We continue to appreciate Carl Bialik's column at the WSJournal web site.  Here are some valuable reminders from this week.

THE NUMBERS GUY By CARL BIALIK

Measuring the Child-Porn Trade
April 18, 2006

Unlike, say, the soft-drink or airline industries, the
child-pornography industry doesn't report its annual sales to the
Securities and Exchange Commission.

Yet in a press release¹
ahead of a recent House of Representatives hearing aimed at curbing the
industry, Texas Republican Joe Barton said, “Child pornography is
apparently a multibillion … my staff analysis says $20 billion-a-year
business. Twenty billion dollars.” Some press reports said the figure
applied only to the industry's online segment. The New York Times reported²,
“the sexual exploitation of children on the Internet is a $20 billion
industry that continues to expand in the United States and abroad,”
citing witnesses at the hearing. (The Online Journal's Real Time column³ also quoted the estimate from the hearing.)

My efforts this week to track down the number's source — and free-lance journalist Daniel Radosh's similar quest⁴
on his blog — yielded lots of dead ends. It turns out it can be easier
to enter a big number into the Congressional record, and national press
coverage, than to locate its origin. (Numbers Guy reader Brian Flanagan
suggested I look into the estimate.)

What was Rep. Barton's staff analysis? A spokesman for
the House Energy and Commerce Committee told me the source of the
number was the National Center for Missing and Exploited Children, a
group that advocates for the protection of children. When I first
talked with that group's president, Ernie Allen, he told me that
Standard Chartered bank, which has worked with the NCMEC to cut off
funding to child-porn traffickers, wanted a quantitative analysis of
the problem, so it asked for a measurement from consulting firm
McKinsey & Co.

Mr. Allen faxed me an NCMEC paper that cites the
McKinsey study in placing the child-porn industry at $6 billion in
1999, and $20 billion in 2004.

But a
McKinsey spokesman painted a different picture for me: “The number was
not calculated or generated by McKinsey,” he wrote in an email.
Instead, for a pro bono analysis for Standard Chartered, he said,
McKinsey used a number that appeared in a report⁵
last year by End Child Prostitution, Child Pornography and Trafficking
of Children for Sexual Purposes, an international advocacy group.

But the trail didn't end there: That report, in turn, attributed the number to the Federal Bureau of Investigation, as did a report⁶
last year from the Council of Europe, a Strasbourg, France-based
human-rights watchdog. Both of those reports noted that estimates range
widely, from $3 billion to $20 billion.

FBI spokesman Paul Bresson told me in an email, “The
FBI has not stated the $20 billion figure… . I have asked many people
who would know for sure if we have attached the $20 billion number to
this problem. I have scoured our Web site, too. Nothing!”

I went back to the NCMEC Monday and shared what I
found. In an email response, spokeswoman Joann Donnellan said, “If it
is determined that this ends up not being a reliable statistic, NCMEC
will stop citing McKinsey as the source and will also stop citing a
specific number. Rather, NCMEC will revert to what it has said
previously… that commercial child pornography is a multi-billion dollar
industry.”

This isn't the first number from the NCMEC that struck
me as questionable. The group provided the estimate that one in five
children is sexually solicited online, which appeared in public-service ads⁷ distributed by the Ad Council. The stat has received a fresh round of publicity thanks to donated air time⁸ from MySpace, a site popular with teens. As I wrote⁹
last year, the “one in five” estimate was based on research that was
five years old which only covered children who spent time online. The
survey also used a broad definition of sexual solicitation. Yet the
stat persists. The NCMEC told me¹⁰
last July it hoped to have new research by the end of last year. Now,
spokeswoman Tina Schwartz says the group expects new research to be
released in the next couple of months.

* * *

As Congress debates whether to pass new laws
specifically outlawing online gambling, a recent poll appears to show
that the public is strongly against the legislative effort: Almost 80%
of Americans oppose a ban, according to the survey.

The poll was conducted by well-known polling firm Zogby International on behalf of an online gambling trade group. As I've written¹¹
in the past, such sponsored research warrants extra scrutiny from
readers, though the fact that the poll was commissioned by a
special-interest group isn't by itself a reason to dismiss it.

Still, in this case, it appears that the sponsor of
the poll influenced the way it was conducted, particularly in the way
the questions were phrased. Here's one question: “Many gambling experts
believe that Internet gambling will continue no matter what the
government does to try to stop it. Do you agree or disagree that the
federal government should allocate government resources and spend
taxpayer money trying to stop adult Americans from gambling online?”
Some 77% of respondents disagreed.

Here's another question: “More than 80% of Americans
believe that gambling is a question of personal choice that should not
be interfered with by the government. Do you agree or disagree that the
federal government should stop adult Americans from gambling with
licensed and regulated online sports books and casinos based in other
countries?” You probably won't be surprised to learn that after being
told that most Americans don't want the government to interfere, some
71% of the respondents to this question signaled they, too, were
against a government ban.

The results of the poll were posted on the gambling trade group's Web site¹² and emailed to journalists.

The gambling questions “were fair and balanced, and
gave the respondent appropriate choices,” Fritz Wenzel, spokesman for
Zogby International, told me in an email. (Zogby does many political
polls separate from interest-group-backed research, including polling
on the 2004 presidential race and 2006 gubernatorial and Senate races
for the Online Journal.)

Polling experts disagreed when I showed them the poll.
Cliff Zukin, president of the American Association for Public Opinion
Research, a professional association of pollsters, told me the
questions are “loaded and biased.” Prof. Zukin added that if any of his
students at Rutgers University wrote such questions, “I would fail
them.”

Robert Blumenfeld, an El Paso, Texas-based attorney for the Antiguan Offshore Gaming Association¹³,
told me the trade group paid “less than $10,000” for the poll. The
Antiguan group, which represents more than a dozen online casinos,
drafted the questions with guidance from Zogby, Mr. Blumenfeld said. He
disagreed with the suggestion that the phrasing of the questions might
have influenced the results, but said the group would conduct further
polling. “We're willing to put the question in a way that can't be
subject to any kind of criticism,” Mr. Blumenfeld said.

Mr. Blumenfeld said the group is using the results of the poll in its lobbying efforts to fight an online gambling ban.

It's not unusual for pollsters to conduct polls for
hire. Many pollsters make their reputations with political polling, and
make their money with sponsored polling. Still, Zogby's poll didn't
meet certain standards¹⁴
set by the polling professional association headed by Prof. Zukin,
which say, among other things, that pollsters should ask unbiased
questions.

Zogby International and its chief executive, John
Zogby, are well known in the polling world. Yet Mr. Zogby has at times
lent his firm's credibility to polls conducted for sponsors and filled
with leading questions, as a New Yorker profile¹⁵
in 2004 noted. One poll funded by People for the Ethical Treatment of
Animals asked respondents if they would stop eating meat or dairy
products “if you knew that within days of birth, chickens have their
beaks seared off with a hot blade to keep them from pecking each other
in their overly crowded cages?”

Other Zogby polls addressing gambling have had conflicting results. A poll¹⁶
conducted by Zogby on behalf of the New York Council on Problem
Gambling, in 2004, found that 67% of respondents said that expanding
gambling by the State of New York will definitely or probably increase
the number of people with gambling problems.

The Zogby poll wasn't the only recent survey on online
gambling to include what I'd consider leading questions. A Harris
Interactive online poll¹⁷,
conducted in February and mentioned in several news outlets (including
the Online Journal) found — among other things — that 27% of
respondents strongly agreed with the statement “since there is no
effective way to regulate or control Internet gambling, it should
remain illegal,” and 27% of respondents somewhat agreed.

The phrasing of that question seems to make an
assumption (the impossibility of regulation) that could have influenced
responses. Humphrey Taylor, chairman of the Harris Poll, told me that
the question was deliberately designed to “see how different arguments
played.” He said he wouldn't use the response to that particular
question, which he called “projective,” to determine whether people
support legalizing online gambling. “In any release we do, we are fair
and balanced, but any single projective question may not be,” Mr.
Taylor said, adding that the poll wasn't sponsored.

* * *

Several readers wrote in about my column¹⁸ last week on the number of illegal immigrants in the U.S. Here's an excerpt from one letter:

As a result of my own experience attempting to
estimate the number of workers employed in California seasonal
agricultural work during my Ph.D. research, I feel qualified to doubt
the validity of any estimation method based on the U.S. census of
population. While attempting to use census data from rural California
counties to estimate the accuracy of two different state employment
service reports on agricultural employment, I encountered discrepancies
among the three sources of as much as 300%, with the census data always
being the lowest. Conversations with friends in the urban Chicano
community confirmed my suspicion that illegal residents were
effectively avoiding enumeration. Perhaps the data collection has
improved, but it's doubtful that people who want to avoid government
scrutiny will make themselves available.

Several readers also wrote in about my comment that
scientific notation is likely to be adopted soon after the U.S. adopts
the metric system:

We've lost enough dollars and lives because of
continuing confusion between our systems of inches vs. millimeters and
pints vs. liters, but the idea of mass re-education of the entire
American public and mass retooling of manufacturing is frightening.

I would like to suggest a column on the status of
metric conversion in the U.S. Is there hope or are we going to continue
to bury our heads in 10 tons of sand (which is 20,000 pounds or 10,000
kilograms which, in turn, is 22,000 lbs)?

And finally, I got this letter about an Associated Press article¹⁹ on a Malaysian man who received a $218 trillion phone bill.

Of course it's amusing that the man received an absurdly high phone bill. But the really
funny part, in my opinion, is the AP journalist's analysis: “It wasn't
clear whether the bill was a mistake, or if [the] phone line was used
illegally.” $218 trillion?!? Hmm… I think it's pretty clear.

Write to Carl Bialik at numbersguy@wsj.com²⁰

Corrections & Amplifications:

A letter in an earlier version of this column incorrectly stated the number of pounds, or kilograms, in a ton.