Evaluation Analytical Support 51 (2000) (“2000 TIFA Study”). The same is true of
• After deriving hourly fatigue-related crash-risk estimates using its curve,
TIFA Study at 51.
by the increased weekly driving and working hours permitted under the 34-hour
restart provision.
required by the APA. Air Transp. Ass’n of Am., 169 F.3d at 7 (citation, internal
crashes at Hour 13 and beyond at Hour 17. The TIFA studies, by contrast,
number of the methodology‘s critical elements. To satisfy the APA’s “arbitrary
and capricious” standard, an agency must “articulate a satisfactory explanation
for its action including a ‘rational connection between the facts found and the
reasoned decisionmaking,” id. at 52, 103 S.Ct. 2856.
[9] Link to KeyCite Notes 1. Public Citizen notes that the TIFA data, upon which
FMCSA’s time-on-task multipliers were ultimately based, indicates that “the risk
of fatal-crash involvement more than doubled from the 10th hour to the 11th.”
Public Citizen Br. 48-49 (citing 2005 RIA at 45 (J.A. 1665)). The actual time-on–
First, as explained above, instead of using the crash risk figures for each hour of
driving that the TIFA Study had calculated directly from the actual crash data,
FMCSA derived a cubic curve of crash risk as a function of time on task. To
derive the curve, FMCSA first plotted the TIFA figures for Hours 1 through 12,
and then used an aggregate measure for Hour 13 and beyond. It did not,
curve would have produced a significantly higher estimate of the risk of a fatigue–
related crash at Hour 11-a figure close to that which the TIFA Study had
calculated directly. See Public Citizen Br. 49-50, A-3 (displaying alternative
curve).
[10] Link to KeyCite Notes[11] Link to KeyCite Notes FMCSA’s decision to plot
Phase-Down Task Force v. EPA, 705 F.2d 506, 535 (D.C.Cir.1983)). Although
the agency’s brief defends the use of a cubic curve on the ground that the margin
of error in the underlying TIFA data was relatively large,FN3 this again misses
the point: the issue is not whether a curve should have been used, but why the
agency chose to draw the curve by plotting the Hour 13+ data at Hour 17. On
determined that the TIFA data “represent[s] the only recently-published data
available for considering” time–on-task effects, and chose to rely on it as the
“prudent” and conservative course. Id. at 50,052.
FN4. FMCSA’s counsel did attempt to explain the derivation of the Hour 17 figure
at oral argument and in a post-argument letter to the court. Counsel represented
the merits of the agency‘s averaging methodology, we cannot affirm on the basis
of a post-hoc explanation by agency counsel. See SEC v. Chenery Corp., 332
U.S. 194, 196, 67 S.Ct. 1575, 91 L.Ed. 1995 (1947); Public Citizen, 374 F.3d at
1218.
Second, after deriving an estimate of crash risk for each hour of driving using a
risk at the 11th hour by an average that includes that heightened 11th–hour risk.”
Public Citizen Br. 50. We express no view on the validity of FMCSA’s statistical
method, but we agree with Public Citizen that-once again-the agency offered no
explanation for its decision during the rulemaking and failed even to respond to
the petitioner’s argument in its brief. Although we apply a deferential standard of
FN5. FMCSA conducted a sensitivity analysis that concluded that an 11-hour
daily driving limit remained more cost-effective than a 10-hour limit, even if the
agency assumed a higher risk of fatigue-related crash in the 11th hour than it
employed in its operator-fatigue model. See 2005 Rule, 70 Fed.Reg. at 50,046–
47; 2005 RIA at ES-7 to -8 (J.A. 1634-35). FMCSA does not argue, however, that
in weekly hours was likely “ ‘an important aspect of the problem,’ ” and that the
“agency’s failure to address” it made “the rule’s rationality questionable.” 374
F.3d at 1222-23 (quoting State Farm, 463 U.S. at 43, 103 S.Ct. 2856). Indeed, in
the 2000 NPRM, FMCSA expressed concern about evidence that many drivers
were exceeding the weekly limit of 60 hours in 7 days. See 2000 NPRM, 65
statements by the agency, Public Citizen argues that the operator-fatigue model
should have taken into account the increased crash risk caused by “cumulative
fatigue” associated with the increased driving and working hours that it would
permit.
FMCSA’s counsel responds with a single conclusory sentence denying that the
effect that is independent of that caused by insufficient sleep. There is no
indication that the operator-fatigue model considered this kind of “cumulative
fatigue,” and FMCSA offers no explanation for the omission.
Of course, it could be that “cumulative fatigue” due to longer weekly service
hours will not constitute a significant problem because, for example, the number
conclusion that the 34-hour restart provision will have economic benefits. See id.
at 50,049; 2005 RIA at ES-3 to -4, 68-69 (J.A. 1630-31, 1688-89). FMCSA
concedes as much. See 2005 RIA at 18 (J.A. 1654) (“[M]any drivers work and
drive longer hours than the averages.”); id. at 67 (J.A. 1687) (“[M]ore than half of
for-hire operations, and somewhat less than half of private fleet operations, are
may not supply a reasoned basis for the agency’s action that the agency itself
has not given.” State Farm, 463 U.S. at 43, 103 S.Ct. 2856 (internal quotation
marks omitted); see SEC v. Chenery Corp., 332 U.S. 194, 196, 67 S.Ct. 1575, 91
L.Ed. 1995 (1947). Accordingly, the agency’s failure of explanation renders the
restart provision arbitrary and capricious. Cf. U.S. Air Tour Ass’n, 298 F.3d at
that it failed to provide an explanation for the elements of that methodology that
Public Citizen disputes. Because the model is the basis for the cost-benefit
analysis that led FMCSA to adopt the two provisions of the 2005 Rule that Public
Citizen challenges-the increase in the daily driving limit from 10 to 11 hours, and
the 34-hour restart provision-we must vacate those provisions. And because
hour daily on-duty limit is arbitrary and capricious, because FMCSA failed to
consider its negative effects on driver health and safety. Finally, the petitioner
objects to the modification of the sleeper-berth exception on several grounds.
The following subparts address each of these challenges.
A
unloading operations …).” 49 U.S.C. § 31136 note. Loading and unloading
operations contribute to driver fatigue primarily by prolonging drivers’ hours.
Because most drivers are paid by the mile or by the load rather than by the hour,
it costs a shipper or receiver nothing to keep drivers waiting to load or unload
their vehicles. The resulting delays associated with loading and unloading can
unloading operations for the problem of driver fatigue. In fact, FMCSA adopted
the 14-hour daily on-duty limit, and eliminated the provision in the pre-2003
regulations that extended the daily limit when the driver took breaks, in part
because of those concerns. As the agency explained:
Under the pre-2003 rules, drivers were allowed a 15-cumulative-hour duty period
or unload. FMCSA concluded that, as a consequence, the 2005 Rule “prevents
shippers, receivers, and companies from abusing the off-duty hours and forcing
drivers to use them as unpaid time.” Id. at 50,013.
In support of this conclusion, the agency reviewed comments from drivers and
industry associations regarding how the limit was working in practice. (It was
driver on-duty time not devoted to driving ….” (citation and internal quotation
marks omitted)).
We conclude that, by expressly considering fatigue-related issues pertaining to
loading and unloading operations, and by reasonably identifying the
nonextendable 14-hour limit as responsive to those issues, the 2005 Rule
decision to make the 14-hour daily on-duty limit nonextendable through the use
of breaks. Specifically, OOIDA contends that this provision “discourages drivers
from taking short rest breaks and naps,” because they can no longer exclude
such breaks from their daily on-duty limit. OOIDA Br. 43. In support of this
argument, OOIDA cites a study of its members indicating that 60% of drivers
number of the methodology‘s critical elements. To satisfy the APA’s “arbitrary
and capricious” standard, an agency must “articulate a satisfactory explanation
for its action including a ‘rational connection between the facts found and the
reasoned decisionmaking,” id. at 52, 103 S.Ct. 2856.
[9] Link to KeyCite Notes 1. Public Citizen notes that the TIFA data, upon which
FMCSA’s time-on-task multipliers were ultimately based, indicates that “the risk
of fatal-crash involvement more than doubled from the 10th hour to the 11th.”
Public Citizen Br. 48-49 (citing 2005 RIA at 45 (J.A. 1665)). The actual time-on–
First, as explained above, instead of using the crash risk figures for each hour of
driving that the TIFA Study had calculated directly from the actual crash data,
FMCSA derived a cubic curve of crash risk as a function of time on task. To
derive the curve, FMCSA first plotted the TIFA figures for Hours 1 through 12,
and then used an aggregate measure for Hour 13 and beyond. It did not,
curve would have produced a significantly higher estimate of the risk of a fatigue–
related crash at Hour 11-a figure close to that which the TIFA Study had
calculated directly. See Public Citizen Br. 49-50, A-3 (displaying alternative
curve).
[10] Link to KeyCite Notes[11] Link to KeyCite Notes FMCSA’s decision to plot
Phase-Down Task Force v. EPA, 705 F.2d 506, 535 (D.C.Cir.1983)). Although
the agency’s brief defends the use of a cubic curve on the ground that the margin
of error in the underlying TIFA data was relatively large,FN3 this again misses
the point: the issue is not whether a curve should have been used, but why the
agency chose to draw the curve by plotting the Hour 13+ data at Hour 17. On
determined that the TIFA data “represent[s] the only recently-published data
available for considering” time–on-task effects, and chose to rely on it as the
“prudent” and conservative course. Id. at 50,052.
FN4. FMCSA’s counsel did attempt to explain the derivation of the Hour 17 figure
at oral argument and in a post-argument letter to the court. Counsel represented
the merits of the agency‘s averaging methodology, we cannot affirm on the basis
of a post-hoc explanation by agency counsel. See SEC v. Chenery Corp., 332
U.S. 194, 196, 67 S.Ct. 1575, 91 L.Ed. 1995 (1947); Public Citizen, 374 F.3d at
1218.
Second, after deriving an estimate of crash risk for each hour of driving using a
risk at the 11th hour by an average that includes that heightened 11th–hour risk.”
Public Citizen Br. 50. We express no view on the validity of FMCSA’s statistical
method, but we agree with Public Citizen that-once again-the agency offered no
explanation for its decision during the rulemaking and failed even to respond to
the petitioner’s argument in its brief. Although we apply a deferential standard of
FN5. FMCSA conducted a sensitivity analysis that concluded that an 11-hour
daily driving limit remained more cost-effective than a 10-hour limit, even if the
agency assumed a higher risk of fatigue-related crash in the 11th hour than it
employed in its operator-fatigue model. See 2005 Rule, 70 Fed.Reg. at 50,046–
47; 2005 RIA at ES-7 to -8 (J.A. 1634-35). FMCSA does not argue, however, that
in weekly hours was likely “ ‘an important aspect of the problem,’ ” and that the
“agency’s failure to address” it made “the rule’s rationality questionable.” 374
F.3d at 1222-23 (quoting State Farm, 463 U.S. at 43, 103 S.Ct. 2856). Indeed, in
the 2000 NPRM, FMCSA expressed concern about evidence that many drivers
were exceeding the weekly limit of 60 hours in 7 days. See 2000 NPRM, 65
statements by the agency, Public Citizen argues that the operator-fatigue model
should have taken into account the increased crash risk caused by “cumulative
fatigue” associated with the increased driving and working hours that it would
permit.
FMCSA’s counsel responds with a single conclusory sentence denying that the
effect that is independent of that caused by insufficient sleep. There is no
indication that the operator-fatigue model considered this kind of “cumulative
fatigue,” and FMCSA offers no explanation for the omission.
Of course, it could be that “cumulative fatigue” due to longer weekly service
hours will not constitute a significant problem because, for example, the number
conclusion that the 34-hour restart provision will have economic benefits. See id.
at 50,049; 2005 RIA at ES-3 to -4, 68-69 (J.A. 1630-31, 1688-89). FMCSA
concedes as much. See 2005 RIA at 18 (J.A. 1654) (“[M]any drivers work and
drive longer hours than the averages.”); id. at 67 (J.A. 1687) (“[M]ore than half of
for-hire operations, and somewhat less than half of private fleet operations, are
may not supply a reasoned basis for the agency’s action that the agency itself
has not given.” State Farm, 463 U.S. at 43, 103 S.Ct. 2856 (internal quotation
marks omitted); see SEC v. Chenery Corp., 332 U.S. 194, 196, 67 S.Ct. 1575, 91
L.Ed. 1995 (1947). Accordingly, the agency’s failure of explanation renders the
restart provision arbitrary and capricious. Cf. U.S. Air Tour Ass’n, 298 F.3d at
that it failed to provide an explanation for the elements of that methodology that
Public Citizen disputes. Because the model is the basis for the cost-benefit
analysis that led FMCSA to adopt the two provisions of the 2005 Rule that Public
Citizen challenges-the increase in the daily driving limit from 10 to 11 hours, and
the 34-hour restart provision-we must vacate those provisions. And because
hour daily on-duty limit is arbitrary and capricious, because FMCSA failed to
consider its negative effects on driver health and safety. Finally, the petitioner
objects to the modification of the sleeper-berth exception on several grounds.
The following subparts address each of these challenges.
A
unloading operations …).” 49 U.S.C. § 31136 note. Loading and unloading
operations contribute to driver fatigue primarily by prolonging drivers’ hours.
Because most drivers are paid by the mile or by the load rather than by the hour,
it costs a shipper or receiver nothing to keep drivers waiting to load or unload
their vehicles. The resulting delays associated with loading and unloading can
unloading operations for the problem of driver fatigue. In fact, FMCSA adopted
the 14-hour daily on-duty limit, and eliminated the provision in the pre-2003
regulations that extended the daily limit when the driver took breaks, in part
because of those concerns. As the agency explained:
Under the pre-2003 rules, drivers were allowed a 15-cumulative-hour duty period
or unload. FMCSA concluded that, as a consequence, the 2005 Rule “prevents
shippers, receivers, and companies from abusing the off-duty hours and forcing
drivers to use them as unpaid time.” Id. at 50,013.
In support of this conclusion, the agency reviewed comments from drivers and
industry associations regarding how the limit was working in practice. (It was
driver on-duty time not devoted to driving ….” (citation and internal quotation
marks omitted)).
We conclude that, by expressly considering fatigue-related issues pertaining to
loading and unloading operations, and by reasonably identifying the
nonextendable 14-hour limit as responsive to those issues, the 2005 Rule
decision to make the 14-hour daily on-duty limit nonextendable through the use
of breaks. Specifically, OOIDA contends that this provision “discourages drivers
from taking short rest breaks and naps,” because they can no longer exclude
such breaks from their daily on-duty limit. OOIDA Br. 43. In support of this
argument, OOIDA cites a study of its members indicating that 60% of drivers