone using these tradable quota systems. In the United States, there are a lot of species that are managed this way too, although it's been much more controversial, and it's come, and it's gone, and it's been banned, and it's been unbanned, but Gulf Snapper right now, Pacific Halibut, those are two very good examples and pretty well-studied examples. Water quality trading, I mentioned the Minnesota River program and the Tuspeak Bay, nitrogen phosphorus markets for a lot of reasons that we can go into, if anybody's interested. These things have not taken off the way we thought they might, perhaps due to some of the skepticism that Jim expressed yesterday about how ready we are for these things, but that's another set of examples. And there's a reason that I've sort of listed them in this order. Wetlands banking that came up yesterday, tradable development rights, I would put this as sort of, maybe with individual tradable fishing quotas at the top, but then I would sort of put this list in this order as sort of moving away from the easy cases toward the much harder cases, and the cases where it's not clear that we have actually done sort of generated net benefits within the environment in comparison to what we would have done with more of a traditional prescriptive approach, and I think it's very hard to say. I don't think we know nearly enough, right, about what sort of what's happened, but if you think about the the the essential problem here, as we sort of move from the easy cases to the hard cases, I'm just going to skip through this and sort of just go to my summing up slide. What we're doing is we're sort of moving away from those clearly, you know, sort of uniformly mixed problems, right? The problems where it really doesn't matter where the control takes place, or in the resource case, sort of doesn't really matter where the conservation takes place, or it doesn't really matter, right? One kind of piece of a wetland is tradable, fully tradable and equivalent, right, to a piece someplace else or a different kind of wetland. So, right, there's all these things to like about market-based approaches. We like that they're cost effective, we like that they have these long-run incentives for technological change. We like that we see examples in practice where empirically we've actually been able to estimate what have been the cost savings relative to other approaches, right, what have been the environmental impacts. We like all that, okay, but there are these applications on the frontier. There's this impulse, right, that when all I have is a hammer, everything looks like a nail, right, that to say, well, it worked, right, for air pollution control, and so now we should have wetlands mitigation banking, and we should trade development rights and species habitat, and so on. But as we move to these newer kind of, and I don't mean new environmental problems in the sense that the problem is new, but, right, that I'm sort of taking the market-based approach to a new set of problems that I haven't tried before, then we have these new challenges. And it's not that I, I think I may be somewhat more optimistic, for example, in the water pollution case than, than what Jim expressed yesterday. I think there's been a ton of progress made in this interdisciplinary work, right, trying to understand how these things work, how they might work better, and really trying things on the ground. But I also think, right, that the key problem is that for these non-uniform damages, the markets have to have constraints, and they have to be designed really carefully, right, in order to not make things worse, right, not just make things more costly, but actually make things worse in terms of benefits. And so the design of those constraints requires these inputs from the natural and physical sciences. It requires, right, this kind of working together across disciplines in ways that I don't think have happened either in regulatory agencies or in academics, or both, to the degree that, that they would really need to in order to make these things fly pretty well.