Thanks, Bill. It looks to me on the Archive as though it's all there, but for some reason there were no carriage returns. Perhaps this message has the same flaw.
-----Original Message-----
From: Bill Williams, UMKC [mailto:w.d.williams@EMAIL.RO]
Sent: Wednesday, December 18, 2002 1:19 PM
To: CSGNET@LISTSERV.UIUC.EDU
Subject: Giffen Data?
Mike,
Somehow your email came through with the message clipped. Take a look at
how it looks on the CSGnet archive.
Bill
[From Bruce Gregory (2002.12.18.2040)]
Rick Marken (2002.12.18.0900)
I'm also not sure that one's imagined place in the income
distribution has much to
do with the "interests" that determine how one votes. I think that,
as long as
things aren't going too badly for too many people, people probably
just vote for
the candidate who seems to be most like oneself in intelligence and
self-righteousness;-)
The best explanation I've heard.
--
Bruce Gregory, bruce@hammersofhell.com on 12/18/2002
A stranger in the Land of the Clueless, Bruce Gregory lives with the
poet and painter Gray Jacobik in self-imposed exile in Pomfret,
Connecticut.
people misreporting their standing
in the income distribution.
[From Bill Powers (2002.12.18.1920 MST)]
It can’t be “just ignorance.” Something is going on
which results in
It could also be that they are correct. People who listen to NPR may be
more highly educated than average, plus other factors which combine to
put them in higher income categories. So if NPR listeners tend toward the
high end of various distributions, it could be true that they all
are in the top 1/5 of the income category, relative to the general
population. The sample is skewed.
Best,
Bill P.
[From Bill Williams UMKC 19 December 2002 1:10 AM CST]
Bill Powers wrote: "The sample is schewed."
This may indeed account for what was reported, they only way to know
would be to replicate the study by controlling for the subjects actual
income level. It was my understanding that Those presenting the report had
interviewed people as a part of voter exit polling. Voters are themselves
a schewed self-selected group. However, if it turns out that half of the
population actuall does think they are in the top 1/5 or 1/10 of the income
distribution then we'd have an explaination for why economic measures that
are of benefit to so few have the support of voters.
Or, maybe in line with, or similiar to something Rick suggested, people may
think that those who have high incomes are so admirable that they ought to
have even higher incomes.
Best
Bill Williams
interviewed people as a part of
voter exit polling.
[From Bill Powers (23002.12.19.1016 MST)]
Bill Williams UMKC 19 December 2002 1:10 AM
CST
It was my understanding that Those presenting the report had
Oh, not NPR listeners. That makes a difference. Voters are more likely, I
would think, to be representative of a broad range of income. Anyone need
an old used theory?
Best,
Bill P.
Remember the Time proclamation "Dewey beats Truman!"
It was a well conducted poll with one minor exception. It was a telephone poll and in 1948 most people did not have telephones. Only the 'well to do' could afford them. Coincidentally, the 'well to do' tend to be Republican. Thus the poll was skewed.
I don't recall the source of this threads poll. But before we charge off trying to explain it we should verify it's accuracy.
Steve O
-----Original Message-----
From: Bill Williams, UMKC [mailto:w.d.williams@EMAIL.RO]
Sent: Wednesday, December 18, 2002 4:15 PM
To: CSGNET@LISTSERV.UIUC.EDU
Subject: Giffen Data?
[From Bill Williams UMKC 18 December 2002 2:40 PM CST]
[ Rick Marken 2002.12.18.0900] wrote about most people reporting that their
income was in a top bracket,
I don't think it's a delusion, really. It's just ignorance.
It can't be "just ignorance." Something is going on which results in
people misreporting their standing in the income distribution. It's like
the kids on the Prarie HOme Companion who are "all above average." But, it
is something that wouldn't be all that difficult to replicate. Or, who
knows maybe I was delusional when I only thought I heard it on NPR.
best
bill Williams
[From Mike Acree (2002.12.20.1020 PST)]
Stephen O'Shaughnessy (2002.12.20.1016)--
I don't recall the source of this threads poll. But before we charge off trying to
explain it we should verify it's accuracy.
For the record, the notorious telephone poll was conducted by _Literary Digest_ for the 1936 election, not the 1948, as Steve said, or 1932, as Bill Williams said.
Mike
[From Fred Nickols (2002.12.20.1635)] --
[From Mike Acree (2002.12.20.1020 PST)]
Stephen O'Shaughnessy (2002.12.20.1016)--
>I don't recall the source of this threads poll. But before we charge
off trying to
>explain it we should verify it's accuracy.For the record, the notorious telephone poll was conducted by _Literary
Digest_ for the 1936 election, not the 1948, as Steve said, or 1932, as
Bill Williams said.
Which notorious telephone poll. Certainly not the Dewer-Truman poll. To
which one are you referring?
Fred Nickols
nickols@safe-t.net
[From Mike Acree (2002.12.20.1357 PST)]
[Fred Nickols (2002.12.20.1635)] --
Which notorious telephone poll. Certainly not the Dewer-Truman poll. To
which one are you referring?
The _Literary Digest_ telephone poll in 1936, which predicted a landslide by Alf Landon (because Republicans owned most of the phones then), has become a textbook example of how not to conduct a survey. Willams and O'Shaughnessy were correct in remembering the nature of the problem, just wrong about the year; and O'Shaughnessy had asked about the source.
Mike