« Republicans overcome image gap | Main | Twitter roundup »

November 01, 2010


According to Chait, the Douglass Hibbs structural factors predict a 45-seat Democratic loss in the House. (Chait thinks that's 5-10 seats too low, but I'm interested in what the models predict, not what Chait thinks.) My question is, does Brendan accept that 45-seat loss as the amount attributable under the best available model to structural factors? If not, what is the predicted loss according to the model Brendan prefers?

I ask these questions because once we know how many lost seats are attributable to structural factors, we can then compare results on Tuesday against that structural factor baseline and consider which other factors (some or all of which may be listed by Brendan) account for the variance.

Thanks for the bingo chart, Brendan. With your permission, can I print a handful of them off and give them to my nieces and nephews as Christmas gifts?

Go for it -- they also make fine Hannukah and birthday presents.

The one thing you didn't list is the "I'm mad as hell and I'm not going to take it anymore" reason. All seats gained over 30-40 should be attributed to that! If you got out among the people you would know that!

Brendan's chart suggests that he may be spending more effort on defending his particular model than on improving it.

Note that the term "structural" is used in a special way. IMHO it's a kind of spin. It's supposed strengthen one's belief in Brendan's favored model by conveying the idea that economic factors are somehow part of the fundamental structure of an election. Of course, that's not the the case. Economic factors can be shown to correlate pretty well with election results. They're convenient because they can be measured objectively and numerically. However, economic indicators are not a part of the structure of an election.

Brendan wants us to believe that "non-stuctural" elements are not significant. The best way to do that would be to include these elements in his model and verify that their inclusion doesn't improve the model's accuracy. Instead, he resorts to mockery.

One missing "non-structural" element that I believe is significant is media bias. E.g. ABC News had invited a conservative speaker to be part of their on air election analysis on ABC TV. In the face of left-wing objections, ABC News rescinded that invitation. (As usual, he was baselessly accused of being a racist.)

ITSM that when a major mainstream TV station allows the left-wing to exercise a veto over its guest list, that must have an effect on the many viewers who depend on ABC News for their worldview. Put another way, would it not make a significant difference if mainstream media allowed conservatives equal access? I don't know how to measure media bias numerically and objectively, but I believe if one could do so, it could be used to improve the election models.

Brendan, do you think the Tea Party has hurt Senate candidates in blue states (e.g., Linda McMahon) who hasn't associated his or herself with them?

Oh, and thanks for the bingo cards. I promise you in no time flat, I'll be the uncle all the kids hate the most.

I'm not sure it's my ideal model, but it's the one I've seen that includes the fewest indicators related to contemporary politics (generic ballot, presidential approval, etc.). In any case, the model probably won't hit the number perfectly, but we should avoid making up stories about why that is without considering how your variable of choice relates to the model's predictions for all years in the data (if possible).

Although Brendan's clever chart represents a wide spectrum of pundits, ISTM it leaves out some conservative explanations for the prospective Republican landslide. They basically add up to accusations of bad laws and irresponsible government.

Number 1 is the Health Care Law. It was shoved down the throats of the American people against their wishes. It was passed by means of backroom deals and political payoffs, rather than legislative improvements. The fact the no legislator read this ground-breaking bill remains mind-boggling. ( Here Eugene Robinson imagines that racism accounts for Tea Partiers wanting to "reclaim their government." IMHO it's the way the Health bill was passed that led Tea Partiers to believe that the government had been taken away from them.)

The failure of the Stimulus Bill despite its enormous cost is second. Americans know we are suffering now from a contining recession and we will suffer in the future to pay back the enormous debt.

Third is the level of the deficit. Everyone agrees it's unsustainable, but nobody knows how to fix it. People worry that the deficit and national debt may become so large as to cause some sort of economic catastrophe.

Fourth is the irresponsible attitude of Congress and the President in the face of potential economic catastrophe. They've been going ahead and spending as if they had all the money in the world. We've seen no serious cost control measures come out of Washington. Congress's failure to pass a budget and their failure to decide what to do about the expiring tax cuts are other examples of irresponsibility. The President should have been haranguing the Congress to take action on these vital matters.

From the conservative POV, these substantive failures are major reasons why the voting public wants to make a change.

If the average voters really were angry about a bill being "showed down their throat" (after more than a year of negotiation and compromise on major principles, e.g. public option ... and being functionally identical to Newt Gingrinch's 1995 plan) or even actually worried about W's last budget's $1.4 Trillion deficit (yes, Virginia, the 2009 budget was written in the Fall of 2008), then there would be no need for attack ads paid for by parties unknown ... certainly including domestic carbon barons and possibly (but unknowably ... thanks to GOP's blocking the DISCLOSE Act) including foreign enemies of our United States, via soveriegn wealth funds.

That's the story of this election, but it's not in the "lamestream media" which to a large extent is owned by the same people buying the election (NewsCorps' largest shareholder is foreign born and 2nd is a Saudi).

So blame the Democrats for style or whatever, but buy all means, don't "follow the money". That's sooooo Nixon-era.

Thanks David in Cal for filling in my bingo card slot for "too liberal". One slot I would have liked to see would be blaming the Citizens United case for the Dems loss. I'm willing to believe money has an influence but I think the changes made by this case slide into the margins.

The comments to this entry are closed.