What is Environmental Economics?

Economics is the study of the allocation of scarce resources. o Note that the theories of economics can be applied to any scarce resource, not just traditional commodities. o Economics is not simply about profits or money. It applies anywhere constraints are faced, so that choices must be made. o Economists study how incentives affect peoples behavior. Environmental and natural resource economics is the application of the principles of economics to the study of how environmental and natural resources are developed and managed. o Natural resources resources provided by nature that can be divided into increasingly smaller units and allocated at the margin. o Environmental resources resources provided by nature that are indivisible. o Natural resources serve as inputs to the economic system. Environmental resources are affected by the system (e.g. pollution).

Why Study Environmental Economics?

In general, prices reflect the relative scarcity of goods. o However, in environmental economics, markets, and thus prices, often do not exist. What aspects of environmental and natural resource economics make it unique? 1. Market failures When market failures exist, government intervention may be appropriate. 2. Dynamics The decision to consume a good today typically does not affect the ability to consume it tomorrow. However, the decision to use natural resources today does affect what will be available tomorrow. Note that prices will influence this. Higher prices both provide incentives to conserve resources, encourage exploration for new sources, and the development of technologies to better obtain resources. 3. Irreversibility Damage to natural resources has long-term effects. For example, if the Grand Canyon were flooded, future generations would be unable to enjoy its beauty. This is not as large a problem for normal consumer goods. 4. Linkages between the economic and ecological system An interdisciplinary understanding of the environment, political science, etc. necessary to be a good environmental economist.

I.

The Invisible Hand

The goal of today's lecture is to discuss why the free market may fail, and to discuss the consequences when that happens. To begin, we must consider how the market should function.

Adam Smith, who coined the phrase "The Invisible Hand," was the first to notice that a free market of individuals acting in their own self interest leads to a socially-desirable result. Why does this occur: o Demand = Marginal Benefit (MB) o Supply = Marginal Cost (MC) o In equilibrium, P = MB = MC No further beneficial transactions are possible. Normally, a free market brings us to this point. However, there are times when private marginal benefits or costs are not equal to social marginal benefits or costs. When this occurs, the market is unable to allocate resources efficiency. We call this market failure. o The welfare lost because beneficial transactions do not occur is known as deadweight loss.

II. Market Failures in Environmental Economics

Most of today's lecture dealt with the problems that various types of market failures cause. Of course, most of the course will focus on externalities. However, it is useful to be aware of other types of market failures and how they are relevant for environmental economics. We discussed three in class today. A. Imperfect Information

The market depends on perfect information, so that everyone knows all of the options available to them. If this is not possible, people may not make optimal choices. o Note that imperfect information is when different parties have different levels of information. If no one realizes an activity is bad (e.g. Mercury pollution in Onondaga Lake in 1950s), imperfect information is not the problem. It may be that the result is uncertain. Uncertainty is an important consideration for environmental policy, which we will discuss later in the course. However, if all sides have the same knowledge, even if uncertainty exists, imperfect information is not a problem. How is this relevant to environmental economics? o People may have imperfect information about things such as health risks or the dangers of pollution. What can be done? o Information can be provided by the government or by private individuals (e.g. "dolphinfirendly tuna"). o The government may provide services that are not provided by the market because of imperfect information (e.g. insurance). B. Public Goods

Public goods have two key features: 1. non-rival -- one person enjoying the good does not keep others from enjoying it 2. non-excludable -- people cannot be kept from enjoying the good

Leads to free-rider problem. Because the goods are non-rival, efficiency requires that the sum of each individual's marginal benefit equal marginal cost. Underprovision results when public goods are provided by a free market. How is this relevant to environmental economics? o climate o clean air o parks What can be done? o The government can provide public goods and finance them with taxes. This helps to alleviate the free-rider problem. However, it still may be difficult to get people to reveal their true preferences for the good.

C. Externalities

An externality is an activity of one entity that affects the welfare of another and is not reflected in market prices. o A key feature of this definition is that the welfare of others is not reflected in market prices. To find the efficient level of activity, we need to know the marginal social cost. o Marginal social cost is the sum of marginal private costs and the marginal external costs, which represent the damage done by the externally. o Note that, without policy, the free market will not lead to an efficient solution. Prices will reflect private costs, but not the additional external costs. o Individuals equate MPC and MB. Since MSC > MPC, over provision results.

III. Open Access Resources

Open Access Resources (also called an common property resources) are resources or facilities that are open to uncontrolled access by individuals who wish to use the resource. The problem is a lack of property rights. Consider someone using a public grazing area. o Individuals equate MB = MC o The benefits are the value of using the grass. o There are two costs: 1. The cost of obtaining the resource 2. The opportunity cost of not being able to use the resource later o The user bears all of the first cost, but only part of the second cost, as it is shared by all users. As a result, MPC < MSC Thus, MB = MPC < MSC This leads to over-utilization of the resource. Fisheries provide a good example of open access resources. Everyone shares the costs when fish are caught, as it makes it depletes the stock, making future fishing difficult.

Note the dynamic nature of the problem which, as we noted last week, makes natural resources unique. This leads to overfishing. Management options Annual quotas These restrict the number of fish that can be caught in a season Leads to a "race to fish" Everyone wants to catch fish early, before others have caught them. Technology restrictions Only certain technologies can be used to catch fish This makes fishing more difficult, but doesn't directly address the commons problem. Individual Transferable Quotas (ITQ) Used in Australia, New Zealand, Canada, Iceland, and Alaska Divides an annual quota among fishermen as a long-term right This provides incentives to manage the fishery properly Shares can be traded In a recent study, fisheries managed by ITQs were half as likely to collapse as other fisheries The Alaskan halibut and king crab fisheries provide examples Before ITQ, the fishing season lasted 3 days Now, it lasts 8 months Not only is it safer, but fish are sold fresh, and at higher prices, since there isn't a one-time glut on the market. Challenges Hard to use in international waters Allocating shares is difficult Shares provide economic rent to those who receive them

IV. Hardin: Tragedy of the Commons

Hardin's article, "The Tragedy of the Commons", relates the open access resource problem to the environment. o How does Hardin relate the Tragedy of the Commons to the environment? With pollution, people are putting something in to the commons, rather than taking something out. Nonetheless, each individuals share of the cost is small. Note that it is the lack of property rights that causes the problem. For example, there are no property rights for air. At the same time, Hardin notes that private property encourages pollution. Says people think it is their right to dirty air/water by their property. Hardin notes different ways to ration scarce goods. Note how they depend on different criteria. o Note that auctioning spots uses the market. However, using the market is just one way in which scarce resources can be rationed. Other methods achieve the same goal. o One important criterion for choosing the method depends on our ethics toward the resource. For example, we may be more comfortable using wealth as a way to ration cars than as a way to ration national parks.

In this case, some other method is necessary to ration the good.

Modeling Pollution

We began by finishing our discussion of the commons. These include the means for rationing scarce resources, which are on the last lecture notes, and strategies for successful commons management. These include the following: o Successful commons usually have elaborate conventions over who can use resources and when they can be used o What one takes out must be proportional to what is put in o Usage must be compatible with the common's health o Everyone must have a say in the rules o In successful commons, more attention is usually paid to monitoring and conflict than on sanctions and punishment. As we discussed, applying these lessons to "new commons" such as the global climate, would be difficult.

I.

The Marginal Damage Function

Our goal, of course, is to develop policies that lead to an optimal level of pollution. o Recall that the optimal level of pollution is not zero. o We need to consider the marginal costs and marginal benefits of pollution. The marginal damage function shows the damage done by an additional unit of pollution. o It is upward-sloping. o The slope normally gets steeper as emissions increase. However, it may level off if there is a point where no more damage can be done (e.g. all the pond life is dead). o We can look at damage in one of two ways: 1. Emissions damage functions -- the marginal damage done by additional flows of emissions. 2. Ambient damage functions -- the marginal damage done by additional concentrations of pollution in the ambient environment. The area under the marginal damage function shows the total damages. The type of pollutant affects the shape: An example where it gets steep very suddenly is a threshold effect. Flow pollutant A pollutant that the environment can absorb. As a result, only the amount that occurs at a specific point in time matters (e.g. waste flowing into the river). For a flow pollutant, the MDF does not change over time (if other things remain equal). Stock pollutant A pollutant that the environment cannot absorb. The level of the pollutant in the environment grows over time as the pollutant is accumulated. Example: CO2 emissions (take 200 years to decay). MDF is flat, and shifts up over time, because the stock keeps getting larger. Things that affect the position of the MDF include: Population

Time of year II. The Marginal Abatement Cost Curve Marginal abatement costs can be: 1. The costs of reducing pollution (e.g. costs of scrubbers, labor needed to maintain them, etc.). 2. The opportunity costs of lowering consumption or production. The marginal abatement cost curve is downward sloping, and equals zero at the level of unconstrained emissions. o Firms choose the easiest ways to reduce pollution first. o It may flatten if economies of scale are present. Things affecting the position of the MAC: o Technology To add together the MAC of individual firms, we use a horizontal summation. See Figure 5-5 in Field. o Note that this follows from what Field calls the equimarginal principle to minimize total abatement costs, choose the lowest marginal abatement costs first, even if it means one firm does more than the other. Intuition: we do the simplest (the cheapest) abatement first. III. The Efficient Level of Pollution The optimal level of pollution is where the MDF and the MAC curves intersect. Here, the additional benefits from pollution control are just equal to the additional costs.
o o

In these examples, the marginal benefits are the marginal damages avoided by increased abatement. Note that this is not where total benefits equal total costs. If that were the case, net benefits would be zero. Rather, we maximize net benefit by equating marginal benefits and marginal costs. Some examples: 1. How would the desired level of pollution control change if a new technology is discovered that improves the efficiency of scrubbers for power plants? A new technology lowers the marginal abatement costs curve. Since abatement is cheaper, we should do more of it. The efficient level of pollution falls. 2. How does the desired level of pollution change between summer and winter if the pollution leads to greater problems in the summer (e.g. ground level ozone)? Here, the marginal damage function is higher in the summer than in the winter. As a result, we want less pollution (e.g. more abatement) in the summer.

The Coase Theorem

I. Enforcement Costs

For any environmental policy, we also need to consider the costs society pays to enforce and administer the policy. o These can be modeled as increasing the marginal abatement cost, which decreases the desired level of abatement. Thus, an important policy consideration is the level of enforcement. Enforcement can be continuous or random. o For example, some EPA air regulations require installation of a device to constantly measure emissions (continuous emissions monitoring systems, or CEMS). o Alternatively, random spot checks can take place. The problem is to balance out the cost of monitoring and the punishment o For a regulated firm: MB of compliance = avoided penalty = penalty for cheating * probability of getting caught MC of compliance = marginal abatement costs o Thus, the government can increase compliance by either raising the penalty for cheating or increasing the probability of getting caught. Raising the penalty is less costly for the government, but it must be practical.

II. Evaluating Environmental Policy

Analysis of marginal damages and marginal abatement costs assumes that efficiency is the criterion for evaluating environmental policy. However, this is just one of several. Our list included (sorry if I've forgotten some): 1. Efficiency Here, we discussed using cost-effectiveness as an alternative. A cost-effective policy meets a given goal for the least possible cost. However, criteria other than efficiency (such as minimizing health impacts) may be used to set the goal. 2. Enforceability 3. Equity 4. Long-run considerations How well a policy encourages innovation is important here. 5. Regulatory capacity 6. How well does the policy internalize the externality? 7. Ethics/moral considerations 8. Political considerations The Coase Theorem

III

The Coase Theorem is the notion that an efficient solution will be achieved independently of who is assigned property rights, as long as someone is assigned the rights. o Coase implies that once property rights are established, no government intervention is necessary. o Note that the distribution of income in the final outcome will vary based on who is assigned the rights. The Coase Theorem doesn't simply mean that assigning property rights to a polluter will cause the pollution to continue. A deal could be struck among both parties to bring about a more desirable solution. o However, the decision on property rights will affect the distribution of income in the final outcome. Coase's main points: 1. Externalities are reciprocal in nature. Not only does the pollution cause an externality, but also the presence of the victims harms the polluter. If no one were harmed, there would be no problem. 2. The economic problem is to maximize the value of production. Thus, you need to determine which activity has the higher value. Since externalities are reciprocal, Coase argues that the highest value option should be preserved. 3. Victims should not be compensated Because of the reciprocal nature of externalities, compensation would lead to too many people living in harms way.

Command and Control Policies for the Environment

I. The Coase Theorem (continued)

Our discussion of the policy examples on the reading list suggested several limitations to the Coase Theorem. These included:

1. Costs of bargaining and transactions costs o Negotiation won't work when large numbers of people are involved, or when the victims aren't well defined (e.g. endangered species). o Similarly, different groups may have different bargaining power, affecting the distribution of the final outcome. 2. Need to be able to clearly establish who causes the harm. 3. Willingness to pay and willingness to accept are different. o Because of income effects, you may not be willing to pay as much to avoid damage as you would require in compensation to accept it. 4. Definition of property rights might affect the number of participants. o More people will come to the problem if they are likely to be compensated. 5. There is no guarantee that bargainers will reach an agreement. II. Liability Law

An example of how the Coase Theorem applies to environmental policy is liability law. o If a firm will be held liable for the damages from its pollution, it has incentive to avoid pollution when the marginal abatement cost is less than the marginal damage. By avoiding damage, the firm lowers its liability. Note that the government does not need to know the marginal costs of the firm in order to achieve the desired level of pollution. Complications o The burden of proof may be difficult in court. Need to know both who causes the harm and what the damages are. o Will courts be willing to hold polluters liable for damages that they cannot pay? o Litigation is costly.

III. Command and Control (CAC) Regulation

Command and control regulation uses the setting of standards. o A standard is a mandated level of performance that is enforced in law. o A standard simply makes excessive amounts of pollution illegal. In principle, the government can set the standard to yield the efficient level of pollution control.

IV. Types of Standards

Ambient Standards o Regulates the amount of pollutant present in the surrounding (ambient) environment. o Examples: Parts per million (ppm) of dissolved oxygen in a river Sulfur dioxide (SO2) in an airshed Ground level ozone levels (ppm) o Measures are often an average (e.g. over a 24 hour period, or per year) This is important, as concentrations vary by time of day and by season (e.g. due to changes in weather) o Note that the level itself cannot be directly enforced. Rather, the sources of the pollution must be found and regulated to be sure that the ambient standard is met. A good example is the U.S. Clean Air Act. The federal government sets ambient standards for six criteria pollutants in a region. If a region is in violation, they must come up with a plan to attain compliance. Emission standards o Regulates the level of emissions allowed o Examples Emissions rates (pounds of SO2 per hour) Concentration (ppm of biochemical oxygen demand (BOD) in wastewater)

Total quantity of a pollutant Residuals per unit of output (SO2 per kWh of electricity) Residual content per unit of output (sulfur content of coal) Percentage removal of pollutant (90% of SO2 scrubbed) o Note that emissions standards do not guarantee a specific ambient level of pollution Weather conditions affect the concentrations Human behavior affects pollution levels Technology standards o Require polluters to use certain technologies, practices, or techniques. o Whereas emissions standards require polluters to meet a goal for the level of pollution, but give the polluter freedom to choose the technology used, technology standards require a specific technology. o Examples Until 1990, electric utilities were required to install scrubbers with 90% efficiency ratings. U.S. requires catalytic converters in autos The 1972 Water Pollution Control Act Amendments set a goal of zero discharges by 1985. Used technology based effluent standards (TBES) EPA determines the best practicable technology and sets standards assuming that firms are using that standard. For enforcement, polluters must have a discharge permit issued by an EPA-backed state permitting program. o Often, as in the Clean Air Act, the government mandates that the Best Available Control Technology (BACT) be used. However, BACT is often not clearly defined.

V. Economic Analysis of Standards

Setting the standard o The first question, of course, is deciding at what level to set the standard. o Efficiency calls for setting the standard where MAC = MD In practice, it is difficult for the regulator to know MAC and MD o Alternative guides to setting regulation: Zero-risk Requires protecting everyone, no matter how sensitive, from damage. Appropriate for cases in which there is a threshold. o Allow reasonably small damages Raises the question of how to decide what is reasonable. Should abatement costs be considered? The recent Supreme Court case (American Trucking Company v. Browner) raises this question. Should standards be applied uniformly?

Federal standards apply throughout the U.S. Is it appropriate to have uniform standards across regions? Are the needs of rural and urban areas similar? If marginal damages differ across regions, a uniform standard cannot be efficient in both jurisdictions. However, having different standards increases costs to the government. Note in this case the issue is variation in the marginal damage function. o Should standards be the same across firms? Efficiency is achieved when MAC is equal across firms, which wont happen with uniform standards unless the MAC curves are the same. By efficiency, we mean that any given abatement level is achieved at the lowest cost possible. Vintage differentiated regulation can help address differences in MAC from command and control. However, it raises additional issues, particularly with dynamic investment decisions. o With vintage differentiated regulation (VDR), standards depend on the entry date of each unit. Newer units face more restrictive regulations. Older units are often exempt (grandfathering) o Examples Clean Air Act New Source Review Emissions standards for automobiles Clean Water Act effluent limits for water treatment plants o Why use VDR? Efficiency Costs are lower for newer units Relates to equimarginal principle Holding all plants to the same standard is not cost effective In principle, CAC could mimic an efficient standard if each plants regulations varied depending on MAC. However, this is hard to observe. If MAC correlates with vintage, using vintage (which is easily observed) to differentiate regulations makes CAC more efficient. Equity Rules aren't changed in midstream Politics Easier to pass regulations if don't harm existing firms Potential for economic rent for existing firms if VDR makes entry into the market harder. o Effect on investment Firms invest if NPV of benefits (net of O&M costs) > cost of investment VDR makes investment more costly Both initial costs and O&M costs higher As a result, investment falls, and capital is kept longer

In extreme cases, VDR could lead to more emissions in the short run, as older, less efficient equipment is kept longer than before. Note that market structure matters The effects will be largest in competitive markets. Plants facing rate of return regulation find it easier to pass costs on to customers. For example, studying the effect of NOX trading markets, Fowlie (2007) finds that plants in deregulated markets are less likely to invest in capital intensive equipment. Evidence: Automobiles VDR for automobiles: Emission standards Safety standards Fuel economy regulations Note that retrofitting older vehicles would be very expensive and hard to enforce. Thus VDR makes sense for automobiles. Evidence from 1981 emissions standards increase Using price elasticities and estimates of cost increases from new emission standards, Gruenspecht found that VDR led to a 2-4% decrease in sales in the first 5 years after new standards Effect on emissions CO emissions up 1%/year for first 4 years HC emissions up 2%/year for first 5 years NOX emissions fell immediately Note that, by 1990, CO emissions down 5.3%, and HC emissions down 16% Alternatives: Pay people to scrap old vehicles. Greunspecht suggests a $250 bounty for scrapping cars more than 15 years old. Gas taxes instead of fuel economy standards Evidence: New Source Review New Source Review (NSR) is part of the Clean Air Act Regulations apply only to new sources However, existing sources that make major modifications must also comply. Several studies find that NSR lowers investment, thus extending the life of power plants. NSR can also raise the cost of operating newer plants. As a result, newer plants may be idled first. Stanton (1993) found that plants with weaker regulations were used more intensely. NSR can discourage investment at older plants. However, Wolfram and Bushnell (2008) find that this effect is small.

They find that NSR reduces capital expenditures at existing plants, but they find no change in operating costs, fuel efficiency, or emissions. Incentive to do innovation o Command and control provides little incentive to innovate o There are incentives to avoid the costs of regulation. However, there are no incentives to exceed the level of regulation. For example, even after costs are lower, firms have no need to control more pollution. Additional regulation would be needed. Enforcement (We won't discuss this in class, but I've put this here for completeness. The details are covered in the text.)
o

In the U.S., self-monitoring is often used. Firms keep their own records on emissions, and are subject to surprise audits. Even when regulations are set at the federal level, enforcement is often left to local governments. For example, water discharge permits are given out by state agencies with EPA backing. Enforceability helps determine which types of standards are appropriate. For non-point sources, technology standards (e.g. requiring or prohibiting certain methods) is easier to enforce than direct monitoring of emissions.

Emissions Fees
I. Pigouvian Taxation

Pigouvian tax A tax levied on each unit of a polluters output in an amount equal to the marginal damage that it inflicts at the efficient level of production. o The goal is to set the tax so that the polluter incorporates the social cost. Note that Pigouvian taxation is a second-best solution, as weve taxed the output. It would be better to tax the pollution directly. Modern economic solutions to pollution, which we will discuss later, take this into account. o It is a second-best solution because, although the level of output is correct for the technology being used, the firm doesnt have the correct incentives to use the appropriate technology (e.g. pollution control, more efficient machines, etc.) because there is no price placed on pollution. o However, there may be times when this is the best we can do. For example, we cannot measure the actual emissions from cars, so we instead tax gasoline consumption, since pollution is a by-product of gasoline consumption. The Pigouvian tax works by internalizing the cost of the externality. We can do the same thing with a subsidy. o In this case, the opportunity cost of polluting is losing the subsidy.

Types of subsidies: An abatement equipment subsidy would pay a firm for adopting a specific abatement technology. A per unit subsidy pays a firm for each unit of pollution reduced below some predetermined level. Problems with subsidies Very different distributional effects The polluter receives money from the government, rather than paying Firms may enter market, so that total pollution increases Need to raise taxes to pay for subsidies Ethics? Should we have to pay to avoid pollution? Subsidies are often politically motivated, and can be difficult to remove when no longer needed.

II. Emission Fees

Recall that the problem with externalities is that they are not reflected in prices. o The government can rectify the problem by setting a price for pollution. o The goal is to set the fee so that the polluter incorporates the social cost. If MAC is known, simply set the fee equal to MAC at the optimal level of pollution. o The firm will find it beneficial to abate up to this point, since abating is cheaper than paying the fee. o After this point, paying the tax is cheaper than abatement, so no further abatement occurs. o Note that since MAC = MD at the optimal level, the firm is taking into account the value of the damage it is doing. o If MAC is unknown, the fee should be based on the expected value (the best guess of MAC). The main advantage of emissions fees is that, when there is more than one polluter, they achieve a given level of pollution control at the lowest possible cost. o Thus, economists say that emissions fees are an efficient environmental policy. o An efficient solution is found when the marginal abatement costs are equal across all firms. At this point, there is no way to shift abatement responsibilities among the firms and achieve a lower total cost. However, the cost to each individual firm is greater, since the firms pay both abatement costs and the fees. Thus, emissions fees are politically unpopular. You may download a spreadsheet with the numbers I used in class by clicking here. Another potential advantage of fees over CAC is that fees encourage innovation. o Once youve met a CAC regulation, you have little incentive to do better. o However, if you lower your MAC, you can abate more, and pay less in fees. See, for example, figure 12-7 in Field.

Emissions Fees

the disadvantages of emission fees. Issues raised included: o Uncertainty Compared to command and control, emission fees provide more certainty on costs, but less certainty on the final level of emissions. o Geographically-varying damage Market-based policies guarantee an overall goal, but they don't guarantee which firms will reduce and which firms won't. If firms near an urban center choose to pay the fee rather than reduce emissions, damages may remain high. Varying the fee based on potential damages can help address this. o Distributional issues Concerns about equity might make some environmental taxes politically unpopular. For example, lower income families spend more of their income on gasoline, making a gas tax a regressive. While all policies raise the possibility of costs being passed on to consumers, there are more costs to be passed on here, as firms pay both for abatement and the fee for the remaining units of pollution. o Monitoring costs To charge a fee per unit of pollution, all pollution must be monitored and measured. o Flexibility Compared to regulations that mandate a specific abatement technique, fees are more flexible In addition, if economic conditions change, fees allow responses to adjust to new conditions.

I. Implementation Issues

We began class by discussing practical issues for implementing an environmental tax. o Note that most U.S. taxes that affect the environment were not implemented primarily for environmental reasons. Other concerns, such as raising revenue, tend to be more important. Questions for designing an environmental excise tax: 1. What is to be taxed? That is, what is the tax base. May be direct (e.g. CFCs, emissions), or indirect (e.g. gasoline) 2. Who is to be taxed? Here were focusing on administration, not ultimate incidence. Recall that economic incidence is independent of legal incidence. Ideally, we would like to tax users directly. However, it can be difficult to know the users.

For example, to prevent non-point pollution, we may tax fertilizer, since we cant tax each farmer for his or her individual contribution. Also, there may be many users. For CFCs, it was easier to tax production than tax each user. 3. What tax rate to impose? This is where most of the economic analysis comes in. Problems: Knowing MD is difficult. Therefore it is hard to set the tax correctly. Uncertainty could be bad if mistakes costly. As we'll see in a couple of weeks, in these cases (e.g. threshold effects), regulation may be better. Although we also might not know about MAC, taxes may help us learn about the MAC of firms. They will choose to pay the tax when tax < MAC. 4. Are their ancillary policy goals? Taxes are not enacted in a policy vacuum. Multiple goals often conflict. Common conflict: revenue vs. abatement Taxes are a source of revenue. If an environmental tax is successful, it lowers emissions, thus lowering the tax base. Therefore, if revenues are important, we might adjust the rate. What should be done with the revenue? Economists have noted that these taxes not only help the environment, but that they can also improve economic efficiency. The theoretical argument is as follows: Most taxes cause distortions in the economy. However, environmental taxes correct a distortion. Since the revenues from environmental taxes can be used to lower other taxes, the economy benefits. However, most economists find these effects to be small. The main reason is that the tax base for environmental taxes is much smaller than the tax base for more general taxes, such as income taxes. Thus, even a large environmental tax only raises enough revenue to allow a small cut in a more general tax. II. Examples

Although many taxes may affect the environment, frequent use of tax policies directly focused on environmental issues began in the 1970s (e.g energy conservation tax credits, gas guzzler tax) Environmental excise taxes generally fall in one of three categories:

1. User fees These are taxes designed to raise revenue for a specific purpose. Usually, the purpose is a public good. Examples: Excise taxes on sport fishing equipment and motorboat fuels fund programs such as fish hatcheries designed to ensure the survival of fish species. The federal gas tax is essentially a user fee, since revenues go into the Highway Trust Fund. Note that reducing pollution is not a goal of the federal gas tax. 2. Insurance premia in mandated risk pooling Taxes collected by an industry to pay for liability against possible damages. Examples: Oil Spill Liability Trust Fund tax of 5 cents per barrel of oil Superfund tax on chemical companies. Revenue pays for cleanup of future hazardous waste sites. 3. Pigouvian taxes Few environmental taxes explicitly aim to internalize external damages. Some that do include: Tax on ozone-depleting chemicals Passed in 1989 to help eliminate chlorofluorocarbons (CFCs) A result of US support of the Montreal Protocol. Production of CFCs is taxed per pound 1990 level was $1.37 per pound The amount has increased over time Gas guzzler tax on vehicles below a certain MPG Applies only to cars (not trucks) Applies to cars below 22.5 MPG Ranges from $1,000 to $7,700, depending on fuel efficiency Examples of environmental taxes (source: EPA Report: The United States Experience with Economic Incentives for Protecting the Environment, chapter 4) o Emission fees in the South Coast Air Quality Management District (SCAQMD) in Southern California Facilities exceeding their allowable emissions limits must pay fees o Solid waste disposal fees, such as stickers for trash Reduce weight of trash in landfills, but can also lead to more illegal dumping. Lesson: beware of unintended consequences. o Several states have taxes for hard to dispose of items (tires, oil, refrigerators) o In response to suburban sprawl, several New Jersey counties assess open space taxes 1 to 5 cents per $100 assessed value of property Money must go to trust funds to be used for purchasing open space, developing parks, etc.

Examples of subsidies in U.S. (source: EPA Report: The United States Experience with Economic Incentives for Protecting the Environment, chapter 7) o Income tax credits in Colorado for conservation easements o Brownfield development grants (also include changes in liability law) For example, in New Jersey, developers get a 10 year property tax exemption for remediating a site in accordance with state standards and return it to industrial or commercial use. o Loans for pollution control expenses The state of California issues tax-exempt bonds to provide low interest loans to small businesses for pollution control devices & solid waste recovery projects. o Farming and land preservation The Conservation Reserve Program, part of the U.S. Food Security Act of 1985 (1985 Farm Bill) pays farmers for taking environmentally sensitive land out of cultivation.- The Conservation Reserve Program, part of the U.S. Food Security Act of 1985 (1985 Farm Bill) pays farmers for taking environmentally sensitive land out of cultivation. o Tax credits for alternative fuels Available for alternative fueled cars Available for renewable energy (1.6 cents per kWh in 1995 for biomass and wind) Estimated cost to government $970 million o Income tax credits for easements in Colorado Income tax credits are available for 50% of the fair-market value of the easement, up to $375,000. For those with less taxable income, such as ranchers, there are two options: If there is a budget surplus, the credit is refundable. If there is no surplus, the credits can be sold to buyers who have more taxable income. The credits sell for 80-85 cents per dollar of credit. There is little regulation of the valuation of easements Some transactions have been found to be fraudulent Colorado doesn't audit the credits. This is left to the Internal Revenue Service. Impact Total acres protected: 2000: 350,000 2005: almost 1 million Cost to government in lost revenue: 2001: $2.3 million 2005: $85.1 million Proponents argue that, despite lost tax revenue, the preserved open space raises the value of neighboring land. o Note that all these subsidies have a cost: there is lost tax revenue. Developing country examples

(from Randall Bluffstone, Environmental Taxes in Developing and Transitional Economies, Public Finance and Management 2003, 3(1), 143-175).
o

China Has used emissions fees for air and water since the early 1980s. Uses a two-tiered system Charge a basic fee per unit, plus a pollution levy for industrial air or water pollution that exceeds annual limits. Firms are only charged for emissions of their worst pollutant. Fees are assessed at national levels, but implemented by local environmental protection boards. Rates vary by locality. Highest in richer urban areas. Uses self reporting by firms, with spot checks by local environmental officials. 80% of air pollution revenues to go pollution reduction programs. Central and Eastern Europe Many Central and Eastern European countries, as well as the former Soviet Union, implemented environmental taxes in the late 1970s and early 1980s. Goal: to give incentives for state owned enterprises to reduce emissions. Rates typically not market based (e.g. considering marginal abatement costs), but rather based on complex engineering formulas. Could lead to odd results. For example, prior to 1999, Lithuania charged several million dollars per ton of SO2. Not surprisingly, non-payment was an issue. Malaysia Uses a licensing fee for water pollution from palm oil production. Uses a two-tiered system Emissions above concentration standards are charged 10 times more. Firms have used a variety of techniques to reduce wastewater emissions, suggesting that there may be differences in abatement costs across firms. Columbia Began water effluent charges in 1974 Rates increased in 1993 Prior to this, rates were low, and typically only covered administrative costs. The new rates vary by region, and are linked to damages. Rates change every 6 months! Revenues go to local environmental agencies. Nationally, waste water emissions are estimated to have fallen by about 20-25% in response to the fees.

III. Comparison of Policies: NOX Charges in France and Sweden

The article by Millock and Sterner looks at two fee systems for NOX emissions, in France and Sweden. o Both build off existing command and control systems. o Both driven by the United Nations Economic Commission for European Convention on Long-Range Transboundary Air Pollution, which sets specific NOX limits to be met by 2010. Based on these goals, the EU set national emission ceilings in 2000. France needed 54% reduction from 1990 levels Sweden needed 56% reduction from 1990 levels o Potential technologies: Changes in combustion process Fuel switching Selective catalytic reduction for industry, catalytic converters for cars o Measurement of emissions important to properly calculate fees. Sweden o Acid rain important issue, because Swedish bedrock has little calcium, and thus does not provide buffering capacity. Goal: reduce NOX emissions by 30% by 1995, compared to 1980 levels. o Uses a refunded emissions payment (REP) for industrial boilers producing 25 GWh/yr Tax of 40 SEK/kg ($4,000/ton) NOX Rate based on engineering estimates of abatement costs, which ranged from 3 to 84 SEK/ton. Covers approximately 250 plants with 365 boilers o What happens to revenues? After administrative costs, remainder of charge is returned to plants Returned in proportion to output of useful energy Thus, if cleaner than average, can get back more than pay in Note that this can avoid one political drawback of fees, in that firms avoid needing to pay twice (for what they abate and for what they emit). The cost is that the government can not use the revenue for other purposes. However, the tax would be unlikely to pass without it. o Monitoring and compliance Units self-report annual emissions, using real-time monitoring Main costs to firms were for monitoring and abatement Average abatement costs between 12-25 SEK/kg Monitoring costs about 2-3 SEK/kg (about $250/ton) Note this is greater than the charge itself in France, suggesting that real-time monitoring would not make sense there. For about of firms, there was a net loss after refund. For another , a net gain after refund. Administrative costs at plant about 1.2 SEK/kg Low administrative costs for government: just 0.5% of revenues o Effect of the tax Emissions target not met until 1998

The problem was mobile sources, where emissions only fell by 13% Stationary sources, covered by the tax, fell by more than 50% Comparison of plants across countries US EPA compared similar plants (coal-fired plants with selective catalytic reduction, or SCR) in different countries. Swedish plants have half the emission rates of plants elsewhere In fact, emission rates are below the Swedish standard Note that the emissions charge gives plants incentives to do better than the standard. Differences across industries Figure 5-2 shows that some industries gained, while other lost Pulp & paper hurt the most, while energy gained the most Introduction of tax closely linked to European Community policy on acid rain. Applies to all units with 20 MW power capacity Higher threshold than the Swedish program Tax has consistently been lower than Sweden: 1990: 150 FF/ton NOX ($23/ton) 1995: 180 FF/ton NOX 1998: 250 FF/ton NOX ($38/ton) What happens to revenues? Pays for subsidies for abatement measures Compliance No fine for failure to pay Government said not needed, because there was 97% compliance Emissions data self-reported, but easily checked against fuel consumption data Administrative costs 6% of total tax revenue Effect of the tax Government reports emission reduced by 27,000 tons Subsidies for new abatement technologies less effective Many just accelerated what would have happened anyway Because of low tax rate, French policy is similar to a command and control policy Firms need to comply with national standards, but there is little incentive for continual improvement Note that average abatement costs in Sweden about 1/3 to 1/2 of charge, or $1,333 - $2,000/ton. Clearly, the French charge does not induce additional abatement at such costs.

France
o o o o

o o

IV. Greening the Tax Code

Note that taxes can not only correct externalities, but can also provide incentives that exacerbate problems. Often these come from tax expenditures (e.g. deductions). Hanson and Sandalow provide examples.

Percent depletion allowance Oil and gas producers can deduct a fixed percentage of gross income each year. This was established in 19090 to stimulate domestic production. Is it still needed? Repealing would save $900 million over five years. o Similarly, oil and gas companies can treat exploration and development as an expense, rather than depreciating over a number of years. Repealing would save $17 billion over five years. Encourages greater extraction of virgin materials, discourages recycling. o Sport utility vehicle deduction Businesses can deduct $25,000 of purchase of large vehicles (> 6,000 lbs) in the first year. No expensing allowed for light vehicles. Originally intended for large equipment (e.g. construction) Repealing would save $700 million over five years. Hanson and Sandalow also propose new green taxes. How would these work? o Water pollution tax Tax BOD in discharges from industry and wastewater treatment plants. Already monitor BOD, so administrative costs would be low. o Nitrogen fertilizer tax Tax fertilizer that leas to nutrient overload and dead zones in water. Currently, as much as 20% of fertilizer applied ends up in waterways. Tax would encourage more efficient use. o Carbon tax Tax fuels based on their carbon content. Encourages use of cleaner fuels (e.g. natural gas vs. coal).

Transferable Discharge Permits

I. Transferable Discharge Permits

Today's class was an exercise in permit trading. The class was divided into six firms, and each firm was given an initial allocation of permits. Based on the marginal abatement costs of each group, firms could offer to buy or sell permits. As expected, the final allocation lowered the variance of marginal abatement costs across firms. o Firms with high marginal abatement costs bought permits, so that they could pollute more, and firms with low marginal abatement costs sold permits, and polluted less. o Thus, the abatement was done by the firms who could achieve abatement at the lowest cost. Note that the permit market enabled us to learn the marginal cost of abatement. Firms did not know the costs of other firms, but we still reached an equilibrium in which costs were lower.

I have placed the results of the market exercise on a spreadsheet, along with some discussion. In the first round, the final marginal abatement costs were lower, and total abatement costs fell from $114 to $105. This is only $1 greater than the most efficient allocation of permits, as can be seen on the sheet labeled "Abatement Schedules". Round 2 had an efficient outcome, where the total abatement costs were as low as possible. You can download the spreadsheet by clicking below. In general, I've found that although costs usually fall in both rounds, we rarely get the lowest costs possible in round 1, as firms are still learning about the market. As a result, sometimes the first couple of trades take place at too high or too low a price. Nonetheless, even with imperfect information, trading generally leads the class to a more efficient outcome, as it did here. Results of permit trading exercise

II. How Trades Work

How permits work: 1. Government begins by setting the desired level of emissions (considering MAC and MD). Thus, like command and control policies, the government has control over the final amount of pollution. 2. Firms are issued permits to emit pollutants. Only the desired number of permits is issued. Thus, the quantity is assured. (e.g. if goal is 1000 tons of emissions, may give 100 firms permits for 10 tons each). 3. Firms can buy and sell permits. Firms with higher MAC will be willing to buy permits from firms with lower MAC. If the price paid is less than the MAC of the high-cost firm, it is better off. Similarly, if the price is greater than the MAC of the low-cost firm, it is better off. It can take the money it gets from selling the permit, use it to reduce pollution, and still have some left over. Such trades are possible until MAC is equal across firms. Thus, permit trading allows a given level of pollution control to be achieved for the least possible cost. Economists consider this least-cost solution to be efficient.

Transferable Discharge Permits

I. Permits: Implementation Issues

Initial allocation of permits o To begin a permit trading system, firms need to have permits to trade. The initial distribution can be done in several ways. The government can auction permits to highest bidder.

At least initially, additional trading shouldnt be needed, as permits go to firms willing to pay the most. Raises revenue for the government. Equal distribution among firms. May seem fairer, but what if firms are of different sizes. Historical emissions rates (more permits to bigger polluters). For example, if want to reduce pollution by 10%, give each firm permits equal to 90% of their current emissions. However, this penalizes early actors. Should firms that have already reduced get fewer permits? Also, combined systems are possible (e.g. hold back some permits for auction). o Note that firms will prefer getting the permits for free, as it gives them an additional asset. Auctioning permits makes the plan more like a tax. Thus, free distribution is more politically palatable. o Also, note that if the market is competitive, the market should yield an efficient solution no matter what the initial allocation. However, the effects on individual firms (e.g. who benefits by selling permits, versus needing to buy them), will be different. Establishing trading rules o For a market to work, transactions costs must be low. o However, at the same time, monitoring and enforcement will be necessary. Need to track both emissions and the number of permits each firm has. o Who should be able to participate? Should environmental groups or private individuals be able to buy permits and then not use them? The EU Emissions Trading Scheme (EU-ETS) is a good example of this. The EU faces an overall cap on carbon emissions from Kyoto. From this, the EU has specified the industries that will participate in trading (electric utilities, oil refineries, coke ovens, iron & steel, cement kilns, glass manufacturing, ceramics manufacturing, and the pulp and paper industry) Each country gets allowances based on its national cap in Kyoto. It faces two allocation decisions: 1. Total allowances are spread between the trading & nontrading sector 2. Permits in the trading sector must be allocated among individual firms. Issues: Difficult to make projects about price, because supply of permits was not known until countries made allocations between the trading and non-trading sectors. Also makes it hard for an individual country to control emissions from its trading sector, because market prices will be determined by the allocation decisions of all countries.

Countries have incentives to give more allowances to industries that trade goods, so that they do not have a competitive disadvantage with firms from other countries. Thus, likely to require more reductions from industries like electric utilities. In general, countries have given more allowances to the trading sector than expected, which has kept prices low. Geographic considerations: For some types of pollution (e.g. CO2), where it is emitted doesnt matter. For others, (e.g. carbon monoxide in a city), location does matter. A tax system would deal with this by charging higher fees in areas where pollution is a bigger concern. Ways for permit system to deal with geographic concerns: Ambient-based permit system: permits needed for pollution as measured at each receptor. E.g. a firm downwind might need to buy two permits from a firm upwind to be able to emit one unit of pollution. Limit trading to within regions Limits trades to areas where the emissions have the same effect. For example, New York state has tried to prohibit NY power plants from selling SO2 permits to plants in neighboring states. Clear Skies divides the country into two regions (East and West) for NOX trading However, such rules may prohibit some beneficial trades. Also, this limits competition, which might keep the market from working correctly. Examples of how the EPA has dealt with the locality issue Since the 1970s, the EPA has used permits to allow economic development in areas that failed to meet ambient air quality standards. 4 types of trades 0. Offsets included in 1977 CAA No new emission source can be located in areas where air quality standards are not met unless existing emissions are reduced at least as much. Trades can be internal or external. There have been over 10,000 offset trades. 90% in California 90% of offset trades have been within a firm. 1. Bubbles began in 1979 An imaginary bubble placed is over a plant Individual firms can sum all emissions from individual sources As long as the total doesnt exceed their limit, they are O.K. Both internal and external trades (external if bubble over a region) Bubbles set up between 1979 and 1986 saved $315 million.

2. Banking -- allowed in 1977 CAA Firms can save emissions reductions above and beyond permit requirements for future use in trading. Not used often (only 24 emissions banks by 1994) 3. Netting added to EPAs offset policy in 1980 Sources undergoing modification can avoid new source review if they can demonstrate plant-wide emissions dont increase significantly. Thus, firms can create new sources of emissions by reducing emissions elsewhere in plant. Internal trading only Most widely used of these programs Cost savings from being classified as a major source can range from $100,000 to $1 million per application. The Pennsylvania water trading article is a good example of offsets, and also raises other implementation issues The problem: nitrogen from manure and phosphorus from fertilizer lead to algae growth, with blocks sunlight and kills underwater grasses. Because the runoff is a non-point pollutant, it is difficult to track runoff from specific farms. Pollution control options: Farmers can adopt best practice methods (BPM), such as Barriers to contain runoff Planting crops year-round, so that storms don't wash soil away In urban areas, waste from sewage treatment plants can be filtered Pennsylvania has established a trading system where farmers receive credits for adopting BPM The value of the credit is based on a formula that considers the impact of improvement and the distance of the farm from Chesapeake Bay As discussed in class, a problem here is that the BPM must be certified by the state to receive credit. This provides uncertainty for farmers. A typical credit is worth $2 - $9, and reduces about 1.6 pounds of pollution. Developers need to purchase credits to offset new sewage treatment plants. A 100 house development would require about 700 credits. So far, there has been only one trade between farmers and developers in the first three months. Incentives for innovation The incentives for innovation are the same as with an emissions fee (see figure 13-4). Consider two cases: 0. A firm has enough permits to cover its pollution. The opportunity cost of polluting is that it cannot sell a permit. Thus, innovation not only lowers marginal abatement costs, but allows the firm to sell more permits. 1. A firm does not have enough permits to cover its pollution.

The opportunity cost of polluting is that the firm must buy a permit. Thus, innovation not only lowers marginal abatement costs, but saves the firm from the need to buy additional permits. Keep in mind that although individual firms have more incentives to develop technologies than under command and control, the total level of emissions need not fall, since the permits that are sold may be used by someone else. It may, however, allow new sources to come on-line, so that more output is produced for the same level of pollution. II. Examples of Permit Trading

Water pollution on Fox River in Wisconsin o Unsuccessful, only 1 trade in 6 years. o Firms were similar, located in clusters and were opposed to the plan. Lead trading o Background: lead caused health problems and doesnt work with catalytic converters. o During phase-out of lead from gasoline, refineries were allowed to trade lead credits o Timetable: 1982: credits established average level of 1.1 grams/gal allowed 1985: banking allowed average level of 0.5 grams/gal 1986: average level of 0.1 grams/gal 1987: terminated o How trades worked: Refineries that produced gasoline below the allowable limits received credits that they could sell to refineries that could not meet the limits. o Lead trading was very successful 50% of refineries participated in trading Reduced costs by at least 10% (several hundred million dollars) Because some refineries could not lower lead content quickly, not allowing trading would have increased the price of gasoline Mitigation banking o Before allowing a project that will damage wetlands, the Army Corps of Engineers requires that developers restore or create wetlands elsewhere. o Has led to a market for mitigation credits. Tradable green certificates o Used in Europe, Australia, Texas o The program begins with a target level for percentage of renewable energy use. o Producers get a certificate for each unit of renewable energy supplied to the grid. o Customers or distributors must show that they use at least that percentage of renewable energy. They do this by purchasing permits. Since producers of renewable energy sell the permits, they are compensated for the extra cost of producing renewable energy.

The Regional Greenhouse Gas Initiative (RGGI) includes 10 northeastern states that have committed to reducing CO2 emissions from power plants, with a goal of capping emissions at current levels in 2009, and reducing emissions 10% by 2019. o The states are the 6 New England states, plus NY, NJ, MD, and DE. Originally proposed by Gov. George Pataki of NY o Allowed emissions are 188 million tons CO2 annually through 2014. This then falls 2.5% per year through 2018, for a total 10% reduction. o How it works States are apportioned their share of allowances from the overall cap. States set limits for carbon dioxide emissions, and can decide how many to auction versus freely allocate. Most states auction nearly all permits. Delaware auctions the fewest, starting at 60% in 2009. This percentage increases 8% per year until 100% in 2014. New York auctions 100%. Utilities bid for allowances, which can sold later. Covers power plants of 25 MW or more burning at least 50% fossil fuels. Auctions are held quarterly. Proceeds from permit auction will be used for energy-saving and renewable energy programs in each state. Sources may choose to reduce their emissions early and bank their excess allowances for future use. Compared to EU, RGGI is more restrictive about the number of outside credits plants can use to offset emissionsOnly 3.3% of a plants emissions, or about half of their obliged reductions, can be met using offsets. o Auction results Two "pre-compliance" auctions were held: 9/25/2008 & 12/17/2008 9/25 price: $3.07/ton CO2 12/17 price: $3.38/ton CO2 Most recent auction was 12/2/2009, with a price of $2.05/ton CO2 Prices are low because emissions cap overestimated current CO2 output, which fell between 2005 & 2006. Estimated 2007 emissions were 172.4 million tons CO2, although cap for 2009 is 188 million tons CO2. As a result, carbon futures prices have also fallen (e.g. on Chicago Board of Trade). Nonetheless, all allowances have been sold above the floor price of $1.86/ton. Because of these low prices, early predictions are that the program will raise consumers monthly bills by no more than $1.

Uncertainty in Environmental Economics

Uncertainty is a problem for all policy decisions. o Often dealt with using expected values.

Given this, what makes uncertainty different for environmental problems? 1. Uncertainty for environmental problems is highly non-linear Damages may be barely noticeable for low levels, but become severe above some uncertain threshold, or tipping point. For policy, this is important because we dont want to go beyond the tipping point. However, the tipping point itself is often uncertain. Uncertainties may interact. For example, the benefits of climate change reductions depend on: Expected GHG levels without abatement (BAU) How rapidly CHG concentrations grow in response to emissions Effect of concentrations on temperatures Economic impact of higher temperatures Similarly, cost of abatement may be low for low levels of abatement, but be very high for total abatement (e.g. climate change) While these may be well-known for some pollutants (e.g. SO2, NOX), for others, costs are uncertain For climate change, the level of tax needed to meet an emissions target depends on how responsive energy use is to prices. There are decent estimates for short-run. However, for the long-run, this is more challenging. Our ability to substitute in the long-run depends on how well alternative energy sources substitute for fossil fuels. This is highly dependent on technological change. It may be that large reductions are very costly. 2. Irreversibilities Two types of irreversibilites: Environmental damage often irreversible. Cant undo cutting down a virgin forest or extinction. Takes many years to remove carbon emissions Thus, adopting a policy now has a sunk benefit (e.g. a negative opportunity cost) Suggests traditional cost-benefit analysis biased against policy adoption. Abatement efforts involve sunk costs Often in the form of capital expenditures However, may just be lost expenditures Thus, there is an pportunity cost to adopting a policy now, rather than waiting for more information This works in the opposite direction, suggesting that traditional cost-benefit analysis biased for policy adoption. 3. Long time horizons Makes results very sensitive to the choice of discount rate.

We will discuss this in the cost-benefit section of the course. Policy implications (we'll discuss these more next Monday) o Policy timing Should we wait until we learn more, or should we act quickly? When do irreversibilities matter? Only if there is uncertainty. If we know with certainty how future generations will value the changes, we can incorporate that into our cost-benefit analysis. Only affects current decisions if it would constrain future choices under plausible conditions. E.g. if, in the future, we would want negative carbon emissions to reverse climate change, we might emit less today, since negative emissions in the future are impossible. One key here is long-lived effects In a good news scenario, we can roll back policy However, in a bad news scenario, we cannot correct the situation. Policy intensity o Choice depends not just on expected values, but on variance o That is, on the shape of the curves (what Pindyck calls convexity) Choice of policy instrument IV.

The Role of Uncertainty: Choice of Policy Instruments The diagrams drawn in class were intended to show how the regulated level of emissions varies from the optimal level of emissions when taxes or command and control regulations are used and the marginal abatement costs (MAC) are uncertain. You may download a copy of the handout by clicking here. The key points were: o When the MDF is uncertain, both command and control and taxes result in the same errors. Thus, policy makers have to decide how to deal with uncertainty (e.g. base policy on expected values, try to avoid the worst outcome, etc.), but the results will not depend on the policy chosen. o When MAC are uncertain, policy matters: With regulation, the regulated level of emissions is too high if MAC are higher than expected (the regulation is too strong), and too low if MAC are lower than expected (the regulation is too weak). With fees, the resulting level of emissions is too low if MAC are higher than expected (because the tax is too low), and too high if MAC are lower than expected (the tax is too high). We will continue with this analysis

The Role of Uncertainty: Theory (continued)

Using the results we covered Monday, we began class deriving rules for the desired policy based on the slope of the marginal damage function (MDF): o When MDF is steeper, mistakes with quantity are more costly. Thus, regulation is a better option. o When MDF is flatter, a tax is better, because the tax represents MDF well. A third option is to combine permits with a safety valve. o One problem with the above analysis is that fees are rarely used. o Permits are more practical politically, but restrict quantity even if costs are high. o This proposal combines both features. o Implementation: Along with emissions targets, the government sets a trigger price (e.g. $25 per ton of carbon). If the price of permits rose above the trigger, the government would sell extra permits to maintain this price. o Political analysis Conservatives view it as a tax. Environmentalists fear stringent standards will not be met.

Voluntary Environmental Compliance

I. What is Voluntary Compliance?

Policy makers and environmentalists have paid increasing attention to voluntary environmental programs, in which firms improve environmental performance beyond what is required by regulation. Key questions include: o Why do firms participate? o Do programs increase environmental protection above and beyond what would have happened anyway? o What is the economic impact of voluntary programs? Types of voluntary programs: 1. Unilateral commitments: business-led environmental programs. Here, industry is the source of action. Examples: Chemical Manufacturers Association Responsible Care Program McDonalds replacing Styrofoam containers General Electrics commitment to reduce CO2 emissions Note that GEs CEO thinks this will be profitable. Why? GE hopes to be a technology leader if regulations are put in place. Perceived demand for green technologies as India and China develop.

Here, we discussed what it means to be "green" Is it just being cleaner than your competitor, or should certain standards be met? Does this lead to a need to regulate "green" claims (potentially pushing us from a firm-led program to a governmentled program, such as the product labels used in Europe)? 2. Public voluntary schemes: Participating firms agree to standards developed by public bodies, such as the EPA. Here, the public agency moves first, and tries to induce firms to comply. Used more in Europe than the United States Examples: EPA 33/50 program Green Lights program 3. Negotiated agreements: typically agreements of firms and government. May be legally binding. Unlike the other two, both firms and government are active participants. Again, used more in Europe Examples: French agreement on end-of-life vehicles Swedish program for packaging Here, compliance was required by the Swedish Ordinance on Producer Responsibility for Packaging, which set goals to be met by 1997. The agreement was established by industry representatives to control the system of collection, re-use and recycling. Registered companies pay a fee in return for using the REPAs recycling and recovery system. EPA Common Sense Initiative Example: Chemical Manufacturers Association Responsible Care Program o Example of unilateral commitment o Started by the Canadian Chemical Producers Association in 1985, in response to the 1984 Union Carbide storage tank leak in Bhopal, India. US Chemical Manufacturers Association and British Chemical Industries Association began similar programs in 1989. o Motivation: To restore public trust in the industry To minimize government regulations in response to the accident o How it works Participants of British program agree to six guiding principles All 200 existing firms signed, and new firms must agree to the principles to join the British Chemical Industries Association However, little public data are available for public to monitor.

Contrast with US, where TRI makes it possible to monitor firm activity Example: EPA 33/50 Program for Reducing Toxic Chemical Emissions o Example of public voluntary scheme o Began in 1991 o Goal: voluntary scheme to get firms to reduce emissions of 17 key toxic chemicals 33% by 1992, and 50% by 1995 Measured relative to a 1988 baseline o How it works EPA provides publicity and limited technical assistance No penalties for failure TRI data makes public monitoring possible o Results: From 1991 to 1994, emissions of these 17 chemicals fell 42%, compared to 22% for other TRI chemicals Raises question of substitution: if other chemicals are used more, what are the risks? Example: Agreement on End-of-Life Vehicles o Example of negotiated agreement o Began in France in 1993 o Participants: French Ministry of Industry French Ministry of Environment 2 French car manufacturers 12 French automobile importers 8 trade associations for dismantlers, recyclers, etc. o Goal: have total weight of vehicle destined for landfill be less than 15% of original weight by 2002. After 2002, new models must allow for 90% recovery, reuse, or recycling. o No specific enforcement mechanism, but there was a threat of legislation if program did not succeed. In particular, Germany was considering regulation on end of life vehicles (ELV) in 1992. The French effort was an attempt to pre-empt this. o Monitoring Initially done by agreement participants Concern over lack of public scrutiny led to NGOs attending some meetings as observers. Little useful data have been generated. II. Why Do Firms Voluntarily Comply? To better understand voluntary agreements, we first need to understand what the motives of participants are. That is, why do firms comply, even when not compelled by regulation? Alternative theories o Improving corporate responsibility

Since pollution is waste, some argue that reducing pollution reduces costs. Commonly cited example is 3Ms Pollution Prevention Pays program. Relates to the Porter hypothesis, which we will discuss later in the course. Key questions: Why didnt firms do this before? Are there new opportunities, or did firms previously ignore profitable options? Responding to green consumers and investors Some consumers are willing to pay a premium for green products. Voluntary compliance may simply be a reaction to this demand. Allows firms to differentiate products Examples: Dolphin-friendly tuna Organic foods McDonalds replacing Styrofoam packaging Socially responsible mutual funds Avoid investing in irresponsible companies. Reduces supply of capital to such companies Note that voluntary behavior in response to consumer demand need not be socially optimal. The products involve externalities. Thus, consumer preferences are unlikely to completely internalize these externalities. It is important to differentiate marketing from performance. That is, would these projects be commercially viable anyway? Strategic behavior to avoid binding regulation Preempting tougher regulation Here, the goal is to anticipate future regulation, and provide enough improvement so that the government deems such regulation unnecessary. In such cases, does this improve welfare? Presumably better than doing nothing, but is it better than having regulation put in place? May be if it avoids the costs of a legislative fight. If consumers didnt feel better off after voluntary agreement, they could still fight for legislation to do more. Thus, transactions costs are a key here. Concerns: Corporate lobbying may reduce effectiveness of consumer and/or environmental groups

Regulators objectives may be different than welfare-maximization. Legislators may delegate too much power to regulators with voluntary agreements. Theory predicts voluntary action more likely when threat of regulation greater (e.g. because citizens more active, or costs of information low). Tests of this hypothesis are discussed later. Weakening forthcoming regulation Even if regulation cannot be avoided, voluntary actions might help shape the regulations that are ultimately set. For example, industry involvement in the EU carbon trading scheme led to freely allocated permits and favorable allocations for industry. Example ARCO introduced reformulated gasoline between 1989 & 19991. While this garnered positive publicity for ARCO, it also slowed down a policy push for vehicles using methanol fuel. Note that this also gives ARCO a first mover advantage. Reducing regulatory monitoring Here, the goal is to reduce the stringency with which the firm is treated by regulators. Commit to higher level of compliance with existing regulations in return for lower monitoring rate or laxer permitting scrutiny from regulators. EPAs Project XL is an example of this. Firms that participate receive relief from EPA enforcement in acknowledgement of ongoing efforts to improve environmental performance. Intel received a waiver to allow for routine changes without permits as long as total emissions did not exceed a plant-wide cap. Caps are more stringent than federal standards. The XL program has had few participants, as they are still vulnerable to third-party lawsuits. Works best if the firm can make an irreversible investment (e.g. abatement equipment) that commits them to better performance. Government can then shift its attention to competitors. Is this welfare-enhancing? Depends on penalties for violations. If penalties are too high, firms have incentive to overinvest in compliance.

Encouraging anticompetitive regulations If fixed cost of compliance are high, small firms may be forced out of the industry, reducing competition. Note that costs of compliance are often uncertain, particularly to regulators. A voluntary effort by a large firm may convince regulators the costs are achievable. DuPonts voluntary CFC reductions are an example. DuPont had developed substitutes for CFCs. By showing they could achieve reductions voluntarily, they convinced regulators that a phaseout of CFCs was feasible.

Valuing Environmental Benefits: Revealed Preference Approaches

I. What is Value?

Up to now, we have found the optimal level of pollution by considering the marginal damage function, which we have taken as given. Now, we discuss how to place a value on environmental amenities. o We can use data from firms to measure MAC. o Measuring benefits is more difficult, because they are not typically part of market transactions. We can measure damages directly, by looking at damages and the value of what is lost, or we can infer damages indirectly from the behavior of individuals. To begin, consider what makes up the value of environmental amenities: o Use value -- the benefits people get from direct use of a good. For most consumer goods, this is what we care about. For environmental goods, this can include: The value of recreation at a site The value of open land near a home The value from better health The value of ecological services provided (e.g. by a wetland) o Non-use value For environmental goods, not all value is use value. For example, we have expressed concern in class about drilling for oil in the Alaskan National Wildlife Refuge. How many in the class are opposed to drilling? How many in the class will actually go there? Examples of non-use value: Option value the amount a person would be willing to pay to preserve the option of being able to experience a particular environmental amenity in the future. Even if you wont go to the Grand Canyon this year, preserving it may have value to you so that you can visit in the future.

Existence value a willingness to pay simply to help preserve the existence of some environmental amenity. Protection of endangered species is an example. Bequest value a willingness to pay to leave behind environmental quality for future generations. Stewardship value a value placed on preserving the environment not for human use, but rather to maintain the health of the environment for all living organisms. To measure value, economists focus on willingness to pay. o We can see willingness to pay from a demand curve. o It is the area under the demand curve. Recall that the difference between what consumers actually pay and the actual price is the consumer surplus. Willingness to pay includes actual expenditures and consumer surplus. Thus, simply using a direct measure of expenditures ignores the consumer surplus, and underestimates the value. o We need to consider the same issues for costs and supply curves. For example, changes in policy may lead to changes in costs (see figure 7.1 in the text). Thus, we look at changes in producer surplus. o Putting these together, the net value is the sum of consumer surplus and producer surplus. o Since policy analysis should focus on marginal analysis, we want to ask how these change as we have an incremental change in pollution.

II. Revealed Preference Approaches

Economists typically use one of two approaches to measure the benefits of environmental quality: 1. Revealed preference approach infer the value of environmental goods from other market transactions Note that revealed preference approaches get at use values, but not non-use values. 2. Stated preference techniques ask individuals hypothetical questions about their willingness to pay. A. Aversion Costs

Note that, in reaction to environmental harms, people may undergo expenses to remedy the problem. o Examples: Filters for drinking water Air conditioners so that windows can remain closed Medication to mask symptoms of health effects Shrubbery to hide a polluted neighboring site By studying how much people spend on averting expenditures, we can estimate the benefits they would receive if the harm were removed.

Example: a study of Los Angeles in 1986 found that people would pay $0.97 to avoid shortness of breath to $23.87 to avoid chest tightness. B. Travel Cost Method

The travel cost method looks at how far visitors travel to come to a site. By placing a value on the cost of travel, we can infer the value of the site. o The travel cost includes both direct costs (e.g. airfare) and indirect costs (e.g. the opportunity cost of travel time). o We can infer the value of a change in quality by looking at demand during different days (e.g. in different types of weather). o Example: Economists surveyed 826 tourists in Florida: How many days did you use the beach? Where are you from? How much did it cost to get there? What is the total length of your stay? From this, they could derive a demand curve Results Average tourist spent 4.7 days at the beach Average daily expense = $85 To calculate consumer surplus, they need to know the slope of the demand curve. Using regression analysis, they found that demand for beach days was inelastic: A 10% increase in price leads to a 1.5% decrease in time on the beach. Using this, they could plot the demand curve and found that the consumer surplus for 4.7 days = $179 (average of $38/day) To get a total value of beaches, they note that tourists spend 70 million tourist days/year on the beach => total value of Florida beaches = $2.37 billion. Potential problems with the travel cost method 1. What is the opportunity cost of time? 2. Only measures value of those that use the amenity. We need to account for substitutes. For example, do all beach users in Florida come for the beaches, or are they there for other reasons (e.g. Disney)? 3. Quality is not always measured. 4. Sampling bias in surveys. C. Hedonic Pricing Techniques: Housing

Hedonic pricing techniques look at the value that people place on the attributes of a good. o That is, it assumes that people don't value a house itself, but rather the features of a house (e.g. number of rooms, location, is there a fireplace) o One such feature is environmental quality. Using regression analysis, we can find the correlation between housing prices and environmental quality in an area.

Most studies find an elasticity of housing prices with respect to pollution that is around 0.1. o That is, a 1% decrease in pollution leads to a 0.1% increase in housing prices. o A sample study is "The Determinants of Residential Property Values with Special Reference to Air Pollution," By Ronald G. Ridker and John A. Henning (Review of Economics and Statistics, May 1967). If you are interested in reading the article, you can click on the title to be taken to it in J-Stor. D. Hedonic Pricing Techniques: Wages

Another application of hedonics is with wages. Examples: 1. People will choose to live in cities with positive characteristics. Differences in wages can be seen as the value of these characteristics. 2. People need to be compensated to be willing to take riskier jobs. Differences in wages represent the value of a human life.

Valuing Environmental Benefits: Stated Preference Approaches

I. How Much is a Life Worth?

The most controversial aspect of cost-benefit analysis is placing a value on human life. o What is the value? Is it merely the opportunity cost (e.g. foregone wages)? Are there other values (perhaps non-market values) that need to be considered? o Concepts of the value of a life The most commonly used value is the value of a statistical life. We don't know who will die, but we expect someone will. The value of environmental protection is lessening the risk of someone dying. Note that specific deaths capture the attention of individuals. However, that is not what a statistical life focuses on. We are valuing changes in the probability that a random individual will die, by asking what is the willingness to pay for changes in risk. For policy, this is the most appropriate measure, because policy does not prevent death, but rather changes the probability that death will occur. Can also be put into annual figures: value of a statistical life year (VSLY) Contrast this with the optimal insurance and compensation of accident victims.

Here, things such as the opportunity cost of foregone wages and medical expenses make sense, since now we are focusing on a specific loss. However, does this mean that rich people are more valuable? The make whole principle is similar. However, after death or serious injury, is it possible? Viscusi suggests that in some cases, it simply isnt possible to restore welfare to the level that it was before. Risk-risk analysis Finally, Viscusi notes that tradeoffs among risks should be considered. For example, chemicals used to make flame retardant pajamas for children increased the risk of cancer. How to measure the value of a life: 1. Revealed preference approaches Expenditures to reduce risk For example, how much more will people spend for a car with airbags, or for bottled drinking water? Challenge: separating value assigned to changing risk to other characteristics (e.g. other features of the car, or better taste for bottled water) Hedonic wage approach People need to be compensated to be willing to take riskier jobs. Differences in wages represent the value of a human life. To do, we regress wages on job characteristics, worker characteristics, and risk Suppose a study revealed workers would accept an additional 1 in 1000 (0.1%) annual risk of dying for $1000/year in wages. What is the value of life? If 1000 people took the job, one would be expected to die. The value needed to pay these people is $1 million ($1,000 x 1,000) Therefore, the willingness to be compensated for one life is $1 million (= $1,000 * 1,000) Note that this assumes peoples preferences are linear. Does $1,000 for 1/1000 => $100 for 1/10,000 Issues: Requires people to have perfect information about risks, and to be able to evaluate this information properly. Do people take risks knowingly and willingly? True locally (that is for marginal changes), but for large changes in risks might not be appropriate. 2. Ask people: stated preference methods Estimates of the value of a life in the United States range from $900,000 - $20.2 million, and cluster around $6-$7 million (all figures 2005 dollars).

The E.P.A. uses $6.9 million (2008 dollars) F.D.A. uses $5 million DOT uses $3 million Issues for valuing life o How do we deal with different groups? Reducing risk extends ones life expectancy. Given this, should we place different weights on the lives of children? Should we place less value on protecting the elderly? In some cases (e.g. air pollution), it is these high risk groups who are most affected by a policy. OMB guidelines advise against adjusting VSL for age. o Extrapolating results across groups can be a problem. Many VSL studies look at job risk in middle-aged men. However, the young and old tend to be most vulnerable to pollution. Estimating VSL for children particularly difficult. Cannot ask them directly Parents often willing to pay more to reduce risk to children than to themselves. o Should the value of human life vary by income? Studies such as those focusing on lost income will place more weight on high-income lives. Given this, should we pay more to reduce risk if high-income people are affected? For example, the average income of air travelers is higher than for the population as a whole. Moreover, the costs of increased air safety will be passed on to these passengers via higher ticket prices. Given this, should the standards for air safety be higher? The EPA does not adjust VSL for incomes within a cross-section (e.g. doesnt say VSL higher for richer neighborhoods than for poorer ones). However, EPA adjusts VSL upward over time to account for higher incomes. o What about people in other countries? These approaches suggest a lower value for lives in developing countries. Does ability to pay matter? Does it matter if we are considering a plan to be paid for by the Chinese government versus one sponsored by the U.S. to aid China? If China is paying, their ability to pay constrains what they can do. o Control matters Smoking vs. workplace safety Driving vs. airline safety o Values can change over time Not having lifeboats for lower class passengers on the Titanic would be unacceptable now. Note that this may mean values also vary across culture or country.

o o o

II.

Stated Preference Approaches: Contingent Valuation So far, we have studied revealed preference approaches to valuing environmental amenities. Those approaches look at actual market transactions to infer value. Unfortunately, there aren't always market transactions that can serve this purpose. For example, how do we value protection of endangered species. In these cases, economists simply ask people for their valuation. The most common technique is contingent valuation (CV). Policy background o The Comprehensive Environmental Response, Compensation, and Liability Act of 1980 (CERCLA) gave government the right to sue for damages to natural resources for which they are trustees. In 1986, the Department of Interior said nonuse (largely existence) values were recoverable if use values were not measurable. Sanctioned the use of CV In 1989, a federal court of appeals directed the Department of the Interior to redraft its regulations with equal weight on use and nonuse values. o After the Exxon Valdez crash in March of 1989, DOI regulations meant that Exxon would have to pay for non-use damages. CV estimate of damages: $3 billion o In response to the Exxon Valdez, Congress passed the Oil Pollution Act of 1990. In addition to reducing the likelihood of future spills, the act provided for damage recovery. The Department of Commerce was to draw up regulations on damage assessment. The Dept., through the National Oceanic and Atmospheric Administration (NOAA), assembled a blue ribbon panel to evaluate CV. Question to panel: Is the CV method capable of providing estimates of lost nonuse or existence values that are reliable enough to be used in natural resource damage assessments? How contingent valuation works: 1. Ask people their willingness to pay (WTP) to bring about a specific environmental improvement. The problem must be described carefully. It must be specific. Types of questions: open-ended -- ask respondents for maximum WTP. close-ended -- ask respondents whether they are WTP a certain amount. This amount is varied across respondents. bidding games -- ask respondents whether they are WTP a certain amount. If they say yes, ask them about a higher amount, until you find the highest amount they are WTP. 2. A payment mechanism is specified. The mechanism must be believable (e.g. increase sewer fees to improve water quality). It should not be controversial (e.g. property taxes).

Because the survey shouldn't serve as a referendum on the type of payment mechanism chosen. 3. Information about the respondent is gathered. E.g. income, age, education... Allows verification of results, estimation of income elasticities, etc. Examples of Results: o First CV: Davis (1963) found the value of outdoor recreation opportunities in the Maine woods to be between $1 and $2 per day. o Valuing clean air: WTP to avoid health problems: $10 per asthma day reduced WTP for better visibility: A 10 percent increase in visibility has been valued as between $7 and $101. Increased visibility at the Grand Canyon has been valued between $5 and $10. Valuing cleaner water: WTP to improve water from "boatable" to "fishable": $12.30 WTP to improve water from "boatable" to "swimmable": $29.60 III.

Evaluating Contingent Valuation Do surveys lead to true values? (content validity) o Hausman and Diamond point out flaws in surveys: Protest zeros (usually thrown out by researchers) Respondents may be expressing value for a class of goods (e.g. expressing general preference for higher environmental quality). Warm glow from supporting good causes. People may be reacting to news (e.g. oil spills) Do people care about the resource (e.g. Alaskan coast) or preventing the action (e.g. clean up oil spills)? o Hannemann and the NOAA panel (as mentioned in Portney) note that these problems can be dealt with by careful construction of surveys. All surveys have potential problems. The key is careful construction. For example, verbal protocol analysis can determine why people answer as they do. Verbal protocol analysis asks respondents think out loud. When they do this, they often say irrelevant things: If we all do our part it wont cost much. Pretesting of surveys can avoid these problems. Survey samples should be based on probability sampling, not simply stopping people on the street. The question should involve a firm sense of commitment. o Hannemann responds to the critique that the survey creates values: People often dont make decisions until the time they are forced to do so. Why should CV be different?

As we discussed in class, lots of issues come and go from the political landscape. He points out that the key question is whether the results are stable, which he says they are (that is, they are consistent across studies). Are the results of CV consistent with theory? (construct validity) o Are the results similar to results from revealed preference approaches? Sellar et al. (1985) used CV and travel cost at 4 Texas lakes. At 3 of the 4, the 95% confidence intervals overlapped. However, means did vary (TC $102.09 at one lake, CV $35.21, but within 95% interval). Brookshire et al. (1982) compare hedonics and CV on air quality. o Are the results similar and consistent with theory? Values increase when income increases. Environmentalists WTP more. Hausman and Diamond argue that would happen despite flaws in survey. For example, if measuring general taste for environmental quality. o Problems noted by Hausman and Diamond (to be covered Wednesday): Anchoring Whether ask for value of a seal and then a whale, or a whale and then a seal, the first value given is around $100. Therefore, WTP depends on order of questions. Embedding Suppose 1 group asked about X A second group is asked about Y A third group is asked about X + Y WTP of (1) + (2) should equal (3) However, (1) + (2) often > 3 Example: Desgouves (1982) WTP to protect migratory birds from oil waste pools by covering them with nets. Said will protect 2,000, 20,000, or 200,000 birds from dying Values didnt increase Do people care about the numbers saved, or just the act of saving birds? Hausman says only the final number of birds should matter. Hannemann replies that economic theory says that other things can enter utility (e.g. altruism) Hausman and Diamond argue that embedding is evidence of a warm glow. Its not just that people value X or Y, but that people feel good about supporting the environment.

When you ask the values individually, the warm glow is added up twice. Thus, adding up results of individual studies leads to double counting of benefits, making aggregation of results for policy difficult. Defenses of embedding Hannemann argues that this results from diminishing return. Protecting one lake is valuable. Protecting a second lake isnt as important. Income effects could be a factor. However, WTP is a small percentage of income. There could be large substitution effects between X and Y

Benefit-Cost Analysis
I. Introduction to Benefit-Cost Analysis

Goal: Maximize total net benefits (= total benefits - total costs). Benefit-cost analysis calculates the costs and benefits of a project and finds the total net benefits. Note that some costs and benefits can be observed directly from market data. Others will need to be inferred from data. Steps to benefit-cost analysis 1. Specify clearly the project or program. For environmental economics, this is usually a physical project such as a dam or wastewater treatment plant, or a regulatory program, such as pollution control standards. 2. Determine quantitatively the inputs and outputs of the program. Can be difficult for example, general equilibrium effects. Also, it is important to distinguish between transfers of resources due to substitution and the creation of new resources. For example, jobs created by a project should normally not be included as a benefit. Jobs created are a transfer of resources. If the project wasn't done, the workers could have been used elsewhere. 3. Estimate the social costs and benefits of these inputs and outputs. 4. Compare these costs and benefits. Here, one can also include other considerations, such as equity. For students who would like a review of the nuts and bolts of benefit-cost analysis, you can view my lecture notes from the topic in my managerial economics class at: http://classes.maxwell.syr.edu/pa723/723lect.html. The last set of lectures cover benefitcost analysis. Dealing with Uncertainty

II.

The first step in dealing with uncertainty is risk assessment. Risk has two components: 1. stochastic depends on chance 2. systematic depends on circumstances (e.g. a smoker is more likely to get cancer) In addition, assessing risk involves two concerns: 1. the probability of an event occurring 2. how serious the event will be Risk assessment o First, we focus on finding the probability of an event occurring. Historical data Risk can be determined by looking at past records. However, it is important to be aware of changes that occur over time. For example, increased safety features reduce the risk of death from auto accidents. This is a change in systemic risk. New technologies Component analysis is often used to assess the risks of new technologies. Problem: components may be related. Risk by analogy Often, time lags make perceiving risk difficult. For example, cancer may be caused after exposure to a toxin, but only after many years. As a result, studies on animals are often used to extrapolate human risks. Problems: Animals are exposed to unrealistically high doses of toxins in the laboratory. Need to extrapolate risk of humans from low exposure from calculated risk based on high exposure. Physiology of animals and humans may be different. The risk may be different for different people. A complication for policy makers is that, even after risk assessment is complete, people have a hard time perceiving risk accurately. o Risk assessments are hard to understand, as they typically involve low probability events. o Control is important Consider that most people worry more about air travel than auto travel, although the likelihood of dying in a car accident is greater. o People may pay more to avoid unpleasant deaths. Example: what should the standard for ammonium perchlorate in groundwater be? o The example over the ammonium perchlorate standard shows how different criteria for risk assessment suggest different answers.

EPA:

Proposes a standard of 1 part per billion (ppb) They look at sensitive populations, including the risk to fetuses. They also make use of laboratory studies. o Pentagon: Proposes a standard of 200 ppb. Note that this standard would release the Pentagon from most cleanup responsibility. They base their figure on a study of exposure to human adults. o Thus, one study focuses on more sensitive populations, whereas the other focuses on exposure to a typical person. Both can be justified. Which is correct? Once risk has been assessed, policy makers face several alternatives for using the information: o Benefit cost analysis For BCA, pieces of information needed to deal with risk include: The risk probability The government often uses conservative estimates (e.g. 95% percentile). The population exposed For example, Superfund regulations consider possible future populations on a site. The value of a life o Avoid upper bound of risks Government agencies, such as EPA, often use conservative risk estimates (e.g. 95% percentiles) However, this adds up If use 95% percentiles for several estimates, actual percentile is above 99% (.95 x .95 x .95) Consider two chemical hazards: Chemical A poses a known risk of 2 in 100,000 Chemical B is uncertain 9 out of 10 scientists believe no risk 1 out of 10 believe risk is 6 in 100,000 Government policy says risk of B is greater, since focuses on upper bound that is, greatest potential risk. However, chemical A has a higher expected value of risk. o Cost-Effectiveness analysis Rather than compare costs and benefits, simply show that the agency has adopted the cheapest way possible to achieve its goal. Takes the policy objective as worthwhile. We can than ask if the costs justify the benefits received, without needing to place a dollar value on the benefits. o Risk-risk analysis Compare risk after regulation to risk before.

Notes that regulation will affect behaviors, and could even increase risk. Issues: Substitution risks If substances that replace banned substances are also risky, net gain from banning the substance is not as great as it seems. Risk of other economic behavior For example, what is the additional risk of industrial accidents from workers who manufacture scrubbers for power plants? Opportunity costs of diverted resources What do we give up in other expenditures (e.g. health care) by spending more on regulation?

Benefit-Cost Analysis
I. Discounting

The costs and benefits we've discussed often occur at different times. To compare them fairly, it is important to discount costs and benefits that occur in the future. o The idea is to compare a flow of benefits and costs into a single value. The present value of a future amount of money is the maximum amount you would be willing to pay today for the right to receive that money in the future. o Present value accounts for the opportunity cost of not investing the money elsewhere. o Example: You have $100 now If you put it in the bank, you will get 5% interest Next year, that money is worth (1 + 0.05)x100 = $105 2 After two years, it is worth (1.05)(1.05)(100) = (1.05) (100) = $110.25 o General rule: FV = future value, PV = present value, r = interest rate t FV = PV(1 + r) o As a result, you wouldnt give up $100 now for $100 next year, because you could invest the money and get $105 next year. The present value of $100 next year is the most you would give up today to get $100 next year FV => PV(1.05) = $100(1.05) PV = FV/r = 100/1.05 = $95.24 o General rule PV = FV/(1 + r) o For a stream of payments: 2 t PV = x + X/(1+r) + X/(1+r) + + X/(1+r) o For payments forever: PV = X/r o Rule of 70

To get the number of years needed to double an investment, divide 70 by 100 times the growth rate. Example: Invested @10%, money doubles every 7 years [= 70/(0.1x100)] To proceed, we need to know what value to use for r. This is the discount rate. The discount rate reflects the relative value a person places on future consumption compared to current consumption. o Lower values show a greater preference for future consumption. If your discount rate is greater than the interest rate, you will be willing to borrow money. A high discount rate says that current consumption is important to you. If your discount rate is lower than the interest rate, you will be willing to loan money. A low discount rate says that future consumption is important to you. o Since the market interest rate reflects an equilibrium of lenders and borrowers, we can use the market interest rate as a measure of the discount rate. Why the discount rate matters o Discounting affects the value placed on future benefits and costs. o Higher discount rates place less importance on future returns. o Note, for example, how this is a particular problem for long-term problems such as climate change. A very low discount rate suggests we would give up virtually all consumption today to protect the future. However, a higher discount rate suggests very distant benefits have little weight in decision making. There are several market interest rates. Which should we use? o Typically, economists use a risk-free rate. Investors looking for a safe return invest in U.S. Treasury bills. Thus, the return on T-bills is a measure of the nominal risk-free rate. To purchase assets that are riskier, investors need to be compensated with a higher rate of return. This additional return is known as a risk premium. Note also how the discount rate relates to economic growth theory o discount rate = pure rate of time preference + growth rate of income x elasticity of marginal utility for income The first term captures the relative weight placed on the future versus today Involves ethical judgments The second term acknowledges that, due to economic growth, we expect future generations to be richer If the marginal utility of income falls as we get richer, than additional money is less valuable when we are richer Might the social discount rate deviate from the market rate?

The above estimates use market data to determine the discount rate. Are their reasons to believe that the market rate is flawed? o Some economists argue that the opportunity cost of foregone future consumption might differ from the opportunity cost revealed in the markets. In this case, it might make sense to use a social discount rate which is lower than the rates observed in the marketplace. The social discount rate represents the willingness of society to trade off present and future consumption. If there are market failures, this may differ from discount rates observed from market behavior. Why might market rates not be appropriate? Long term projects involve benefits or costs for future generations. However, future generations are not represented in the market. People may be myopic, and thus not save sufficiently. There may also be other externalities that cause the market rate of return on investments to deviate from the social discount rate, such as positive externalities from research and development. Uncertainty may be a concern Therefore, risk aversion may justify using a lower discount rate. However, uncertainty is not an excuse to do nothing. What can be done? o Compromise view: Use the market rate for the first 30 years of a project, and a lower social discount rate afterwards. The intuition is that, for the first 30 years or so, the market rate is a reasonable guide to individual preferences. However, since the market rate may ignore future generations, a lower rate is used for benefits and costs affecting future generations. o Alternatively, analysis can be done without discounting. Note that this is a controversial view in economics. Most mainstream economists would disagree with a zero discount rate.

II. Concepts of Cost

I've included my complete set of notes on estimating costs below. Note that we did not cover all of these issues in class today. Important concerns: o Establishing the baseline We want to compare costs with regulation versus without regulation, not before and after regulation. Even without regulation, we expect some things to change over time. o Distributional issues Note that costs will often be focused on a few individuals (e.g. affected firms or communities).

Benefits are more likely to affect a wider range of people. Thus, equity concerns will be an issue.

Types of costs 1. Opportunity costs the value of the best forgone opportunity. It is what we give up by using a resource for this use, rather than the next best alternative use. Example: an opportunity cost of going to school is foregone salary. Important to distinguish between costs and transfers. 2. Environmental costs because most regulations focus on a single pollutant, regulating one pollutant may increase the use of another pollutant. Example: using scrubbers to clean SO2 emissions leaves behind a sludge that must be disposed of. 3. Enforcement costs 4. Implicit costs nonmonetary costs of inconvenience, time searching for substitutes, lost product variety, etc. Few studies include implicit costs

III. Estimating Costs

Costs of a single facility o For some projects, such as building a wastewater treatment plant, we need only consider the cost of a single facility. o In other cases, we may need to aggregate the costs per facility to get industry totals. o Estimation depends primarily on engineering estimates. o After regulations are in place, survey data is available. o Types of costs: Fixed cost of building a facility (considered capital costs by accountants) Expenditures for plant, equipment, construction, and making process changes. Wont change as the level of abatement changes. Variable costs of operation (operating and maintenance costs) Costs incurred in the operation and maintenance of abatement processes, such as labor, energy, and R&D. Will vary with the level of abatement. The present value of costs over the life of a facility are needed. What about the cost of lost jobs, lost income to merchants, etc.? o If the workers will be absorbed into the economy, these are not a cost to society. o Certainly the workers themselves are affected, but the resources are still being employed. o For larger scale regulations (e.g. an entire industry), we cannot assume that resources now unused will be used elsewhere. o The Department of Labor surveys firms for the reasons for mass layoffs. Less than 0.1% are attributed to environmental regulations! Abatement cost data

The US spent 121.8 billion on Pollution Abatement and Control Expenditures (PACE) in 1994 About 1.7% of GDP. How it is spent 96.6% on abatement 8.0% personal spending 62.9% business 25.6% government 1.8% on monitoring 1.6% on R&D A 1999 pilot survey focused on PACE in manufacturing industries: $5.8 billion on PACE capital expenditures $11.9 billion on PACE operating and maintenance expenditures.

IV. How Accurate are Cost Estimates?

Sources of error o One difficulty is that errors can come in many ways. o The most obvious is incorrectly estimating the costs of control. o However, even if the cost of control is estimated correctly, predictions about emissions levels, number of plants, etc. can also be wrong. See examples on pages 11-12 of Harrington et al. Evaluation of estimates o Harrington et al. study 25 estimates of the cost of regulation. o They compare pre-regulation estimates to actual costs after the regulation is in place. o Label an estimate as accurate if it within 25% higher or lower than the actual costs. o Results: Unable to Determine 2

Accurate Overestimate Underestimate Quantity Reduction Unit Pollution Reduction Cost Total Cost 10 7 5 9 12 12 4 6 2

Costs more likely to be overestimated Discussion of results: The two underestimates were for rather minor regulations: EPA aldicarb ban and OSHAs powdered platform regulation. EPA and OSHA tend to overestimate reductions, but not per unit costs. As a result, total costs are overestimated. State and foreign agencies were more likely to overestimate per unit costs.

For market-based polices, seven of the eight estimates overestimated costs! Note that this is where we would expect technological innovation to be most important. Why do errors occur? Many estimates ignore the possibility of technological innovation. Regulators have an obligation to identify a means of complying with the regulation, which usually means considering current technologies. Future technologies are much harder to predict. Quantity errors: misestimating baseline emissions However, keep in mind that this also means that benefits are wrong. For example, overestimating emissions reductions overestimates costs, but also overstates the benefits. Regulations may change during the public comment period. Thus, cost estimates arent for the final regulation. Estimates may focus on maximum values, rather than means. Especially if rely on industry for data. Asymmetric correction of errors. Firms are likely to bring underestimates to the attention of regulators. There is no similar group with strong incentives to bring overestimates to the attention of regulators.

Growth and the Environment

I. The Macroeconomy

Up to now, the analysis in class has focuses on microeconomic issues. o That is, how do individual actors behave, and what incentives affect this behavior? o This analysis allows us to find the optimal level of various activities. However, all of these activities affect the economy as a whole. o Consider, for example, the effect of higher energy prices on the economy during the 1970s. Also, we need to know how economic growth affects the environment. o Is economic growth good or bad for the environment? To begin, we must distinguish between economic growth and economic development. o Growth refers to increases in aggregate level of output. o Development refers to increases in per-capita output. Thus, population is important. We begin with some macroeconomic basics. o The traditional measure of macroeconomic performance is Gross Domestic Product (GDP). o GDP is the sum of the money values of all final goods and services produced in the domestic economy during a year.

Does not include sales of intermediate goods and services. Only includes work done in the United States. Only market activity is included. Since there isnt a market for most environmental goods, they are not included!!! In contrast, defensive expenditures (e.g. repairs after an auto accident) are included. Problems that result: The value of environmental amenities is not included in GDP. Depleting a stock of natural resources (e.g. oil, minerals, forests) increases GDP, since it results in new sales. For comparison, depleting the capital stock hurts, as capital wont be there for future generations (measured in net domestic product, or NDP). Similar reasoning should follow for natural resources. Example: Repetto recalculated Indonesias growth rate including natural capital. The growth rate is 7%. Recalculating to include the degradation of resources such as timber lowers the growth rate to 4%. Growth that comes from the consumption of capital, including natural capital, is not sustainable. Defensive expenditures are included. Defensive expenditures are expenditures made to eliminate, mitigate, or avoid damages caused by other economic activity. For example, after a car accident, the costs to repair the car, provide medical treatment to the victims, pay for lawyers, etc. all add to GDP. But, has the car accident really increased welfare? A study of Germany found that 10% of the countries GDP consisted of defensive expenditures.

II. How Does the Environment Affect the Economy

To see how the environment can affect the economy, we proceed with a more detailed model of the economy. Economists usually model output by using a production function. From this, we can derive a growth equation, as we discussed in class. The basic form of the equation is: o Rate of growth of output = A + a*rate of growth of capital + b*rate of growth of labor + c*rate of growth of energy + d*rate of growth of materials. In this model, A represents changes in technology. With this model, we can address two questions: 1. How does environmental policy affect economic growth?

Environmental regulations divert inputs from the production of output to other goals, such as reducing emissions. Since environmental benefits are not measured in GDP, they are not part of measured output. Thus, resources diverted to environmental protection cannot be considered in the equation. GDP falls, so economic growth, as traditionally measured, slows. In 1994 (the last year with available data), PACE were $121.8 billion, which was 1.7% of GDP Environmental regulations may also prohibit certain resources, such as timber, from being used at all. This lowers M. Environmental policy can affect technological change, which affects A. Whether this is good or bad for growth is a debated topic, which we discussed with the Porter hypothesis. 2. How does the environment benefit economic growth? There are positive effects of the environment on the economy. Even though most environmental benefits are not included in GDP, some benefits to have tangible dollar values. For example: Environmental resources are an input to production. If clean water is not available, it cannot be used. Environmental regulations that protect water thus benefit GDP. Environmental quality affects the quality of other inputs. For example, reduced air pollution makes agriculture more productive and provides healthier workers. Not only are healthy workers more productive, but health care is a large portion of GDP spending. If average costs are rising, reducing the number of sick people reduces the costs of treating them. Note that this raises the question of how to incorporate the environment into GDP. Well discuss that during the lecture on sustainable development. III. Income and the Environment

Case study: the environment in East Asia o The article from Environment magazine focuses on the evolution of environmental policy in three East Asian countries: Japan, South Korea, and China. o History of environmental policy in Japan Early concerns arose in 1868, in response to damage from copper mining. Government did not pay serious attention to the environment until after WWII. National economic development was the central governments top priority. Pollution was originally thought of as a local government problem. Rapid growth in the 1950s led to increases in pollution.

In response to disease outbreaks from water pollution, the National Diet passed two water quality laws in 1958. These are the first laws at the national level. International pressure played a role. Before hosting the Tokyo Olympics in 1964, there was international pressure to improve water quality in the Sumida River in Tokyo. Led to the creation of the Pollution Control Division of the Ministry of Health and Welfare (1964) & the Basic Law for Environmental Pollution Control (1967). More recently, policy focus shifted to global issues. Global warming became part of agenda in mid-80s. There is now more of a focus on quality of life issues. History of environmental policy in South Korea Not an issue until after Korean War During 1960s, Korea was under an authoritarian regime (President Park). Focus was on Five-Year Economic Development Plan of 1962 In 1963, Pollution Prevention Act passed, but it was ineffectual due to lack of resources. Rapid growth lead to greater water and air pollution in 1970s. Environmental Preservation Act passed in 1977. Set water quality standards in 1978, and SO2 standards in 1979. Reducing air pollution became a priority before the Seoul Olympic games in 1988. Government supplied lower sulfur oil beginning in 1981. Strengthened vehicle emission standards in 1987. Made clean fuels such as liquefied natural gas mandatory for large cities in 1988. History of environmental policy in China Under Mao Zedong (beginning in 1949), the government insisted pollution was a capitalist problem that did not exist in socialist countries. In 1972, water pollution began to attract the governments attention, so the government sent a representative to the 1972 Stockholm Conference on the Human Environment, and a national conference was held in 1973. Environmental protection was introduced into the Chinese Constitution in 1978. The Constitution stated that protecting the environment was the responsibility of the state. Environmental Protection Law passed in 1979. Included the responsible that polluters should be held responsible for pollution treatment, including a polluter pays fee system. Despite increased institutional attention, positive results for the environment are limited by a lack of resources. Convergence of policy In each country, economic growth came first, followed by a reaction to resulting environmental problems. International pressure played a role in each case.

The Olympics were important for Japan and South Korea, and are now important for China Note that for all three countries, attention first focused on water and air pollution on a sector by sector basis. Broader environmental regulation came later. Differences Latecomers had the advantage of learning from other nations' environmental policies. In Japan, local governments played a bigger role in policy. This makes it easier for citizens to influence the decision making process. Market forces played a bigger role in Japan Levels of democracy differ in each country.

IV. The Environmental Kuznets Curve

Empirical studies of the relationship between per capita income and pollution typically find one of three patterns: 1. For some problems (e.g. access to drinking water), income growth always leads to less of a problem. 2. For many pollutants (e.g. SO2), the level increases as per capita income begins to grow, but then falls as income continues to grow. (See Figure 19-2 in Field for graphs) Initially, growth leads to industrialization that causes pollution. As income grows more, the country becomes more willing to devote resources to pollution control. 3. For still other problems (e.g. CO2 emissions), the problem gets worse as income grows. How is the Environmental Kuznets Curve (EKC) tested? o Typically, studies look at cross-country data using Global Environmental Monitoring Systems data (GEMS). As noted by the earlier graphs, seems to only fit for a subset of pollutants. Results are sensitive to the specification. o Issues: Data availability is a problem. Current research focuses on criteria pollutants, because that is where the best data is available. What about toxins? As noted in class, results are sensitive to the inclusion of higher-order polynomial terms. More time series studies needed. Cross-section studies do not capture dynamics. Time series studies of Netherlands, Germany, the UK and the US find that economic growth has a positive effect on emissions of CO2, NOX, and SO2 (meaning emissions increase), although technological change may offset this.

Theoretical requirements for EKC: 0. Marginal utility of consumption falling or constant So that cost of giving up consumption falls as incomes are higher 1. Marginal disutility of pollution rises As problems get worse, it is more valuable to do something about them 2. Marginal damages rising 3. Marginal abatement costs rising It is expensive to do a lot. Implications of the EKC: o While EKC relationships have been observed using the data above, critics raise concerns that suggest other possibilities: Revised EKC: proposes that the curve could shift downward over time Might the peak be lower for newly developing countries, since they can use technologies first developed elsewhere? Race to the Bottom: promoted by globalization. Intuition: once everyones income rises, where does the pollution go? New Toxics Over time, countries shift away from traditional pollutants, but use more of pollutants that are more dangerous. o Note: there is little empirical evidence for the last two theories.

Discussion: The Environment in Developing Countries

I. The Environment in Developing Countries

Today's class was a discussion of the environment in developing countries. One of the main points was to discuss what was different about environmental problems in developing countries, as compared to developed countries. I have highlighted some of the major points of the discussion below, and have tried to remember as many of the points raised in class as possible. I apologize for any comments that have been left out. We began with a discussion of the Economist article on the Environmental Sustainability Index (ESI). o Unlike the environmental Kuznets curve literature we discussed last week, this article attempts to address causation -- what factors lead to better environmental quality. o Income is particularly important when there are immediate health effects. o It is less important when the impacts are long term, or when they do not directly impact human health. o Good governance is important. We continued our discussion by noting the causes of environmental problems in developing countries. Our list included: 1. Population growth 2. Industrialization

Developing countries are moving from agrarian societies to industrialized societies, which results in more pollution. 3. Increasing urbanization As industrialization occurs, people move to the cites, as that is where jobs are. Environmental problems become problems when people are close together, so that their actions affect others nearby. Also, infrastructure to support increased populations is lacking in many cities. 4. Weak governance Not only are regulations often weaker, but when they do exist enforcement limits compliance. Corruption and lack of democracy are also problems. 5. Lack of awareness/education Demand for environmental regulations depends on awareness of the problems 6. Poverty/pressures for economic growth As a result of lower incomes, developing countries place less weight on future considerations With lower incomes, current needs take precedence. This may lead to weaker regulation as an attempt to promote industrial development 7. Infrastructure Lower incomes often mean that infrastructure is inadequate 8. Poorly defined property rights 9. Technological constraints Technologies used may be less efficient than those used in developed countries, leading to more energy and resource use. 10. Different priorities The environmental problems of concern to developing countries may differ from those promoted by NGOs and developed countries providing aid. Water o Access to clean water is a major concern in developing countries. 90% of sewage in developing countries is discharged without treatment. o Causes of the problem The costs of the necessary infrastructure are high. The price of water is subsidized. As a result, infrastructure is allowed to decay and utilities are reluctant to connect new customers. Question: Although pricing water at its true cost is efficient, is it fair? One suggestion is to have a low base price for a basic unit of water, but to price additional access at a higher rate. Another alternative may be to price water correctly, but also provide subsidies to low-income people.

Water utilities may be bureaucratic, inefficient, and corrupt. However, these problems may continue with privatization, since the companies have monopoly power. o A possible solution to the subsidy problem is block pricing. Charging a lower rate for the first units of water used, but high rates to those that use more. Note that this requires infrastructure to monitor usage. Air pollution o The most important air pollution problems in developing countries are: 1. Indoor air pollution One-third of energy in developing countries comes from burning wood, crop residues, and animal wastes in stoves. 2. Lead emissions 3. Small particles o Note that subsidies are a major problem. Because electricity is subsidized, utilities are reluctant to hook up new customers. If few people have access to electricity, they turn to burning fuel in stoves for energy. o In addition, transportation policy is also important. Timber and deforestation o Problems caused by deforestation: Soil erosion Water quality and supply both decrease Loss of biodiversity Loss of animal habitat Carbon emissions increase Because forests serve as carbon sinks o Note that the last three are likely to be important to people in developed countries, but less so to developing countries. Developing countries are reluctant to devote resources to problems such as biodiversity or global warming, as the benefits are long-term. If people in developed countries want to protect these resources, they need to make it desirable for developed countries to do so. Demonstrating local benefits, such as reducing soil erosion, is important. o Reasons why deforestation is a problem: Pressures from population growth and migration Lack of income Exporting timber products provide income. Developing countries are more concerned with current consumption than future consumption. Sustainable policies are not desirable if they don't provide enough income in the present. Lack of other energy supplies leads to using timber as a fuel. Lack of property rights People who do not own the land have little incentive to preserve it.

Note that, despite laws in Brazil to discourage deforestation (e.g. limiting harvest to 20% of trees on the land), enforcement is difficult due to poorly defined property rights. Just 14% of privately owned land in the Amazon is backed by a secure title deed Most land is "acquired" by right of settlement Subsidies For example, Brazil offered tax breaks for development of forested land (such as for cattle ranching). Enforcement is also a problem because it is costly. Countries do not have the resources necessary to enforce existing environmental regulations. One estimate: 73% of forest products imported by the EU from Indonesia in 2008 were from illegally harvested timber Indonesia claims this is only 10% Decentralization makes enforcement difficult Local governments often resent the enforcement efforts of the national government Demand from developed countries o One solution that might help enforcement is labeling Using barcodes, logging firms can prove that timber has been harvested sustainably. This allows firms to differentiate themselves in market. Currently, the price premium is low: about 2-3% Also makes it possible for developed country consumers to differentiate. For example, developed countries can ban illegal timber for construction projects financed by government. Cannot ban all imports of uncertified timber due to WTO rules. However, legal logs are more expensive, because royalties must be paid to governments. Illegal logging cost governments $15 billion in 2002. Certification can be time consuming in developing countries. Land reform in Brazil is another potential solution Property rights for small plots would be given to apparent owners. The state would reclaim property rights for large plots Resource management (yet to be discussed) o Many developing countries have abundant natural resources, yet their populations are poor. Resource-rich countries grow more slowly than other poor countries, even after controlling for other variables. This paradox is known as Dutch Disease. Increase in natural resources (e.g. oil) leads to appreciation in domestic currency. This makes non-oil sectors less competitive on world market.

As a result, the oil sector dominates the economy. o Moreover, oil-rich countries do worse on issues such as child mortality, nutrition, and education. Oil employs few unskilled workers. Volatility of oil prices hurts poor the most, as they are unable to hedge risks. Ownership of resource concentrated. Thus, revenues pass through few people, making it easy to redirect funds. Oil revenues allow governments to keep taxes low. Thus, population has less incentive to demand change. o Possible solutions Set aside revenues when prices are high Alaskas fund that is redistributed to households is an example. However, easy to dip into fund for other purposes. Particularly a problem in developing countries, such as Zambia & Venezuela in 1970s. ExxonMobil has invested in an oil project in Chad and Cameroon in which funds are deposited in an offshore escrow account. An oversight committee evaluates proposed spending from the account. Ten percent are to be held in a trust fund for future generations. Eighty percent will be devoted to education, health and social services, rural development, infrastructure, and environmental and water resource management. However, there have been reports of Chad diverting funds from the account. Tony Blair proposes the Extractive Industries Transparency Initiative (EITI) Voluntary effort between governments and oil companies to promote transparency. From their website (http://www.eitransparency.org/): The EITI aims to ensure that the revenues from extractive industries contribute to sustainable development and poverty reduction. Over 20 countries have joined. Agriculture (we skipped this in class today, but I thought I'd include my notes on it) o Since WWII, agriculture in developed nations has become increasingly intensive, using more capital and fertilizer. o This system has spread to developing countries. Has succeeded in producing more abundant, less expensive food. o Costs of increased productivity Soil degradation Soil is compacted from machines being used on it. Water and wind erosion Depletion of minerals from overplanting and overgrazing. Pollution

Runoff from fertilizers and pesticides is a major non-point source pollution. Pesticide use has doubled over the past 30 years. Pesticides such as DDT are still used in developing countries. Even organic materials, such as manure, cause pollution problems. Water scarcity 40% of the worlds food comes from 5% of the agricultural land that is irrigated. Water is being pumped from the ground faster than it can be replenished. Inefficient use leads to waste of water. Part of the problem is that water is usually provided below cost. Biodiversity loss Intensive agriculture leads to species loss. For example, 13 million hectares of forest are lost to agriculture each year. 7,000 crop species are available for cultivation. 90% of the worlds food comes from 30 of them. Focus is on high-yielding, pest-resistant crops. o Can biotechnology be a solution? Growth of biotech crops has been rapid (44% increase in 1999). Most of the use is in the U.S. Crops are bred to resist herbicides or to be pest resistant. Note that, where used, herbicide applications per acre have fallen by 9% to 14%. However, total applications up because acreages have also expanded. Yields have not increased. Concerns: Access to seeds GM seeds can be designed so that they cannot be reused. Farmers would thus need to purchase new seeds each year. Reductions in biodiversity Genetic resistance will the surviving weeds be resistant to herbicides? Can the gene be transferred to other crops? Lack of consumer demand Some countries are reluctant to use GM seeds because they are afraid that consumers will not buy crops from their country. General lessons o Throughout these examples, we observed a couple of common problems Lack of property rights leads to overexploitation of resources.

Subsidies encourage inefficient use of resources, but, because of equity concerns, may be difficult to remove. Poverty is part of the problem. Many of these problems require large capital expenditures. The pressures of population growth often contribute.

Sustainable Development
I. What is Sustainable Development?

The Brundtland Report (1987), part of a World Commission on Environment and Development, proposed the following definition of sustainable development: o Sustainable development is development that meets the needs of the present without compromising the ability of future generations to meet their own needs. o Note that this definition is quite vague. How do we know what future generations will need? What are the needs of the current generation? Quotes from UNESCO report: o every generation should leave water, air, and soil resources as pure and unpolluted as when it came on earth. o each generation should leave undiminished all the species of animals it found existing on earth. These definitions imply no use of natural resources. Is this possible? If not, can you have a moral obligation to do something impossible? Solow's definition: an obligation to conduct ourselves so that we leave the future the option or the capacity to be as well off as we are. o Solow's definition is an example of weak sustainability. This is the approach taken by most neoclassical economists. Weak sustainability any loss of natural capital should be balanced out by creation of new capital of at least equal value. Assumes natural capital and man-made capital are substitutes. Herman Daly (Beyond Growth 1996): o For renewable resources: Keeping the annual offtake equal to the annual growth increment (sustainable yield) is equivalent to maintenance investment. o For nonrenewables: The general rule would be to deplete non-renewables at a rate equal to the development of renewable substitutes. o These can be summarized as: Never reduce the stock of natural capital below a level that generates a sustained yield unless good substitutes are currently available for the services generated. (Goodstein 1999) o These guidelines are examples of strong sustainability natural systems should be maintained whenever possible. Critical natural capital should be preserved under all circumstances. Assumes natural capital and man-made capital are compliments. This approach is taken by most ecological economists. Neoclassical vs. ecological economics o Neoclassical economists view man-made capital and natural capital as substitutes.

Neoclassical economists view sustainability as a need for dynamic efficiency. Thus, a measure of welfare needs to be maximized. o Ecological economists view man-made capital and natural capital as compliments. Note that Daly does acknowledge technological progress, but says that knowledge does not simply substitute capital for nature. Rather, he argues that it makes the uses more efficient. Thus, a key concern for ecological economists is the carrying capacity of the environment. Economics is driven by scarcity. The key question, according to Daly, is how big is the economic subsystem relative to the natural subsystem. Daly claims we are moving to a world in which natural capital is the limiting factor. As we noted in our discussion, clearly defining a sustainable path is difficult. o Three dimensions of sustainability are important: The existence of a positive sustainable level of welfare, The magnitude of the ultimate sustainable level of welfare vis vis current welfare levels, and The sensitivity of the future welfare to actions by previous generations. o Both intergenerational and within-generation equity are also issues Solow notes a paradox: Sustainability says that you should be thinking about poor people today, but helping poor people today requires more consumption, which hurts sustainability.

II. Will Sustainable Development Occur in Market Economies?

Valuing environmental benefits o In some cases, markets are beginning to appreciate the values provided by ecosystems o Good values depend both on good science and good economics Early studies often gave implausibly high values, leading to the work being discredited. Example: Panama Canal o The issue is that deforestation is harming the canal. o Water supply has been drying up Most freshwater stored upstream in Gatn Lake, an artificial lake created during the canals construction, where it accumulates from rainfall 52 million gallons of water is released into the ocean each time a ship uses the canal. Deforested slopes around the lake do not absorb rainwater. Thus, rain runs into the lake too quickly, causing it to overflow and run into the sea. This is important, because while the area gets 10 feet of rain per year (3x more than Seattle), most rain comes in the rainy season from May to December.

If the land were forested, the water would be absorbed and would flow into the lake more slowly. o Much of the deforestation has occurred since the 1950s, when a highway made the land accessible to loggers. Sediment and nutrients get into the canal Nutrients lead to growth of weeds Leads to expensive dredging o There have been some reductions in deforestation since the 1990s. Leading Panamanian bankers stopped financing cattle ranchers who cut down forests for pasture. When the canal was turned over to the Panamanian government in 1999, government agencies had incentive to protect the watershed. o Still, the agencies do not have enough money for thorough monitoring and enforcement, and reforestation needs to be done. o As a result, a private insurance firm, ForestRe has stepped in with a proposal The plan would have shipping companies that use the canal underwrite a 25 year bond to pay for reforestation. These companies would ask their big clients, such as WalMart and the Asian auto companies, to buy the bonds. These companies would get discounts on insurance premiums they currently pay to insure against closure of the canal. Other examples o Catskill Mountains, NY New York City gets its water from the Catskill Mountains watershed. In 1997, the city faced building a water filtration plant to clean water from the watershed. This had an up-front cost of $4-6 billion, plus $250 million/year operating costs. Instead, the city made payments to preserve the Catskill watershed Spent $250 million to buy land to prevent development Spends $100 million/year to farmers to minimize pollution. o Cape Town, South Africa Found that it was cheaper to restore the local watershed than to divert water supplies from elsewhere or build reservoirs. o Cauca Valley, Columbia Large agriculture producers pay fees for watershed management projects Note how the fees address the public goods problem o Costa Rica Hydro-electric producers, private customers, and the government contribute $57 million/year to protect a local watershed Services provided For hydro-electric producers: Stream-flow regulation Sediment retention Erosion control For private consumers

Irrigation Government Water supply for towns Maintain scenic beauty for recreation and ecotourism o A study found that natural pollination of insects raised productivity of one coffee farm by $60,000. o France Perrier-Vittel restored parts of a heavily farmed watershed and paid farmers to switch to organic farming to preserve the quality of some of its products. o Muthurajewela wetland sanctuary, Sri Lanka The World Conservation Union calculated the value of ecological services provided by the sanctuary at $8 million/year This amounts to $260,000 per square km. These services include: Cleaning sewage and waste Flood attenuation Support of downstream fisheries But, note that in this case, the article does not discuss how these benefits are supported financially. That is, who pays to preserve the wetland sanctuary. What to do with this information? o In some cases, there is a clear, well-defined, small number of beneficiaries (e.g. NYC). o In these cases, collective action and compensation is possible (e.g. Panama) o In other cases, the beneficiaries are more widespread (e.g. insect pollination) What can be done then? The article discusses cap and trade as a solution The Clean Development Mechanism for climate change fits here Provides a market for undertaking activities to reduce carbon emissions in developing countries Note that the principle of additionality is important: would the emissions reduction occur without the credit? Wetland mitigation trading is an example Developers who use wetlands in the U.S. must buy credits from a mitigation bank. Note, however, that unlike the above examples, this does not protect a specific wetland. Audits for resource-rich countries o These focus on providing information to allow market forces to work Like many other policies we've discussed, the effectiveness of the audits depends on the reactions of the developed world o The proposal would subject resource-rich countries to financial audits of the revenues and resulting spending from sales of natural resources

o o

While corrupt countries may not wish to be audited, the hope is that pressure from consumers would encourage multinational companies to insist on audits. Records of revenues raised would make it easier to save gains from boom years for later use. Harder to steal funds if there is a public record.

Sustainable Development/Trade and the Environment

I. How to Incorporate the Environment in Economic Analysis

As mentioned earlier, GDP does not give a complete picture of the effect of the environment on the economy, since it ignores the value of natural resources. Desired solution: adjust GDP to account for natural resources Green GDP. o First attempt: Nordhaus/Tobin (1972) o Adjusted GDP to account for amount of welfare-reducing environmental damage done by pollution. Measured as a portion of the difference in wages between wages in urban and rural areas. o They also make other adjustments to GDP: Add value of leisure Subtract capital purchases, add capital services used. Subtract non-beneficial spending, such as commuting to work. o End result: measure of economic welfare (MEW) MEW grew at 1/2 the rate of NNP from 1929-1965. China has begun to incorporate green GDP into its decision making o The motivation is that previous targets focused on economic growth, leading local officials to ignore environmental concerns Local leaders are evaluated based on an elaborate point system. Points are rewarded for meeting specified targets (e.g. a certain level of GDP growth). GDP growth is a veto target: failure to meet the target ensures that the cadre is considered underperforming. Has led to abuse of environment and human rights to meet GDP growth goals. o A pilot program in 10 regions began in February 2004 The hope is that a green GDP target will help local leaders focus on environmental concerns. The challenge is coming up with green GDP numbers One study said growth from 1980-2000 falls from 9.6%/yr to 6.8%/yr if green GDP used. However, the National Bureau of Statistics is skeptical about whether appropriate numbers can be devised. o Because of difficulties, abandoned the effort in 2006. o Instead, China will use what is known as a satellite system.

Separate accounts that try to integrate environmental and economic measures. Green accounting data will be provided alongside GDP data. Guidelines for satellite systems published by UN in 1993. o Satellite systems relate economic activity, measured in cash terms, to environmental magnitudes measured in physical units. E.g.: tons of CO2 emitted by each sector of the economy. Measures the effect of economy on environment, but doesnt adjust values. French President Nicholas Sarkozy recently appointed a commission to modify national accounts. The report noted three issues: 1. Problems with what GDP leaves out Depreciation of capital This could include both natural and physical capital Environmental benefits Government provided services are either imputed or left out, since there are no market values available. For example, private education has a market value, but the value of public education must be imputed The value of leisure time The report suggests that statistics should focus on household income, consumption, and wealth, rather than total production. It compared US and French GDP per capita using traditional measures and the suggested alternative. With traditional measures, French per capita GDP is 73% of US per capita GDP With alternative measures, French per capita GDP is 87% of US per capita GDP 2. Quality of life People in richer countries do not report being more "happy" The country of Bhutan includes happiness in its official statistics However, happiness is hard to quantify 3. Well-being of future generations In addition to natural and physical capital, human capital is also important. Finding a single measure to capture all of this will be difficult. How should Green GDP be measured? o The goal is to measure what is enjoyed or consumed. Traditional GDP uses quantities of goods and services, and prices of these. o Some resources, such as timber or minerals, have market values. However, most environmental benefits do not have market values. o The analog for the environment is ecosystem services. Ecosystem services arise from and depend on the broader sets of ecological components, processes, and functions but are different: they are the aspects of the ecosystem that society uses, consumes, or enjoys to experience those benefits. (p. 7 of Boyds article) o 5 principles to measuring:

1. Services are natures end products, not everything in nature. For example, we measure a car in GDP, not the steel, tires, leather, and workers used to produce the car. The value of the car embodies all of these other values. Thus, what we want for the ecosystem is what matters directly to people. Its not that other things arent relevant, but that they should be embodied in this value. 2. Ecosystem-services are benefit-specific. Flood protection from wetlands should be counted. They substitute for flood control. Wetlands are not services for the water quality they provide. Instead, this quality should be valued directly. Including it again as a wetland service would be double counting. 3. Counts what we can count, not what should be counted. Boyd argues this is what we do for other goods. We consider the price of the car, not the satisfaction from owning it. That is, we use price, rather than consumer surplus, as the value. 4. Ecosystem services should be ecological. Recreation itself is not a service. Things in nature that make recreation possible are. 5. Ecosystem services should be counted with greatest possible spatial and temporal resolution. People benefit in specific places and times. For example, when and where clean water are available matters. The amount of pollution will have different effects depending on location. May make placing dollar values on these difficult. Note that key to these principles is marginal analysis. The question isnt the cost of destroying all trees, but the cost of destroying one additional tree

II. The Benefits of Free Trade

Before discussing the links between trade and the environment, we first consider what benefits economists perceive from free trade. Goals of free trade: o To expand markets for goods for which a country has a comparative advantage in production, and to provide greater opportunities to procure goods for which the country has a comparative disadvantage. Free trade and comparative advantage o Comparative advantage a person has a comparative advantage over another if that person can produce that good at a lower opportunity cost than someone else.

Comparative advantage makes international trade beneficial, even between seemingly different countries. Ex: 2 goods: computers and TVs 2 countries: Japan and US Output per year of labor:

Computers TVs

U.S. Japan 50 10 50 40

In this example, the US has an absolute advantage in both goods. Its workers can make both more computers and more TVs than Japanese workers. Nonetheless, trade is beneficial: The U.S. has a comparative advantage in computers. To make 1 computer, the US makes 1 less TV In Japan, to make 1 computer, 4 less TVs are made. The opportunity cost is higher. Conversely, Japan has a comparative advantage in TVs. For each TV made in Japan, only of a computer is lost. Thus, the US can trade computers to Japan and both countries will be better off. Suppose there are 1 million workers in each country. If both countries specialize in what they have a comparative advantage in, Japan makes 40 million TVs, the US makes 50 million computers. This is more than can be produced if they each do some of each. Suppose workers assignments are split in half: US makes 25 million TVs, 25 million computers. Japan makes 5 million computers, 20 million TVs. Total is 45 million TVs, 30 million computers Example of trade. If US makes 1,000 more computers, 1,000 less TVs are made. However, Japan only needs to stop making 250 computers to make 1,000 more TVs. Thus, we have the same number of TVs and 750 more computers! Critique of comparative advantage:

o o o o

Once trade begins, a country that becomes specialized has no choice to continue. Thus, critics of globalization ask if diversification has value. Daly notes that the basic principle of comparative advantage depends on immobile factors. If factors are mobile, absolute advantage matters. E.g. all capital can go to the country with lowest costs. Response: The quality of a factor (e.g. skilled vs. unskilled labor) also matters. Factor mobility increases the opportunity set, so should make the country better off. However, a country with monopoly power in world factor markets can make itself better off at the expense of the rest of the world by limiting factor mobility.

III. Pollution Haven Hypothesis

Question: Does free trade make protecting the environment more difficult? Pollution haven hypothesis the idea that firms will move from countries with strong environmental standards to those with weaker standards. o Developing countries may even be able to attract industry with low standards. Esty notes that, while expanded trade and economic growth need not hurt the environment, there is no guarantee that it wont. o Three effects of economic growth on trade: 1. Technique Higher incomes => cleaner production processes. 2. Composition effects Higher incomes => preferences for cleaner goods. 3. Scale effects Higher incomes => increased pollution due to greater consumption. o If scale effects dominate, the environment will be worse off. Empirical evidence of the pollution haven hypothesis is weak. o Example of a test of pollution haven hypothesis: The Heckscher-Ohlin model of international trade states that countries will have a comparative advantage in goods produced with endowed factors that are in relative abundance. Countries with hydroelectric resources trade goods that require a lot of electricity to produce (e.g. aluminum) Countries with cheap labor produce labor-intensive goods (e.g. clothing) Countries with large capital stock produce capital-intensive goods (e.g. autos) Thus, the theory predicts that countries with a large capacity to assimilate pollution will produce pollution-intensive goods. As well as the physical ability to assimilate pollution, income matters.

Citizens of wealthier countries are more likely to demand cleaner environments. To determine if lax environmental regulations attract dirty industries, economists use the Heckscher-Ohlin model to explain trade in goods associated with pollution. Controlling for other resource endowments, one can test to see if environment variables are significant. Net Exportsij = ai + b1Ej1 + b2Ej2 + +bKEjK + dRj + eij Ejk = endowment in country j of factor k (e.g. capital, labor, land, natural resources) Rj = strictness of environmental regulations in country j (e.g. pollution control expenditures) In general, economists fail to find a significant relationship between environmental regulation and trade. One criticism of such models is that stocks of capital change slowly, as they represent years of accumulation. Thus, measuring flows of capital may be a better alternative. This is done by using foreign direct investment (FDI) as the dependent variable. FDIij = ai + b1Fj1 + b2Fj2 + +bKFjK + dRj + eij Fjk = level of variables affecting FDI in country j, such as tax policy Rj = strictness of environmental regulations in country j Using such a model, Xing and Kolstad (1997) find that foreign direct investment (FDI) for dirty industries, such as chemicals, does flow to countries with lax environmental regulations, but find no effect for low polluting industries, such as electronics. There has been more growth of toxic-intensive industries in developing countries. A recent paper by Ederington, Levinson and Minier offers explanations for small effects They argue that aggregate data misses effects in specific industries. Their work proposes three reasons why others find little effect Most trade is between similar countries (North/North) If look specifically at North/South trade, find an effect. They divide countries into low and high environmental costs When US environmental costs rise, net imports to low cost countries increase. Elasticity is 0.2 (10% increase in US costs => 2% increase in imports) Not all industries are mobile Industries are mobile if: Low benefits to agglomoration

Low transport costs Indstries that are mobile are more sensitive to environmental costs Not all industries are pollution intensive Surprisingly, they do not find a bigger effect of environmental costs on pollution intensive industries, unless they control for mobility. It appears that pollution intensive industries are also less mobile! However, this also means that the argument that there is no pollution haven effect because most PACE costs are insignificant is not sufficient. Other explanations for the weak evidence of pollution haven hypothesis In general, other factors are more important in choosing location: Skills of workers Proximity to markets Political stability Availability of materials As a result, even when evidence of the pollution haven hypothesis is found, the magnitudes are often weak. The Bureau of Labor Statistics collects data on total layoffs and the reasons. Few mass layoffs are due to environmental regulations.

2000 2001 2002 2003

Total Total people layoff affected events 5,620 1,170,427 8,350 1,751,464 7,295 1,546,976 7,346 1,502,825

Layoffs for Total people affected by environmental environmental reasons reasons 7 1,142 3 445 3 718 5 1,044

The most frequent reason for mass layoffs is completion of seasonal work (2,370 layoffs in 2003) and internal restructuring (1,437 layoffs in 2003). Other major reasons are contract completion and slack work. In 2003, 111 were due to import competition.

IV. International Agreements

The World Trade Organization (WTO), and its predecessor, the General Agreement on Tariffs and Trade (GATT), provides a framework of rules and procedures to be followed in international trade relationships. o GATT began in late 1940s. The WTO was established in 1995 after the Uruguay round of GATT (1986-1994) GATT covers trade in goods, and still exists as a subset of WTO.

o o o

Aims to reduce barriers of trade. Does allow exceptions to protect human, animal or plant life and to conserve natural resources. Such import restrictions must be done in a non-discriminatory way. Examples: GATT upheld the US tax on luxury cars and gas guzzlers because they applied equally to all autos (Europe protested the taxes). Denmark placed a ban on nonrefillable drink containers. Imports of such containers were prohibited. Other European countries felt that this was done to give Danish beverage producers an advantage, rather than protect the environment. The European court ruled in favor of Denmark, since it was non-discriminatory. In the 1990s, the U.S. banned imports of tuna caught in nets that kill dolphins. Mexico complained to GATT & won. GATT accepted Americas aim of protecting dolphins, but objected to the use of discriminatory trade sanctions. Suggested labeling of dolphin-friendly tuna instead. The WTO also ruled against a U.S. ban against shrimp harvested with technologies that harm sea turtle because it gave preference to Western Hemisphere countries. The ban gave Western Hemisphere countries more time to comply and provided financial aid for new technologies. India, Malaysia, Pakistan, and Thailand protested to WTO. However, the WTO noted it was the preferential treatment that violated GATT article 20. The ruling made clear, however, that countries did have the right to take trade action to protect the environment, as long as it is nondiscriminatory. The issue becomes cloudier when we consider pollution from production. Consider a firm in a country such as the U.S. that has strong environmental standards, but has a competitor in a nation with weak standards. The competitor has lower costs, so it has an advantage. However, it isnt clear that the U.S. can do anything legally under GATT. GATT gives the U.S. the authority to protect the health of its citizens. However, imposing stronger environmental restrictions on the second country affects the health of those citizens, not Americans. As we discussed in class, an important question is whether we should do something. Whose standards should matter? A World Trade Organization (WTO) report in the fall of 1999 admitted that trade can harm the environment.

Report says that environmentally damaging subsidies for farming, fishing, and fossil fuels should be eliminated. Report says more product labeling should be allowed. This leads to a bigger question: is it fair for developed countries to expecdt developing countries to strengthen environmental regulations?

Exhaustible Resources
I. Optimal Extraction of an Exhaustible Resource

Three classifications of exhaustible resources: 1. current reserves -- known reserves that can be profitably extracted at current prices. 2. potential reserves -- reserves that could be recovered at higher prices. 3. resource endowment -- the entire geological supply of resources (including those not yet discovered). Decisions to use exhaustible resources are dynamic decisions, because future availability of the resource depends on what is used today. Thus, exhaustible resources should be treated as an asset. A. The Costs of Extraction

There are two costs to using a resource, such as oil, today: 1. Extraction cost How much does it cost to obtain the resource? Obviously, only sell if P >= MEC (marginal extraction cost). 2. User cost -- the opportunity cost of not having the resource to sell in the future As a result, the price of the resource will be greater than the MEC. The owner of a resource, such as oil, has two options to make money for next year: 1. Sell all the oil now, and invest the profits at interest rate i. 2. Wait and sell the oil next year. Case A: Expected price next year rises less than the rate of interest: o Present value of marginal profits next year is less than current value this year: P1 - MEC > (P2 - MEC)/(1+i) o The owner of the oil is better off selling the oil now and investing it. o Leads to lower prices now (greater supply) and higher prices next year (lower supply). Case B: Expected price next year rises faster than the rate of interest: o Present value of marginal profits next year is greater than current value this year: P1 - MEC < (P2 - MEC)/(1+i) o The owner of the oil is better off waiting to sell the oil next year. o Leads to higher prices now (lower supply) and lower prices next year (higher supply). Prices adjust whenever one option (case A or B) looks better. Thus, equilibrium is reached when the expected price of the oil rises at the rate of interest. o P1 - MEC = (P2 - MEC)/(1+i), or o (1 + i)(P1 - MEC) = (P2 - MEC) Marginal user cost (MUC) -- the present value of the opportunity cost of the last unit of oil used not being available in the next period. P(t) = MUC(t) + MEC(t)

If marginal extraction costs are constant, the marginal user cost rises at the rate of interest. o This implies that the present value of marginal user cost remains the same! o Note that the price of a resource is greater than the MEC. Thus, higher prices are not, by themselves, evidence of abuse of market power. Rather, they simply represent economic rent due to scarcity. The mathematical example shows how the marginal user cost increases as scarcity is more of a problem. For those who would like more practice working with these concepts, you can download a spreadsheet with a numerical example by clicking here: energy_worksheet.xls. Changes in the marginal extraction cost o Up to now, we have assumed the MEC is constant. o If marginal extraction costs rise over time, the marginal user cost will fall. Intuition: MUC represents the opportunity cost of using the resource now. If it will be more costly to use the resource in the future, the opportunity cost is not as high.

B. Backstop Technologies

This theory describes how the price should change over time. But what price should we start at? o We want to run out of the resource at the highest price that people are willing to pay. o This is the backstop price. It is determined by a backstop technology. o A backstop technology is a technology that is available in vast quantities at the backstop price. E.g.: solar energy Note that the backstop price is constant, since the backstop technology is not exhaustible. o Note that, if several fuels are available, we will begin by using the one with the cheapest extraction cost, and continue, using up each fuel, until the backstop technology is reached. Note that, since these prices do not account for negative externalities, such as pollution, all of the dirty fuel supplies are used up!

Technological innovation on the backstop technology could lower its price and speed the transition. However, if the backstop won't be usable for many years, markets might not support this research.

II. Energy Prices Over Time

These theories predict that energy prices should rise over time. In reality, they have fallen until recent years. For example, consider the price of a gallon of gasoline in the US, in 2000 dollars: (http://www.eia.doe.gov/emeu/aer/txt/ptb0524.html) 1950: $1.62 1960 $1.48 1970: $1.30 1974: $1.53 1980: $2.20 leaded, $2.30 unleaded 1985: $1.72 unleaded 1990: $1.43 1995: $1.25 1998: $1.10 2000: $1.51 2005: $2.04 2006: $2.23 2007: $2.33 ($2.80 in today's dollars) While our model predicts that energy prices will rise gradually over time, they instead fell for much of the 20th century, and have risen more dramatically in recent years. What other factors have influenced energy prices? o Factors leading to higher prices Higher demand, particularly in the United States and China Risk premium
o o o o o o o o o o o o o

Concerns over security in the Middle East, political uncertainty in Latin America, and corruption in Russia all raise the risks of future shortages. Lack of refining capacity Even if more oil comes into the U.S., there isn't additional capacity to refine it into gasoline more quickly. Price volatility has reduced investment in recent years. OPEC has kept supplies tight, to avoid a return to cheap oil

Factors leading to lower prices Resource scarcity If a resource is not scarce, the user cost is near zero, so the price is set at the marginal costs of extraction. Hamilton (2008) finds evidence that scarcity rents were 0 before 1997. Early price spikes were due to supply fluctuations. Scarcity rents have come a factor as demand increases, leading to slowly rising prices. New discoveries have limited the importance of scarcity in oil prices. The model assumes constant marginal extraction costs Technological change has lowered extraction costs. Lower extraction costs increase economic reserves => lower MUC as well. Higher prices in recent years have made it feasible to extract energy from places with higher extraction costs.

Alternative Energy Technologies

I. Energy Prices Over Time (continued)

We concluded this topic with a brief discussion of peak oil. One question is whether we have reached a peak. o Optimists point to the role of improved technologies and substitutes. These become more viable, and receive more investment, when prices are higher. o Pessimists note that consumption has outpaced production over the past 20 years. New sources are harder to find, and are in less stable regions of the world.

II. The Role of OPEC

What role does OPEC play? o The Organization of Petroleum Exporting Countries (OPEC) is a cartel of oil producing countries.

They control the price of oil by agreeing on how much oil each member nation will produce. Formed in 1960. o OPEC can be analyzed using a dominant firm model. Note that if the supply of oil in the rest of the world increases, OPEC's price will fall. This has happened since the oil embargo of 1970. In 1979, OPEC provided 50% of the world's oil. By 1986, OPEC supplied only 30%. In 2009, OPEC supplied 40% of the world's oil. Higher prices increase the supply of the rest of the world. Because marginal extraction costs are higher elsewhere, non-OPEC producers will not be profitable when prices are low. Marginal extraction cost per barrel: Middle East $2 Venezuela $7 Gulf of Mexico $11 North Sea $11 Russia $14 Global supply and demand o 2009 demand for oil Global: 84.04 million barrels/day US: 18.69 million barrels/day o 2009 oil production Global: 84.17 million barrels/day US: 9.06 million barrels/day Persian Gulf region: 22.89 million barrels/day 9.76 million barrels/day come from Saudi Arabia alone. OPEC: 33.88 million barrels/day

III. Where Does Our Energy Currently Come From?

Most energy used in the US is consumed by industry and transportation (handout). Of the 99.30 quads consumed in 2008, 83.44 were fossil fuels. Just 7.30 quads were renewable. o Of the renewables, most were hydropower and wood. Just 7% of renewables were wind, and 1% of renewables were solar. o Power generation is dominated by coal, nuclear and natural gas. In 2006: 49% coal 21% natural gas 20% nuclear Worldwide, renewables made up 12.7% of total primary energy supply in 2006 (TPES). o Most of this is combustible renewables and renewable waste, which makes up 9.9% of total primary energy supply.

Biomass is particularly important in developing countries. Thus, they have higher shares of renewable energy. Share of renewables in TPES, 2006: Africa 49.0% Latin America 30.2% Asia (except China) 28.1% China 14.1% Non-OECD Europe 10.2% OECD 6.2% Former USSR 3.0% Middle East 0.8% o Hydro is another 2.2%. o The remainder is geothermal, solar, wind, and oceans. The oil problem o Oil is 40% of US primary energy supply, and 34% of global supply o US imported 65.4% of its oil in 2005 o Nearly all transportation powered by oil. o Concerns Dependence on oil leaves countries vulnerable to price shocks. Note that it isnt just dependence on foreign oil. In a global market, increased prices affect every country. Military conflicts to ensure supply Here, domestic production can help The climate change problem o Electricity generation is the main source of carbon emissions (24.5% of carbon emissions globally), followed by deforestation (18%) and transportation (13.5%). o Potential mitigation targets frequently discussed 550 ppm atmospheric CO2 Emissions path: Emissions begin to decline around 2020 Peak at 11 billion tons in 2040 Decline to 7 billion tons by 2100 BAU is about 20 billion tons Gradual decline to 3 billion tons by 2200 o Would lead to warming of about 3 C above pre-industrial values 450 ppm atmospheric CO2 Advocated by those concerned that even moderate warming currently experienced (about 0.8oC) has already had effects o This would lead to warming of about 2 C Emissions path: Emissions begin to decline around 2012 Peak at 9 billion tons in 2020 Decline to 3.5 billion tons by 2100 Decline to 2.5 billion tons by 2200 Without reductions, projected emissions over 21st century are around 1400 billion tons

These reductions would allow emissions of 500-800 billion tons Thus, we need to remove 600-900 billion tons. How can this be done? Geogengineering Change environment to offset emissions Seen as costly and risky Seeding clouds over oceans would increase their reflectivity One cost estimate is $4 billion Another more common proposal is projecting sunlight-deflecting sulfate particles into the atmosphere Sulfates are already in the atmosphere, and have a known cooling effect Consider, for example, weather changes after large volcanic eruptions. Amounts of sulfates in the atmosphere would need to increase by 15-30X to offset climate change. National Academy of Science panel suggests this could be done for pennies per ton reduced. What are the risks? Ocean acidification Aerosols could destroy the ozone Spatial climate patterns would change International considerations Geoengineering could be done unilaterally Raises potential for conflict Changes in climate may benefit some countries at the expense of others. Increase removal of CO2 Currently, vegetation holds 500-700 billion tons carbon Increasing by 20% (which seems unlikely) would reduce only 100-150 billion tons more. Thus, mitigation will play a large role.

IV. Alternative Sources

Note that we are not running out of energy. However, this does not mean that cheap energy will always be available. Rather, energy is a difficult issue for the following reasons 1. Multiple economic, environmental, and security aims Limiting costs increases usage (and emissions) and reduces funds for investment in infrastructure. Increasing domestic supply unpopular if it involves drilling in environmentally-sensitive places. Increasing supply via nuclear raises security concerns. 2. All sources face some limitations These are discussed in detail below. 3. Large embodied capital investment and long turnover times of worlds energy supply Replacement cost of todays global supply system is $12 trillion Typical turnover times are 30-40 years The potential and problems of energy sources o Conventional oil and gas What is available? Emissions are a concern o Coal, tar sands, oil shale Usage increases carbon emissions Coal has a high carbon content. The process of extracting oil from tar sands is more energy intensive. For example, in the oil sands in Canada, the oil must be melted out of rocks. Because this requires more energy, more carbon is released For example, a gallon of gasoline from tar sands releases 3x more carbon than a traditionally produced gallon. Integrated gasification combined-cycle (IGCC) is a way to reduce CO2 emissions from coal Turns coal into a gas before burning CO2 is separated from hydrogen when gasified CO2 is stored, and hydrogen is used to generate electricity Avoids the technical challenge of separating CO2 from other flue gasses. The technology is expensive. Needs a carbon price of at least $30/ton to be viable. o Carbon capture and storage (a/k/a carbon sequestration) Can be done before combustion (removing carbon from fuel) or afterwards (removing from waste gases) Some capture is already done (e.g. to produce fertilizer) However, once captured, the CO2 is then released Because of economies of scale, only appropriate for large emitters, such as power plants. Carbon is stored by injecting it into the ground Oil & gas reservoirs, deep saline aquifers, and un-minable coal beds are options.

Must be stored in formations with impermeable cap rock to avoid leakage. Eventually will dissolve in water. Thus, safety has been a concern for some. Space is an issue Storing 60% of the CO2 produced in the U.S. would fill all the space from oil consumed in the U.S. Cost estimates suggest a carbon price of $40 - $90/ton C needed. Retrofits would be more expensive. Biomass can also be used for storage (e.g. forestation) Scale will be an issue Increased re-forestation will reduce land available for agriculture. If biomass as fuel were combined with carbon capture and sequestration, it could have negative emissions. However, this is currently costly ($50-110/ton C) Biofuels Currently, this is the largest source of renewable energy. However, much of this is low-technology uses in developing countries. Presumably usage of these fuels will fall as countries grow. Other fuels include things such as ethanol. Is there enough farmland to grow the needed feedstocks as well as supplying necessary food supply? Recent concerns over corn prices is an example here Short term challenges Supply of feedstocks Medium term challenges Developing new feedstocks and conversion methods. Carbon content of biofuels depends on how produced A project in Georgia uses waste from cut trees. If the project receives credit for using material that would decompose anyway, the project could be carbon neutral. In contrast, using corn as a feedstock uses fossil fuels in production. Note how a CO2 tax would affect the incentives for different types of feedstocks. Hydropower Used for 16% of world electricity production. Does not require technological breakthroughs. However, political acceptance is an issue. Small hydro is cost competitive Geothermal Uses heat from the earth, which is captured as steam or used to heat water that is piped below the earth.

The technology is mature, but cost reductions are needed to make it competitive. Costs of wind fell by a factor of four between 1981-1999 Wind is now competitive in favorable locations. Now about 5-8 cents/kWh Competitive with traditional fuels with a $25/ton CO2 tax Barrett cites a study showing wind is competitive at $38/ton CO2 near Chicago, and could be situated further away with a price of $76/ton CO2. Distance from center decreases intermittency, but increases transmission losses. Because wind is intermittent, storage is an issue. For instance, excess power could be used to produce hydrogen stored under pressure in a reservoir. Currently feasible at about $93/ton Denmark and Norway work in tandem to provide power. When winds are favorable, Denmark exports wind energy to Norway. When not, Norway exports hydropower to Denmark. Essentially, the hydropower not used when wind energy is exported is stored energy. R&D needs include: Continued cost reductions Understanding extreme wind conditions Integrating wind turbines to the electric grid Storage Are there enough acceptable sites? Good sites have sufficient wind or solar resources, are near where energy demanded (to avoid transmission losses) and are not ruled out politically. Offshore sites take advantage of stronger, more consistent winds. However, these are more expensive and require better technologies. Barrett cites a source saying that wind could provide 100x the necessary power for the world. However, I would argue that there is not universal agreement on this. Finding appropriate sites is a limitation.

Wind

Alternative Energy Technologies

I. Alternative Energy Sources (continued)

The potential and problems of energy sources (continued) o Solar Solar is the most expensive of currently used renewable sources.

In addition to improving technology to lower cost, storage of solar energy is also an issue. As with wind, are there enough acceptable sites? However, because high pressure areas have fewer clouds and less wind, solar is most abundant in places where wind energy is scarce. Concentrated solar uses mirrors to produce heat, which turns a turbine. In prime locations, could be competitive at $35/ton C. However, many of the best locations are in developing countries. Who will pay this price? Solar photovoltaics Remain very expensive Useful for modular locations (e.g. remote lighting, signs, etc.), but not for mainstream use. Ocean energy Uses waves or tides to produce energy Costly and environmentally damaging Nuclear fission Hugely sensitive to capital costs In US, competitive with carbon prices above $27/ton C. More difficult for developing countries, because large capital costs mean that large scale plants (> 1000 MW) are most profitable. Safety concerns Has improved since Chernobyl. Passive reactors automatically shut down during an accident. Waste is an issue Long-term storage has yet to be implemented Currently, waste is stored on site. Reprocessing waste, to make it available in an easily handled form, increases the risks of nuclear proliferation. Particularly an issue if nuclear spreads to developing countries Uranium is an exhaustible resource Currently, 400 plants in operation Reserves could power 1,000 new reactors over the next 50 years. Uranium prices will rise if nuclear expands Could lead to new discoveries. Nuclear fusion Poses no accident risk Fuel is abundant Waste is less risky Still requires more energy to run than is produced, even after 50 years of research. Hydrogen Needs to be extracted from hydrocarbons or water Both of these currently use more energy than is produced.

To use for vehicles requires new infrastructure & new vehicles Costs may exceed $272/ton C o Hybrid vehicles Plug-in hybrids could serve as a transition technology If recharged from electricity grid, CO2 emissions depend on the type of power used. As usage increases, infrastructure for both charging and additional power generation will be needed o Improved energy efficiency Cheapest, cleanest, surest, and most rapidly expandable option, but lack of knowledge limits diffusion. In the IEAs greenest energy projection, energy efficiency accounts for 2/3 of averted emissions Many profitable measures currently exist Could earn average returns of 10-17% Some investments have been made Energy intensity falling 2%/yr in US, 1/5%/yr globally Potential concern is the rebound effect Higher efficiency makes using energy cheaper Thus, demand for services increases For example, drive more when cars use less gasoline Two British studies suggest the rebound effect cancels out 26-37% of the gains from energy efficiency The role of innovation o As the previous section makes clear, all clean technologies face technological hurdles. Overcoming these will lower costs, and make these technologies more competitive. Until the past few years, energy R&D efforts have remained relatively flat since the 1970s. o Current efforts $5-6 billion/year in US This is 1% of what US spends on electricity and fuels $3 billion comes from the federal government Revkin notes that government R&D funding for health and the military has grown much more rapidly. Note that much of this R&D, particularly from industry, focuses on traditional fossil fuels. Global efforts around $15-20 billion This is 0.5% of energy expenditures, and about 0.03% of world GDP Only Japan has increased R&D efforts recently

II. Technological Change and the Environment

The process of technological change includes three steps:

1. Invention the birth of an idea 2. Innovation commercialization of an idea 3. Diffusion Adoption and utilization of the innovation Note that technological change is uncertain. o We dont know whether research will be successful, or which projects will be successful. o While some patents are worth billions of dollars, most have little commercial value. o This suggests that a diversified strategy is desirable. Picking winners can be costly E.g. synfuels in the 1970s. Technological change and the environment is complicated by the presence of multiple market failures. o Of course, one concern is environmental externalities. Even if R&D markets functioned perfectly (which they dont), firms will not have incentive to develop environmentally-friendly products if the costs of pollution are not internalized. o In addition, market failures affect the process of technological change more generally. Market failures for knowledge o Knowledge is a public good. Alternatively, we can consider the results of innovation a positive externality. Once an idea is in the public domain, others can make use of it. As such, the inventor is not able to capture all of the social benefits of the innovation. As a result, the social returns to R&D are greater than the private returns to R&D. Studies typically find that the social returns to R&D are about 4X higher than the private returns to R&D. Implications: 1. Underprovision of R&D. Firms only care about the private returns. They invest in R&D until the marginal private rate of return equals the marginal cost. At this point, the marginal social rate of return will be higher than the marginal cost. Thus, even if environmental externalities are corrected, there will still be insufficient R&D. 2. Opportunity costs are important This high social rate of return is true for all R&D, not just environmental R&D. Thus, if we design policy to enhance environmental R&D, we must consider where those resources come from. At least in the short-run, resources available to do R&D are inelastic. Firms may face revenue constraints.

More importantly, R&D requires highly-skilled scientists and engineers. Goolsbee (1998) finds that one of the chief beneficiaries of R&D tax subsidies are scientists and engineers, who receive larger wages when subsidies are increased. In Popp (2004), I estimate that approximately one-half of the energy R&D spending that took place in the 1970s and 1980s came at the expense of other R&D.

Policy issues

Because of the public goods nature of knowledge, government policies are used to foster invention and innovation: Intellectual property rights (e.g. patents, copyrights) Give inventors a temporary monopoly, which enables them to capture more of the returns to their invention. In return, the patent document makes the invention public. As such, not every inventor chooses to patent an invention. Because of the temporary monopoly, patents encourage innovation, but slow diffusion. Concern over the high price of patented drugs, as compared to generic drugs, is an example. Government R&D funding The government can provide research funding to firms and universities, or can perform research itself in government laboratories. Many of the government laboratories are for the Department of Energy (DOE). In 2007, the US government provided $112.8 billion of federal R&D funding. Of that: $24.7 billion performed directly by govt. $9.6 billion performed by Federally Funded Research and Development Centers (FFRDCs) $46.5 billion performed by industry $25.0 billion performed by universities $5.8 billion performed by nonprofits Government funding gives the government more control over the type of R&D done. However, broader policies (e.g. supporting a range of options), is preferable to picking winners.

Government funding is particularly useful when spillovers are large. For example, basic research that cannot be patented and/or embodied in a proprietary product. Basic research can complement research done by firms. For example, DOE labs often include public/private partnerships to help commercialize new technologies.

Tax credits Tax credits lower the cost of R&D for firms. However, they give the government less control over the projects done. Firms will still choose to do the most profitable projects first, so tax credits are unlikely to stimulate basic research. o Incomplete information Uncertainties for R&D are particularly large. This makes raising capital to invest in projects difficult. This may be a particular problem for projects with long term payoffs, such as basic research. Also problematic for long-term environmental problems like climate change. o Adoption externalities Adoption may also include positive externalities. This is particularly a problem for energy efficiency People demand high returns (greater than 30%) to invest in energy efficiency Demand payback periods of 2-3 years One key issue is price Energy is a small portion of the average households budget Energy efficiency is higher in counties with higher energy prices In the U.S., energy consumption in a state is about 7% lower for each cent/kWh by which electricity prices exceed the national average. Potential adoption market failures Information As more people use a technology, others learn about it (epidemic effects) There are transaction costs to learning about new technologies. However, recent research suggests firm characteristics are more important than epidemic effects in explaining adoption.

Learning by doing & learning by using As firms or consumers gain experience with a product, costs may fall. If this learning benefits others as well, there is a positive externality. Principal-agent problems E.g. do landlords have incentives to improve building efficiency if tenants pay their own energy bills? Do tenants have incentives to conserve if landlords pay the bills? Lock-in Switching to new technologies can be expensive Thus, to adopt, the technology must not only be beneficial, but the benefits must justify the costs of switching. Thus, slow diffusion of capital goods may be rational. E.g. even if gas prices up, dont get rid of your current car right away. Lock-in is particularly problematic when there are network externalities. Network externalities are when one persons usage of a product affects others. For example, hydrogen-powered cars arent practical without filling stations. But filling stations arent profitable without many cars to serve. Moreover, switching from fossil-fuel based filling stations would be expensive. As a result, Paul David suggests asking whether we can identify cases in which society would have been better off if another technology had been chosen. In addition, he suggests delaying irreversible commitments. What matters most for energy efficiency? Studies suggest that households are more sensitive to up-front costs of investment than to energy prices (and thus, the potential savings) This suggests that subsidizing energy efficiency purchases would help Consider, for example, the case of energy-service companies (ESCOs) described in The Economist These companies borrow money to make energy efficiency improvements in a building Earn profits by retaining the resulting cost savings Most customers are government, schools, hospitals, and universities Policy options

Investment subsidies Deal with concern over up-front costs Product labeling Energy Star labeling is an example Deals with the information problem Product standards (e.g. product efficiency standards) Forces consumers to make choices that they are not currently making For instance, Australia has proposed banning incandescent light bulbs Tradable white certificates Projects that improve energy efficiency are certified Utilities required to have minimum investments in energy efficiency Can buy and sell certificates to meet requirements Utility regulation Because operate in regulated markets, utilities face little incentive to encourage efficiency One way to do so is to decouple sales and profits Regulators forecast demand and set a price that earns profits at that price If demand is lower than expected, regulator lets price rise If demand is higher, regulator cuts prices

III. How to Get There: Policy Options

Recall that there are two market failures at work (environmental and innovation). Calls to increase federal R&D spending address innovation market failures, but not environmental market failures. Environmental market failures require environmental policies such as taxes, permits, or command and control policies. o Types of policies used Many policies designed to encourage alternative energy are very specific, rather than broad-based policies such as a carbon tax. Renewable energy targets Many EU countries and US states have targets for a percentage of energy that should be generated by renewable resources by a certain date. In some countries, such as Australia and Japan, these are binding constraints. E.g. all wholesalers must get 2% of their electricity from renewable sources. In other cases, these are policy goals that are accompanied by other policies to help meet these targets.

Price guarantees Some EU countries guarantee a higher price for electricity generated from renewable sources. This helps make these sources competitive with other fuels. Examples include feed-in tariffs in Germany Germany guarantees a price of 55/kWh for solar, and 8.4/kWh for wind Tradable green certificates (Energy Policy 2003) Used in Europe, Australia, Texas The program begins with a target level for percentage of renewable energy use. Producers get a certificate for each unit of renewable energy supplied to the grid. Customers or distributors must show that they use at least that percentage of renewable energy. They do this by purchasing permits. Since producers of renewable energy sell the permits, they are compensated for the extra cost of producing renewable energy. Investment subsidies Examples are tax credits for installation of solar panels, energy efficient appliances, etc. U.S. has a 1.9/kWh production tax credit Encourages wind production, since that is closest to being competitive Uncertainty is an issue, since the credits need to be renewed frequently Policy considerations Even if current technologies make large scale reductions costly, dont we want to provide incentives for some basic reductions now? It will be more costly to do more later, as we will have missed low cost options that are currently feasible. Gradual phase-in is useful, as it gives time for the capital stock to turn over. Environmental policies provide incentives for increased R&D. Policies are needed to provide rewards for green innovation. However, consider what types of R&D encouraged by these incentives. Will projects with only long-term payoffs (e.g. solar PV) be encouraged? Note, for example, that feed-in tariffs encourage R&D on solar, since they guarantee a return on solar investments. In contrast, in countries using renewable portfolio standards, R&D efforts focus on wind. Because the costs of wind are lower, it is the alternative energy source chosen if a specific technology is not mandated by policy.

Finally, note that higher energy prices help encourage investment in alternatives, but they are not a substitute for environmental policy. Higher energy prices also encourage the search for more fossil fuels. Some of these, such as tar sands, even produce more carbon emissions. In contrast, policies addressing emissions change the relative price of fossil fuels, so that cleaner sources become more competitive. Innovation market failures require government support for R&D. o Basic research Recall that this is particularly important for basic research. Basic research will be important for discovering new technologies. Examples include using solar energy to produce hydrogen fuel. Also needed when payoffs are in distant future. o Applied research This is developing technologies for the ma

Climate Change
I. The Problem of Climate Change

Climate change comes from the accumulation of greenhouse gases in the atmosphere: o carbon dioxide (CO2) major contributor: 49% of the proportionate effect is from CO2 o methane (CH4) 18% of proportionate effect o nitrous oxide (N2O) 6% of the proportionate effect o water vapor As these gases accumulate in the atmosphere, they trap infrared radiation (heat) that would otherwise escape into the earths atmosphere. Hence, the name greenhouse effect. o The effect is natural. Without the greenhouse effect, the earths temperature would be -18oC (-0.4oF), rather than +15oC (59oF). o The question is the effect that human activity has on the greenhouse effect. Concentrations of these gases have been growing. o Were between 270-290 ppm for thousands of years before the Industrial Revolution By 1998, concentrations were at 365 ppm. Today they are at 390 ppm for CO2 only, and 430 ppm CO2 equivalent. o Although some emissions are natural, the problem is that human activities have increased the rate of growth. Sources of emissions: o CO2: Burning of fossil fuels Deforestation (because forests are carbon sinks) o Methane:

Released from wetlands and other areas where anaerobic decay of organisms occurs. Emissions from ruminants gases from cud-chewing animals, such as cows Leakage from natural gas pipelines.

N2O:

Burning of fossil fuels Agricultural fertilizer Part of the problem is that some greenhouse gases (particularly CO2) remain in the atmosphere for a long time. In addition, different gases have different radioactive forcings (heat absorbing potential). o For example, methane has a greater radioactive forcing than carbon dioxide, but does not persist for nearly as long in the atmosphere.

II. Estimating the Damages from Climate Change

To begin, we want to consider the economic costs of climate change. Then we can address relevant policy issues. What consequences will climate change bring? Predicted consequences to global warming include: o A rise in sea level 50 cm rise expected by 2100 Leads to erosion of shoreline and loss of habitat. o Increased intensity of storms o Changes in agriculture Can be both positive and negative. For example, increased precipitation will help. Overall effect expected to be negative. However, may be positive in some areas. Crop locations will change. Developing nations expected to be hurt the most. Agriculture makes up just 3% of GDP in developed economies, but is often 30-50% of GDP in developing countries. o Changes in energy use Less need for heating, but more for air conditioning. o Changes in ocean currents Shifts in ocean currents could dramatically cool the European continent. o Increased strain on fragile ecosystems o Health effects For example, tropical diseases may spread. There are two options for dealing with global warming: o Mitigation taking steps today to reduce emissions to avoid or limit future problems. Examples of mitigation: Using less fossil fuels. Pollution control devices.

Reforestation trees act as carbon sinks. Enhancing the ability of oceans to absorb carbon. Adaptation efforts of future generations to adjust in ways that reduce the negative impacts of climate change. When valuing the damage done by climate change, we need to consider the cost of adaptation. The cost of a sea level rise to the island of Manhattan is not the cost of Manhattan being under water, but the cost of preventing flooding on the island. The readings discuss how adaptation is easier for developed countries. Consider Malawi, which lacks irrigation and depends on a single crop (maize). Compare to Australia, which uses desalination to cope with drought. Raises the question of whether raising incomes in developing countries is a better solution. However, is adaptation possible for developing countries? Not only do developing countries have fewer resources, but they are often smaller, and thus in only one climate zone. Thus, substitution of activities among different regions of the country would be impossible. Methodologies for estimating damages o Studies must make assumptions about o Future emissions Extent and pattern of climate impacts Economic value of these damages o Types of studies Nordhaus (1994) interviewed experts Frakhauser, Nordhaus, and Tol use enumerative method. Begin by estimating physical effects from natural science research Give each physical effect a price and add up E.g. calculate effects of agriculture by modeling how crop prices change as crop outputs change Advantage Based on literature from natural science Disadvantage Concerns about extrapolation Studies from limited areas extrapolated to rest of word Results from recent past extrapolated to distant future Are assumptions about adaptation realistic? Additional adaptation would lower costs Mendelsohn and his co-authors use a statistical approach.

Direct estimates of welfare impacts using observed variations in prices and expenditures Done for selected countries, added up, and extrapolated. Advantages Based on real-world observations, rather than extrapolated differences Disadvantages Differences in values across places attributed to climate. Might there be other unobserved variables that matter. Much of the research looks at cross-section variation Appropriate because climate is about long-term trends, not year to year variation. Thus, cross-section variation gives you variation in climates However, some important aspects, such as sea level rise, do not have much spatial variation. Table 1 in Tol presents results for warming between 1-3oC. These are total cost estimates. o o Estimates range from no effect to a loss of 4.8% of GDP for 3 C warming Equivalent to about one years GDP o Figure 1 fits a regression line to the estimates Note low levels of warming may have small benefits o Turning point occurs around 1.1 C Much of this warming will occur anyway, as early emissions are sunk costs. More CO2 reduces water stress in plants, allowing them to grow faster. Most economic activity in temperate zones, where warming reduces cold weather problems More people live in tropical areas, but there is less economic activity there Raises interesting equity issues Damages get more substantial as temperature increases More recent studies tend to be less pessimistic, as they incorporate more adaptation opportunities Uncertainty is skewed to the right More likely to experience disaster than for climate change to have large benefits o Regional differences The table also lists the areas from each study with the smallest and largest effects. Africa is typically the worst off region, with losses around 5-10%, and one estimate for sub-Saharan Africa at 23.5%. Note that the 2010 World Development Report from the World Bank says that 2oC climate change will cost Afric 4% of GDP, and India 5%.

Why are developing country losses larger? In tropical regions, thus already hotter Less able to adapt to climate change because of lack of resources and less capable institutions. Inadequate housing and health care systems leave poor people more vulnerable to natural disasters Example: After Hurricane Mitch in Honduras in 1998, poor households lost 15-20% of their assets, while rich households only lost 3%. 10 of the 15 largest cities are in low-lying areas prone to flooding Impact on natural systems must be considered. Nations that depend more on unmanaged ecosystems (that is, developing countries) will be impacted more. These nations are more dependent on naturally occurring rainfall, runoff, and/or temperatures. Estimates of marginal damage o Estimates of marginal damage give us a social cost of carbon. This can be used to guide policy, as a carbon tax should equal the marginal social cost. o Methodology Estimates begin with a total cost estimate Of the marginal cost estimates reviewed by Tol, all use one of 9 total cost studies as a starting point. Different assumptions about discount rates, future emissions, effects of carbon cycle, etc. lead to different marginal cost estimates from these 9 studies. o E.g. for a study with 3 C warming, need to assume when the warming will occur, whether the damages are linear or quadratic, etc. o Table 2 summarizes results Mean is $105/ton C Median only $29/ton C Mode only $13/ton C o Higher with lower discount rates th At a 3% discount rate, the 99 percentile is just $45/ton C o What are these studies missing? Both positive and negative effects may be ignored. Examples include Costs Saltwater intrusion into groundwater due to sea level rise Increased damages from storms Extreme events are unknown (e.g. changes in ocean currents) Biodiversity loss is hard to quantify Benefits Higher wind speed lowers future cost of wind energy Reduced polar ice opens shipping lanes

Warmer weather in cold regions reduces costs of heating and disruptions from snow storms. Most studies stop at 2100 Longer term effects not as well known. As an example, we looked at tables from Nordhaus (1999), which estimates the global impact to be about 1.5% of GDP. o Effect by region (% of GDP): US: 0.45% European Union: 2.83% Japan 0.5% India: 4.93% China: 0.22% o We also looked at other studies from the US, which found damages ranging from -0.1% of GDP (a net benefit) to 1.5% of GDP. o Both of the tables used in class are available from William Nordhaus' web site for his book, Warming the World: Economics Models of Global Warming (with Joseph Boyer). Both tables used in class come from chapter 4. They are tables 4-10 and 4-11.

III. The Costs of Dealing With Climate Change

What are the costs? They depend on: o How much we can improve energy efficiency o How fast the price of renewable technologies falls o The availability of technological solutions such as carbon sequestration o Willingness of consumers to change behavior o Policy instruments used Costs are higher if inefficient policies (e.g. those not encouraging the lowest cost reductions first) are used o How fast will reductions occur Can we replace existing capital with more efficient equipment as it wears out, or will we need to scrap existing plants? o Assumed effects of climate change, such as: The feedback between emissions and temperature change The effect of temperature change Because of the uncertainty surrounding global warming, many assumptions are needed in these models. o Examples: The rate of technological change The feedback between emissions and temperature change The effect of temperature change Examples from IPCC: cost of Kyoto varies by country. Allowing for trading within the EU, but not elsewhere, Kyoto leads to a 0.5 to 7% loss of GDP, with most countries clustering around 1%. Puting targets in perspective:

Reducing warming to 2 0C requires concentrations no larger than 450 ppm CO2 equivalent. o Present concentrations are 430 ppm CO2 equivalent (390 CO2 only) o o A 550 ppm CO2 equivalent target yields warming around 2.5-3 C. The Stern Review has generated much discussion o It estimates damages from emissions ranging from 5-20% of GDP Note that this is higher than the estimates in Tols review o It estimates the cost of abatement (for stringent abatement) would be about 1% of GDP per year Would limit concentrations to 500-550 ppm o Weynat (2008) notes that IPCC estimates costs for stabilizing greenhouse gas concentrations ranging from -1% (slight benefit) to 5.5% of world GDP in 2050. o Critiques Uses a low discount rate Optimistic about future technologies High damage estimates ignore opportunities for adaptation Assumes perfect policy implementation (e.g. lowest cost emissions reductions done first, on a global scale) Finally, in class I presented data on carbon taxes necessary to achieve various goals. The idea was to convert the costs of various policy options into an easy to understand numerical framework. o A carbon tax would be based on the carbon content of fuels. Thus, fuels with high carbon contents, such as coal, are taxed more heavily. Examples of the effect of carbon taxes on prices:
o

Oil Absolute price increase Current price (2007/08) Percent increase

Natural Gas

Coal

Electricity

Gasoline

$1.18/barrel $0.15/mcf $5.19/ton $0.0017/kWh $0.024/gal $108.91 1.1% $13.01 1.2% $34.26 ('06) 15% $0.0914 1.9% $3.03 0.7%

What is the cost of various policy options? o The following table presents the carbon tax necessary to achieve each of the following policy options. Note that the tax levels increase over time. As population increases, the marginal costs of reducing emissions increase. o The results are taken from William Nordhaus' DICE model of global warming. The complete study including these results is available from his web site: http://www.econ.yale.edu/~nordhaus/homepage/dicemodels.htm. o The "optimal policy" is one which equates marginal benefits and marginal costs. Policy Option 2010 2025 2100

Optimal policy 1.5x pre-industrial concentrations (420 ppm) 2x pre-industrial concentrations (560 ppm) Limit temperature increase to 2.5oC
o

$33.80 $53.90 $202.40 $189.70 $421.92 $761.20 $39.60 $58.97 $445.50 $60.20 $102.25 $863.40

Effect on emissions (reduction from a no-policy baseline projection): Policy Option Optimal policy 2015 2025 2095

15.9% 18.5% 40.4%

1.5x pre-industrial concentrations (420 ppm) 42.8% 58.3% 87.9% 2x pre-industrial concentrations (560 ppm) Limit temperature increase to 2.5oC
o

16.6% 19.5% 61.3% 21.5% 26.5% 97.8%

Note that, because carbon emissions persist in the atmosphere for a long time, there is little difference in the effect of each policy on temperature over the next 100 years. In these models, temperature increase by 2100 in optimal policy is 2.61o C. It is 1.61o C in the limit to 1.5x case, and 2.48 in the limit to 2x case.

The Political Economy of Environmental Policy

Today's class was a discussion of the role that economics plays in environmental policy. The notes below highlight some of the key points raised in today's discussion, as well as summarizing some of the important points from the reading. I. What role does economics (and economists) play in environmental policy?

An interesting point raised is that environmental policy is dominated by lawyers and economists. o One thing we discussed was that this is true at the stage of policy implementation, but not necessarily when goals are set. Natural scientists play more of a role bringing issues to policy makers attention and deciding on the targets to be met. Why might economists play an important role. o In the U.S., the most visible role for economics is for cost-benefit analysis. This has led to a more normative use of environmental economics. Environmental economics contributes to policy by clarifying the choices available. By placing values on potential benefits, it can help clarify what we are getting for these costs. o The development of cost-effective policy is also important.

Economists play two roles here: Estimating compliance costs Designing cost effective policies. This affects both the type of policy and the timing of policy. New policy instruments (e.g. tradable permits) have gained increased acceptance among policy makers. Still, market-based policies are not often used in U.S. environmental policy. Considering the players involved in environmental policy may help explain why. Industry Prefer policy instruments with low costs to the firm (not necessarily to society as a whole) E.g. grandfathered permits vs. auctioned permits Taxes hurt firms because they must pay both for the reductions they make and pay a fee for the reductions they do not make. Regulations can generate rents Firms will support regulations that offer new opportunities for them. For example, firms that produce pollution control equipment benefit from technologybased standards that require their product. In the 1980s, DuPont favored CFC reductions because it had developed a substitute. Note that transfers of wealth are important. Even socially efficient regulations may impose costs on individual firms. Thus, firms have incentive to lobby against regulations. Consumers Individual consumers can influence firms by their purchasing decisions. However, there are limitations to this: Information: do individuals understand the environmental effects of their purchases? Free-riding: people may feel that individual actions will have little effect. They note that individual consumers or firms receive little reward for lobbying Note also that, while the costs of policy may fall on a few affected firms, the benefits will be shared by a larger group of citizens. Concentrated costs and diffuse benefits make it more difficult to generate support for regulation. Interest groups

Individual actors, whether they are firms, consumers, or environmentalists, have more power when organized as interest groups. Environmental organizations care about: Likelihood that the instrument considered will be chosen Increasing their influence For example, being identified with successful legislation could increase fundraising Most groups oppose market-based policies Market based policies provide less control over final outcomes than command and control policies. Note also that, in a democratic system, the preferences of environmental groups likely differs from those of the general population. One take-away from this, that is highlighted in the readings, is that the individual motives of each group are important.

II. What role should economics (and economists) play in environmental policy?

A concern was raised that cap-and-trade provided a new concept of property. o Should the government be able to create property rights? o If not, what are the implications of environmental regulation for existing property rights?

III. After taking a semester of environmental economics, what key insights does economics offer to the study of environmental problems?

One of the key lessons of economics is that individual behavior matters. o Economists study how people react when their behavior faces constraints. o These constraints typically come from scarcity, which leads to opportunity costs. However, government policy can also lead to constraints. o Understanding how people will react to policy is important to achieving policy goals. A key point here is that incentives matter. Economics is about how people respond to incentives. Environmental economics can both tell us when the incentives of the marketplace are insufficient to achieve an efficient solution, and also tell us how individuals will respond to the incentives given to them by various policy mechanisms. This can often be helpful for environmentalists to remember. Economics is not about how people should behave, but how they do behave. Since the preferences of environmentalists may differ from others, it is important

to realize that not everyone will choose to protect the environment without incentives to do so. Another important insight was the effect of regulation on costs over time. o Regulations provide incentive to innovate, making the costs of future regulation lower. This can make it easier for developing countries to regulate. As remaining environmental problems become harder to solve, this becomes more important. We discussed, for example, that in developed countries, many of the easy problems have been solved, and that the benefits of remaining environmental problems often occur in the future.

IV. Are there weaknesses to the economic approach to environmental policy?

One point that came up early in class (regarding ecosystems) is that some benefits are more difficult to quantify. o Do they, in turn, get less attention from the policy community? Another related to discounting o How does economics deal with issues where costs occur today, but the benefits are delayed?