|
Got this from someone I know from Alliance for Democracy. It is a preliminary analysis by Lehto, one of the authors of the report.
Preliminary Analysis of Electronic Voting at Penny Creek Elem. School,
Snohomish County, Washington State: Rossi Does Twice as Well As Gregoire on Touch Screen Voting as on Paper Absentee and Provisional Ballots
Question: Based on the data presented below for the five precincts located at Penny Creek Elementary School in Snohomish County, Washington, why do Republicans do so much better on the electronic touch screen voting machines compared to both absentee and provisional ballots?
Snohomish County residents, like all Washington state residents, cast three different types of ballots: Electronic, absentee, and provisional ballots. Both provisional and absentee ballots are reported by the Snohomish County auditor as "absentee ballots" on its canvassing reports, and I confirmed with Elections Dept. staff that "election day" ballot totals consist solely of electronic balloting on Sequoia, Inc. voting
systems, and not of any absentee ballots, even if dropped off at the actual precincts. Because Snohomish County has all Sequoia touch screen voting systems for use only on election day, and all absentee or "early" voting along with provisional balloting is handled via optical scanning of paper ballots, the two methods of tabulating vote totals are on distinct lines in the final canvass report, and we can isolate from the canvassing reports vote totals produced by touch screen methods versus vote totals produced by optical scanning of paper ballots.
(Note: While this paper examines differences in 5 precincts at one
particular voting location and finds substantial differences in the touch screen voting, it does not follow that optical scanning of absentee ballots is foolproof, primarily because the central tabulation of those ballots is done on a single computer, so only one computer needs to be defective or altered to alter the result of the election substantially).
On Election Day, I was a volunteer attorney at the Penny Creek Elementary School, and my understanding was that it contained precincts that trended Democratic. During the day there, I observed provisional balloting rates of very close to 10% of the electronic ballots cast that day, and there were also 144 absentee ballots dropped off at the polling location. This means that in addition to the "election day" touch screen totals, there were about 91 provisional ballots from Election day and 144 absentee
ballots from election day that were nevertheless reported as "absentee" ballots, and not as "election day" ballots, which consist solely of the touch screen electronic ballots. Thus, "paper ballots" subject to optical scan tabulation computers scanned close to a total of 235 paper ballots from Election day. (I say "close" because the provisional poll book number of signatures did not match the number of provisional ballots, so there is some uncertainty as to the exact number of provisional ballots). Thus, the total at Penny Creek Elementary School was close to 922 votes, based on the
above 235 absentee and provisional ballots plus 687 votes recorded on the touch screen systems there, of which there were six.
With these totals in mind, we look at the results of the Presidential and Gubernatorial elections for the years 2000 and 2004: Note: when the term "absentees" is used below with regard to figures from official canvassing reports, it means BOTH absentee votes and provisional votes, regardless of where or when received, so long as postmarked by election day.
PRESIDENT: (Summary: Gore wins in 2000 817 to 748, but Bush wins 2004 over Kerry by a margin of 1060 versus 980).
In the five precincts located at Penny Creek Elementary School, Kerry is listed as having obtained 980 votes (versus 817 for Gore in 2000), split between 655 "absentees" and 325 Election Day votes. Meanwhile, Bush is listed as having received 1060 votes (versus only 748 in 2000), split between 702 absentees and 358 Election day votes. Put another way, Bush edged Kerry by 7.18% in the absentees, while Bush also edged Kerry by around the same amount or 10.15% in the election day voting. However,
Bush's improving over 2000 is more dramatic: Despite heavy turnout, the democratic presidential candidate managed only an increase from 817 votes to only 980 votes, an increase of 163 votes or 19.9% over 2000 performance, while the same Republican presidential candidate (Bush) went from 748 votes all the way up to 1060, an increase of 312 votes, or an increase of 41.7% compared to the last non-electronic voting presidential election PLUS a change in the overall winner in these five precincts.
GOVERNOR: (Summary: Locke beats Carlson 911 to 691, while Gregoire
actually gets fewer votes than Locke at 878 and Rossi skyrockets to 1116 votes). Moreover, there is a large difference between canvassing totals for "absentees" and canvassing totals for Election Day electronic votes.
In the same five precincts located at Penny Creek Elementary School in
Snohomish County, Washington, Democratic gubernatorial candidate Christine Gregoire is listed as having obtained 878 votes (versus 911 for Locke in 2000), split between 599 absentees and 279 Election Day votes, meaning Democrat Gregoire got 33 FEWER votes than Democrat Locke despite heavy turnout, for a performance of - 3.8%. Meanwhile, Republican gubernatorial candidate Dino Rossi is listed as having received 1116 votes (versus only 691 for Carlson in 2000), split between 726 absentees and 390 Election day votes, an increase of 425 votes for the Republican gubernatorial candidate, an increase of 38.1% over 2000 Republican gubernatorial performance. Put another way, Rossi's total exceeded Gregoire's by 21.20% in the absentees, but almost doubled that margin to exceed Gregoire by 39.78% in the election day (electronic only) voting.
In these precincts trending Democratic and still very tight when a Democrat
runs against a heavily entrenched incumbent Republican (like the
Cantwell/Sen. Gorton matchup) the Republican gubernatorial candidate in an open seat is walking away with a landslide 425 extra votes representing a 38.1% increase while the Democrats actually Lose votes for Governor relative to 2000, both percentage wise and on an absolute number of votes level.
ANALYSIS: Obvious errors tend to be caught, because they are very large and stick out, such as the Ohio machine that registered negative 25 million votes. However, moderate to slight errors (whether intentional or not) will tend not to be caught necessarily because it is hard to distinguish improper data from changes in the voter demographic or late-breaking changes in campaigns. Any such slight to moderate errors
that exist would be susceptible of rational explanation by means other than their true cause (i.e. a Republican tide came in, changes in demographics, last minute attack ads, etc.). Thus, the only "errors" that will make it through will be moderate to subtle ones. Another way of putting this "rule" is that only relatively close elections can be "stolen".
CONCLUSIONS: There are always going to be explanations that can be
proffered to explain any set of numbers (weaker candidates, war,
etc.). However, these explanations are not at all sufficient to explain why there is such a large difference in favor of Republicans ON THE ELECTRONIC VOTING MACHINES in precincts that Gore and Locke won in 2000, relative to the paper ballots. A simple "Republican tide" would not explain why the totals on electronic machines differ from "absentee" totals of absentee and provisional ballots by almost double. FOR SOME REASON, more Republican votes are recorded on the Sequoia machines. This could be due to a programming error, deliberate alteration of election day Sequoia data, or that Democrats avoided the touch screen systems in large numbers, by disproportionately voting early.
Note however, that any theory that suggests that latebreaking voters broke for Rossi suffers from the problem that there are 235 paper ballots that came in from the polling location on Election day but were reported in the canvass as "absentees". If there was any late break for Rossi, 235 of the 920 late breaking ballots were placed in the "absentee" column, so the late break for Rossi would have to have been even larger than it appears from the numbers here. Thus, it does not appear very likely that latebreaking voters explains the continuing gap (doubling up from 21.2% to 39.8%) between Rossi and Gregoire on electronic voting relative to paper ballots scanned optically.
A working hypothesis is that something in the counting of the votes,
whether it be a defective computer algorithm, an altered total, a hacking of the system, or defective math caused the Republican gain differential based on the type of voting technology used. It is also possible that Democrats disproportionately (and heavily so) used absentee ballots relative to electronic voting, either because they were more motivated to vote early compared to Republicans, or because they wished to avoid electronic machines in large numbers. If it is the former, and Democrats used absentee ballots much more than Republicans, it is a complete reversal of the general historical rule that Republicans are better at absentees, though this "rule" has admittedly eroded in recent years. But has it eroded to the point of democratic dominance of absentees? If it is the latter (namely that Democrats don't trust the touch screen machines) then there is a serious problem of confidence in the electoral system that has
to be addressed regardless of whether or not a computer defect is found, or even findable.
In any case, the discrepancies between both pre-electronic 2000
performance, and the discrepancies between absentee ballots and electronic voting both argue for more investigation into the general election of 2004, and the possibility that something is wrong regarding the touch screen electronic voting machines can not be ruled out.
Paul R. Lehto, Attorney at Law
LEHTO & PENFIELD, PLLC
2829 Rucker Avenue, 3rd Floor
P.O. Box 1091
Everett, WA 98206-1091
425-257-2297 (voice)
425-258-5041 (fax)
Paul@lehtopenfield.com
|
Printer-friendly format
Email this thread to a friend
Bookmark this thread
(1000+ posts)
Send PM |
Profile |
Ignore
