United States Court of Appeals
For the First Circuit
No. 08-2303
UNITED STATES OF AMERICA,
Appellee,
v.
MOISE WILLIAM,
Defendant, Appellant.
APPEAL FROM THE UNITED STATES DISTRICT COURT
FOR THE DISTRICT OF NEW HAMPSHIRE
[Hon. Steven J. McAuliffe, U.S. District Judge]
Before
Lynch, Chief Judge,
Boudin, Circuit Judge,
and Saylor,* District Judge.
Judith Mizner, Assistant Federal Public Defender, Federal
Defender Office, with whom Martin J. Vogelbaum, Assistant Federal
Public Defender, Federal Defender Office, was on brief for
appellant.
Seth R. Aframe, Assistant United States Attorney, with whom
Aixa Maldonado-Quiñones, Assistant United States Attorney, and
Michael J. Gunnison, Acting United States Attorney, were on brief
for appellee.
April 22, 2010
*
Of the District of Massachusetts, sitting by designation.
BOUDIN, Circuit Judge. On the evening of July 21, 2007,
Moise William's car was halted at a sobriety checkpoint on New
Hampshire Route 28 manned by Auburn, New Hampshire, police
officers. The checkpoint had been authorized by a state judge on
petition of the town's police department. All vehicles (except
tractor trailers due to parking-space problems) were stopped
briefly. When William rolled down the driver's side window, the
officer who had approached smelled marijuana and noticed that
William's eyes were glassy and bloodshot. The officer asked
William to step out of the car.
In the ensuing search of the car, marijuana was found.
During the search, a fireman standing nearby saw William throw a
package on the ground; later testing revealed it to contain crack
cocaine. A further search with a drug-sniffing dog revealed more
marijuana in the car. William was arrested and indicted on two
counts of possessing a controlled substance, one for marijuana and
the other for "cocaine base ('crack')." 21 U.S.C. § 844(a) (2006).
The government also alleged that the crack cocaine weighed three or
more grams and filed an information pursuant to 21 U.S.C. § 851
alleging that William had a prior conviction for possession of
crack cocaine.
William moved to suppress all of the drug evidence on the
ground that the drugs were the fruits of an illegal stop. The
district court held a hearing, heard testimony, and concluded that
-2-
the stop of William's car was lawful. In response to a second
motion to suppress, the court also held that the car search was
lawful, but that issue is not before us on this appeal. Thereafter
William was tried on both counts.
Among the items seized when William was stopped was a
wrapper containing eleven small bags. At trial a government expert
testified that these contained crack cocaine weighing approximately
3.6 grams; but pursuant to laboratory policy, the expert had
weighed only five of the eleven bags, determined the average weight
per bag, and multiplied by eleven to estimate the total weight of
all of the bags. The five bags tested weighed 1.64 grams, making
the expert's estimate 3.6 grams for all eleven.1
Although the weight went only to penalty, the issue was
submitted to the jury--as required by Apprendi v. New Jersey, 530
U.S. 466 (2000)--because a weight exceeding three grams would raise
the maximum possible penalty. 21 U.S.C. § 844(a). The trial judge
expressed unease about the chemist's method, given that the margin
was small (3 grams versus 3.6) and the standard of proof was
"beyond a reasonable doubt." However, the trial court denied
William's motion seeking a judgment of acquittal based on the
weight issue, Fed. R. Crim. P. 29, and the jury convicted William
1
The chemist also testified that he tested only five of the
eleven bags to determine whether the substance inside was cocaine
base in the form of crack cocaine, but that all of the bags
contained "an off-white powdery chunky material."
-3-
on both counts and found by special verdict that the total amount
of crack cocaine exceeded three grams.
In sentencing William on the crack cocaine conviction,
the trial court applied a statutory provision, 21 U.S.C. § 844(a),
setting a five-year mandatory minimum sentence for possession of
more than three grams of cocaine base when the defendant has a
"prior conviction. . . under this subsection." The court sentenced
William to the five-year prison term, to run concurrently with a
one-year sentence for possession of marijuana. William now appeals
to challenge both his conviction and sentence on the crack cocaine
charge.
The challenge to the conviction rests solely on the
lawfulness of the stop. Although individualized suspicion is
normally required for a car stop and "probable cause" is required
for an arrest, the Fourth Amendment rubric is "reasonableness."
City of Indianapolis v. Edmond, 531 U.S. 32, 37 (2000); United
States v. Almeida, 434 F.3d 25, 28 (1st Cir. 2006). Under this
rubric, the Supreme Court has permitted vehicle checkpoints and
very brief inquiries of all drivers for certain purposes and with
certain safeguards: one of the allowed uses is for sobriety
checkpoints, the principal authority being Michigan Department of
State Police v. Sitz, 496 U.S. 444 (1990).2
2
See also Illinois v. Lidster, 540 U.S. 419, 423-28 (2004)
(upholding checkpoint to locate witnesses to a hit-and-run); United
States v. Martinez-Fuerte, 428 U.S. 543, 545 (1976) (upholding
-4-
The threshold requirement under Sitz and Edmond--that
sobriety concerns be the primary purpose of the checkpoint--is met
in this case. See Edmond, 531 U.S. at 47-48; Sitz, 496 U.S. at
451. New Hampshire has a general procedure for authorizing
sobriety checkpoints; and in this case a plan was submitted and
approved by a state judge, and the directions to the officers were
consistent with operating a sobriety checkpoint. The police were
aware that other crimes might come to light; thus, a drug-sniffing
dog was kept in reserve but brought forward to William's car only
after drugs had been initially found due to William's appearance
and the odor of marijuana.
William's argument that the checkpoint's primary purpose
was something other than detecting impaired driving consists mostly
of objections to the quality of the statistical data cited in the
petition seeking approval for the checkpoint. But such weaknesses
give no reason in this case to think that the district court
clearly erred in concluding that the government was in fact
conducting a sobriety checkpoint. United States v. Green, 293 F.3d
855, 859 (5th Cir. 2002) (reviewing district court findings on
primary purpose for clear error); United States v. Davis, 270 F.3d
977, 980 (D.C. Cir. 2001) (same).
William cites two decisions from the District of Columbia
Circuit that have found traffic checkpoints doubtful in particular
checkpoint near border to detect illegal aliens).
-5-
cases, but in both cases, the evidence suggested that the stops
were not primarily vehicle registration checks, as the officers
claimed, but rather general anti-crime stops implemented as part of
broader crime-control task forces. See United States v. Bowman,
496 F.3d 685, 693-94 (D.C. Cir. 2007) (remanding for further fact-
finding on primary purpose); United States v. Davis, 270 F.3d 977,
981-82 (D.C. Cir. 2001) (same). The reasons for allowing one kind
of use and barring another can be debated; but the short answer is
that the Supreme Court has thus distinguished and that is the end
of the matter. Edmond, 531 U.S. at 47-48.
Other cases William cites likewise involved affirmative
evidence that the checkpoint's stated purpose was pretextual.
United States v. Huguenin, 154 F.3d 547, 555-56 (6th Cir. 1998)
(checkpoint funded by drug interdiction sources was supervised by
narcotics officer); United States v. Ramirez-Gonzalez, 87 F.3d 712,
715-16 (5th Cir. 1996) (remanding for an evidentiary hearing based
on similar evidence). Unlike those cases, William does not show
that the Auburn checkpoint was part of some task force or program
aimed at some purpose other than sobriety.
Beyond assessing the legitimacy of the purpose, Sitz and
other checkpoint cases have used Brown v. Texas, 443 U.S. 47, 50-51
(1979), as the source of a further balancing test for determining
the reasonableness of the checkpoint. Several formulations have
been used but a helpful summary is provided by a Fourth Circuit
-6-
decision, United States v. Henson, No. 08-4221, 2009 WL 3792435, at
*2 (4th Cir. Nov. 13, 2009) (citations omitted):
If the primary purpose was valid, the court
must then judge the checkpoint's
reasonableness on the basis of individual
circumstances. . . . This requires balancing
"'the gravity of the public concerns served by
the seizure, the degree to which the seizure
advances the public interest, and the severity
of the interference with individual liberty.'"
. . . Factors to weigh intrusiveness include
whether the checkpoint: (1) is clearly
visible; (2) is part of some systematic
procedure that strictly limits the
discretionary authority of police officers;
and (3) detains drivers no longer than is
reasonably necessary to accomplish the purpose
of checking a license and registration, unless
other facts come to light creating a
reasonable suspicion of criminal activity.
In the present case, Sitz establishes the "gravity of the
public concerns" with the death and injury toll taken by drunken
drivers and the fact that sobriety checkpoints can "advance the
public interest" sufficiently to make such a checkpoint reasonable.
496 U.S. at 451. William argues, however, that the data offered to
support the use of this checkpoint at this location was inadequate
and that the results in arrest numbers were unimpressive.
Certainly one can imagine a purported sobriety checkpoint whose
location or timing was demonstrably unlikely to be of use.
But in Sitz, Chief Justice Rehnquist went out of his way
to say that whether and where to establish a stop is primarily a
judgment for state or local officials. 496 U.S. at 453-54. In
addition, using the number of drunk driving arrests resulting from
-7-
a specific checkpoint has at least two problems: one is that the
reasonableness of the effort is primarily a forward-looking
exercise (in fact, the percentage of arrests in Sitz was very low,
id. at 455); and the other is that sobriety checkpoints likely have
a deterrent value apart from immediate detentions resulting from
the stops.
In this case, the police petition and affidavit
supporting the court order showed that the checkpoint was on a
regularly used highway and that the days and the hours selected
were nights in or around weekends when party-going would be most
likely to generate drunk driving. Much of the data offered in the
petition in this case as to drunk driving and the utility of checks
related to New Hampshire rather than the location chosen; but in
Sitz most of the data was nationwide. 496 U.S. at 451. This is
far from the case of a checkpoint on a dead-end road.
The other concerns in Brown v. Texas center around the
intrusiveness of the search and its potential for abuse. This
begins with the distinction drawn by the courts between fixed
checkpoints and random "roving" stops that are highly stressful for
the innocent individual pulled over by flashing lights and siren.
Sitz, 496 U.S. at 452-53; Martinez-Fuerte, 428 U.S. at 558. An
orderly checkpoint usually reveals itself as a general stop of all
vehicles and is deemed less alarming. Martinez-Fuerte, 428 U.S. at
559.
-8-
William argues that the checkpoint in this case was not
well-marked, that it was operated in a sloppy fashion, and that the
handling of William's drug evidence was badly documented. But the
checkpoint constrained officers' discretion in the areas most
important to the constitutional analysis: the checkpoint occurred
at a fixed location, the officers stopped every car passing
through, and further investigation occurred only if individualized
suspicion developed. The concerns with roving stops and unlimited
officer discretion were absent. Sitz, 496 U.S. at 452-54; Delaware
v. Prouse, 440 U.S. 648, 661 (1979).
In his opening brief, William also challenged his
sentence. In sentencing William for the cocaine base conviction,
the trial court applied a provision of 42 U.S.C. § 844(a) fixing a
five-year mandatory minimum sentence for possession of more than
three grams of cocaine base where, as alleged here, the defendant
has a prior conviction under that subsection. William argues to
us, as he did to the trial judge, that the evidence of weight was
inadequate and warranted a judgment of acquittal as to weight.
The main argument is that the weighing of five bags out
of eleven was insufficient absent a basis for the jury to be sure
the bags were similar in weight, and the expert gave no specific
reason to support the assumption. The trial judge, as noted,
shared the defendant's doubt but left the issue to the jury.
William points out that, given the sentencing jurisprudence, the
-9-
weight issue required a finding beyond a reasonable doubt to raise
the permissible maximum above the one-year default. See 21 U.S.C.
§ 844(a); United States v. Eirby, 262 F.3d 31, 36-37 (1st Cir.
2001).
The problem is an ongoing one. Reasonable estimation
techniques are permitted;3 it is perhaps implicit here that the
expert deemed the bags comparable in weight, and, in addition, the
bags were available to the jury who could judge from appearance.4
Indeed, the bags may well have been professionally prepared for
sale, which would support the inference; William had three
cellphones and lots of cash in his possession when arrested.
Still, the margin (3.6 grams as against 3) was not large and only
a couple of light weight bags among those unweighed could have made
the difference.
We need not pursue the issue--one could say much more on
both sides--beyond agreeing with the trial judge that this was a
close case and therefore a warning to prosecutors in future cases:
it would have been easy to shore up the estimate as the case headed
for trial. However, both sides agree that on appeal a clear-cut
3
United States v. Correa-Alicea, 585 F.3d 484, 489-91 (1st
Cir. 2009); United States v. Rodriquez, 525 F.3d 85, 107-09 (1st
Cir. 2008).
4
Although defense counsel points out that the tested drugs
from five of the bags were no longer in lump form, uniformity in
size of the lumps in the remaining six bags would still permit an
inference that all of the bags were roughly uniform.
-10-
error in sentencing has now come to light that precludes the use of
the higher sentence even if weight exceeded three grams and that
William has already served what may well be the maximum sentence
for his offense.
The provision applied to William--a five-year minimum and
a twenty-year maximum--required not only more than three grams of
cocaine but also a "prior conviction . . . under this subsection."
21 U.S.C. § 844(a). William's prior conviction, both sides now
agree, was for a state drug offense and does not meet the
requirements of § 844(a) for a sentence above two years.
William's prior state law conviction did increase the
statutory range to more than the default maximum of one year: where
the defendant has "a prior conviction for any drug, narcotic, or
chemical offense chargeable under the law of any State," the
minimum term is fifteen days and the maximum term is two years.
Id.. Weight is irrelevant to this provision and, although the
issue was not raised in the district court, the government concedes
plain error in a sentence exceeding two years.
When this concession was made, William had served more
than two years in jail and so might have been released on bail
pending review, but he rejected the proposal for reasons relating
to exposure to possible deportation proceedings. Accordingly, we
now affirm his conviction, vacate his sentence and remand the case
-11-
to the district court for re-sentencing. The mandate will issue
forthwith without prejudice to any petition for rehearing.
It is so ordered.
-12-