Pubdate: Mon, 18 Sep 2000
Source: Insight Magazine (US)
Website: http://www.insightmag.com/
Feedback: http://207.238.36.125/feedback/
Address: 3600 New York Ave., NE, Washington, DC 20002
Contact:  Mark Davis

IS THE DRUG CZAR SKIRTING THE LAW?

The head of the Office of National Drug Control Policy is under fire for
manipulating data in a report to Congress to cover shortcomings in his
federal antidrug program.

Bill Clinton’s drug czar, Barry McCaffrey, has had no shortage of trouble
recently. First, he provoked outrage by paying TV producers to let him edit
scripts to promote the White House Office of National Drug Control Policy’s,
or ONDCP’s, antidrug message. Then he got into even more trouble when Salon,
a liberal Internet magazine, discovered that he was paying publishers to run
antidrug editorials. And McCaffrey’s problems only got worse when it was
discovered that his office’s Website was allowing advertisers to store
“cookies” on visitors’ computers, potentially allowing advertisers to track
what other Websites they visit.

Now McCaffrey’s office is in more hot water. Insight has discovered that
ONDCP manipulated data in a formal report to deceive Congress, a likely
violation of federal law. The move concealed from congressional budget
makers the shortcomings of ONDCP’s $195 million per year media campaign to
promote antidrug awareness. The campaign is the only program ONDCP directly
manages.

The doctored document, “Performance Measures of Effectiveness: 2000 Report,”
is supposed to fulfill ONDCP’s obligations under Public Law No. 105-277,
which seeks to reduce waste by requiring ONDCP to set “quantifiable and
measurable” goals and then file annual reports with Congress detailing
progress. If goals are not met, Congress can eliminate programs that don’t
work.

To this end, ONDCP includes in its report a section called “Progress at a
Glance,” two pages that color code each of its goals: green for goals that
are “on target” and red for those “off target.”

It’s here that the deception began. In the section describing its media
campaign, ONDCP listed its goal to “increase the percentage of youth who
perceive drug use as harmful” in green. This meant that it was on target to
increase the proportion of American youth who see drug use as harmful to 80
percent by 2002. But was ONDCP really on target?

In 1996 only 59.9 percent of 12th-graders — the group that ONDCP used to
measure youth perception — saw regular marijuana use as harmful. By 1998,
the proportion had fallen to 58.5 percent and, by 1999 it was at 57.4
percent.

For three consecutive years, risk perception among 12th-graders fell below
the 1996 levels. Given these bleak numbers, ONDCP should have marked this
goal in red, indicating that it was not on target. To get the numbers to
come out right, officials made two changes (which they did not point out in
the report) in the way they calculated risk perception. First, they changed
a figure called the “base year,” from which they judged progress. The
original base year was 1996; the new one was 1998.

By making the change, they accomplished two things. First, they made the
downward trend — one that might have jeopardized funding for their program —
look less severe. Second, since the most recent data came from 1999, a 1998
base year left less room for long-term comparison. Using a 1996 base year
would show a trend for three years: 1997, 1998 and 1999. But by starting in
1998, officials could now report that, although there had been a decline, it
had been only for one year.

But even this more modest decline never made it into the report. That’s
because McCaffrey’s officials made a second change: They started to use
eighth-grade data instead of the 12th-grade data. Conveniently,
eighth-graders typically see drug use as more risky than 12th-graders,
hovering in the lower 70 percent range compared to the upper 50 percent
range for 12th-graders. By changing the data source, officials found they
could boast that they were less than 7 percent from their 2002 goal,
compared to the 23 percent shortfall with the original 12th-grade data.

ONDCP officials cooking the report to Congress now waived their magic wand a
third time by misreporting data which they obtained from a University of
Michigan study called “Monitoring the Future.” The numbers they reported
from the study should have indicated 73.3 percent of eighth-graders saw
regular marijuana use as harmful. Officials tripled their improvement by
reporting a higher number.

Could these have been well-intentioned mistakes? When Insight confronted
ONDCP officials about the changes, they refused to speak on the record. On
background, however, officials sheepishly defended themselves. “There was no
intention to hide anything,” says one official who spoke only on condition
of anonymity. “This does not say anything about ONDCP.”

The source speculates: “The change in base year could be a misunderstanding.
The graph should have started with 1996. I can’t imagine why that happened.
It’s a very large enterprise to put together.”

And what of the improvement that ONDCP tripled when it reported data to
Congress? “That was a typo,” the source insists.

But officials were less eager to admit error when it came to changing
12th-grade data to eighth-grade data. Those changes, they claimed, were
appropriate because they said the media campaign targets middle-school
children. One official, who was in charge of data for the report, says: “We
weren’t trying to pull anything sneaky here.”

The evidence seems to contradict these claims. Federal law requires ONDCP to
point out any changes it makes to the reporting system, and the 2000 annual
report did note other legitimate changes. If officials weren’t trying to
hide their failure, why didn’t they note these changes?

An official who spoke on the condition that his name not be used responded
that the report’s authors simply forgot. “We missed that,” he says. “It’s a
200- or 300-page report. We missed that one, so shoot me.”

Besides, he continues, “This to me was a very minor change to make because I
’m also aware of what goes on in the media campaign.”

This official’s supervisor, the first source we quoted, tried to back him
up. “It was an oversight,” the supervisor first told Insight. But later she
indicated that officials might have chosen not to document a “minor” change
such as this. “Changes are indicated only if it’s a major change,” the
supervisor says. “This was only a minor change.”

But that’s not what the law requires. Public Law No. 105-277, which
appropriates money for ONDCP’s media campaign, requires “a description of
any modifications to the performance-measurement system.” That certainly
seems clear enough.

And the opinion that this is a “very minor change” doesn’t seem to be
universal. Insight contacted the office of Republican Rep. Jim Kolbe of
Arizona, a member of the Task Force for a Drug Free America and chairman of
the House Appropriations subcommittee that allocates funding for ONDCP.
According to Kolbe spokeswoman Fran McNaught, switching data samples in a
report that attempts to measure progress is deceiving: “We’re looking at
tracking true performance, so if you give us apples (one set of data) one
year and oranges (another set) the next, that’s not an effective tool!”

And what of ONDCP’s claim that the changes were appropriate? If these
changes were legitimate, there should be someone willing to take
responsibility for the changes and explain why they approved them. So
Insight asked ONDCP who made the decisions. Several officials blamed an
“interagency working group.” One said directly, “That was a decision that
the interagency working group made and that we went along with.” When
Insight pressed for names of the members of this working group, the source
referred us to his supervisor.

But, according to the supervisor, the decision was “not officially brought
up at the working group.”

When Insight confronted the original official about the discrepancy, he
changed his story, saying, “I don’t recall that I took this to [the working
group] at all. This to me was a very minor change to make because I’m also
aware of what goes on in the media campaign. I’m aware that their target is
the ‘tweens’ (middle-school children). I said, ‘You know we’re not getting
an accurate measure.’ I said, ‘eighth-graders are a better measure.’”

Each time Insight talked with this official, the story changed. But always
there were discrepancies.

While it is true that many ONDCP commercials target middle-school children,
the media campaign, according to ONDCP’s Website and brochures, targets
“youth ages 9 to 18.” Given this, why would eighth-grade data be better? Is
it a better indicator of whether children grow up to become addicted to
drugs? Did it better reflect the cost of drug use to society? “I couldn’t
give you any definite correlation,” he says.

Insight asked some other experts about the differences, just to be sure. A
second ONDCP official, who had not been involved in the decision to change
data, says, “Eighth-grade data simply does not reliably portend future
12th-grade drug-use patterns. And 12th-grade drug-use patterns are the best,
if imperfect, indicator of the future burden drug abusers will impose upon
society.”

Lloyd Johnston, a professor at the University of Michigan who conducts the
“Monitoring the Future” study, agrees that it would be inaccurate to call
one set of these data “more accurate” than another. “I think they’re
probably both important,” he says, although he also indicates that he might
expect eighth-grade data to change first. “The eighth-graders, it appears,
are in a sense the first to respond to new influence in the environment,” he
explains.

Insight also asked him whether the eighth-grade data were a reliable
predictor of the future cost to society — whether eighth-grade attitudes
strongly correlate to drug addiction later in life — since this is what
ONDCP is tasked to reduce in the long run. “I don’t know if I can make that
connection,” Johnston says. But, he responds, “We’ve done a lot of our
analyses on the perceived risk on 12th-graders and can show pretty
conclusively that that’s one of the determinants of whether they use or
continue using.”

Either way, the most important question remains unanswered. If ONDCP’s
changes were legitimate, why did officials not report them to Congress, and
why did they refuse to talk about them on the record? Congress may be
getting ready to ask those questions.
- ---
MAP posted-by: Don Beck