What Sorta Joint

September 26, 2003

This one’s neat. Concerning a Josh Marshall “Talking Points Memo entry”:http://talkingpointsmemo.com/sept0304.html#0924031029pm where Josh says:

bq. But the key here is that the president’s numbers seem to be in something close to free-fall. His approval ratings have fallen roughly 20 percentage points in four months. Even with all the context which may be fairly provided (like the fact that the 70+ numbers were part of a post-war spike), the president’s rapid descent is undeniable. And it’s not clear he’s hit bottom.

The Wall Street Journal’s “James Taranto writes”:http://www.opinionjournal.com/best/?id=110004063 [ scroll down to "get a grip"]:

bq. Marshall’s analysis represents a triumph of hope over arithmetic. He cites a current poll showing Bush’s job-approval rating at 49%. If we extrapolate using his other figure, a drop of 20 percentage points in four months, we see that this is unsustainable … [ chart showing -11% approval rating by September 2004 ] Mathematically, then, Bush’s “free fall” has to end at some point, with his ratings at least leveling off. And it seems likely that his “bottom” is a lot closer to the current 49% than to zero, for the simple reason that his own party remains united behind him.

So it can’t be free fall because at some point it has to end? Let’s go to our vocabulary list, shall we?

bq. free fall: The ideal falling motion of a body that is subject only to the earth’s gravitational field. [ American Heritage Dictionary ]

“Free Fall” describes the current state of motion of a body, not the ultimate end state of that body. In short, whether or not there’s a sudden stop — or even a bounce — at the end makes no difference. If the fall is currently unrestrained, then it’s free fall.

Even so, the question of whether or not the fall will continue at this rate in perpetuity is not one Marshall has concerned himself with. All Marshall has said is the fall in the president’s approval ratings in the last four months have been falling rapidly. An undeniable fact, not a “triumph of hope over arithmetic.” The arithmetic and extrapolation all belongs to Taranto, who appears to be attacking an argument Josh didn’t make.

Woe be to those who don’t go back to the original source. There’s no hopeful math but what Taranto has made.

In the process of creating a new argument for Marshall that suits Taranto better than the one Marshall actually is making, Taranto engages in some wishful thinking of his own. He predicts where Bush’s non-free-fall will level out:

bq. It seems likely that his ‘bottom’ is a lot closer to the current 49% than to zero, for the simple reason that his own party remains united behind him. Every incumbent president to be defeated in modern times–Ford, Carter and Bush pére–faced an intraparty split before losing the White House…The Democrats, by contrast, are divided.

Let’s see. Ford was elected in 1976. The next election is 2004. In between we’ve had five presidential elections. You’re going to do a historical analysis on the trend of five elections? Are you sure that sample size is large enough? And how does it let you predict a “bottom” number within a tolerable margin of error? In other words,

Look! _Behind you!_

Using “exit polls from the 2000 election”:http://www.cnn.com/ELECTION/2000/epolls/US/P000.html to derive the number of people who call themselves Republican and bother to vote, Taranto only gets around 35%. Assuming, of course, all those Republicans vote for Bush a second time (they didn’t all vote for him the first).

But these numbers are pretty meaningless, too, because — let’s face it — they’re exit polls. You can’t blame me too much, however. It’s damn hard to do a numerical analysis on James’ numerical conclusion, because he doesn’t tell us where he’s getting the numbers from except some mumbo-jumbo about historical trends based on to narrow a sample to be considered predictive, if one assumes that these historical trends have any predictive power on their own — which they probably don’t, seeing as there are an awful lot of other variables involved. Like how many people like the incumbent.

Taranto’s 49% bottom, just like his claim that Marshall is engaging in hopeful math, is just wishful thinking. So not only has he *failed* to even address his thesis (”Marshall’s analysis represents a triumph of hope over arithmetic”), he’s engaged exactly the kind of behavior he’s condemning.

Way to prove your point, James.

[BTW: Josh "defends himself as well":http://talkingpointsmemo.com/sept0304.html#0926031251pm, but not enough in my opinion.]

Posted in Archived Unsorted