fertile ground - world resources institutepdf.wri.org/fertile_ground.pdf · fertile ground:...

F E R T I L E G R O U N D :N U T R I E N T T R A D I N G ’ S P O T E N T I A L T O C O S T -

E F F E C T I V E L Y I M P R O V E W A T E R Q U A L I T Y

P A U L F A E T H

W O R L D R E S O U R C E S I N S T I T U T E

W A S H I N G T O N, D C2 0 0 0

R O B E R T L I V E R N A S HS E N I O R E D I T O R

H Y A C I N T H B I L L I N G SP R O D U C T I O N M A N A G E R

A G R I C U LT U R A L R E S E A R C H S E R V I C E ,U S D AC O V E R P H O T O

Each World Resources Report represents a timely, scholarly treatment of a

subject of public concern. WRI takes responsibility for choosing the study

topics and guaranteeing its authors and researchers freedom of inquiry. It

also solicits and responds to the guidance of advisory panels and expert

reviewers. Unless otherwise stated, however, all the interpretation and

findings set forth in WRI publications are those of the authors.

Copyright © 2000 World Resources Institute. All rights reserved.

ISBN: 1-56973-197-7

Library of Congress Catalog Card No. 00-104473

Printed in the United States of America on chlorine-free paper with recycled content of 50%, 20% of which is post-consumer.

FERTILE GROUND: NUTRIENT TRADING’S POTENTIAL TO COST-EFFECTIVELY IMPROVE WATER QUALITY

iii

Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . v

Foreword. . . . . . . . . . . . . . . . . . . . . . . . . . vii

1 INTRODUCTION . . . . . . . . . . . . . . . . . . 1

Is Trading the Answer? . . . . . . . . . . . . . . . . . . . . 2

How Does Nutrient Trading Work?

An Illustration . . . . . . . . . . . . . . . . . . . . . . . . 3

2 WATER QUALITY TRENDS ANDISSUES. . . . . . . . . . . . . . . . . . . . . . . . . . 5

Causes of Impairment. . . . . . . . . . . . . . . . . . . . . 5

Water Quality Policy Issues . . . . . . . . . . . . . . . . 9

3 WHAT ABOUT TRADING? . . . . . . . . . 13

Point-Nonpoint Trading . . . . . . . . . . . . . . . . . . . 15

The Potential for Pollution Trading. . . . . . . . . . 16

4 NUTRIENT TRADINGOPPORTUNITIES IN THREE WATERSHEDS OF THE UPPER MIDWEST . . . . . . . . . . . . . . . 19

The Cases: Nutrient Trading in Minnesota,

Michigan, and Wisconsin. . . . . . . . . . . . . . . 19

Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

Point Sources and Control Options. . . . . . . . . . 23

Point Source Controls and Costs . . . . . . . . . . . 24

Nonpoint Sources and Load Reduction

Options. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26

Total Load Estimates . . . . . . . . . . . . . . . . . . . . . 30

5 ECONOMIC RESULTS . . . . . . . . . . . . . 31

The Least-Cost Solution . . . . . . . . . . . . . . . . . . . 31

Policy Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32

6 SUMMARY ANDRECOMMENDATIONS. . . . . . . . . . . . 39

About the Author . . . . . . . . . . . . . . . . . . . 44

Notes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45

References . . . . . . . . . . . . . . . . . . . . . . . . 47

C O N T E N T S


iv

FIGURES

Figure 1. Proportion of Surface Waters that

Support Designated Uses . . . . . . . . . . 6

Figure 2. Fertilizer Use 1965–1995 . . . . . . . . . . 8

Figure 3. Map of Case Study Watersheds . . . . . 20

Figure 4. Distribution of Point Source

Discharges in Case Study Areas . . . . 23

Figure 5. Phosphorus Loads by Source and

Watershed. . . . . . . . . . . . . . . . . . . . . . 30

Figure 6. Least-Cost Model Solutions . . . . . . . . 32

TABLES

Table 1. Leading Sources and Pollutants

of Impaired Waterways Affecting

Two or More Categories of Surface

Waters. . . . . . . . . . . . . . . . . . . . . . . . . . . 7

Table 2. Nitrogen and Phosphorus

Discharges to Surface Waters

From Point and Nonpoint Sources

in the United States. . . . . . . . . . . . . . . . 8

Table 3. The Top 20 States by Number of

Waterways Impaired by Nutrients . . . 12

Table 4. Sub Basin Vulnerability Indices

by Watershed . . . . . . . . . . . . . . . . . . . 22

Table 5. Number of Plants in Each

Watershed by Phosphorus

Treatment Type. . . . . . . . . . . . . . . . . . 25

Table 6. Land Use for Each Watershed. . . . . . 26

Table 7. Soil Erosion from Cropland and

Delivery Ratios by Watershed . . . . . . 29

Table 8. Cost for Phosphorus Control Under

Different Policies . . . . . . . . . . . . . . . . 33

Table 9a. Scenario Results—Minnesota

River Valley . . . . . . . . . . . . . . . . . . . . 34

Table 9b. Scenario Results—Saginaw Bay . . . . 34

Table 9c. Scenario Results—Rock River

Watershed. . . . . . . . . . . . . . . . . . . . . . 35


v

I have many people to thank at the stateagencies, on the advisory panels, andamong my research partners in each of the

three case studies covered in this study. Fromthem, I have learned a great deal about the vari-ous issues involved in water quality and market-based mechanisms.

I owe special thanks to Dave Batchelor of theMichigan Department of Environmental Quality.Dave began life as a regulator, committed toimproving the environment through enforce-ment of the law. While still maintaining thatcommitment, he is now working to improveenvironmental performance by identifying waysto appropriately increase flexibility and reducecosts. Dave has vast practical experience, which Ifound enormously valuable. In the course of thiseffort we spent many hours over the phone or abeer, thrashing through the issues, sometimesagreeing, sometimes not, but always workingtoward a better understanding and hopefully,better results.

Mahesh Podar at the U.S. Environmental Pro-tection Agency also deserves a special mention.His office not only provided financial supportfor this project, but Mahesh led EPA’s effort toexplore trading in the nation’s watersheds. Thatwork helped to get a lot of people thinking aboutthe possibilities and challenges and informedthe analysis presented here.

In each of the three case studies, I workedclosely with the lead state regulatory agency. Wealso formed advisory panels to review the workas it progressed. These panels provided invalu-able critiques and helped to bring the analysisdown to earth.

In Michigan, various individuals from severalgroups participated, including the MichiganAgricultural Stewardship Association, theMichigan State University Institute for WaterResources, MSU Extension Service, the NaturalResources Conservation Service, ConsumersEnergy Company, the Michigan Farm Bureau,the City of Wyoming, and the Huron RiverWatershed Council.

In Minnesota, lead partner Norman Senjem ofthe Pollution Control Agency was generous insharing his insights with me. Our advisorypanel in Minnesota included the MinnesotaRiver Joint Powers Board, the Minnesota RiverAgriculture Team, the Natural Resources Con-servation Service, the Minnesota Farm Bureau,the Metropolitan Council, the University ofMinnesota, the Institute for Agriculture andTrade Policy, the Board of Soil and WaterResources, the Land Stewardship Project, andthe Friends of the Minnesota River.

Our work in Wisconsin was completed in col-laboration with the Wisconsin Department of

A C K N O W L E D G M E N T S


vi

Natural Resources (WDNR) and the Rock RiverWatershed Partnership, a group of mostly point-source dischargers interested in cost-effectivemeans of attaining clean water objectives. SanjaySyal and Danielle Valvassori were the leads forthe WDNR; Roger Sherman was the chair ofthe Rock River Partnership.

Two key partners helped complete the analy-sis. At the MSU Institute for Water Resources,Jon Bartholic and Da Ouyang provided hydro-logic information needed to convert erosionand runoff estimates into loadings. One partic-ularly vexing problem was solved by MikeDoran of Strand Associates, who developedup-to-date cost curves for phosphorus controltechnologies.

Thanks go to those who participated in thepeer review process: Will Anthony, Dave Batche-lor, Rob Day, Debbie Farmer, Al Hammond,Tony Janetos, Richard Kashmanian, JonathanLash, Mahesh Podar, and Marc Ribaudo. BobLivernash provided review, editing, and produc-tion help. Kim Kusek and Andi Thomas deservethanks for providing programmatic assistance.Hyacinth Billings, as ever, did a stellar job ofseeing the report through production.

No effort can be accomplished without fund-ing, and here a special acknowledgment isreserved for the Joyce Foundation, the KelloggFoundation, the McKnight Foundation, the OakFoundation, and the U.S. Environmental Protec-tion Agency.

P . F .


vii

A fter nearly three decades of effort, theUnited States can take considerablepride in answering the mandates of the

1972 Clean Water Act. At great expense, muchof the municipal and industrial pollution ema-nating from pipes, or “point” sources, has beenreduced. But the largely voluntary effort toreduce “nonpoint” pollution from farms andelsewhere has been less successful. As we entera new millennium, progress in improving waterquality seems to have leveled off and our hopesof meeting the goals of the Clean Water Actseem to be receding.

Some 3,600 waterways across the nation arelisted as either impaired by nutrients or by algalblooms, which are typically caused by excessnutrients. The U.S. Environmental ProtectionAgency is locked in a contentious legal battlewith many states over additional requirementsto improve water quality.

Part of the problem is the complex nature ofwater pollution and the difficult challenge ofcontrolling nonpoint pollution. Clearly anotherpart of the problem is the question of costs.States are heavily constrained by cost considera-tions and reluctant to embark on another costlyround of stricter point source controls.

In Fertile Ground: Nutrient Trading’s Potentialto Cost-Effectively Improve Water Quality, the

Director of WRI’s Economics Program, PaulFaeth, proposes a way out of this dilemma.Using case studies in three states, he developsa framework to assess the cost-effectivenessof various policies and combinations of poli-cies to reduce phosphorus loads in specificwatersheds. In all three cases, policyapproaches incorporating nutrient tradingprograms are dramatically less expensive thanconventional approaches and can achieve com-parable benefits.

With this report, WRI continues over adecade of work on the environmental implica-tions of agricultural policies and practices inthe United States and abroad. In GrowingGreen: Enhancing the Economic and Environ-mental Performance of U.S. Agriculture, Faethand the team he managed integrated volumi-nous amounts of data into an analytic frame-work that assessed the profitability and envi-ronmental impacts of alternative croppingsystems. In Agricultural Policy and Sustainabil-ity: Case Studies from India, China, the Philip-pines, and the United States, Faeth and severalco-authors found that farm policies are usuallystacked against resource-conserving farmingmethods.

WRI maintains a continuing commitment inthis area. Using the methodology developed inthis report, a study is currently underway that

F O R E W O R D


viii

will focus on the heavily polluted “Dead Zone”in the Gulf of Mexico. WRI has also developed aprototype nutrient trading web site to help sup-port pilot nutrient trading programs across thecountry.

A N T H O N Y J A N E T O SS E N I O R V I C E P R E S I D E N T F O R P R O G R A M

W O R L D R E S O U R C E S I N S T I T U T E


1

T he Clean Water Act of 1972 has largelysucceeded in reducing pollution fromindustrial, municipal, and other “point”

sources. But the job of achieving clean water inthe United States is only about half done.Roughly 40 percent of the nation’s surfacewaters remain at least partially impaired, andsurveys suggest there has been little improve-ment recently.

Further progress now depends on developinglow-cost, innovative approaches that can effec-tively cut pollution from “nonpoint” sourcessuch as agriculture and urban runoff. But workstill needs to be done to test new approachesand to determine how they can be implementedto offer cost-effective and quantifiable improve-ments in water quality.

This study amounts to a search for such new,cost-effective approaches. It is based on twopremises. First, excessive nutrient loading is thesingle largest cause of water quality impairmentin the United States; and second, attacking the

worst pollution problem should yield significantbenefits.

Using case studies of watersheds in threestates—Michigan, Wisconsin, and Minnesota—this study develops a framework to explore thecost-effectiveness and environmental perfor-mance of various strategies to reduce phospho-rus loads in specific watersheds. It describes theimpact of point and nonpoint sources on phos-phorus loads, the technology options for reduc-ing those loads, and the costs of implementationusing various technologies and policies. Further,it compares several policy approaches with anestimate of the least-cost outcome, giving policy-makers an opportunity to compare costs relativeto a theoretical minimum.

In all three cases, the analysis finds that pol-icy approaches utilizing nutrient trading aredramatically less expensive than conventionalpoint-source performance requirements. In theMichigan case, for example, one variation utiliz-ing trading is estimated to cost about $2.90 perpound of phosphorus removed, compared toalmost $24 per pound for conventional point-source requirements.

Based on the analysis, we recommend that

• The Environmental Protection Agency(EPA) and state environmental agencies

I N T R O D U C T I O N / O V E R V I E W

1

Excessive nutrient loading is the singlelargest cause of water quality impairmentin the United States.


2

allow trading among industrial and munici-pal point sources whenever new nutrientreduction requirements are put in place.

• EPA and the U.S. Department of Agricul-ture provide much greater regulatory andfinancial support for state efforts to developpoint-nonpoint source pilot trading pro-grams and standards.

• Concentrated animal feeding operations betreated the same as industrial and munici-pal point sources, which means that theyshould face involuntary permittingprocesses and tight enforcement.

• Sufficient money from farm income subsidyprograms be shifted to conservation pro-grams so that environmental problemscaused by agriculture can be remediatedusing market-based approaches.

• Monitoring of water quality be vastlyincreased.

• Federal and state agencies explore opportu-nities for trading in large watersheds.

IS TRADING THE ANSWER?

Experience to date is limited, but it appears thattrading can be part of the answer to achieve bet-ter water quality. Trading has been successfullyused to obtain cost-effective reductions in otherareas of environmental concern, including lead,sulfur dioxide, and other air emissions. In addi-tion, trading is the leading option proposed toaddress the build-up of greenhouse gas emis-sions that could cause climate change.

When tighter standards are put in place, trad-ing increases flexibility and reduces costs byallowing dischargers with new obligations theoption of adapting their own facilities or financ-

ing comparable reductions by others. Tradingmakes it profitable for sources with low treat-ment costs to reduce their own effluentsbeyond legal requirements, generate a credit,and sell these credits to dischargers with highertreatment costs. This flexibility produces a lessexpensive outcome overall while achieving—and even going beyond—the mandated environ-mental target.

This analysis examines trading as an option toachieve phosphorus reductions in three water-sheds in the upper Midwest: the Saginaw Baywatershed in Michigan, the Rock River water-shed in Wisconsin, and the Minnesota River Val-ley. In each of these cases, variations of nutrienttrading are compared with traditional regulatoryand subsidy approaches. The results, presentedin Chapter 5, show great potential for cost sav-ings by applying market-based strategies towater quality regulations and agricultural con-servation subsidy programs. In Michigan andWisconsin, pilot trading programs are beingpursued, and in Minnesota, ad hoc trades haveoccurred. This report describes some of theexperiences to date with these efforts and others.

While trading is the economically preferredapproach for nutrient management problems,there are a number of significant concerns thatmust be addressed to ensure that trading is anenvironmentally effective and equitable solu-

When tighter standards are put in place,trading increases flexibility and reducescosts. This flexibility produces a lessexpensive outcome overall whileachieving—and even going beyond—themandated environmental target.


3

tion. Some of the measures that have been sug-gested also affect the cost-effectiveness of theoutcome, and these tradeoffs are important torecognize. A key issue in nutrient trading isequivalence. The biggest economic and environ-mental opportunity appears to be tradingbetween “point” and “nonpoint” sources. Butpoint discharges are relatively easy to quantifyand monitor, while nonpoint discharges areepisodic and cannot be directly measured, onlyestimated. However, nonpoint source reduc-tions can also have the additional advantage ofdecreasing sediment loads. The apparent bene-fits and risks of nutrient trading as a strategy tocost-effectively support the goals of the CleanWater Act point to the need for a cautious butalso encouraging approach. However, meansare available to increase the certainty of out-comes in ways that still produce major savings.This and other issues of program implementa-tion are discussed in Chapter 6.

We start by examining some of the majortrends and policy issues in Chapter 2 and thenexplain how trading could address some of theproblems in Chapter 3. In Chapter 4 wedescribe the three case study watersheds. Thissection highlights the water quality problemseach watershed faces, as well as the analyticalmethods we used to produce our results.

The intent of the analysis is to develop andimplement a policy tool to explore the cost-effectiveness and environmental performance

of various strategies to improve water quality inspecific watersheds. In doing this, we hope toilluminate the economic issues, to facilitate thedevelopment of successful pilot nutrient tradingprograms, and to better understand the oppor-tunities and barriers to improving water qualityunder each of the policy approaches considered.

HOW DOES NUTRIENT TRADINGWORK? AN ILLUSTRATION

Assume there is a river somewhere in the UnitedStates and the state environmental agency deter-mines that the river is “impaired.” There is toomuch phosphorus getting into the water, causingalgal blooms that die and leave too little oxygenin the water for the fish to survive. The statedecides to impose a limit of 1 part per million(ppm) of phosphorus in the waste stream ofmunicipal sewage and industrial waste treatmentplants. The state agrees to let these dischargershave the option of trading to meet the regulatoryrequirement. There are three towns along theriver and a number of factories. The area alsohas many farms that contribute to the problem,but these aren’t regulated.

The intent of the analysis is to develop andimplement a policy tool to explore the cost-effectiveness and environmentalperformance of various strategies to improvewater quality in specific watersheds.

B O X 1 W H AT T R A D I N G I SA N D I S N ’ T

• Trading is an adjunct to regulation, not asubstitute for it.

• Trading is used to improve water quality,not degrade it.

• Trading is used between sources in thesame area of impact, not outside it.

• Trading is used to reduce net loads, notincrease them.


4

The first town’s sewage treatment plant is oldand due for an upgrade, so that town decides torebuild its plant and add the necessary technol-ogy to meet the standard. When bundled intothe upgrade, the addition of phosphorus controlis cheap, so the town’s costs to meet the regula-tion are low.

The second town rebuilt its plant five yearsago, when the state didn’t require phosphorustreatment. Upgrading a relatively new plantwould be expensive now, and the town wants tospend its money on a new school, not moresewage treatment. However, there is a factorynearby that also has to meet the new standard.The factory offers to reduce its phosphorustreatment to 0.5 ppm, low enough to providethe necessary credits to meet the town’srequirement, if the town will split the cost ofthe factory’s more expensive upgrade. Splittingthe factory’s cost is much lower than the cost ofupgrading the town’s facility, so the town agreesto the trade. The town and the factory both savemoney and when considered together, the newrequirement is met for both. Since both the fac-tory and wastewater treatment plant dischargeto the river from a specific point, i.e., a pipe,this is a “point-point” trade.

The third town also upgraded its plantrecently. A member of the town council gets agroup of farmers to agree to put assorted con-

servation practices in place if the town will paythe expenses. These conservation practices notonly reduce the amount of phosphorus gettinginto the river, but also sediment and pesticides.The state gives its approval, but the agriculturalloads are variable since they only occur when itrains, so the state makes the town buy threetimes as much as the load reduction require-ment for the plant. This “trading ratio” willprovide a margin of safety for the river. Sincethe load of the plant is 10,000 pounds of phos-phorus a year, the town must buy 30,000pounds of reductions from the farmers orother sources that don’t come from a specificpoint like a pipe. The cost of these “nonpoint”trades is much less than upgrading the plant,even with the extra 20,000 pounds. Ratherthan get all of its credits from farmers, thetown council also decides to give some moneyto a local conservation group to restore anearby wetland. The conservation group buysup several farms on low-lying areas of the riverand plants the land with water-loving plantsthat grow in the region and provide habitat forwaterfowl and other species. As the wetlandmatures, the plants trap phosphorus-carryingsediments from the water, reducing the loadand improving the quality of the river for peo-ple and wildlife. The town takes credit for thereductions provided by the wetlands andapplies one third of them against the loadreduction mandated by the state.


5

T he principal source of information aboutthe nation’s water quality is the Environ-mental Protection Agency’s National

Water Quality Inventory. This is a compilation ofinformation submitted by states, Indian tribes,territories, interstate water commissions, andthe District of Columbia.

For the 1996 report, states and other jurisdic-tions surveyed 19 percent of the nation’s rivermiles; 40 percent of lakes, ponds, and reser-voirs; 72 percent of estuaries; 6 percent ofocean shoreline waters; and 94 percent of theGreat Lakes shoreline. The survey found that:

• For rivers, 56 percent were rated good (fullysupporting all uses), 8 percent were ratedgood but threatened for one or more uses,36 percent were rated impaired for one ormore uses, and in less than 1 percent useswere not attainable.

• For lakes, ponds, and reservoirs, 51 percentwere rated good, 10 percent were goodbut threatened, 36 percent were impaired,and in less than 1 percent uses were notattainable.

• For estuaries, 58 percent were rated good,4 percent were good but threatened, 38 per-cent were impaired, and in less than 1 per-cent uses were not attainable.

• For ocean shoreline waters, 79 percent wererated good, 9 percent good but threatened,and 13 percent were impaired.

• For the Great Lakes, 2 percent were ratedgood, 1 percent good but threatened, 97 per-cent impaired, and in less than 1 percentuses were not attainable.

These results are generally similar to the 1994inventory. In that survey, surface waters onlypartially supporting or not supporting their des-ignated uses included 36 percent of surveyedrivers and streams, 37 percent of lakes, ponds,and reservoirs, and 36 percent of estuaries.

Figure 1 shows the results from the EPA’ssurface water quality assessments undertakenbetween 1988 and 1996. For rivers, lakes andestuaries, the proportion of the waters surveyedthat fully support their designated uses hasdeclined, while those waters not supporting orpartially supporting has increased. Althoughthere are notable problems with this dataset,1

the trend shown here nevertheless suggests thatthe United States is failing to restore the qualityof the nation’s water.

CAUSES OF IMPAIRMENT

The principal cause of surface water impairmentis agriculture, which contributes to problems in

W A T E R Q U A L I T Y T R E N D SA N D I S S U E S

2


6

70 percent of the impaired rivers, 49 percent ofthe impaired lakes, and 27 percent of theimpaired estuaries. (See Table 1.) Industrialand municipal discharges are also significantcauses of water quality problems; in the nation’sestuaries, they are the leading contributors toimpairment. The major pollutants from theseagricultural sources are nutrients, silt and oxy-gen-depleting substances (primarily organicwastes), metals, and bacteria. Metals have beenidentified as one of the leading sources of lakeimpairment, because of the widespread detectionof mercury in the tissues of fish (USEPA, 1998).

Excessive inputs of nutrients impair waterthrough a process called eutrophication, whichcauses a wide variety of problems (Smith,1998). These include

• Increased biomass of phytoplankton;

• Shifts in phytoplankton to bloom-formingspecies that may be toxic or inedible;

• Decreases in water transparency;

• Taste, odor, and water treatment problems;

• Oxygen depletion;

• Increased incidence of fish kills;

• Loss of desirable fish species;

• Reductions in harvestable fish and shellfish;and

• Decreases in aesthetic value of the water body.

Nonpoint sources, particularly croplands, areby far the largest source of nutrients in Ameri-

1988 1990 1992 1994 1996

0

10

20

30

40

50

60

70

80

90

100

Per

cen

t

Rivers

Lakes

Estuaries

F I G U R E 1 . P R O P O R T I O N O F S U R F A C E W A T E R S T H A T S U P P O R TD E S I G N A T E D U S E S


7

can waterways. (See Table 2.) One study foundthat nonpoint sources account for 82 percent oftotal nitrogen discharges and 84 percent of totalphosphorus discharges. Croplands aloneaccount for 39 percent of all nitrogen loads and30 percent of all phosphorus loads. In contrast,point sources account for 18 to 19 percent ofnitrogen and phosphorus discharges (Carpenteret al., 1998).

Farmers have dramatically increased nitrogenfertilizer use since the mid-1960s. (See Figure 2.)Nitrogen use has increased by about 150 per-cent, from 4.6 million tons in 1965 to about 12million tons in the mid-1990s. Phosphorusapplications in the form of P2O5 were about 25percent higher in 1995 than in 1965; use peakedin 1979 at 5.6 million tons (USDA, 1997a).

Plants do not use a large percentage of thenutrients applied. A National Research Councilanalysis of nutrient use in the United Statesfound that up to 40 percent of nitrogen andabout 60 percent of phosphorus is not taken upby crops (National Research Council, 1993). Inthe case of nitrogen, which is soluble, most ofthe residual is readily available and contaminates

T A B L E 1 . L E A D I N G S O U R C E S A N D P O L L U T A N T S O F I M P A I R E DW A T E R W A Y S A F F E C T I N G T W O O R M O R E C A T E G O R I E SO F S U R F A C E W A T E R S

Percentage

Rivers Lakes Estuaries

Sources of ImpairmentAgriculture 70 49 27

Industrial point sources/discharges 14 – 56

Municipal point sources 9 18 44

Urban runoff/Storm sewers 13 21 20

Land disposal of wastes – 11 19

Hydromodification 14 14 –

Pollutants/StressorsNutrients 40 51 57

Siltation 51 25 –

Oxygen-depleting substances 29 21 33

Bacteria 32 – 42

Metals 16 51 –

Habitat alterations 19 – 14

Source: USEPA, 1998. National Water Quality Inventory.

Nonpoint sources, particularly croplands,are by far the largest source of nutrients inAmerican waterways.


8

T A B L E 2 . N I T R O G E N A N D P H O S P H O R U S D I S C H A R G E S T O S U R F A C EW A T E R S F R O M P O I N T A N D N O N P O I N T S O U R C E S I N T H EU N I T E D S T A T E S ( I N T H O U S A N D S O F M E T R I C T O N S P E R Y E A R )

Source Nitrogen Phosphorus

Nonpoint sources

Croplands 3,204 615

Pastures 292 95

Rangelands 778 242

Forests 1,035 495

Other rural lands 659 170

Other nonpoint sources 695 68

Total nonpoint discharges 6,663 1,658Total point sources 1,495 330Total discharge (point + nonpoint) 8,158 2,015

Nonpoint as a percentage of total 82 84

Source: Carpenter et al., 1998.

1965 1969 1973 1977 1981 1985 1989 1993

0

5

10

15

Mil

lion

Ton

s P

er Y

ear

Nitrogen (N)

Phosphate (P2O5)

F I G U R E 2 . F E R T I L I Z E R U S E 1 9 6 5 – 1 9 9 5


9

the environment through runoff, leaching, andvolatilization to the atmosphere. Phosphorusbinds to the soil and reaches surface watersthrough soil erosion processes. So much phos-phorus has been used in the past that in manyareas extension agents recommend that farmersstop applying it.

Animal manure is another key source ofnutrient use on croplands and pastures. In thelast 30 years, and particularly in the last 10, live-stock populations have shifted away from cattleand toward poultry, resulting in greater nutrientconcentration in wastes. Poultry litter has amuch higher concentration of nutrients perunit than cattle manure, roughly 150 percentmore nitrogen and 200 percent more phospho-rus. The total population of dairy cows andother cattle is down by 12 percent since 1972,while the population of poultry layers and broil-ers is up by 134 percent. The number of turkeyshas increased by 180 percent.2 The volume ofanimal wastes has not changed between 1972and 1996, but the nutrient levels in waste havegone up significantly. Using livestock numbersand standard waste and nutrient coefficients,we estimate that the total amount of nitrogen indomestic livestock wastes is up by 17 percentand phosphorus by 18 percent. 3

The total availability of nitrogen and phospho-rus from fertilizer and animal wastes in 1995was 21.4 million tons and 7.4 million tons,respectively. This represents no change in thecase of phosphorus, but a 30-percent increasein the availability of nitrogen for cropland appli-cations from these sources since 1972.

A contributing factor in water pollution fromanimal wastes is that manure is typicallyapplied well beyond crop needs. About 90 per-cent of the manure from animal operationsdoes not leave the farm where it is produced(Bosch and Napit, 1992); because of high trans-

portation costs, it is usually applied only to theland on the farm. It typically takes an area ofcropland about 1,000 times greater than thefeedlot to distribute manure at rates comparableto crop requirements (NRC 1993). If manure isapplied excessively, phosphorus builds up in thesoil, and without proper erosion control willrun off into surface waters. Excess nitrogen notonly runs off into waterways, but can also leachand contaminate groundwater or escape to theatmosphere as nitrous oxide (N2O), a gas thatcontributes to global warming.

Another trend contributing to excessive nutrientflow into the environment is lagging sewage treat-ment. Because of population growth, there arenow more people who are not served by someform of sewage treatment than in 1980. The per-centage of people served from 1980 to 1996increased from 70 to 72 percent (USEPA, 1996and 1980 Needs Surveys). However, populationincreased over that same time by 38 million peo-ple, or 17 percent (U.S. Department of Commerce,1998). According to our calculations, about 75 mil-lion people—up by 5 million since 1980—do notcurrently have access to sewage treatment.

WATER QUALITY POLICY ISSUES

The Clean Water Act (CWA) set a national goalof achieving zero discharge by 1985. But mostof the evidence suggests that we are movingfurther from this goal, rather than closertowards it.

The Clean Water Act set a national goal ofachieving zero discharge by 1985. But mostof the evidence suggests that we are movingfurther from this goal, rather than closertowards it.


10

Why are we failing to achieve the goals of theAct? One reason is that the problem has grown.With more people, a bigger economy, and moredemand on natural resources as receptors ofpollution and providers of water, recreation andother services, the challenge has grown larger.

A second major reason concerns policy for-mulation. In the water quality area, policy isfractured, expensive, and inconsistent. Both theEnvironmental Protection Agency and theDepartment of Agriculture (USDA) addresswater quality problems at the federal level. As abroad generalization, given the relative levels ofeffort, the focus of these two arms of the gov-ernment can be characterized as regulation toachieve point source control and subsidies forvoluntary nonpoint source abatement, thoughboth use regulation, grants, and subsidies tosome extent.

When the Clean Water Act was first passed,several dramatic events—such as the burning ofthe Cuyahoga River in Ohio—prompted anational focus on controlling industrial andmunicipal sources. Under the National Pollu-tion Discharge Elimination System (NPDES),state environmental agencies with USEPA over-sight grant permits to point source dischargerssuch as municipal and industrial wastewatertreatment plants. These permits control the dis-charge of pollutants into the nation’s waters.The permits specify control levels that must beachieved. The agencies also monitor sources forcompliance and can take a variety of actions toforce compliance when permitted discharge lev-els are violated.

A great deal of money has been spent in theUnited States on municipal and industrialwaste treatment and agricultural conservationpractices. For example, in the original 1972 Actthe federal government agreed to pay up to 75percent (later reduced to 55 percent) of the con-

struction and design cost for municipal treat-ment plants. Between 1974 and 1994, about$96 billion was spent through the federal con-struction grants program for new municipalconstruction and upgrades for point source con-trol. Local governments have added another$117 billion (AMSA/WEF, 1999).

Over the next 20 years, EPA estimates thatalmost $140 billion in capital costs will beneeded for municipal treatment works andrelated needs. The Association of MetropolitanSewerage Agencies (AMSA) and the WaterEnvironment Federation say that another $190billion will be needed by local governments toreplace aging facilities and collection systems,not including operation and maintenance costs(AMSA/WEF, 1999).

The approach taken for nonpoint sourcesstands in stark contrast to that for pointsources. Abatement programs for nonpointsource pollution occur mainly through subsidyprograms provided by the USDA and EPA. Thelion’s share of the funds come through agricul-tural legislation to farmers for land retirementand cost-share programs, primarily for erosioncontrol. In recent years, the USDA has spentaround $3.5 billion per year on conservationprograms, extension, administration, andresearch. EPA spends about another $800 mil-lion on its voluntary nonpoint programs

Over the next 20 years, EPA estimates thatalmost $140 billion in capital costs will beneeded for municipal treatment works andrelated needs . . . another $190 billion willbe needed by local governments to replaceaging facilities and collection systems.


11

(USDA, 1997a). Approximately half of theUSDA’s money goes to the ConservationReserve Program (CRP), which was conceivedprimarily as a means to keep land out of pro-duction to support crop prices. Environmentalbenefits of the CRP were initially focused onreducing soil erosion; in this respect, the pro-gram has been effective. Since 1982, soil ero-sion in the United States has been cut by 40percent (USDA, 1994a). Adjustments in theprogram have attempted to make it moreresponsive to water quality needs, but about 60percent of the acreage in the program is con-centrated in the Northern Plains, SouthernPlains, and Mountain States, where water qual-ity benefits from the program are relativelysmall (USDA, 1997a).

There is some measure of involuntary regula-tion of nonpoint sources provided in the CWA,particularly for farms larger than 1,000 animalunits.4 However, even for such large operatingunits, regulation is not automatic in moststates. The state first has to make a determina-tion that the source is discharging to a waterbody and contributing to a water quality prob-lem; only then will the farm be required toapply for a permit. Of 1.1 million farms in theUnited States, just 10,000, or 1 percent, havepermits under the CWA (Adler, et al., 1993).

As a result, wastes from animal operationsare not controlled with anywhere near the samerigor as human wastes, even though, in the

United States, waste from livestock is about 130times greater than that from humans. Takentogether, the 1,600 dairies of California’s Cen-tral Valley produce waste equivalent to 21 mil-lion people. The 600 million chickens of theDelmarva Peninsula on the eastern side of theChesapeake Bay generate as much nitrogen as acity of 500,000 (U.S. Senate, 1997).

Yet, there is a wide variance across the statesregarding nonpoint source regulations and

Wastes from animal operations are notcontrolled with anywhere near the samerigor as human wastes, even though, in theUnited States, waste from livestock is about130 times greater than that from humans.

B O X 2 T H E C L E A N W AT E R A C T

Federal water pollution control lawattempts to protect water quality by con-trolling the discharge of pollutants. Con-gress passed the Clean Water Act (CWA)in 1972 with the basic objective “to restoreand maintain the chemical, physical andbiological integrity of the Nation’s waters.”The CWA sought to have “fishable andswimmable waters” by 1983, “zero dis-charge” of pollutants by 1985, and to elim-inate the release of “toxics in toxicamounts.” The law made the EPA respon-sible for setting national standards for thedischarge of effluents on an industry-by-industry basis, considering both the capa-bilities of pollution control technologiesand the costs of implementation. This leg-islation makes it illegal to discharge pollu-tants to surface waters without a permit.The legislation specifically focused onpoint sources. The law was extensivelyamended in 1977 and 1987 to expandEPA’s powers and to address nonpointpollution through voluntary programs.

Adapted from: Adler et al., 1993, and Arbuckle

et al., 1993.


12

their implementation (Environmental LawInstitute, 1997). As of 1996, only 16 statesrequired nutrient management plans, and thesewere usually in areas suffering from ground-water contamination. Eighteen states requiredpractices for controlling soil erosion, but onlywhen a complaint is filed by a citizen or govern-ment agency. Implementation in response tothe complaint is only required if cost-sharefunds are available from the government(USDA, 1997a).

It is clear that the nation’s existing water qual-ity policy framework is inadequate. Not only isthe nation failing to meet the goals set out byCongress, but forward progress seems to beslow and difficult. In many cases, environmen-tal groups have resorted to litigation in anattempt to make state and federal agenciesenforce the basic requirements of the CWA.

States are charged with identifying waterswithin their jurisdiction that do not meet desig-nated uses and developing Total MaximumDaily Loads (TMDLs) to address the problems.TMDLs are plans to establish and allocate load-ing targets. This typically means additionalrequirements for point sources and more subsi-dies for nonpoint sources. Where states do notaccomplish these tasks, EPA must do it forthem. There are currently 12 states where EPAis under court order to establish TMDLs if thestates do not, another 15 states with pending lit-igation, and 4 states where notices of intent tosue have been filed (USEPA, 1999).

The list of impaired waterways is extensive.Section 303(d) of the CWA requires states toidentify waters that are not fishable or swim-mable and to develop TMDLs for these waters.Nationwide, there are 3,456 waterways listed asimpaired by nutrients and another 141 impaired

by algal blooms, typically caused by excessnutrients. (See Table 3.)

T A B L E 3 . T H E T O P 2 0 S T A T E S B YN U M B E R O F W A T E R W A Y SI M P A I R E D B Y N U T R I E N T S

Illinois 634

Florida 539

Mississippi 469

Oklahoma 218

Pennsylvania 217

Ohio 204

Montana 156

Maryland 145

Delaware 138

Massachusetts 135

North Dakota 98

Tennessee 60

California 58

Kentucky 52

Wisconsin 46

Connecticut 36

Rhode Island 35

Michigan 35

Iowa 25

Nebraska 24

Subtotal 3,324Total 50 States 3,456

Source: USEPA, 1999.

Nationwide, there are 3,456 waterwayslisted as impaired by nutrients andanother 141 impaired by algal blooms,typically caused by excess nutrients.


13

M any studies on the economics of envi-ronmental regulation emphasize theimportance of allowing flexibility in

addressing environmental problems.

Trading, as an adjunct to regulation, can helpprovide the flexibility to keep compliance costsdown while meeting or exceeding water qualitygoals. For example, discharge permits, a com-mon regulatory approach for reducing pointsource pollution, specify the amount and typeof allowable discharge to the nation’s waters.In the case of nutrients, permits are used tolimit discharges that were previously uncon-trolled or tighten prior limits, causing permitholders to incur significant capital and operat-ing costs to comply. For several reasons,including size, control technology, and charac-teristics of the incoming waste stream, dis-chargers can face vastly different costs of com-pliance, but have little flexibility to seekcost-effective solutions without programs suchas trading.

There are two fundamental types of tradingprograms for any type of discharge: open andclosed. Closed trading programs are far morecommon and are an extension of traditional reg-ulation in that they begin with a mandate toreduce discharges. Often called “cap and trade,”these systems include a mandatory “cap” onemissions or discharges and individualallowances to sources within a defined tradingarea. A cap is established by the regulatoryagency to achieve or maintain ambient air,water, or other environmental quality standards.A TMDL—Total Maximum Daily Load—is theprocess under the Clean Water Act that estab-lishes and allocates the cap within a watershedor impaired area. Caps may be fixed or decliningdepending on the attainment or nonattainmentstatus. A cap can be increased without under-mining the integrity of the water quality objec-tives if sources not controlled by the cap opt inand reduce their loads by at least the amountthat the cap is extended. For example, if a capwere on point sources only, nonpoint sourcescould make load reductions that could be usedto increase the point source cap. Allocations inthe form of allowances are established for eachdischarge site within the trading area based on apercentage of overall reductions that must beachieved from a baseline of actual or allowedhistorical discharge levels. A trading system issaid to be “fully closed” when all discharge sitesare controlled under the cap and the cap is equal

W H A T A B O U T T R A D I N G?

3

Trading, as an adjunct to regulation, canhelp provide the flexibility to keepcompliance costs down while meeting orexceeding water quality goals.


14

to the total permissible load for the watershed(Stephenson and Shabman, 1996).

Closed systems have been developed andimplemented in areas where ambient environ-mental standards are not being met to provide amore cost-effective means of achieving thereductions necessary to attain these standards.

The Acid Rain Trading Program under thefederal Clean Air Act Amendments of 1990 isone example of a closed trading system. Theprogram, which set a cap of 8.95 million tons ofSO2 per year, is enforced through a system oftradable emission allowances. From 1995 to1997 the program exceeded expectations, withfirms over-achieving the reduction target at lessthan half the cost. Industry and EPA initiallyestimated that abatement costs would be in therange of $750-$1,000 per ton, but allowanceprices reached a low of $66 per ton in mid-1995 and have generally remained below $200per ton since the system’s inception. Over thelong run, it is estimated that allowance tradingcould result in savings of $700 to $800 millionper year compared to a command-and-controlapproach with a uniform emission standard(Anderson, 1999).

The states of California and Illinois have alsoimplemented closed air emission trading pro-grams designed to attain the national ambientair quality standard for ozone.

Open systems are usually voluntary. Whereambient standards are being met, they are usedto maintain or improve environmental qualitywhile allowing for economic growth and devel-opment. A number of states are looking intothe development of open air and water qualitytrading programs. Michigan’s air emission trad-ing program took effect on March 16, 1996,and the states of Florida and Texas are alsoworking to implement similar programs.

Open trading systems rely on existing regula-tions to establish a baseline; reductions fromthe baseline generate a reduction credit. Creditscan be banked, traded, or used to comply withdischarge limits established by an applicableregulation. Open systems offer greater opera-tional flexibility and improvements in environ-mental quality without establishing a manda-tory cap and allowance for each discharge site.Reductions must be real, surplus, and enforce-able to be used or traded under an open system.

In a few instances, regulatory agencies haveallowed trading between polluters. In thesecases, those sources with high remediationcosts have been allowed to purchase unusedemissions allowances provided under anotherfacility’s permit. The most important determi-nants of success have been cost and sufficiencyof trading options, limited regulatory restrictionon trading, low transaction costs, certainty inestablishing pollution levels, and anticipatedimprovement in environmental quality (Hahnand Hester, 1988).

Regulators have allowed trading in a smallnumber of cases for water quality management(Bacon, 1992; Hahn and Hester, 1988; USEPA,1996). While effluent trading programs havebeen very successful for lead, air emissions, andsulfur dioxide trading, the results in the waterquality area are not impressive.5 In the firstinstance, the state of Wisconsin gave sourcessuch as wastewater treatment plants, and pulpand paper mills the right to meet state waterquality standards through the trading of effluentrights. The state initiated biological oxygendemand (BOD) trading on a 35-mile stretch ofthe heavily industrialized Fox River and on 500miles of the Wisconsin River. It allowed tradesbetween the paper mills and municipal waste-water treatment plants. However, regulatorswould only allow trades if permitted dischargelevels could not be met through the adoption of


15

technological standards. Cost reduction was notconsidered a legitimate reason for trading. Since1981, only one trade has occurred. Several fac-tors are thought to have contributed to the lim-ited activity, including the severe restrictionsimposed by the state on the ability of sources totrade, the vulnerability of the program to legalchallenge, and the fact that the dischargersdeveloped a variety of compliance alternativesthat had not been foreseen when the regulationswere drafted (Anderson, 1999).

The Environmental Protection Agency is cur-rently considering extending point-point tradingfor copper dischargers in San Francisco Bay andfor dischargers of nitrogen and suspendedsolids in Tampa Bay.

POINT-NONPOINT TRADING

In the point-nonpoint trading area, programsare currently in place at the Dillon and CherryCreek Reservoirs in the Denver, Colorado, area;in North Carolina’s Tar-Pamlico Basin; and arebeing considered in many other places.6

Dillon Reservoir provides about half of thecity of Denver’s water supply. Four municipalwastewater plants discharge into the reservoir.Studies in the 1980s indicated that phosphorusdischarges would have to be reduced to main-tain the reservoir’s water quality and accommo-date future growth, and that reductions wouldhave to include nonpoint sources such asrunoff from lawns and streets, and seepagefrom septic tanks. Using 1982 phosphorus dis-charges as the baseline, the plan established acap on total phosphorus loadings, set load lim-its for the four treatment plants, and allowedtrading of phosphorus loadings with nonpointsources (Anderson, 1999).

In this case, nonpoint source control wasallowed as a means to supplement point source

management, provide for regional growth, andimprove water quality in the reservoir. Pointsources can increase discharges above approvedlimits only if they reduce loads from nonpointsources in the watershed. In order to deal withthe uncertainty of nonpoint source reductions,only half of the expected reduction from non-point sources can be applied against pointsource loads. Since 1984, point sources in thearea have increased their efficiency and reducedtheir phosphorus loads significantly. Pointsources have generally been able to keep theirloads below regulatory limits, so only threetrades have taken place. Future trades areexpected as development in the area proceeds(Bacon, 1992; Michigan DEQ, 1998a). A similarprogram is in place at Cherry Creek Reservoir—also a source of water for the Denver region—but no trades have taken place to date becausephosphorus loadings at municipal treatmentfacilities are still below the limits set by the Col-orado Water Basin Commission.

In 1989, the North Carolina EnvironmentalManagement Commission designated the Tar-Pamlico Basin as “nutrient-sensitive” after stud-ies found that algae blooms and low dissolvedoxygen posed a threat to the basin’s fisheryresources. The formal designation required thestate Division of Environmental Management(DEM) to identify the nutrient sources, setnutrient limitation objections, and develop anutrient control plan. The DEM analysisshowed that most of the basin’s nutrient load-ings (primarily nitrogen, but also phosphorus)came from agricultural runoff and other non-point sources. Other sources include municipalwastewater treatment plants and industrial andmining operations (Anderson, 1999; MichiganDEQ, 1998a).

In the Tar-Pamlico, an association of pointsource dischargers is allowed to trade with oneanother under a cap. If the cap cannot be met,


16

members of the association can pay into a fundto support a government-managed programthat encourages the adoption of best manage-ment practices by farmers in the watershed.Tar-Pamlico is a hybrid of a trading programand an effluent tax, since the credits are pur-chased at a fixed price and there is no directconnection between the credits needed by thepoint sources and the credits generated by thenonpoint source program. In the first phase ofthe program, association members improvedthe efficiency of their facilities and successfullymet defined standards. Allowable dischargesare being gradually reduced during the secondphase, which began in 1995 (USEPA, 1992;Michigan DEQ, 1998a).

A number of other trading programs are inexistence or being developed in Colorado, Con-necticut, Michigan, Minnesota, and Wisconsin.However, in the existing cases, only limitedmarkets for nutrient trading have emergedbecause the programs are not fully functionalor allowable discharge limits have not beenreached. Some programs have actuallyincreased administrative and transaction costs,making trading difficult.

THE POTENTIAL FOR POLLUTION TRADING

On a strictly physical basis, the opportunitiesfor nutrient trading appear to be enormous. Aconservative estimate suggests that point-non-

point trading could be applicable in as many as900 watersheds in the United States (USEPA,1992a). Nonpoint sources, in particular agricul-ture, are major contributors to water pollutionand remain largely unregulated, creating a vastopportunity for significant reductions in pol-luted runoff. The U.S. National Research Coun-cil estimated national nitrogen “residuals”—that is, nitrogen not taken up by plant mater-ial—on cropland at 6.6 to 9.1 million metrictons and phosphorus residuals at 2.9 millionmetric tons, much of which represents excessuse. Other researchers have shown that asmuch as three quarters of excess nutrient usecan be avoided through fairly simple and inex-pensive techniques such as soil testing and fer-tilizer banding (NRC, 1993).

From an economic standpoint, the opportuni-ties for cost savings from point-nonpoint sourcetrading are large. One estimate suggested thecost of point-source reduction could be 65times higher than nonpoint source reduction(Bacon, 1992). EPA estimates that if back-ground pollution from agriculture werereduced, the need for tertiary (advanced) watertreatment could be avoided—providing a netsaving of $15 billion in capital costs for tertiarytreatment (USEPA, 1992).

Nutrient trading between point sources makessense as well. Wastewater treatment, like manyprocesses, is characterized by significanteconomies of scale; larger facilities can processwastewater at much lower per-unit cost thansmall facilities. Particularly in cases involving alarge number of small treatment works, tradesinvolving a few large operations that overcomplyand sell the credits to the smaller ones can dra-matically lower overall compliance costs whileachieving the same environmental objective.

An analysis of President Clinton’s 1994 CleanWater Initiative (CWI) suggested that nutrient

The opportunities for nutrient tradingappear to be enormous. A conservativeestimate suggests that point-nonpointtrading could be applicable in as many as900 watersheds in the United States.


17

trading could lower the cost of implementingthe initiative by $658 million to $7.5 billion,based on total incremental costs estimated torange from $5 billion to $9.6 billion. The bulkof these savings (75 to 92 percent) would berealized from point-nonpoint trading, while the

remaining savings were expected to come frompoint-point trading (USEPA, 1994).

Two other possible options for nutrient trad-ing include pretreatment trading, where amunicipal plant pays a facility that is discharg-ing into its system to reduce its contributionrather than upgrading the system itself. Non-point-nonpoint trading, another option, is possi-ble in theory, but practical examples do not exist.

In short, nutrient trading may represent avast untapped reservoir to resolve the nation’swater quality problems, but significant imple-mentation and performance issues must betackled.

An analysis of President Clinton’s 1994Clean Water Initiative suggested thatnutrient trading could lower the cost ofimplementing the initiative by $658million to $7.5 billion.


19

N utrient trading is widely thought toreduce the costs of improving waterquality, but there is little economic

analysis comparing trading with other policyapproaches. Of the work done to date, moststudies simply compare average costs of variousforms of nutrient reduction for point and non-point sources.

In the course of this study, we developed acomprehensive analytical framework to com-pare the economic and environmental perfor-mance of alternative policy strategies to reducenutrient loads. In particular, we explored oppor-tunities to reduce phosphorus loads in threewatersheds of the Upper Midwest: the SaginawBay in Michigan, the Rock River in Wisconsin,and the Minnesota River Valley. (See Figure 3.)

The intent of the analysis was to develop andimplement a policy tool to explore the cost-effectiveness and environmental performanceof various strategies to improve water quality in

specific watersheds. In doing this, we hoped toilluminate the economic issues, to facilitate thedevelopment of successful pilot nutrient tradingprograms, and to better understand the oppor-tunities and barriers to improving water qualityunder each of the policy approaches considered.

THE CASES: NUTRIENT TRADINGIN MINNESOTA, MICHIGAN,AND WISCONSIN

These cases were chosen because each area hassignificant water quality problems and theresponsible state agency was interested inexploring nutrient trading as a policy alterna-tive. The study is a joint effort of the WorldResources Institute; the Michigan, Wisconsin,and Minnesota state environmental protectionagencies; and various stakeholders in eachwatershed, including point source operators,farmers, environmental groups, and academics.In each of these three states, there has been sig-nificant movement toward the use of trading.

BackgroundNotwithstanding requirements to upgrade pointsource discharges and the development of non-point source programs, nutrient pollution in theMinnesota River has been a chronic problem foryears. In 1988, the Minnesota Pollution ControlAgency (MPCA) established a Total MaximumDaily Load (TMDL) on the lower 25 miles of the

N U T R I E N T T R A D I N GO P P O R T U N I T I E S I N T H R E E

W A T E R S H E D S O F T H E U P P E RM I D W E S T

4

These three cases were chosen because eacharea has significant water quality problemsand the responsible state agency wasinterested in exploring nutrient trading asa policy alternative.


20

Minnesota River, limiting chemical and biologi-cal oxygen demand (BOD) and ammonia nitro-gen (USEPA, 1992b). Phosphorus control isnow under active regulatory consideration by theMPCA. A study of point source treatment foundthat incremental costs for phosphorus controlrange from about $15 million to $39 million peryear (Metcalf and Eddy et al., 1993). A prelimi-nary study done for EPA in 1987 identified thebasin as a good candidate for a trading program(Industrial Economics, 1987).

Recently this interest materialized in the formof a trade tied to a new permit. In January1997, Minnesota Governor Arne Carlsonannounced a new discharge permit for the RahrMalting Company that allows the company tobuild its own wastewater treatment facility on atightly regulated stretch of the Minnesota Riverin exchange for reducing nonpoint source dis-charges upstream from the plant (MPCA,1997a). Under the terms of the agreement,Rahr can support agricultural conservation

F I G U R E 3 .


21

practices including soil erosion controls, live-stock exclusion from waterways, rotational graz-ing, critical-area set asides, and wetland treat-ment systems for nutrient removal. Credits thatRahr receives from these practices can beapplied against the load of its new facility.Detailed papers outline how the amount ofcredit from various practices is to be calculated(MPCA, 1997b, 1997c,1997d).

Michigan has developed rules for a voluntarystatewide “water quality” trading program andis conducting a demonstration project on theKalamazoo River.7 The Kalamazoo project isintended to help the state and a group of stake-holders obtain design information for develop-ment of the statewide program (Michigan DEQ,1998a). The Water Quality Trading Group inMichigan has drafted an extensive Water Qual-ity Trading Rule to enable it to move forwardand implement the demonstration program(Michigan DEQ, 1998b). The draft has beensubmitted to the Region V office of the USEPA;discussions are continuing on the merits of thedraft rule.

Though Michigan ultimately chose to imple-ment a pilot program in the Kalamazoo Riverwatershed, Saginaw Bay was chosen as themost appropriate case for this study. SaginawBay has been the focus of water qualityimprovement efforts by the state of Michiganfor more than 15 years. The state has sponsoredseveral studies and made major efforts to

reduce nutrient loads; the latest effort is theSaginaw Bay Soil Erosion and SedimentationControl Program. Because Saginaw Bay hasbeen the subject of prior studies, the informa-tion necessary to complete a research effort onnutrient trading was generally available andcomparative analysis was possible. According toDepartment of Environmental Quality (DEQ)staff, the analysis presented here was a key fac-tor in the state’s determination to undertake astatewide program (Batchelor, 1999).

According to both the Wisconsin Departmentof Natural Resources and the U.S. Environmen-tal Protection Agency, the Rock River is a waterbody that does not attain its designated uses. Assuch, the river is subject to a limit on the allow-able load of pollutants it can receive, includingphosphorus.

In response to a proposed rule to require allthe point sources in the Rock River watershed toadopt phosphorus controls, a group of pointsource dischargers called the Rock River Partner-ship petitioned the secretary of the WisconsinDepartment of Natural Resources to grant theman extension so that a nutrient trading programcould be tried in the watershed. The partnershiphas assessed its membership according to theirsize in order to raise funds to pay for hydrologi-cal modeling and monitoring of the basin. In1996, the governor’s budget included funds tostudy nutrient trading in four watersheds, theRock River among them. These pilot programsare intended to help develop a statewide frame-work for trading (Michigan DEQ, 1998a).

Table 4 summarizes EPA’s vulnerability indi-cators for the 23 sub-basins found in the threewatersheds in the present study. The data showthe nature of the water quality problems. Justone of the sub-basins have “best” or “better”condition, while three quarters have “moreserious” conditions (a five or six rating, with six

Michigan has developed rules for avoluntary statewide “water quality”trading program and is conducting ademonstration project on the KalamazooRiver.


22

being the worst possible rating). Just one sub-basin is highly vulnerable to urban runoff,while 19 are highly vulnerable to agricultural

runoff. A few sub-basins are vulnerable to “pop-ulation change” (meaning growth) and the indi-cators show that many are moderately vulnera-

T A B L E 4 . S U B - B A S I N V U L N E R A B I L I T Y I N D I C E S B Y W A T E R S H E D

Index ofAquatic Urban Agricultural

Overall Species Runoff Runoff Population HydrologicSub-Basin Name Condition at Risk Potential Potential Change Modification

Saginaw Bay WatershedAu Gres-Rifle 1 Low Low Moderate High Moderate

Kawkawlin-Pine 5 n/a Low Moderate Low Low

Pigeon-Wiscoggin 5 Moderate Low High Low Low

Tittabawassee 3 Low Low Moderate Moderate Moderate

Pine 5 Low Low High Low Moderate

Shiawassee 5 Moderate Low High Moderate Moderate

Flint 4 n/a Low High Low Moderate

Cass 5 Low Moderate High Low Low

Saginaw 5 Low Low Moderate Low n/a

Minnesota River WatershedUpper Minnesota 3 Low Low High Low Moderate

Pomme De Terre 5 Low Low High Low Moderate

Lac Qui Parle 5 n/a Low High Low Moderate

Hawk-Yellow Medicine 3 Moderate Low High Low Moderate

Chippewa 5 n/a Low High Low Moderate

Redwood 5 n/a Low High Low Low

Middle Minnesota 5 Moderate Low High Moderate Moderate

Cottonwood 5 Moderate Low High Low Low

Blue Earth 5 Moderate Low High Low Moderate

Watonwan 5 n/a Low High Low Moderate

Le Sueur 5 n/a Low High Low Moderate

Lower Minnesota 6 High Low High High Moderate

Rock River WatershedUpper Rock 4 Moderate Moderate High High HighCrawfish 5 Low Moderate High Moderate Moderate

Source: Environmental Protection Agency, “Surf Your Watershed” website.

Notes: The overall condition describes the health of the aquatic resources for each watershed. The score is the result of combining 15 indi-

cators of watershed condition and vulnerability, zero being the best water quality or condition, and 6 being the most serious condition.


23

ble to hydrologic modification such as dams.Not shown here, Kawkawlin-Pine is highly vul-nerable to conventional loads over permittedlevels. Most other basins have low vulnerabilityto toxic or conventional loads over permittedlevels. For various reasons, only one sub-basinattains its designated uses.

Point Sources and Control OptionsThe three watersheds have a number of thingsin common and a few important differences.Three key differences are the number of pointsources in each watershed, the variation in theirsize distribution, and their level of treatment.Figure 4 shows the number and size distribu-tion for each watershed. In Minnesota, 211 facil-ities release phosphorus into the Minnesota

River. They range in size from a few thousandgallons per day to 26 million gallons per day(mgd), with a total flow of 130 mgd. The twolargest municipal plants in the region haveflows of 26 mgd and 21 mgd. The vast majorityof the plants on the river are considerablysmaller; most of the flow is concentrated in afew facilities. Of the total, 199 have a flow ofless than 1 mgd and 126 have a flow less than0.1 mgd. Only the two largest plants currentlyhave phosphorus control.

The average concentration of phosphorus ineffluent from all point sources in the watershedis 1.76 parts per million8 (ppm), according toestimates made using Minnesota Pollution Con-trol Agency (MPCA) data. The total phosphorus

Minnesota River Saginaw Bay Rock River

0

20

40

60

80

100

Per

cen

tage

s

< 0.10.1 to 1.01.0 to 5.0> 5.0

211 plants 69 plants 60 plants

(Flow in millions of gallons per day)

F I G U R E 4 . D I S T R I B U T I O N O F P O I N T S O U R C E D I S C H A R G E R S I N C A S ES T U D Y A R E A S B Y F L O W R A T E


24

load to the watershed from these sources isabout 348 tons per year.

Sixty point source dischargers release phos-phorus into the Rock River in Wisconsin. Theyrange in size from 30,000 gallons per day to 40million gallons per day at Madison, which is byfar the largest plant. The majority of the plantson the river are small, but there is a muchgreater proportion of large plants than in Min-nesota. Of the total, 16 have a flow of more than1 mgd and 35 have a flow less than 0.25 mgd.Only Madison’s municipal wastewater treatmentplant has phosphorus control. Under the pro-posed state regulation (NR217), plants with loadsless than about 1,300 pounds per year would nothave to reduce P loadings. This means that 25 ofthe 60 dischargers would have no obligation tocomply under the proposed rule.

The average concentration of phosphorus ineffluent from point sources in the watershedsubject to the regulation is about 2.7 ppm,according to estimates derived from WisconsinDepartment of Natural Resources (WDNR)data. The total phosphorus load to the water-shed from these sources is about 401 tons peryear. The point sources that are not subject tothe regulation contribute an additional 14 tonsper year of phosphorus, or 3 percent of the totalpoint source load.

The Minnesota and Rock Rivers drain intothe Mississippi, but Saginaw Bay is part ofthe Great Lakes system. For this reason, onlythe facilities in that watershed have had toadopt treatment for phosphorus removal.There are 69 point source dischargers withpermits to release phosphorus into the Sagi-naw Bay Watershed. They range in size froman average of a few thousand gallons per dayto 89 million gallons per day. The largestmunicipal wastewater treatment plants in theregion are Saginaw (25 mgd) and Flint (28

mgd). There are few plants smaller than 0.1mgd in the basin and a large proportion ofrelatively large facilities.

By law, the standard concentration limit forphosphorus in effluent that flows into the GreatLakes is 1 ppm, though some plants havetighter restrictions or have voluntarily achievedlower limits. The average concentration ofphosphorus in effluent from all point sourcesin the watershed is 0.74 ppm, according to datareported to the Michigan Department of Envi-ronmental Quality (MDEQ) by dischargers. Thetotal phosphorus load to the watershed fromthese sources is about 321 tons per year.

Point Source Controls and CostsFor sources that have been subject to phosphoruslimits for some time, as in Michigan, the stan-dard method of control has been chemical phos-phorus removal. In this method, metal salts ofaluminum and iron are added to waste streamsto form compounds that precipitate phosphorusfrom the wastewater. For effluent limitations of 1ppm, metal salt additions in conjunction withconventional clarifiers do an acceptable job. It ispossible to reduce phosphorus levels to 0.5 ppmin the effluent by adding more metal salt, but toachieve limits of 0.2 ppm it is necessary to usesecondary filtration with either chemical or bio-logical treatment. (See Table 5.)

Biological phosphorus removal is a relativelynew technology. There are at least six differenttypes of biological phosphorus removal sys-tems; all rely on biota to remove excess phos-phorus from the waste stream during theirgrowth processes. For most point sources withexisting chemical treatment, biological phos-phorus removal cannot be technically accom-modated, so these plants are locked into highercontrol cost options. Because costs are lower,most new plants use biological treatmentprocesses.


25

For this analysis, we characterized six cate-gories of treatment:

1. no treatment;

2. standard chemical phosphorus removal;

3. maximum chemical phosphorus removalpossible without filtration;

4. chemical phosphorus removal with filtration;

5. biological phosphorus removal; and

6. biological phosphorus removal withfiltration.

For each of these treatment options, we usedcost curves developed especially for this studyto estimate costs for operation, maintenance,and sludge removal (OM&R), as well as capital

costs to upgrade plants (Doran, 1997). Biologi-cal phosphorus removal is the least expensiveoption for point sources that currently have notreatment. For a 1 mgd plant, the cost of build-ing a chemical phosphorus removal system isapproximately $1.25 million. For the same sizeplant, a biological process would cost about $1million. For both of these types of plants, costsper unit of treatment as measured by water flowdecline significantly as the size of the plantincreases. The costs of adding filtration tochemical or biological phosphorus removalprocesses is the same, about $2.5 million tocapital costs for a 1 mgd plant.

Biological phosphorus removal processes havean advantage in total operation, maintenance andsludge removal costs, primarily because there isless sludge produced with biological processes.For a chemical process, the annual costs in thiscategory are about $60,000 per mgd plant size

T A B L E 5 . N U M B E R O F P L A N T S I N E A C H W A T E R S H E D B YP H O S P H O R U S ( P ) T R E A T M E N T T Y P E

Minnesota River Valley Saginaw Bay Rock River

No treatment 209 21a 34b

Standard chemical 0 21 0

Maximum chemical 0 17 0

Chemical with filtration 0 10 0

Biological 2 0 1

Biological with filtration 0 0 0

Total for Watershed 211 69 60

Average concentration of P in

effluent (ppm) 1.76 0.74 2.7

Total Point Source P Load 348 321 401

Notes:

a. These sources meet the discharge standard without treatment.

b. Of this number, 25 plants are too small to meet the minimum compliance requirements.


26

with no economy of scale. For biological removal,the comparable cost is $35,000 per year, withdeclining costs per unit of water treated as plantsget larger. The additional OM&R costs for filtra-tion for either process are about $80,000 peryear, with a small decrease in unit costs as plantsget bigger.9

Nonpoint Sources and Load Reduction OptionsThe watersheds of Saginaw Bay, the MinnesotaRiver Valley, and the Rock River are quite simi-lar in the dominance of agriculture in land useand phosphorus loads. As Table 6 shows, agri-culture accounts for 47 to 76 percent of thetotal land use in each case. Pasture, urban, andother uses make up relatively small shares of

land. Saginaw Bay has a heavily forested north-ern end, but in the other two cases forest landis not a large land use category. The MinnesotaRiver Valley is the most rural of the three cases,with just 2 percent of the land in urban uses.

Corn for grain or silage is the dominant cropin each watershed, with almost 4 million acresin the Minnesota River watershed, 1 million inSaginaw Bay, and just over 600,000 acres inthe Rock River. The Minnesota River Valley isthe northern end of the Corn Belt, and thecorn-soybean rotation is common there. Threemillion acres of soybeans were grown there in1992. Soybeans are important in Saginaw Bay,with about 450,000 acres grown, but are muchless important in the Rock River watershed,where pasture and hay are the second mostimportant agricultural land uses. In all thesecases, wheat, oats, barley, and other cropsaccount for much smaller shares of the landused for agricultural production.

Phosphorus loads originate not only frompoint source discharges, but also from a varietyof nonpoint sources such as agriculture, other

The watersheds of Saginaw Bay, theMinnesota River Valley, and the Rock Riverare quite similar in the dominance ofagriculture in land use and phosphorus loads.

T A B L E 6 . L A N D U S E F O R E A C H W A T E R S H E D ( P E R C E N T A G E S I N P A R E N T H E S E S )

Land Use (1,000 acres)

Cropland Pasture Forestry Urban Total

Minnesota River 7,636 692 279 231 10,842

(70) (6) (3) (2)

Saginaw Bay 2,246 409 1,522 476 5,592

(40) (7) (27) (9)

Rock River 1,413 203 159 235 2,420

(58) (8) (7) (10)

Source: 1992 National Resources Inventory (USDA, 1994).

Note: Watershed totals include other land use categories not shown such as transportation, water, federal, minor and

miscellaneous uses.


27

land uses such as forestry, and urban runofffrom golf courses, pets, and septic systems. Aprincipal mechanism by which phosphorusreaches surface waters is through erosion, asphosphorus attaches to soil particles and is car-ried away in surface runoff from the land.

Agricultural land use is the major source ofphosphorus for each of the watersheds includedin this study. These results agree with EPAwater quality assessments (USEPA, 1998). (SeeTable 4.) Analysis of erosion and expected phos-phorus runoff from nonagricultural land (usingstandard technical coefficients) suggests thatnonagricultural nonpoint sources are relativelyminor contributors to the P load in these water-sheds compared to other sources (Horner et al.,1994).

Because of agriculture’s dominance in waterquality problems, we used a more detailedmethodology for estimating loadings from agri-cultural sources. This methodology was devel-oped and peer reviewed as part of an earlierstudy done by the World Resources Institute toexplore national policy options to enhance agri-culture’s economic and environmental perfor-mance (Faeth, 1995). The methods and datawere adapted and extended to make them suit-able for a watershed-level study.

To capture the impacts from agriculture, welooked at land use within the sub-basins of

each watershed and developed alternatives forcrop production systems commonly usedwithin the region. Just as for point sources, thedata collection was intended to describe thefinancial and environmental characteristics ofcropping practices used in the watershed. Thisincludes the principal crops grown, the rota-tions in which they are grown, and the tillagepractices used.

The information used for this analysis camefrom a variety of USDA databases, includingthe Cropping Practices Survey, which is a state-level survey of farmers and the practices theyuse. From this information we identified allimportant production inputs and their levels ofuse for each crop rotation and tillage practice.Unfortunately, this source of information onlycovers the major field crops: corn, wheat, soy-beans, hay, silage, barley, and oats. There is nosimilar source of production data for vegetables,sugar beets or other minor crops, so these arenot included in the analysis. This is not a seri-ous problem, however, because the crops cov-ered account for 80 to 92 percent of cultivatedcropland and erosion rates are similar for thosecrops included and excluded. To account for allloadings from agricultural land, we inflated theacreage of the crops included to equal the totalfor all cropland. From the 1992 NationalResources Inventory, we have been able to iden-tify the acreage of various cropping rotationsand the acreage of each crop grown in eachsub-basin (USDA, 1994b).

Tillage is critically important in determiningthe amount of soil that erodes from croppingactivities, so we also represented five differenttillage types:

• mold board plowing, which completelyoverturns the soil and leaves little or noresidue to prevent erosion by wind orwater;

A principal mechanism by whichphosphorus reaches surface waters isthrough erosion, as phosphorus attachesto soil particles and is carried away insurface runoff from the land.


28

• conventional tillage, which uses a variety ofimplements such as disks and chisels butexcludes the mold board and leaves someresidue, but not much;

• mulch tillage, which leaves over 30 percentof the surface covered by residue to preventrunoff;

• no-til, which avoids tillage altogether, leavingall of the crop residue on the surface, cut-ting erosion by up to 90 percent; and

• ridge tillage, which uses built-up ridgeswithin the field to intercept runoff andencourage percolation.

Tillage data were obtained from the Conserva-tion Tillage Information Center county-level sum-maries and combined with the NRI crop rotationdata to characterize the production alternativesavailable in each watershed. In total, we represent10 rotations and four tillage types in 18 combina-tions.10 Each of these combinations is also repre-sented by the average nutrient use levels obtainedfrom survey data and a nutrient managementalternative that reduces fertilizer use to levels rec-ommended by county extension agents.

The amount of conservation tillage, mulch,no-til, and ridge tillage varies for each water-

shed. About 15 percent of cropland is in someform of conservation tillage in the Rock Riverwatershed, compared to 25 percent for the Min-nesota River Valley and Saginaw Bay. The RockRiver and Saginaw Bay have appreciably higherrates of mold board plow usage.

Environmental impact data such as soil ero-sion rates and nutrient loadings were estimatedfor each cropping alternative using a biophysi-cal simulation model, EPIC, developed by theUSDA (Williams et al., 1989). The results fromthe simulation model were calibrated to datafrom the county and sub-basin level. Cropyields and crop acreages were calibrated usingcounty averages from the state agricultural sta-tistics services and the 1992 U.S. Census ofAgriculture (USDA, 1997b).

Soil erosion rates for erosion by water (sheetand rill) and by wind were also calibrated so thatthe simulated totals for each sub-basin matchtotals estimated by the USDA’s NationalResources Inventory. Table 7 shows these ero-sion estimates, which use 1992 data. Accordingto USDA, cropland soil erosion by wind is muchlarger than sheet and rill erosion for the Min-nesota River Valley and Saginaw Bay. Values forthe Rock River were not reported by the NRI.

USDA (Ribaudo, 1992) has estimated dam-ages caused by sediment from sheet and rillerosion in 10 regions. Using a value for the

The value of damages caused by just sheetand rill erosion is $84 million per year inthe Minnesota River Valley watershed, $13million in Saginaw Bay, and $29 millionin the Rock River.

Tillage is critically important indetermining the amount of soil that erodesfrom cropping activities. About 15 percentof cropland is in some form of conservationtillage in the Rock River watershed,compared to 25 percent for the MinnesotaRiver Valley and Saginaw Bay.


29

Great Lakes region of $5.07 per ton of soileroded (1995 dollars), the value of damagescaused by just sheet and rill erosion is $84million per year in the Minnesota River Valleywatershed, $13 million in Saginaw Bay, and $29million in the Rock River.

Soil erosion is the principal means by whichphosphorus from agricultural sources reachesrivers. However, not all soil erosion is a cause ofenvironmental harm, because most eroded soilmoves a short distance and does not reach aditch, creek, or river where it can cause a prob-lem. The Institute of Water Research at theMichigan State University (MSU) developedestimates for this study of the amount of sheetand rill erosion in each sub-basin that reaches awaterway (Ouyang and Bartholic, 1997). Thereare five methods for estimating these deliveryratios. MSU generated estimates using all five.We used the mean value of the five for theanalysis. These “sediment delivery ratios” wereused to estimate loads to the river by multiply-ing the acreage in a given practice by theamount of phosphorus attached to sediment forthat practice. The range of delivery ratios for thesub-basins in each watershed are shown inTable 7.

Soil deposition by wind erosion is potentially amajor source of phosphorus loading, given thatwind erosion in the two cases where data exist ismore than double sheet and rill erosion bywater. Unfortunately, there is no method to esti-mate deposition of soil eroded by wind. For theMinnesota and Rock Rivers, we assumed a sedi-ment delivery ratio for wind of 3 percent, whichis approximately the area covered by water inthese watersheds. For the Saginaw Bay, the pre-vailing winds flow from west to east over thebay, meaning that erosion from wind is likely tohave a significant impact on the load. One studyfor a similar situation, the Chesapeake Bay,showed wind erosion deposition to the water-shed to be 9 percent. We estimated that 10 per-cent of wind-eroded soils were deposited in theSaginaw watershed. Local soil conservationistsbelieve this is a conservative estimate.

As noted earlier, animal production is alsobelieved to be a large source of water contami-nation. Unfortunately, there are no good meth-ods to estimate the environmental impacts ofdifferent forms of animal production or wasteremediation options explicitly. In practice,manure from animal production is spread onagricultural fields as a nutrient source, and

T A B L E 7 . S O I L E R O S I O N F R O M C R O P L A N D A N D D E L I V E R Y R A T I O S B YW A T E R S H E D

Sheet and Rill Sediment DeliveryErosion Ratio for Sheet and Wind Erosion Total Soil Erosion

Watershed 1,000 tons per year) Rill Erosion (1,000 tons per year) (1,000 tons per year)

Minnesota River 16,484 0.138 – 0.206 34,543 51,027

Saginaw Bay 2,640 0.171 – 0.216 6,702 9,342

Rock River 5,793 0.157 – 0.171 No estimate 5,793

Source: Erosion data, 1992 National Resources Inventory, USDA; delivery ratios, Ouyang and Bartholic, 1997.


30

erosion control, while not the only factor, isimportant in determining the ultimate disposi-tion of the waste. Assuming that animalmanure is all applied to cropland, then the loadfrom animal operations should be reflected inthe agricultural cropland load estimates.

Total Load EstimatesPutting the point source and nonpoint sourceinformation together and simulating the base-line land use, we produced a baseload estimatefor each watershed by sub-basin. The estimatesby sub-basin vary considerably depending onthe number and size and treatment of pointsources and land use. The total and breakout bysource category for each watershed is shown inFigure 5.

In each of the three case studies, the phos-phorus loads from agriculture are the domi-nant source, ranging from 51 to 69 percent of

the total estimated load. The total phosphorusload for the Minnesota River Valley was cali-brated to estimates made by the MPCA. Sincethe point-source loads are reported by eachfacility, the adjustment was made to the non-point source loads. For Saginaw Bay, ourresults compare well to a prior study done byLimno-Tech (1983) that estimated total phos-phorus loads to the Bay at 989 tons per year.For the Rock River, the approach we used com-pared well with per-acre load averages esti-mated by hydrological models.

In each of the three case studies, thephosphorus loads from agriculture are thedominant source, ranging from 51 to 69percent of the total estimated load.

Nonpoint Agricultural Source LoadOther Nonpoint Source Load Point Source Load

Minnesota River Saginaw Bay Rock River

0

200

400

600

800

1,000

1,200

1,400

Ton

s p

er y

ear

F I G U R E 5 . P H O S P H O R U S L O A D S B Y S O U R C E A N D W A T E R S H E D


31

W orking with our partner organizations,we used a modeling framework devel-oped for this study to analyze the eco-

nomic and environmental potential of alterna-tive policy approaches. Some of the assumptionswere tailored to the data available in each water-shed. The scenarios reported here are the sameto allow cross-watershed comparisons. Sincephosphorus is the principal pollutant of interestin each of these cases, we focus on phosphorusload reductions.

THE LEAST-COST SOLUTION

In the first set of scenarios, we asked the sim-ple question, “What is the least-cost solution fora given level of phosphorus load reduction?”The cost curves developed through this exerciseprovide a benchmark to measure the cost-effec-tiveness of other policy tests. From an economicpoint of view, the least-cost solution presentsthe ideal. The goal is to identify those policiesthat meet the environmental objectives and areas close as possible to the least-cost result.

Figure 6 shows least-cost curves for eachwatershed. These were obtained by running themodel at total load reductions of 10, 33, 50 per-cent and at the maximum potentially attainablefor each watershed. The maximum load reduc-tion is different in each case because of theunique characteristics of each watershed. Foreach of these runs the model found the pointsource and cropping practice options that mini-mize the total cost to produce a load reductionof the specified amount; no policy constraintswere applied.

As shown in Figure 6, significant load reduc-tions are possible at very low costs. Costsincrease slowly until a reduction of 50 percentis achieved, and then rise rapidly thereafter. Thedifference in costs is attributable to the fact thatat lower levels of load reductions there are anumber of relatively minor agricultural changesthat can dramatically cut phosphorus loads. Forexample, for a cut of up to 10 percent, the prin-cipal shift would be from moldboard plowing toconventional tillage without the moldboard. Theelimination of moldboard plowing would dra-matically cut erosion—and therefore phospho-rus loads—at a very low cost. At further loadreductions, conventional tillage would bereduced and conservation tillage practiceswould become dominant. There would also bean increased move to nutrient managementpractices as greater reductions are demanded.

E C O N O M I C R E S U L T S

5

We asked the simple question, “What isthe least-cost solution for a given level ofphosphorus load reduction?”


32

On the point source side, at relatively lowreduction requirements, a few of the largersources that have no systems treatment adoptbiological or chemical phosphorus removal. Asloads decrease, more plants adopt a removalprocess and those that have chemical removalreduce their loads by adding more of the chemi-cal to the process, incurring only somewhatgreater OM&R costs. Finally, at the maximumachievable reduction level, all point sourceshave adopted filtration, incurring the highestlevel of capital upgrade and greater OM&Rcosts. Reductions beyond the endpoint of thecurves are not possible because at this point allpoint sources and agricultural options to reduceloads are employed to the fullest extent speci-fied by the dataset. The opportunities for reduc-tions are exhausted earlier for Saginaw Baybecause a 1 ppm standard for phosphorus isalready in place. The Minnesota River showsthe highest eventual costs because of the pre-

dominance of small treatment facilities and thehigh unit costs of additional treatment.

POLICY TESTS

We tested several policy options that would typi-cally be considered for situations like those pre-sented by the three case studies. These wereintended to contribute to the discussion ontrading and were not endorsed by any stateagencies. These tests include:

• A point source performance requirement. Pointsource controls have been the first avenueof attack to correct water quality problems.This scenario asks: “What if we do more ofthe same?” This scenario is a policy baselinefor comparative purposes. We assumed inthis scenario that all point sources would beforced to adopt a new standard for phospho-rus, except in the case of Wisconsin, where

0 8 12 16 20 24 28 32 36 40 44 48 52 56 60 64 68 72 76 80 84 88

0

10

20

30

40

50

60

70

Cos

t p

er p

oun

d of

ph

osp

hor

us

redu

ctio

n (

$/l

b)

Percent Reductions in Total Phosphorus Load

Minnesota River

Saginaw Bay

Rock River

F I G U R E 6 . L E A S T - C O S T M O D E L S O L U T I O N S


33

the smallest dischargers would be exemptedas a cost-saving measure. In Minnesota andWisconsin, the standard would be 1 ppm.Michigan already has a standard of 1 ppmand would go to 0.5 ppm.

• A conventional subsidy program for agricul-tural conservation “best management practices”(BMPs). Instead of regulatory controls onagriculture, most of the policy effort in theU.S. has focused on providing subsidies tofarmers to help reduce the costs of imple-menting best management practices(BMPs). In this vein, this scenario providesa subsidy for mulch tillage, no-til, and nutri-ent management. We adjusted the subsidylevel (unique in each case) until it inducedan agricultural load reduction for the water-shed equal to that for the point source per-formance requirement.

• A point source performance requirement cou-pled with trading. What if point sourcescould trade with other point and nonpointsources to meet the new standard? This sce-nario adds point and nonpoint trading flexi-bility to the first scenario.

• A trading program coupled with performance-based conservation subsidies. This scenariocombines elements of the second and thirdscenarios, where the burden of reductions isshared between point source and nonpointsources. A key difference however, would bethat the conservation subsidies would bebased upon the attainment of least-cost loadreductions however achieved, not the adop-tion of any particular BMP. For nonpointsource reductions that are applied to non-point source obligations, there is no tradingratio applied.

Table 8 shows the cost results for all the casestudies. Additional details for each case areshown in Tables 9a through 9c. In each case,the strategy of tightening point source perfor-mance requirements is the most expensive

T A B L E 8 . C O S T O F P H O S P H O R U S C O N T R O L U N D E R D I F F E R E N T P O L I C I E S( U S $ P E R P O U N D O F P H O S P H O R U S R E M O V E D )

Conventional TradingSubsidies for Point Program with

Point Agricultural Source Performance-Source Conservation Performance based

Performance Practices Requirement Conservation Least-costRequirement (BMPs) with Trading Subsidies Solution

Minnesota River 19.57 16.29 6.84 4.45 4.36

Saginaw Bay 23.89 5.76 4.04 2.90 1.75

Rock River 10.38 9.53 5.95 3.82 3.22

Source: The levels of phosphorus reduction from the base are different for each watershed.

In each case, the strategy of tighteningpoint source performance requirements isthe most expensive option.


34

T A B L E 9 a . S C E N A R I O R E S U LT S — M I N N E S O T A R I V E R V A L L E Y

Point Source Conventional Trading Performance Subsidies for Point Source Program withRequirement Agricultural Performance Performance-

(1ppm for Conservation Requirement based Conservation Least-costphosphorus) Practices (BMPs) with Trading Subsidies Solution

Total P Load

Reduction (percent) 20 20 29 20 20

Point Source P Load


Agricultural P Load


Total Costs ($M/yr) 9.58 8.06 5.01 2.21 2.15

Average Cost per

Pound ($/lb.P) 19.57 16.29 7.10 4.50 4.36

T A B L E 9 b . S C E N A R I O R E S U LT S — R O C K R I V E R W A T E R S H E D

Point Source Conventional Trading Performance Subsidies for Point Source Program with Requirement Agricultural Performance Performance-


Total P Load


Point Source P Load


Agricultural P Load


Total Costs ($M/yr) 5.9 5.4 4.5 2.2 1.8

Average Cost per

Pound ($/lb.P) 10.38 9.53 5.95 3.82 3.22


35

option. Costs across the studies vary consider-ably, however. Saginaw Bay results are the high-est because there is an existing requirement of1 ppm in place and the requirement simulatedhere is for 0.5 ppm instead of 1 ppm as in theother studies. Minnesota River results are alsorelatively high because there are so many smallsources. In contrast, in the Rock River thesmallest sources are exempt and the costs arequite a bit lower for the same level of control.

Tables 9a-c show estimates for load reduc-tions under each policy test. The new regulatoryrequirements on point sources provide a cut inthe point source load of 70, 71, and 49 percent,respectively for Minnesota River, the RockRiver, and Saginaw Bay. These reductions workout to cuts in the total load of 20, 30, and 16percent. Again, less potential is available frompoint source reductions in Saginaw Bay becausephosphorus is already controlled.

A conventional conservation subsidy programfor agricultural BMPs is somewhat better atachieving the same result, but is still relativelyexpensive except in Saginaw Bay, where winderosion is such a problem and soil conservationhas a greater benefit. The costs of achieving thesame amount of load reduction through untar-geted agricultural subsidies for conservationtillage practices is lower than the point sourceregulation approach, but still more expensivethan other options. Costs are comparativelygreater in the Minnesota River because there ismore adoption of conservation tillage in thiswatershed and wind erosion, unlike Saginaw, isless of a problem. There is also greater use ofthe moldboard plow, the most erosive practice,in the Rock River. These characteristics provideless expensive remediation opportunities.

If agriculture achieved the same absolute cutsas new point source performance requirements,

T A B L E 9 c . S C E N A R I O R E S U LT S — S A G I N A W B A Y

Point Source Conventional Trading Performance Subsidies for Point Source Program with Requirement Agricultural Performance Performance-


Total P Load


Point Source P Load


Agricultural P Load


Total Costs ($M/yr) 7.47 1.80 2.14 1.06 0.55

Average Cost per

Pound ($/lb.P) 23.89 5.76 4.04 2.90 1.75


36

the percentage load reduction would be lessbecause the agricultural load in each case com-prises a larger share of the total. We comparedthe cost per acre required in the model to bringin more conservation tillage with the costs actu-ally paid under existing government programs.In each case, the results were quite close to theactual.

Compared to the least-cost solution, both ofthese approaches are expensive. The least-costsolution relies on performance objectives toachieve the desired result and is otherwiseunconstrained. The first two policy tests areobviously quite far from the most cost-effectiveresult. In the case of new point source require-ments, the reason is that there are many smallpoint sources with diseconomies of scale andtherefore expensive remediation costs com-pared to agriculture. The agricultural conserva-tion subsidy program favors certain practiceswithout regard to performance and is thereforerelatively inefficient. The least-cost solution rep-resents the equivalent of a highly targeted per-formance-based subsidy program. The least-costsolution ranges from 69 to 93 percent lowercompared to tighter regulatory standards onpoint sources and from 66 to 73 percent lowerthan untargeted BMP subsidies.

In contrast, point source performance require-ments, coupled with flexibility to decide whetherto treat or trade with point or nonpoint sources,would lower costs considerably. In this case, theperformance requirement acts as the “cap” in a

“cap and trade” system. This would not repre-sent a “fully closed” cap and trade programbecause not all sources are covered under thecap. Nevertheless, a point source standard withtrading gets much closer to the least-cost solu-tion because it allows the point sources to takeadvantage of the least expensive remediationopportunities, wherever they may be found. Thisscenario assumes that only one third of the loadreduction from nonpoint sources could beapplied against point source requirements (a 3:1trading ratio). The rest of the load reduction pro-duced is essentially an “environmental credit” toassure the achievement of water quality goalsand account for the greater uncertainty inherentin nonpoint source loads. The trading ratio pro-duces a greater total load reduction than anyother policy, an additional 10 percent for eachcase. Even with the environmental credit for theuncertainty of the nonpoint source load appliedto the point source obligations, there is still asignificant savings over the strict regulatorycase. For Saginaw Bay, the costs drop by nearly$20 per pound—82 percent for the pointsources. The other case shows similar thoughless dramatic cost reductions, primarily becauseof the phosphorus requirement already in placein Michigan.

This trading scenario assumes that existingpoint sources would pick up the bill to helpclean up threatened rivers, lakes, and estuaries,raising the question of equity. Is it fair to ask

Point source performance requirements,coupled with flexibility to decide whetherto treat or trade with point or nonpointsources, would lower costs considerably.

Is it fair to ask point sources to pay forremediation simply because that is whereregulatory control is the strongest andpolitically easiest? Should broader social orsectoral responsibility be sought?


37

point sources to pay for remediation simplybecause that is where regulatory control is thestrongest and politically easiest? Should broadersocial or sectoral responsibility be sought?These questions was raised in each of our localadvisory panels.

In response to these issues, we constructed ascenario that mimics burden sharing. This sce-nario, “trading with performance-based conser-vation subsidies,” assumes that the burden forreductions would be borne evenly by point andnonpoint sources. In this case, point sourceswould be responsible for half of the reductionlevel imposed in the other tests. The remainingreduction is assigned to the agricultural sector.We assumed for this scenario that the cost foragricultural reductions would come from con-servation subsidy program funds. However, thesubsidies would be performance-based, so thatthose farmers able to produce the cheapest loadreductions would receive the available funds.This would mean that farmers closest tostreams, with the most highly erodible soils, orwho had not yet adopted conservation practiceswould be the first to receive program funds.Further, we assumed that point sources couldstill purchase nonpoint source reductions.Because agriculture has its own obligation,

however, and would not trade away its cheapestreductions, the results show very little point-nonpoint source trading, but quite a bit ofpoint-point source trading. This scenario effec-tively provides a cap for all sources and so iscloser to a “fully closed” trading system.

Here, both point and nonpoint sources con-tribute evenly to the reduction of phosphorusloads. While most of the total costs would stillbe paid for by point sources, their costs wouldbe cut considerably compared to the previoustrading scenario and even more compared to aregulatory standard. For example, in the RockRiver, the cost would be just $3.82 per pound.The total annual cost estimate to produce thisreduction is about $2.2 million per year, lessthan half of the point source trading programby itself. Of this total cost, about $600,000would be borne by farmers or by public subsidyprograms. The total cut in the phosphorus loadis less because the trading ratio is not applied toreductions made by agriculture on its ownbehalf, as in the case of the trading programwith caps on point sources only. This scenariois the closest to achieving the economicallyideal least-cost solution. Point sources still bearthe majority of the costs, but these are thelowest of any scenario.


39

S U M M A R Y A N DR E C O M M E N D A T I O N S

6

F rom the analysis presented here, it is clearthat if water quality programs aredesigned more efficiently, significant

water quality gains can be achieved at relativelylow cost. A comparison of alternative programsshow that there is a wide range in the cost-effec-tiveness of different approaches, with conven-tional strategies showing the least benefit perdollar spent. More flexible approaches canpotentially provide greater improvements inwater quality, over a larger range of reductions,and at much lower cost.

This is not to say that conventional regula-tory approaches have been a failure in achiev-ing improvements in water quality; they havenot. But our analysis shows that pushing onpoint sources alone would be a relativelyexpensive approach when other sources con-tribute more to the problem. Conversely, in theagricultural sector the opportunities for inex-pensive gains are great if conservation subsi-dies were to be based upon performance, to the

extent that we are able to estimate it. Further,point sources are not the largest contributor tothe problem in the watersheds we looked at, oron a national basis. This means that any strat-egy to reduce the level of loads to restore sur-face waters must include agriculture, not onlybecause of economic efficiency arguments, butalso for the sake of fairness. This may not betrue in every watershed, but it appears to be adominant situation in many rural and urbaniz-ing U.S. watersheds.

Because there is a large differential betweenremediation costs for conventional approachesand programs that involve trading of someform, trading has potential in the watershedswe considered. One would also expect that thereis potential for trading in many other water-sheds as well, because the same reasons forcost-effectiveness would apply. And, as Table 3shows, there are many watersheds in theUnited States that are impaired by nutrients.

More flexible approaches can potentiallyprovide greater improvements in waterquality, over a larger range of reductions,and at much lower cost.

Any strategy to reduce the level of loads torestore surface waters must includeagriculture, not only because of economicefficiency arguments, but also for the sakeof fairness.


40

While a regulatory mandate on point sourcescoupled with a flexible trading program appearsto have merit, an even better program wouldcouple these elements with a strategy directlyinvolving agriculture. In such a program, point-point and point-nonpoint trading would beallowed, but nonpoint sources would have ashared responsibility to undertake remediationactions not coupled to point source regulatoryrequirements. This could take one of severalforms. In one case, agricultural conservationsubsidies, or some share of them, couldbecome part of the pool of funds availablethrough a joint trading program. Farmers,municipalities, or industrial sources who gener-ated credits could sell them through a singleprogram sponsored by the government oranother broker in conjunction with pointsources who wished to purchase and applycredits. Credits purchased by government con-servation funds would be retired.

As a second option, farmers could only gener-ate credits after they had met a minimum stan-dard for agricultural practices. For example, afarmer who used a moldboard plow or appliedmanure to frozen fields in the wintertime couldnot generate credits by abandoning what arebroadly acknowledged to be environmentallyunsound practices. But, farmers moving fromconventional tillage to conservation tillage, orfarmers who currently had sound manure man-agement practices but wished to install moreadvanced ones, could generate credits. Alterna-tively, farmers moving from practices that werebelow standard to ones well above standardcould generate a partial credit. While such aprogram requirement would be entirely volun-tary, it would have the benefit of rewardingfarmers who undertake sound management ontheir own, and would still provide an incentivefor farmers who had not yet made the change.It would have the disadvantage, however, ofreducing cost-effectiveness.

Finally, state agencies could apply a perfor-mance requirement to all sources, includingagriculture, and allow point sources and farm-ers the option of meeting the requirementthrough trading. Point and nonpoint sourceswould have access to the same pool of credits,and conservation subsidy funds could beapplied as before to offset some part of farmers’costs or the cost of operating the program.Farmers who had previously undertaken conser-vation practices would be in compliance andhave no further obligation. Others who couldmake inexpensive reductions would do so, andperhaps do more than their obligation to sellcredits and recover their costs. Some farmerswith high costs or high-valued crops whowished to continue their current practices couldpurchase credits from point or nonpointsources to meet their obligation.

While trading has economic potential, thereare some uncertainties associated with tradingthat need to be acknowledged and accountedfor. The first and perhaps most importantaspect of trading that would involve nonpointsources is that there is a great deal of uncer-tainty involved because the loads are tied toweather events. While point sources producefairly regular flows across seasons and evenyears, nonpoint sources do not. Loads are high-est during rainy seasons and years with highprecipitation, and conversely lower at othertimes. For this reason, a reduction in the loadfrom a nonpoint source may not be equivalentto that from a point source. Therefore, it is

While trading has economic potential,there are some uncertainties associatedwith trading that need to be acknowledgedand accounted for.


41

important that water quality is monitored tomake sure that expected improvements are real-ized and water quality goals are met.

A key advantage of the nonpoint sourcereductions, however, is that they have otherwater quality benefits in addition to the reduc-tion in phosphorus loadings. For example, inthe point source cap-with-trading scenario, atthe same level of phosphorus reductions neces-sary to achieve the P load performance therewould also be a one-third reduction in sedimentloads and nitrogen loads for agriculture.

Issues regarding liability must also be care-fully considered. When a point source fails tomeet a legal requirement, the responsible regu-latory agency has the ability to force theoffender to comply. Any water quality programthat employs trading must similarly provide alegal remedy for those instances when someonesells or applies a credit that has no environmen-tal value. One way to do this is to have sellersvoluntarily accept the regulatory obligations ofthe purchaser for that amount of the load.Alternatively, the buyer can make the seller aloan to be paid in credits. If the credits are notdelivered on the specified date, the loan is duewith interest. If the seller cannot provide thecredits himself, he can purchase them on themarket and turn them over.

Another consideration is that trading pro-grams can be expensive to put in place andoperate if poorly designed. Regulatory paper-work, information gathering and the process ofidentifying partners to trade with can createtransaction costs that are prohibitive and makea trading program ineffectual. Administrativeoversight needs to be sufficient to ensure goodperformance, but not so burdensome as toinhibit trading. Registration of trades should beefficient so that partners can easily hook up,report their trades, and get approval. When

numerous nonpoint sources are involved, somesort of broker—for example, a cooperative—needs to be organized to coordinate the sale ofcredits and to verify them using standard tech-niques. For all these reasons, trading shouldoccur within a regulatory program where rulesand methods are standardized and appropriatereview can be cost-effective, not permit by per-mit, which is expensive.

RECOMMENDATIONS

The potential of market-based policies to pro-duce cost savings and environmental progressappears large enough that much more effortshould be focused on this front. The followingrecommendations could help further theseobjectives.

1. The Environmental Protection Agency andstate environmental agencies should allowtrading among industrial and municipalpoint sources whenever new nutrient reduc-tion requirements are put in place. Theissues with point-point trading are fairlystraightforward and can be effectively man-aged. With the right set of standards, largecost savings can be achieved, as this studydemonstrates. The standards developedshould be written into regulations andensure that the environmental results from atrade are the same, if not better than thatexpected under conventional regulations.Standards should only allow trading within

Trading should occur within a regulatoryprogram where rules and methods arestandardized and appropriate review canbe cost-effective, not permit by permit,which is expensive.


42

the watershed that defines the problem to beaddressed and should also set trading ratiosthat encourage reductions upstream. Tradingshould only result in reduced loads andimproved water quality. In cases where newrequirements are put in place because of awater quality problem, trading ratiosbetween point sources should be close toone. However, if a source wants to increaseits emissions because of expansion orgrowth, higher trading ratios should apply sothat a net gain in water quality results. Wastetreatment facilities should not be able to usetrading to stop using capacity already inplace, but trading could be used to allowfacilities to cover inadvertent and infrequentpermit violations. Facilities that have gener-ally poor records of environmental compli-ance should be prohibited from trading.

2. EPA and the U.S. Department of Agricultureshould provide much greater regulatory andfinancial support for state efforts to developpoint-nonpoint source pilot trading programsand standards. Pilot trading programs shouldbe viewed as laboratories to learn how thepotential of point-nonpoint source tradingcan be achieved. Such pilots, if properly sup-ported, could provide cost savings and otherbenefits, such as sediment reductions, thatcan come from the implementation of betteragricultural practices. Pilots in various set-tings should be supported through appropri-ate regulations and aggressively pursued withtechnical and financial resources from fed-eral agencies. Federal and state agencies andother local stakeholders should participate ascooperative partners. EPA regional adminis-trators should provide oversight of these pro-grams, but allow sufficient flexibility to trydifferent approaches and help find ways tomake these programs work. Evaluation proce-dures should be built into pilot programs. Ifagreed goals are not met within a reasonable

period, such as one permit period of fiveyears, the pilots should be terminated andstandard regulatory approaches applied inthat situation.

3. Concentrated animal feeding operationsshould be treated the same as industrial andmunicipal point sources, which means thatthey should face involuntary permittingprocesses and tight enforcement. Currently,confined animal feeding operations smallerthan 1,000 animal units are not regulated inmost states under the Clean Water Act. This isroughly equivalent to a city of 25,000 to50,000 people. Cities much smaller than thisdo not escape regulatory obligations, and nei-ther should farmers. Given that domestic live-stock produce 130 times the waste of humansin the United States, it is remarkable that only5,000–10,000 of the 1.1 million farms withanimals have to control their livestock wastes.With livestock waste trends going in thewrong direction, the nation will never be ableto improve water quality if such a large sourcecontributing such a great share of the prob-lem is left uncontrolled. Minimum animalwaste standards need to be put in place for allconcentrated operations. These would elimi-nate the most egregious practices such asdrainage and overflow of holding pondsdirectly into waterways, application of wastesonto frozen fields, animal defecation directlyinto streams, and overapplication of manurebeyond the ability of crops to use the nutri-ents. Without exception, large concentratedanimal operations should be treated like pointsources and be subject to the same standardsthat apply to municipal and industrial facili-ties of the same size. They should receiveeffluent limitations and be required to adoptpractices to meet these permit-based limits.

4. Sufficient money from farm income subsidyprograms should be shifted to conservation


43

programs so that environmental problemscaused by agriculture can be remediatedusing market-based approaches. In 1996,farm income support programs werechanged so that payments were based on his-torical payments rather than production. Cut-ting the link between commodity productionand income payments was a useful changethat will encourage the adoption of croppingrotations and less monoculture, which is agood outcome environmentally. However, thenext step needs to be taken—a much greatershare of the billions of dollars spent on unre-stricted farm income support should be tiedto environmental remediation efforts so thatpolluted runoff from agriculture can be con-tained. Farmers and the nation should cometo a compromise to continue support pay-ments, but link the payments to environmen-tal improvements. The programs shouldemploy market-based mechanisms such asauctions, which ensure that the greatest envi-ronmental benefit is achieved for the moneyspent.

5. Monitoring of water quality should be vastlyincreased. A common comment regardingpilot trading programs is that monitoring inthose watersheds should be increased so thatenvironmental performance can be assessed.There is sound logic in increased monitoringto ascertain the results of trading pilots. How-ever, this logic should not be restricted just toinnovative programs, but should extend to allwater quality programs. The scant evidenceavailable suggests that the nation’s waterquality policies are not sufficient to makeprogress toward the goals of the Clean WaterAct. Given the tens of billions of dollars thatare spent on infrastructure and the hundredsof millions on conservation subsidies, it ispenny-wise and pound-foolish to provideinsufficient funding for water quality moni-toring. With the current levels of funding for

monitoring, we will never be able to ade-quately assess the effectiveness of theseinvestments and make appropriate adjust-ments with confidence. There is a great dealof interest in water quality among the generalpublic. Some states and localities have usedthis interest to develop volunteer monitoringprograms.

6. Federal and state agencies should exploreopportunities for trading in large water-sheds. To date, trading has generally beenconsidered in the context of relatively smallwatersheds entirely within state boundaries,such as the ones we examined in this study.However, there are several situations in theUnited States where trading could providemajor cost savings in large basins. Two thatmerit such attention are the Chesapeake Bayand the Mississippi Basin. Drainage fromthe Mississippi is the cause of one of thenation’s biggest water quality problems, the“dead zone” in the Gulf of Mexico caused byoxygen depletion from nutrient enrichment.Trading programs in large basins would pre-sent additional problems, mainly becausetrades would cross state boundaries andhave to be managed at the federal ratherthan the state level. The allocation of loadsand assignment of caps—something neces-sary to set up a trading program—wouldhave to include major and minor watershedsand resemble a large TMDL process. Theadvantage however, would be that for thefirst time policy would be forced to accountfor large-scale performance objectives in aconsistent manner that was ecologically rele-vant. The contributions of various programsto the final objective would have to beassessed as part of a larger frameworkinstead of piecemeal, as is done now. Thiscould do more to force improvements inwater quality than conventional approachesnow being considered.


44

CONCLUSION

The Clean Water Act has succeeded in control-ling some sources of pollution, but with counter-vailing increases from other sources the resulthas been something akin to running in place. Inmany instances the nation has still not achievedthe law’s goals, and the job of cleaning up ourwaterways remains unfinished. One reason we

have not moved to correct our water quality prob-lems is that the conventional remedies are per-ceived as too expensive. Less expensive optionsinvolving other dischargers can produce signifi-cant improvements in water quality at low cost.Trading can provide a flexible alternative andencourage the use of least-cost options.

P A U L F A E T H is Director of the Economics Program at the World Resources Institute. He currently directs sev-

eral efforts: analysis of the cobenefits of climate policies and water quality; an assessment of policies to address the

dead zone in the Gulf of Mexico; collaborative work with industry on climate change; the development of Internet-

based trading for nutrients; ecosystem services and biodiversity protection; and an assessment of trade and its

impact on the environment in Latin America. Faeth was WRI’s Liaison to the Sustainable Agriculture Task Force of

the President’s Council on Sustainable Development organized by President Clinton. Previously, Faeth worked with

the International Institute for Environment and Development, where he applied methods of systems analysis to

examine development projects to improve their economic and environmental performance. He has also worked for

the USDA’s Economic Research Service on issues related to agricultural trade policy.

A B O U T T H E A U T H O R


45

N O T E S

1. The states and other jurisdictions do notuse identical survey methods and criteria torate their water quality. Furthermore, theymay survey different waterbodies every twoyears. As a result, EPA cautions againstcomparing water quality information sub-mitted during different time periods.Though EPA has discouraged aggregatingsurveys because of methodological differ-ences, this is currently the only trend dataavailable.

2. Data for 1970 to 1993 are from livestockand poultry inventories in the USDA Eco-nomics and Statistics System:

Cattle—(http://usda.mannlib.cornell.edu/data-sets/livestock/95131/ct7084.exe, and/ct8595.exe)

Chickens—Layers (http://usda.mannlib.cornell.edu/data-sets/livestock/89007/2/table006.wk1)

Chickens–Broilers—(http://usda.mannlib.cornell.edu/data-sets/livestock/89007/3/table065.wk1)

Hogs—(http://usda.mannlib.cornell.edu/data-sets/livestock/95131/hg4q6778.exe,/hg4q7987.exe, and hg4q8894.exe)

Turkeys—(http://usda.mannlib.cornell.edu/data-sets/livestock/89007/5/table131.wk1).

Data for 1994 to 1997 are from UnitedStates Department of Agriculture-NationalAgricultural Statistics Service, AgriculturalStatistics 1998. Washington, DC: UnitedStates Government Printing Office, 1998.(http://www. usda.gov/nass/pubs/agr98/acro98.htm).

3. Manure production and nutrient contentparameters for this analysis came fromLander et al., (1998) and USDA, 1998b.

4. An animal unit is defined as a cow withcalf, or the equivalent of 1,000 pounds ofliveweight.

5. There is a significant body of literature onemissions trading in general and nutrienttrading in particular that largely describesthe same small set of examples. For paperson emissions trading, see Hahn, 1988. Forsummaries of existing or potential nutrienttrading activities, see Michigan DEQ, 1998,or USEPA, 1996.

6. See USEPA (1996) and the MichiganDepartment of Environmental Quality


46

website (1998a) for more complete lists ofwhere nutrient trading is being considered.

7. The Michigan DEQ prefers to use the term“water quality trading” rather than nutrienttrading, because the explicit intent of theirprogram is to use trading to improve waterquality beyond regulatory requirements.

8. This average is weighted by flow. Only thetwo largest facilities have some phosphoruscontrol, so the majority of plants have aconsiderably higher concentration of phos-phorus in their effluent, although theirshare of the flow is small.

9. The cost curves developed by Doran (1997)are as follows (where Q is flow in mgd). Forcapital costs (in millions of dollars): stan-dard chemical treatment (1.25*Q)0.67; bio-logical treatment (1.0*Q)0.67; for filtrationadd (2.5*Q)0.67. For operation and mainte-nance (in millions of dollars): standardchemical 0.6*Q; biological treatment(0.035*Q)0.7; for filtration add (0.08*Q)0.9.

10. Not all rotations employ each tillage practiceso there are not 40 combinations, whichmight be expected. For example, the soy-bean-wheat rotation is predominantlygrown using conventional tillage only.


47

Adler, Robert W., Jessica C. Landman, andDiane Cameron. 1993. The Clean Water Act 20Years Later. Natural Resources Defense Council.Washington, DC: Island Press.

Arbuckle, J. Gordon et al. 1993. EnvironmentalLaw Handbook, Twelfth Edition, Rockville, MD:Government Institutes, Inc.

Association of Metropolitan Sewerage Agencies(AMSA) and the Water Environment Federation(WEF). The Cost of Clean. Washington, DC,March 1999.

Bacon, Elise F. Use of Economic Instruments forWater Pollution Control: Applicability of PointSource/Nonpoint Source Trading for PollutantDischarge Reductions to Washington State.Apogee Research, Inc. Bethesda, MD: Septem-ber, 1992.

Batchelor, David. Water Quality Program Trad-ing Specialist, Michigan Department of Envi-ronmental Quality, Lansing, MI. Personal com-munication, May 6, 1999.

Bosch, Darrell, Keith O. Fuglie, and Russ W.Keim. 1994. Economic and EnvironmentalEffects of Testing for Fertilizer Management.Staff Report. Economic Research Service,USDA. Washington, DC.

Bosch, Darrell and Krishna Napit. 1992.“Economics of Transporting Poultry Litter toAchieve More Effective Use as Fertilizer.”Journal of Soil and Water Conservation 47(4):335–347.

Carpenter, S.R., N.F. Caraco, D.L. Correll, R.W.Howarth, A.N. Sharpley, and V.H. Smith. 1998.Nonpoint Pollution of Surface Waters with Phos-phorus and Nitrogen, Issues in Ecology. Washing-ton, DC: Ecological Society of America.

Doran, Michael D. POTW Phosphorus ControlCosts. Unpublished memorandum. April 27,1997. Madison, WI: Strand Associates.

Environmental Law Institute. 1997. EnforceableState Mechanisms for the Control of NonpointSource Water Pollution.

Faeth, Paul. 1995. Growing Green: Enhancing theEconomic and Environmental Performance of U.S.Agriculture. Washington, DC: World ResourcesInstitute.

Hahn, Robert W. and Gordon L. Hester. “Mar-ketable Permits: Lessons for Theory and Prac-tice.” Ecology Law Quarterly Vol. 16:361.

Hoag, Dana L. and Jennie S. Hughes-Popp.1997. “Theory and Practice of Pollution CreditTrading in Water Quality Management.” Review

R E F E R E N C E S


48

of Agricultural Economics. Volume 19, Number2, pp. 252–262.

Horner, Richard et al. Fundamentals of UrbanRunoff Management: Technical and InstitutionalIssues. Washington, DC: The Terrene Institute.August, 1994.

Industrial Economics, Incorporated. 1987. Pointand non-point source pollution reduction tradingalong the Lower Minnesota River: Project PlanningBriefing. Unpublished paper prepared for theRegulatory Innovations Staff, Office of Policy,Planning, and Evaluation, USEPA. November16, 1987.

Lander, Charles H., David Moffitt and KlausAlt. 1998. Nutrients Available from LivestockManure Relative to Crop Growth Requirements,Resource Assessment and Strategic PlanningWorking Paper 98-1. Washington, DC: U.S.Department of Agriculture, Natural ResourcesConservation Service. http://www.nhq.nrcs.usda.gov/land/pubs/nlapp1b.html.

Limno-Tech, Inc. Cost-Effectiveness of Conserva-tion Tillage Practices for Improving Saginaw BayWater Quality. Ann Arbor, MI. June 1983.

Metcalf and Eddy in association with Bonestroo,Rosene, Anderik, and Associates, Inc.,Dahlgren, Shardlow, and Urban, Inc., Poquetteand Associates, and Martinez Corp. 1993.Centralization/Decentralization Planning Study.Draft Final Report. Volume 1. MWCC ProjectNo. 920200. Metroplitan Waste Control Com-mission, Twin Cities Area.

Michigan Department of Environmental Qual-ity (DEQ). 1998a. Great Lakes Trading Network,Second Semi-annual Report to the Great LakesProtection Fund. Water Quality Trading—HomePage. Surface Water Quality Division. http://www.deq.state.mi.us/swq/.

Michigan DEQ. 1998b. Part XXX. Water QualityTrading—Draft #9, 11/98. Lansing, MI: Depart-ment of Environmental Quality, Surface WaterQuality Division, Water Quality Trading Work-group Discussion Document. http://www.deq.state.mi.us/swq/.

Minnesota Agricultural Statistics Service.1996. Minnesota Agricultural Statistics, 1996.St. Paul, MN: Minnesota Department ofAgriculture.

Minnesota Pollution Control Agency (MPCA).1997a. Governor Carlson Lauds Innovative Pollu-tion Controls for Rahr Malting Company. PressRelease, January 28, 1997.

——. 1997b. National Pollutant Discharge Elimi-nation System and State Disposal System PermitMN 0031917. St. Paul, MN: State of Minnesota,Minnesota Pollution Control Agency, WaterQuality Division. January 8, 1997.

——. 1997c. Point-Nonpoint Source TradingSummary: Finalized for Rahr Malting Permiton January 8, 1997. St. Paul, MN: State of Min-nesota, Minnesota Pollution Control Agency,Water Quality Division.

——. 1997d. Nonpoint Source Trade CreditingCalculation: Finalized for Rahr Malting Permiton January 8, 1997. St. Paul, MN: State of Min-nesota, Minnesota Pollution Control Agency,Water Quality Division.

National Research Council, Committee onLong-Range Soil and Water Conservation, Boardon Agriculture. 1993. Soil and Water Quality: AnAgenda for Agriculture. Washington, DC:National Academy Press.

Ouyang, Da and Jon Bartholic. 1997. EstimatingSediment Delivery Ratios for Three MidwesternDrainage Basins. Unpublished white paper


49

prepared for the World Resources Institute bythe Michigan State Water Resources Lab.

Ribaudo, Marc. 1992. Water Quality Benefitsfrom the Conservation Reserve Program. Agricul-tural Economic Report Number 606. Washing-ton, DC. Resources and Technology Division.Economic Research Service, USDA.

Stephenson, Kurt and Leonard Shabman. 1996.Effluent Allowance Trading: A New Approach toWatershed Management. Water Science Reporter.Blacksburg, VA: Virginia Water ResourcesResearch Lab, Virginia Tech, for the NationalInstitute for Water Resources.

United States Department of Agriculture(USDA). 1994a. Summary Report. 1992 NationalResources Inventory. Washington, DC: SoilConservation Service.

——. 1994b. 1992 National Resources Inventory,CD #3 Midwest Region. Fort Worth: SCSNational Cartography and GIS Center.

——. 1997a. Agricultural Resources and Environ-mental Indicators, 1996–97. Washington, DC:Economic Research Service, Natural Resourcesand Environment Division. Agricultural Hand-book No. 712.

——. 1997b. United States Agricultural Census.Information obtained from http://www.nass.usda.gov/census/ for general information andhttp://govinfo.library.orst.edu/cgi-bin/ag-state?Minnesota (or Michigan or Wisconsin) to findcounty by county data for Minnesota. Informa-tion downloaded during 1997.

——. 1998a Agricultural Statistics (variousyears). National Agricultural Statistics Service.http://www.usda.gov/nass/pubs/agstats.htmhttp://www.usda.gov/nass/pubs/agr98/acro98.htm

——. 1998b. National Engineering HandbookPart 651: Agricultural Waste Management FieldHandbook. Washington, DC: Natural ResourcesConservation Service (http://www.ncg.nrcs.usda.gov/tech_ref.html)

United States Department of Commerce. 1998.Statistical Abstract of the United States: 1998.118th Edition. Washington, DC.

United States Environmental Protection Agency(USEPA). 1980. 1980 Needs Survey: Cost Esti-mates of Construction of Publicly-Owned Waste-water Treatment Facilities. EPA430/9-81-001.Washington, DC: Office of WastewaterManagement.

——. 1992 Report to Congress.

——. 1992a. Incentive Analysis for Clean WaterAct Reauthorization: Point/Nonpoint Source Trad-ing for Nutrient Discharge Reductions. Washing-ton, DC: Office of Water and Office of Policy,Planning and Evaluation.

——. 1992b. TMDL Case Study: The Lower Min-nesota River. EPA841-F-92-011. Washington DC:Office of Water (WH-553).

——. 1994. President Clinton’s Clean WaterInitiative: Analysis of Costs and Benefits. EPA-800-R-94-002. Washington, DC: Office ofWater.

——. 1996. Draft Framework for Watershed-Based Trading. Washington, DC: Office ofWater. EPA-800-R-96-001.

——. 1997. Index of watershed indicators.http://www.epa.gov/surf/hucinfo. Visited onOctober 31, 1997.

——. 1999. 1998 TMDL Tracking System DataVersion 1.1.


50

http://www.epa.gov/owow/tmdl/trcksys.html.Visited on July 12, 1999.

United States Senate. 1997. Animal Waste Pollu-tion in America: Environmental Risks of Livestockand Poultry Production. Report Compiled by theMinority Staff of the United States Senate Com-mittee on Agriculture, Nutrition, and Forestryfor Senator Tom Harkin (D-IA), Ranking Mem-ber. Washington, DC.

Williams, J.R. et al. 1989. EPIC—Erosion/Productivity Impact Calculator: 2, User Manual.Technical Bulletin No. 1768. A.N. Sharpley andJ.R. Williams, eds. Washington, DC: USDA.

ADDITIONAL REFERENCES

Anderson, 1999 (see EPA website)

WRI’s Board of DirectorsWilliam D. Ruckelshaus, ChairmanJulia Marton-Lefèvre, Vice ChairManuel Arango

Frances G. Beinecke

Bert Bolin

Robert N. Burt

David T. Buzzelli

Deb Callahan

Michael R. Deland

Sylvia A. Earle

Alice F. Emerson

José María Figueres

Shinji Fukukawa

David Gergen

John H. Gibbons

Paul Gorman

William M. Haney, III

Denis Hayes

Cynthia R. Helms

Samuel C. Johnson

Calestous Juma

Yolanda Kakabadse

Aditi Kapoor

Jonathan Lash

Jeffrey T. Leeds

Jane Lubchenco

Mahmood Mamdani

William F. Martin

Matthew Nimetz

Ronald L. Olson

Yi Qian

Peter H. Raven

Florence T. Robinson

Jose Sarukhan

Stephan Schmidheiny

Bruce Smart

Scott M. Spangler

James Gustave Speth

Ralph Taylor

Alvaro Umaña Quesada

Pieter Winsemius

Wren Wirth

Jonathan Lash

PresidentMatthew Arnold

Senior Vice President and Chief Operating OfficerAnthony Janetos

Senior Vice President and Chief Program OfficerMarjorie Beane

Vice President for Administration and Chief Financial OfficerLucy Byrd Dorick

Vice President for Development

Kenton R. Miller

Vice President for International Development and ConservationDonna W. Wise

Vice President for Communications

The World Resources Institute (WRI) is anindependent center for policy research andtechnical assistance on global environmentaland development issues. WRI’s mission is tomove human society to live in ways thatprotect Earth’s environment and its capacityto provide for the needs and aspirations ofcurrent and future generations.

Because people are inspired by ideas,empowered by knowledge, and moved tochange by greater understanding, theInstitute provides—and helps otherinstitutions provide—objective informationand practical proposals for policy andinstitutional change that will fosterenvironmentally sound, socially equitabledevelopment. WRI’s particular concerns arewith globally significant environmentalproblems and their interaction witheconomic development and social equity atall levels.

The Institute’s current areas of work includeeconomics, forests, biodiversity, climatechange, energy, sustainable agriculture,resource and environmental information,trade, technology, national strategies forenvironmental and resource management,and business liaison.

In all of its policy research and work withinstitutions, WRI tries to build bridgesbetween ideas and action, meshing theinsights of scientific research, economic andinstitutional analyses, and practicalexperience with the need for open andparticipatory decision-making.

WORLD RESOURCES INSTITUTE10 G Street, N.E.Washington, D.C. 20002, USAhttp://www.wri.org/wri

World Resources Institute

fertile ground - world resources institutepdf.wri.org/fertile_ground.pdf · fertile ground:...

Documents