Michael Barone's very bad column

Blog ››› ››› JAMISON FOSER

Slate's David Weigel catches Michael Barone "phoning it in":

I'm not even sure that Michael Barone woke up before writing this one about "the Democratic party shrinking back to its bicoastal base."

Now we see Barack Obama campaigning at the University of Wisconsin at Madison, in Dane County where he won 73 percent of the vote in 2008, chiding students for their apparent apathy. Sen. Russ Feingold, who lives in Middleton, four miles away, was unable to make it -- and it's not the first Obama event in Wisconsin he's skipped.

Actually, Feingold did make it to the rally, announcing his intentions in a Tuesday afternoon tweet and giving a brief speech from the stage. I noticed the mistake in the free print edition of the paper, but it's still online.

Barone's column contains another flaw that suggests it was thrown together with no more than eight minutes of thought: a reliance on the tired old gimmick of pretending that the GOP's edge in "land mass represented" is meaningful.

Here's an exercise for some evening when you're curious about big nationwide trends in this year's elections.

Get an outline map showing the 50 states and take a look at the latest poll averages in pollster.com in each race for senator and governor. Color in the percentage (rounded off; no need for tenths) by which either the Republican or Democratic candidate is leading (I use blue for Republicans, red for Democrats) in each state.

The results are revealing, even breathtaking.

The map of the Senate races shows Republicans leading over almost all the landmass of America.

I imagine most readers stopped right there. I mean, who hasn't read this exact column a couple hundred times over the past decade? (I guess Barone thinks he's keeping things fresh by switching around the now-standard blue/red indicators. Bold!) Is there anyone who still believes -- actually believes, not just pretends to believe -- that maps that give parties credit for landmass represented are anything other than wildly misleading?

Several paragraphs later, Barone admits this, sort of:

Now, the geography can be a little misleading. The Democrats' Northeast and Pacific Coast bases are heavily populated, and the states where they're leading in Senate races cast 136 electoral votes in 2008. But the states where Republicans are leading cast 274 electoral votes.

A little misleading? No: It can be completely misleading. You know what else is misleading? Using electoral votes as a proxy for population. And Barone is wrong about the electoral vote totals, anyway: States in which Republicans are leading cast 257 electoral votes in 2008, not 274. By using electoral votes, a misleading proxy for population, then inexplicably awarding the GOP a bonus of 17 electoral votes, Barone makes it look like the Republicans have a 2-1 advantage.

If, on the other hand, you look at each state's population, you find that Republicans lead in states containing a total of about 144 million people, and Democrats in states with a total of about 82 million people. That's still a sizable gap, but considerably smaller than a 2-1 margin. (And then there's the fact, not acknowledged by Barone or included in his calculations, that there are two Senate races in New York, both of which Democrats are leading. Include both races, and the gap shrinks to 144-101.)

Long story short: Beware the columnist who misleads you in his disclaimer acknowledging that he misled you earlier.

Washington Examiner
Michael Barone
We've changed our commenting system to Disqus.
Instructions for signing up and claiming your comment history are located here.
Updated rules for commenting are here.