My (Awful) 2010 House Predictions — And Questions About Bias

It seems as if every time I post the results of another close congressional contest, especially one that resulted in a Republican pickup, there is invariably a comment or two from someone wanting to know why — or gleefully pointing out — that I was way off in many of my calls in House elections.

For the record, yes, it's true, I said the GOP would gain 45 seats in the House (while losing four).  As of this writing, the Republican net pickup is 63.  My numbers were not even close.  So the ribbing is well-deserved.

Where did I go wrong?

Paul Cook, a reader who seems to have little more to do than remind me of this fact in several blog comments, points out, rightfully so, that many other analysts were closer to the mark.  He notes that Dick Morris predicted a GOP pickup of 73 seats, Fred Barnes 60, Larry Sabato 55, Nate Silver 54, Charlie Cook 50-plus, Stu Rothenberg 55-65, etc. etc. etc.

This is all true.

He also suggested that I underestimated the extent of the GOP sweep because of my pro-Democratic bias.

That part is not true.

Let me first address how I approach calling the races.

What I did this election, and what I do every election, is call every single House race.  I don't just throw out a number of net gains/losses.  And I especially don't, as many of the above-listed analysts do, say there are 40 or 50 or whatever "Tossups."  I call winners in every contest.

Two of my favorites, and I mean this with the utmost respect, are Charlie Cook and Stu Rothenberg.  They are both smart, on the mark, and fair.  They know their stuff.  And they are friends.

But in his final pre-election chart, Charlie listed 50 races as Tossups.  Stu was less reticent; he had just 18 pure Tossups, plus six Tossups that were "leaning Democratic" and 24 Tossups that were "leaning Republican."

This is not a criticism of Stu or Charlie.  I just want to point out that yes, many were calling for GOP gains of 50 seats or more.  But not everyone went out on a limb and called a winner and loser in every single race.  Which is what I always do.

Sometimes better than other times.  Obviously, this was not my best time.

Yes, I was off on a lot of calls.  But it's silly to say that I was wrong because of a "bias," that I was rooting for Democrats to win.  I can't imagine anyone who has ever looked at Political Junkie over the years — a column (later blog) that began in 1998 — and said I am in the Democrats' camp, or the Republicans' camp.  In 2008, for example, I called every Senate race correctly save one (the Minnesota contest, and that one needed a recount and a court intervention, and it took some eight months to be resolved.)  Also that year, I went wrong on just TEN House races.  Out of 435.

And, for the record, in 2004, I was even better:  I missed only FOUR House races, out of 435.

Was I not biased then but biased now?

In calling every House race, I put my reputation on the line.  Calling a race because I was rooting for one side to win is just ridiculous.

Did I see the tsunami-like magnitude of the GOP sweep in the House this year?  Clearly, I did not.  Yes, a Republican House.  But nowhere near the size of what it was.

But how many of those analysts who were cited specifically predicted the defeats of, say, Ciro Rodriguez and Solomon Ortiz in Texas, or Mike McMahon in New York, or Lincoln Davis (Tenn.), Gene Taylor (Miss.), Ron Klein (Fla.), Charlie Wilson (Ohio), Bob Etheridge (N.C.), Jim Oberstar (Minn.), Walt Minnick (Idaho), etc.?  Yes, these Dems were vulnerable, and yes, in a GOP wave, they could/would (and did!) go down to defeat.  But how many people specifically said they were going to lose?

I'm not making excuses for my calls.  My House numbers, plain and simply, sucked this year.  My Senate numbers were better; I missed just two — Colorado and Alaska.  In both races, I thought the Republican nominees would win.  Does that mean I had a Republican bias in the Senate and a Democratic bias in the House?  Gimme a break.

Oh wait, I also predicted Harry Reid would win in Nevada, though most people did not.  Did I have a GOP bias in Colorado and Alaska but a Democratic bias in Nevada?  My head is already spinning.

Helen Highwater, another persistent commenter on this subject, asks these questions:

1) Was there personal bias involved? (Not really expecting an admission, even if there was).
2) Is his opinion trustworthy in future elections?
3) Was there any pressure from NPR or its management?

I already answered #1.  As for #2, that judgment is up to the readers; Helen has already made up her mind.  I would like to think my blog offers more than simply predictions.  But my record, especially in the House, had been pretty exemplary in previous election cycles.  And as for #3, that is so patently ridiculous it's probably not even worth responding.  But there are those who are convinced that everything that happens at NPR is because of ideological pressure or interference from above.

Besides, I thought we were Nazis, not Democrats.  Oh, wait.  "Left-wing Nazis."  My mistake.

I especially enjoyed this comment from Helen in a previous post:

If your boss asked you to develop a budget for a major project, and you underestimated by 55%, would you be asked for an explanation? Would you still have a job?  My theory is that he gave those 22 races to the Dems for perhaps personal/political reasons, so that NPR voters (heavily Dem) wouldn't be discouraged from going to the polls. (It obviosuly didn't help).

My dear Helen, I appreciate your tenacity on this subject, but my job doesn't depend on my predictions in House races.  I didn't get a bonus or a gold star in 2004 and 2008 because my House calls were excellent, and I won't get fired in 2010 because my House calls stunk.  Your "theory" is just nonsense.

But heck, people will believe what they want to believe.  I know in this day and age of decreasing tolerance for things they don't agree with there are conspiracy theories and theorists everywhere; the fact that some have gone out of their way to post numerous (though repetitive) comments about the "real reasons" for my predictions should not be surprising.  Heck, if I made predictions based on bias, I'd have the Yankees winning the World Series every year.

Look, this was an amazing election.  The Republican gain of 63 House seats (and counting) was nothing short of historic, the largest gain for any party since 1948. Most people got it right.  I did not.  Most people did not call each and every House race.  I did.  The eventual results show that this year's calls were not my proudest moment.  But I laid it on the line for people to see.  It's not as if I predicted the Democrats would keep the House.  Still, if some think that my decades of covering elections is little more than about my bias, then there really isn't much more I can say.

And that's the last I have to say on this subject.  See you in 2012.

Comments

 

Please keep your community civil. All comments must follow the NPR.org Community rules and terms of use, and will be moderated prior to posting. NPR reserves the right to use the comments we receive, in whole or in part, and to use the commenter's name and location, in any medium. See also the Terms of Use, Privacy Policy and Community FAQ.

Support comes from: