The Gadfly Bytes -- November 1, 2004
Defeating a "Touch-Screen" Theft of the Election
|"The right of voting for representatives is the
primary right by which all other rights are protected. To take away
this right is to reduce a man to slavery.."
John F. Kennedy
There has been mounting criticism among concerned citizens and computer professionals of the reliability and confirmability of paperless “touch-screen” voting machines (“e-votes”). These machines, the critics charge, present wide-open opportunities for undetectable election fraud.
Thus, if Kerry loses in a close election – in particular, if crucial states with “e-voting” machines give the election to Bush – is it at all possible to determine if “the fix was in”?
Perhaps it can be done through a comparison of separate tallies, first the votes on auditable voting machines, and next of the reported votes on the paperless, secretly coded touch-screen machines.
If the result is a validated landslide for Kerry in the auditable machines, and victory for Bush from an offsetting and unverifiable landslide in the GOP manufactured and coded “e-vote” machines, then our worst fears will be confirmed: We the People of the United States have been had and, unless this crime is remedied, our democracy will be finished.
Fortunately, election returns are open, public documents. All that is required to assemble the separate totals on the paperless e-voting machines (about a third of the votes), and on all the other voting devices, is time and labor. If the result of that study indicates e-vote theft, more refined and compelling statistical analysis might follow, as I will explain below.
THE MENACE OF E-VOTING:
Ever since the essential “workings” of the e-vote machines have become widely understood and appreciated, there have been relentless demands for independent and verifiable “audit trails” of the votes cast on these machines. Corrective responses to these demands have been few and thus insufficient. While California has decertified the Diebold machines, and Nevada has required paper verification of e-votes, legislative remedies, most notably by Congressman Rush Holt (D NJ), have died in the GOP controlled congressional committees.
And so, Californians and Nevadans excepted, the American people have been asked to rely on the “trustworthiness” of the manufacturers and code-writers. There will not be, because there cannot be, independent validation of their votes, if these votes are cast on paperless touch-screen voting machines.
Even so, while validation of individual votes and individual machines are impossible (the Republicans have seen to that), statistical evidence of the validity of e-votes in general are obtainable. I will explain this shortly. But first, let’s review the objections to e-voting machines.
Because the machines produce no independent paper record
of the voting, it is impossible to validate the tally with a recount.
The software that collects, totals and records the votes
is “proprietary” – i.e., secret and the exclusive property of the
manufacturers of the machine. There is no independent assurance that the
vote totals are not systematically altered.
The machines can be easily “hacked” – vote totals changed,
leaving no evidence of the tampering. This is not speculative: several
Digital files from the individual precincts are then
collected in tabulating centers, where there are still further
opportunities for undetectable partisan tampering with the returns.
The owners and manager of the three leading e-voting companies – Diebold, ES&S and Sequoia – are all partisan Republicans.
For a more comprehensive criticism of “e-voting,” see my
Greatest Story Never Told.” A sizeable collection of articles on
the topic can be found at The Crisis Papers’
“Electoral Integrity” page. . If you are still not convinced, watch the
This link will take you directly to it. Warning: this movie could cause
some sleepless nights between now and Tuesday.
STATISTICAL ANALYSIS OF E-VOTING
Republican legislators in the states and in the US Congress have blocked attempts to require independent auditing and validation of e-votes. Nonetheless, while individual votes can not be validated, a collective assessment of validity of e-votes is still possible through statistical analysis.
Exit Polling: This method, which was mysteriously and suspiciously abandoned early in the 2002 vote counting, could be an early-warning of an e-voting “fix.” But it is important that the exit polls be compared with the votes cast on election day. Because of the Democrats’ justifiable suspicions of touch screens, many have opted to avoid e-voting by casting early ballots. Accordingly, the election-day balloting will be skewed toward the Republicans. Comparison of exit polls with totals including early ballots would therefore sound a false alarm.
The "quick and dirty method:" Very soon after the election, a comparison could be produced of the Bush-Kerry vote split in the 30% e-voting states, counties and precincts, on the one hand, and the split in the 70% of "other" returns. This method would not be definitive, for if the e-voting returns yielded sharply higher percentages for Bush, it could be claimed that these machines were in more Republican areas. Nonetheless, sharp discrepancies between the returns of e-voting machines and all others would present a compelling case for a closer statistical analysis.
The September 16th edition of Economist.com points the way. There we find a very helpful color map of the United States, indicating the distribution of the various voting methods. Some states are completely committed to e-voting (Georgia, Nevada, New Jersey, Kentucky). But many are "patchwork" with e-voting among several other methods. Of these, the "battleground states" should be given close attention. The most likely candidates for study, then, are Iowa, Missouri, Arkansas, and those most important states, Pennsylvania, Ohio and Florida.
Refined Statistical Analysis. This method is patterned after classical statistical studies of the deterrent effect of capital punishment. They are basically of two types: parallel and longitudinal.
In a parallel study, two states are selected, alike as possible in terms of socio-economic and cultural factors, differing most notably in that State A has the death penalty and State B does not. If capital punishment is a deterrent, this would show up in lower murder rates in the state with the death penalty.
In a longitudinal study, a single state that has adopted, or alternatively has abolished, capital punishment is examined. If the death penalty deters, then murder rates will drop when capital punishment is adopted, and will rise when it is abolished.
(Most studies, parallel and longitudinal, have indicated little or no deterrent effect. But let''s steer clear of that controversy. We have other fish to fry here).
Now let's apply these techniques to the e-voting question.
Parallel: We select two clusters of precincts or counties which have very similar socio-economic and cultural profiles and which, before e-voting (preferably the 2000 election), had very similar election results. One cluster, ""the control,"" still employs auditable voting equipment (punch cards, optical scanners, paper ballots, etc.). The other cluster uses paperless e-voting. Compare the results.
Longitudinal: Select precinct or county clusters that used auditable methods in 2000 and have adopted e-voting for this election. Still better if they were polled before both elections. There should also be a minimal amount of shift in the population profiles in the ensuing four years. Then compare the results.
Now consider the following scenario: Counting only the 70% non e-voting returns, Kerry scores a 55-43 landslide (2% for ""others""). The e-voting machines alone give Bush a twenty point advantage. Combining both methods, each candidate takes his respective "safe states," and Bush wins, by the slimmest of margins, just enough battleground states to top 270 electoral votes, and to take the election. Kerry, like Gore in 2000, wins the popular vote.
Add to this, continuing job losses, declining median income, soaring health and fuel costs, continuing deficits with offsetting rises in interests rates, a revival of the draft and mounting casualties in foreign wars, international ostracism -- all this the likely result of a second Bush administration and a President stubbornly "staying the course." How long would the public put up with this, knowing that it had been cheated in two consecutive national elections -- that, in effect, their democracy had been stolen from them?
After all, ours is a society that has known freedom and prosperity for the middle class.
SO WHY VOTE?
If the electronic e-vote fix is in, then why bother to vote on Tuesday?
First of all, we don’t know this for certain. So it would be wise to vote on the chance that the unverifiable, paperless voting machines are, for all that, on the up-and-up.
Second, whenever possible, opt for the paper ballot rather than the e-vote machine. (We can do this in California. Presumably elsewhere as well).
Third, if your ballot is auditable (not an e-vote), then by all means vote so that we can rack up a large auditable Kerry majority. The larger that majority, the more conspicuous the fraud built into the e-voting software. The more conspicuous the fraud, the less legitimate the stolen election.
Finally, if the Diebold-ES&S-Sequoia cheaters underestimate the strength of the Kerry tide, their "fixes" might not withstand the flood of votes.
So by all means, get out and vote on Tuesday! If the e-voting “fix” is in, your vote will enhance the statistical evidence of a stolen election. If the e-voting is fair and accurate, then, well, a vote for Kerry is a vote for Kerry.
Then, post election, let’s all do our damndest to outlaw those diabolical non-auditable voting machines.
Copyright 2004, by Ernest Partridge
Conscience of a Progressive: A book in progress.
Partridge's Scholarly Publications. (The Online Gadfly)
Dr. Ernest Partridge is a consultant, writer and lecturer in the field of Environmental Ethics and Public Policy. He has taught Philosophy at the University of California, and in Utah, Colorado and Wisconsin. He publishes the website, "The Online Gadfly" and co-edits the progressive website, "The Crisis Papers".