Innumeracy, MSNBC, and the “liberal” media
Posted by metaphorical on 25 June 2007
In one of the worst cases ever of throwing numbers against a wall and then writing them up as news, MSNBC published a “study” last week that purported to show a liberal bias in the media by looking at public campaign-contribution records.
Whether you sample your news feed from ABC or CBS (or, yes, even NBC and MSNBC), whether you prefer Fox News Channel or National Public Radio, The Wall Street Journal or The New Yorker, some of the journalists feeding you are also feeding cash to politicians, parties or political action committees.
MSNBC.com identified 143 journalists who made political contributions from 2004 through the start of the 2008 campaign, according to the public records of the Federal Election Commission. Most of the newsroom checkbooks leaned to the left: 125 journalists gave to Democrats and liberal causes. Only 16 gave to Republicans. Two gave to both parties.
It’s hard to know where to begin to complain about the shoddy analysis here.
The methodology was to look at the Federal Election Commission records, but not any state ones, for job titles and the names of news organizations.
Our first search of the records used job titles: “editor,” “anchor” and so on. Because often no job title is reported, we also searched using the names of news companies. Smaller companies were not checked; for example, we checked only the company names of the 200 largest newspapers, out of more than 1,400 dailies in the nation.
So that’s one possible bias introduced in the data: smaller organizations may be more likely to be outside big cities, and therefore more likely to be Republican or conservative.
Only line editors and reporters were included, not other newspaper staff. It’s well known that while reporters are more likely to vote Democratic, media management are more likely to vote Republican — and to have far more influence over the direction of coverage and editorial policies.
MSNBC also only looked at
Federal candidates, PACs and parties in the records of the Federal Election Commission, not the separate state campaign records
which might introduce another bias.
They ended up with 300 records but only used 143 for reasons I can’t even possibly begin to explain or understand.
Then, with a list of about 300 apparent journalists, we tried to contact them all. The list published here includes only those who either confirmed that they made the donation or did not respond. Many journalists who changed jobs since the donations were not contacted and are not included here.
As it turns out, that’s 300 out of at least 100,000, by MSNBC’s own estimate.
The final list represents a tiny percentage of the working journalists in the nation. Daily newspapers alone employ about 60,000 full-time journalists. Approximately 30,000 work in television news jobs and 10,000 in radio news.
How MSNBC can presume any kind of statistical significance is beyond anyone’s understanding.
(I’m also a little miffed that they don’t include magazine journalists such as myself and the New Yorker’s George Packer, about whose modest campaign contributions they certainly complain loudly enough).
Having a bunch of numbers doesn’t make for an analysis. It’s inexcusable that MSNBC didn’t kill this story and seriously question the basic journalistic competence of its authors.
Finally, there’s the hopelessly vague “Democrats and liberal causes” at the heart of the statistic. One shudders to think what happens statistically when we let MSNBC define which causes are liberal.
For example, it included VoteVets.org, and organization whose goal “is to put Iraq and Afghanistan War Veterans in Congress who are critical of the execution of the war in Iraq and representative of the principles of VoteVets.org PAC in the US Congress.” It’s hard to see that as a necessarily liberal organization.
Looking at its “Board and Advisors”, there’s some people who look like Democrats, but also a number of Republicans, including one member ran for Congress in Indiana as a Republican. Another is an investment banker.
There’s a kind of Orwellian conflation of “liberal” and “Democratic” throughout the “study.” After all, a number of Democratic candidates are pretty conservative, even if “liberal” and “conservative” were useful terms in surveying the political landscape, which they’re not. (Which party is more fiscally conservative, for example? What are we to make of big business’s increasing support for universal health care, traditionally a “liberal” cause?)
There may well be more liberals than conservatives in the ranks of reporters; it’s not my goal—or task—to argue otherwise. It’s much harder to make a case for liberal bias in news reporting itself. Certainly the media rolled over for the Republicans for at least three years after 9/11.
And then there’s the question of what counts as liberal and what counts as, not to put too fine a point on it, merely rational? In a country where 55 percent of the public believes in creationism, how is one to report on attacks on evolution? Is it liberal bias to give short shrift to the for-hire opinions of a tiny minority of fourth-rate scientists who disagree with the near-unanimous consensus on human involvement in climate change?
In fact, we should only be so lucky as to see a bias toward “liberal” rationality in the news media. Certainly, rationality is in fact becoming an increasingly prized commodity in television news. There’s little at CNN as it tries to catch up to Fox, where there’s been none for quite some time. And here MSNBC has shown its lack of regard for anything that might count as the scientific method.
A thank-you to Andy Patrizio for putting me onto the topic (even though as far as I know he and I disagree more or less 100 percent on it).