|
STATEMENT TO TEXAS HOUSE COMMITTEE ON ELECTIONS
Richard Hayes Phillips, Ph.D.
April 27, 2009
My name is Richard Hayes Phillips. I am an educated man, but that is not
what makes me an expert on elections. I was the leading investigator of the
fraudulent 2004 election in Ohio. We photographed or photocopied 126,000
ballots, 127 poll books, and 141 voter signature books from precincts of my
choosing in 18 suspect counties. I analyzed them all myself. My book,
“Witness to a Crime: A Citizens’ Audit of an American Election,” has sold
more than 1000 copies and turned a profit with no grants, no subsidies,
and no bailouts. Never before in American history has an election been
so scrutinized by citizen volunteers without the help or concern of the
government or the news media. Much of what we found in Ohio is relevant
in Texas today.
Electronic voting machines were used in seven counties in Ohio in the 2004
election. There was no paper record of the votes in these counties, accurate
or otherwise. But accounts from the voters themselves, often in the form of
sworn affidavits, were borne out by detailed analysis of the official vote
count at the precinct level.
In Youngstown, Ohio, some electronic voting machines were programmed to
default to candidate Bush. The Board of Elections reported that twenty to
thirty machines were shut down for this reason, but the fraudulent vote count
could not be corrected. And these were the machines that were not only
rigged but were also malfunctioning, on which the default choice of the
computer actually appeared on the touch screen. In one precinct, three
voters complained that their votes hopped from Kerry to Bush, on two
different voting machines, and three other voters complained that they never
saw a presidential choice appear on the screen. And yet there were no
“undervotes” – that is, votes cast with no choice for president. Every voter
who touched the screen had a vote counted for President. The fraud was
proven only through precinct results. I never saw the computer tapes, and
I would not have been able to decipher them anyway. The “recount” staged in
December 2004 consisted of pushing the “print” button one more time.
In Columbus, Ohio, some electronic voting machines would not always record
a vote for Kerry. Many voters reported that when they pushed the button for
Kerry, a light would appear next to his name, and then would fade from the
screen. Precinct analysis showed that the rate of “undervotes” in the 2004
election was much higher than in 2000, which is consistent with the
complaints of the voters. On one machine, no votes were recorded at all, for
Bush or Kerry. “Don’t worry,” a poll worker was told, “they’ll figure it out
downtown.” I don’t know how they could have figured it out, as there was no
paper record of the votes.
Optical scanners were used in thirteen counties in Ohio in the 2004 election.
We audited two of these counties. We photographed actual ballots, poll
books, voter signature books, and ballot accounting charts, and compared the
primary evidence to the official results.
In Miami County, the official results failed to match the poll books and
the voter signature books in all 82 precincts. The number of ballots cast,
officially, according to the optical scanners, always differed from the
number of actual voters. Sometimes the difference was one or two votes,
but there were discrepancies of 38, 52, 53, 69, 137, 139, and 263 votes in
certain precincts. A comparison with the computerized database of voter
histories turned up two other precincts with discrepancies of 104 votes and
119 votes. If the wrong number of votes have been recorded, the vote count
for the candidates cannot be right.
Also in Miami County there were 570 “remakes.” We photographed them all.
These “remakes” were ballots filled out by the Board of Elections, ostensibly
to duplicate ballots which could not be fed through the optical scanner,
although no audited county could produce the original ballots which their
“remakes” allegedly duplicated. Among the 570 “remakes” were 79 ballots with
no choice for president, an impossibly high rate of 13.9%, twenty times
higher than the countywide rate of 0.67%. The optical scanner cannot
distinguish between fake ballots and real ones.
In Clermont County, the favorite method of vote rigging was to take a ballot
marked for Kerry, cover the mark with a little white sticker, fill in the
mark for Bush, and then run the ballot through the optical scanner. Three
observers to the “recount” staged in December 2004 observed these “stickered
ballots” and confronted the Board of Elections with sworn affidavits. By the
time we were allowed to photograph the ballots a year and a half later,
nearly all of the “stickered ballots” were gone, replaced by pristine remakes
masquerading as valid ballots. The Board of Elections had to dip into their
stash of unused ballots in order to do this, which forced them to destroy
their unused ballots because there were not enough of them remaining to
survive an audit.
Also in Clermont County, “overvotes” – that is, ballots with two or more
choices for president, in this case for Bush and Nader, were counted for
Bush. At the time that Nader was stricken from the ballot by the courts,
many counties, including Clermont County, had already printed their ballots,
so they were allowed to post signs in the polling places announcing that
votes for Nader would not be counted. But instead of programming their
optical scanners to count the Nader column as zero, as was done in the punch
card counties of Ohio, Clermont County programmed their optical scanners not
to recognize a mark for Nader. This enabled election officials to fill in a
mark for Bush on the same ballot, knowing that the optical scanner would read
it as a vote for Bush and not as an “overvote.” If an optical scanner can be
programmed not to read a mark for a candidate, it can be programmed not to
read every tenth mark, or every twentieth mark, and thus throw an election.
There is no way to verify the results without hand counting the ballots.
In four counties in Ohio, in nearly every precinct audited in those counties,
we photographed more ballots than the total number issued to the voters. Let
me repeat that. We photographed more ballots than the total number issued to
the voters. Even if every absentee ballot issued was returned by the voters
in a timely manner; even if every provisional ballot issued was ruled valid
and counted; and even if every voter signature was legitimate – there were
still too many ballots. This is called “ballot box stuffing.” Some of these
ballots were never touched by voters. And if some of the ballots are fake,
how do we know that any of the ballots are real? This happened in counties
that used punch cards and in counties that used optical scanners. The
problem was not the paper ballots themselves, but breaks in the chain of
custody. Someone had a chance to alter the ballots. In one of these
counties, Darke County, we photographed punch card ballots with absolutely
nothing printed on the back side – not the name of the precinct, not even the
name of the county. These were “wild card” ballots that could be used on an
as-needed basis in a scheme of old-fashioned ballot box stuffing.
In Warren County, Ohio, under cover of a fake homeland security alert, punch
card ballots were brought to an unauthorized warehouse for three to four hours
after the polls closed, during which time some of these ballots were shifted
from precinct to precinct, which altered the vote count because, in Ohio, by
law, the sequence in which candidates’ names are listed on the ballot must
rotate from precinct to precinct. Only then were the ballots brought to the
Board of Elections and run through the electronic tabulators, in secret, with
the county building locked down and surrounded by police cars and bomb-sniffing
dogs. This is not theory. We know from e-mails and eyewitnesses that the
lockdown was planned in advance by county officials; we know from the ballot
accounting charts that the last ballots left the polling places three hours
before the last ballots arrived at the Board of Elections; and we know from our
own photographs that ballots were issued at the polls with no precinct printed
upon them, thus making the fraud possible. Again, there was a break in the
chain of custody. Someone had a chance to alter the ballots. The problem lies
also with central tabulation. If the ballots had been counted at the polling
places, in full public view, the fraud would not have occurred.
Herein lies the fundamental problem with random audits, or even with full
recounts. If the optical scanners and electronic tabulators are programmed
honestly and functioning properly, and we merely recount the same fraudulent
ballots, we will get the same fraudulent results. Unless we first examine
the voter signature books, we do not know if we are looking at the right
number of ballots. Unless we first examine the ballot accounting charts,
and count the unused ballots, we do not know if all of the unused ballots
are accounted for. If not, then the opportunity existed to substitute fake
ballots for real ones, either before they were run through the optical
scanners or electronic tabulators, or in an after-the-fact attempt to get
the ballots to match a rigged tabulator count.
In the 2006 mid-term election, punch cards were no longer in use in Ohio.
All 88 counties used Diebold touch screens or ES&S (Election Systems and
Software) optical scanners. Both of these were programmed – not hacked,
programmed – to produce two different sets of numbers for total ballots cast.
Imagine how the IRS would react if you told them that you keep two sets of
books. The two parameters produced by the machines are called “Cards Cast”
and “Times Counted.” The smaller of these numbers may or may not be correct.
The higher number is derived by counting absentee ballots twice, thus opening
a window of opportunity to alter the vote count. If the higher number is
reported as total ballots cast, and these numbers do not carry over to the
vote count for the candidates, we are left with undervote rates as high as
25%, or with reported voter turnout above 100%. These numbers are reported,
with a straight face, by Boards of Elections. Cuyahoga County reported that
25% of the voters made no choice for United States Senator, and that voter
turnout was above 100% in sixty precincts. Summit County reported 110% voter
turnout countywide. Cuyahoga County used Diebold touch screens, and Summit
County used ES&S optical scanners. If we do not know how many actual voters
there were, then we do not know if the right number of votes was counted, and
we certainly do not know if the votes were assigned to the right candidates.
If these phony numbers do carry over to the vote count for the candidates,
then the numbers reported will not appear to be irregular, and we have no way
to verify the count without examining each and every voter signature book and
each and every ballot, which raises the fundamental question: Why not
dispense with the machines and go with hand counted paper ballots in the
first place? No other voting method on the table in Texas can produce a
verifiable count.
Our elections must be free, fair, transparent and verifiable. Elections
belong to the people, not to the government, and certainly not to the vendors
of voting machines.
Having an “open source code” for electronic voting machines and electronic
tabulators will not suffice. This leaves election oversight entirely in the
hands of computer programmers. Election transparency requires that any
citizen should be able to audit the election by examining the actual marks
of the voters, on paper, in an easily understandable format.
We have a right to verify the count, immediately, whenever it is questioned.
Before any candidate claims victory or concedes defeat, we need to know that
all votes have been counted as cast, because every vote matters, and because
elections belong to the people.
I am irrevocably opposed to electronic voting machines, to optical scanners,
and to centralized tabulation of results. We need paper ballots, counted by
hand, at the polling place, in full public view, on election night, no matter
how long it takes. Nothing else will ensure a verifiable vote count.
Without PAPER BALLOTS, no election results can be verified by auditors.
“Voter-verified paper trails” do not suffice, because: (a) not all voters
will take the extra time to “verify” each and every vote for a multitude
of offices and initiatives; (b) there is no way to ensure that the vote count
reported by the machine actually matches the “paper trail” without looking at
the paper record; (c) contentious court cases will be required in order to
look at the paper record during the brief time period within which an
election can be challenged; therefore (d) the unverified machine count will
likely become official.
The ballots must be COUNTED BY HAND. I have examined more than 126,000
ballots from the 2004 general election in Ohio, and have found large
discrepancies, errors, and outright fraud that were not detected by the
machine counts, including: (a) hundreds of consecutive ballots cast for
the same presidential candidate; (b) outright alteration, such as marks for
certain candidates covered with white stickers; (c) far more, or far fewer,
ballots than the number of actual voters; (d) “wild card” ballots with no
precinct printed upon them; and (e) fake ballots substituted for real ones.
The ballots must be counted AT THE POLLING PLACE, because: (a) hand counting
of paper ballots is not manageable at the county level; (b) the counting
takes place before “chain of custody” issues have arisen; (c) all ballots
will be counted in the correct precincts.
The ballots must be counted IN FULL PUBLIC VIEW, because: (a) nothing else
will restore public confidence in the veracity of the vote count; (b) public
monitoring of the actual counting of ballots is an exciting opportunity for
citizens to engage in participatory democracy; and (c) if observers are
required to be local residents, contentious challenges from partisan lawyers
can be averted.
The ballots must be counted ON ELECTION NIGHT, NO MATTER HOW LONG IT TAKES,
because: (a) again, the counting takes place before “chain of custody” issues
have arisen; (b) the voters have already endured a lengthy campaign, and can
wait a while for the unofficial results; (c) the right of voters to have
their votes counted correctly, the first time, trumps the desire of poll
workers to go home early; and (d) if poll workers are tired after a long day
at the polls, a new team can be brought in to count the ballots.
Do not fall for the counter-argument that hand counted paper ballots is not
practical in urban counties. The ballots need to be hand counted at the
polling place by poll workers, as was common practice not so long ago. It
is no more difficult to count 700 ballots in an urban precinct than to count
700 ballots in farm country.
Do not be deceived by the lure of modern technology. Elections belong to the
people, not to the voting machine vendors and the computer programmers. Do
not be dissuaded by the notion that hand counted paper ballots are old
fashioned. I am a mountain man. I have been hiking the Adirondack Mountains
since I was four years old. And I learned as a young boy that when you are
on the wrong path, a step backward is a step in the right direction. The
Texas state legislature is to be commended for being in the forefront on this
issue, for seriously considering hand counted paper ballots, for having the
courage to look forward by looking backward.
All the forms of ballot rigging that we found in Ohio were due to breaks in
the “chain of custody.” Crime scene investigators look for three things:
motive, means, and opportunity. There will always be a motive to rig an
election and win the count. There will always be a means – whatever voting
method is used. Our only hope is to stop the opportunity.
That is why paper ballots need to be hand counted at the polling place,
before any “chain of custody“ questions have arisen, before the opportunity
arises for ballot alteration, ballot substitution, ballot box stuffing, and
ballot destruction, all of which we documented in Ohio. Paper ballots are
the solution, not the problem – but only if counted by hand, at the polling
place, in full public view, on election night, no matter how long it takes.
NOTE: After having taken one bus and three planes to get to Austin, Texas,
I was not allowed to read this testimony. Officer Joy, in the hallway,
explained to me how things are done in Texas: “You have the right to be
angry, but you don’t have the right to free speech. It’s the chairman’s
prerogative.”
|
|