Site Map
Voting News
Contact Us
About Us

is NOT!
associated with

S.1487: A Deconstruction
By Robert Bancroft, 6/19/07

Printable pdf
I am not a lawyer, nor a politician. I am neither an anchorman, nor a journalist. My qualifications to write this are few, it would seem. I am, at least, a citizen of this great nation, a citizen who does not appreciate his public servantsí attempts to meddle with our right to vote. It is my modest hope that this, alone, will warrant further reading.

In their own words
According to the authors of S.1487, the principal purpose of the bill is to modify the Help America Vote Act (2002) "to require an individual, durable, voter-verified paper record [Ö] and for other purposes." Other purposes include a "moratorium on acquisition of certain direct recording electronic voting systems", the promotion of accuracy and integrity in the voting process, the requirement of manual audits, the establishment of new grants "to replace or retrofit" non-compliant equipment, and imposing "additional requirements for Federal Elections." It all sounds good.

What is a voter-verified paper record?
For the purposes of this bill, such a record is essentially a mock-receipt. Voters first verify their vote on a computer screen, and then receive a printout which they can also verify, ultimately casting their ballot when ready. In the case of a recount or audit, the printout is considered the "true and correct record." Yet, the printout may be ignored, at the discretion of the State, whenever there is the vaguest fear of compromise by "damage or mischief or otherwise." (SEC.201.(A)) The American people have serious concerns about this charlatanism.

It is misleading to suggest that, because a voter has verified a paper printout, that he or she has verified the vote itself. Under the proposed regime, the voter interacts with a user interface (touch screen), which displays a representation of the ballot. Based on secret code, the machine translates what is on the screen into binary data, unknown and unreadable to the human voter. Later, the machine, interpreting this data a second time, generates a printout, a second representation.

Think of it as an artistís rendering of a vote. Rather than verify the vote itself, which, as it turns out, would be impossible, the voter simply compares two representations. That is what it means to vote using a direct record electronic (DRE) machine. The ideal of an open or transparent election is expressly precluded, and the very process of vote counting is declared intellectual property, a trade secret.

It ainít just semantics.
To understand the distinction between actual votes, and representations of votes, consider the report entitled "DRE Analysis for May 2006 Primary, Cuyahoga County, Ohio", published by the ESI in August that year. The report compares a hand count of receipts with: printed totals, electronic totals, and a manual inspection of memory. If the prescription of this bill is reasonable, then those figures should all agree. After all, each receipt was "voter-verified." Instead, this is what the ESI found:

16% of the time, printed totals do not match a hand count of the paper receipts
72% of the time, printed totals do not match the electronic totals
26% of the time, the electronic totals do not even match the memory

In the end, an inspection of the votes in memory differed from a tally of the printed receipts a staggering 76% of the time. If the printout will match the real vote only one time in four, how could Congress possibly believe that a voter-verified paper trail offers any safeguard at all?

Who is my Commission?
Most Americans are not familiar with the Elections Assistance Commission (EAC). It was established in 2002 to usurp the Office of Elections Administration, an independent agency that had served the people in a limited role since 1974. This new Commission, comprised of four individuals appointed directly by the President, absorbed the old FEC appendage, and gained several new responsibilities. Yet the Commission was made temporary, tasked primarily to help States comply with new standards and administer federal grants for new equipment.

In this bill, Congress proposes to greatly increase the Commissionís scope, and install it as a permanent fixture with vast discretionary power over federal elections. One might think that allowing a President to directly appoint those officials that may oversee his own re-election is inherently perilous. Yet, this bill does exactly that. Here is a partial list of some of the new powers Congress has in store for its Commission-unseen:

Set standards for equipment accuracy (SEC.201.(a)(4))
Set different benchmarks for different peoples (SEC.201(a)(4))
Determine whether resources are distributed fairly (SEC.304.)
Determine which machines and software may be used (SEC.201.(b))
Determine which labs are used for testing (SEC.201.(c))
Control the certification process (SEC.201.(b))
Ignore the certification process (SEC.201.(b))
Decide who is allowed to observe elections (SEC.307)
Decide who is allowed to see the secret source code, within the machines (SEC.201.(b))
Work closely with corporations, to protect their interests (SEC.201.(b))

In reviewing the list, remember that: (a) these would be new powers granted to the Commission by this bill, (b) the Commission is accountable to no one, and (c) to-date the Commission has a poor track record, fraught with failure, corruption, and secrecy.

Jim Crow, meet Mark Bench
When it comes to accurate vote counts, the very first thing this bill does is to delete the standards set in law today. Instead, the bill suggests that the Commission can decide what error rate is acceptable. It can change its mind too, any time it likes, and for no reason at all. In other words, there are no standards, anymore.

The bill also defines a "residual ballot performance benchmark" (SEC 201.(a)(4)). This measure includes not only errors that are directly machine-attributable, but such anomalies as overvotes, undervotes and spoiled votes. The authors are wise to include these additional forms of error in the benchmark. But then things turn sinister.

The authors opine an adjustment to this, which they call an "intentional undervote." Setting aside, for a moment, that Congress is incapable of knowing the "intentions" of any particular voter, let us consider this bizarre social phenomenon. Why would droves of American citizens take time away from their work and their families, brave the perils of rush-hour traffic, spend hours in line, and then, when their moment arrives, intentionally cast a legally invalid vote? Nevertheless, historically, roughly 1 in 200 Americans do exactly this, and vote no one for President (www.votersunite.org).

Rather than include the established 0.5% figure, this bill allows the Commission to subtract whatever sort of undervote correction it likes from the benchmark. Unfortunately, the machines that Congress continues to push on constituents are notorious for producing high undervotes, sometimes greater than 20%. With this little loophole, the Commission can sweep that unpleasant fact right under the rug.

The authors further explain that "Congress finds that there are certain distinct communities [Ö] that have historically high rates of intentional undervoting." What? It goes on to establish special "treatment of certain distinct communities." If the Commission discovered that black voters in a particular precinct had suffered an unusually high undervote, rather than investigate that as an apparent problem, it could instead declare that any vote by an African-American, anywhere in the country, is thereafter exempt from any standards at all (SEC.201.(a)(4)).

Notwithstanding the troubling racial undercurrent of this venomous yet sophisticated "Jim Crow" law, there remains a practical question. The bill affirms that "States may not exceed" whatever benchmark is established for them. Indeed, what should be the remedy in the case of a State that does exceed its benchmark? Would its votes not be considered in the federal election at all? Would Congress call a "do-over"? The bill provides no answers. Perhaps the authors hope that the Commission will figure it out.

Letís have an Audit
It is difficult to provide much commentary on the billís audit standards because the bill does not have many. It proposes the weakest standards of any of the voting reform legislation under consideration (2% of all precincts must be audited). While it suggests that the precincts should be selected in a random manner, it does not preclude States from determining ahead of time which precincts will be audited, creating an obvious loophole. For the most part, the details are all left to the Commission. Even though the need for such rigorous auditing largely arises from Federal pressure to use unreliable electronic devices, neither Congress nor manufacturers will lend a dime to offset these new costs.

About Dimes
One cannot help but notice a pattern in the fiscal priorities set by this bill.

Money to help States develop training programs: none
Money to help States establish early voting programs: none
Money to help States perform rigorous audits: none
Money for the Commission to give to manufacturers for research: $4,000,000
Money for States to give to manufacturers for machines: $600,000,000

The bill is quick to clarify that the $600 million grant in no way prevents States from also tapping the $650 million granted by HAVA. That puts the total amount of American taxpayer dollars siphoned off to voting-machine manufacturers at $1.25 billion. Are the authors certain that the American people agree with these priorities? Even after handing so much of our money over to these business interests, the bill still does not provide the one thing we have requested: a paper ballot.

If you donít have anything Nice to sayÖ
While the bill recklessly cedes control of federal elections to a group of four Presidential appointees, prioritizes corporate interests above the integrity of the voting process, and legalizes racism at the polls, there are some positive changes introduced by the bill, and most of them are listed under Title III Ė Improving Federal Elections. These include:

Guarantees for absentee voting
Support for 3rd-party registration assistance
Training of poll workers
Equitable allocation of voting resources
Limitations on conflicts of interest between chief election officials and campaigns
Reasonable protection from voter purging
Some assurance of election observers
Requirements for early voting
Requirements to count provisional ballots
Reasonable protection for ballots cast by uniformed and overseas citizens

Unfortunately the assurance of election observers is a superficial one, since it is left completely to the Commissionís discretion who can observe and who cannot (SEC.307). Similarly, it is left to the Commission to ultimately determine what constitutes an "equitable" distribution of resources, and the bill explicitly authorizes the Commission to use past inequities ("voting patterns") to justify future unfairness (SEC.304.(b))

There is also a troubling omission from the descriptions of all of these improvements. When do ballots get counted? Polling results have an influence on turn out, as voters begin to form impressions of who the "winner" is. If the results of early voting were made public, it would almost certainly influence the election. Likewise, when are provisional and absentee ballots counted? After the television has already declared a "winner"? The question of timing is not inconsequential to the integrity of the process. This leads some to question whether there is any merit to early voting at all.

Whoís your Daddy?
Because S.1487 is essentially a modified version of Rep. Holtís H.R.811, it is instructive to compare this bill to its predecessor, and note what was changed. The devil is in the details, or in this case, the changes:

Votes are less safe, and less likely to be counted, in S.1487:
o S.1487 replaces most references to "paper ballot" with "paper record", a stinging reminder that the two are not equal.
o S.1487 removes the test of "clear and convincing evidence" with regards to tampering, making the paper records easier to ignore.
o S.1487 allows the paper ballots of an entire precinct to be ignored, if there is any hint of "mischief", while H.R.811 suggests that any tampering be considered on a machine-by-machine basis
o S.1487 deletes language in H.R.811 that requires prominent reminders for voters to double-check their paper records before casting the vote.
o S.1487 adds dangerously racist "residual benchmarks", and deletes accuracy standards established by HAVA.
o S.1487 alters disclosure requirements, to offer improved protection for corporate "trade secrets", at the expense of vote integrity.
o S.1487 permits some machines used in vote tabulation to be connected to the internet, while H.R.811 does not.
o S.1487 deletes language in H.R.811 that would require polling stations to offer real, paper ballots, as an alternative to electronic voting.
H.R.811 makes some minimal effort to constrain the role of the Commission, but S.1487 goes hog-wild:
o H.R.811 relies on the Director of the National Institute of Standards and Technology to develop best practices, but S.1487 hands that responsibility over to the Commission.
o H.R.811 would like States to play a role in the certification process. S.1487 hands control to the Commission. It even permits the Commission to break its own rules on a whim ("emergency certification").
o H.R.811 requires disclosure between manufacturers, labs and state election officials, but S.1487 makes the Commission the central command.
o H.R.811 mandates that the Commission shall select labs at random "to the greatest extent practicable", but S.1487 weakens the language.
o H.R.811 requires the Commission to inform the public if it has "credible evidence of significant security failure at an accredited laboratory", but S.1487 deletes this, preferring such knowledge remain a secret between the lab and the Commission.
o H.R.811 tries (albeit timidly) to define some limits on the role of the Commission, but S.1487 has all such limitations deleted.
o H.R.811 allows the Director of the National Science Foundation to determine who is eligible to receive grants for research, but S.1487 gives this responsibility to the Commission.
H.R.811 has a lot more to say about audits:
o H.R.811 requires that States may not give any advance notice as to which precincts will be selected for audit. S.1487 has this line deleted.
o H.R.811 allows audits to be skipped if the winning candidate received 80% or more of the total votes, S.1487 does not.
o H.R.811 requires the entity conducting the audit to "meet the standards established by the Comptroller General to ensure the independence" of all parties, and requires that audits be performed "under generally accepted government accounting standards." S.1487 deletes all of this.
o H.R.811 attempts to outline several additional requirements for the audit, including that at least 10% of all precincts be audited in the case of a particularly close race. S.1487 suggests 2%.
o H.R.811 provides States the opportunity to develop their own audit standards, so long as the National Institute of Standards and Technology verifies the proposed method is at least as accurate as the method that H.R.811 outlines. S.1487 deletes this; why consult NIST when one can simply defer all judgment to the Commission?
o H.R.811 requires random audits, but also insists that at least one precinct from each county be audited. S.1487 deletes this safeguard.
o H.R.811 insists that the Commission adopt model audit procedures before the next Presidential election. S.1487 is content to wait until 2010.
o H.R.811 appropriates $100,000,000 to assist states in paying the cost of a rigorous audit regime. S.1487 tells States to go lay an egg.
Other assorted differences:
o S.1487 deletes several legal protections for "aggrieved persons", who have been disenfranchised, to seek remedy.
o S.1487 cuts the total research spending by $2,000,000.
o S.1487 cuts the total grants for purchase of equipment by $400,000,000.
o While H.R.811 bases funding allocation on voting age population, S.1487 bases it on the number of precincts.
o S.1487 expands and clarifies the protection of voters who speak certain languages other than English, referring to 1965 Voter Protection Act.
o S.1487 pushes back most deadlines, many not effective until after the next Presidential election.

The Essence of Democracy
There are many debates that occupy our media, and indeed our nation faces difficult questions. But on this perhaps we all agree, that our right to vote is the bedrock of a vibrant democracy. And yet, where is the great debate over those things that impact this most fundamental of rights? The time has come for conversation.

The groundbreaking Voting Rights Act of 1965 set a high standard for lawmakers aspiring to resolve this issue. I believe that any legislation that succeeds in restoring voter confidence, ballot integrity, or is indeed worthy of that Act, will include the following three essential components:

1. Voters must be able to privately and independently verify their actual vote. Practically speaking, this means paper ballots. Indirect comparison of the representation of votes is insufficient to promote a healthy democracy.

2. The Commission must have a narrow scope. There are several missions for this body, including gathering and dissemination of statistical data and the administration of federal grants. Oversight must not be among its tasks.

3. Open source code is required for voting software. The process by which the votes in our democracy are counted can never be a trade secret. In exchange for over a billion of the taxpayersí dollars, programmers of voting software must accept this fact, as must our lawmakers realize it.

In addition, I believe the following points are important:

1. Voting devices must remain secure. This means no wireless, no infrared, and no internet connection, anywhere in the process. It also means that voting data cannot be transmitted via the internet either.

2. Audits should be mandatory and meaningful. Specific and practical standards are needed. Federal funding could help encourage a rigorous regime.

3. Every vote should count. The American people were disturbed to learn that over 3.6 million votes were cast, but not counted, in 2004 (Greg Palast, Armed Madhouse, 2007). Absentee ballots, provisional ballots, and especially those ballots cast by our brave soldiers overseas must be staunchly protected.

the voting system shall
permit the voter to verify
(in a private and independent manner)
the votes selected by the voter on the ballot
before the ballot is cast and counted;
~ HAVA Section 301(a)(1)(A)(i)

Election Problem Log image
2004 to 2009


Accessibility Issues
Accessibility Issues

Cost Comparisons
Cost Comparisons

Flyers & Handouts

VotersUnite News Exclusives

Search by

Copyright © 2004-2010 VotersUnite!