Polls will drive you nuts

T.A. Barnhart

Polls rule the political world, of course, so it's a good idea to understand how the damn things work. One way to avoid being driven nuts by the polls is to look at as many as possible. Stick with a single poll and you might get the wrong idea about what people are really thinking. Last week, for example, Newsweek showed Obama leading in Iowa. Great news for this supporter, but the other polls still had whats-her-name in front (barely). It takes many polls to find something close to the truth of the moment.

Take the Congress. Please. (Badda boom.)

Everyone hates Congress right now. This is understandable. The Republicans are scam-ridden, cowardly, they got us into this war, and they can do nothing but obstruct and try to hide their bribes. The Democrats meanwhile are inept and impotent.

Have I summed it up pretty well?

The polls back up this summary, albeit with less confrontational language. Congress' approval ratings vary from 23% to 27%, while their disapproval rate starts at 63% (Fox News, oddly enough) to 71%. That's even worst than Bush (29% - 36% approval, 58% - 64% disapproval).

Looking towards next year's elections, I found those numbers scary. What if the voters, in their anger and disgust, gave the Congress back to the GOP? Thankfully, I found these numbers and breathed a sigh of relief:

The voters have consistently positive views of the Democratic Party and negative views of the Republicans. Latest polls from September show the Dems getting approving marks ranging from 50% to 52%. (The Pew Center did ask for "neutal" views and the Dems came in at 27%.) Negatives for the Dems ranged from 38% to 41%.

The GOP is in sad shape, however. Approvals went from 38% to 44%; negatives were 47% to 59% (Gallup; and the Pew poll found a small "neutral" result, 20%).

On the important question of terror, the Democrats are now seen (according to poll from Gallup and AP/AOL News) as stronger. This is a drastic change from 5 years ago, before the war. They are also seen as more likely to keep the country prosperous (Gallup). The AP/AOL poll found more Americans felt the Democrats represented their beliefs more closely (49%-39%). But there is one final question that should give both parties concern: 57% of respondents in mid-September think a third party is needed.

And on and on and on. The polls will drive you nuts if you let them. To avoid this affliction, take these two steps. First, read numerous polls from various sources — and take them all with grains of salt. The more polls you read, the more you'll see that the reported opinions reflect the fact that most people are having a tough time making up their minds. Step two: stop reading so many frikkin' polls! Go do something instead. Politics is not a spectator sport; it's the way we make our civic life reflect our beliefs. Reading polls fixes the political situation just as much as checking the stats moves your team up a place in the standings.

Polls will drive you nuts. Participation will keep you sane.

(And over at DailyKos, Dem from CT seems to have felt a similar need to blog on the polls.)val).

  • pdxskip (unverified)

    Nothing in the polls matters right now, except as conversation pieces over a beer. Way too early.

    As for the approval ratings of the Dems -vs- the GOP trash...all it will take is one terrorist incident inside the US that kills a few hundred or more and everything reverses for us Dems in a very very bad way. We over stretched our vocal opposition to the Patriot Act and I worry it may come back to haunt us. We weren't going to get anywhere on that issue until our guy moved into the White House....but we keep it in front of the voters. Pray they use the Patriot Act powers effectively until Nov 08 and thwart all attacks. THEN we are home free.

    Sigh..........we always seems to over reach.

  • (Show?)

    Thank you for your concern troll, Skip. However, the unconstitutional and/or extremely ill-advised portions of the Patriot act have never shown any actual ability to stop terrorism. How exactly are we made safer by the police infiltrating a bunch of little-old-lady peace activists in California (as documented in Michael Moore's Fahrenheit 9/11) anyway? You really think a sewing circle is hiding a dirty nuke under the needlework?

    The purpose of this unconstitutional spying has been for domestic political purposes and fat payoffs to Republican political cronies. All the real catches have been handled with the old fashioned combination of (moderate Muslim) informants and police work.

  • (Show?)

    A key point TA makes is to look at trends, not single polls. Dunno if everyone knows about it, but Pollster.com has a fantastic trendline for all the polls--national, congressional, primary states. And they show trends, which is key.

    Methodology is also important, but I doubt seriously anyone will dig down that deeply. Although cell phones are starting to mess with telephone surveys (they slightly under-report the youth vote, they're still far more accurate than internet surveys. That will change, but should be accurate through the election next year. Finally, the gold standard in polling is Pew. Take those to the bank.

  • (Show?)

    Don't forget Poland!!!

    Oh wait... you were saying polls not Poles... nevermind.

    /Emily Latella

  • (Show?)

    Posted by: pdxskip | Oct 1, 2007 2:25:54 PM


    So if a terrorist attack occurs while the Patriot Act is in place, that will prove the Democrats are bad on security because the Patriot Act failed to prevent the attack?

    Your argument makes zero sense is is self-contradictory.

  • Bill R. (unverified)

    Polling is largely smoke and mirrors, both in attitudinal and candidate polling. The voter screens are rarely consistent, and the sampling methodology today that excludes wireless phones is very suspect. Having participated in a number of telephone polls the phrasing of questions can be subtly leading. And, of course, much of the electorate seems to ignore elections until the last minute. And the pundits who cite these polls all go to the same bar and parrot the same lines to each other that they then repeat on the same talking heads shows. Everyone had Howard Dean annointed as the Dem nominee just days before the Iowa primary, this time four years ago. It's all smoke and mirrors and an attempt to manipulate expectations.

  • David (unverified)

    I think polls are very useful, if used correctly. It just isn't one poll by itself that matters, but several polls (I would say over a six month period) together, averaged out, give a good indication of developing trends.

  • Robert G. Gourley (unverified)

    How many people want more money than they get now? How many people want less?

    More wins!

    See, polling isn't so hard.

  • (Show?)

    Bill R, that's just a silly thing to say. your example of Dean makes no sense because the MSM couldn't stand the guy. thus the fake scream and its incessant (nearly 800 times) in a week following NH. the trouble with a lot of polls is they ask crappy questions: "If you were given $1000 to spend on anything, what would you buy?" and people then give "good" answers knowing they'd go on a 3-day spree with hookers in Las Vegas. read the questions as carefully as the results (and check who they poll). as Jeff said, what we're looking for are trends, and comparing the results of multiple organizations is the best way to view those trends -- as you reach for your sack full of grains of salt.

  • Robert G. Gourley (unverified)

    Polls will drive you nuts

    There are those for whom it is more of a putt than a drive.


connect with blueoregon