On August 1st, The Brennan Center for Justice released: Post-Election Audits: Restoring Trust in Elections. The entire report is quite readable without requiring knowledge of statistics, voting laws, or computers. I recommend it as a compurehensive introduction to the issues of auditing elections.
This post will discuss the report’s relation to and implications for Connecticut’s new election
audit law.
First, let me thank all of those involved in creating and contributing to this report. The Brennan Center, The Samuelson Law Clinic, and the members of the ‘Audit Panel’. The audit panel included several individuals from Verified Voting and staff from the Office of the Secretary of State of Connecticut. The Connecticut panelists alone were four of the entire panel of seventeen. As the report states, “Opinions, findings, and conclusions or recommendations … are soley of the authors and the Brennan Center”.
While we seldom agree on everything, the the staff of the Office of the Secretary of the State have always been responsive and available. The Secretary has also always been responsive and I have no doubt that she is committed to fair elections with integrity. When the Secretary changed course to specify optical scan machines in early 2006, she demonstrated the kind of flexibility and openness that will be necessary to apply the conclusions contained in this report and the reports from California to the benefit of Connecticut voters.
Perhaps the most basic findings of the report and the most challenging to recently enacted Connecticut law are:
Of the few states that currently require and conduct post-election audits, none has adopted audit models that will maximize the likelihood of finding clever and targeted software-based attacks, non-systemic programming errors, and software bugs that could change the outcome of an election…
Based upon our review of state laws and interviews with state election officials, we have concluded that the vast majority of states conducting audits are not using them in a way that will maximize their ability to improve elections in the future.(emphasis added)
Strickingly, these finding were based on an earlier version of Connecticut’s law that was proposed by the Secretary but later watered down considerably. From the Brennan report: “Connecticut, California, and Illinois check all races on the ballot during a post-election audit”.
As originally proposed by Secretary Bysiewicz, the bill would have audited 20% of districts and all races. However, the final bill approved by the legislature and fully supported by the Secretary audits 10% of districts and a minimum of three or 20% of races. So, on a strict percentage basis the current law audits about 1/8th as many votes as the one judged inadequate by the authors and the Brennnan Center.
In a blog session at MyLeftNutmeg, The Secretary explained this past Thursday:
our office recommended to the Legislature that 20% of all precincts be audited. After discussions with Dr. Schvartsman(sic) at UCONN and our work with the Brennan Center, we were advised that a 10% threshold would ferret out fraud. We also had to deal with significant objections from municipalities that thought it would be too costly and burdensome to administer at the 20% level.
In light of the Brennan report’s conclusions, one has to ask: Was there a mis-communication? Has the Brennan Center changed its opinion? Perhaps, Dr. Shvartsman included or assumed qualifications and meant that the ferreting out would be sufficient for a state wide race, that was actually chosen to be audited?
While I share the concern for municipal costs they could be covered by the state or, at least intially, by federal HAVA funds. The right questions to ask is “How much is it worth per vote to insure fairness, confidence, and integrity in Connecticut elections?” Is it worth $0.20, $0.50, $1.00 a voter?
Consider this from the Brennan Center:
Had they paid the market rate of $7-10 per hour for counters and $20 per hour for managers, the 2006 Walpole, New Hampshire election night hand count of 1,574 fourteen-race ballots, with nine two-person counting teams, would cost about four cents per ballot, per race.
For example, if Connecticut had an audit law that mandated a 10% random audit in all elections, plus one district per municipality we would audit about 1/3 of the ballots in the state. If elections had an average of 15 races or questions, at $.04 per race per ballot, it would cost $0.60 per ballot audited, or about $0.20 per voter. We could reduce these costs by using adjustable-percentage auditing for state wide and congressional races or auditing only 5% in state wide races, 10% in congressional races, and skipping the audit alltogether in non-contested races. That is in the range of the cost of printing the paper ballots and would make them, as they say, priceless. (Note: I am not proposing this specific level of audit, just demonstrating the high end potential costs of sufficient random audits. One flaw in most of the recent election audit reports, in my opinion, is that they do not deal sufficiently with the challenge of small elections on the order of 1, 10, or 25 districts. A topic for this blog to address on another day.)
And as the Brennan Report also concludes:
Regardless of the audit model a jurisdiction implements, there are several simple, practical, and inexpensive procedures that it can adopt to achieve the most important post-election auditing goals, without imposing unnecessary burdens on election officials.
The Brennan report is also consistent in several points with our concerns with Connecticut’s law, posted on this site. Here are some relevant excerpts from the Brennan report:
what constitutes a “sound†audit actually centers on disagreement over the purpose of an audit.
additional factors that jurisdictions will probably want to consider… whether their effectiveness will depend heavily on the subjective judgments of election and other public officials in charge of the audit (something jurisdictions should generally want to avoid).
Implement Effective Procedures for Addressing Evidence of Fraud or Error… jurisdictions must adopt clear procedures for addressing discrepancies between the paper records and electronic tallies when they are found. Without protocols for responding to discrepancies, the detection of fraud or error will not prevent it from successfully altering the outcome of an election.
The Report also suggests additional solutions, some of which the Secretary may be able to implement in the near term by regulation or procedure:
Allow Candidates To Select Precincts or Machines To Be Audited.
Freeze and Publish Unofficial Election Results Before Selecting Precincts or Machines to be Audited.
Conduct “Blind†Manual Counts. While unofficial totals should be made available to the public so that they can verify the accuracy and fairness of the audit, manual counters should be “blind†to the unofficial election results for the machines they are auditing to ensure that knowledge of the unofficial results does not influence their counting.
Don’t Just Match – Count! (Record and Publicly Release Meaningful Data on Votes Cast). Audits that record and detail the overvotes, undervotes, blank votes, spoiled ballots,…Rather than only matching paper and electronic tallies, election officials should record and publicly release this meaningful data, which should be useful for improving elections in the future…We are aware of only one state – North Carolina – that has collected and made this type of information publicly available
Generally, our review of state laws and interviews with election officials revealed that the majority of jurisdictions do not have adequate, detailed, and practical procedures for action to be taken when unexplained discrepancies are found. Jurisdictions should conduct a transparent investigation of all machines where the paper records and electronic tallies do not match to try to determine the cause of any discrepancies.
The warehouse [for voting machines] has perimiter alarms, secure locks, video surveillance and regular visits by security guards.
There is much more in the report. My attempt here is to highlight its major implications and value for Connecticut.
Let me conclude with a final quote from the report that I interpret must apply to members of the “Audit Panel” from Connecticut:
Our interviews with election officials have left us convinced that many are eager to use the information developed by academics and election integrity experts to improve their post-election audits.
