I am honored that Eva Waskell has entrusted me to present The Privatization of Our Democracy, a work that I regard as her Profile in Courage. For 25 years she has labored to correct what is possibly the most significant public policy failure of the computer age—the privatization of vote counting carried out under the rationale that computers are simply automatic calculators that can tabulate votes more cost effectively than old analogue machines. I have known her for 19 of those years.
. . . . . .
People think they know that something is wrong with the way elections are conducted in this country. They are correct. There is. But readers only now will get access to a full history of the abuse of public trust by the elected politicians of the United States of America. That’s a large claim to make, but see for yourself.
I believe that there are multiple publications here in what Eva has to say. The scholarly monograph. An Elections for Dummies paperback. A paperback of humorous tabby cat photos where the kitties are running elections. Eva is a national treasure and I am proud to be able to use the Internet to make her story known. Fortunately, there are many, many public spirited citizens left.
Table of Contents and Selected Quotes Below the Line –
A SPECTACULAR Piece of Work
Foreword: An Abdication of Trust
Waskell’s concise verdict: “Our public servants gave up their ability to understand, and hence control and effectively manage, the vote counting process.” “Privatization was packaged and sold as modernization, efficiency, cost savings, a way to speed up ballot processing and the release of election
results, and cost savings. No one seemed to notice that control of the vote counting had now silently and irrevocably passed from our public servants who could be held accountable to the anonymous and unaccountable employees of a private company.”
1. “You really have your work cut out for you, don’t you?
My feelings about elections started with sheer anger and a deep sense of betrayal by the public servants who were in charge of our public elections and had let me down. For about the first five years I was very, very angry. And outraged at the neglect and/or complacency that was endemic at so many levels of decision-making. Nobody cared.
2. Before Elections: Looking for Unchartered Frontiers
3. US Election Administration: People, Processes, and Technology
An October 2001 GAO report entitled ELECTIONS: Perspectives on Activities and Challenges Across the Nation notes that the states permit the local jurisdictions “considerable autonomy and discretion in the way they run elections.” This highly decentralized nature and lack of uniformity, even within a state, is always something to keep in mind when talking about election reform.
According to a Congressional Research Service survey Election Reform and Local Election Officials:
Results of Two National Surveys that was updated February 27, 2008, “the typical LEO [local election official] is a white woman between 50 and 60 years old who is a high school graduate. …and earns under $50,000 per year. …and she believes that her training as an election official has been good to excellent.” The 60-page report also says that this description “does not capture the diversity within the community surveyed.”
4. Origins and Early Evolution of Computerized Voting Systems
Privatization was packaged and sold as modernization, efficiency, cost savings, a way to speed up ballot processing and the release of election results, and cost savings. No one seemed to notice that control of the vote counting had now silently and irrevocably passed from our public servants who could be held accountable to the anonymous and unaccountable employees of a private company.
5. You can’t have democratic elections if you have vote counting that’s a secret
COOK Report: And the issue of proprietary code and its impact on elections was something you realized right from the very beginning? Waskell: Absolutely. That’s one of the core issues. You can’t have democratic elections if you have vote counting that’s a secret.
Keep these three things in mind. First, it’s difficult to trust a process that’s secret. Second, privatization is incompatible with transparency. Third, trust but verify must be the guiding principle behind every critical step in a democratic election.
The election management system, the ballot layout software, the voter registration system, the software that runs
the electronic poll books, the software that processes and tracks absentee ballots, and the signature verification software used in absentee and provisional ballot processing is all proprietary, unexamined and unregulated.
6. Quoted on the Front Page of the New York Times
There were toggle switches on the front that allowed easy, direct access to memory and the ability to reprogram the software. I couldn’t imagine how anyone could possibly justify having that feature on a voting system!
COOK Report: Give me one example from West Virginia. Waskell: You could use a summary card, which was the same size and shape of a punch card ballot, to add votes to the running totals of a selected candidate.
7. Entertaining Ourselves to Death?
Privatized vote counting, flawed voting equipment and a lack of transparency were not generally perceived as a problem.
Journalists have become stenographers. They write down whatever an election official says and don’t ask any informed follow-up questions. That’s bad news because you can’t have a democracy without an informed electorate.
8. By 1975, Roy Saltman Had Identified the Issues and Proposed Solutions
I’d even go so far as to say that my observation, based on 25 years of experience, it that practically all of the essential and critical procedures in elections are run on an elaborate honor system that’s facilitated by insiders in the private and public sectors who blindly trust each other. It’s a world of don’t ask, don’t tell. This is the dirty little secret of American elections in my view.
Unfortunately, this de facto honor system hasn’t been squarely dealt with…or even publicly acknowledged.
To do so opens you to immediate and brutal criticism. It’s the third rail of election reform.
The only question that counts is, Are state and local election officials serving the interests of private companies
or the public’s right to know? The answer to this one basic question will ultimately determine the course of events as election reform moves ahead.
9. Getting to Know Saltman in Person and Witnessing the Subsequent Policy Flaws
NIST, the government entity charged with setting standards for computer technology, was not initially,
directly and continually involved in setting and updating standards for voting systems because voting systems were exempt from the Computer Security Act of 1987.
And there’s another catch. How do we know that the software in escrow is identical to the software that runs on the election computer on election night? And there’s another version of this question, How do we know that the software that is certified is the software that is running on the election computer on election night? Key question. A very key question. The answer involves identifying the proper methodology to use for software verification, i.e. how do you compare two pieces of software to prove they’re identical? Unfortunately, the answer is a bit murky to election officials because the vendors tell the LEOs one thing, and the mainstream computer scientists and security experts have a somewhat different answer regarding the proper methodology.
10. Elections in the Great State of Texas
Many of the discrepancies Elkins found were only uncovered because she had those PBC tapes from all of the precincts with the precinct totals. Lesson learned: precinct totals are vital to any kind of meaningful post-election audit. One of the first election reforms that Dallas County soon implemented was to get rid of the PBCs—some of them were reportedly borrowed from Chicago—and go to a centralized counting system for the punch cards instead.
Texas taught me that the technology was only part of the story. It helped change my emphasis from the flawed and vulnerable voting systems to the people involved and the election process itself, i.e. the procedures and policies that were in place. I quickly realized how complicated a management problem an election was and that sloppy recordkeeping was pretty routine. And it didn’t help that elections were generally run on a shoestring budget and pollworker training was almost always in need of great improvement.
11. Policy Lesson: Don’t Confuse the Simple with the “High Fallutin”
Rush Holt’s Voter Confidence and Increased Accessibility bill (HR 2894) also protects the number one item on the vendors’ wish list, their trade secret software.
For any audit that attempts to verify the correctness and integrity of election results, you must have precinct results on election night. That means precinct results publicly posted at the polling place. All too frequently this does not happen and there are no consequences, even though it’s a violation of the law. So the precinct totals are the building blocks on which the totality of the integrity of the election depends.
12. Show Me the Ballots!!
Public election data. Remember that it’s supposed to be public data—from a public election. And that means accessible to the public. Timely access. There are interim/cumulative reports and the unofficial election results (They’re called by a variety of different names.) that are published on election night. There are audit logs, event logs and error reports that also need to be made available—to the public. This kind of data, more correctly, this combination of comprehensive election data has never been routinely released on election night before. Not that I know of. And remember that the most important, the most important and fundamental piece of election data in
terms of independent verification of election results, is the voted ballot. That’s essential. And it’s the cornerstone of restoring public trust in our elections.
The 197 missing ballots didn’t change the results, but that’s not the point. The point is that counting the voted ballots with open source software was more accurate than the counting done with the proprietary software.
13. The Role of the FEC in the Development of Voting System Standards
According to Saltman, “Once the FEC commenced the development of the standards in 1984, it did not request the assistance of NIST, although this agency offered its expertise.”
The FEC had hired a consulting firm with little to no experience in elections, and then put them in a room full
of vendors for a full morning—literally behind fully closed doors—so the vendors would have an opportunity to tell the consulting firm…what?? We want you to get tough with us?!
Strauss, David Stutsman, Rebecca Mercuri and others all strongly criticized the standards for allowing the vendors and their proprietary software to control elections, not to mention the many specific technical flaws that were cited by these critics related to, for example, the troublesome accuracy requirement and the lack of clear, specific performance test requirements for compliance with the standards.
Before I make a final comment about the process, let me say something about the people involved. It’s important to note that the small group of consultants and technical advisors to the Clearinghouse and the election community in general have remained the same over the decades. These advisors have conversations among themselves and with election officials…in private. We’re talking about a well-established, and unexamined, relationship that has grown over many, many years. All of these people trust each other. However, the work products and reports authored by this small group of consultants and technical advisors have never been through the gauntlet of peer review.
14. Testing and Certification of Voting Systems
None of the state testing included a line-by-line examination of the vote-counting software because the states respected the vendors’ rights to keep their software a secret. And once a piece of software was state certified, the state had little if any control or effective oversight of what happened to the software at the local level.
There was no requirement in place that would routinely guarantee that the software that ran on election night was the software that was certified. This point is often overlooked by the media when they inaccurately report that the escrow system is somehow a security feature that helps prevent election fraud.
In 1984, Illinois became the first state to systematically and exhaustively test the voting systems in all of its 111
jurisdictions. In one county, 42% of the precincts had errors. And with results like that, what do you think happened? I’ll tell you what happened. The testing came to a halt because the state legislature didn’t provide any more funding for the project.
Over 40 years after computers were first used to count votes, two out of the 50 chief election officials in the United States of America finally got up the political courage to ask for the first genuine and independent review of the voting systems that determine the winners in all local, state and federal elections. Two out of 50.
15. The Importance of Location
The next step is to get this Election Transparency Resolution passed and added to the SARA Resolution (that our voting systems be Secure, Accurate, Recountable and Accessible) at the LWV national convention.
16. Post-1988: Computer Professionals for Social Responsibility Picks Up the Ball
I helped put together an inch-thick source book on computerized voting systems. In April 1993 I mailed copies of the source book to the 50 secretaries of state, and about 10-14 days later I started calling their offices. No one had read it, no one was interested.
Many election officials I talked with referred derogatorily to Dugger’s article as alarmist, and they blithely and reflexively dismissed what he wrote with a scowl. “We have checks and balances in place.” “Ronnie Dugger has never run an election.” “It’s all very theoretical.” “The pre-election logic and accuracy test ensures that the software is counting correctly.”
17. The 1990’s: A Leap into Full-Time Work on Elections
But on October 2 he called to say that 60 Minutes wasn’t going to do the story. He said he agreed with me philosophically but there was no smoking gun. I tried to persuade him that it’s impossible to get a smoking gun in any election and that’s the real story. That’s how dysfunctional our election system.
There’s never been any legal proof or hard evidence of vote rigging because it’s impossible to get your hands on it. I will show that this impossibility is really an indictment of our election system, not a vindication of it.
18. In 1997, Jenny Appleseed Writes a Planning Grant Proposal
COOK Report: How would you characterize the July 1985 Burnham article? Waskell: In a word…ground-breaking. It was the first in-depth article in the mainstream press about these very important topics and it was critical in raising the public’s awareness—albeit only temporarily—regarding the vulnerabilities of proprietary voting systems. Prior to this article, anyone with legitimate complaints about the vulnerabilities of trade secret vote-counting software or the inadequacies of the pre-election logic and accuracy testing or the weaknesses of the checks and balances was either totally ignored and/or marginalized or dismissed with official claims that
no reporter ever bothered to substantiate. I would also note that this general dynamic continues unabated to this very day…with a few notable exceptions.
Willis Ware of RAND wrote in 1987, “There is probably a Chernobyl or a TMI waiting to happen in some
election, just as a Richer-8 earthquake is waiting to happen in California.” I would say now, and I said then, that there are and have been many election Chernobyl’s of various intensity going on all over this country for decades at the county and state level.
Let’s just get rid of the inauditable touchscreen systems altogether. You can’t trust a machine that audits itself.
Would the IRS let you audit yourself?
20. To the Victors Go the Spoils
And since there are no perfect elections, as everyone knows, there is always a list of flaws to choose from. The losers cherrypick the flaw that can be manipulated to work in their favor. But after all is said and done, the bottom line is that election fraud and/or a rigged vote count only happens to the losers; it never happens to winners.
I’d also call many secretaries of state and ask for thecertification reports on their voting systems. I’d contact counties and ask for a copy of the report on their acceptance testing and fire a bunch of friendly questions at them. How did you decide on this particular system?
21. January 2000
22. Internet Voting and other Public Policy Issues after 2000
The citizens’ reaction to this since November 2000 has been this huge collective Aha Moment! As a result, they’re mad as hell and they’re not going to take it anymore.
Unlike the technology used in any other critical industry, like banking for example, the operating but unspoken assumption in the election industry is that the insiders who control their equipment can be trusted completely; to
suggest that insiders might cheat at vote counting in an election is not only totally unacceptable socially and politically, it’s unpatriotic.
The systemic issues like lack of transparency and accountability were present not only in Florida, the state with
the gold standard for testing and certifying voting equipment, but in states all over the country.
23. Election Reform after 2004
Acting quickly behind the scenes were the vendors who weren’t going to let this disaster go to waste. The vendors and their lobbyists saw a golden opportunity to increase sales not only of voting equipment, but of the much more lucrative, very pricy service contracts that typically accompanied a sale.
Many election laws carry no real penalty. So if there’s no enforcement and no penalty, is it really a law? While election laws also don’t cover what really matters, like bona fide independent verification of election results, they get very specific about certain other things, like the requirement to perform the pre-election logic and accuracy test, which proves very little but has a great feel-good effect on the uninformed.
24. First Define the Problem
Below the surface were the bigger and more fundamental problems of privatized election software, transparency and the public right to know, accountability, observer access, inauditibility. These core problems were systemic and were not seriously and sufficiently addressed by the academics, the public interest organizations, the election community, or elected officials.
“The most important political office is that of private citizen,” wrote Justice Louis Brandeis. And citizens in the election integrity movement take their political office pretty seriously. They’re living it. And don’t forget, they’re angry citizens…who have the potential to have a powerful and lasting effect on shaping election reform.
25. How the November 2004 Ohio Recount Was Rigged
COOK Report: How exactly was the hand recount rigged? Waskell. Election staff members used cheat sheets given to them by one of the vendors so that during the hand recount they would know what vote totals to report, i.e. vote totals that “matched” the vote totals from the computer. Not a very nice thing to do but not surprising. And the only reason it became an issue was because a citizen videotaped it!
As I recall, and I’m on shaky factual ground here, I admit, but I think it takes about 30 years for a movement to become mature.
The Creekside Declaration of March 22, 2008 is one attempt by about a dozen election integrity advocates (including election official Ion Sancho of Leon County, Florida) to define a common purpose: “Our mission is to encourage citizen ownership of transparent, participatory democracy.”
“After almost two years of deliberations, Germany’s Supreme Court ruled in March that e-voting was unconstitutional because the average citizen could not be expected to understand the exact steps involved in the recording and tallying of votes.”
26. Applying the Principles of Three Cups of Tea
I can’t emphasize enough how critical face-to-face listening is as a first step. Right now I don’t see much listening
and mutual respect at the local level. Citizen activists and election officials are largely throwing stones at each other. But election officials must be held accountable if they violate the law.
If you remain desk-bound and fail to do enough of the right kind of fieldwork, you’ll never become familiar with local political cultures and traditions and their enormous impact on election procedures as practiced, not as they’re preached.
Many secretaries of state and LEOs also don’t take the activists seriously and consider them to be operating on the fringe. Keep in mind that the kind of broad-based and intense scrutiny these public servants are getting from citizens is their first genuine encounter with accountability.
COOK Report: In the sense of Malcolm Gladwell’s book The Tipping Point, you were truly performing the role of the connector for all those years and probably just about the only one doing this. Waskell: Yes, I was and am a connector. But I never consciously thought of myself that way. I just saw two things or people or ideas that were separated and felt they would be more effective if they were connected. The sum of the parts is always greater than the whole. And because elections are so vast and complex and arcane in many respects, there was an endless amount of connecting that had to be done. But I think it’s now widely recognized that any work toward solutions to the challenges we face has to be an inter-disciplinary endeavor. The only problem is that well-informed citizens
are still left out of this equation. They still don’t have a seat at the table. And that’s got to change.
27. Because it was so very important…I just kept at it
28. VoteWatch and the Election Science Institute
Ohio election officials didn’t fully comply with the public records requests. Records were missing in quite a few categories of documents and for some counties there were whole categories of documents that were missing. This wasn’t surprising at all. Even so, the records that were turned over were quite revealing.
I also want to get my election files and historical materials from the 1970s, 1980s and 1990s online. It’s important to see where we’ve been, what we missed, and what mistakes were made in order to move ahead in a direction that will produce the needed improvements in elections.
29. Evolution of My Thinking
It was always necessary to shift your focus and your awareness during any election to ensure that you saw the whole problem in context and in relationship to all of the other parts of the election system. The trick in doing this was that some of the election context was visible and some of it was either invisible or a taboo topic you didn’t talk about. At the same time, you had to know what was important and keep your eye on that. You had
to understand what you were looking at. This is a key factor in election monitoring that’s overlooked all the time. You can’t effectively monitor a process you don’t understand.
… agencies in Nevada and California know more about the employees in casinos and the people who collect money from parking meters than any election official at any level of government anywhere in the country knows, or has even tried to know, about the people who own, work, consult with and are tied to the companies that manufacture voting systems.
Then over the course of a few decades, I spoke with numerous attorneys across the country who were involved in a variety of election litigation at the local level. The common theme that emerged here was that the courtroom was a dead end.
And once in power, the winners truly do not want to question the technology that put them there. They are not under any circumstance going to do a genuine comprehensive investigation of either the technology or the dysfunctional election system that got them elected.
The political parties will continue to focus on the issues they’ve been focused on since the presidency of George Washington, i.e. voter suppression and increasing turnout. You can read all about it in Tracy Campbell’s Deliver the Vote: A History of Election Fraud, an American Political Tradition—1742-2004.
I have a somewhat different perspective on how to handle the counting. Let’s cut to the chase and go right to the strongest validation possible, which is the direct evidence, i.e. the voted ballots. The votes on the voted ballots are the most important and fundamental pieces of data in a democracy in terms of trustworthy election results. Is this approach achievable in the short term? I believe we must achieve this is the shortest time possible by fast-tracking the only projects I’ve seen that put the voted ballots online on election night, i.e. the Humboldt Election Transparency Project and the Clear Ballot Group. Perhaps the open source community could put its full weight behind creating more pilot projects like the one in Humboldt County, California. Perhaps the Clear Ballot Group will be able to obtain enough funding for their commercial venture that they’ll have a handful of pilots.
I wish them both super-duper success. These pilots are all doable for the November 2010 elections if we get started right now. They’re both already past the proof of concept stage. And think about what we could have in place for November 2012 to minimize chances of another election meltdown.
30. Bottom-up Data Collection in Riverside County, CA
SAVE R VOTE’s first report was on the June 2006 election and pulled no punches with the title Riverside
County CA Voting System Operations: Expensive, Insecure, Illegal, Unqualified and Unaudited. It was followed up with the hard-hitting November 2006 election report. These reports will give your readers a true sense of the level of professionalism, detail and dedication of this particular group of election integrity advocates.
The main point I wanted to make when talking about the excellent work done by SAVE R VOTE is that if we want an accurate and comprehensive measure of the performance of LEOs, including accountability, we need a grassroots bottom-up approach to gathering election data to complement the government and the academics’ top-down approach.
But all I want to say about San Diego is that the Registrar of Voters there is Deborah Seiler. Remember her? She was the person in the California Secretary of State’s office that I talked to way back in May 1985. After many years in Sacramento, Seiler went to work for Sequoia Voting Systems, and then as a sales representative for Diebold from 1999 to 2004. She sold $30 million in Diebold equipment to San Diego. Since May 2007,
she’s been the trusted public servant in San Diego County who runs elections. Need I say more?
31. Inside the Last Bastion of our Democracy: the Central Computer Room
Let’s be clear about exactly what takes place in the central computer room on election night. This is where the consent of the governed is given, or taken away. This is the room where the keys to the kingdom are passed from one candidate to another. This is the room where it’s ultimately decided whether our government will be legitimate, or not. And it’s the computer operator who holds the position of supreme power in this room. This is the person who knows the details of the full range of capabilities of the proprietary software that resides on the computer on election night, and this person knows how to use them in real time. Absolutely no one in the state’s department of elections or secretary of state’s office has anywhere near this type of hands-on knowledge
about the proprietary voting systems that the state certifies.
There may or may not be a surveillance camera in the computer room. But there certainly is no complete, permanent and publicly inspected record at all of the most critical events that take place in the room on election night. Airplanes have flight recorders so we know what happened in the cockpit in case of a plane crash. There is no equivalent comprehensive record—or anything closely resembling it—of what the computer operator does in the cockpit of elections on election night. None. There is no black box or other device to record audio and, say, all of the keyboard keystrokes. The only black box in the room is the election computer.
32. Framing a Solution for a Seemingly Intractable Problem
It certainly can be framed as a human rights issue, i.e. the public’s right to know, the public’s right to a public vote count, the public’s right to observe all critical aspects of an election. This is a something that’s not being discussed. My present thinking is that this is also an ethical problem. I don’t want to get philosophical here, so
I’ll just say that by ethics I mean a consideration of what a person does when no one is looking. For years I’ve closely followed Larry Lessig’s blog and his critical thinking. He’s absolutely brilliant in how he frames and talks about the issues he feels passionate about. I especially like his definitions of independence and dependency. I’ve learned an awful lot from reading his books and watching his Keynote presentations. He speaks about lobbyists and the economy of influence, the pernicious influence of money in political campaigns, and he sees solving the corruption of Congress as “the first problem.” He is correct in saying this…but only partially correct because I think that solving the election integrity problem, specifically the integrity of election results problem, precedes what Lessig calls “the first problem.” It’s really the preamble to the first problem. Election integrity is a much, much harder problem to talk about openly and honestly. It’s the crazy aunt in the basement who surfaces whenever there’s an election. We hope and pray no one will notice her. Heaven forbid we confront her about her condition and insist she get help!! If someone does notice
her, we call that person crazy and/or simply ignore them. Once the election results are certified and any election challenges are put to bed, the crazy aunt descends again into the nation’s cellar of dirty secrets.
So who are the regulators of elections? Let’s first remember that there is no effective national or state regulatory
body per se for elections. None. The Election Assistance Commission is definitely not a regulatory body in any sense of the word, and neither is the EAC’s Technical Guidelines Development Committee as the vendors falsely claim.
But the local level is precisely where the real power and control over elections takes place and this is where to start, i.e. with local election officials taking an ethical oath of office. A rewritten oath of office I should say, a modification of the oath of office they already take. In short, the duties of local and state election officials would better serve election integrity and increase trust in election results if these public servants would truly and consistently represent the interests of the voters and not the interests of the private companies that provide election equipment and services.
33. Progress Toward Solutions and How to Overcome Obstacles
What Congress could do right now—but won’t—is require states to have uniformity in things like processing provisional ballots, ballot design, the matching formula used with voter registration databases, or standards for what a vote is (like Washington state already has). Congress could also require the states to have detailed rules for recounts, like Minnesota already has, and require these rules to be continually updated and refined.
There is no one solution. No silver bullet. It’s a complex problem and it will require a complex suite of solutions. The problems are too vast, complex, and deeply entrenched in the culture and political traditions at the state and local levels. What it will take is never-ending, incremental improvements to continually solve the problems that themselves will continue to change. But first things first. Let’s honestly define the problem. We haven’t done that. And it’s vital that all stakeholders be at the table on an equal footing when this problem definition process happens. Stakeholders include well-informed, dedicated citizens who need to be equal partners in this process They have valid concerns that have yet to become an integral part of the election reform dialogue going on primarily among election officials, computer experts, academics, and public interest organizations. “We the People” are nowhere in sight.
Where is our money going? Let’s compare this year’s Pentagon budget of about $660 billion to this year’s budget for all of the election departments in the 10,000 local jurisdictions and the 50 state offices around the country. … The difference between the two budgets is like the difference between Tyrannosaurus rex and a subatomic particle. I’ve made my point.
In my opinion, a combination of honesty, transparency and accountability is essential if we’re going to restore and retain the public’s trust that has been eroded by 45 consecutive years of incredible neglect and mind-boggling complacency.
“Democracies die behind closed doors.”
– Judge Damon Keith in a 2002 federal appeals court ruling