Notice: This opinion is subject to formal revision before publication in the
Federal Reporter or U.S.App.D.C. Reports. Users are requested to notify the
Clerk of any formal errors in order that corrections may be made before the
bound volumes go to press.
United States Court of Appeals
FOR THE DISTRICT OF COLUMBIA CIRCUIT
Argued April 6, 2010 Decided May 14, 2010
No. 09-1011
COALITION OF BATTERY RECYCLERS ASSOCIATION,
PETITIONER
v.
ENVIRONMENTAL PROTECTION AGENCY,
RESPONDENT
Consolidated with No. 09-1012
On Petitions for Review of Final Rule
of the United States Environmental Protection Agency
Michael B. Wigmore argued the cause for petitioner
Coalition of Battery Recyclers Association in No. 09-1011.
Dennis Lane argued the cause for petitioner The Doe Run
Resources Corporation in No. 09-1012. With them on the briefs
were Robert N. Steinwurtzel and Sandra P. Franco.
2
Eric G. Hostetler, Attorney, U.S. Department of Justice,
argued the cause and filed the brief for respondent.
Before: SENTELLE, Chief Judge, ROGERS and GARLAND,
Circuit Judges.
Opinion for the Court by Circuit Judge ROGERS.
ROGERS, Circuit Judge: Upon review of the air quality
criteria and national ambient air quality standards (“NAAQS”),
the Environmental Protection Agency revised the primary and
secondary NAAQS for lead. See National Ambient Air Quality
Standards for Lead, 73 Fed. Reg. 66,964 (Nov. 12, 2008)
(“Final Rule”). The revision was designed to provide, in light
of recent science, the requisite protection of public health and
welfare by revising the permissible level of lead in ambient air
and revising the averaging time over which the level must be
met. The final rule also revised data handling procedures and
emissions inventory reporting requirements and provided
guidance on implementation. The Coalition of Battery
Recyclers Association and the Doe Run Resources Corporation
(together “petitioners”) contend EPA action was arbitrary and
capricious in multiple ways. Upon review of the rulemaking
record, we are unpersuaded and we deny the petitions for
review.
I.
Pursuant to sections 108 and 109 of the Clean Air Act, 42
U.S.C. §§ 7408–09, EPA regulates air pollutants, “emissions of
which, in [the Administrator’s] judgment, cause or contribute to
air pollution which may reasonably be anticipated to endanger
public health or welfare,” id. § 7408(a)(1)(A). Lead (“Pb”)
emitted into the air can be inhaled or ingested and then absorbed
into the bloodstream, potentially leading to a broad range of
3
adverse health effects including adverse neurological effects in
children. See Final Rule, 73 Fed. Reg. at 66,972–73, 66,975–76.
In 1978 EPA established primary and secondary NAAQS for
lead of 1.5 micrograms of lead per cubic meter of air (µg/m3)
averaged over a calendar quarter. See National Primary and
Secondary Ambient Air Quality Standards for Lead, 43 Fed.
Reg. 46,246 (Oct. 5, 1978) (“1978 Lead NAAQS”). At this
time, adverse neurocognitive effects in children had not been
shown for blood lead levels below 50 micrograms of lead per
deciliter of blood (µg/dL), and the 1978 NAAQS aimed to
prevent most children from exceeding a blood lead level of
30 µg/dL. See id. at 46,246, 46,252. However, later studies
showed adverse neurocognitive effects in children with blood
lead levels below 10 µg/dL. See Final Rule, 73 Fed. Reg. at
66,975–76, 66,984.
EPA began reviewing the NAAQS for lead in 2004,
considering some 6,000 studies and concluding “there is now no
recognized safe level of Pb in children’s blood.” Id. at 66,984.
As part of its review, EPA produced a “Criteria Document”
assessing the latest scientific information regarding health
effects associated with lead in the ambient air. See id. at
66,966–67. EPA’s review “shift[ed] focus from identifying an
appropriate target population mean blood lead level and instead
focuse[d] on the magnitude of effects of air-related Pb on
neurocognitive functions.” Id. at 66,984. EPA developed an
“evidence-based framework” that examined published studies
addressing the relationship between IQ loss in children and air
lead levels. Id. EPA relied to a lesser extent on risk estimates
derived from risk assessment models. Id. at 67,006. Through its
review, EPA sought to identify an air lead level “that would
prevent air-related IQ loss (and related effects) of a magnitude
judged by the Administrator to be of concern in populations of
children exposed to the level of the standard.” Id. at 66,997.
4
To relate IQ loss to air lead levels, EPA used its evidence-
based framework to determine relationships between air lead
levels and blood lead levels (the “air-to-blood ratio”) and
between blood lead levels and IQ loss (the “concentration-
response” relationship). EPA concluded that for each µg/m3
increase of lead in air, children’s blood lead levels increase by
5–10 µg/dL, i.e., the air-to-blood ratio ranged from 1:5 to 1:10.
Id. at 67,002. EPA selected an air-to-blood ratio of 1:7 “as a
generally central value within this range.” Id. at 67,004. EPA
also concluded that the concentration-response relationship is
nonlinear, with greater incremental IQ loss occurring at lower
blood lead levels, and thus that analyses of children with blood
lead levels closest to those of children in the United States today
were most relevant. Id. at 67,002. EPA determined that the
most recently measured mean blood lead level of U.S. children
five years old and younger was 1.8 µg/dL, id., and selected four
study groups involving children with mean blood lead levels
between 2.9 and 3.8 µg/dL rather than groups with higher mean
blood lead levels, id. at 67,003. To “avoid[] focus on a single
estimate that may be unduly influenced by one single analysis,”
each of the four selected groups was from a different study. Id.
Using each group’s reported mean IQ point decrease per µg/dL
increase in blood lead levels, i.e., using the slope of the
concentration-response relationship for each group, EPA
calculated the median concentration-response slope to be
-1.75 µg/dL. Id.
After considering public comments and the
recommendations of its independent scientific review
committee, the Clear Air Scientific Advisory Committee
(“CASAC”), see 42 U.S.C. § 7409(d)(2), as well as “the
uncertainties in the health effects evidence and related
information” and the role of IQ loss in its evidence-based
framework, EPA concluded that an allowable airborne
lead–related loss of two IQ points should be used to set the
5
NAAQS standard. Final Rule, 73 Fed. Reg. at 67,005. CASAC
had stated that “a population loss of 1–2 IQ points is highly
significant from a public health perspective” and that such loss
should be prevented, recommending an air lead level standard
of 0.20 µg/m3 or less. Id. at 66,999–67,000. Combining the
blood-to-air ratio of 1:7, the concentration-response slope of
-1.75 µg/dL, and the allowable air-related IQ loss of 2 points,
EPA concluded that an air lead level standard of 0.15 µg/m3
“would be sufficient to protect public health with an adequate
margin of safety” and “is neither more nor less stringent than
necessary for this purpose.” Id. at 67,007.
EPA also concluded that the appropriate averaging time for
the air lead level standard is a rolling three-month period with a
maximum (not-to-be-exceeded) form evaluated over a period of
three years. Id. at 66,996. EPA had initially proposed an
averaging time of either a calendar quarter or a calendar month,
National Ambient Air Quality Standards for Lead, 73 Fed. Reg.
29,184 (May 20, 2008) (“NPRM”), with CASAC recommending
a monthly averaging period, see Final Rule, 73 Fed. Reg. at
66,993. EPA based its conclusion on scientific evidence
indicating that blood lead levels increase quickly in response to
increased lead exposure, that blood lead levels measured at the
same time as an IQ test (“concurrent blood lead levels”) are
most strongly associated with IQ response, and that these
concurrent blood lead levels reflect lead exposure over the past
one to three months. Id. at 66,992–93. On November 12, 2008,
EPA published the final rule, revising the primary and secondary
NAAQS for lead to 0.15 µg/m3 averaged over a rolling three-
month period. Id. at 66,964.
II.
Petitioners assert that the revised primary lead NAAQS is
overprotective, contending that (A) EPA did not provide
6
sufficient record support for basing the standard on preventing
a decrease of more than two IQ points, (B) reliance on particular
studies relating blood lead levels and IQ was arbitrary and
capricious, and (C) selection of a lead standard of 0.15 µg/m3
was arbitrary and capricious when measured as an average over
a rolling three-month period. Consistent with our standard of
review, see Lead Indus. Ass’n v. EPA, 647 F.2d 1130, 1145–48
(D.C. Cir. 1980); 42 U.S.C. § 7607(d)(9), we conclude these
contentions lack merit because there is substantial record
evidence to support EPA’s conclusions that the population of
children exposed to air lead levels above the revised NAAQS
could suffer, and should be prevented from suffering, average
losses of more than two IQ points, that greater incremental IQ
loss occurs at lower relative blood levels and the more relevant
IQ analyses are those of children with blood levels closest to
today’s population of children, and that a standard of 0.15 µg/m3
measured as a three-month rolling average is required to protect
public health with an adequate margin of safety.
A.
Sensitive populations; focus on IQ decrements. Petitioners’
assertion that the revised lead NAAQS is overprotective because
it is more stringent than necessary to protect the entire
population of young U.S. children ignores that the Clean Air Act
allows protection of sensitive subpopulations. Primary NAAQS
are those “which in the judgment of the Administrator, . . .
allowing an adequate margin of safety, are requisite to protect
the public health.” 42 U.S.C. § 7409(b)(1). In Whitman v.
American Trucking Ass’ns, 531 U.S. 457, 475–76 (2001), the
Supreme Court interpreted “requisite to protect” as “not lower
or higher than is necessary . . . to protect the public health with
an adequate margin of safety.” This court, en banc, cited the
Senate Report accompanying the Clean Air Act explaining that
EPA should set standards providing “a reasonable degree of
protection . . . against hazards which research has not yet
7
identified.” Natural Res. Def. Council v. EPA, 824 F.2d 1146,
1152 (D.C. Cir. 1987) (en banc) (quoting S. REP. NO. 91-1196,
at 10 (1970)) (emphasis and omission in Natural Res. Def.
Council). And so this court has held that “NAAQS must protect
not only average healthy individuals, but also ‘sensitive
citizens’” such as children, and “[i]f a pollutant adversely affects
the health of these sensitive individuals, EPA must strengthen
the entire national standard.” Am. Lung Ass’n v. EPA, 134 F.3d
388, 389 (D.C. Cir. 1998) (quoting S. REP. NO. 91-1196 at 10);
see also Lead Indus. Ass’n, 647 F.2d at 1152–53.
In the Final Rule EPA explained that the scientific evidence
showing the impact of lead exposure in young children in the
United States led it “to give greater prominence to children as
the sensitive subpopulation in this review,” Final Rule, 73 Fed.
Reg. at 66,975, and to focus its revision of the lead NAAQS on
“the sensitive subpopulation that is the group of children living
near [lead emission] sources and more likely to be exposed at
the level of the standard,” id. at 67,000. Given the recent
scientific evidence on which it relied, EPA’s decision to base the
revised lead NAAQS on protecting the subset of children likely
to be exposed to airborne lead at the level of the standard was
not arbitrary or capricious.
Petitioners’ suggestion that EPA failed to explain
adequately its shift in focus, from blood lead levels in the
original 1978 lead NAAQS to IQ decrements in children in the
revised NAAQS, is without merit. In the rulemaking EPA
explained that although the 1978 NAAQS was based on
determining a maximum safe blood lead level for children,
current scientific evidence no longer recognized a safe blood
lead level, and EPA consequently adopted a different focus
when revising the lead NAAQS. See NPRM, 73 Fed. Reg. at
29,229; Final Rule, 73 Fed. Reg. at 66,984. EPA further
explained that epidemiological studies of cognitive effects and
8
lead exposure commonly use IQ scores and that the scientific
literature supports the conclusion that lead exposure causes IQ
loss in children. See Final Rule, 73 Fed. Reg. at 66,976. As
EPA noted in its brief, EPA does not view blood lead levels as
adverse health effects under the Clean Air Act, and both the
1978 and 2008 lead NAAQS focused on preventing adverse
health effects such as neurocognitive effects. See id. at
66,983–84; 1978 Lead NAAQS, 43 Fed. Reg. at 46,252–53.
Petitioners further contend EPA’s use of IQ decrements to
revise the lead NAAQS was arbitrary and capricious because IQ
measurements are more uncertain than blood lead level
measurements. They assert that confounding factors such as
environmental factors affect IQ scores and that the population
significance of IQ loss is imprecise. However, EPA explained
that a large number of high quality studies support the inference
that lead exposure causes population IQ loss, and that animal
studies in which confounding factors are not present show that
low levels of lead cause neurobehavioral effects. See Final
Rule, 73 Fed. Reg. at 66,984 n.56. Petitioners claim that EPA
has acknowledged the standard error of measurement for IQ is
between three and four IQ points, and so assert that an IQ
decrement of two points therefore “cannot be detected at the
level of an individual.” Pet’rs Br. 32. This assertion confuses
the “‘critical’ distinction between population and individual
risk,” wherein a small change in IQ at the level of an individual
is a substantial change at the level of a population, as noted by
EPA when citing the discussion of the differences between
individual-level and population-level data in the Criteria
Document. See Final Rule, 73 Fed. Reg. at 66,976. Thus any
suggestion by petitioners that EPA did not adequately respond
to comments on the IQ test error rates, assuming this issue was
not forfeited by their failure to resubmit the comments as
directed in the notice of proposed rulemaking, see NPRM, 73
Fed. Reg. at 29,190, is without merit.
9
Petitioners also protest, while conceding that CASAC
concluded a population loss of two IQ points “is [a] highly
significant” public health problem, Final Rule, 73 Fed. Reg. at
67,000, that EPA did not provide adequate record support for its
decision to protect against a population loss of more than two IQ
points. In the Final Rule and the Criteria Document EPA
explained that a mean population loss of two IQ points would
cause both a substantial decrease in the percentage of the
population achieving very high IQ scores and a substantial
increase in the percentage achieving very low scores. See id. at
66,976. In addition to citing CASAC’s conclusion, EPA cited
the comments of the American Academy of Pediatrics and state
health agencies that such a loss should be prevented. See id. at
67,000. These explanations sufficiently support EPA’s decision
to prevent a population loss of more than two IQ points.
Regardless of whether EPA sufficiently explained its decision
not to follow the recommendation of CASAC and others to
further reduce IQ loss, a deficiency there would not lend support
to petitioners’ contention that the revised lead NAAQS is
overprotective.
Additionally, while petitioners suggest EPA ignored record
evidence that population IQ has not increased commensurate
with decreasing blood lead levels over the past several decades,
EPA’s response to comments on the proposed rule stated there
is evidence that IQ scores have increased by an average of three
points per decade in the United States over the last several
decades and that IQ tests are routinely renormalized to adjust for
these increases. See EPA Responses to Significant Comments
on the 2008 Proposed Rule on National Ambient Air Quality
Standards for Lead, at 30–31 (Oct. 2008) (“Response to
Comments”). EPA’s response also explained these increases
may be due to improvements in nutrition and to other societal
and environmental factors as well as to reduced exposure to
lead, and concluded from an analysis of the scientific evidence
10
that the decline in blood lead levels over the past several
decades contributed to the increase in IQ over that time period.
For these reasons we conclude that EPA adequately justified
its decision to prevent a loss of more than two IQ points in the
population of children exposed to the level of the revised
NAAQS. Petitioners’ reliance on In re Permian Basin Area
Rate Cases, 390 U.S. 747 (1968), to support their contention that
the revised lead NAAQS is overprotective, is misplaced. The
record shows EPA met its “oblig[ation] at each step of its
regulatory process to assess the requirements of the broad public
interests entrusted to its protection by Congress,” id. at 791, by
considering what lead NAAQS “is requisite to protect public
health, including the health of sensitive groups, with an adequate
margin of safety,” Final Rule, 73 Fed. Reg. at 67,006.
B.
Selection of scientific studies. Petitioners contend that the
record does not support EPA’s decision to exclude from its
determination of the concentration-response slope three
additional study groups involving children with blood lead
levels greater than those of current U.S. children but below
10 µg/dL. The four study groups EPA used to determine the
concentration-response slope were (1) a group from the
“Lanphear” study analyzing blood lead levels below 7.5 µg/dL,
(2) a group from the “Téllez-Rojo” study analyzing blood lead
levels below 5 µg/dL, (3) the “Canfield” study analyzing blood
lead levels below 10 µg/dL, and (4) the “Bellinger and
11
Needleman” study analyzing blood lead levels below 10 µg/dL.1
See Final Rule, 73 Fed. Reg. at 67,003 (Table 3). Like the
Canfield and the Bellinger and Needleman studies on which
EPA relied, the three study groups petitioners wanted EPA to
use, namely an additional group from the Lanphear study and
two additional groups from the Téllez-Rojo study, involved
blood lead levels below 10 µg/dL. Petitioners contend that by
excluding the three study groups EPA did not rely on the latest
scientific knowledge, that the studies’ concentration-response
slopes do not continuously become steeper as blood lead levels
decrease, that there is a larger difference in the steepness of such
slopes above and below 10 µg/dL than above and below
5 µg/dL, and thus that EPA should have included the three
additional study groups because they involved blood lead levels
below 10 µg/dL.
However, EPA explained that the scientific evidence, as
well as CASAC’s recommendations, supported its conclusion
that the concentration-response relationship is nonlinear, in the
sense that a given increase in blood lead levels would cause a
greater IQ loss in a population with low blood lead levels than
a population with higher blood lead levels. See id. at 67,002.
EPA reasonably concluded from this evidence that studies
involving children with blood lead levels closer to the mean
1
See B.P. Lanphear et al., Low Level Environmental Lead
Exposure and Children’s Intellectual Function: An International
Pooled Analysis,” 113 Envtl. Health Perspectives 894 (2005); M.M.
Téllez-Rojo et al., Longitudinal Associations between Blood Lead
Concentrations < 10 µg/dL and Neurobehavioral Development in
Environmentally-Exposed Children in Mexico City, 118 Pediatrics
e323 (2006); R.L. Canfield et al., Intellectual Impairment in Children
with Blood Lead Concentrations below 10 µg per Deciliter, 348 New
Engl. J. Med. 1517 (2003); D.C. Bellinger & H.L. Needleman et al.,
Intellectual Impairment and Blood Lead Levels [letter], 349 New Eng.
J. Med. 500 (2003).
12
blood lead level of today’s population of U.S. children would
provide a more representative estimate of the concentration-
response slope. Id. EPA’s decision not to rely on the three
studies identified by petitioners is practically self-evident from
the table included in the rulemaking, see id. at 66,978–79 (Table
1), and EPA’s explanation for not relying on these three studies
was reasonable. EPA explained that it had included from the
Lanphear and Téllez-Rojo studies only the study group with “a
mean blood Pb level closest to today’s mean for U.S. children.”
Id. at 67,003. EPA further explained that it had “identified four
different studies” to use in determining the concentration-
response slope and thus “avoid[ed] focus on a single estimate
that may be unduly influenced by one single analysis.” Id.
No less, contrary to petitioners’ view, did EPA reasonably
explain why it relied more on the evidence-based framework
than on the risk assessment model results, results that petitioners
assert contradict EPA’s conclusion that a given increase in blood
lead levels would cause a greater IQ loss at lower blood lead
levels than at higher blood lead levels. EPA explained it
considered its risk assessment model results but gave them less
weight than the results in the published studies considered in its
evidence-based framework, because the risk assessment models
were associated with “important uncertainties and limitations”
related to modeling air lead dispersion, lead exposure pathways,
and other factors. Id. at 66,981, 67,006. In addition, EPA noted
that the results obtained from the risk assessment models were
“roughly consistent with and generally supportive of” the air-
related IQ loss estimates EPA determined using its evidence-
based framework. Id. at 67,006 & n.84. To the extent
petitioners rely on the location-specific urban case studies from
the risk assessment models, EPA noted that the population
exposures modeled in the general urban and primary lead
smelter sub-area case studies “relate more closely to the air-
related IQ loss evidence-based framework” than the location-
13
specific urban case studies. Id. at 67,006 n.84. To the extent
petitioners rely on the lower ends of the ranges of IQ losses
estimated by the general urban case study, EPA observed both
that these lower ends “do[] not fully represent the risk associated
with all air-related pathways,” id. at 66,981, and that at a level
of 0.2 µg/m3 the range of IQ loss estimates provided by the
general urban and primary lead smelter sub-area case studies
“are inclusive of the range of estimates” derived from the
evidence-based framework for that level, id. at 67,006 n.84.
Petitioners fare no better in contending EPA unreasonably
assumed the relationship between blood lead levels and IQ was
linear by determining a single number (-1.75 µg/dL) for the
concentration-response slope. EPA concluded from “the weight
of the current evidence” that the concentration-response
relationship is “nonlinear, with steeper slopes at lower blood Pb
levels,” but “recognize[d] uncertainty in the quantitative
characterization of the nonlinearity in the blood Pb–IQ loss
relationship.” Id. at 67,003. Petitioners’ reliance on Natural
Resources Defense Council, 824 F.2d at 1165, where the court
stated that “there is no particular reason” to think straight-line
extrapolation from known to unknown harm levels provides an
accurate prediction of harm, undercuts rather than supports their
contention. Even if EPA were using the value of -1.75 µg/dL to
extrapolate an IQ at a lower blood level rather than to estimate
an IQ loss at a particular air lead level, the court there concluded
from the limitations of extrapolation not that EPA could not
extrapolate, but that “by its nature the finding of risk is uncertain
and the Administrator must use his [or her] discretion to meet
the statutory mandate” of the particular Clean Air Act provisions
involved. Id.
EPA thus adequately justified its decision to rely on
analyses of IQ in children with blood lead levels closest to those
of today’s population of children when revising the lead
14
NAAQS. To the extent petitioners contend the connection
between IQ loss and lead exposure is quantitatively uncertain,
EPA has fulfilled its obligations to “engage in reasoned
decision-making,” Am. Lung Ass’n, 134 F.3d at 392, and to “err
on the side of caution by setting primary NAAQS that ‘allow[]
an adequate margin of safety,’” Am. Trucking Ass’ns v. EPA,
283 F.3d 355, 369 (D.C. Cir. 2002) (quoting 42 U.S.C.
§ 7409(b)(1)) (alteration in Am. Trucking).
C.
Non-conversion with three-month rolling average.
Petitioners contend EPA was arbitrary and capricious not in
selecting a rolling three-month averaging period, but in failing
to convert the 0.15 µg/m3 air lead level from an annual basis to
a three-month basis. This contention lacks merit. EPA based its
selection of a rolling three-month average on different studies
than its selection of the 0.15 µg/m3 standard. Compare Final
Rule, 73 Fed. Reg. at 66,991–96 with id. at 66,996–67,007.
From the studies considered in determining the averaging
period, EPA concluded that the scientific evidence “does not
specify the duration of a sustained air concentration associated
with a particular blood Pb contribution” but does “support[] the
importance of time periods on the order of three months or less.”
Id. at 66,994. Petitioners’ characterization of the 0.15 µg/m3
standard as the product of a unit conversion error is thus inapt.
EPA did rely on at least two published studies referring to
annual or yearly averages: a World Health Organization study
providing an air lead guideline of 0.5 µg/m3 averaged over an
annual period, and a study by S.R. Hilts reporting results of
15
annual blood lead level tests of children living near a smelter.2
EPA relied on both studies in deriving an air-to-blood ratio
rather than in determining the relationship between changes in
blood lead levels and IQ. See Final Rule, 73 Fed. Reg. at
66,973–74. Petitioners fail to demonstrate either study
determined the length of exposure to a particular average
ambient air lead level that resulted in a particular blood lead
level response, see id. at 66,994, and so fail to show that EPA’s
conclusions regarding the lead NAAQS level are valid only for
exposures to lead averaged over a period of one year.
Furthermore, petitioners do not dispute that EPA explained that
the studies on which it based its selection of the averaging
period indicated that adverse health effects may result from lead
exposure over a period of one to three months. See id. at
66,993.
Petitioners erroneously conclude that because EPA’s risk
assessment models could accept inputs only of annual average
ambient air lead concentrations, see id. at 66,980 & n.45, EPA
calculated the 0.15 µg/m3 level based on annual average air lead
exposure rather than three-month exposure. As detailed in
EPA’s Criteria Document, the annual average values were used
in the risk assessment models to estimate blood lead levels
rather than IQ responses to blood lead levels. However, as EPA
explained, the use of annual average ambient air lead
concentrations in the risk assessment modeling was an artifact
of the models themselves, and EPA adjusted the input annual
concentrations to correspond to monthly or quarterly averaging
times. See id.; see also id. at 66,995; Response to Comments at
9.
2
See World Health Organization, Air Quality Guidelines for
Europe 33 (2d ed. 2000); S.R. Hilts, Effect of Smelter Emission
Reductions on Children’s Blood Lead Levels, 303 Sci. Total Env’t 51,
52 (2003).
16
The rulemaking thus demonstrates that EPA adequately
explained that it did not determine the 0.15 µg/m3 air lead level
by assuming exposure to that level over a period of one year,
and that EPA reasonably concluded and adequately explained
that a lead NAAQS of 0.15 µg/m3 measured as a three-month
rolling average is requisite to protect public health with an
adequate margin of safety.
III.
The Lanphear study investigated the concentration-response
relationship between blood lead levels and IQ changes, and
provided what EPA described as “the most compelling
evidence” for effects of lead on IQ at blood lead levels below
10 µg/dL and for the nonlinearity of these effects. Final Rule,
73 Fed. Reg. at 66,977. Petitioners contend the Lanphear study
contained such errors that EPA acted arbitrarily and capriciously
in relying on results from the study without first obtaining and
making public the underlying data for the study. However, in
American Trucking, 283 F.3d 355, this court rejected the notion
that EPA had improperly failed to obtain and make public data
underlying studies on which it had relied during a NAAQS
rulemaking, holding that “[t]he Clean Air Act imposes no such
obligation” and that “requiring agencies to obtain and publicize
the data underlying all studies on which they rely would be
impractical and unnecessary.” Id. at 372 (quotation marks
omitted).
Petitioners attempt to distinguish their request on the
ground that in American Trucking the court was addressing
requests for data underlying several studies, while they request
only that EPA obtain and make public the data underlying the
Lanphear study. This distinction finds no support in the
reasoning of American Trucking. Rather than distinguishing
between an agency’s burden in obtaining data from one versus
17
many studies, the court distinguished EPA’s reliance on a
study’s results from its reliance on the raw data underlying such
results, noting that raw data often is unavailable due to
proprietary interests of a study’s scientific investigators or
confidentiality agreements with study participants. See id.
Petitioners do not contend EPA possessed the underlying data
but failed to include it in the rulemaking record. Cf. Am. Radio
Relay League, Inc. v. FCC, 524 F.3d 227, 237–38 (D.C. Cir.
2008).
Petitioners’ reliance on EPA’s regulation concerning
requests for data under the Freedom of Information Act, 5
U.S.C. § 552 (“FOIA”), is forfeited. This regulation provides
that in response to a FOIA request, EPA shall request and make
public “research data relating to published research findings
produced under an award” of an EPA grant, subject to privacy
and other considerations. 40 C.F.R. § 30.36(d). Petitioners
cited this regulation for the first time during rebuttal oral
argument, affording EPA no opportunity to respond. See Ark
Las Vegas Rest. Corp. v. NLRB, 334 F.3d 99, 108 n.4 (D.C. Cir.
2003). Nor did petitioners preserve this argument by
mentioning their separate FOIA request for the Lanphear study
data in a footnote of their brief, without citing 40 C.F.R. §
30.36(d). See, e.g., Nat’l Mining Ass’n v. Mine Safety & Health
Admin., 599 F.3d 662, 671 (D.C. Cir. 2010); NSTAR Elec. &
Gas Corp. v. FERC, 481 F.3d 794, 800 (D.C. Cir. 2007). In any
event, the issue raised by petitioners in the instant case involves
whether EPA has a general duty to obtain and make public
underlying data as part of a rulemaking, rather than any specific
duties related to FOIA releases; more generally the regulation
concerns EPA’s reservation of rights in intangible property
related to its grant awards, and the court has no occasion to note
more than that the regulation does not apply here.
18
Similarly unavailing is petitioners’ contention that EPA did
not respond adequately to comments concerning errors in the
Lanphear study. EPA responded to petitioners’ comments by
noting that errors in Table 4 and two typographical errors in the
Lanphear study had been corrected and explaining that these
corrected errors did not affect the portions of the study on which
EPA had relied. See Response to Comments at 72. EPA also
explained in response to petitioners’ questioning of Figure 3 of
the Lanphear study how petitioners had misinterpreted the
statistical methods involved in the figure and that EPA’s
conclusions from the Lanphear study did not depend on the
figure. See id. at 24–25. EPA further observed that its
confidence in the Lanphear study had been reaffirmed by the
“Rothenberg” study, which had re-analyzed the Lanphear data
and confirmed the nonlinear relationship between IQ and blood
lead levels shown in Figure 3 of the Lanphear study.3 See id. at
25.
Consequently, petitioners have failed to identify errors in
the Lanphear study that would make EPA’s reliance on it
arbitrary and capricious, and EPA thus appropriately considered
the Lanphear study and was not required to obtain and make
public the data underlying the Lanphear study.
IV.
Finally Doe Run Resources, Inc., separately contends that
EPA erroneously concluded it lacked statutory authority under
the Clean Air Act to consider the bioavailability of lead sulfides
when determining compliance with the lead NAAQS. Doe Run
asserts that lead sulfides are less bioavailable than other forms
3
See S.J. Rothenberg & J.C. Rothenberg, Testing the
Dose-Response Specification in Epidemiology: Public Health and
Policy Consequences for Lead, 113 Envtl. Health Persp. 1190 (2005).
19
of lead and thus less likely to contribute to health risks, and that
EPA’s NAAQS compliance evaluations consequently should
treat low-bioavailable lead sulfides as “policy relevant
background” with “in effect, a waiver from the general NAAQS
rule.” Pet’rs Br. 55–56. EPA declined to determine compliance
with the lead NAAQS based on the bioavailability of lead
sulfides, explaining that it “must determine compliance with the
standard” pursuant to Clean Air Act § 109(b), 42 U.S.C.
§ 7409(b). See Response to Comments at 65.
The term “policy-relevant background” does not appear in
the Clean Air Act, and in the context of the lead NAAQS, EPA
defines “policy-relevant background” as natural emissions of
lead into the air from non-anthropogenic sources as well as lead
from non-air sources. See NPRM, 73 Fed. Reg. at 29,192. In
responding to comments EPA stated that evidence indicates the
bioavailability of lead sulfides increases over time and that EPA
considered this variable bioavailability in revising the lead
NAAQS. See Response to Comments at 10. In its brief on
appeal, EPA noted that concentrations of ambient lead near lead
mines or smelters are not naturally occurring background
concentrations. Resp’t Br. 57 n.27. Under the circumstances,
EPA reasonably concluded Doe Run’s request to treat the
bioavailability of lead sulfides differently was effectively a
request for a waiver from the lead NAAQS.
Applying the familiar two-step analysis under Chevron,
U.S.A., Inc. v. Natural Resources Defense Council, Inc., 467
U.S. 837 (1984), see Natural Res. Def. Council v. EPA, 489 F.3d
1364, 1371 (D.C. Cir. 2007), we conclude Doe Run’s contention
fails at both steps. First, Congress provided in the Clean Air Act
that primary NAAQS “shall be ambient air quality standards the
attainment and maintenance of which in the judgment of the
Administrator . . . are requisite to protect the public health.” 42
U.S.C. § 7409(b)(1). EPA is required to designate “any area that
20
does not meet (or that contributes to ambient air quality in a
nearby area that does not meet) the national primary or
secondary ambient air quality standard for the pollutant” as
“nonattainment.” Id. § 7407(d)(1). Further, EPA “may not
promulgate a redesignation of a nonattainment area (or portion
thereof) to attainment unless” it “determines that the area has
attained the national ambient air quality standard.” Id.
§ 7407(d)(3)(E)(i). Additionally, Congress addressed the
circumstances under which attainment could be waived, e.g., as
with certain particulate matter sources, see 42 U.S.C. § 7513(f),
and with emissions emanating from outside the United States,
see 42 U.S.C. § 7509a, but provided no authorization for EPA
to waive NAAQS attainment requirements in the manner
requested by Doe Run. Even assuming the Clean Air Act was
ambiguous with regard to whether EPA was empowered to grant
other waivers, EPA’s interpretation of its authority under the
statutory scheme is permissible under Chevron step two, 467
U.S. at 843, and entitled to deference by the court.
Accordingly, we deny the petitions.