Offset Markets for Nutrient and Sediment Discharges in the Chesapeake Bay Watershed Policy Tradeoffs and Potential Steps Forward


NCEE 0

NATIONAL CENTER FOR

ENVIRONMENTAL ECONOMICS

Offset markets for nutrient and sediment discharges
in the Chesapeake Bay Watershed: Policy tradeoffs
and potential steps forward

Andrew Manale, Cynthia Morgan, Glenn Sheriff and
David Simpson

Working Paper Series

Working Paper # 11 -05
August, 2011

^e.0 sr^ U.S. Environmental Protection Agency
g	ra National Center for Environmental Economics

s	z 1200 Pennsylvania Avenue, NW (MC 1809)

^ Washington, DC 20460
sP http://www.epa.gov/economics

ProT^-

-------
Offset markets for nutrient and sediment discharges in
the Chesapeake Bay Watershed: Policy tradeoffs and
potential steps forward

Andrew Manale, Cynthia Morgan, Glenn Sheriff and
David Simpson

NCEE Working Paper Series
Working Paper # 11-05
August, 2011

DISCLAIMER

The views expressed in this paper are those of the author(s) and do not necessarily represent those
of the U.S. Environmental Protection Agency. In addition, although the research described in this
paper may have been funded entirely or in part by the U.S. Environmental Protection Agency, it
has not been subjected to the Agency's required peer and policy review. No official Agency
endorsement should be inferred.

-------
Offset markets for nutrient and sediment discharges in the Chesapeake
Bay Watershed: Policy tradeoffs and potential steps forward

Andrew Manalea, Cynthia Morgana, Glenn Sheriff1, and David Simpsona'b

June 2011

We are grateful to our EPA colleagues Richard Colbert, Kevin
DeBell, Amelia Letnes, Carol de Marco, Jeff Potent, and Bob Rose
for helpful suggestions on earlier drafts. Mark Conway and Sarah
Hinkfuss provided excellent research assistance. The authors are
responsible for any errors, however. The opinions expressed here
are those of the authors only, and do not necessarily express the
views of the United States Environmental Protection Agency.

a US EPA National Center for Environmental Economics.

b Corresponding author. 1200 Pennsylvania Ave. NW, Washington, DC 20460, phone: (202)
566-2356, email: simpson.david@epa.gov.

-------
Offset markets for nutrient and sediment discharges in the Chesapeake
Bay Watershed: Policy tradeoffs and potential steps forward

Abstract

Considerable interest has been expressed recently in prospects for water quality trading markets
between nutrient sources in the Chesapeake Bay Watershed. Allowing such flexibility in
response to the terms of recently announced total maximum daily load (TMDL) restrictions
might considerably decrease costs of compliance with the TMDLs. Before an effective and
efficient market for offsets can be established, however, certain preconditions must be met. In
particular, there must be means by which nutrients can be measured, allowances can be assigned,
and limits on nutrient discharges enforced. In this paper we consider some factors that may
affect the realization of these preconditions. A recurrent theme is that there are tradeoffs in
policy design. A regime that imposes tight restrictions on those who are eligible to trade may
also limit the cost savings that might be realized from trading. On the other hand, a regime that
maximizes market participation might fail fully to achieve the environmental goals of an offset
or trading policy. We conclude with some recommendations for steps that might be taken to
initiate limited markets in nutrient and sediment discharges. These markets might then be
expanded as experience is gained and methods developed to assure improved market
performance.

Key words: water quality trading, offsets, transactions costs, adverse selection, leakage,
additionality, monitoring, Chesapeake Bay, nutrients

Subject matter classification: 2) water pollution, 26) land use, 43) non-point source pollution
JEL codes: Q15 - agriculture land use etc.; Q53 - water pollution etc.

-------
Introduction

Water quality trading is a much-discussed policy instrument for reducing pollution in the
Chesapeake Bay watershed. Executive Order 13508, issued in May 2009, heightened interest in
the topic by directing federal agencies "to protect and restore the health, heritage, natural
resources, and social and economic value of the Nation's largest estuarine ecosystem and the
natural sustainability of its watershed" (US GPO 2009). In theory, water quality trading achieves
environmental benefits at least cost. In practice, decisions made regarding details of
implementation rules can affect the magnitude and direction of both benefits and costs. Several
jurisdictions, both in the United States and in other countries, have implemented water quality
trading programs during the past three decades. Here, we draw from the literature analyzing this
experience to highlight tradeoffs in offset market design and offer suggestions regarding
potential steps forward for a water quality trading program in the Chesapeake Bay.

Improving water quality in the Chesapeake Bay is a daunting task. The Bay's three principal
pollutants, nitrogen,1 phosphorus2, and sediments, issue from a variety of sources. Some are
"point sources," with identifiable ends-of-pipes or means of conveyance. Examples include
municipal and industrial wastewater treatment plants (WWTPs), and municipal separate storm
sewer systems (MS4s). Under the Clean Water Act (CWA), the major statute addressing water
quality in America's streams, rivers, lakes, and estuaries, point sources are required to obtain and
comply with National Pollution Discharge Elimination System (NPDES) permits. Concentrated
animal feeding operations (CAFOs) are also point sources. Only CAFOs that discharge into
jurisdictional waters, however, require an NPDES permit.

Nonpoint sources, which are not regulated under the CWA, are less easily monitored. About 60
percent of the nutrients and sediment reaching the Bay originate in farms, private septic systems,

unregulated stormwater flows, or as airborne emissions (Batiuk 2010). Ideally, nonpoint sources
would face discharge limits and be free to determine the most cost effective means of meeting
them. In practice, it would be difficult to measure actual discharges from such sources and
determine the contribution of each to overall environmental degradation.

If regulating nonpoint source discharges4 directly is not feasible, regulators can take alternative
approaches to reducing pollution. One is to regulate polluting inputs, rather than the pollution
itself. Examples include setting animal stocking levels, setting limits on feed imports into a
geographic area, and regulating the composition of the animal feed. Compared to regulations
controlling point source discharges, input regulations are imprecise. Nutrients fed to an animal,

1 More precisely, contamination arises from "reactive" nitrogen, i.e, nitrogen bound in molecules with atoms of
other elements, in contrast to the inert form of nitrogen (N2) comprising about 80 percent of the atmosphere. Since
the latter is benign, pollution control strategies often "denitrify" wastes by converting reactive nitrogen to N2. In this
document, any references to "nitrogen" as a pollutant indicate its reactive form.

2 By phosphorus, we mean bioavailable phosphorus and not forms of phosphorus that are bound to other elements
such that they are not available for biological reactions.

3 This figure is based upon the 2009 Chesapeake Bay model, and excludes contributions from POTWs and forested
lands.

4 Our use of the term "discharge" may differ from the more restrictive definitions given to the term in other
contexts. By "discharge" we mean simply "something that is discharged - i. e., emitted or released - from some
source". We do not necessarily suppose that a "discharge" is prohibited or regulated by any existing law or rule.

-------
for example, have a different impact on Bay water quality depending on various characteristics
such as soil type, location, weather, production practices, and time of year.5 Another
disadvantage to input-based regulation is that it may not be cost effective; it may be cheaper for
polluters to reduce the damage to the bay by controlling effluent rather than inputs.

Another alternative is to require abatement actions at the farm or field level designed to reduce
nutrient and sediment discharges. Examples include manure and fertilizer handling, crop choice
and timing, tillage, and riparian buffers. Like input-based regulations, these Best Management
Practices, or BMPs, have the disadvantage of being inexact instruments for discharge control.
Topography, weather, soil conditions, and other factors determine their effectiveness. It is
impractical for regulators to factor in location and circumstance with enough specific
information to determine the precise relationship between BMP use and emissions for every
source. Also, BMPs may not be cost effective if one can reduce discharges more cheaply by
controlling polluting inputs.

Polluting input reductions and BMPs may either be required or rewarded. Rewards might come
in the form of payments for participation in voluntary programs such as the US Department of
Agriculture's Conservation Reserve and Environmental Quality Incentive Programs, or as
payments for credits sold in environmental markets.6

Environmental markets could allow regulated point sources to meet their NPDES permit
obligations by purchasing a credit from an unregulated nonpoint source that agrees to reduce

polluting inputs or undertake BMPs to reduce loadings by an equivalent amount. The economic
argument for such transactions is straightforward, although the analysis becomes more
problematic the farther one moves from textbook ideals toward real-world applications. A cap-
and-trade program such as the Sulfur Dioxide (Acid Rain) program comes close to the ideal. The
"cap" means that there is an overall ceiling on allowed loadings, and that the aggregate
numerical quantity is allocated among all dischargers. No participant can increase its loading
beyond its allotment without obtaining someone else's discharge permit.

With large numbers of entities participating in a permit market, there should be competition
between buyers and sellers, trading institutions should arise, and the ultimate outcome should be
cost effective. That is to say, the overall cap on discharges is not violated, and entities that can
reduce their discharges at a lower cost sell their rights to those who would face a higher cost of
reduction. The overall cap would thus be met at the least total cost. The intuition behind this
outcome is simple: if any participant could reduce its loading at lower cost than another, it would
have an incentive to sell its permit for at least one unit. Since such trading is possible, all such
cost-reducing transactions should be consummated.

5 Advances in modelling may help reduce the degree of uncertainty for some of these factors.

6 A disadvantage to the reward approach is that it makes the polluting sector more profitable, which can encourage it
to expand over time.

7 Of course environmental markets can and do also allow regulated point sources to trade with one another. In this
paper, however, we focus on the prospect of point-nonpoint source trades, however, as conventional wisdom has it
that the latter as a class typically have considerably lower marginal costs of abatement than do the former, and hence
the potential gains from trade would be greatest between point and nonpoint sources.

-------
Although a total maximum daily load (TMDL) has been established for the Chesapeake Bay
watershed, for the purposes of a water quality trading program it is a quantitative target rather
than a strictly enforceable cap. A strict cap would imply that every participant in the trading
system would require a permit to discharge. As currently envisaged only point sources covered
by NPDES permits have a waste load allocation (WLA) stipulating their allowed disharge. State
regulators assign a load allocation (LA) to nonpoint sources as a group, without restricting
individual sources. The Clean Water Act, as noted above, lacks a mechanism to ensure that
individual nonpoint sources reduce their discharges below a specific level. EPA requires that
states provide "reasonable assurance" in their Watershed Implementation Plans that nonpoint
sources will collectively meet the LA. Moreover, states may require monitoring of nonpoint
sources that participate in a trading program. Individual farms that meet the regulatory definition
of a non-point source, however, do not need to acquire a permit to expand their operations or
change production practices in a manner that increases discharges to the Bay.

The remainder of this paper discusses policy design issues for an "uncapped" offset market. We
begin by describing the theoretical benefits of water trading in the Bay. We then discuss practical
implementation decisions policy makers face, identifying implicit tradeoffs between program
cost and environmental performance. We conclude with suggestions for productive first steps in
developing an effective water quality trading program.

Prospects for water quality trading in the Chesapeake Bay Watershed

The Bay's restoration goals require nitrogen and phosphorus reductions of roughly 15 percent to
meet 2017 interim annual goals, and by 24 percent to meet the 2025 objectives (US EPA
2011).Under the CWA, EPA has direct regulatory control over point sources. From 1985 to
2008, WWTPs reduced annual nitrogen loading from direct discharge by 44.1 percent and
phosphorus loading by 68.2 percent, despite serving growing populations (US EPA 2009). While
further reductions may be possible by upgrading treatment technologies at these facilities, such
upgrades will likely prove expensive. Moreover, WWTPs now account for 52 million pounds of
nitrogen entering the Bay each year, while an annual reduction in nitrogen loading of 84 million
pounds is sought. Thus, even if it were possible to eliminate WWTP loading completely,

substantial reductions would be required from other sources to meet the goal.

8 These figures for reductions from WWTPs may overstate actual reductions since they represent reductions from
the effluent streams of direct discharge. Except for the relatively small amount of reactive nitrogen that has been
converted to gaseous form, nitrogen and phosphorus remain in the biosolids extracted from such treatment plants. At
present much of the residual is applied to agricultural land in the Chesapeake Bay Watershed. If the nutrients are not
appropriately managed and the biosolids do not displace fertilizer that would otherwise have been applied, they can
contribute to the load of nutrients entering the Bay from the land on which they are applied. Phosphorus in biosolids
is relatively non-soluble and hence less likely to move to water unless soils have reached saturation levels. The same
is not true for nitrogen which can more readily be converted to a highly water soluble form. Hence, there can be a
tradeoff between more effective reductions of nutrients in the direct discharge of WWTPs and the greater volume of
nutrients in biosolids disposed of land that can potentially become a non-point source with concomitant lesser
regulation under the CWA.

-------
Stormwater runoff from MS4s contributes about 10 percent of annual nitrogen and 31 percent of
annual phosphorus loads to the Bay. Increases in developed land have caused runoff loadings to
increase over the last 25 years (US EPA 2010), while those from other sources have declined.
Options for reducing runoff are limited, however, as stormwater management is among the most
expensive ways to reduce nutrient loading in the Bay (Jones et al. 2010).

Due to these limitations on further reductions from WWTPs and MS4s, reductions in agricultural
loadings are necessary for meeting the objectives of EO 13508. Agriculture now contributes 46
percent of the yearly nitrogen load to the Bay and 48 percent of the yearly phosphorus load (US
EPA 2010). Agriculture may present more cost-effective alternatives for nutrient loading
reductions than do WWTPs and MS4s. The World Resources Institute estimates that the cost of
reducing a pound of nitrogen from agriculture is one-tenth of the cost of eliminating the same
pound from a WWTP and one-fortieth the cost of eliminating the pound from an MS4 (Jones et
al. 2010).

With the exception of some CAFOs, the CWA does not give EPA authority to regulate most
agricultural operations. In the absence of authority to impose mandatory reductions or to force
agricultural sources to pay for the right to pollute, the only option left is to pay them to reduce
their pollution. The question then becomes one of financing. One alternative would be to use
state or federal tax revenues to pay nonpoint sources to undertake reductions. Such an approach
would be similar to existing USDA and state programs to pay for land set asides and adoption of
BMPs. Another option is to place stricter limits on WWTPs and MS4s, but to allow those
sources to meet their obligations by paying agricultural sources to adopt BMPs, i.e., encourage
water quality trading.

In theory, a water-quality trading program has the advantage of cost effectiveness; the
environmental objective is met and competition among buyers and sellers ensures that those
sources that can reduce their emissions at the lowest cost do so. In addition, an offset program
creates incentives for technological innovation to lower the cost of reducing nutrient loading. In
contrast, the standard rule-making process can be slow and conflict-ridden as regulated entities
try to affect regulatory design. A trading system also requires less knowledge by the regulator of
abatement costs to be effective.9

The academic economic literature has long noted the advantages of market-based incentives such
as fees, taxes, and pollution trading markets (Pigou 1920; Crocker 1966; Dales 1968;
Montgomery 1972; Kneese and Schultz 1975). However, it has also noted the limitations of
trading systems and the differences between market-based incentives in theory and in practice
(see, e.g., Hahn 1989). In particular, there is a voluminous literature identifying potential
problems in water-quality markets (e.g., Woodward 2002; King and Kuch 2003; Abdalla et al.
2007; Breetz and Fisher-Vanden 2007; Morgan and Wolverton 2008; Smith et al., 2010;
Shabman and Stevenson 2011; Horan and Shortle 2011).

Based on observations of existing water quality trading programs, many researchers have come
to pessimistic conclusions. Abdalla et al. (2007) find "the physical and regulatory context for

9 For more details regarding the theoretical advantages of market based mechanisms such as pollution trading, see
Freeman and Kolstad (2007).

-------
agricultural NPS pollution does not match the conditions that economic theory suggests are
needed for widespread trading to occur," while Woodward (2002) concludes that "programs to
date have not generated a convincing record. Given the investment of agency time required to set
up and monitor water quality trading programs, there is certainly reason to wonder if this
regulatory approach is worth the effort." Horan and Shortle (2011) advise that "it is clearly
incorrect to assume that the gains promised by idealized textbook models will simply materialize
in similarly constructed water quality trading markets: if you build it, they may not come."

This pessimism generally stems from the cost of trades. It takes a considerable effort to develop
and maintain the institutions required to operate an efficient and verifiable market. A trading
program in environmental performance, like any market, needs a number of features in order to
function smoothly. These features include:

• enough traders to prevent any individual from exercising monopolistic power;

• substantial differences in abatement costs among potential traders;

• ways for prospective traders to identify each other;

• systems for monitoring, recording, and certifying trades; and

• a legal framework to enforce contracts.

Establishing such features can be costly. Costs in these categories are collectively referred to as
"transaction costs," and if they are too high, they can erode potential efficiency gains from
trading, or even make trading infeasible.

For market efficiency to be realized, trades must occur. Due to high transaction costs, water
quality trading markets have had low trading volumes between point and non-point sources.
Ribaudo and Gottlieb (2011) find that no trades at all had occurred in 11 of 15 water quality
markets allowing nonpoint source trades, and only two markets had more than four trades.
Morgan and Wolverton (2008) report similar results.

There is a tension between reducing transaction costs and maintaining the environmental
integrity of a trading program. In the next section we highlight several policy decisions necessary
for implementing a trading program, discussing implicit tradeoffs between program cost and
environmental benefits.

Tradeoffs in policy design

There is not a simple policy prescription that solves all potential issues with a water
quality trading program. Rather, practical policy design involves balancing a program's
effectiveness at reducing costs with its effectiveness at improving water quality. Here we
highlight the tradeoffs involved in the key policy decisions.

Limits on eligible suppliers of credits

The larger the set of sources eligible to sell offset credits, the greater the possibility for
efficiency-improving trades, and the lower the expected cost of purchasing a credit. The less
stringent the eligibility requirements, however, the greater is the likelihood that some offsets are

-------
not true reductions in nutrient loadings. Leakage and additionality are two eligibility concerns
that can result in trades that actually increase net discharges.

Leakage. Leakage (the term "slippage" is also used to describe the same phenomenon,
particularly in literature on agricultural land retirement programs) refers to the potential for
conservation programs to give farmers an unintended incentive to convert land to agricultural use
or to increase the intensity of crop or livestock production on existing agricultural land. Leakage
typically occurs due to indirect price effects. In the case of the USDA Conservation Reserve
Program (CRP), for example, researchers have found evidence that paying farmers to idle
cultivated land causes an increase in commodity prices that makes it profitable to bring new land
into production (Wu 2000).10

Due to the relatively small size of crop production in the Chesapeake Bay watershed, the effects
of an offset program on crop commodity prices is likely to be limited. Given the vertical
integration of the livestock industry, however, and the regional demand at the processor level for
intermediate livestock products, an offset program could generate significant leakage. If a farmer
were to reduce livestock production as a result of selling a credit, the integrator may respond by
providing additional animals to another farm.11

Another problem can arise from the price of the offsets themselves. Farmers will only sell offsets
if it is profitable to do so. Therefore, the existence of an offset market may increase the
profitability of cropland relative to other uses, such as woodland. Absent eligibility requirements
on previous land use, an offset program could induce landowners who would have earned
slightly less profit from other land uses to convert land to crop production and sell offsets - with
the potential net effect of increasing discharges. As noted by the USDA National Resource
Conservation Service, nationwide "land use is .... surprisingly dynamic, with .... constant shifts
... among cropland, pasture, range, and forest land" (US NRCS 1997). Thus, this type of leakage
is theoretically possible, although determining its potential magnitude would require further
research.

Such policy-induced land use changes have been a concern in other contexts as evidenced by the
fact that USDA agricultural programs often contain provisions to avoid land use changes arising
from the program payments themselves. Commodity program payments, for example, are often
made on the basis of historical "base acres." These provisions reduce the incentive to convert
land to program crops in order to obtain more payments. Similarly, CRP provisions limit
eligibility on the basis of previous land use to reduce incentives to convert forest or pasture to
cropland, and then back to CRP.

Leakage presents a problem only if there is not a strict cap on all sources of discharges to the
Bay. With a cap in place, land use changes would not affect overall discharges since these would
be limited by the total availability of permits. An alternative to a cap would be to require all new

10 The magnitude of leakage in the CRP is controversial; see Roberts and Bucholtz (2005, 2006) and Wu (2005).

11 Most livestock production in the Chesapeake Bay watershed follows a model whereby integrators provide young
animals under contract to farmers who feed and care for them, returning them to the integrator after a specified
amount of time. For details of these arrangements in the poultry sector see MacDonald (2008).

-------
or expanding sources (including nonpoint sources) to obtain offsets. This requirement would
eliminate the incentive to convert land in order to sell offsets, but may not be politically feasible.

An alternative approach, at least for crop agriculture, would be to base eligibility on previous
land use. While reducing the incentive to convert land, this approach has the disadvantage of
eliminating potentially beneficial trades. Landowners convert woodland or pasture to cropland
for many reasons. If farmers decide to clear woodland or convert pasture due to an increase in
commodity prices, for example, it may be efficient and environmentally sound to pay them to
undertake BMPs on this newly cultivated land.

Another approach would be to recognize that a certain percentage of gains from nonpoint source
offsets will be lost due to leakage, and to incorporate this percent into a trading ratio between
point and nonpoint sources. The downside to this approach is that it effectively makes it more
expensive to buy offsets from all nonpoint sources, thus discouraging some potentially efficient
trades.

Additionality. A strict cap is also known as an absolute baseline; there is one limit set on the
total amount of permissible emissions, from which all actual emissions are debited. In contrast,
an offset program typically involves relative baselines; emissions from each source are debited
from hypothetical pollution levels that would have occurred at that source. Relative baselines are
costly to implement since counterfactual emissions must be evaluated before each trade (Jepma
2003; Rentz 1998).

Relative baselines raise the question of "additionality." Are claimed reductions offered as offsets
really additional to what the source would have been doing in the absence of an offset program?
For example, landowners may already be engaged in offsetting behavior for a number of reasons
(they find it profitable, they are paid from some other source such as state or USD A programs,
they choose to be good stewards, etc.). Suppose that a WWTP purchases offset credits from such
a landowner instead of reducing its emissions. If the offsets are not additional, then emissions are
exactly the same as what they would have been without the trade. Both the WWTP and the offset
seller behave exactly as they would have otherwise. While money may change hands, it is a mere
"paper trade"; loading is not reduced.

"Adverse selection" arises if those sources most likely to sell offset credits are those that would
have been most likely to undertake the action in the absence of a program. Adverse selection can
exacerbate the additionality problem. Consider, for example, a policy of rewarding farmers who
maintain riparian buffer zones near streams on their property. For many farmers establishing a
riparian buffer is costly due to foregone earnings from stocking more animals or planting crops
closer to the water. For others, land near a stream might not have been suitable for grazing or
cultivation anyway. As it costs nothing for such farmers to maintain natural vegetation in such
areas, they will do so. If participation in the program is voluntary it is precisely those farmers
who would have undertaken best management practices anyway who have the greatest incentive
to sell offsets.

-------
As with leakage, additionality is a theoretical issue whose practical relevance remains an open
question. Further research on this topic in the context of the Chesapeake Bay would be useful.
The following considerations suggest that it may be imprudent to dismiss this issue:

• Diversity with respect to farm landforms, size, crops and/or animals raised,
ownership, and other factors likely imply a wide range of BMP costs, leading to
potential adverse selection problems.

• The Acid Rain program was not always a pure cap and trade policy. Initially some
unregulated firms were allowed to "opt-in" to the program. Evidence suggests that
those who did so were the emitters whose allotted credits exceeded what they
would have emitted anyway (Montero 1999). Similarly, it has been argued that
the possibility of receiving payments under climate change programs for
industrial plant closures encouraged China to "reduce" emissions from plants that
otherwise would not have been economically viable (Wara 2006).

• Results from EPA's evaluation of the experimental Rural Clean Water Program
(Gale et al. 1993) suggest that farmers participate in programs only if they believe
the programs provide a material benefit to their operations.12 They also only adopt
and continue those practices that they believe to be profitable. Hence, adverse
selection may be a problem not only with regard to which farmers participate, but
also with regard to what practices (among a suite of potential practices) are
implemented on those farms that do participate.

The ideal way to eliminate additionality problems is to have a strict cap in place. If such a cap is
not politically or legally feasible, there are second-best approaches. Similar to leakage, one
approach is to recognize that a certain percentage of offsets are not likely to be truly additional
and to increase the trading ratio (or, equivalently, set a side a proportion of traded land into a
reserve) to account for this. As before, this approach has the disadvantage of treating all nonpoint
sources equally, thus discouraging trading with those sources for which additionality is not an
issue. Nor is determining the trading ratio a trivial problem (see Montero 2000).

Another potential means of reducing the additionality problem is to require that sellers undertake
a minimum amount of BMPs before they are allowed to sell additional offsets. This requirement
could help to ensure that offsets sold are truly additional. To offer credits in Virginia, for
example, farmers are required to implement soil conservation and nutrient management plans,
plant cover crops on cropland, exclude livestock from streams, and install riparian buffers before
selling credits for any further reductions in discharges. Pennsylvania and Maryland have similar
requirements. Such requirements come at a cost, however. The minimum requirements
exacerbate the adverse selection problem by making it more costly for some potential sellers to
participate in the trading market, and they may decide to opt out. The sellers for whom
reductions would have been costly, and, for that reason truly additional have the most incentive
to opt out.

The difficulty in determining a priori the significance of leakage and additionality on pollution
reduction from non-point sources of pollution argues for the adoption of Adaptive Management

12 This program included projects located in the Chesapeake Bay watershed where the major pollutants of concern
were nutrients.

-------
in any trading program. At the core of Adaptive Management is an effective tracking or
monitoring system for activities that potentially contribute to nutrient losses to the environment
(NRC 2011). Although EPA and the states do have some Bay monitoring systems in place, they
would need to be improved in order to identify and measure the consequences of leakage and
additionality from trading.

Multiple payments for one offset

A discharge-trading system in the Chesapeake Bay would not take place in an
environmental policy vacuum. There are other programs currently in place (e.g., the
Conservation Reserve Program) that pay non-point sources for actions that would have the effect
of improving Bay quality. In addition, there are proposals for programs providing payments for
carbon sequestration, wildlife habitat, and other ecosystem services. In this context, the question
arises as to what extent a nonpoint source should be permitted to sell a discharge offset credit for
an activity that is remunerated from another program.

The environmental effect of allowing multiple payments depends on whether the additional
impact of the offset payment induces a nonpoint source to change its behavior. Suppose that the
amount of money received from other sources were sufficient to induce a nonpoint source to
undertake a BMP. If a point source then purchases a credit based on that BMP, total emissions to
the Bay would rise relative to the case in which the point source were not permitted to purchase
the credit.

If, however, the amount of money received from other sources is not sufficient to induce the
nonpoint source to undertake the BMP without the additional payment for the offset credit, then
the discharge reduction is additional and results in real environmental improvement to
compensate for the increased emissions of the purchasing point source. A relatively lax policy
decision regarding payments from multiple sources is likely to result in less environmental
improvement, while a strict policy will result in higher costs since some potentially efficient
trades will be prohibited. The current diversity of state approaches to trading could provide a
basis for researching the importance of these tradeoffs.

Accuracy of geographic trading ratios

The premise motivating trading programs is that dischargers with high abatement costs
can trade obligations with sources with lower costs. The larger the pool of potential trading
partners, the greater the expected differences in costs among them, and the greater the cost
savings from trading. However, the greater the distance between would-be trading partners, the
more likely that the discharges they trade would result in different damage.

For the purposes of the TMDL, the Chesapeake Bay is divided into 92 separate tidal segments.
Under currently envisaged rules and policies, trades cannot result in violation of a TMDL in any
segment. At present, 89 of the 92 segments are impaired (FLC 2010). This constraint is likely to
limit the scope of potential trades, and with it, possible cost savings.

-------
The theoretical solution to the problem of differential impacts arising from discharges from
different locations is to define a system of impact-based trading ratios. While the underpinnings
of market design with multiple geographical restrictions were developed by Montgomery (1972),
it is so complex that such a design has never been implemented in practice (Freeman and Kolstad
2007). The Long Island Sound and Lower Boise River water quality trading programs do specify
a number of different trading ratios between different dischargers depending on location and
other factors (Morgan and Wolverton 2008); however, they do not use the theoretically
appropriate rates. The Chesapeake Bay Watershed comprises 64,000 square miles, some four
times greater than that of the Long Island Sound and 16 times greater than that of the Boise
River. Complications of determining appropriate trading ratios are thus likely to be
correspondingly greater.

The problem of devising accurate trading ratios can be exacerbated by the incentives of the
regulators themselves. The pressure by participants and program goals can make administrators
inclined to use simple uniform trading ratios to facilitate trades at the expense of quality.
Salzman and Ruhl (2007) document this problem in the context of the wetlands banking
program. There, high quality wetlands were often replaced by lower quality ones on an even
acre-for-acre basis.

If calculating the ideal trading ratio is infeasible, policy-makers are faced with a decision
regarding how many ratios to use and how much of a margin of error to incorporate. The cost of
using the wrong trading ratios is that trading may reduce environmental quality if the damage
from increased discharges at one location is greater than the benefits from the offsetting
reductions at another. Attempting to fine-tune ratios, however, may eliminate many potential
cost-saving trades. Multiple trading ratios would entail a potentially costly and cumbersome
certification process prior to each trade, leading to increased uncertainty among traders and
increased transaction costs. Nevertheless, the certification process is necessary for the integrity
of the trading program.

Requiring relatively large reductions from nonpoint sources in exchange for a unit of increased
emissions from a point source helps to ensure that the trade does not result in net environmental
damage. Stringent ratios increase the cost of the program, however, by discouraging some
potentially efficient trades.

Monitoring and Liability

Monitoring discharges and credit-generating activity is an essential component of a
trading program due to the possibility of a credit seller not complying with contract terms. In
conventional markets the consumer has an obvious stake in the quality of the traded good, thus
giving sellers a strong incentive to develop a reputation for quality, even in the absence of third
party monitoring. In environmental markets, this incentive is notably absent. Buyers only care
about quality inasmuch as they fear the consequences of a government audit. If the seller of a
credit did not meet the terms of its contract and did not reduce its own contribution to loading by
the amount it had promised, the buyer of the credit might then find itself liable for penalities.

-------
Monitoring systems are in place for WWTPs. With the possible exception of a few easily
observed practices such as establishing riparian buffers, however, it is notoriously difficult to
verify BMPs for non-point sources. A recent study conducted under the Conservation Effects
Assessment Project that examined farmer compliance with nutrient management plans, for
example, found that about 80 percent of farmers did not implement the practices that were
required or were not aware that they were not in compliance with the conditions of the plan
(Osmond 2010). In the case of the wetlands banking program, the benefits were severely limited
by lack of effective enforcement (Salzman and Ruhl 2007; Shabman 2004).

An effective trading system must be able to punish participants who are not in compliance. A
practical issue concerns whom to punish if a contracted offset is not delivered. It is useful to
consider why a contract may not be fulfilled. If the cause is pure force majeure, then the
assignment of liability is not important as long as someone is obliged to provide the missing
offsets. In this case, requiring a trading ratio greater than one with the extra offset going into a

state-held reserve may be a solution.

If the offset provider can undertake some not-easily-monitored effort to reduce the probability
that an offset fails, then assignation of liability is important. Assigning liability to the purchaser
may create conditions of moral hazard since it does not give the seller incentive to undertake
such preventative action, whereas assigning it to the provider does.

With respect to the last point, purchasers could presumably initiate civil cases against offset
credit providers who did not fulfill the terms of their agreements. The uncertainties inherent in
such litigation result in less of an incentive to undertake preventative action than direct liability
for the seller since courts would first need to find the purchaser in violation, the purchaser would
need to decide that it was worthwhile to pursue the case in civil court, and the court would need
to rule against the seller.

In principle, concern about one's reputation could induce offset credit providers to meet the
terms of their agreements. It is not clear, however, that landowners who had defaulted on earlier
agreements could be effectively "blacklisted" or that contracting opportunities would arise often
enough to make this a compelling consideration.

A key policy decision is thus where to assign liability if offsetting discharge reductions fail to
take place. In most established U.S. water quality trading programs, point sources are liable for
non-compliance in all trades with non-point sources.14 Currently, if a nonpoint source does not
fulfill its reduction obligation, EPA only has authority to undertake enforcement actions against
the point source NPDES permit holder who purchased the credit. This potential liability may
make point sources reluctant to enter into an agreement with a nonpoint source (see, e.g.,
Ribaudo and Gottlieb 2011), thereby reducing the scope for potentially efficient trades and
increasing the cost of the program.

13 Pennsylvania sets aside 10 percent of credits into a reserve pool to cover possible failure of offsetting activity
(Selman, et al. 2009).

14 Morgan and Wolverton (2008) provide examples of several liability arrangements in water quality trading
programs.

-------
Placing Clean Water Act liability on credit sellers could reduce transaction costs associated with
ensuring compliance, thus encouraging more trades. In practice, such an approach could be
problematic if sellers are sufficiently small that the expected punishment is less than the cost of
ensuring that the offset is successful. An alternative approach would be to set up reserve
accounts so that if some promised discharge reductions do not occur additional credits may be
drawn upon (see, e.g., Morgan and Wolverton 2008), although this strategy would not solve
moral hazard issues. Liability could also be shifted to entities that aggregate offsets from many
different suppliers.15 Such aggregators would likely have a stronger concern for reputation than
individual offset providers, and have greater assets to cover potential penalties. These features
may give them an increased incentive to monitor the behavior of the suppliers.

Command and control requirements

NPDES-permitted point sources must meet their technology-based emission limits and
cannot purchase credits to exceed them. Although EPA currently recognizes the possibility that
point sources can purchase credits to meet water quality based effluent limits, such technological
restrictions potentially reduce the cost effectiveness of a trading program. They do, however,
provide a margin of safety regarding environmental benefits since the ultimate performance and
reliability of nonpoint source offsets is not yet established. The crucial question for policy
makers is how to balance the cost of technological requirements (including mandatory BMPs for
nonpoint sources) with their benefits in terms of relatively assured environmental improvements.

Population and economic growth

One of the great challenges in meeting the objectives of EO 13508 is that nutrient
loadings must be reduced against a backdrop of increasing population and economic activity in
the watershed. One means of addressing this issue is to require that all new and expanding
sources of discharges offset their emissions with reductions elsewhere.

Fundamental to the success of such an approach is the definition of "new and expanding." Due to
population growth, biofuel policy, and increased incomes in countries such as China and India,
demand for agricultural outputs produced in the Bay watershed are likely to increase over time.
If the definition of new and expanding sources who require offsets does not include crop and
livestock producers, then even the reductions from a perfectly functioning offset program for
WWTPs and MS4s may be insufficient to improve water quality in the face of increased
emissions from these other sources.

Potential Steps Forward

In this section we discuss some alternatives to the point-nonpoint source water quality
programs currently in development. Based on the previous section, it is clear that many potential
problems with offset trading stem from the lack of a strictly enforceable water shed-wide limit on

15 Since the CWA restricts EPA enforcement authority to point sources (i.e., the credit purchasers), such shifts in
liability would need to be enforced by States under their laws.

13

-------
loadings from all sources. The theoretical benefits of a cap-and-trade program (attaining an
environmental goal at the least cost) crucially depend upon the existence of the "cap." The ideal
option would therefore be to require all sources (new and existing, point and nonpoint) that may
potentially participate in a trading system to have a permit before discharging.

Recognizing that this policy may not be politically or technically feasible, however, we identify
some alternatives that may increase the likelihood that a trading program achieves net
environmental benefits. These options are by no means definitive solutions; some may require
significant regulatory or even statutory changes to be feasible.16 Rather, they are intended to
stimulate further discussion regarding possible market-based mechanisms to reduce discharges to
the Bay that are less susceptible to the policy dilemmas identified in the previous section.

Begin with expanded point source trading

Many potential problems of water quality trading arise from the inclusion of non-point
sources in the trading system. Many of the benefits of trading could be captured while avoiding
the problems mentioned above if trading occurs only between point sources. Since a cap on
WWTPs and MS4s alone would be insufficient to meet water quality goals, the universe of point

sources eligible for trading could be expanded to include CAFOs. Such a program would have
the following potential advantages.

• Precedent. As mentioned above, most of the successful water quality trading
programs only allowed trades between point sources.

• Scope. Just under 40 percent of the Bay's nutrient loading comes from WWTP
and CAFO emissions. Thus a cap over these two sectors could go a long way to
reaching the programs overall emission goals.

• Cost effectiveness. Since many CAFOs currently face comparatively few
constraints on discharges, there is likely to be considerable scope for relatively
low-cost discharge reductions, thereby creating the opportunity for large gains
from trade with WWTPs.

• Additionality. Since point sources fall under EPA's statutory authority under the
Clean Water Act, in principle it could enforce a cap on the entire trading sector,

thus eliminating many environmental concerns implied by additionality.

• Reduced leakage. Price-induced leakage within the two sectors would not be a
problem due to the cap. Moreover, such a system would likely lead to a reduction
(or slower growth) in animal production, thus reducing pressure on local feed
prices and reducing farmer incentives to apply fertilizer or cultivate new crop
land.

16 Treatment of the legal requirements is beyond the scope of this paper. None of this discussion conveys an EPA
position regarding interpretation or implementation of any of the relevant rules or statutes.

17 Existing CAFO rules may need to be modified to allow trading.

18 Many non-trivial legal issues would need to be resolved to expand EPA regulation, i.e., require a greater number
of CAFOs to have permits, and thus enforcement authority over CAFOs.

-------
• Feasible monitoring. Monitoring CAFOs for compliance is likely to be
significantly less expensive and easier than monitoring of non-point sources.

Although such a point-source trading program may realize significant reductions, it may not be
sufficient to attain the desired reductions. It could therefore be complemented by an expansion of
programs such as EQIP, WRP, or the CRP to pay for the production of environmental amenities
that serve to reduce loadings from nonpoint source agriculture. Such programs can reduce the
overall watershed-scale nutrient burden without running the risk that increased loadings from a
particular point source are not being offset by non-additional reductions.

Finally, the experience gained from point-source-to-point-source trading can inform broader
trading program. It may be relatively easy to resolve logistical issues such as determining
geographic trading ratios in point-source to point-source trades. The lessons learned in the
process of bringing the limited program to success could be applied to an eventual expansion to
make non-point sources eligible for participation. Beginning with a point-source only trading
program would also provide time to conduct further research regarding the practical importance
of leakage and additionality in the non-point source sector.

Focus on Integrators

Many farmers in the Chesapeake Bay watershed produce for a few "integrators" - large
companies that purchase farmers' outputs while providing them with inputs. If it is infeasible to
regulate or otherwise involve the small farmers in offset markets, it may be feasible to address
the integrators.

Reduced to its biochemical essence, the problem with nutrient pollution in the Chesapeake Bay
watershed is the imbalance between in and out-flows of reactive nitrogen and bioavailable
phosphorus—that is, more nutrients flow into the confines of the Bay watershed than are
removed through agricultural products exported out of the Bay watershed, immobilized in new
soil, or converted through denitrification. With the exception of atmospheric deposition, the
nutrients originate as protein content in livestock feed, as chemical fertilizer, and biosolids
(sewage sludge) imported from areas outside the Chesapeake Bay watershed (NRC 2011). The
soils and the plants and animals within the watershed can accommodate only so much nitrogen
and phosphorus based upon the climate, exports of food and feed out of the basin or watershed,
or denitrifying capacity of the land. The excess represents the nutrient imbalance that, depending
upon the severity of storm events, becomes the nutrient load to the Bay. Moreover some
nutrients embodied in crop or livestock produced in the Chesapeake Bay watershed are recycled
through sewage sludge that is returned to the land. This recycling adds to the imbalance resulting
from imported nutrients.

Since grain brokers or livestock integrators in large measure control the flow of feed into the Bay
watershed, they control a major control point for nutrients. By reducing feed imports, they
thereby also reduce total animal production - and wastes. Moreover, there are relatively few
grain brokers or livestock integrators, making tracking of nitrogen and phosphorus material
flows easier.

-------
Thus, a regulatory option could be to establish a "deposit-refund" system. Whenever any
nutrients - feed or fertilizer - are introduced into the watershed, the entity introducing them, in
this case the integrators, would need to pay a deposit. A portion of the deposit would be refunded
based on the amount of nutrients that are not released to the environment. The size of the deposit
could be reduced if less fertilizer were used, feed reformulated to reduce nitrogen and
phosphorus content, or actions undertaken on the land to increase net denitrification or reduction
of bioavailable phosphorus, such as wetland rehabilitation or restoration. The size of the refund
could also be increased if more nutrients in animal waste were processed to a chemical form that
can be shipped out of the watershed as fertilizer.

Alternatively, rather than charging for net introductions of nutrients, integrators could be
assigned numerical limits on the net amounts of nutrients they introduce. Such allocations would
form the basis of an environmental market in net nutrients in which integrators might trade with
one another, or with other polluters in the watershed.

While such a system might be more administratively tractable than one dealing directly with
myriad farmers, it does present some challenges. One is that the approach we have described
begs the question of leakage. The advantage of imposing the burden of compliance on the small
number of large integrators is that it would be easier to keep track of what they are doing than of
the farmers with whom the contract. At present integrators contract with farmers because there
are cost advantages to them of doing so. If more independent farmers could avoid regulatory
scrutiny, those cost advantages might be eroded, if not eliminated. It would be important to
verify that this sort of atomization of the industry would not occur if a program of the type we
have suggested were implemented.

Another related consideration is that, while regulators might shift the technical burden of
compliance to large integrators, a large part of the program would still rely on verifying the
performance of individual farms. It would still be important to monitor farm-level performance
and enforce compliance.

Conclusion

There is a widespread consensus regarding two facts: the water quality of Chesapeake
Bay is seriously degraded, and reductions in agricultural discharges are essential for attaining
adequate improvement. The difficult question is how to reduce these discharges short of direct
regulation of agriculture.

One approach would be to severely curtail allowances on current NPDES holders, especially
WWTPs. Although it would be unrealistic for WWTPs to attain these reductions at a reasonable
cost, in lieu of reducing their own discharges they could be offered the opportunity to pay
agricultural sources to undertake relatively inexpensive discharge-reducing BMPs.

In this paper we examine the key policy decisions required to implement such an offset trading
program in practice. We find that many face a dilemma. On the one hand, program rules can be
specified in such a way that facilitates trade. The danger with this approach is that trades may
result in little, if any environmental improvement. At the extreme, this path could result in a

-------
massive transfer of resources from municipalities to agricultural producers while water quality in
the Bay continues to deteriorate.

On the other hand, program rules can be specified in a way that minimizes the risk of non-
environmentally beneficial trades. The danger with this approach is that the cost of safeguards
and administrative hurdles is so great that few if any trades occur. At the extreme, this path could
result in a massive investment by municipalities in further treatment. Due to the relative impact
of municipalities in total discharges, this treatment alone is likely to be insufficient to attain
water quality goals in the Bay.

Navigating policy between these two dangers is not a simple task. Nonetheless, there is a
common element at the root of most of the difficulties. The fundamental issue with the offset
program is that, unlike successful policies such as the Acid Rain trading program, there is no
strictly enforceable cap. We offer some tentative examples of how an offset program might be
modified to incorporate a cap.19 Programs that focus on trading between WWTPs and CAFOs
may be promising, as well as those that focus on the possible role of integrators in the
management of nutrients and animal wastes.

References

Abdalla, C., Borisova, T., Parker, D., Saacke Blunk, K. (2007) Water Quality Credit Trading and
Agriculture: Recognizing the Challenges and Policy Issues Ahead. Choices 22 (2): 117-
124.

Batuik, R. (2010) Introduction to TMDLs with Examples from the Chesapeake Bay

http://www.epa.gov/ocem/frrcc/pdf/2010_rich_batiuk_introduction_to_tmdls.pdf).

Breetz, H.L., Fisher-Vanden, K., Jacobs, H., Schary, C. (2005) Trust and Communication:
Mechanisms for Increasing Farmers' Participation in Water Quality Trading. Land
Economics 81(2): 170-190.

Crocker, T.D. (1966) The Structuring of Atmospheric Pollution Control Systems. In The

Economics of Air Pollution Wolozin, H. ed. New York: W.W. Norton.

Dales, J. H. (1968) Pollution, Property and Prices. Toronto: University Press.

Federal Leadership Committee (FLC) (2010) Strategy for protecting and restoring the
Chesapeake Bay watershed. http://executiveorder.chesapeakebay.net/file.axd?file=
2010%2f5%2fChesapeake+EO+Strategy%20.pdf.

Freeman, J., Kolstad, C.D. (2007) Prescriptive environmental regulations versus market-based
incentives. In Moving to Markets in Environmental Regulation: Lessons from twenty
years of experience, Freeman, J., Kolstad C.D. eds. New York: Oxford University Press.
Ch. 2.

Gale, J.A., Line, D.E., Osmond, D.L., Coffey, S.W., Spooner, J., Arnold, J.A., Hoban, T.J.,
Wimberly, R.C. (1993) Evaluation of the Experimental Rural Clean Water Program.
National Water Quality Evaluation Project, NCSU Water Quality Group, Biological and
Agricultural Engineering Department, North Carolina State University, Raleigh, NC.
EP A-841 -R-93 -005.

19 These ideas are discussion points rather than pronouncements on what is permissible under current statutory
authority.

-------
Hahn, R. (1989) Economic prescriptions for environmental problems: How the patient followed
the doctor's orders. Journal of Economic Perspectives 3(2): 95-114.

Horan, R.D., Shortle, J.S. (2011) Economic and Ecological Rules for Water Quality Trading.
Journal of the American Water Resources Association 47(1): 59-79.

Jepma, C.J. (2003) The EU emissions trading scheme (ETS): how linked to JI/CDM?. Climate
Policy 3, 89-94.3.

Jones, C., Branosky, E., Selman, M., Perez, M. (2010) How nutrient trading could help restore
the Chesapeake Bay. World Resources Institute Working Paper.

King, D.M., Kuch P.J. (2003) Will nutrient credit trading ever work? An assessment of supply
and demand problems and institutional obstacles. Environmental Law Reporter News and
Analysis 33(5): 10352-10368.

Kneese, A., Schultze, C. (1975) Pollution, Prices, and Public Policy. Washington, DC: The
Brookings Institution.

MacDonald, J. (2008) The Economic Organization of U.S. Broiler Production. Economic
Information Bulletin 38. Washington, DC: USD A Economic Research Service.

Montero, J.-P. (1999) Voluntary compliance with market-based environmental policy: Evidence
from the U.S. acid rain program. Journal of Political Economy 107(5): 998-1033.

Montero, J.-P. (2000) Optimal design of a phase-in emissions trading program. Journal of Public
Economics 75: 273-291.

Montgomery, W.D. (1972) Markets in Licenses and Efficient Pollution Control Programs.
Journal ofEconomic Theory 5: 395-418.

Morgan, C., Wolverton A. (2008) Water quality trading in the United States: trading programs
and one-time offset agreements. Water Policy 10(1): 73-93.

National Research Council (2011) Achieving Nutrient and Sediment Reduction Goals in the

Chesapeake Bay: An Evaluation of Program Strategies and Implementation National
Academies Press, Wsahington, DC.

Osmand, D.L. (2010) USDA water quality projects and the National Institute of Food and
Agriculture Conservation Effects Assessment Project watershed studies. Journal of Soil
and Water Conservation 65(6): 142A-146A.

Pigou, A.C. (1920) The Economics of Welfare. London: Macmillan.

Rentz, H. (1998) Joint implementation and the question of 'additionality': a proposal for a
pragmatic approach to identify possible joint implementation projects. Energy Policy 26:
275-279.

Ribaudo, M.O., Gottlieb, J. (2011) Point-nonpoint trading: Can it work? Journal of the American
Water Resources Association 47(1): 5-14.

Roberts, M.J., Bucholtz, S. (2005) Slippage in the Conservation Reserve Program or spurious
correlation? A comment. American Journal of Agricultural Economics 87(1): 244-250.

Roberts, M.J., Bucholtz, S. (2006) Slippage in the Conservation Reserve Program or spurious
correlation? A rejoinder. American Journal of Agricultural Economics 88(2): 512-514.

Salzman, J., Ruhl, J.B. (2007) "No net loss": Instrument choice in wetlands protection. In
Moving to Markets in Environmental Regulation: Lessons from twenty years of
experience, Freeman, J., Kolstad C.D. eds. New York: Oxford University Press. Ch. 12.

Selman, M., Greenhalgh, S., Branosky E., Jones, CY., Guiling, J. (2009)"Water Quality Trading
Programs: An International Overview" WRIIssue Brief No. 1: 1-16.

Shabman, L. (2004) Compensating for the impacts of wetlands fill: The U.S. experience with
credit sales. In Tradeable Permits: Policy evaluation, design and reform, Tietenberg, T.,

18

-------
Johnstone, N. eds. Paris: Organisation for Economic Cooperation and Development.
Ch.7.

Shabman, L., Stephenson, K. (2011) Rhetoric and reality of water-quality trading and potential
for market-like reform. Journal of the American Water Resources Association 47(1): 15-
28.

Smith, C.M., Peterson, J.M., Leatherman, J.C. (2010) Simulation of Factors Impeding Water
Quality Trading Market Performance. Paper presented at Agricultural and Applied
Economics Association Meeting.

US Environmental Protection Agency (2009). The next generation of tools and actions to restore
water quality in the Chesapeake Bay: A revised report fulfilling section 202a of
Executive Order 13508.

http://executiveorder.chesapeakebay.net/file.axd?file=2009%2fl I%2f202a+Water+Quali
ty+Report.pdf

US Environmental Protection Agency (2011) Chesapeake Bay Program.
http://stat.chesapeakebay.net/?q=node/59&quicktabs_3=2&quicktabs_10=2.

US Government Printing Office (GPO) (2009) Executive Order 13508: Chesapeake Bay
protection and restoration. http://www.gpoaccess.gov/presdocs/2009/DCPD-
200900352.pdf.

US Natural Resource Conservation Service (NRCS) (2010) 1997 National Resources Inventory:
Highlights. http://www.nrcs.usda.gov/technical/NRI/pubs/97highlights.html.

Wara, M. (2006) Measuring the clean development mechanism's performance and potential.
UCLA Law Review 55: 1759-1803.

Woodward, R. T. (2003) Lessons about effluent trading from a single trade. Review of
Agricultural Economics 25(1): 235-245.

Wu, J. (2000) Slippage effects of the conservation reserve program. American Journal of
Agricultural Economics 82(4): 979-992.

Wu, J. (2005) Slippage effects of the conservation reserve program: Reply American Journal of
Agricultural Economics 87(1): 251-254.

19

-------