Straight-Line Projections
Many years ago, I remember that the Washingtonian magazine commissioned the writer who goes by the name of "Joe Bob Briggs", and who had made his name in a Texas paper with his reviews of drive-in and B-movies, to review all the Sunday morning political talk shows. It was a hilarious piece of writing, which unfortunately I cannot find online, but I vividly remember, first, his observation that David Gergen bore an uncanny resemblance to The Cat in The Hat, and second, that everything that purported to be a "prediction" on McLaughlin or Meet the Press or Sam and Cokie was nothing more than "a straight-line projection from the present." Not necessarily a humorous observation, but accurate, and not always obvious.
Reading Elizabeth Bumiller's cold assessment this morning of Bush's futile effort to justify the Iraq war reminded me of Joe Bob's second observation.
It's tempting to play the game of "the press is cowed by the right," or "the press is all a bunch of liberals." The fact is that the main bias of the press is toward the assumption that, however things look now, that's how they will remain. For my money, over the last few years, no reporter has been more "in the tank," more slavishly devoted to the conventional wisdom on Bush's genius and Bush's overwhelming political strength than Bumiller. Part of that was the isolation of the bubble, but more important was that straight-line projection: Bush is politically strong, therefore he will remain politically strong.
Now of course, Bush looks ridiculously weak, so the straight-line projection has him going down the tubes. Bumiller's video presentation on Bush is an even more potent example of her shift over to the alternative projection. As a friend in Iraq reminded me a few weeks ago, things are never either as bad as they look when they're bad nor as good as they look when they're good. Under Bush's apparent strength in 2002-4, there were some incipient weaknesses just as his apparent weakness now disguises some political strengths. The press isn't biased toward the right or the left (generally speaking, with some exceptions), but it is biased toward inertia. That's a factor that's worked hugely to the advantage of Bush and the right, and now it will kill them.
Let's hope so.
March 22, 2006 9:29 AM | Reply | Permalink
It shows the kind of businessman he is - spending his capital on a worthless deal...
March 22, 2006 9:41 AM | Reply | Permalink
The right is getting increasingly desperate in making the case that the war is going well. Among the most preposterous is this little gem that I uncovered on the neanderthal site Little Green Footballs:
Can you top that?March 22, 2006 10:44 AM | Reply | Permalink
You got me curious, so I checked CDC mortality statistics for 2002...broken down by age. 20-24 year olds have a mortality rate of 95.7 per 100,000. Oops...guess that's another 'Iraq ain't so bad' rationale out the window.
March 22, 2006 12:27 PM | Reply | Permalink
Actually, BradThe Dad, I can (almost) top this: I found this little gem of misuse-of statistics in a comment thread on <a href=http://www.washingtonmonthly.com/mt/mt-comments.cgi?entry_id=8477 > Kevin Drum's blog </a>:
Take a look at the actual US Military Casualty figures since 1980. If you do the math, you wil find quite a few surpises. First of all, let's compare numbers of US Military personnel that died during the first term of the last four presidents.
George W. Bush . . . . . 5187 (2001-2004)
Bill Clinton . . . . . . .4302 (1993-1996)
George H.W. Bush . . . . 6223 (1989-1992)
Ronald Reagan . . . . . . 9163 (1981-1984)
Which, one suposes is some sort of attempt to either minimize current US casualties, or play the old "Clinton (GHWB, Reagan) did it too" game. However, another commenter quickly corrected this nonsense with another, more salient set of figures:
The casulaty figures include all deaths to Military Personnel. Let's break the numbers down a bit and include only the numbers killed as a result of combat or terrorist attack:
George W. Bush . . . . . . 1157 (2001-2004)
Bill Clinton . . . . . . . . . . . 56 (1993-1996)
George H.W. Bush . . . . . 172 (1989-1992)
Ronald Reagan . . . . . . . 288 (1981-1984)
This doen't even include those killed during combat in 2005-2006.
(spelling errors not mine)
Amazing what can be done with the manipulation of numbers: especially when there is a political point trying to be made!
March 22, 2006 12:27 PM | Reply | Permalink
Joe Bob Briggs...now that's a name I've not heard in years, as Alec Guinness would say. I remember his heyday, when I was in college. A lot of people just didn't get him at all--thought he was a real redneck instead of a Jewish liberal satirizing reactionary sexist assholes. We could sure use someone like him covering the current bunch of clowns.
March 22, 2006 12:32 PM | Reply | Permalink
Yesterday, Limbaugh used nationwide accidental death numbers (204,000) to show how few troops have died in Iraq. Unfortunately for Rush, when you adjust the numbers for population, approximately 88 Soldiers should have died instead of 2300.
He must have taken EJ Dionne's algebra course.
March 22, 2006 12:34 PM | Reply | Permalink
I think you mean Richard Cohen, not Dionne.
March 22, 2006 12:40 PM | Reply | Permalink
Inertia might explain the behavior and "scripts" of certain reporters, but it only scratches the surface of the lunacy in the offices of certain editors, editorial boards and publishers.
Two words: Ben Domenech
March 22, 2006 12:45 PM | Reply | Permalink
Damn. I think you're right. Some days the folks at the WaPo get me all atwitter.
March 22, 2006 12:46 PM | Reply | Permalink
If you let the timeline run long enough the rate for any given 100,000 of population will reach 100,000 deaths, so what does it matter anyway?
March 22, 2006 12:47 PM | Reply | Permalink
Had to Google Joe Bob. Interesting. I remember him on TNT's "MonsterVision." But I was surprised that he was on the first two years of "The Daily Show." A pretty clever guy - Gergen....ah, hell, I have to spit it out. I can't help thinking Russ Feingold looks just like Soupy Sales.
Neoboho
March 22, 2006 1:42 PM | Reply | Permalink
I agree, and would add that it's also extremely disconnected from average people.
Bush is not getting challenged either about his spin, on his claims about Tal Afar for example. Even on NPR the other day, the reporter didn't offer any counterpoint.
I remember Joe Bob Briggs, we first caught his show on TMC in '83, and I became an avid fan. He followed that up with a similar program on TNT, until Time/Warner/Disney threw the monkey wrench in. I'd recomend his article Joe Bob Goes to the Supreme Court http://www.joebobbriggs.com/misccolumns/supremecourt.asp, it's a hoot and a half.
March 22, 2006 2:35 PM | Reply | Permalink
Yeah, Monstervision.. unfortunately they had to micromanage everything he said until he up and quit.
Re: Soupy Sales.. jeez, we have to be of an age.
March 22, 2006 2:37 PM | Reply | Permalink
Thanks! I don't remember Joe Bob, but I'm very glad "y'all" do. :) Great stuff!
Bee from Connecticut
CSPAN junkies visit http://spannerbackup.ipbhost.com
March 22, 2006 3:49 PM | Reply | Permalink
Briggs' observation goes for not only the Sunday talk shows, but the chattering class in general, which includes the blogosphere. MyDD, for example, projects out trouble for the Democrats based on a less-than-spectacular primary win for Duckworth. To which I can only scratch my head. The McCain will unstoppable meme has taken hold in some other corners, which only makes sense if you think McCain's popularity among independents and Democrats can't be shaken. This kind of punditry is so absurd it reads like a parody.
In December 1990, nobody thought Bush I would lose and few figured that a Governor from Arkansas would be his general election opponent.
It's fun to read tea leaves, but no one should take this stuff seriously.
March 22, 2006 4:30 PM | Reply | Permalink
I remember Joe Bob....musta been way back in '82 (I was 3 <wink wink>) & he used to write the best reviews for the worst (drive-in) movies imaginable (the reviews rarely had anything to do with the movie)...funny stuff then, funny stuff now.
BamBam in Connecticut, late of Arizona.
March 22, 2006 6:01 PM | Reply | Permalink
Yet even as Elisabeth sails with the prevailinfg winds, she still lets slide one of the President's unchallenged "facts."
Kids?
The average of California National Guard members killed in Iraq is 31. One Tennessee Guard member killed was 54 and a grandfather. The average age of regular US Army soldiers killed is 27.
These aren't high schoolers plucked from an inexhaustible senior class (Bush's metaphor, not mine). George Bush has uprooted fathers, husbands, providers, and, yes, grandfathers from their families for a war without purpose and without end and sent them to their death and dismemberment.
These are no "kids." These are men, sir, men engaged in a far noble pursuit than you were at a comparable age, and who deserve far more than your patronizing lip service.
March 22, 2006 10:15 PM | Reply | Permalink
Of course, one of the reasons the straight-line projections last as long as they do is that the conventional wisdom believes in them. And then the corrections are that much more jarring.
March 23, 2006 11:01 AM | Reply | Permalink
J. McCutchen "JmacSF"
San Francisco. CA
The battle between Sunni and Shia Muslims for control of Baghdad has already started, say Iraqi political leaders who predict fierce street fighting will break out as each community takes over districts in which it is strongest.
Many Iraqi leaders now believe that civil war is inevitable but it will be confined, at least at first, to the capital and surrounding provinces where the population is mixed. "The real battle will be the battle for Baghdad where the Shia have increasing control," said one senior official who did not want his name published. "The army will disintegrate in the first moments of the war because the soldiers are loyal to the Shia, Sunni or Kurdish communities and not to the government." He expected the Americans to stay largely on the sidelines.
...Already Baghdad resembles Beirut at the start of the Lebanese civil war in 1975, when Christians and Muslims fought each other for control of the city.
Pointing, "I could never look that Marine in the eye and send him to fight in Iraq if I didn't believe we would suceed and that the mission is vital to the security of the US"
He never DID look that Marine in they eye. He just pointed.
March 25, 2006 12:04 PM | Reply | Permalink
J. McCutchen "JmacSF"
San Francisco. CA
or
5,307,692 US dead in Iraq
March 25, 2006 12:43 PM | Reply | Permalink
Got your message, and although it is an important point, I guess I think that "kids" as cannon fodder for a bogus war are equally horrible as "fathers" and "grandfathers." I know you didn't mean to say that, but the real truth is that George Bush was too scared to send his own sorry ass to Vietnam, so how dare he send any one at all --> kids or not.
Jan Knaus
March 25, 2006 5:05 PM | Reply | Permalink
Not to defend naive projections. I dare say, few posters here have made any projections at all, have more than a "common sense" idea of what a straight line projection is (including Schmitt), or have any hint what would substitute.
I am curious who here has engaged in any forecasting or has even taken a forecasting class. I am also curious what substitutes you have in mind.
Since I have dared bring this up, I will dare make an actual forecast or two.
For the month of November, Bush's Job Disapproval Rate will likely be 63% averaged across all polls. The interval (something like a bootstrap interval, the estimation method doesn't clarify the level) is 53% to 78%.
His approval rate is the mirror image, slightly narrower, with 32% as the predicted value.
This produces a range between the two of 31%. Estimated a different way, the range between the two is 31.5%. The interval is consistent with the other intervals, and since these variables are NOT independent, the intervals are not additive for the range.
Of course, some people will say that an interval of 25 points renders this estimate useless, but forecasts are often quite unreliable. When Bush's job approval tanked its worst last November, he jawboned it back up 10 points over a month of effort. There is evidence he is doing this again. That is what the predicition interval is about.
On the other hand, if he has to jawbone it now, he may not have much effect when jawboning it in September. Too many times to the well. For the record, that's a hope not a prediction.
The 4 year long term trend ("straight line" (not really straight, but straight enough) as most everyone here has put down) is consistent with these results.
If you are going to condemn someone else's projection methods, put your own up for the test.
If it's good for me it must be Good 4 A MericaMarch 26, 2006 3:07 PM | Reply | Permalink
Not to be pedantic, but JBB (John Bloom) was born -- I believe -- Southern Baptist. He's a member of the Trinity Foundation, a nondenominational "first century" Christian church here in Dallas. I don't see eye-to-eye on them on a lot of doctrinal issues (don't assume that Bloom is as liberal as you might think), but they're without a doubt one of the best religious groups around, especially when it comes to working with the homeless. These guys put their money (and homes) where their scripture is.
April 3, 2006 12:02 PM | Reply | Permalink