Please, a small unobtrusive "CaT" diary.
No Martians, no commies at Department of State. Nothing but stat for the CaT. And a worry that Hart InterCivic and other RW investment shops have bought their way in to the voting machine industry.
36% of the registered electorate were recorded for voting.
The lowest percentage tallied since 1942 when we already had millions overseas and in training for WW II.
Presidential turn out:
2000 -- 54.2%;
2004 -- 60.4%;
2008 -- 62.3%; and
2012 -- 57.5%
Congressional turnout:
1998 -- 39%;
2002 -- 41%;
2006 -- 41%; and
2010 -- 42%
Republicans skip voting in 2012. Then Democrats go away en masse in 2014. Really ???
The basis for this piece goes back to June of 2010 when the polling operation for Dailykos -- one Research 2000, Inc., out of Maryland -- was discovered to have been faking its numbers. In short, the numbers "R2K" provided were too good. Too smooth, too well matched.
Statistically not quite believable.
Research 2000: Problems in plain sight
A bit over two weeks ago, a group of statistic wizards (Mark Grebner, Michael Weissman, and Jonathan Weissman) approached me with a disturbing premise -- they had been poring over the crosstabs of the weekly Research 2000 polling we had been running, and were concerned that the numbers weren't legit.
I immediately began cooperating with their investigation, which concluded late last week. Daily Kos furnished the researchers with all available and relevant information in our possession, and we made every attempt to obtain R2K's cooperation -- which, as I detail in my reaction post here -- was not forthcoming. The investigators' report is below, but its conclusion speaks volumes:
We do not know exactly how the weekly R2K results were created, but we are confident they could not accurately describe random polls.
-- kos (June 29, 2010)
The R2K polls matched odd/even pairs, result sets that lacked outliers, chi-square measures on variance in the distribution, and anomalies that violated the usual Law of Large Numbers properties for random samples from normal populations.
Markos had lots of help.
Mark Grebner is a political consultant.
Michael Weissman is a retired physicist.
Jonathan Weissman, a wildlife research technician.
With expert assistance:
J. M. Robins.
J. I. Marden.
They pretty well nailed it for Research 2000.
- - - - - - - - - - - - -
What strikes me right off, looking at the Senate and Governor election voting patterns of 2010 and 2014 is that they also appear to be "well behaved." It's not just that Democrats don't come out and vote. They don't do it anywhere.
The turnout in Georgia resembles the turnout in Kentucky.
Races like Kansas, where Republicans have wrecked the state, still find Democrats A.W.O.L.
A few governors hold, a few Senators but with huge fall-offs in Dem voting.
Out on Long Island we had a Republican, Michael Grimm, running for office while he awaits trial for multiple felonies. Yep, those Democrats didn't show up either.
What's happening, if anything ??? Is there reasonable evidence, statistical measures to support a position that the voting machine industry is producing artificial numbers?
That American democracy is, in a word, history?
Matlab's cheap. I've got SPSS, SAS/ETS and more around here somewhere. MySQL works. Gotta figure a million rows of Excel could handle it. Whatever it takes....
The whole country winked out for Dems ?????
Are all of these figures real voting numbers?
Anyway, I would feel better if our voting systems get a solid round of quality control checks. Bill Clinton saw Democrats win seats in the House and hold even in the Senate in his 6th Year midterm. So don't get lost that 6th Year midterms have a kiss of death attached. (That was Bush in 2006.)
What do we need for a database? Can we do a "Research 2000" pattern of analysis on the 2010 and 2014 vote numbers?
HELP !!