1000 Friends of Wisconsin Inc v. United States Department of Transportation et al

Filing 76

DECISION AND ORDER signed by Judge Lynn Adelman on 4/29/16 that the defendants motion to reinstate the record of decision and enter judgment in their favor 66 is DENIED. Further ordering that defendants motion to strike 75 is DENIED. (cc: all counsel) (dm)

UNITED STATES DISTRICT COURT
EASTERN DISTRICT OF WISCONSIN
1000 FRIENDS OF WISCONSIN, INC.,
Plaintiff,
v.

Case No. 11-C-0545

UNITED STATES DEPARTMENT OF
TRANSPORTATION, et al.,
Defendants.

DECISION AND ORDER
The plaintiff in this case, 1000 Friends of Wisconsin, Inc., brings claims under the
National Environmental Policy Act (“NEPA”) and the Administrative Procedure Act,
challenging a decision of the Federal Highway Administration and the Wisconsin
Department of Transportation (“WisDOT”), to expand to four lanes a 19-mile segment of
Wisconsin State Highway 23. In a prior decision addressing the merits of the plaintiff’s
claims, I concluded that the defendants’ analysis of reasonable alternatives to the 4lane expansion was deficient because the analysis rested on traffic forecasts that the
agencies had failed to fully explain. See May 22, 2015 Dec. and Order, ECF No. 61 I
vacated the record of decision approving the expansion project and remanded the
matter to the agency for further proceedings. On remand, the agencies prepared a
“revised technical memorandum,” in which they further explained how they arrived at
the traffic forecasts that appear in the environmental impact statement for the project.
See ECF No. 70-1. The agencies now move to reinstate the record of decision and for
judgment in their favor. The plaintiff opposes the motion.

1

I. NEPA AND APA STANDARD OF REVIEW
NEPA “declares a broad national commitment to protecting and promoting
environmental quality.” Robertson v. Methow Valley Citizens Council, 490 U.S. 332,
348 (1989). It has been described as a “procedural” or “action-forcing” statute that does
not “mandate particular results” but instead requires agencies to study and describe the
environmental consequences of their proposed actions. Id. at 348–51; Vermont Yankee
Nuclear Power Corp. v. Nat. Res. Defense Council, 435 U.S. 519, 558 (1978). Thus,
under NEPA, if an agency has adequately identified and evaluated the environmental
effects of its proposed action, it is permitted to take that action even if it is expected to
be environmentally costly. Robertson, 490 U.S. at 350. Put differently, “NEPA merely
prohibits uninformed—rather than unwise—agency action.” Id. at 351.
NEPA requires agencies to prepare an environmental impact statement for
“major Federal actions significantly affecting the quality of the human environment.” 42
U.S.C. § 4332(2)(C). This impact statement is “a detailed analysis and study conducted
to determine if, or the extent to which, a particular agency action will impact the
environment.” Highway J Citizens Group v. Minetta, 349 F.3d 938, 953 (7th Cir. 2003).
Requiring an agency to prepare an impact statement serves NEPA’s action-forcing
purpose in two respects.

Robertson, 490 U.S. at 349.

First, “[i]t ensures that the

agency, in reaching its decision, will have available, and will carefully consider, detailed
information concerning significant environmental impacts.” Id. Second, it “guarantees
that the relevant information will be made available to the larger audience that may also
play a role in both the decisionmaking process and the implementation of that decision.”
Id. Thus, in the impact statement, the agency must “articulate why [it has] settled upon a

2

particular plan and what environmental harms (or benefits) [its] choice entails.”
Simmons v. U.S. Army Corps of Eng’rs, 120 F.3d 664, 666 (7th Cir. 1997). The impact
statement must show that agency officials have “[thought] through the consequences
of—and alternatives to—their contemplated acts,” and must ensure that “citizens get a
chance to hear and consider the rationales the officials offer.” Id.
Judicial review of an agency’s compliance with NEPA occurs under the
Administrative Procedure Act, which instructs courts to set aside agency action only if it
is “arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with the
law.” 5 U.S.C. § 706(2)(A); Highway J Citizens Group, 349 F.3d at 952.
II. DISCUSSION
The deficiencies I identified in my prior decision concerned the impact
statement’s discussion of reasonable alternatives. Under NEPA, an impact statement
must discuss alternatives to a proposed action. 42 U.S.C. § 4332(2)(C)(iii). The NEPA
regulations specify that an agency preparing an impact statement must “[r]igorously
explore and objectively evaluate all reasonable alternatives, and for alternatives which
were eliminated from detailed study, briefly discuss the reasons for their having been
eliminated.” 40 C.F.R. § 1502.14(a).
In the present case, the agencies’ evaluation of reasonable alternatives was
based, in part, on traffic projections through the year 2035, which WisDOT prepared in
2012. R. 21414–19. The projections are expressed in terms of daily traffic volumes for
each of eight segments of Highway 23, and there are four projections for each
segment—one for each reasonable alternative—along with the actual traffic counts
measured in 2012:

3

R. 21414. The plaintiff contends that these traffic projections are flawed, in that they
overestimate the amount of traffic that can be expected on Highway 23 through the year
2035.

The plaintiff contends that if accurate traffic projections were used, then

alternatives to expanding the highway to four lanes, such as adding passing lanes to the
existing two-lane highway, would appear more attractive.
In my prior decision, I did not find that the traffic projections were flawed. Rather,
I determined that I could not decide whether the projections were flawed because
WisDOT had not fully explained how it applied its methodology. See Dec. and Order at
9–14.

I also determined that I could not tell whether the projections were flawed

because WisDOT had not explained how recently updated demographic data from the
Wisconsin Department of Administration might affect the projections. Id. at 14–17. The
revised technical memorandum attempts to explain these issues. Below, I consider
whether the technical memorandum cures the deficiencies in the impact statement.

4

A.

Application of Traffic Forecasting Methodology
The methodology that WisDOT purported to use in projecting traffic volumes is

explained in an appendix to the impact statement. See R. 21923–44 (Appendix LS-A).
In this appendix, WisDOT states that traffic forecasts “are prepared and approved
centrally to assure that a consistent methodology is utilized for all forecasts in
Wisconsin.” R. 21938. Under this methodology, WisDOT will use two forecasting tools,
provided that the forecast is for an area in which it is possible to use both tools.
R. 21938–39. Highway 23 is located in an area in which it is possible to use both tools.
The tools are known as the Traffic Analysis Forecasting Information System (“TAFIS”)
and the Northeast Region Travel Demand Model (“TDM”). Id. TAFIS is an automated
procedure and computer program that uses historical traffic count information to project
future traffic through use of regression analysis. R. 21939; see also ECF No. 70-1 at 2.
TDM is a model that attempts to provide a more nuanced understanding of future travel
patterns than TAFIS. It “uses current socio-economic data, roadway networks, trip rates
and other factors to calculate current and future travel patterns.” R. 21939.
TDM and TAFIS will each produce their own, independent traffic projections.
However, in the appendix to the impact statement, WisDOT states that its policy is to
“compare [TDM] growth rates with the TAFIS growth rates using several techniques.”
R. 21940. It then explains that “[w]hen the [TDM] forecast produces a future forecast
year volume greater than 10% over the TAFIS future forecast year volume, a
compromise number is reached.”

R. 21940.

The compromise number “should

generally be at the edge of the accepted 10% range in TAFIS, as long as it is also within
10% of the [TDM] forecast volume.” Id. The methodology states that the reason for

5

reaching a compromise number when TDM produces a forecasted volume that is
greater than 10% over the TAFIS forecasted volume is to protect against the possibility
that the TDM model is poorly calibrated. Id. The methodology states that, to protect
against this possibility, “the decision was made to insure that forecasted volumes in
[TDM] and TAFIS were to be within 10% of each other.” Id. The requirement to reach a
compromise number when the TDM volume exceeds 10% of the TAFIS volume is also
stated in a different part of the appendix. See R. 21928.
In my prior decision, I found that although WisDOT had generally explained its
methodology for projecting traffic volumes in the impact statement, it had not adequately
explained how it applied that methodology. Specifically, I found that WisDOT had not
shown how the raw data it used resulted in the bottom-line numbers that appear in the
impact statement for each of the project alternatives. Dec. and Order at 11. I noted
that, among other problems, WisDOT had failed to identify the specific traffic volumes
that TDM and TAFIS had independently projected for each segment of Highway 23 and
to explain whether WisDOT selected compromise numbers to bring the TDM projections
within 10% of the TAFIS projections. Id. at 11–12.
In its revised technical memorandum, WisDOT has supplied the missing
information. See ECF No. 70-1 at 8–13. However, WisDOT’s explanation reveals that
it did not actually follow the methodology described in the impact statement when
forecasting traffic volumes for Highway 23.

Although WisDOT compared the TDM

projections to the TAFIS projections, ultimately it selected the TDM projection for each
highway segment, even when the TDM projection was greater than 10% of the TAFIS
projection. As the plaintiff demonstrates in an exhibit to its brief, see ECF No. 72-1, the

6

TDM projection exceeded the TAFIS projection by more than 10% at four of the seven
segments for which projections were created.

For these four segments, the TDM

projections exceeded the TAFIS projections by 14.5%, 13.7%, 11.8%, and 18.2%,
respectively, yet no compromise numbers were reached. WisDOT points out that, in
these four cases, the TDM results are “within 10% of the average” of the combined TDM
and TAFIS projections. 1 ECF No. 70-1 at 12. However, nothing in the methodology
described in the impact statement indicates that being “within 10% of the average” is
acceptable.

Rather, the methodology as described in the impact statement clearly

states that it is WisDOT policy to “insure” that forecasted volumes in TDM and TAFIS
are “within 10% of each other.” R. 21940.
In administrative law, where an agency fails to follow its own rules, procedures,
or methodologies, its action is likely to be found arbitrary and capricious. See Motor
Vehicle Mfrs. Ass’n of U.S., Inc. v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 50
(1983) (“It is well-established that an agency’s action must be upheld, if at all, on the
basis articulated by the agency itself.”); Andershock’s Fruitland, Inc. v. U.S. Dep’t of
Agric., 151 F.3d 735, 736 (7th Cir. 1998) (noting that “an agency must follow its own
rules and doctrines until it changes them explicitly”); Natural Res. Def. Council v. U.S.
E.P.A., 735 F.3d 873, 881–85 (9th Cir. 2013) (vacating agency decision because
agency failed to follow its own risk-assessment standards); Lopez v. Fed. Aviation
Admin., 318 F.3d 242, 246 (D.C. Cir. 2003) (noting that Supreme Court has held that
“an agency is bound to the standards by which it professes its action to be judged”). In
1

To determine whether the TDM results were “within 10% of the average,” WisDOT
averaged the TDM and TAFIS output values to determine if the TDM output and the
TAFIS output were between 90% and 110% of the average. See Reply Br. at 4, ECF
No. 73.
7

the present case, WisDOT deviated from the traffic-projection methodology that it
identified in the impact statement when it failed to reach a compromise number for the
segments in which the TDM projection exceeded the TAFIS projection by more than
10%. The explanation that WisDOT offers for using the TDM projections even where
they were not within 10% of the TAFIS projections, i.e., because the projections “were
within 10% of the average,” is not supported by the stated methodology and appears to
be an after-the-fact rationalization. Thus, I cannot find that the traffic projections that
appear in the impact statement were the product of a reasoned application of the
agency’s stated methodology.
In their reply brief, the defendants contend that WisDOT’s failure to follow its
policy of reaching a compromise projection when the TDM projection exceeds the
TAFIS projection by more than 10% was justified because “other reasons” supported
TDM’s reliability. Reply Br. at 5. First, WisDOT points out that TDM projected only a
“modest” growth rate, which was “consistent with anticipated regional growth.”

Id.

However, the compromise projections, which would have been lower than the TDM
projections, would also have been modest and consistent with anticipated regional
growth.

Thus, the fact that the TDM projections were themselves modest and

consistent with anticipated growth does not support WisDOT’s decision to abandon its
stated policy of reaching compromise numbers where TDM exceeds TAFIS by more
than 10%.
WisDOT also points out that, for two of the four segments for which the TDM
projections were not within 10% of the TAFIS projections, the TDM projections
produced growth rates that were at or slightly below 0.5%, which under WisDOT policy

8

is the suggested minimum growth rate. See R. 21927 (stating that “[t]he growth rate
may not be less than 0.5% or greater than 5% unless there is significant change in the
model inputs such as socioeconomic data or the road network”).

For these two

segments, the TDM projections resulted in growth rates of 0.51% and 0.41%. ECF No.
70-1 at 21 (Table 3, column E). However, at the other two segments, the projected
growth rates were 0.99% and 1.85%—well above the suggested minimum. Id. So I fail
to see how the projected growth rates support WisDOT’s decision to abandon its stated
policy of reaching compromise numbers when the TDM projections exceed the TAFIS
projections by more than 10%. 2
Finally, WisDOT points out that it had separately “calibrated and validated the
entire Northeast Region TDM.” Reply Br. at 6. However, nothing in WisDOT’s stated
methodology provides that where a TDM model has been separately validated, reaching
a compromise number where its projections exceed TAFIS projections by more than
10% is unnecessary.

Rather, the methodology described in the impact statement

requires the TDM projection to be “checked against the TAFIS system” and that
compromise numbers be reached regardless of whether the TDM model has been
separately validated. R. 21928, 21940.
Before moving on, I note that WisDOT has not argued that even if it had adopted
compromise projections at the four sites where the TDM projections exceeded the
TAFIS projections by 10%, the analysis of reasonable alternatives would have been the
2

Also, it appears that WisDOT used the TDM projections in the impact statement even
when they resulted in growth rates that were less than the suggested minimum. See
ECF No. 70-1 at 21 (Table 3, column E) (reflecting growth rates of 0.47%, 0.35%, and
0.41%). It is not clear why WisDOT did this, when its stated policy is that the growth
rate may not be less than 0.5% “unless there is significant change in model inputs such
as socioeconomic data or the road network.” R. 21927.
9

same. Nor is it obvious to me that the failure to reach compromise projections was
harmless.

I have reviewed the discussion of reasonable alternatives in the impact

statement, including the detailed discussion of reasonable alternatives that appears in a
technical memorandum attached to the impact statement, see R. 21947–98, and have
been unable to find anything that explains how a change in the traffic forecasts might
affect the viability of the alternatives to a four-lane expansion. However, because it is
clear that the traffic forecasts played an important role in the evaluation of reasonable
alternatives, I cannot conclude that WisDOT’s failure to follow its own methodology and
reach compromise projections was harmless.
For these reasons, I conclude that the traffic projections used in the impact
statement’s evaluation of reasonable alternatives were not produced through a
reasoned application of WisDOT’s stated methodology, and that the agencies’
evaluation of reasonable alternatives was deficient.
B.

Consideration of Updated Population Data
As I noted in my prior decision, when WisDOT built the TDM model that it applied

to Highway 23, it used household population projections prepared by the Wisconsin
Department of Administration to estimate growth in the region through the year 2030.
R. 14179, 14333–35. These projections were based on census data from the year
2000. Shortly before the defendants finalized the impact statement in March 2014, the
Department of Administration released new population projections, through the year
2040, which were based on census data from the year 2010. The agencies included
these population projections in the final impact statement and discussed them in the
context of the indirect and cumulative environmental effects associated with the project.

10

R. 21485. 3 The new projections reflected a lower growth rate than was reflected in the
2030 projections. In my prior decision, I found that nothing in the impact statement
indicated whether the defendants considered revisiting their traffic projections and
evaluation of reasonable alternatives in light of the recently updated population
projections. I noted that population data appeared to be a key input in the TDM model,
and that therefore a significant reduction in expected population growth would likely
produce a significant reduction in expected traffic growth, which, in turn, could affect the
evaluation of reasonable alternatives.

Thus, I concluded that, on remand, the

defendants had to consider whether to supplement the impact statement in light of the
new population data.

See 40 C.F.R. § 1502.9(c)(1)(ii); Marsh

v. Or. Natural Res.

Council, 490 U.S. 360 (1989).
In the revised technical memorandum that WisDOT prepared on remand, it
concludes that the updated Department of Administration population projections “do not
significantly change the final traffic forecasts or impact the alternative selected in the
[final impact statement].”

ECF No. 70-1 at 14.

The reasoning underlying this

conclusion is as follows: (1) TDM does not use general population projections as direct
inputs to produce traffic forecasts. Instead, TDM uses “the type and location of certain
developments,” such as households, employment, and other traffic generators, to
forecast traffic volumes. (2) The updated general population projections do not “indicate

3

The population projections for 2040 that appear in the impact statement are general
population projections, not household projections. Presumably, however, the Wisconsin
Department of Administration released 2040 household projections along with its
general population projections. Both general population projections and household
projections for 2040 are currently available on the Department of Administration’s
website.
See http://www.doa.state.wi.us/forms (click “Population & Household
Projections” under “Demographics”).
11

changes to the location and type of development.”

(3)

Therefore, the population

projections would not affect the traffic forecasts. Id.
This reasoning does not support WisDOT’s conclusion that the updated
population data would not affect the traffic forecasts. First, although it may be true that
the Department of Administration’s general population projections are not direct inputs
into the TDM, the record shows that the Department’s household population projections
are. See ECF No. 70-1 at 14 (stating that Department’s household data is included in
TDM); R. 14179 (stating that Department’s household data was used to create
“Transportation Analysis Zones,” or TAZs, in TDM); R. 14181 (same); R. 14180 (stating
that “households” are variables in TDM database). If the growth rate for the general
population declines significantly, then presumably the growth rate for households will as
well. So a decline in the general population growth rate should at least have caused
WisDOT to check the revised household data to determine if that growth rate had
significantly declined as well.

However, so far as it appears from the technical

memorandum, this was not done.
Second, WisDOT’s observation that the updated population projections did not
affect the anticipated type and location of developments does not, by itself, support the
conclusion that the updated population data would not affect traffic forecasts. Although
I do not doubt that the type and location of expected developments plays a large role in
traffic forecasting, presumably WisDOT will also need to anticipate how much of each
type of development to expect in a given location in order to prepare an accurate
forecast. For example, if WisDOT anticipates that residential housing will be built at
some location along Highway 23, certainly it will also want to know approximately how

12

many households to expect at that location, as the number of households will determine
the number of vehicle trips that will begin and end at that location.

Thus, if the

Department of Administration’s updated projections show a decline in population and
household growth rates, then it is likely that the growth rate for traffic will be affected,
even if the type and location of anticipated development remains the same.
In their reply brief, the defendants point out that the difference between the
Department of Administration’s 2030 and 2040 population projections for the project
area is 5,218 people, which is 6.7% of the projected 2040 project-area population. The
defendants seem to imply that therefore the change in projected growth in the area is
insignificant. However, WisDOT in its revised technical memorandum did not state that
the projected decline in the population growth rate is too small to affect its traffic
forecasts. Therefore, I cannot conclude that the size of the projected decline supports
WisDOT’s decision not to update its forecasts. See State Farm, 463 U.S. at 50 (a court
“may not accept appellate counsel’s post hoc rationalizations for agency action”; the
agency’s action “must be upheld, if at all, on the basis articulated by the agency itself”).
For these reasons, I cannot find that WisDOT, when deciding whether to update
its traffic forecasts in light of the updated population data, “conducted a reasoned
evaluation of the relevant information and reached a decision that, although perhaps
disputable, was not ‘arbitrary or capricious.’” Marsh, 490 U.S. at 385.
III. CONCLUSION
For the reasons stated, IT IS ORDERED that the defendants’ motion to reinstate
the record of decision and enter judgment in their favor is DENIED.
IT IS FURTHER ORDERED that the defendants’ motion to strike is DENIED.

13

Dated at Milwaukee, Wisconsin, this 29th day of April, 2016.

s/ Lynn Adelman
__________________________________
LYNN ADELMAN
District Judge

14