Here’s something to think about when looking at Nate Silver’s column on skewed polls. Going back to 1990, the elections that did not turn out as expected wound up having the worst skew in the polls. Now, that might seem like a tautological observation, since expectations are formulated largely by looking at polling data. But, the premise here is that bias can sometimes be introduced into the polls based on outside factors, including the meta-media narrative about which party is supposed to win or lose.
Going into the 1994 midterm elections, the Democrats had controlled the House of Representatives since 1955, and for all but two single congresses since 1933. It was difficult to imagine the Republicans winning the House, and even harder to foresee the absolute landslide that actually occurred. It’s not surprising that the polls in 1994 were skewed three percentage points in the Democrats’ favor. Was it a modeling problem or a lack of imagination on the pollsters’ part?
The 1998 midterms were dominated by l’affaire Lewinsky, which was more of a Beltway obsession than a national one. The polls skewed four points toward the Republicans, which was the biggest skew (for either party) in the time period we are examining here.
The 2002 midterms are the ones I remember, because the polls were much too optimistic for the Democrats. We seemed to lose every close race, including ones in which we had been narrowly favored. Perhaps the New York/Washington-based media simply failed to understand how popular Bush was at that time in the rest of the country? Perhaps the pollsters couldn’t calculate the impact of war fever? The result was the biggest skew (4%) in the Democrats’ favor in this time period.
The 2006 elections were expected to be good for the Democrats, and they were, but the polls still skewed two points Republican, slightly underestimating the ultimate size of the Democrats’ victory.
And then in 2012, the polls skewed three points Republican, which made the Obama/Romney race seem closer than it ever was.
Silver is correct that the skew goes back and forth, sometimes favoring Democrats and sometimes favoring Republicans, but the one pattern I see is that the polls tend to skew in the direction of the media narrative or, at least, hew closer to what seems imaginable than what actually takes place.
In 1994, it was assumed that the Democrats simply owned the House of Representatives. In 1998, it was assumed that the nation shared David Broder’s revulsion with the president’s adultery. In 2002, the elite media may have lost a feel for the sentiments of the heartland. In 2012, the media may have been willing a competitive race that never really existed.
So, this year, if there is to be a skew, which way is it likely to swing? The media has been giving us a steady drumbeat of predictions that the Republicans will win back the Senate. They predict low voter turnout by Democrats and they suggest that the president is very unpopular and will be a drag on the tickets. It hardly seems possible that they’ve underestimated the Republicans’ strengths .
My prediction: either the polls are accurate or they have a Republican bias. I think we can rule out a Democratic bias, despite Nate Silver’s caveats.
In the twelve elections since 1990, only four elections have seen unskewed polls. That’s thirty-three percent. Does that mean that there is a 67% chance that the polls are skewed toward the Republicans?
That kind of math will probably give Nate Silver an embolism, but I do think the media narrative has had an historical role in biasing the polling data.
Maybe the media influence the way that respondents to polls think they should answer in public conversations.
The reluctance to be honest about voting intentions might not be constant from election to election, depending on the media story. Certainly when North Carolinians were putting on the face of racial progress in the early 1970s, being for Jesse Helms in his first election to the US Senate (over a guy named Nick Galiafinakis) was not something expected in a state still dominated by the Democratic Party. That is why after the election the myth of the Jessecrats was invented, who turned out to be Republican voters still registered as Democrats.
Makes me wonder whether that phenomenon will work in reverse in some red states this year.
That sort of skew can be estimated. More interestingly after the election it can be geographically mapped. What precincts. counties, states departed most from the polling consensus. Obviously for Karl Rove, it was Ohio, but which parts of Ohio screwed the polling.
This is where “gut” can outperform a base model and why my results beat Silver in 2008, 2010,and 2012.
I think we need to look at individual races than the country as a whole. I think Colorado polling is skewing Republican, possibly heavily. I think Kentucky polling is probably accurate. I think Arkansas is slightly skewed Republican. I’m not sure on Louisiana Kansas might actually be skewed a little against the Republicans, but that’s such a weird scenario that it’s tough to tell.
The only races im not really sure of are Iowa and Georgia. Iowa very well might be skewed in favor of the GOP, but it’s also possible to go the other way. I don’t know Georgia’s GOTV operations well enough.
Kilgore has a very encouraging take on these lines: if you look just at states with competitive Senate contests voters are much more enthusiastic and that favors Democrats. I always think of turnout first and the hatred of Congress and lack of hope we see everywhere made me think it would be low, but this is a counterindication.
Perhaps whichever way aggregate public perception and mood skewed nine to sixteen months before election day. Compare the 1994 and 1996 bias with that of 2006 and 2008. Mirror images of each other.
Why did polls skew significantly in the wrong direction in 1994 and 2006? 1) Discounted the mostly manufactured anger against Clinton and legitimate anger against GWB. 2) Developing or late breaking Congressional scandals. Maybe Silver should consider adding secret sauce to the later polls instead of the early ones.
2002 was the second time I paid close attention to the polls. Tossing out the crap polls and giving more weight to trends, either other Democratic and liberals were smoking something or I was. The trends said that the 50/50 electorate of 1996-2000 had broken and the GOP was now favored. Republicans would win most of the too close to call elections. And they did it again in 2004. The opposite was in place for the next two election cycles.
Not seeing anything at the national level that would suggest this is a change election. Democratic Senators that have been labeled as vulnerable aren’t bums. Nor are three retiring Dem Senators in Iowa, WV, and SD. I’m still going with the status quo is favored which means that the polling is skewed to the right. Maybe not as badly skewed as in 2012 but that’s an extremely low bar since anyone with half a brain didn’t even need polls to project the outcome correctly. With the wildcards in play — all of which favor Democrats — Reid might have a pleasant election night.
Because that’s when the national media narrative takes shape? (CNN brings out its Fall Collection)
Something like that — that’s when they write and file their copy and spend the remainder of the election cycle having drinkies with the Beltway and regurgitating their original copy.
Early polls mostly pick up diehard partisans and a smattering of those that value and are tuned into Beltway narratives. Neither changes much during the course of an election cycle. The results of completely unbiased early polling would favor conservatives/Republicans because they really made up their minds decades ago and little factual information that would challenge their position ever seems to impinge on their brains.
2002 was curious as the polls seemed to detect the pre-9/11 narrative — if not the GOP, then the SCOTUS stole the election (that made even some that had voted GOP uncomfortable), the Bush/Cheney tax cuts (so bad a GOP Senator defected, after voting yes, which flipped the Senate from 50/50 plus Cheney to DEM), the secret energy deals, and national reporters didn’t like hanging out in Crawford in August.
There is a vested interested in the never ending search for more and more MONEY that polls get things wrong at time. I have taken the policy of not worrying about any poll results. This is easy to do when every time a supposed poll shows a candidate down it is used to email others to donate MONEY.
The only poll that matters to me is the total votes counted at the end of election day. The outcome of these elections are decided by those that make a choice to not vote. What is needed is a tidal wave of voters to show up
and VOTE. Want to help your candidate write to the local papers and encourage others to get out and VOTE. Go in the neighborhoods door to door encourage others to VOTE. Tell all you know to VOTE.
An interesting phenomenon, and one that may be relevant, is something that everybody and his uncle were talking about up to this year, namely the received wisdom that midterm elections nearly always take seats away from the party in power.
On the one hand, this is another reason for the media narrative. It is ordained that this election just has to be bad for the Democrats. I mean, look at 2010.
On the other hand, it’s been months since I’ve seen anyone actually citing this particular old chestnut. That is encouraging. There is a high level of uncertainty this year, on both sides.
Have to back to 1966 for the last time a DEM POTUS in his second midterm took a licking. Not to bad in the Senate and while the losses were big in the House, Democrats remained in the majority.
Truman lost big in 1946 — but won back both House in ’48 and held them in ’50 with some losses.
Clinton and Obama flubbed their first midterm. Clinton never recovered but avoided further damage in the next two. Expect to see something similar for Obama’s second midterm
Interesting that Carter/Democrats didn’t lose in 1978. The held it off until 1980.