Such is the summary judgment the Post passes on its competitor. The article blasts NBC news for its failure to mention that GE, NBC's parent company, "paid exactly zero dollars in federal taxes" despite earning $14.2 billion in profits.
Yep, there's nothing worse than missing stories.
Speaking of which ...
On March 16, the Post published an article on the popularity in the GOP base of various Republican political figures. I wrote to the Post, pointing out that the opinion poll's methodology was outdated and "fatally flawed."
In the spirit of the "accountability journalism" The Washington Post so deeply cares about, it refused to print my piece.
Was the Post right? Here's what I wrote. You be the judge.
At the moment, the only thing with bigger ground problems than Sarah Palin is The Washington Post's March 16 article, "Palin loses ground among GOP base."
The article is groundless because its opinion poll used outdated polling techniques; baseless -- literally -- because the poll failed to identify adequately the GOP base, the focus group of the article.
The poll was conducted nationwide of people who have conventional and cellular phones. Of course, not every voter has a telephone -- notably Indians on reservations. Also, bored teenagers with a sense of humor will readily claim over the phone to be true blue voters. But I digress.
The Post article presented survey findings gleaned from 414 Republicans and "GOP-leaning Independents." What that means: from the proverbial getgo, the sample was contaminated. The last time I looked, Independents are not Republicans, hence they cannot be part of the "GOP base." If you wish to include nonGOP registered voters in some sort of functional definition of "GOP base," then you must include GOP-leaning Democrats. Such people exist; in fact, they were decisive in more than one election.
You Doubting Thomases out there asking why the Post poll included one group but not the other: I hope you find solace in Aristotle's words that are as timeless as they are inconvenient: "Wonder is the beginning of philosophy."
I submit that the poll did not include GOP-leaning Democrats because the Post is not interested in some functional, abstract or emotive "GOP base." The fact that the poll made GOP political figures (Huckabee, Romney, Palin, Gingrich) the topics of its questionnaire indicates what the Post really wants to know: who is likely to win the next GOP presidential primary?
The question is crucial; billions of dollars are riding on the outcome. Unfortunately, because the poll's methodology is fatally flawed, its road to the answer is paved with double and triple-bottomed boxes:
For starters, how did the poll identify "Republicans"? Answer: It let people on the telephone identify themselves. Don't look now, but the poll just fell into a double-bottomed box. Not only are some people mistaken or lie about their voter registration, others are not even registered to vote. NO PROBLEM: the Post poll blissfully went its merry way and interviewed them even though they were not members of the right population, i.e., the GOP base.
Sorry, I made a mistake -- the box is triple-bottomed. Many people interviewed over the phone in fact are registered Republicans but they will not vote in the upcoming GOP presidential primary. You don't need to read "Hamlet" to know there's the rub ...
To give you folks at home an idea of what's at stake, let's assume the turnout of GOP-registered voters in their upcoming presidential primary will be 50%. If the Post sample were randomly drawn from the right group, i.e., GOP registered voters -- which it was not -- then we could reasonably expect that 207 of the 414 Republicans interviewed would be nonvoters. Stated crudely, fully half of the interviews would be worthless for predicting the GOP primary.
Or would they? What difference does it make? Well, a trainload. GOP primary voters are, among other things, older than GOP registered voters in general. Older people go to bed earlier, have more physical problems answering a telephone questionnaire, etc. Result: they tend to be underrepresented in a random telephone poll. Such realities are why polls of primaries are notoriously less accurate than polls of general elections. Polls of city elections are even more problematic: city election voters tend to be much older and have far longer residency than either the general population or the registered voter population.
In drawing a sample of the GOP base, how can you be sure that someone is (1) registered to vote, (2) registered with the GOP, and (3) here comes the hard part -- will likely vote in the upcoming GOP presidential primary? How, in short, can the Post be sure next time it polls the right population?
Here we arrive at a sampling technique that most polling firms don't know and don't want to learn. Their reluctance is understandable: the technique will double their workload, triple their accuracy -- but halve their profits.
As with most magic tricks, the solution to the enigma of identifying real voters is hidden in plain sight ...
When you vote, what happens? In most places, after you give your name, precinct officials hurriedly search through a document -- a voter roster. The roster is a computerized list of all the registered voters in your precinct. Next, you sign the roster in the blank space beside your name, then step inside the voting booth. Unknown to the general public (and to most political consultants) voter rosters are usually public information retained by Secretaries of State for a legally specified number of years. If you go into the state archives, dig out the rosters, and draw a sample entirely from people who definitely voted in a past GOP primary -- you know they voted by their signatures -- guess what happens. The cause of the astounding accuracy of your poll is simple: the right population has finally, truly, really been sampled. (Sidebar: the Post poll claims a level of accuracy of plus-or-minus five percentage points. Such figures always accompany polls; they exude an odor of science. In truth, those numbers are blue smoke and mirrors because they assume the right population was sampled in the first place.)
To be sure, new people move into a state and minors turn voting age; such people obviously are not on the voter rosters in the state archives. For those experienced in working with rosters, however, such groups do not pose any significant problem; simply put, there are ways of weighting them.
When using actual, real, live voters -- not telephone chatter -- as the source of a poll sample, something intuitively obvious but surprising takes place: the number of undecided/no opinion drops drastically. That outcome has already been explained: polls done the outmoded way of randomly surveying people on the telephone invariably end up interviewing people who are not voters, hence do not care, hence have "no opinion." In that regard, the Post poll findings that 23% of the people interviewed had "no opinion" of Mike Huckabee, 19% had no opinion of Mitt Romney, etc., are as typical as they are predictable.
Blindly polling telephone talkers definitely skews polls. (A case in point: on Indian reservations, telephones belong disproportionately to better-off, more conservative voters). Given such undeniable biases, many readers will condemn the Post poll as corrupt. Others will conclude it was incompetent.
Who is right?
One answer: corruption or incompetence, it makes no difference. In news about our federal government and officials, The Washington Post leads the nation. Its readers and editors deserve better.
Well, there you have it, Dear Reader. If you ever wondered about what is not being printed -- what sort of thing the powers-that-be are forbidding you from seeing -- you now have a concrete case.
Was my missing article too long? Too complex for the "average reader" to understand? One thing I learned after decades of writing: when it comes to why an article or book was not published -- that is to say, why it was censored -- THERE WILL ALWAYS BE A REASON.
Which is why I don't waste time with reasons.
If The Washington Post truly wants to be the moral guardian of journalism in America -- if not worldwide -- it's easy. All the Post has to do is investigate itself.