Some Journalists Don't Play By the Rules

Posted: Oct 26, 2003 12:00 AM

Actually they're not rules, they're laws: the laws of logic. These laws can get complicated. But they boil down to just one: the law of identity.

The law of identity says things are what they are. Framed negatively, it is the law of non-contradiction; the fact, in Aristotle's words, that "the same attribute cannot at the same time belong and not belong to the same subject and in the same respect." A boulder does not have the features of a donut. If you want to know what a donut is like, inspect a donut, not a boulder. (Assume a fresh donut.)

Logic is good for everything, from crossing the street to doing journalism. One thing logic enjoins as you're doing journalism is that you be internally consistent in your depictions and assessments. Also that you report about the real world, not your free-floating fantasy construct. Fiction has its own logic, of course, but not the logic of reports that carry the weight they carry because of the assumption that there is an effort to use facts, not fictions, as the building blocks.

Let's now examine a textbook example of illogical reporting by an alleged reporter: a newspaper article by Michael Dresser that was reprinted under various headlines around the country. The most apt headline--i.e., the one most clearly indicative of the biased, slanted illogic--was the Orlando Sentinel's: "Women's Progress in Politics Has Stalled" (August 17, 2003). Better would have been "Women's Progress in Politics Has Stalled Thanks to Those Dang Term Limits," but you can't have everything.

The article illustrates several logical fallacies of the sort that I puncture periodically in my Common Sense e-letter, read only by the most sophisticated and politically savvy people. Let's consider two of these fallacies.

1. Post hoc ergo propter hoc. Latin for "after this, therefore because of this." This is the fallacy of treating a temporal relationship as if it must be a causal relationship. So let's say that in 1990, voters pass state legislative term limits in California. Later there's an earthquake, mud slide, fire, Gray Davis, or what have you. You cannot then say, "Aha! See what happens when you have term limits?" No, you have to actually supply evidence showing that term limits cause all bad things. Then, and only then, can you argue that term limits cause all bad things. Sequence does not equal causality.

In his article, Dresser makes two related claims that together constitute this fallacy. The first: that "women's progress in winning legislative seats stalled in the 1990s and has yet to recover." (See fallacy #2.) The second, safely attributed to Dresser's cherry-picked sources: "A major culprit [of this stalled progress], according to lawmakers and academic researchers, is the legislative term limits adopted in 17 states."

The first claim is false. The second claim is unintelligible if the first claim is false. But assuming there were data to support the first claim, one would then have to show that the percentages of women represented in state legislatures had stalled especially in term-limited state legislatures, even as it had advanced unimpeded in un-term-limited state legislatures. In order to begin to show causation, one would have to at least begin to show correlation between the alleged cause and the alleged effect. Dresser doesn't bother to do this, of course, because he's merely quoting "experts" in that impartial way that journalists have when they don't bother to quote experts who actually know about the subject.

2. Interpreting a generally ascending trend line by choosing one tiny slumping splinter of the trend line and treating the splinter as if it were the whole trend line. In it's more general form, this is the fallacy of composition--treating the whole as if it were the same as the part. This is also the fallacy of "context chucking." Or "making stuff up as you go along."

To establish that female progress in occupying state legislative seats has "stalled in the 90s," Dresser cites a single isolated datum from a study by the Center for American Women and Politics at Rutgers University. According to this datum, in 2003 women make up only "22.3 percent of the nation's legislative seats--down from a peak of 22.7 percent last year."

But this kind of minor lump variation in a lump average is meaningless. One must look at the whole trend line. After all, if the ability of women to win state legislative seats "stalled in the early 1990s"--you know, when voters started passing those dratted state legislative term limits--why did 2002 represent a "peak" in the allegedly already collapsed trend?

Well, because the number climbed throughout the 1990s. Let's look at the very data that Dresser cites and presumably read. The Center began recording the number of women in state legislatures in the 1970s. Since then, the proportion of women has more than doubled. In 1978 the average percentage was 10.3%. By 1984, 13.4%. By 1992, 18.4%. By 1999, 22.4%. In 2002, 22.7%. Then--in off-election year 2003--we see a toboggan-slide drop of .3 percent (also known as three tenths of one percent). It's easy to get a trend these days.

What about the dampening effect of term limits? In 2002 term limits had taken effect in 11 states. In 2002, the average proportion of female legislators in those states was 24.8%, higher than the national average. Which doesn't prove that term limits increase opportunities for women at a faster rate than they increase opportunities for everybody. But does help punch a moon-sized hole in the conclusions of Dresser's purported reporting.

And then, of course, you have the incidental zooming of women to leadership posts in various now-term-limited Houses and Senates (which happens because term limits equalizes power within the legislatures). A fact noted in the article but which somehow does not inhibit the sweeping conclusion that female legislative advancement has been plunging into reverse because of that term-limit-impelled point-three-percent thing.

Progress in journalism has stalled.