Tyler Sperry wrote:
> I don't know about German exit polls, but in this year's US election
> the pollsters explicitly warned before the election that their results
> would have a large margin of error and were designed to be descriptive
> rather than predictive. Perhaps the US pollsters were using different
> types of questions and a different methodology? Or perhaps the US
> polling firm should hire some German pollsters next time?
Don't confuse the telephone call polls with exit polls. Exit polls means
that you ask people who just have voted. This takes most of the errors
out, except for the postal voters (who can't be exit polled). Our main
source of exit poll errors are these postal voters, since they voted
two weeks before, and they are not a very representative sample of the
population. Absentee vote is up to 10%, and that could dillute the exit
poll accuracy by 1% or so.
> Do you have a source for that assertion? US newspapers performed a
> review for all Florida counties. Using Gore's vote criteria statewide,
> Bush actually picked up votes:
> <http://archives.cnn.com/2001/ALLPOLITICS/04/04/florida.recount.01/>
IIRC, that's true. The recounter however also used the supreme court
criteria, i.e. recount everywhere. It's that part of the supreme court
decision that makes sense: If you recount somewhere, you have to
recount in the whole state. You can't just drop votes here, because you
are not interested, and recount there, because you are. Funny enough,
the supreme's court rule would have put Gore ahead, whereas his own
selfish and selective recount attempt would have put Bush ahead. I
don't have an URL ready, since any search for that is poisoned by too
many pages ;-).
The law is quite clear: voter's intent counts. And games about
recounting yes/no here/there are silly. In Germany, a recount is
mandatory, and noone can stop it. It's the right of every citizen to
have his (or her) vote counted. The recount on our last election, where
both big parties were only 6000 votes apart first did change that
margin to ~8000.
>> but this time, they made
>> sure that noone will be able to recount. They left no paper trail
>> behind.
>
> Which "they" are you talking about?
The voting machines which do not leave a paper trail behind. However,
apparently the scanner voting machines from Diebolds (which allow
manual recount) produced more dubious results than the touch screen
machines.
> Are you aware that the most
> problems in Florida's 2000 election process were found in counties
> that were overwhelmingly Democrat? Or that the so-called "confusing"
> butterfly ballots had in fact been previewed and approved by both the
> county's Democratic party officials and the public at large?
Hm, the lady in question is no member of the Democratic party anymore,
and AFAIK wasn't in 2000, too (she left the democrats before). She
managed to produce a confusing ballot this year, too.
> Pollster John Zogby, a Kerry sup****ter who also forecast a narrow
> Kerry victory, has remarked "the exit polls were terrible."
>
<http://seattletimes.nwsource.com/html/politics/2002081763_exitpolls04.html>
Well, even if the exit polls are inaccurate, there's no explanation why
they did bend all to the same direction. If you have 10 swing states,
where 9 were exit-poll-predicted for Kerry, and an exit poll accuracy
of 3% in each one, the likelyhood that most of these states go to Bush
(as they did) is still very small.
I think this observation is a post fact one, i.e the pollster assumes
that the official votes are accurate, and therefore the exit polls were
terrible. But to have 6 or 7 out of 10 swing states predicted wrong,
the exit polls have to be far worse than terrible. They need a
systematic error.
> Actually, given both the influence of the rural ("red") states
> in deciding the election and the fact that Germans aren't allowed
> to decide who leads Italy or the US, I rather suspect they'd be
> pleased. (Hessians weren't beloved in the US 200 years ago.)
Well, most of the German emigrants in the USA live in the "red" states,
and we have quite some in Bush's team, especially hardliners like
Rumsfeld. I still agree to the US voices that say that you didn't know
what Kerry wanted, because Kerry had an ABB strategy (anyone but Bush),
and tried to embrace everyone and nobody. Actually, Bush used the same
strategy last term, and Gore tried, too, so they met as faceless
lookalikes. Politicians try to do that all the time; we had a good
laugh when our last two candidates even wore the same tie. The only
point were they differed (Irak war no or maybe) was the point that made
the winner. The "maybe" candidate lost.
--
Bernd Paysan
"If you want it done right, you have to do it yourself"
http://www.jwdt.com/~paysan/
|
