Reuters' Global Megaphone Slurs America

In the lead—up to the fifth anniversary of 9/11, a glut of news stories appeared decrying the increase of alleged bias against Muslims in America since that tragic day. One such article that has been reprinted in dozens of papers all over the country and in prominent media outlets overseas was an August 25th article by Reuters reporter Ellen Wulfhorst, 'U.S. Wages of Arab, Muslim men fell after 9/11 — study.'

The story references an unpublished study, 'Backlash: Effects of 9/11 on Immigrants, Muslims and Arabs Living in the U.S.', by Neeraj Kaushal of Columbia University, Robert Kaestner of the University of Illinois—Chicago, and Cordelia Reimers of Hunter College, the City University of New York. The study isn't scheduled for publication until the Spring 2007 edition of the Journal of Human Resources.

According to Wulfhorst's article, the study finds that

'the earnings of Arab and Muslim men working in the United States dropped about 10 percent in the years following the 9/11 attacks, according to a new study. The drop in wages was most dramatic in areas that reported high rates of hate crimes...'

Rhe picture painted for Reuters' worldwide client base is one of an America seething with ethnic and religious hatred, a country which denies opportunities based on anger over 9/11.

Predictably, the usual Islamic victimization—mongers, the Council on American—Islamic Relations (CAIR) and the Muslim American Society, immediately began citing its findings as yet even more proof of the great American anti—Muslim conspiracy.

Flawed Reporting

One curious element to the Reuters article is that entire sections are taken verbatim from a University of Illinois—Chicago press release announcing the study's findings issued a few days before the Reuters article appeared. For instance, compare these two paragraphs, the first taken from the UIC press release; the second from Wulfhorst's Reuters article:

[UIC] The study measured changes in wages of first— and second—generation immigrants from countries with predominantly Arab or Muslim populations between September 1997 and September 2005 and compared them to changes in wages of first— and second—generation immigrants with similar skills from other countries.

[Reuters] The study measured changes in wages of first— and second—generation immigrants, from countries with predominantly Arab or Muslim populations from September 1997 to September 2005. It then compared them to changes in the wages of immigrants with similar skills from other countries.

A quick review of both sources indicates that these nearly identical paragraphs are only one example of several that could be cited. Needless to say, in academia this would be called plagiarism; at Reuters, evidently it is considered reporting.

If Ms. Wulfhorst had actually researched the article, rather than cutting and pasting it together from the press release issued by one of the study author's university public affairs office, she might have discovered that there are problematic elements to the methodology of the study that can be discerned just from the limited information released thus far.

A Flawed Study

Perhaps the most glaring flaw of the present study, which attempts to link the prevalence of hate crimes with the decrease in American Muslim wages, is that according to the UIC press release it uses only one year's worth of hate crime data — 2001. Since the authors used wage data up to September 2005, it is more than odd that the readily available FBI hate crime data for the years 2002—2004 — three additional years worth — was left out of the study altogether. This can hardly be accidental.

The most reasonable explanation for this oversight is that after 2001, anti—Muslim hate crimes plummeted rapidly and has stayed at comparatively low levels ever since, as the FBI figures demonstrate:

2001 481 incidents

2002 155 incidents

2003 149 incidents

2004 156 incidents

By using this one year's stats (2001) to the exclusion of additional and more recent hate crime data automatically calls into question the authors' findings. In statistical jargon, the 2001 FBI hate crime figure is called an outlier. The singular use of this data point severely skews the study and makes the conclusions highly questionable.

Despite the study's unavailability (when I contacted the UIC Office of Public Affairs to obtain a copy of the study, I was told they wouldn't provide copies of the study to the media until it is published next year; an email to Dr. Kaestner requesting a copy of the study went unanswered), there are still several other obvious methodological problems with the findings:

 The wages of Muslims were compared to other immigrants rather than the general population. The reason is, of course, that American Muslim households have higher median incomes than the general population, and any overall comparisons would have shown that despite any decreases in wages, American Muslims are still doing quite well financially after 9/11 — a fact that poses a prima facie challenge to the study's findings of widespread anti—Muslim bias.

 No mention is ever made of the fact that wages for declined across the board post—9/11, and no comparison is apparently ever made to this overall decline in wages due to the post—9/11 economic slowdown.

 The study relies on a ridiculously small sample size of 4,300 out of the millions of Muslims living in America, and the source of this sample isn't identified.

 One statistical correlation discerned from the FBI hate crime data but apparently overlooked by the study authors is that most of the states with the most incidents of anti—Muslim hate crimes (California, New York, New Jersey, Michigan, Illinois, etc.) are dominated by liberal/Democratic administrations and/or legislatures. In fact, one of the most politically liberal areas in the country — Washington D.C. — has one of the highest per capita rates of anti—Muslim incidents in America.

Again, the only thing anyone has to go on regarding the study is the very limited information released by the study authors thus far, but this intentional data cherry—picking should raise immediate suspicions of bias by the study authors.

Agenda—based Research?

In fact, the primary contributor, Neeraj Kaushal, has written scathing denunciations of the Bush Administration and the war in Iraq (which she labels 'the Iraqi quagmire') in her capacity as a columnist for The Economic Times of  India. In a September 2005 editorial, 'Time for the US to leave Iraq,'  she characterized the administration's response to critics of the war in Iraq thus:

'So far President Bush has responded by questioning the patriotism of his critics; vice—president Cheney has gone a step further by calling them dishonest, reprehensible, shameless and corrupt.'

After hailing members of the Congressional Cut—and—Run Caucus, particularly Democratic Congressman John Murtha (who she is obliged to note is 'a 73—year—old Marine veteran, who just recently visited Iraq'), she concluded her editorial with an indictment of the Bush Administration:

'The US needs a leadership that understands the threat of terrorism, has the moral authority to get the support of the US public to proceed in this war and get involved in a constructive effort to build Iraq. The present leadership clearly lacks all this.'

Yep, no political bias here, folks.

Might it be the case that an academic with a political axe to grind against the Bush Administration would jerry—rig the data to ensure a finding that anti—Muslim bias has increased during President Bush's tenure? Just how much the authors manipulated the data will remain to be seen until the study, the data, and the methodology they used are finally published next year, but even someone with ten weeks study in an undergraduate social science statistics class is able to see that the study is fatally flawed.

(In an interesting aside, despite the fact that Kaushal led the study research team, her co—author Robert Kaestner was the first to publicize the study, which prompted the Reuters article. Mention of Kaushal and Reimers was buried in the last paragraph of the UIC press release and Kaestner was the only one quoted by Reuters.

Perhaps in a bit of media envy, Kaushal rushed out a similar press release a few days after the Reuters piece appeared to publicize her involvement and to belatedly capitalize the attention her colleague and the study were receiving.

Notwithstanding the fact that the study isn't available for critical review until next year, the study's authors decided to publicize their findings during the annual ritual of decrying anti—Muslim bias around the 9/11 anniversary. This 'black box' reporting of the study and its findings is also a sure—fire method for evading the detailed scrutiny critics. Such conduct, while hardly surprising coming from academia these days, is grossly irresponsible and makes a mockery of authentic scientific inquiry that actually occurs in universities from time to time.

But one might think that after the notorious Reuters 'fauxtography' scandal this past summer during the recent Israel—Hezbollah conflict, when Reuters photographers were caught red—handed manipulating images and staging photos, that venerable news service would be doubly—cautious about floating fake news while it is still under increased scrutiny.

That experience, however, apparently didn't give Ms. Wulfhorst a moment's pause before cutting and pasting a story together to fit Reuters' typical ideological—driven narrative, with nothing more to go on than a press release to go on. According to a WizBang blog entry earlier this year, this also isn't the first time that Ms. Wulfhorst has used the findings of a bogus study as the sole source for one of her Reuters articles.

If anything, this episode provides further evidence that most of what passes as news spewing forth from the mainstream media is little more than thinly—veiled propaganda following a very carefully crafted script.

Patrick Poole is an author and researcher. He maintains a blog, Existential Space.