A new blog called the Mystery Pollster written by a longtime political pollster Mark Blumenthal has an excellent entry on why the supposedly oversampling of Republicans in major polls isn’t what it seems. He notes, as I have previously that party identification isn’t some figure that remains stable over time:
Let me add a few thoughts. Party Identification is one of the longest tracked and closely examined questions in political polling. “Generally speaking, do you consider yourself a Republican, a Democrat, an independent or what,†has probably been asked on more surveys over more years more consistently than any other question.
The most important thing to remember is that Party ID is an attitude, not a demographic. People can change their views of political parties. They cannot change their age, gender, race, years of education and locale (unless they’ve moved).
However, for many years political science students were taught that Party ID, like religious affiliation, rarely changes among adults. This belief was the result of two famous panel (or “longitudinalâ€) surveys conducted by the University of Michigan’s National Election Studies (NES) that interviewed the same respondents three times over a four-year period. These studies convinced researchers that Party ID was highly stable at the individual level.
One limitation of those studies was that they were conducted during periods (1956-1960 and 1972-1976) when overall levels of Party ID remained virtually unchanged. Later closer examination of the data showed that even then, one of every four respondents had changed their answer to the basic Party ID question over the span of four years. Most of the change involved shifts between independence and one of the political parties (See Samuel L. Popkin, The Reasoning Voter, pp. 53-55).
Indeed party ID is not a stable factor – people can and do change their voting habits all the time. Many Democrats voted for Reagan in 1980 and 1984. Republicans voted for Perot in 1992. Many Democrats have indicated their intention to vote for Bush this year.
This election is likely to be a realigning election. Every 30 years there is a major shakeup in American politics that changes the political landscape in a lasting way. The issue of national security seems to be driving many Democrats away from Kerry and towards Bush. Already we have Randy Kelly, Ed Koch, Zell Miller, Roger L. Simon and others moving from being staunch Democrats to Bush supporters. These “9/11 Democrats” are a voting bloc that may very well tip the scales towards Bush, and they’re one of the major reasons Bush is polling ahead of John Kerry at the moment.
This is where Zogby and others get their models wrong. Zogby deliberately manipulates their data to fit a preconceived notion of where the allegiances of the electorate lies – so if they get more Republicans in a survey, they assume that’s a sampling issue rather than a real movement in partisan affiliation over time. Of course, if the reason why Gallup or the Washington Post are getting more Republicans in their polls are because more people identify themselves as Republicans than their models indicate, they’re essentially skewing their own data. Indeed, polls that do this show a much tighter race than those that do not.
That doesn’t necessarily mean that Gallup’s double-digit lead is correct – a more modest 3-6% Bush lead is probably far more accurate. What that does indicate is that some pollsters are adding an artificial bias to their polls by trying to shove their data into an assumed framework that may or may not be correct. Given the strong possibility for realignment and the shift in other non-weighted polls, the evidence suggests that the assumptions some pollers are making about the partisan balance in the US is simply wrong. They’re trying to fit their data to match a preconception, and that means that the data some polls are giving us are being tainted by the inaccuracy of that preconception.
Even pollsters who don’t manufacture a Republican registration advantage tend to strongly favor the GOP, particularly in this election. The vast majority of reported polls only cite “likely voters”, a term that often applies only to those who have voted in most recent elections (which I assume also means off-year elections like 1998 and 2002). In a race as contested as this one, I’m expecting higher voter turnout, meaning the registered voter models that tend to favor Democrats in general and Kerry in particular are likely to be closer to right than the likely voter models. Furthermore, pollsters don’t call those whose primary phone is their cell phone, disproportionately excluding 18-29 year-olds who strongly favor Kerry. There’s are several reasons why virtually no polls came within arm’s length of predicting the 2000 popular vote outcome accurately. Outdated and downright incompetent polling methodology was most likely near the top of the list.
I don’t have a landline, and I’m voting for Bush. (Big surprise on the latter, I know).
All this talk about how skewed telephone polls are ignore that A:) turnout in the 18-24 bracket is always extremely low B:) that those who don’t have landlines probably don’t answer political poll in the first place and C:) for all the talk about high Democratic turnout, there’s yet to be a poll that shows it.
At the end of the day, most college-age kids have the attention spans of goldfish and the libidos of rutting boars, neither of which are condusive to political activity. Believe me, I’ve tried to run youth voting initiatives and it’s damn hard. Every cycle everyone says how the youth vote will change everything and it never happens – and this year will be no exception.
The youth vote propelled Gore to a popular vote victory in 2000…and it almost definitely sealed the deal for Gore in the same-day registration state of Wisconsin, where Gore did gangbuster numbers in Madison (substantially better than Clinton in ’96). There’s far more at stake at this election, and while there will still be a large number who fail to vote, I suspect the vast majority of those who choose to cast ballots in the final days of the campaign will do so in favor of Kerry. Of course, the election will have to be perceived as close for them to bother. If the headlines suggest Bush has it in the bag on October 20, 18-24’s won’t waste their time.
Mark, see bellow:
————
from RedState
PAYDAY November 2, 2004 By: Yahuti · Section: Diaries
I found this item this morning (September 23, 2004)in the Washington Times. It paraphrases Barbara Stock’s writings earlier this month at the American Daily.com. It told me nothing new, but as a Vietnam vet Stock’s words impacted upon me with surprising power. Kerry’s “Reporting for Duty” posturing at the Democratic Convention DID make my skin crawl. And, yes, I am one of probably millions of vets who sat quietly out here in the either for 30 years, controlled our anger an loathing, and kept our mouths shut.
Let’s begin a true blogo-‘Movement’ here and now. Let’s address all Vietnam Vets, all vets, their families and friends – and declare November 2, 2004 PAYDAY for Kerry. Let’s give the good Senator what he has coming, and has had coming for decades: A humiliating drubbing at the polls.
We need to get word of this movement all over and throughout the ‘blogosphere’ and trumpet it until the polls close a little over a month from now.
I quote the Washington Times article below
Yahuti
xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
“Over thirty years ago they put away their medals and their uniforms. They buried their anger and bitterness and moved on with their lives . . . and they waited.
Most of them didn’t know who or what would be the signal to make their move, but they knew they would recognize it when it happened.
On July29, 2004, it happened. John Forbes Kerry came to the podium at the Democratic Convention and uttered three words that made many Vietnam vets’ skin crawl: “Reporting for Duty!”
Kerry had stripped them of their dignity the day he sat before Congress (in 1971) in his fatigues . . . He had publicly turned on his fellow vets while they were still in harm’s way and American prisoners were still in the hands of the enemy. Kerry accused them all of being out-of-control animals, killing, raping and pillaging Vietnam at will . . .
Kerry denied them their rightful place as heroes and they will deny him his dream of the presidency. Angry Vietnam veterans, silent for so long, will finally have their say. Payment in full will be delivered to John Kerry on November 2, 2004. Revenge is indeed a dish best served cold”
Washington Times, Culture, etc., Thursday September 23, 2004. Paraphrasing Barbara Stock, writing on “Revenge is Dish Best Served Cold”, September 1, 2004 at http://www.americandaily.com
————
This may be off topic (since I haven’t read the above), but on the issue of voting. I was listening to early morning drivetime radio and was informed that Teresa Heinz-Kerry sat down with a Arizona report either yesterday or not to long ago. The reporter was asking why the Kerry campaign was pulling its ads from Arizona.
Are the giving up on Arizona?
Apparently
Yeah, Kerry’s given up on Missouri, Arkansas, Louisiana, and Arizona. I’ve seen surveys with Bush up by 17 in Arizona, so it’s probably not a bad move for Kerry at this point. All these states were won by Bush in 2000, and probably weren’t going to Kerry.
Florida, Ohio, and Colorado are all in play. If Kerry could take one of those it would severely hurt Bush. Then again, if things continue along the current trendlines Kerry will have already lost so many states it won’t matter…
Yea, but I seem to remember that Arizona was narrowly run by Bush. And I’ve read in articles today that Bush’s lead in Arizona isn’t as well defined as 17 points. The reporters of the area are claiming that the lead is 1 point.
Electoral-vote.com has posted a poll from Survey USA showing that Maryland might be an even tie right now. Since that states has 10 electoral votes I’d say it is at least as important as Colorado. I don’t fully believe the poll, but if there is still 16 states and a total of 181 electoral votes up for grabs near November we clearly could have a long night on November 2.
Personally I hope we know who our president is on November 3rd.