UNITED STATES COURT OF APPEALS
FOR THE FIFTH CIRCUIT
No. 98-30095
Summary Calendar
UNITED STATES OF AMERICA,
Plaintiff-Appellant,
VERSUS
ARNOLD KATZ,
Defendant-Appellee.
Appeal from the United States District Court
for the Western District of Louisiana
June 14, 1999
Before DAVIS, DUHÉ and PARKER, Circuit Judges.
PER CURIAM:
The Government brings this interlocutory appeal of a pretrial
ruling excluding evidence in a criminal prosecution that charged
Arnold Katz (“Katz”) with violation of 18 U.S.C. § 2252(a)(2),
receipt of child pornography. We affirm.
I. FACTS AND PROCEDURAL HISTORY
The government alleges that the following facts will be proven
at trial.
1
On November 23, 1994, Katz posted a message on an Internet
bulletin board, stating that he had homemade “pornos” and was
interested in trading with others. An undercover customs agent
responded and arranged to exchange videos with Katz. On April 7,
1995, agents executed a controlled delivery of a package containing
a videotape entitled “Masturbating Lolita” and a computer disk
containing eleven Graphic Image Files (“GIFs”) to Katz at his
residence, which became the subject of Count II (receipt of child
pornography). The Government also seized a videotape entitled
“Dream Teens,” that Katz sent to the undercover agent which became
the subject of Count I (distribution of child pornography).1
At issue is whether the government’s evidence is sufficiently
reliable that a jury could conclude beyond a reasonable doubt that
the models depicted in the evidence were less than 18 years old at
the time the images were produced. Katz filed a Daubert motion
“pursuant to Federal Rules of Evidence 403 and 702, to
exclude all expert witness testimony purporting to
determine the age of the persons portrayed in the
[evidence] upon the basis of the application of the
“Tanner Scale” to review of a visual depiction. The
application of the “Tanner Scale” to a visual depiction
for the purpose of determining the age of the person
depicted is not valid and reliable scientific methodology
and does not comport with the requirements of evidentiary
reliability articulated by the Supreme Court in Daubert
v. Merrell Dow Pharmaceuticals, Inc., [509 U.S.
579](1993). The accused moves for a Daubert hearing on
this issue pursuant to Federal Rules of Evidence 104(a)
1
Count I was dismissed on December 2, 1997. Count II, receiving
child pornography through the mail, is the only currently pending
count.
2
and 702.
The trial court set a Daubert hearing on the motion on March 11,
1997,2 wherein it was developed that the Tanner Scale of Human
Development for females is the recognized scientific test utilized
for determining the age of postpubescent Caucasian females and
consists of separately rating, on scales of 1 to 5, breast
development and pubic hair development, with Stage 1 being pre-
adolescent and Stage 5 being adult. However, the government’s
expert witness testified that he could not use the Tanner Scale
breast development scale for determining the age of the models in
question because the age bands were too wide. For instance, the
Tanner stage 5 breast development band encompasses ages 12 through
19. Further, the Tanner Scale is valid as to Caucasians, but it is
not valid as to all ethnic groups. After hearing testimony, the
parties stipulated and the district court found that the Tanner
Scale has been subject to peer review and publication, that it is
a scientifically valid methodology for determining the age of
individuals, and that the Government’s expert, Dr. Woodling, was
qualified to perform Tanner Scale analysis. Whether the Tanner
Scale analysis could be adequately performed on the images in
evidence remained in dispute. At the close of the hearing, the
district court concluded there was sufficient ability to visualize
2
The district court considered evidence from Counts 1 and 2 at
the March Daubert hearing, as Count 1 had not yet been dismissed.
3
the Tanner Scale criteria to permit the expert to express a
reliable opinion whether the models were less than 18 years old and
preliminarily determined that the videotape and the expert witness
testimony were admissible.
A second hearing was conducted on December 1-4, 1997,
immediately prior to the scheduled trial, to resolve all remaining
evidentiary issues. The district court reaffirmed that the
videotape and government’s expert testimony were admissible, which
ruling is not challenged in this interlocutory appeal. The
government brings this appeal challenging two district court
rulings relating to the inadmissibility of the GIF files.
On the evening of December 1, 1997, the government turned over
to defendant a computer disc containing the GIFs. The government
chose five of the eleven GIF images from the computer disk to
introduce at trial and at the second hearing, labeling them 1-A, 1-
B, 1-C, 1-D, 1-E. Katz objected to the admission of the five
color “photos”3 from the GIF files which the government proposed
using as exhibits because the government had provided only poor
quality black and white versions of these images to the defense
during discovery. The district court ruled that, as a sanction for
failure to timely disclose the color images to the defendant, those
3
The images printed from the computer disk are referred to in the
record as “photos.” The color “photos” were produced by printing
the GIFs from a computer with a color printer. The black and white
images were produced by copying the computer generated color images
on an ordinary office copier.
4
images would not be admissible. However, for purposes of the
pretrial hearing, the district court permitted the government to
use a set of “better quality black and white photos” in place of
the poorer quality images originally turned over to defense counsel
and actually utilized the color versions at various times during
the lengthy hearing.
In its rulings at the close of the December 1997 hearing, the
district court enmeshed its Daubert analysis with a Federal Rule of
Evidence 403 weighing of probative value against potential for
prejudicial effect. After considering the GIF images and the
testimony of the government’s expert, the district court concluded
that the black and white images were inadmissible at trial pursuant
to Federal Rule of Evidence 403 because they lacked sufficient
clarity to determine the models’ ages under the Tanner Scale and
therefore their probative value was outweighed by their prejudicial
effect. Specifically, Dr. Woodling was unable to apply the Tanner
Scale pubic hair analysis to 1-A because the poor quality of the
photo precluded him from determining whether any of the model’s
pubic hair had been removed. Participants in the production of
child pornography may manipulate the appearance of a model’s pubic
hair to make an older model look younger, thus impacting on the
validity of the Tanner pubic hair development scale. The ethnicity
of the model in 1-B was uncertain, and the district court held that
the scientific methodology of the Tanner Scale was not sufficiently
5
verified on non-Caucasian individuals. The district court found
that the poor quality of the images and the models’ position in 1-
C, 1-D and 1-E precluded the application of Tanner Scale pubic hair
analysis. It is difficult to determine from the record which set
of images some of the expert’s testimony referred to. Regardless,
the expert was asked several times whether his testimony would
change if he were to base it on the excluded color photos. He
testified that it would not. As to one of the images in 1-E, the
district court noted that the amount of pubic hair appeared quite
different depending on whether the court viewed the black and white
image or color image, and concluded that this discrepancy
illustrated the lack of reliability of the images in depicting the
actual appearance of the person shown. In summary the district
court concluded that problems with visibility attributed to the
angles of the photos and the quality of the prints precluded
utilization of the Tanner scale and thereby greatly reduced the
probity of the exhibits. Probity and reliability became
inextricably linked which, when balanced against prejudice, tilted
the scale toward excluding not only Dr. Woodling’s opinion
testimony relating to the GIFs but the exhibits themselves.
The district court granted the government’s motion to stay the
trial pending the interlocutory appeal of orders excluding the GIF
images.
We would be remiss if we did not note that we are troubled by
6
the amount of judicial resources that were devoted to the Daubert
hearing. In a case capable of being tried start to finish in a day
and one half, not only the court but the lawyers were engaged for
the better part of five days in a hearing to determine the
reliability of testimony and potential prejudice of exhibits
involving a well known test that is applied in a quite
straightforward manner. Daubert hearings in cases much more
complex than this one are customarily conducted with dispatch
consuming only a few hours at best.
II. DISCUSSION
A. SANCTION FOR LATE DISCLOSURE OF EVIDENCE
The government challenges the district court’s ruling
excluding the color versions of the GIF images. The district court
found that the government’s failure to disclose the “photographs”
to the defendant in the identical form it intended to produce them
at trial was either an attempt to “sandbag” the defense or highly
unprofessional conduct and therefore limited the government to the
use of black and white images.
We review remedies for discovery violations imposed by a
district court for abuse of discretion. See United States v.
Bentley, 875 F.2d 1114, 1118 (5th Cir. 1989). In exercising its
discretion, the district court “should consider factors such as the
reasons why disclosure was not made, the prejudice to the opposing
party, the feasibility of rectifying that prejudice by granting a
7
continuance, and other relevant circumstances.” Id. The district
court “should impose the least severe sanction that will accomplish
the desired result – prompt and full compliance with the court’s
discovery orders.” See United States v. Sarcinelli, 667 F.2d 5, 7
(5th Cir. Unit B, 1982).
First, the government reiterates its explanation for the delay
presented at the hearing. Prior to the return of the superseding
indictment, the black and white copies had been given to defense
counsel as potential Rule 404(b) evidence. After the superseding
indictment, which elevated the GIFs to intrinsic rather than
extrinsic evidence, defense counsel never requested better images
or copies of the computer disc. The district court rejected this
explanation, finding instead that the reason disclosure was not
made was the government’s attempt to sandbag the defense or highly
unprofessional conduct. This finding is not clearly erroneous.
Second, the district court made repeated inquiry into whether
its order would result in prejudice to the government by asking the
government’s expert whether his testimony would be different if he
were to base his answers on the color photos rather than the black
and white photos under consideration at the pretrial hearing. The
expert testified repeatedly that it would not.
Third, although the district court made no specific findings
on this factor, potential for prejudice to the defendant was high
because of the unique circumstances of this case: the computer disc
8
had never been in Katz’s possession, so the defendant had no
information about the contents of the images other than what he
learned during discovery. Further, Katz’s defense was premised on
expert testimony concerning the age of the models and it was
necessary for his expert to examine the evidence and formulate an
opinion prior to trial.
All three of these factors weigh in favor of affirming the
district court’s ruling. However, the government argues that a
continuance would have been an appropriate and less severe sanction
than exclusion of the evidence. See Sarcinelli, 667 F.2d at 7.
Given the findings concerning the government’s motives, which are
not clearly erroneous, the testimony of the government’s expert
that viewing the excluded “photos” would not have changed his
answer to the questions posed at the pretrial hearing, the high
potential for prejudice to the defendant and the ruling that the
“photos” are inadmissible on alternative grounds, see infra, we
hold that the district court did not abuse its discretion in
excluding the evidence.
B. EXCLUSION OF EVIDENCE UNDER RULE 403
Federal Rule of Evidence 403 provides:
Although relevant, evidence may be excluded if its
probative value is substantially outweighed by the danger
of unfair prejudice, confusion of the issues, or
misleading the jury . . . .
We review district court rulings excluding evidence for abuse
of discretion. See United States v. Pace, 10 F.3d 1106, 1115 (5th
9
Cir. 1993). The district court granted Katz’s motion in limine
excluding the five GIF images, concluding that the government’s
expert was unable to render a reliable opinion as to the age of the
individuals depicted.
Implicit in the district court’s ruling is the finding that
the age of the models had to be determined by expert testimony.
The ruling was no doubt influenced by the government’s position
embracing the need for, and advocating the admissibility of, expert
testimony on this issue. On appeal, the government changes its
position and argues that a lay jury could determine the age of the
post-puberty models without any assistance from its own expert,
citing United States v. Lamb, 945 F. Supp. 441 (N.D.N.Y.
1996)(declining to require the government to prove the age of the
persons depicted by expert testimony); United States v. Gallo, 846
F.2d 74 (4th Cir. 1998)(table), 1988 WL 46293 at 4 (“expert
testimony as to age, while perhaps helpful in some cases, is
certainly not required as a matter of course.”); United States v.
Villard, 700 F. Supp. 803, 814 (D.N.J. 1988)(noting that “the jury
can examine the photographs in question and determine for itself
whether the individual is under eighteen years of age.”)
The threshold question – whether the age of a model in a child
pornography prosecution can be determined by a lay jury without the
assistance of expert testimony – must be determined on a case by
case basis. As the government correctly points out, it is
10
sometimes possible for the fact finder to decide the issue of age
in a child pornography case without hearing any expert testimony.
See United States v. O’Malley, 854 F.2d 1085 (8th Cir.
1988)(defendant’s letters describing the models in the pictures as
a “twelve-year-old girl” and “younger than [nine],” combined with
the pictures themselves, sufficient to sustain a child pornography
conviction). However, in other cases, the parties have been
allowed to present conflicting expert testimony. See United States
v. Anderton, 136 F.3d 747, 750 (11th Cir. 1998)(Government’s
expert, a medical doctor with expertise in adolescent growth and
development, testified that the models were between eleven and
fifteen and a half. Defendant’s expert, a clinical psychologist
and sex therapist, testified that the ages of the models could not
be determined.) In yet other cases, one party presents expert
testimony, while the other does not. See United States v. Broyles,
37 F.3d 1314, 1316 (8th Cir. 1994)(Government presented the expert
testimony of a pediatric endocrinologist and Broyles presented no
evidence.) A case by case analysis will encounter some images in
which the models are prepubescent children who are so obviously
less than 18 years old that expert testimony is not necessary or
helpful to the fact finder. On the other hand, some cases will be
based on images of models of sufficient maturity that there is no
need for expert testimony. However, in this case, in which the
government must prove that a model, who is post-puberty but
11
appears quite young, is less than eighteen years old, expert
testimony may well be necessary to “assist the trier of fact to
understand the evidence or to determine a fact in issue.” Fed. R.
Evid. 702.
In addition, the government argues the district court erred in
limiting its expert to opinions based on the Tanner Scale pubic
hair development. Although the expert testified that he could not
perform the Tanner Scale pubic hair analysis on the GIF images, he
was willing to give an opinion concerning the models’ ages based on
breast development and general body habitus (the body’s shape, size
and distribution of body fat). However, the expert also testified
that the Tanner Scale breast stages are not scientifically useful
in determining the age of the models because the range of ages for
each stage was too broad and extended beyond the age of 18. The
district court did not abuse its discretion in determining that the
probative value of the images was compromised by the inability of
the government’s expert to determine the age of the models using
the portion of the Tanner scale which his own testimony advocated
as scientifically valid and reliable. We do not mean to imply
that, as a matter of law, the Tanner Scale pubic hair development
scale is the only reliable basis for judging the age of models in
child pornography cases. Rather, we hold only that a fair reading
of the extensive record in this case reveals that Dr. Woodling’s
expertise was linked to Tanner scale methodology, and that the
district court did not abuse its gate-keeping function in limiting
12
his opinion testimony to that methodology. However, the government
is not precluded from attempting to persuade the district court
that some other witness can express a reliable opinion concerning
the age of the models using scientifically valid methodology that
is not dependent on the Tanner Scale.
Finally, the government argues that because the GIF images are
res gestae, they are particularly probative, and the district court
erred in performing the weighing task required under Rule 403. The
indictment alleges violation of 18 U.S.C. § 2252(a)(2), which
requires proof that defendant knowingly and intentionally received
visual depictions of children under the age of eighteen engaged in
sexually explicit conduct. The government argues that the evidence
is probative of elements of the crime charged other than the age of
the models, specifically, that the defendant received depiction of
individuals “engaged in sexually explicit conduct.” Because that
element is not disputed and because the district court ruled
admissible a videotape that fulfills the government’s burden on
that element, we cannot say that the district court abused its
discretion in rejecting the government’s argument that the res
gestae nature of the images in question gave overwhelming weight to
their probative value in spite of their prejudicial nature. See
Campbell v. Keystone Aerial Surveys, Inc., 138 F.3d 996, 1004 (5th
Cir. 1998)(When the probative value of evidence is tenuous and the
risk of prejudice substantial, the district court does not abuse
13
its discretion in excluding the evidence.)
In sum, it was not an abuse of discretion to exclude images
that, according to the government’s own expert, depict models whose
ages are not susceptible to evaluation using the scale that the
same expert advocates as scientifically reliable.
III. CONCLUSION
Based on the foregoing, we affirm the district court’s
exclusion of the GIF images.
AFFIRMED.
14