NC0000272_BenefitsCostsColorReductionReport_19880401

ieiria
NATIONAL ECONOMIC RESEARCH ASSOCIATES,INC.
CONSULTING ECONOMISTS
BENEFITS AND COSTS FROM THE REDUCTION
OF COLOR EFFLUENT FROM THE CHAMPION MILL
INTO THE PIGEON RIVER
PREPARED BY
NATIONAL ECONOMIC RESEARCH ASSOCIATES, INC.
AT THE REQUEST OF
CHAMPION INTERNATIONAL CORPORATION
REVISED APRIL 1988
A MARSH&MCLENNAN COMPANY
WHITE PLAINS,NY•WASHINGTON•LOS ANGELES•PALM BEACH•ITHACA,NY•LONDON

TABLE OF CONTENTS
Page
I. INTRODUCTION AND SUMMARY OF RESULTS 1
A. Introduction 1
B. Summary of the Economic and Employment 2
Impacts of Discharges from the Champion Mill
C. Evaluation of the Bach and Barnett Analysis 4
D. Organization of the Report 4
II. BENEFITS 4
A. Overall Approach 5
B. Rafting and Floating 7
C. Fishing and Sightseeing 13
D. Property Values 14
E. Apportioning Benefits 16
F. Nonuse Value 18
III. COSTS 21
A. 50 Percent Reduction 21
B. 50 Units at the Tennessee Border 22
C. 50 Units at the Mill 24
IV. COSTS AND BENEFITS 25
V. EMPLOYMENT CONSEQUENCES 27
VI. EVALUATION OF THE BACH AND BARNETT REPORT 30
A. Discounting 30
B. Indirect Benefits 31
C. The Costs of Color Reduction 34
D. Financial Impact on Champion 34
International Corporation
REFERENCES 35
1Le/1'/d®

BENEFITS AND COSTS FROM THE REDUCTION
OF COLOR EFFLUENT FROM THE CHAMPION MILL
INTO THE PIGEON RIVER
I. INTRODUCTION AND SUMMARY OF RESULTS
A. Introduction
Champion International Corporation owns and operates a paper mill at
Canton, North Carolina that is sited on the Pigeon River. The mill is about 37
miles upstream of the Tennessee and North Carolina border. The United States
Environmental Protection Agency (USEPA) has proposed an NPDES permit for the
mill that would significantly reduce the amount of color allowed to be discharged
from the mill into the Pigeon River. One of many submissions in this permit
proceeding was a report by Bach and Barnett entitled, An Economic Impact Analysis
on the Recreational Benefits of a Restored Pigeon River (see Reference [1]). This
study concluded that levels of reduction consistent with the April 1987 USEPA
proposal permit would be economic. In December 1987, USEPA published a second
proposal which limited color discharges even more severely than the April proposal.
At the request of Champion, National Economic Research Associates, Inc.,
an economic consulting firm specializing in energy and environmental economics,
evaluated the economic impacts of three color reduction strategies and also reviewed
and evaluated the Bach and Barnett report. First, we estimated the benefits and
costs under three strategies proposed to reduce color discharges from the Champion
mill. One of the plans was proposed by Champion; two were proposed by USEPA.
The most stringent plan (i.e., the December proposal by USEPA) would certainly
lead to plant shutdown. The less stringent USEPA plan would likely cause plant
shutdown. All the plans had costs significantly greater than benefits, and both
USEPA plans would have substantial unemployment impacts. However, Champion's
proposal yielded benefits that were closest to cost and exhibited small net gains in
employment. Second, based on our review and evaluation of the Bach and Barnett
n/e/r/w

-2-
analysis, we have concluded that there were several fundamental errors that
completely undermine their conclusion that substantial reductions in color discharges
could be justified on economic grounds.
B. Summary of the Economic and Employment Impacts of Reducing Color
Discharges from the Champion Mill
Table I summarizes our results. This table assumes that Champion would
take all actions technologically feasible to keep the plant open. The plan proposed
by Champion would reduce color discharges from the mill by 50 percent on a long-
term average basis. The costs associated with this program would be $38.4 million.
This is the present value of the costs over a 10-year period expressed in 1988
dollars. These costs would be over three times the benefits of $11.3 million. Thus,
an expenditure of one dollar achieves only 30 cents in benefits.
The two USEPA proposals would remove more color from the effluent
discharge but at substantially higher costs. Also, costs would exceed benefits by a
far greater margin than under Champion's proposal. Under the less stringent of the
two, which would consistently achieve less than 50 color units in the Pigeon River
at the Tennessee border, the cost would be $164.5 million, with corresponding
benefits of only $18.0 million. This results in a benefit of 11 cents per dollar of
expenditure. The more stringent proposal, contained in the second draft USEPA
permit, would limit discharges at the mill to 50 units of color. This proposed
effluent limit cannot be achieved with current technology and therefore would
require that the mill be shut down. The direct cost of this proposal would then be
the cost of building and operating a new facility in a different region net of any
operating and maintenance savings attributable to the new investment. We have
estimated these costs at approximately $290.0 million. The benefits under this
program would be $18.3 million. Thus, for every dollar of expenditure, this plan
would yield benefits of about one cent. These cost estimates do not include the
Tlih/I'ra

-3-
substantial economic and social cost impacts on the city of Canton, western North
Carolina or eastern Tennessee.
There are several reasons for the striking divergence between costs and
benefits. On the benefit side, only the relatively small stretches of the Pigeon
River which are immediately downstream of the Tennessee border have good
recreational potential. The whitewater rafting run of five miles is comparatively
short. The river is usable for rafting only when the `hydroelectric electric
generation demands at the Walter's Lake plant of Carolina Power and Light (CP&L)
result in adequate releases. Also, other factors unrelated to color tend to make the
river less attractive: the most scenic part of the river in North Carolina has
virtually no water in it at all and is not affected by color discharges from the mill
because of the CP&L water diversion tunnel; and low summer river levels,
particularly in North Carolina, restrict its suitability for most recreational activity.
On the cost side, the plant is already a low color producer in terms of
pounds of color effluent per ton of product. The incremental removal of color
requires increasingly more expensive and untested methods. The low river flows
make the achievement of the lowest level of color impossible without a complete
shutdown of the mill.
Employment is another measure of the impacts of the discharge reduction
plans. The Champion proposal would increase recreational-related employment in
Tennessee by 77 jobs. The other two plans would see net recreational employment
increases of about 123 jobs. Based upon U.S. Department of Commerce figures, the
50 percent reduction plan would increase the number of jobs in North Carolina by
176 jobs. If the mill were to shut down, as would be required under the USEPA
December proposal and would be likely under the earlier proposal, the loss of
employment in North Carolina would total over 12,000 jobs, with an added loss in
nle✓rlw

4-
Tennessee of over 800 paper-related jobs. Thus, under either USEPA proposal,
Tennessee would experience a net loss of almost 700 jobs.
C. Evaluation of the Bach and Barnett Analysis
Bach and Barnett estimate the direct and indirect recreational benefits
from reducing color discharges from the Champion mill to meet 50 color units at
the state line and the costs of achieving these benefits. Based on their analysis,
they conclude that reducing the color discharges to that level is justified on
economic grounds.
We have reviewed and evaluated their analysis and found it to be
seriously flawed. There are several key errors that cause them to incorrectly
conclude that the level of color reduction they evaluated would be justified on
economic grounds. They have failed to discount benefits and have overstated and
miscalculated indirect economic benefits. In addition, they have utilized company-
wide rather than plant-specific data to evaluate local impacts.
D. Organization of the Report
The remainder of the report is divided into five sections. Section II
describes the calculation of benefits. Section III discusses the costs of each plan.
Section IV compares the costs and benefits. Section V calculates employment
impacts. Section VI evaluates the Bach and Barnett analysis.
II. BENEFITS
This section discusses how the benefits from reducing color discharges
into the Pigeon River are calculated. We begin by describing our overall approach.
This is followed by a discussion of our estimates of the benefits from greater
recreational use and improved property value on the assumption that color
discharges are eliminated entirely. We then determine the benefits associated with
each of the color discharge reduction programs we evaluated. The section concludes
ia/ea'ao

-5-
with a discussion of other possible sources of benefits--option values and existence
values.
A. Overall Aonroach
There are two issues regarding the relationship between reductions in
color discharges into the Pigeon River and improvements in recreational benefits
and property values. First, it is not clear to what extent reductions in color
discharges from the Champion mill would be perceptible in the river. It is even
less clear that incremental improvements, say from 100 to 50 units on average,
would be noticeable in the river. Second, even if these changes were noticeable,
there is no direct evidence that they would affect either recreational activity or
property values.
Bach and Barnett allege, however, that color levels in excess of 50 units
reduce the attractiveness of the Pigeon River for recreational uses (whitewater
rafting in particular) and may also reduce property values for land adjacent to the
river. It is important to note that these supposed adverse effects are entirely
hypothetical. Despite an extensive review of the literature we can find no empirical
studies which relate color to recreational activity. Indeed, we have limited evidence
for certain naturally very dark rivers (the Androscoggin River in Maine for
example) on which whitewater rafting and fishing are quite popular activities.
While there is some limited support in the economic literature for a relation
between color and property values, this may simply reflect the statistical association
between color and other unrelated aspects of water quality.
Despite our reservations about the actual impacts of reducing color
effluents, for the purposes of this study we have adopted Bach and Barnett's thesis
and have assumed that color reduction will enhance recreational activity and
rL4EV /W

-6-
property values on the Pigeon River and have attempted to quantify the maximum
possible impacts of color changes.
The time available for this study has made a direct investigation of the
Pigeon River infeasible. Consequently, we have relied on data for other rivers to
estimate the benefits from reducing color discharges into the Pigeon River. In our
effort to transfer other results to the Pigeon River we have always allowed for the
maximum benefit potential. The actual benefit levels are quite likely to be lower.
Several recreational activities are feasible on the Pigeon River, including
whitewater activities (rafting, canoeing and kayaking), floating (primarily rafting
although it includes "inncrtubing," canoeing and kayaking), fishing and general
sightseeing. To calculate the increase in usage for these activities in the absence
of color effluent we used estimates of current usage on Tennessee rivers in the
vicinity with no perceived elevated color levels, particularly the Ocoee and
Hiawassec Rivers. Adjustments must be made to these estimates since these rivers
are different from the Pigeon River in respects other than color.
To estimate the value of a visit we have utilized a variety of published
sources. For the most part, they are the results of contingent valuation studies.
In these studies participants in an activity are asked to estimate the value of that
activity to them in excess of the costs incurred. These estimates, known as
consumer surplus, serve to define the value of a day. Bach and Barnett also employ
consumer surplus valuation. Multiplying the value per day by the increased number
of days gives our estimate of benefits from reducing color discharges.
Property value benefits presumably derive from scenic advantages of being
along a river of reduced color as well as the enhanced levels of wildlife and
recreation the river can support. Several studies [2,3,4] have related color and
general water quality to property values along coastlines and lakefronts. We have
n/e/ram

-7-
calculated current values of river-front property along the impacted portion of the
Pigeon River and have used these studies to estimate the increase in property
values attributable to a river with no added non-natural color.
B. Rafting and Floating
Whitewater rafting is feasible on a 4.5 mile stretch of river about
37 miles below the Champion mill--from River Mile (RM) 26 at the CP&L
powerhouse to RM 21.5 in Hartford, Tennessee (see Figure 1). As recently as last
summer, a commercial supplier offered whitewater rafting trips on this stretch. The
river below this point has too low a gradient to support whitewater rafting, while
over the 37 miles between the mill and the power plant, both the gradient of the
river and the volume of water are insufficient to support rafting.
Rafting on the Pigeon River is further limited for two reasons. First,
the climate is conducive to rafting only from April to October. Second, during this
period rafting is limited to periods of discharge from the CP&L power plant. When
the plant is not operating (or is operating only one of its three turbines) the
depths are insufficient to support rafting. CP&L's obligation as a public utility is
to generate power so as to minimize electric generation costs without regard to
potential rafting activity. However, for the purposes of this study we have assumed
that CP&L generates electricity during as many daylight hours as water levels in
Walter's Lake permit during the rafting season.
The relatively short length of the Pigeon River run makes whitewater
rafting less attractive than it would be on longer rivers. First, no more than a
half-day trip is possible, while other rivers offer full day trips. Second, since
congestion limits the number of potential rafters per mile, shorter rivers will have
lower capacities.

. -8-
To estimate the number of potential whitewater rafting trips for the
Pigeon River we used data for the Ocoee River, which is the closest comparable
river to the Pigeon River. The whitewater rafting portion of the, Ocoee River, like
the Pigeon River, is five miles long, allowing only half-day trips. In addition, it is
also dam-constrained. Total usage on the Ocoee River for 1987 was about 120,000
visits [5]. Following Bach and Barnett, we take this figure as a limit to usage on
the Pigeon River. We assumed that there is no current recreational usage on the
Pigeon River and that over five years usage for whitewater activities would rise to
120,000 visits per year. Thus, we have assumed that all recreational use of the
Pigeon River would be attributable to reductions in color discharges. This obviously
tends to overstate the benefits from lower discharges, since there is currently
recreational usage of the river. The attached brochure (Figure 2) illustrates the
recreational benefits of the Pigeon available today. These visit estimates are
displayed in column (1) of Table II.
To estimate the value for a rafting day we used a study conducted in
Colorado in 1978 [6]. This study found that the consumer surplus from a
whitewater rafting trip averaged $10.94 per person per day. Once again, this
represents the difference between what people would be willing to pay and what
they actually had to pay for a rafting trip. The estimate was made by eliciting
from rafters the amount of money in addition to what they had paid that day that
would cause them to forego that trip on these rivers. Converting this value to 1988
dollars, we find that the benefits are $19.28 per tourist day. Other whitewater
activities, for example, kayaking and canoeing, had similar values.
The equations from which these estimates were drawn show consumers to
be quite sensitive to congestion on the river. Since congestion levels on the Ocoee
n/e✓r/W

-9-
River are much higher than those observed in Colorado, we have lowered our
estimate to $13 per trip.
In addition to the five-mile stretch of whitewater between the power
plant and Hartford, there is a more placid stretch from RM 21.5 to RM 16.5 on
which floating is feasible during periods of higher water. On this stretch the river
resembles the Hiawassee River. We have estimated floating usage for the Pigeon
River based on data for the Hiawassee River. While this activity may be expected
to be constrained somewhat by power generation, water levels further downstream
are less affected by dam activities because of the lower gradient. Nonetheless,
river levels are probably too low to achieve any significant recreational activities in
August or September. Given that these months generate 35 percent of
noncommercial activity, we have reduced the Hiawassee River floating usage of
80,000 visits [7] by 35 percent to 52,000.
Unlike whitewater rafting, which has been growing at about 6 percent per
year on the rivers for which we have visitation data over time, floating activity
seems to be on the decline. On the Hiawassee River, usage has been declining by
about five percent per year since 1978. In estimating floating usage for the Pigeon
River, we have conservatively assumed no decline. Again, we assumed growth to
capacity would occur over a five-year period. The resultant demands appear in
column (2) of Table II.
Bach and Barnett assume a value of $12.94 per tourist day for floating
activity, based on Water Resources Council criteria. These values seem quite
severely overstated. Based on our visits to the river, a score of 36 would seem to
be a better indication of the value than their assigned score of 62. In particular:
1) Bach and Barnett assign a score of 10 out of 30 for recreation
n/e/rra°

-10-
experience, which measures crowding. The large crowds projected
here should reduce this value to 5.
2) Bach and Barnett assign a score of 12 out of 18 for availability of
opportunity. A 12 would indicate no floating opportunities with one
hour, which seems to ignore the French Broad and Hiawassee Rivers.
We find 4 to be a more objective classification.
3) Bach and Barnett assign a score of 10 out of 14 for carrying
capacity. This rating reflects "optimum facilities to conduct activity
at site potential." These facilities are simply nonexistent. A 5,
representing basic facilities, seems quite generous.
4) We accept Bach and Barnett's accessibility rating of 12 out of 18.
5) Their rating of an 18 out of 20 for outstanding esthetic quality seems
highly inflated. The river would still receive municipal waste
treatment discharges and both agricultural and urban runoff. We
have lowered their rating to 10.
The score of 36 translates to a value of $8.30 in 1982 dollars, converted
to $9.76 in 1988 dollars. We have used this value in our estimate for floating
activity.
Table II describes the annual benefits and the present value of these
benefits in 1988 dollars. The present value reflects the annual benefits over the
next 10 years discounted to reflect the time value of money. We used a 10-year
time frame since that is about the expected life of the investments for color
discharge reduction. Discount rates reflect the rate at which benefits can be traded
between years. We have discounted the benefits at an annual rate of 6.79 percent
in excess of inflation. The rate used reflects the after tax cost of money to
Champion (13.2 percent) net of a presumed six percent future inflation rate. The
nm✓r/aa

. -11-
Champion rate reflects the marginal cost of capital in the private sector, since
funds spent cleaning up the Pigeon River would otherwise. earn these private sector
rates. The benefits of cleanup should consequently be discounted at the same rate.
Using this method, we derive total benefits over the next 10 years of $8.9 million
for rafting and $2.9 million for floating. The annual benefits and the present value
of the benefits for whitewater activity and floating are given in columns (5) and (6)
of Table II.
We are quite confident that these estimates overstate the actual changes
in value on the Pigeon River. There are at least six reasons why we expect the
value of additional whitewater activity and floating visits to be substantially lower.
Fib we have assumed that there is currently no recreational usage on
the Pigeon River. In fact, there is a commercial operation on the river today.
Kayaking and canoeing have also been observed on the river. However, we have
been unable to secure estimates of current usage and have therefore ignored them.
Since our goal is to calculate increases in usage, the value of all current usage
should be' subtracted from our value estimates. Our failure to do so clearly
overstates the number of trips that can be attributed to reductions in color
discharge and thereby overstates benefits.
Second the benefits per trip used for whitewater activity were estimated
on much larger and more striking rivers in Colorado. The set of rivers on which
this estimate was based includes, for example, the Yampa River, which "is among
the few rivers in the U.S. on which rafters can spend from three to five days
without encountering roads, private land, or other evidence of civilization." [6,
page 8] The rivers sampled averaged 42 miles in whitewater length, over eight
times the length of the Pigeon River run. Trips on these rivers averaged 14 miles
per day. The shortest of the rivers was almost three times the length of the
nje/Tids

12
Pigeon River. Such unique characteristics would be expected to make this set of
rivers much more valuable than the Pigeon River.
Third the $13 consumer value assumes a very modest increase in
congestion. The value of a whitewater rafting day declines sharply with the number
of people encountered on a trip. The Colorado study users encountered an average
of only forty people on a trip. With visits of 120,000 per year, Pigeon River
whitewater rafters might well encounter hundreds of people per day. In the
Colorado study, a doubling of encounters reduces the benefits per day by over
40 percent. Thus, fairly moderate increases in usage could lower the benefit per
day estimate substantially.
Fourth of the two methods traditionally used to estimate the value of
recreational activities, we relied on the contingent valuation method which usually
leads to higher values. It is subject to a number of well-known biases in which
people tend to overstate the value of services. Alternatively, travel cost models
which estimate the demand for a service by the distribution of distances people
were willing to go to participate in an activity, typically give lower values than
contingent valuation.
Fifth the current physical facilities at both the dropoff and takeout
areas are grossly inadequate to the task of accommodating the estimated number of
rafters or floaters. We assume that these facilities could be developed over time if
demand warrants but have not included the costs of developing these facilities.
These costs would properly be a subtraction from the benefits associated with these
activities.
Sixth, increased recreation visits on the Pigeon River may consist largely
of visits currently made to other rivers. Thus, in equilibrium, the total increase in
whitewater rafting trips might be much less than the 120,000 visits per year
Tl�L/Zi`d,'a

-13-
projected on the Pigeon River even if that many people do raft on the Pigeon
River. Each visit "cannibalized" from the Ocoee, Hiawassee, or any other river
represents less new value than a new trip; most of its value reflects a shifting of
value from one site to another.
For these reasons we consider our estimates of both usage and value per
trip to be quite generous and would expect to see far less rafting and floating
usage of the Pigeon River.
C. Fishing and Sightseeing
For these activities we have used both the visit estimates and value
estimates of Bach and Barnett as a base. They have scaled fishing visits to usage
estimates for the Norris Dam tailwaters put-and-take fishery. Their sightseeing
visits are scaled to estimates for the Hiawassee River. The number of visits for
these activities are given in columns (3) and (4) of Table II.
Our only adjustment is to lower fishing values to account for the costs
of the presumed put-and-take fishery. Failure to include these costs overstates
potential net benefits. One study [61 suggests that these costs can range from
$4.50 to $11.00 per visitor day depending on the length of the fishery. We have
used the average of this range to reduce Bach and Barnett's estimate of the net
benefits for a fishing day of $17.00 to $9.25.
Further, while we have used Bach and Barnett's estimates for the number
of additional fishermen expected, the source of their estimate is unclear. A proper
estimate should take into account the fact that trout are not a viable species in the
Pigeon River below the Canton Mill discharge and that put-and-take fisheries are
generally less efficient in smaller bodies of water. We have accepted their
estimates for lack of better data.

-14-
It should be noted that fishing is a viable activity on the more scenic
portions of the river only when whitewater rafting is unavailable. Excessive stream
flow and heavy boating usage are detrimental to fishing quality. We do not regard
this as a very serious constraint given the limited time available for whitewater
rafting and have therefore made no adjustment.
The annual benefits and the present values for fishing and sightseeing are
described in columns (7) and (8) of Table II. The present value of fishing benefits
is about $1.2 million over ten years, while sightseeing accounts for about
$0.7 million.
D. Proyerty Values
It has been argued that reductions in color can enhance residential
values. A 1979 study [2] in Pennsylvania, which focused on overall water quality
rather than color, indicated that a completely clean river raises adjacent residential
property values about 28 percent. Similar values were obtained for apparent color
in a 1985 Michigan study [3] on lakefront property and for a combined index of
turbidity and distance in a 1980 study [4] of beachfront property in Massachusetts.
We have used the 28 percent figure as an upper bound on residential
property improvements on the Pigeon River. We have further assumed that all
property along the Pigeon River not currently blocked by roads, public land, or
Champion property is transformed to residential use. This is obviously an extreme
assumption and causes us to overestimate the impacts of reduced color discharge on
property values.
Our calculation of increased property values had three steps: 1) an
inventory of property along the Pigeon River, 2) a calculation of its value for
residential purposes, and 3) a calculation of the maximum increase in value
attributable to reduced color.
n/e✓r/W

. -1s-
1. Much of the land adjacent to the Pigeon River is publicly owned.
Route I-40 follows one side of the river while public forest land borders the river
in many areas. The land between the dam and the powerhouse which contains no
water due to the powerhouse diversion tunnel and is, therefore, unaffected by color
was also excluded. Only about 50 percent of the river is adjacent to usable private
land. We have also excluded plots belonging to Champion and the Cocke County
School District. Since the usage of these properties cannot be expected to change
with reductions of color in the Pigeon River, we have excluded this land from our
assessment.
2. We had assessed property values for land in Tennessee only and have
assumed that similar values prevail in North Carolina. Residential land and property
was valued at $52.49 per river-foot of frontage. We have assumed that all land
along the river would have this value in residential use. In fact, nonresidential
land and property in Tennessee had a value of only $32.01 per river-foot.
To calculate increases in the value per river-foot of frontage, we used
the 28 percent figure. This increases the value of all potential residential property
by $14.64 per river-foot to $67.13. The results of this calculation are given in
Table III. Property value increases, which are assumed to be instantaneous, total
$2.52 million in North Carolina and $2.05 million in Tennessee.
Again, we feel it quite likely that this estimate overstates potential
benefits. Fi= we have applied residential-level benefits to agricultural and
commercial land. This is based on the assumption that this land may ultimately be
converted to residential use. The available studies deal exclusively with residential
values. The links between property value and color are far less obvious for land
used for agricultural, forestry, or industrial purposes. However, it is unclear that
I1�2/Y'id,�

-16-
private land along the Pigeon River currently used for agricultural or industrial
purposes will be converted to residential use.
Second the 28 percent increase cited is partly attributable to the fact
that more expensive homes are constructed on more desirable land. The net
benefit, which would subtract the cost differentials, is not included here.
Third color may not be the true measure of property desirability but may
instead be a proxy for other measures of environmental quality. A 1973 study [8]
found color much less important to property owners than the ability to sustain fish
and other wildlife. The Michigan study of lakeside property owners found that
turbidity was an important factor in value but notes that turbidity is highly
correlated with other pollution variables which decrease both fishing and swimming
uses. To the extent that these variables are not correlated here, the effects might
well be less.
Fourth we have included several plots of land currently inaccessible by
road. Several areas of vacant land doubtless have no developed residential
infrastructure--for example, sewage, telephone and water access. The costs of
providing such facilities should be subtracted from any property value increases.
Given these caveats, we find it quite likely that the combined property
value effects will be lower than what we have estimated.
E. Atmortionina Benefits
The recreational and property benefits we calculated in Sections II.B.
through II.D. total $18.3 million over 10 years. These estimates are based on the
complete elimination of color discharges from the Champion mill. The next step is
to determine the proportions of these benefits that will accrue from each of the
specific proposed reduction plans.
11�/IYc1i°

-17-
Under the most stringent reduction plan, discharges of color would be
limited to no more than 50 units at the pipe. This strategy entails a reduction of
311,000 pounds of color per day on average. We have adopted the Bach and Barnett
assumption that 50 units is the limit of color acceptability. Consequently, we have
assumed that the 50 units at the pipe strategy achieves all of the $18.3 million in
benefits.
For the other two strategies, we have assumed that the benefits achieved
are proportional to the reduction in discharges. Further, for the 50 units at the
border scenario, we have assumed that the plant remains open. Thus, the strategy
•that achieves 50 units at the border would achieve all the potential benefits in
Tennessee. This includes all of the recreational benefits as well as the increase of
$2.05 million in property values in Tennessee. In addition, since this plan also
achieves 88 percent, of the discharge reductions under the 50 units at the pipe
strategy (274,000 divided by 311,000), we assume it achieves 88 percent of the
property value benefits in North Carolina. As a result, the total benefits from the
50 units at the border strategy are $18.0 million.
The 50 percent reduction plan removes 172,000 pounds of discharge per
day on average. This achieves 63 percent (172,000 divided by 274,000) of the
Tennessee benefits and 55 percent (172,000 divided by 311,000) of the North
Carolina benefits. Thus, the total benefits from this program are $11.3 million.
This proportionality assumption is somewhat arbitrary. In the absence of
specific recreational studies on color, it is unclear whether a 50 percent reduction
in effluent will garner more or less than 50 percent of the benefits. Instead, there
might be threshold, effects. That is, below a certain color level, all benefits accrue;
above a certain level, no benefits accrue. While 50 units in the stream may
represent the lower threshold, we have no such upper threshold. Further, even the
I1�E',/ZSd�

' -1s-
50 unit lower threshold is controversial. However, we have tested several different
upper thresholds in benefits. The assumption of proportionality yields larger
incremental benefits than any other set of assumptions tested.
F. Nonuse Value
It is often asserted that there are values other than those accruing to
users of the. river. Bach and Barnett cite a study which estimates those benefits at
five times the direct benefits. We reject the notion that there are significant
nonuse benefits for the Pigeon River. Fib it is unclear whether nonuse values for
the Pigeon River are positive or negative. Second even if these benefits are
positive, they are not likely to be large.
The economics literature considers two components of nonuse values:
option values and existence values. Option value is similar to a nonrefundable
admission price that must be paid before one knows whether or not he or she will
use the river. It represents for consumers the difference between (1) the
willingness to pay to shield themselves from uncertainties about their demands for
the river and its availability and (2) the expected value of their benefits from the
river.
Substantial theoretical research indicates that option value, this
difference between willingness to pay and expected value of benefits, is as likely to
be negative as positive. Further, in some important cases where it is positive, it is
likely to be small. Whether option values are positive or negative depends on the
nature of the uncertainties confronting users and nonusers. At least five sources of
uncertainty can create option values. These include: (1) supply of the resource,
(2) income, (3) taste, (4) price of the resource and (5) prices of substitutes or
complements. Examples can be constructed where option value is positive, negative,
or zero for reasonable uncertainty assumptions. Further, V. Kerry Smith's analysis
n/e/r/W

-19-
[9] shows the importance of uniqueness and irreplaceability in assigning a positive
option value. Given that the Pigeon River has no unique characteristics (especially
given the abundance of substitutes) and that the level of color is reversible at any
time, his analysis implies that option values are unlikely to be positive.
Regarding the magnitude of the option value in the case of uncertainty
about taste, Freeman [10] points out that the conditions for a large option value
are quite stringent. Option value is unlikely to be large unless "the probability of
demand is low, the expected consumer surplus is large, and the individual is highly
risk averse." [10, p.11] For the Pigeon River, these specific conditions are unlikely
to occur. In particular, rafting on the Pigeon River is quite unlikely to have a
large expected consumer surplus relative to income given the abundant potential
substitutes in the area, both for whitewater recreation and for entertainment in
general.
Finally, the empirical evidence on option value is generally unpersuasive.
While several studies (including the one cited by Bach and Barnett) have concluded
that option values were positive and substantial, these studies have been poorly
done and cannot be used to infer large positive values. There are various reasons
for this, including the fact that the survey questions addressed the wrong issues
and were posed in a confusing manner. As Smith [11] points out:
Empirical efforts to measure option price (and option value) have not
been clear in specifying (or attempting to determine the individual's
perceptions on) the terms of access to the resource and the time
horizon for future use. . . .[T]he terms of access and the individual's
ability to adjust to demand uncertainty . . . have not been clearly
described (or elicited) from the respondents involved in the
surveys. . . .[T]his limitation is especially important when the option
price is used to measure the value of changes in either features of
the resources or in the uncertainty itself. [11, p.8]
As a result, these studies have generated estimates of surplus much too large to be
easily explained on a theoretical basis.
11/e/t'/W

_20_
Further, applying estimates taken from these other studies to the Pigeon
River would be even more tenuous. Even if the wild and scenic rivers of Colorado
may have significant option values, the transfer of those results to the Pigeon River
is unwarranted.
Considering all of these issues, we have-not included any option values in
our estimates. Our best reason for doing so is stated by Professor Richard
Schmalansee [12] of the Massachusetts Institute of Technology, one of the original
contributors to the theory of option value:
Individuals' option prices may exceed or fall short of the expected
value of the contingent surpluses they would derive from a price
change, and it is not obvious how one might judge in a real situation
which was more likely. This suggests that when tastes are the main
source of uncertainty, the expected value of consumer's surpluses
ought to be employed as the best available approximation to the sum
of their option prices . . . . Benefits will be sometimes
underestimated and sometimes overestimated by this procedure, but
there would appear to be no practical way to obtain superior
estimates. 112, p.823]
Existence values represent the utility of knowing of the existence of a
resource for a person who knows he will never use it. We believe that this does
not represent a significant source of benefits on the Pigeon River and have
excluded them from our estimates for two reasons. Fib existence value is linked
to uniqueness of the resource. A resource with features available nowhere else (for
example, the Grand Canyon) may have substantial psychic value to nonusers. We do
not believe that nonusers place a large value on the Pigeon River given the
existence of nearby alternatives.
Second a full specification of existence values would of course include
foregone opportunities for the people of Canton. Those who would place a high
utility on lower color levels in the Pigeon River should also place high disutility on
the social disruption caused by the shutting of the Canton Mill. In scenarios in
I1/evr/ad

-21-
which the mill must be closed, a positive existence value.would mean that people
put less value on the disrupted lives of Canton than for color levels; in the Pigeon
River. Thus, in this case existence value is the sum of two effects: one positive
and one negative. It is not clear a priori which effect would predominate.
Third little is known empirically about existence value. Testing for
existence value is fraught with difficulty, since people not actually called upon to
pay for improvements may systematically overstate, for any number of reasons, their
willingness to pay. The most comprehensive study of existence (and option values)
suggests that together they might increase benefit estimates by as much as
50 percent. Even these estimates have come under attack. We find that an
increase of this size at the Pigeon River to be highly unlikely. However, we note
in passing that even a tripling of all benefits in Table I still results in an
unfavorable benefit-cost ratio for all reduction plans. The benefit cost ratios for
the more stringent USEPA proposals require assumptions of option and existence
values of ten to twenty times the direct effects.
III. COSTS
Colored effluent is the byproduct of the removal of lignin from wood
chips in the pulping process. Currently, the mill is one of the lowest producers of
pounds of color per unit of output in the United States.
A. 50 Percent Reduction
The primary source of color discharge is the caustic extraction stage of
the bleach plant. Improvements proposed by Champion to cut this contribution by
90 percent would achieve an overall reduction of 50 percent. This reduction is
proposed to be accomplished by the addition of oxidation chemicals.
The costs of the 50 percent reduction proposal have been estimated by
Champion at $10 million dollars in capital cost and $4 million per year in operating

-22-
and maintenance costs. We have converted these costs into a ten year stream and
discounted them at Champion's after-tax cost of money. This yields a discounted
present value of expenses over a 10-year period that is directly comparable with the
benefit estimates. The operating and maintenance expenses were discounted using
the 6.79 percent rate. The present value of the costs of this plan is therefore
$38.4 million. The costs are summarized in Table V.
B. 50 Units at the Tennessee Border
Other technologies that would treat more than the caustic extraction
stream require far greater investments. Champion submitted to the USEPA, at the
Agency's request, seven reports detailing methods of reduction sufficient to achieve
no more than 50 color units at the border at any time. Of these, the USEPA
deemed as potentially suitable the lime, polyamine and alum removal systems. Each
technology has a different capital and operating cost, with the alum removal system
being by far the lowest cost. Further, the systems have different risks associated
with them. Champion's engineers have supplemented the estimates of probable cost
of each of these technologies with a probability distribution of costs. The proper
cost to use ex ante in comparing the plans is the expected cost of the technologies,
i.e. where the costs are weighted by the probabilities that they occur. Under this
criterion, the polyamine technology has the lowest cost. The present value of its
costs are $164.5 million, as opposed to $192.5 million for the alum reduction plan
and $209 million for the lime reduction plan. Since the polyamine system is the
least expensive of the three technologies that would achieve 50 color units of the
border at any time, we have used it in the analysis.
We should note that the polyamine removal process is likely to understate
the cost, possibly greatly so. EjEaL no polyamine process on this scale has ever
been implemented before. Large scale experiments with new technologies often have
n/esrral

-23-
substantial unexpected cost increases. Second byproducts of the polyamine process
would almost certainly violate existing environmental standards. In particular,
operation of the sludge dryer would cause air quality deterioration in particulate
matter, hydrocarbons, SOZ, and NO, The cost of meeting existing standards have
not been included in these estimates.
In calculating the results in Table I, we have assumed that the plant
'remains open at this level of expenditure. Champion has submitted to USEPA
workpapers based on the USEPA gross margin test. This test compares the ratio of
net operating revenues to annual pollution control costs. Where this ratio exceeds
the average ratio of investment returns for the industry, "the test results indicate
that pollution controls would impose severe economic impacts, [and] a more detailed
plant closure analysis would be necessary." [14, page 92] Champion's submissions
indicate that the 50 unit at the border strategy fails to pass this test. This raises
the possibility of mill shutdown even under this plan. Were this the case, the quite
substantial resource and employment costs of the more stringent December USEPA
proposal would result. Only the 50 percent reduction plan would have costs within
a factor of 50 times benefits.
We have analyzed this case from a different perspective and find a
persuasive case for shutdown. Our analysis suggests that the internal rate of
return on an investment in pollution control (combined with the $200 million
necessary to keep the plant commercially viable) is considerably less than the
20 percent hurdle rate necessary to justify an investment for Champion. Even were
the investment of average risk for the company, the project would likely not be
justified. In fact, the risks involved in both the pollution cleanup and oxygen
delignification production process are substantial, requiring internal rates of return
in excess of the hurdle rate to warrant the investment. We therefore regard the
ri/evr/ i°

_24_
50 unit at the border proposal as likely to lead to shutdown of the Canton mill.
Again, were this the case, the true costs and benefits of the 50 units at the border
plan would be identical to those from the more stringent limit of 50 units at the
pipe.
C. 50 Units at the Mill
Engineering judgment has determined that these color treatment schemes
cannot simply be scaled up to achieve a discharge level of less than 34,000 pounds
per day on the average, the amount sufficient to assure less than 50 units at
Canton. Also, the mill would not be able to operate on pulp shipped in. Thus, the
only method of achieving 50 units at the mill is the shutdown of the plant. In this
case, the real resource cost of shutdown is the cost of constructing a new mill with
replacement capacity elsewhere in the country. Champion estimates that the
creation of a new mill capable of producing 600,000 tons of pulp per year would
have capital costs of $1.4 billion. To properly characterize the total resource costs,
we should subtract from this estimate the present value of savings in operating and
maintenance provided by a new plant.
Champion has proposed a $200 million investment in an oxygen
delignification process which would allow the plant to continue operations. As a
proxy for the operating and maintenance cost savings, we have assumed that the
$1.4 billion investment creates a profit stream which yields an internal rate of
return of 13.2 percent, i.e., represents an investment of average risk for Champion.
This profit stream is then subtracted from the profit stream arising from the
$200 million switch to oxygen delignification at the Canton mill. The present value
of changed profits must represent the present value of cost savings associated with
the new technology. The present value of resource losses thus totals about
$290 million. The additional impacts of a shutdown are discussed below.
Ili/r/w

-25-
IV. COSTS AND BENEFITS
The economic criterion for undertaking any investment, including
environmental enhancement programs, is that the benefits exceed the costs. Even
where environmental benefits are themselves large, undertaking a project whose
costs exceed its benefits is a waste of society's resources. Further, the difference
between incremental costs and benefits should guide the choice between projects.
While none of the three programs pass the benefit-cost test, the
50 percent reduction plan comes the closest by a wide margin. Under the
50 percent reduction plant the present value of the costs over a 10-year period are
$38.4 million. These are $27.1 million greater than the benefits of $11.3 million.
These costs and benefits translate into a benefit-cost ratio of 0.30 (11.3 divided by
38.4). The interpretation of this benefit-cost ratio is that a one dollar of
expenditure yields 30 cents in benefits. The comparison of benefits and costs
benefits are summarized in Table I. This is substantially less than one, and
consequently would not pass the economic criterion for undertaking the investment.
The plan to reduce discharges to 50 units at the border is even less
justifiable. Its costs are $164.5 million, and they exceed the benefits of
$18.0 million by $146.5 million. This results in a benefit-cost ratio of 0.11--nearly
half of what could be achieved under the 50 percent reduction plan.
The most costly program--50 units at the mill--achieves the poorest
benefit-cost ratio. The costs would be much greater than either of the other plans
because it would involve shutting down the Champion mill in Canton and
constructing a new one. Under this program the costs would be $290 million.
These are $271.7 million more than the benefits and would yield a benefit-cost ratio
of .06.
ri/2/r'rw

-26-
A very useful way to look at the alternative programs is by comparing
the increase in costs and benefits that occurs from imposing increasingly stringent
discharge removal plans. This is a comparison of the incremental costs and benefits
and describes the additional benefits and costs that accrue from removing additional
color. If the additional costs exceed the additional benefits, then the incremental
reduction in color is uneconomic.
As we have noted above, the least stringent of the programs--the
50 percent reduction plan--increases costs by $38.4 million and benefits by
$11.3 million for a benefit-cost ratio of 0.30. The incremental cost of removing
additional color under the 50 units at the border plan is $126.1 million--the
difference between $164.5 million and $38.4 million. This added expenditure removes
an additional 102,000 pounds of color per day (274,000 minus 172,000). However,
this additional color reduction only increases benefits by $6.7 million (18.0 minus
11.3). Thus, the incremental benefit-cost ratio is 0.05 (6.7 divided by 126.1).
The incremental comparison is even more unfavorable under the 50 units
at the pipe plan as proposed by the USEPA. This program would increase costs by
$125.5 million--from $164.5 million to $290.0 million. The added expenditures would
achieve an additional reduction of color discharge of only 37,000 pounds per day.
However, this additional reduction would only increase benefits by $300,000. As a
result, the incremental benefit-cost ratio would be virtually zero.
These incremental results depend crucially on the assumption of 50 color
units as an objective perceptibility standard. If, in fact, there is no perceptible
difference between 50 and 100 color units, then the conclusions are even more
clear. Since the 50 percent reduction plan essentially achieves 100 color units or
less at the border over 95 percent of the time, this plan would achieve all benefits,

-Z7-
no matter how large. Each of the more stringent plans, therefore, has incremental
benefits of zero and cannot be justified on any rational criterion.
In estimating the benefits and costs we have discounted the annual
impacts at Champion's after-tax cost of money. Some have argued that the proper
discount rate to use for the evaluation of projects should be a riskless discount
rate. The basis for this argument is that since the benefits are spread over the
population at large, the risk premium associated with private investment should not
be included. We disagree with this argument. The reason is that funds spent on
this project will deny Champion the ability to undertake other projects capable of
earning the private rate of return. Hence the rate we have used is appropriate.
We would agree, however, were this proposal financed with new money, but that
would not be the case.
Nonetheless, we have calculated the benefits and costs using a real
discount rate of three percent, which is intended to represent a riskless rate. This
substitution does not alter the results, and therefore our conclusions remain
unchanged. Using a lower discount rate raises both the costs and benefits. For
each case, the benefit-cost ratios are slightly less favorable. Table VI compares the
results at the alternative discount rates.
V. EMPLOYMENT CONSEQUENCES
We have also estimated employment impacts from the three color
discharge reduction strategies. Employment impacts are treated separately from the
benefit-cost impacts because they represent different issues. Jobs created cannot be
examined in isolation. For example, the cost side of the benefit-cost calculations
reflects the employment impacts of the discharge reduction plans. Higher cost
levels always create more jobs, whether or not the costs were incurred efficiently.
Djevr/w

-28-
Thus, the employment consequences of a particular strategy cannot be added to
benefits. Instead, they represent an alternative perspective.
The 50 percent reduction strategy would likely have modest employment
impacts of about 250 jobs. The 50 units at the border strategy, if implemented,
would yield about 600 jobs. Given the probability that the plant would shut down,
however, both this plan and the 50 units at the pipe strategy would impose very
substantial employment losses totalling over 12,000 jobs.
Jobs created and lost can be broken down into direct and indirect
components: the direct component consists of the jobs required either to provide
recreation or to maintain color reductions. Indirect employment gains and losses
relate to the employment in other sectors required to support the direct work force.
The U.S. Department of Commerce has issued state- and industry-specific estimates
of total direct and indirect job impacts for given changes in output in various
industries [13]. For recreation in Tennessee, we have used the hotels, lodging
places and amusements multiplier. It indicates that for an increase of $1 million of
expenditure in this industry group, direct and indirect employment will increase by
60.2 jobs. For discharge reduction activities, we have used the North Carolina
values for the paper and allied products classification and the Tennessee value for
forestry and fishing products. A $1 million expenditure in North Carolina provides
30.1 jobs, while a $1 million expenditure in Tennessee provides 54.5 jobs.
The direct employment gains in Tennessee follow from the staffing
requirements of the whitewater rafting facilities. Fishing, kayaking and sightseeing
generate no direct employment gains, since they have no direct costs other than the
costs of transportation. Using Ocoee River figures which show commercial rafting
to account for about 85 percent of whitewater activity and using the Bach and
Barnett figures of $20 per trip, we derive maximum expenditures in Tennessee for
Tl/e/r/am

_29_
whitewater rafting of $2 million (20x.85x120,000). This assumes the 120,000 visits
per year associated with complete color removal. The multiplier of 60.2 direct and
indirect jobs per $1 million in expenditures implies an increase of 123 jobs.
The employment impacts are summarized in Table VII. Under either the
50 units at the border plan or the 50 units at the pipe plan, the concentration of
color will be at the minimum levels of acceptability. Consequently, under either of
these plans there will be the full employment impact of 123 jobs in Tennessee.
The 50 percent reduction plan would have less of an employment impact
because the benefits in Tennessee are less. Recall that we assumed that the
benefits are proportional to the color reduction. Since with the 50 percent
reduction plan the recreational benefits are 63 percent of those under complete
color removal, the employment impacts also would be 63 percent of 123 or 77 jobs.
The plan to provide 50 units at the pipe suggested by USEPA causes mill
shutdown. This shutdown reduces Tennessee commercial output by $15 million as
cited by the Canton Mill Operations Manager, Oliver Blackwell. This in turn causes
a loss of 816 jobs, for a net loss of 693 jobs.
For North Carolina the employment impacts depend on the annual capital
and operating costs of a discharge reduction plan. The operating costs plus an
annualized capital charge represent additional employment at the mill. Under the
50 units at the border plan, the annual capital and operating costs would be
$17.91 million. This translates into 539 increased jobs (direct plus indirect) given
the 30.1 multiplier. Under the 50 percent reduction plan the annual costs would be
$5.85 million which would result in 176 more jobs.
Shutdown of the plant under the USEPA standard would have substantial
employment effects in North Carolina. The mill currently employs over
2,100 workers and produces $400 million of paper products per year. Using the
11/G/1/W

-30-
Department of Commerce multiplier of 30.1 jobs per million dollars of output for
paper and allied products in North Carolina, a total loss of over 12,000 jobs can be
projected. The direct losses to the community include the $160, million currently
paid to local factors.
Thus, in North Carolina and Tennessee combined, the USEPA standard
would lead to a loss of over 12,000 jobs as well as widespread social costs from the
withdrawal of Champion's substantial support both directly and indirectly through its
employees and suppliers and through Champion's support of the Western North
Carolina-Eastern Tennessee infrastructure through tax payments.
VI. EVALUATION OF THE BACH AND BARNETT REPORT
We have reviewed and evaluated the Bach and Barnett analysis and found
it to be seriously flawed. There are four key conceptual errors that undermine
their conclusions that it would be economic to substantially reduce color discharges
from the mill. First they have failed to take into account the fact that benefits
occurring in different years are not of equal value. Second they have included a
category of benefits (indirect benefits) which have no place in a properly designed
benefit-cost study and which have been misestimated in any case. Third, their
examination of costs was entirely hypothetical and bears no relationship to the
benefits achieved. Fourth they have employed the wrong standard in evaluating the
possibility of mill shutdown. The appropriate way to evaluate discharge reduction
strategies is on the basis of their impacts on the Canton mill. Bach and Barnett
evaluate the strategy on a Champion-wide basis. We will address each of these
issues in the following sections.
A. Discounting
The primary distinction in direct benefits between our analysis and that
of Bach and Barnett involves discounting. Bach and Barnett simply sum the year-
nj ✓I'/w

-31-
by-year benefits to yield a total. This is obviously incorrect. A dollar of benefits
today is worth more than a dollar of benefits tomorrow, just as a dollar of income
today is worth more than a dollar of income tomorrow. Discount factors which
reflect the relative utility of current and future consumption must be applied to
accurately measure the value of future benefits.
Indeed, once appropriate discount rates are applied, the Bach and Barnett
estimates of recreational benefits are slightly lower than ours. Application of our
6.79 percent discount rate lowers the Bach and Barnett direct benefit estimates from
$18.3 million to $13.1 million. Our corresponding estimate of direct recreational
benefits is $13.72 million (see Table II).
B. Indirect Benefits
Bach and Barnett add to the recreational benefits indirect benefits
reflecting increased expenditures in the community resulting from the recreational
activity. We have strong objections to this technique. Most importantly, this type
of benefit has no place in a properly done benefit-cost study. Secondly, even were
these benefits to be included, they have overstated both the level of expenditures
to which these indirect multipliers should apply and have overstated the multiplier
to be used.
The proper standard for a benefit-cost study is the real resource costs
and the real benefits. In comparing the real resource costs and benefits of the
specific proposals mentioned here, we have a criterion which is independent of the
indirect benefits. All programs have indirect benefits. Removing those indirect
benefits can therefore have no effect 'on the analysis. Since the money spent on
this project will be spent in some other way, the indirect benefits are essentially
independent of the program undertaken.

-32-
Indirect economic multipliers represent a flawed concept taken in
aggregate. The problem is that money spent on whitewater rafting does not reflect
new income but is instead the shifting of income from another purpose. A dollar
spent on whitewater rafting is a dollar not spent on other outdoor activities or
other recreation. The economic multiplier of the dollar spent on whitewater rafting
is offset by the loss of the same multiplier from the foregone activities. In
aggregate therefore, these indirect benefits should be ignored. Each dollar of
indirect benefit represents an indirect dollar of loss to someone else.
If one is interested in the net economic benefits for a smaller region,
this can be calculated. However, we must be careful to subtract out expenditures
which cancel out within the region. To the extent that persons from outside Cocke
County are induced to spend money in Cocke County, those dollars will create
indirect benefits in Cocke County. Only expenditures by those outside Cocke
County create these spillover effects within Cocke County.
However, the number of such visitors is likely to be small. Unpublished
data on a nearby river indicate that 41 percent of all river users travelled less than
one hour to the river. Using these data, we have defined the region of direct
economic interest is that within one hour of the Pigeon river. Thus, multipliers
would apply to only 59 percent of all expenditures.
Using the average expenditure figure cited by Bach and Barnett of $20
per day, we find that a total of $8.1 million (in present value terms) should have a
multiplier attached.
This value likely overstates the benefits. For those travelling in excess of
three hours, it is highly unlikely that they would take a trip as short as those on
the Pigeon River. The abundance of alternative fishing, floating and sightseeing
areas and the comparatively low value of these activities make it unlikely that any
IL�P-✓Ifd,®

-33-
long distance visitors would use the Pigeon River for these purposes. Either the
Chattooga or French Broad on which longer trips are available should be preferred
by visitors travelling long distances, especially given the uncertainty of water
supply on the Pigeon River.
Bach and Barnett use a multiplier of three to calculate the indirect
benefits. This estimate reflects the spillover effects for the whole country, not just
the region. The correct value is much smaller. For example, assume an individual
spends $500 on a raft. If the raft was not manufactured in Tennessee, the spillover
benefits will be created wherever payment for the raft goes. By the theory
outlined above, inclusion of these benefits as a spillover is illegitimate, since these
spillovers are exactly offset elsewhere. The U.S. Department of Commerce estimates
state-by-state multipliers for hotels, lodging and amusements. For Tennessee they
have calculated the direct and indirect benefits to the region as 2.1 times the
dollars spent. Indirect benefits therefore are 1.1 times direct dollars expended.
This is the 2.1 less the 1.0 for the direct effect.
Thus, instead of using an estimate of indirect benefits equal to three
times the direct benefits, a more realistic appraisal would be indirect benefits equal
to at most 65 percent of the direct costs, which is 1.1 times 59 percent of the
costs. Total indirect benefits would therefore total about $8.9 million.
Further, these indirect benefits have offsetting indirect costs in the mill
shutdown case. Given the net output losses of $400 million based on expected 1988
mill sales and using the Department of Commerce multiplier of 2.2, we project
indirect losses of $480 million for the most stringent plan, resulting in combined
indirect losses of over $470 million for both states.
Given the importance of the mill to the local economy as the largest
employer, it is possible that these values substantially understate losses associated
I1�L✓ZYd,�

-34-
with termination of pulping facilities. For instance, the mill pays almost
$1.7 million in taxes per year to counties and municipalities over and above taxes
paid by its employees. The impacts of these local tax losses on social services and
education are not calculated here.
C. The Costs of Color Reduction
Bach and Barnett assume a capital cost of $60 million and an operating
and maintenance cost of $6 million. The present value of this expenditure,
$102.5 million, is insufficient even to achieve 50 units at the border. Further, we
note that even by their own analysis, the costs of cleanup far exceed the benefits.
D. Financial Impact on Champion International Corporation
In evaluating the impact of these expenditures on Champion rather than
on the financial viability of the mill, Bach and Barnett have used the wrong
standard. No matter how small the impact on Champion's overall balance sheet, the
economics of the Canton mill must stand on its own merits. An incremental
investment which fails to achieve a suitable rate of return is not justified.
Champion has submitted to USEPA documentation supporting the fact that
the 50 units at the border plan fails to pass the plant gross margin test. Under
this standard, plant economics do not support even this less stringent plan.
Further, our analysis indicates that the returns to this investment are insufficient
to warrant continued operation of the plant.
Tl/,/Y9d®

-35-
REFERENCES
(1] Bach, Orville E. and Barnett, William H., "An Economic Impact Analysis on
the Recreational Benefits of a Restored Pigeon River and A Financial Analysis
of Champion International Corporation's Ability to Provide for a Clean Pigeon
River", unpublished mimeo., May 9, .
[2] Epp, Donald J. and Al-Ani, K.S., "The Effect of Water Quality on Rural
Nonfarm Residential Property Values," American Journal of Agricultural
Economics 61 August 1979.
[3] Brashares, Edith Nevins, "Estimating the Instream Value of Lake Water Quality
in Southeast Michigan," Ph.D. Dissertation, University of Michigan, 1985.
[4] Feenberg, Daniel and Mills, Edwin S. Measuring the Benefits of Water
Pollution Abatement. New York: Academic Press, 1980.
[5] Data supplied in correspondence with Eastern Professional River Outfitters
and confirmed in a conversation with Bob Allen, Tennessee Department of
Conservation.
[6] Walsh, Richard G., Ericson, Ray K., Arosteguy, Daniel J., and Hansen, Michael
P., "An Empirical Application of a Model for Estimating the Recreational
Value of Instream Flow," Completion report OWRT Project No. A-036-COLD.
(Colorado Water Resources Research Institute: Fort Collins Colorado, October,
1980).
[7] Telephone Conversation with Bob Allen, Tennessee Department of
Conservation.
[8] Dornbusch, David M. and Barrager, Stephen M. Benefits of Water Pollution
Control on Prooerty Values. Prepared for the Office of Research and
Monitoring, U.S. Environmental Protection Agency, Socioeconomic
w ✓r/ao

-36-
Environmental Studies Series. EPA-600/5-73-005. (Washington: U.S.
Environmental Protection Agency, October 1973).
(9] Smith, V. Kerry, "Option Value: A Conceptual Overview," Southern Economic
Journal Volume 49, Number 3 (January, 1983).
(10] Freeman, A. Myrick, "The Sign and Size of Option Value," Land Economics
Volume 60, Number 1 (February, 1984).
[11] Smith, V. Kerry, "Nonuse Benefits in Benefit Cost Analysis," Southern
Economic Journal. Volume 54, Number 1 (July ).
[12] Schmalansee, Richard, "Option Demand and : Valuing Price Changes Under
Uncertainty," American Economic Review Volume 65, September 1975.
[13] Bureau of Economic Analysis, U.S. Department of Commerce, Regional
Multipliers: A User Handbook for the Regional Input-Output Modeling System
(RIMS II). (Washington: May 1986).
[14] Office of Analysis and Evaluation, U.S. Environmental Protection Agency,
"Workbook for Estimating the Economic Effects of Pollution Control Costs,"
(Washington: USEPA, November 1983).
rl/ev at

FIGURES
I1�P/Tfd®

FIGURE 1
DIAGRAM OF THE PIGEON RIVER
BETWEEN CANTON, NC AND NEWPORTs TN
FRENCH BROAD RIVER
•u '
MKINFORT,TN
COfaT CREEK a FIILLOTI CREEK
Tennessee
FL RM°aer-J North Carollna
ov RM 26.06.0
T
CREEK
WALTERS DA
WALTERS LAKE
CREEK
am HMO BMDGE
ROW
RREE
ramu
CIaDCE
JONATHAN Cam
FORrau
0
FaeaLANo CREEK LL OUTF
RM 63.1
CLTOE
RORFARM O
ADDmow
T

FIGURE 2
k
C;O! with the P. R.-O.S
P.O. BOX 592
GATLINBURG,
f" P1GEON TENNESSEE
° 37738
° 615,RIVEK '
i436.5008
•`
OlJ. T®OORS
• ..
S O TO
•RRA• HO
32
C 4 + PIGEON FORGE FOOTHILLS
pARXWAY
Sr --- HARTFORD
+ e COSBY
TRAFFIC N
p :r";kfG1Tq 1\
G :f - 1n(7
J ••e3 �2:' ..i�':J' �LAFITRi
a '``•f'%'�'' GREAT'SMOKY MOUNTAIN
With .;.;cNAT1ONAL PARK =:
e s.•_� ._, s.: .v: 0 CTOBER
:.� �..�tz-., �:,._ K•._ ,r._�_.. w . ., gar '`..aie •�..,APRtL-
,,.. +. ��::•.�:.�;ti:j ;t'RFSERVATIONS •�. . ....�,.u::
T•r• 'l. `y - es[ ou hone for reservations(even •:
• �: '"To U134A:I trip with us we suR Y p
:^.: "-'. •.'`.K.onl an'hour or-two aMadll• Refunds will be given if we are
P6GE0�1• _ V,: _+;,•; Y m riur•conditlons,.or if given JO or more days
Ri�E R . ...ryaMr m raft err.
.,•�`C-..:.,.:.;/_:,v�:..__�•.: T: ..Rain ed eaneailatton..
1• CLi'vWting,WNITEWATER RAFTING TRIPS am on a scenic
• 'S '•dF;•�1, ! d•`SihUr stretch of that BIG PIGEON RIVER. Trips Isere TWICE
0��� ®C • '�?_ '.DAILY from our ougoa on h E. -in ere You
Me foot of the
�ORJ ',.•: V. •r+•; sRoanWro,we.Y+uttis you m tl+s paten there You receive inatfes.
..:.y.:,.. . -i.,..,�v.•. ;_.:•; :••3 •lion in whitowuer technique and ufety procedures.With a Drofes•
w _.;;y-_ .',."`•=;-="-�-��:' .'':.._- ...u ++'�:• Clonal guide in each raft and a kayak safety boat to lead the way.we
LA through TWELVE CLASS 35 AND TWO CLASS 4'SI
. — • Sp SH E BIG PIGEON RIVER has beautiful scenery,lots of splashes.�. .. _LOCATED �N A. .and LS LOTSOF,FUN!
• 1....� Ouc trftis Ire fun for• dI•- BEGINNERS To EXPERIENCED
4,'` PADDLERS-FUN FORTHE WHOLE FAMILY!We emDhasiae...
• �i.`..' ca•MO.EXPERIENCEISNECESSARYI •ear��alm®d,and
. +'2"'•1=•' a ali safety f
t `:PIGEON RIVER OUTDOORS omvm
•:••.;-"'Bea'Iackam;Pwasr..waar Swathing m est wet in. and TENNIS
Li`C+•.:SWE51 For.iooler day wrprovidS watarpeoef lacksts.
' . •`••" �'*p&iD:SERVICE'b iialiable.with a PACKAGE DEAL�toorneor'
Pit '.:•
., bdr thr•rapids bY_We also sell hand-designed T-shire with our BIG
• IGEON RIVER RAFTING ROUTE.
3y2 HOUR TRIPS _ _
1'I�RTHER INFORMATION
'WI of ... . for fiirthsr information on our one-daY rafting MPS,
••� •• I wsrrnlghf trios, kayak and canoe lesions,group rates,v+Aat m wear,
• accommodations,and age!knits. -....

TABLES
n/e✓r/av

TABLE I
COSTS AND BENEFITS OF PLANS TO REDUCE
COIAR DISCHARGES INTO THE PIGEON RIVER
Total Benefits and Costs Marginal Benefits and Costs
Discharge Benefit/Cost Benefit/Cost
Reduction Plan Benefitsl Costsl Ratio Benefits2 Costs2 Ratio
-(Millions of -(Millions of
1988 Dollars)- 1988 Dollars)-
(1)/(2)- (4)/(5)
(1) (2) (3) (4) (5) (6)
50 Percent Reduction $11.3 $ 38.4 0.30 $11.3 $38.4 0.30
50 Units At Border 18.0 164.5 0.11 6.7 126.1 0.05
50 Units At Pipe 18.3 290.0 0.06 0.3 125.5 0.00
Sources and Notes
1 Total benefits are taken from Table IV. Total costs are taken from
Table V.
2 Marginal benefits and costs are the difference between the figures of a
plan and the plan immediately above it.
T1h�./T'icl.'

TABLE II
Page 1 of 2
ESTIMATES OF THE MAXIM BENEFITS FROM
A PIGEON RIVER WITH NO COLOR DISCHARGES
Visits Consumer Surplus
White- WhiteJ.
-
Year water Floating FishinR3 Other3 water Floating FishinE3 Other3
--------------(Thousands)----------- --------(Thousands of Dollars)
(1) (2) (3) (4) (5) (6) (7) -(8) -
1988 24 10 13 15 $ 312 $102 $117 $ 68
1989 48 21 14 16 624 203 126 73
1990 72 31 15 18 936 305 136 79
1991 96 42 16 19 1,248 406 147 85
1992 120 52 17 21 1,560 508 159 92
1993 120 52 19 22 1,560 508 171 99
1994 120 52 20 24 1,560 508 185 107
1995 120 52 22 26 1,560 508 200 116
1996 120 52 23 28 1,560 508 216 125
1997 120 52 25 30 1,560 508 213 135
Present Value4 $8,886 $2,892 $1,226 $711
Total Recreation Value
(Millions of 1988 Dollars) $13.72
Property Enhancements5
(Millions of 1988 Dollars)
Tennessee 2.05
North Carolina 2.52
Total Benefits
(Millions of 1988 Dollars) $18.29
n/eirr

TABLE II
Page 2 of 2
ESTIMATES OF THE MAXIMUM BENEFITS FROM
A PIGEON RIVER WITH NO COLOR DISCHARGES
Sources and Notes
1 Whitewater visits are set at a capacity of 120,000. This is in line
with 1987 visits to the Ocoee River of 124,449 visits supplied in
telephone conversation with Bob Allen at the Tennessee Department of
Conservation. Value of $13 per visit derived in text.
2 Floating visits are set at a capacity of 52,000. This was set from
current usage on the Hiwassee (80,000) reduced by 35 percent to account
for low Pigeon River water levels in August and September. Value of
3 $9.76 derived in text.
Fishing and Other taken from Bach and Barnett. Fishing value derived in
text.
4 Present discounted values over 10 years at a real rate of 6.79 percent,
derived by taking Champion's 13.2 percent after-tax cost of money and
netting out an assumed 6 percent inflation rate.
5 Property value adjustment taken from Table III.
Il�/Ifd'

TABLE III
ESTIMATES OF THE MA=MDM IMPROVEMENTS IN
PROPERTY VALUE FROM A PIGEON RIVER
WITH NO COLOR DISCHARGES
(1988 DOLLARS)
Increase in
Value per Foot Total
Miles of Valuel from Reduced Increase-
Frontage Per Foot Dischar eg s� in Value
-------(D'ollars/Foot)------- --(Thousand
Dollars)-
(1) (2) (3) (4)
North Carolina:
Residential 4.33 $52.49 $14.64 $ 335
Other Private 28.32 18.78 14.64 , 2,189
Public Land3 38.614 NA 0 0
Total $2,524
Tennessee:
Residential 5.53 52.49 14.64 427
Other Private 20.96 18.78 14.64 1,620
Public Lands 14.66 NA 0 0
Total $2,047
Total of Tennessee
and North Carolina $4,100
Sources and Notes
1 Assumes same value per river foot in North Carolina as in Tennessee.
2 For residential land this value represents an increase of 27.9 percent as
estimated by Epp and Al-Ani, reference [2] . For all other private land,
the same absolute value per foot was applied.
3 North Carolina Public includes: Interstate 40, Highways 19 and 23,
forest, mapping recon, unassigned territory, and Champion property.
4
Includes 9.34 miles to estimate the distance on missing maps. The
calculation was done by noting that there are 12.5 miles /from the dam to
the powerhouse and subtracting from this the distances on maps we have
between the dam and the powerhouse.
5 Tennessee public includes: Interstate 40, Edwina Rd. , forest, Cocke
County, unassigned territory, and publicly owned property.
I1/e/I7cl7

TABLE IV
ESTIMATES OF THE BENEFITS OF PLANS TO REDUCE
COLOR DISCHARGES IN THE PIGEON RIDER
50 Percent 50 Unit Maximum
Reduction At Border At Pipe
(1) (2) (3)
Pounds Per Day Removed 172,000 274,000 311,000
Proportion of Benefitsl
Tennessee 0.63 1.00 1.00
North Carolina 0.55 0.88 1.00
Total Benefits $11.33 $17.99 $18.29
Benefits Per Pound of
Color Per Day Removed2 66 66 59
Marginal Benefit
Per Pound Per Day3 66 66 8
Sources and Notes
1 Proportion of benefits is the proportion of total state benefits which
would be observed under each plan. Tennessee benefits include all
recreation benefits in Table II.
2 Total benefits divided by pounds per day removed.
3 Difference in benefits divided by difference in pounds per day removed
from less stringent to more stringent removal.
T1�✓IYd.'

TABLE V
ESTIMATES OF THE COSTS OF PLANS TO REDUCE
COLOR DISCHARGES IN THE PIGEON RIVER
50 Percent 50 Unit Maximum
Reduction At Border At Pipe
(1) (2) (3)
Pounds Per Day Removedl 172,000 274,000 311,000
Capital Costl
(1988 Dollars) $10,000,000 $95,120,000 $290,000,000
Operating and
Maintenance Costl
(1988 Dollars Per Year) 4,000,000 9,783,000 0
Present Value of
Total Cost2
(1988 Dollars) 38,366,000 164,496,000 $290,000,000
Cost Per Pound Per Day
of Color Removed3
(1988 Dollars) 223 601 933
Marginal Cost Per Pound
Per Day of Color Removed4 223 1,244 3,392
Sources and Notes
1 Pounds removed, capital and operating costs provided by the Company. For
50 units at the border plan, capital costs have been adjusted to reflect
probability distribution of capital cost equations.
2 Total cost taken by adding capital cost to operating and maintenance cost
stream discounted at 6.79 percent real discount rate for 10 years. Real
discount rate assumes 6 percent inflation applied to Champion after-tax
cost of money of 13.2 percent.
3 Total cost divided by pounds per day removed.
4 Difference in cost divided by difference in pounds per day removed from
less stringent to more stringent removal.
Ile/T9a.'

TABLE VI
Y
COMPARISON OF COST AND BENEFITS OF
PLANS TO REDUCE COLOR DISCHARGES
IN THE PIGEON RIVER AT ALTERNATIVE DISCOUNT RATESI
6.79% Discount Rate 38 Discount Rate
Discharge Benefit/Cost Benefit/Cost
Reduction Plan Benefits Costs Ratio Benefits Costs Ratio
-(Millions of -(Millions of
1988 Dollars)- 1988 Dollars)-
(1)/(2) (4)/(5)
(1) (2) (3) (4) (5) (6)
50 Percent Reduction $11.3 $ 38.4 0.30 $13.0 $ 45.9 0.28
50 Units at Border 18.0 164.5 0.11 20.7 196.0 0.11
50 Units at Pipe 18.3 290.0 0.06 21.0 343.0 0.06
Sources and Notes
1 All calculations are as in notes to Table I.
TLp✓IPd,•

TABLE VII
ESTIMATES OF EMPIDI!MENP IMPACTS OF PLANS
TO REDUCE COIDR DISCHARGES IN THE PIGEON RIVER.
Changes in Output Changes in the Number of Jobs
Discharge North North
Reduction Plan Tennessee' Carolina2 Tennessee3 Carolina3 Total
(Millions of 1988 $) -------(Number of Jobs)--------
(3)&(4)
(1) (2) (3) (4) (5)
50 Percent Reduction 1.29 5.86 77 176 253
50 Units At Border 2.04 17.94 123 539 662
50 Units At Pie Recreation 2.04
p Forest 15.00] -385.00 -693 -12,040 12,733
Sources and Notes
' Changes in output for Tennessee are derived by multiplying increased visits per
year (120,000) , the proportion of expenditures that are commercial (.85) , the cost
per trip (20) , and the proportion of benefits achieved (from Table IV) . The loss
of $15 million of output in the last case is taken from Oliver Blackwell's
testimony before USEPA, page 27.
2 Changes in output for North Carolina are calculated as follows: for the first two
plans they are derived by adding the operating and maintenance cost of a plant and
the capital cost multiplied by the capital recovery factor (.186) . For the 50
units at the mill plan, the loss in output is equivalent to the current output of
the mill.
3 Multipliers are derived from RIMS manual, reference [131 .
Ilip./IYd'

CHRONOLOGY
CHAMPION INTERNATIONAL
CANTON, NORTH CAROLINA
June 25, 1984 EMC Public Meeting for Champion' s
Temperature Variance
October, 1984 Temperature variance issued by EMC not to
exceed temperature increase at Fiberville
Bridge ( . 4 miles downstream of discharge)
of 13 . 9 degrees Centigrade . Maximum
temperature of 29 degrees Centigrade
October thru June and 32 degrees Centigrade
July thru September
May 14 , 1985 North Carolina issued NPDES Permit Number
NC0000272 to Champion International, Canton
Mill .
November 13 , 1985 - EPA assumes control of Champion' s NPDES
Permit after determining that it (EPA)
did not have sufficient time (90 days) to
review the permit prior to N. C. issuance
and in response to color concerns of
downstream residents in Tennessee.
January 17, 1986 - Champion challenged EPA' s claim of
authority over the permit process in US
District court .
March 24 , 1986 - North Carolina filed court action to
block EPA action.
March 26, 1986 - Champion asks for temporary restraining
order to block EPA action until
existing suit is settled.
March 31, 1986 - US District Court Judge Sentelle denies
Champion request for temporary
restraining order.
April 11, 1986 - Champion requests NPDES Permit from EPA.
May 2, 1986 - Motion by EPA to dismiss Champion
suit denied.

Page 2
December 1, 1986 - US District Court Judge Sentelle ruled
that EPA had jurisdiction to assume
authority over Champion' s NPDES Permit .
March 4, 1987 - Champion appeal of Sentelle ruling.
April 9, 1987 - First EPA Draft (public notice) of
Champion NPDES Permit with 50 unit color
limit applicable to Champion immediately
downstream of discharge at Fiberville
Bridge.
May 22, 1987 - Second EPA public notice for permit
issuance .
January 14, 1988 - First Public Hearing in Asheville Civic
Center for NPDES Permit issuance.
January 21, 1988 - First Public Hearing in Knoxville,
Tennessee Civic Center.
March 9, 1988 - Tennessee, North Carolina, Champion, and
EPA agree on change in proposed permit
from 50 Apparent Color units at NC/TN
state line (proposed as applicable at
the Fiberville Bridge immediately
downstream of the Champion Mill) to 85
True Color Units at the State line.
North Carolina and Tennessee to grant
variance from color stream standard to
allow permit issuance.
May 5, 1988 - Champion formally requests water quality
color standard variance from EMC NPDES
subcommittee at public hearing in Raleigh.
June 24, 1988 - Champion appeal of Sentelle decision
denied by three judge panel of 4th US
Circuit Court of Appeals - verifying EPA
authority over Champion' s NPDES permit .
Ron Levine, Director, Division of
Health Services, NC Dept . of Human
Resources issued advisory against eating
fish from Pigeon River as per studies
showing a presence of dioxin in fish and

Page 3
sediment downstream of the mill; albeit
.not in Champion' s effluent . Dioxin not
regulated to date in the proposed NPDES
Permit .
July 13 , 1988 - EMC issues water quality variance from
North Carolina color standard allowing
NPDES Permit for 85 True Color Units to be
placed in proposed NPDES Permit.
August 18-19, 1988- Tennessee color variance hearing in
Newport, Tennessee .
December 23, 1988 - Tennessee denies Champion' s color variance
request from standard, preventing
agreement reached 3/9/88 from being
formally enacted.
March 15, 1989 - EPA Issued a draft Discharge Permit for a
reconfigured Canton Mill, allowing it to
operate during reconfiguration.
July 12, 1989 - Second draft NPDES Permit (public notice)
for Champion with color limit consistent
with EMC variance and Tennessee
limit of 50 units at State Line . Permit
also including additional requirement for
chloroform and dioxin. Permit provided
for reduction in flow from 48 . 5 MGD
to 29 MGD and three (3) year schedule for
compliance (down from 5 year schedule in
initial draft) .
August 17, 1989 - Newport, Tennessee public hearing on EPA
NPDES Draft Permit .
August 24, 1989 - Asheville public hearing on permit .
September 8, 1989 - DEM issues Section 401 Water Quality
Certification for proposed Champion Permit.
September 25, 1989- Champion NPDES Permit issued by EPA.

Page 4
October 14, 1989 - Permit appealed first by Environmental
groups and then by Champion. EPA
administrative law judge grants hearing.
Hearing to be scheduled.
March 27, 1990 - Champion announces implementation of
modernization project despite pending legal
issues . Estimated cost of project is $250
million.
1990 - 1991 - Legal issues continue to be fought before
EPA Administrative Law Judge . Champion
continues modernization project.
April, 1991 - EPA Administrative Hearing held in Atlanta,
Georgia.
Early 1992 - Administrative Law Judge issues final
decision essentially leaving permit as is .
No appeal requested.