Our broken national media
Tuesday, Nov 27, 2007 - Posted by Rich Miller
* This item in a Boston Globe story tells us a lot about trusting the national media’s polls in Iowa…
In making [caucus turnout] projections, campaigns rely above all on their “hard count,” a tally of voters who have pledged to support them, and a list of previous caucusgoers made available for sale by the state party.
But no media organization is believed to have purchased such a list, so instead of knowing who has participated in past caucuses - considered the best indicator of turnout - pollsters are random-dialing households and asking voters whether they have voted before and how interested they are in the current race.
Iowa’s goofy process is drastically different than a traditional primary. You have to go to a neighbor’s house and openly declare your support for a candidate. So you’d think the media’s polling methods would be different. They’re not. Oops.
Thoughts on the race so far?
- Buckeye - Tuesday, Nov 27, 07 @ 10:08 am:
It is goofy, and to have the election process begin (and due to enormous attention paid by the media)and unduly influenced by this state is not a good/healthy idea for our body politic. Media and everyone else guesses (although it passes as “Polling”)….Then, WALLA, an unexpected winner! Any wonder most people are turned off by this process?
- Greg - Tuesday, Nov 27, 07 @ 10:17 am:
Agreed about Iowa being an absurd process.
Political futures markets currently have (in Iowa) Obama trailing Hillary by only a few percentage points, and Romney and Huckabee trading dead even, and all other Republicans essentially zero chance. Seems about right. These markets consistently outperform polls and pundits.
- Kiyoshi Martinez - Tuesday, Nov 27, 07 @ 10:36 am:
I think it’s interesting that despite all the hype of how the Internet’s going to change this election and presidential politics, that we’re seeing very little evidence that it is.
Outside of fundraising, all these videos, petitions, Facebook groups and MySpace profiles aren’t translating into effective support for candidates. I think the CNN/YouTube debates are the Web 2.0-politics equivalent of jumping the shark.
- Poli-Sci Geek - Tuesday, Nov 27, 07 @ 10:36 am:
I have caucused in Iowa. The process is a lot more thoughtful than running into the booth and punching a name on the machine. The caucus participants are much more fluid because they can be convinced to change their vote at the caucus, this dynamic makes any polling suspect.
- Ghost - Tuesday, Nov 27, 07 @ 10:43 am:
An interesting tidbit to this mix is the number of young people who seem to be attracted to Obama. This group, traiditionaly, has been very apathetic about voting. If Obama can get his supporters to show up and vote at the caucas, I think he will do better then the random phone polling predicts.
- cermak_rd - Tuesday, Nov 27, 07 @ 10:48 am:
Iowa’s process is goofy, though sometimes it’s at the local high school and not a neighbor’s house. I find the lack of anonymity to be disturbing as I believe the secret ballot to be fundamental to a functioning democracy even though I know, a primary is not really a part of the democratic process, it still makes me squeamish.
- Anon - Tuesday, Nov 27, 07 @ 10:51 am:
My other issue in Iowa is the number of people who participate. It’s very small.
Thanks to Jimmy Carter, that small number now exert enormous national influence, thanks to our screwed up 24/7 media.
It’s just wrong.
- Ghost - Tuesday, Nov 27, 07 @ 10:57 am:
===thanks to our screwed up 24/7 media====
I disagree. We as individuals need to take responisibility for our own decisions and stop trying to place blame. The media will stop covering stories and reporting when we stop watching. We need to stop blaming the media for our decisions to view and follow news stories. People choose to follow these sotries and polls, and thus the media cataers to our political vouyerism. In the end, if we stopped watching the would stop reporting.
- Anonish - Tuesday, Nov 27, 07 @ 11:52 am:
I agree with PSG. I think the caucus process calls for a more thoughtful and invested participant.
As for anonimity, the caucus is a function of the party, not the state election board.
What should really annoy people is that while the individual precinct delegates to the county are decided on January 3, the county and state deligates are not decided until much later and they can change candidates along the line. So the binding state decision isn’t even that night.
I am personally very excited to go out in January and work a caucus room for my candidate.
- GoBearsss - Tuesday, Nov 27, 07 @ 11:58 am:
Ghost, I noticed an error in your post, and so I corrected it and reposted below:
“An interesting tidbit to this mix is the number of young people who seem to be attracted to Dean. This group, traiditionaly, has been very apathetic about voting. If Dean can get his supporters to show up and vote at the caucas, I think he will do better then the random phone polling predicts.”
- Poli-Sci Geek - Tuesday, Nov 27, 07 @ 11:59 am:
cermak rd, the use of the secret ballot in US elections didn’t happen until after 1890.
- ZC - Tuesday, Nov 27, 07 @ 12:16 pm:
Goofy or no, for Obama supporters out there, be down on your knees thanking God Iowa works the way it does.
There’s another side to Iowa being so small: according to a late Nov. survey by the Washinton Post, over half the likely caucus-goers have reported that they have attended a campaign event. That may well rise in December. Can you point to any other election, anywhere in the U.S., where over half of the electorate has personally seen and evaluated at least one of the candidates?
My feeling is lay off the Iowans. The caucus-goers take their job responsibly, they are independent-minded, and they are one of the few examples where the political news media clearly take their leads from the voters, rather than the other way round. We could do worse.
- Dan Vock - Tuesday, Nov 27, 07 @ 12:45 pm:
While you’re on the subject of the problems with polling, you might want to check out this post. It talks about all the potential pitfalls of trying to poll New Hampshire accurately before its Jan. 8 contest: http://marcambinder.theatlantic.com/
archives/2007/11/polling_over_the_h
olidays_a_ca.php
- L.S. - Tuesday, Nov 27, 07 @ 12:50 pm:
I’m very curious to see what (if any) supporters switch to Obama and Clinton in the last minute. Four years ago, if I recall correctly, Kusinich supporters switched to Edwards and gave him a strong showing. It could be a very good indicator as to where some of the lower tier candidates will throw their support as they inevitably run out of money.
- cermak_rd - Tuesday, Nov 27, 07 @ 1:46 pm:
Poli-Sci geek,
Wow I didn’t know that. I would consider US democracy to have really started in 1919 and only been really involved in the 1960’s with the civil rights acts that got rid of the poll taxes etc. Anything up to those points was more poorly functioning plutocracy than anything.
- VanillaMan - Tuesday, Nov 27, 07 @ 1:48 pm:
It is already too much!
When will we have a candidate become the nominee so soon that they implode before their party’s convention? I can easily see a time when we end up with one candidate for each party by March and then see a disaster occur before the conventions in the summer. What then?
The earlier the nominating process becomes, the riskier each party allows time to unravel their presidential campaigns.
It just wouldn’t surprise me to see Clinton end up the nominee and watch as polls start showing voter’s remorse by September. It could happen to any of the candidates running.
This has happened before. In 1976, we watched Carter lose 30 points in the polls and barely win the election due to voter’s remorse. Within six months in office, this continued to the point where then-President Carter was only 28% approval rating, even lower than Bush’s today.
We are screwing up our election system by catering to a bunch of geeks and nerds demanding political news before it is real news.
- kimsch - Tuesday, Nov 27, 07 @ 2:38 pm:
ZC,
How does “half the likely caucus-goers” equal “half the electorate”?
Actual voters
- kimsch - Tuesday, Nov 27, 07 @ 2:56 pm:
part of the comment didn’t show up:
Actual Voters < Likely Voters < Registered Voters < Adults legally allowed to register to vote
- train111 - Tuesday, Nov 27, 07 @ 3:29 pm:
When can we all say enough!!!
The whole 24/7 365 day 4 year Presidential Campaign is starting to really get on my nerves. Good God is it so hard to land a real job for people to do nothing but run for political office permanently.
Iowa/Schmiowa–wake me up sometime in July so I don’t have to hear anymore.
train111
- Calling V-man - Tuesday, Nov 27, 07 @ 6:37 pm:
Once again, Vanilla hits it on the head.
Iowa is only important because Jimmy Carter said it was.
It is not a microcosm of our nation, it is not representative of the primary process nor are there many electoral votes one way or the other.
That said, on a semi-related note, I wish people, including some contributors to this site, would stop saying that they are “registered” Republicans or Democrats. We have open primaries in Illinois that do not require party registration like some other states. Why people think they are “registered” is beyond me.
Maybe they want to think so.
- Lyrl - Tuesday, Nov 27, 07 @ 7:32 pm:
I believe only Iowa’s Democratic party requires open support, and the Iowa Republican party votes by secret ballot at the caucus.
How accurate is past participation in caucuses as a predictor of participation this year? Or, turned around, how many caucus goers in a typical year are “new”? If 70% of caucus participants are repeat goers, that would be large enough to make those lists very valuable to campaigns, but small enough to make polls of only that group somewhat unreliable.
The Boston Globe article has an excellent point about polls that just ask people if they vote. I read somewhere that 95% of registered voters polled will say they are likely to participate in caucuses or primaries - but only about 10% actually participate. That’s a BIG margin of error.
I’m not convinced there is a good way to accurately predict who will participate in the caucuses and primaries. Which means it really bothers me that the media puts so much reliance on the polls. It’s a vicious feedback loop of media attention generating name recognition generating poll numbers generating more media attention.