Mary Ann Gould Interviews OH Secretary of State Jennifer Brunner
Audio file of this interview: http://voiceofthevotersarchive.org/VV121907.mp3
MAG - Good Evening Secretary Brunner. Thank you for being a guest and
thank you for initiating the Everest study. Would that more states
act to investigate and get the facts on the security of our voting
system, and we really appreciate that you have taken that action. I'd
like to start with the findings of Everest—by the way, what a great
acronym and name. It's definitely apropos for Ohio and perhaps the
entire United States. In your opinion, what were the three most
critical findings that are of most concern, and why?
SoSB - Thank you Mary Ann. To sum up, the stream of critical findings
were that the security in our voting systems, whether it's in the
software, servers, workstations, or at the Boards of Elections, or
the voting machines, do not contain the industry standard, or even a
minimum standard of security that we are used to and that we expect
in our other computer applications that we use for things like
banking or communications. That leads to the vulnerability to
viruses being entered in to the system even through the voting
machines in the polling places, and while we don't think that there
would be very many people who would do that, in our computer security
protocols that we use in our other systems, it's already guarded
against by the engineering of the system.
Second, we were very disappointed to learn that what we inherited
from the previous administration is a system where it's not
documented as to the configuration of the software that goes from
system to system among the counties of the state of Ohio. And, the
other thing that we learned is that the performance in some
instances, as these voting machines get used more and more again,
they're going to quickly wear out and not perform and start to
malfunction, and we've only been using them since the earliest of
November 2005. I probably should probably cover the fourth area of
the testing which was the internal operations and controls. That
allowed us to understand what a disparity there is of documentation
that's used by the Boards of Elections to operate the systems and the
work that the Secretary of State needs to do to provide instructions,
guidance, information to help the Boards of Elections do what they
are trying to do which is to do a good job.
MAG - Is there any point in particular that was especially shocking
to you that you did not expect?
SoSB - Well, overall I had hoped that we would see some bright spots
in all the equipment that we tested, and what was extremely
disappointing was that none of the systems tested well. So, from the
statewide standpoint, it leaves us in a position where we have to
look at-what are the risks? because we know we can't mitigate all the
risks, but how do we craft a solution that's going to allow us to use
equipment that we're not entirely happy with, in such a way that we
can satisfy the needs of the voters?
MAG: Especially for 2008.
SoSB: — a lot of bad choices, really.
MAG: Bottom line: from the initial landmark Harri Hursti hacking test
in Florida, through many other tests, including the California top-to-
bottom, now yours, the findings are that the machines have so many
ways to be compromised, either by mistake, or deliberate, that many
have deemed them "fatally flawed", beyond repair, that basically,
they were not designed with security as a priority. Would you agree
with that statement?
SoSB: I would agree that the security was lacking and that part of
the responsibility lays with what was established by the Help America
Vote Act by Congress, in that the testing protocols to bring these
machines to certification for federal elections didn't include the
type of security review that, for instance, California performed, and
the state of Ohio performed, and that's troubling.
MAG: And I hope what the two states have done gets performed
elsewhere. Let's move to recommendations. Who actually made the
recommendations of the problems found?
SoSB: the recommendations were actually sketched out by myself
working in connection with my staff, consulting with the researchers,
our testers, but then what we did, we took those recommendations to a
bipartisan group of election officials, twelve in total, six
republicans, six democrats, all directors and deputy directors of
Boards of Elections throughout the state that represent a variety of
different types of voting systems. And, with those officials we were
able to hammer out the finer points and we gave them a number of
options and said, "which do you think is going to work better?" We
started to assign costs to those options because this has to be paid
for if we're going to make the changes. And in the end the
recommendations were essentially my recommendations but reached with
the consultation of the election officials who actually provided a
lot of value to how these recommendations would work and could be
MAG: Are these recommendations final or will there be a chance to
have them reviewed, revised, have citizen input?
SoSB: These are just recommendations. They are not final... the
reason that we're actually taking this to the legislature, besides
the need for perhaps assistance with funding, is that we want people
to be able to express their points of view, to give us their ideas,
their suggestions. We want the beauty of a free speech dialogue like
we have in a democracy in our country so that we can make this a
better process and a better proposal, a better solution to the
problems that we found in our report.
MAG: I assume then you would agree with one of my favorite quotes
from Abraham Lincoln, "Elections belong to the people. It is their
decision" and we need time to get them back in the process.
SoSB: I like that saying and we actually have a poll worker
recruitment brochure that has on it's cover "elections belong to the
people." and we've got pictures of Suffragists and pictures of people
marching during the civil rights movement. So clearly elections do
belong to the people.
MAG: I'd like to just touch base on overarching principles. The
beauty of our country was the creation of a government based upon
separate and independent checks and balances. Would you agree that
that should also apply to our election system?
SoSB: That and transparency create reliability, accuracy and trust in
MAG: And I would assume openness and provability to the original vote?
MAG: O.K. Given the above, let's discuss the recommendations because
there seems to be some confusion about the use of the precinct-based
optical scan. Could you explain at this point in time, how you see
the precinct-based optical scan being used?
SoSB: What we're recommending with the proposal that we've given to
the legislature and to the Governor, is that we, and we got to kind
of look at this from a total picture...we're looking at creating vote
centers that would be in size from five to ten precincts that would
accommodate early voting fifteen days before the election through
election day, and we're looking at people being able to vote a paper
ballot, or if they need assistance because of a disability, would be
able to insert a paper ballot into a ballot marking device, such as
an Automark, that would mark their ballot for them. Now, the Help
America Vote Act talks about having a second chance, for a voter to
have a second chance to review their ballot to check especially for
overvotes, and also for undervotes. Although the HAVA, the Help
America Vote Act says that if you have central counts it's
permissible to deal with this opportunity for second chance by
adequate amounts of public education, we would like to take the
existing precinct-based optical scanners, use them in the vote
centers as a scan, or a check on the ballots to allow the
voter to insert their ballot into the optical scanner, be alerted to
an overvote or an undervote and then be able to reject that ballot
and create a new ballot that doesn't contain the overvotes that could
cause invalidation of their vote. Once that's done they would place
their ballot into a ballot box and then that would be transported to
a central counting location at the Board of Elections. Now, at the
Board, we would provide for a server for the count...
MAG: Now, just a moment, at the precinct then. Would the precinct
based optical scan still keep a number count?
SoSB: Would they still be tabulating?
SoSB: Based on the security results of the study that we saw, the
answer is "no." Not at this time. It doesn't mean that they wouldn't
ever. But with the current configuration of the software and the
firmware that's in the equipment and the lack of control so that
right now with the current precinct-based optical scanners someone
can actually turn off the memory. It will still allow the ballots to
scan but there'll be no tabulation, and that can be done for a short
period of time, different times throughout the day and it would only
be discovered or detected if there were a full hand count of the
ballots, or in a recount with a limited number of ballots, so that
can be corrected. We do not want to subject the voters to that risk.
MAG: Many people have raised the question: Wouldn't it be good to at
least have the count at the precinct so that should anything happen
to those ballots between there and the central counter, you do have a
count, and secondly it would be a back-up.
SoSB: That may be a good idea. In the case of the early vote centers
we would just have to ensure that with the technology used that the
scanners have different functions between scanning and tabulating. In
Ohio, two of the types of scanners that are used do, but one
manufacturer's scanner doesn't. And that scanner both scans and
tabulates all in one so that we would have to ensure that whatever
type of equipment was certified if we use that tabulation at the
precinct level as a back-up, had a differentiation in function so
that we weren't having tabulating occurring before election day.
MAG: Um-hmm. That would be important. So, you are open at this point
to having a back-up number at the precinct?
MAG: O.K. Why do you think the central counter would be more
accurate? Isn't that exposed to the same problems?
SoSB: This was actually something that we saw as a recommendation of
our academic researchers that that was the more secure from attacks.
The reason that we see that as being more secure is that we can
control who has access to it, better than we can with machines placed
all over the state in about fourteen thousand precincts.
MAG: What would you say to those who are saying, well, number one,
somebody else will be handling the voter's ballot: Number two, you
have a chain of custody question with the transfer of the ballots
from the precinct to wherever the central counter will be?
SoSB: Well, chain of custody is something that I know very well from
the days serving as a judge in the Common Pleas Court in Franklin
County. And chain of custody can certainly be documented and that
would be a very worthwhile thing for us to put in our procedures that
we would set forth for the Boards of Elections and those who were
working in the vote centers. So, I think that's the first part of
your question. Was there a second part, Mary Ann?
MAG: No, It's the concern that the chain of custody, be at least at
the precinct base and yes, I recognize the problems there, but it's
the only person handling that ballot is the voter themselves whereas
you would be having somebody else transfer those ballots potentially,
they may not have been counted if you don't have a counter at the
precinct. Therefore you may have a lost chain of custody.
SoSB: O.K. a couple things to answer that. First of all with each
voting machine where tabulation occurs at the precinct there's a
memory card that transfers those tabulated votes to a central server.
Those can be lost in the process. Those have been lost in the
process. In some instances it's taken half the night or more to
recover those cards. Second of all, with the paper ballots...go back
to your original question. I'm sorry.
MAG: So, let's just really get to the question of concern that the
central based (tabulation) may not be secure. What makes (central
tabulation) more secure?
SOSB: What makes the central-based more secure...is that, first of
all, I'm not saying that our high speed optical scanners are going to
be perfect but we know from our study some of the significant
vulnerabilities and we can deal with most of those through procedures
and policies and through documentation. For instance, workstations
that are connected to the server where ballot definition or other
programming goes on, are subject to an audit log. As long as the
audit log is turned on, which we would require be done in the
procedures we're going to be able to tell who did what with what was
going on that night and we also have procedures already in law and
plan to continue to support those procedures for observers to be able
to observe the count. And then finally we're looking to make
recommendations for a post-election audit procedure to check that.
And the other point I was going to make earlier. With a paper ballot
system as I am sure you know there is the opportunity for
reconciliation. So that we know how many ballots were printed for a
particular precinct. We know how many ballots were voted; how many
were spoiled and how many were unvoted, and those should add up to
the number of ballots that were originally printed to allow
reconciliation before we ever get to the steps of auditing.
MAG: We'll come back to that in a moment but you did raise an
important point on the cost. Since these systems have so many
vulnerabilities, why aren't we going after the vendors?
SoSB: Mary Ann, I'm starting to lose you. I'm sorry.
MAG: Why not approach the vendors to recap some of the cost?
SoSB: That's a very frequent question that we're asked, and in this
process, whatever recommendations are ultimately adopted by the state
legislature with the input of lots of people I think we're going to
be in the position where we're going to need to purchase new
equipment. I think that based on the time frames we have and the
negotiations and the contract work that's going to need to go into
it, we're in a good position to be able to seek concessions based
upon problems that we have with the machines and where we need to go.
We think that hopefully will be a superior course to any kind of
MAG: Now, time frame: Do you think that this can be done in 2008? and
especially with the primary?
SoSB: We don't think that we're going to be able to enact wholesale
changes for the primary. But looking at the example of New Mexico,
Governor Richardson signed a bill to convert that state to optical
scan in March and by November they had it in place statewide. We
think that it can be done, but we will proceed with caution because
the last thing we want to do is to rush a solution that creates more
MAG: Have you decided on how you would audit the system?
SoSB: We're still looking at the some of the best ways to do that.
What's interesting is that with the advent of the new equipment
because of the Help America Vote Act there's been quite a bit of
literature and research on audit procedures. I think it's going to
take some time and some experience to develop what we would call best
practices for auditing. I have been intrigued by audits that would
allow us to actually test... if we were using machine voting to test
the votes on every machine but with the model that we're proposing I
think an audit would be somewhat simpler. We would end up working
with experts statistically on how to randomly select ballots so that
we weren't hand-counting an entire state.
MAG - Now that raises the entire question, by the way, I had worked
with Dr. W. Edwards Demming who was the foremost quality statistical
expert in the world, and the basic premise is that when a system is
out of control, unpredictable, audits lose their validity, and you
may have to go to almost a 100% count until the system is under
control. Would you be willing to consider that?
SoSB - I'm not sure what under..."out of control" means.
MAG - "Out of control" means that there are so many problems that it
is not a statistical-controlled system; that the problems can come up
randomly at unusual times, unusual places and there is no way to
know. Therefore, a normal auditing system which is based upon a
normal operating system, doesn't seem to work.
SoSB - Would we be willing to go with a "statewide handcount?"
MAG - If that's necessary.
SoSB - I can't tell you at this point....Again, the goal is to assure
an accurate vote count. We would need to see what that is going to
take and go from there.
MAG: In any way are you heading for Vote-by-mail?
SoSB: One of the things we suggested was to allow individual
counties, like they do in the state of Washington, to actually put
the issue on the ballot of whether or not they can vote by mail, and
I don't know how popular that recommendation will be with the state
legislature, but again, what we're trying to do as part of this
proposal is move Ohio into the twenty-first century by offering early
voting, and voting on more than one day. We're looking at seven days
a week, twelve hours a day, except on Sunday, seven hours. The option
of allowing a county to vote by mail is a way for our voters to test
some new systems and new ways of voting because essentially voting in
Ohio and many places in the country hasn't changed for about forty or
more years. But again the caveat with that is that a state has to
ensure that it's voter database is intact, is reliable, is in good
shape and what we inherited from the year before we came into office—
It was a system that still needs some work. We've done quite a bit.
We've hired a voter database coordinator with a lot of background and
experience statewide working with database systems. But we know that
our system is not yet perfect. We know that what the Boards of
Elections have in their records oftentimes is better voter history
than what we have in ours. And when we notify a county that it has a
duplicate not all counties are responding as quickly as we'd like to
eliminate those duplicates,
MAG: You also inherited a history of caging, or voter roll cleansing.
Are there any steps to reverse that? We understand that in certain
places many people turned up at the polls and lived in the same
location for many years and were found that they were no longer on
SoSB: What is troublesome for me as Secretary of State in Ohio, is
the first step in vote caging is actually sending a notice to voters
at an address that is in the Board of Elections records and when that
notice comes back, for whatever reason, if there was an error in the
data entry, or if in fact the voter did move, then that was used in
2004 to systematically challenge voters. What happened in Ohio is the
Republican legislature built into our statute the expense of mailing
that notice into the statute...essentially the government is mailing
that notice out, and when that notice comes back, because the statute
is very specific that it can't be forwarded if the person's not
there, that first part has been paid for by political operatives who
would choose to use that. That we would love to see changed. That may
take some time until we see a change in the make-up of our legislature.
SoSB: You've taken a strong stand on ethics and you have sent out I
think a notice to the workers at Boards of Elections, those who work
at the polls to sign such an agreement and I laud you on that, but
you included a confidentiality component. How would confidentiality
of information to do with voting be supportive of an open and
SoSB: Mary Ann, I need a little more description of what you're
talking about with a confidentiality component.
MAG: I believe, I don't have it right in front of me, but there was a
section that was mailed out that included that all information was to
be kept confidential. Are you saying then that if those at the Board
of Elections, those working at the polls, see a problem that they are
free to disclose that?
SoSB: Oh, certainly. We operate in a system of public records in
Ohio, So I have personally reviewed that policy about three times now
and I'm having trouble discerning or recalling what exactly that
applies to and I apologize for that. So I think we'll to have to look
at it again and maybe revisit it.
MAG: What can citizens do to help you?
SoSB: Citizens should make their voices heard. If they live in Ohio
and they have an opinion on this they should contact heir legislator.
If they like parts of our recommendations or all of them, I think
that the legislature should want to hear from them. What's been
interesting for me is that the unsolicited comments, emails, faxes,
letters and phone calls that we receive from everyday citizens have
been overwhelmingly positive. I think it's exciting for us to look at
where voting can go in Ohio and what we can do to create a system
that's going to allow more people to participate. The last thing that
I ever want to come from these findings is for people to say that
they have no confidence in the system. The sense that I'm getting is
almost a sense of relief because people have had so many questions
for so long and at least our study answers those questions and it
gives us a launching point to go far beyond what any of us could have
ever imagined our election system could be.
MAG: Now for those places that ran into very long lines in the last
presidential election, will they be able to just go and get a paper
ballot and use it?
SoSB: What I'm suggesting, especially for the March Primary, where we
still have to use the DRE's and the precinct-based optical scans in
the polling places, is that for situations with the DRE's that if a
person wants a paper ballot, they can have a paper ballot. Now anyone
in Ohio who wants to vote by absentee ballot is able to do so without
giving any reason whatsoever and they will always be given a paper
ballot to do that, unless of course they go to the Board of Elections
and vote in-person, early, at the Board as an absentee, and in some
cases they may vote by machine.
MAG: Now, how would you deal with the fact that some of the studies
have found that a precinct-based optical scan can read a ballot one
way and a central (tabulator) may not read it the exact same way. So
you may end up with two systems reading two different ways.
SoSB: Can you elucidate for me what study that is?
MAG; That's in fact by Lehigh University, Dr. Dan LoPresti. He has
been doing quite a bit of study on the use of paper ballots and
SoSB: I would need to look at that study but one of my concerns is
that, are we saying that the precinct optical scan was the correct
tabulation? and central wasn't? Or, visa versa?
MAG: You don't know. They can read differently. And in fact certain
optical scans can read certain sections and not read another section
well, Which may, depending upon where a party is located (on the
ballot), leave a party out.
SoSB: Now, some of what we learned in our study, because we looked at
more than security. We looked at performance and we looked at
internal operations, and controls and configurations, was that the
limitations in some of the machines include a ballot that's printed
where the selections have to be in a certain location, for instance,
in the right-hand column or the left-hand column, and I can't recall
whether it is right or left off the top of my head. That's a
limitation that's built into the engineering and performance of it.
That could be one effect. Another effect is the actual ballot
definition, the actual creation of the ballot so that it can be read
by a scanner, and as you may recall in Cuyahoga County in 2006 the
first time they used their new Premier system which was Diebold at
the time, one of the problems they ran into with their optical scan
ballots, which were absentee ballots were that they had them printed
at a printer who didn't create the ballots in a way that they could
be read. So, there are a lot of variables that could play into this
and we'll certainly look at those issues. As you suggested, allowing
a check through, allowing the precinct-based optical scanners to
actually tabulate on election day all of the results that have been
scanned into it and comparing that with the central optical scan, may
be a good solution to explore that.
MAG: That certainly would and it certainly meets the criteria of
separate and independent checks and balances and is a back-up. And if
there are differences that usually a signal to investigate. Now, if
recounts are needed how will you ensure that the selection is truly
random, rather than what has occurred in the past? Known ahead.
SoSB: Like cherry-picked?
MAG: (laughs) Yes.
SoSB: We have already issued a directive, and I'm sure it's going to
be subject to further change on randomizing the recounts. Right now,
with recounts in Ohio there are situations where automatic recounts
occur when the margin is very close as specified in statute. We take
three percent of the precincts in that particular race and we recount
the votes by hand in those precincts. We did set out a rather
methodical way to randomize the selection. There's been some
criticism that we could make it more random, but we tried this our
for the November election. It worked. It didn't slow down the
process. But we're willing to look at more sophisticated ways to make
this even more statistically random.
MAG: Now, in summary, many people have said that raising this
question will scare people away from the polls. What say you?
SoSB: I go back to the issue that you said, that election belong to
the people. The way that we have a voice in what our government does
is by participating. We know that as these machines were originally
engineered that they were supposed to count the votes correctly. We
have procedures that will allow us to check and to double check in
the short term for the March primary. For November we plan to have a
system in place that will minimize the risks; that will maximize the
instructions and the uniformity of the providing of rights to people.
What we're saying is don't give up on this system. We have many
people who are working hard day and night to make this work because
we understand that voting is the mechanics of how democracy gets
done. Without people's participation we don't have a full and robust
debate. Participation is the fullest measure of our democracy and I
know that each and every one of us believes in that democracy and we
can show that by ensuring that we inform ourselves and that we stick
with the process and allow the improvements to take place as the year
unfolds next year.
MAG: How can citizens get involved/ How can they become poll workers?
SoSB: Citizens can contact their local county Board of Elections and
indicate that they would like to be a poll worker and the Boards
desperately need that help. And they can also go to our website. We
have information on how to contact all the Boards of Elections at the
website. And, we also have something that we created specifically for
high school students who are graduating this year. It's called Grads
Vote 2007. It's under the Secretary of State's website which is
www.sos.state.oh.us and there's actually a place on online to sign up
to be a poll worker and we will transmit that information to that
voter's particular county.
MAG: How can citizens comment on the recommendations. Will there be
some open hearings?
SoSB: There should be hearings in the Ohio legislature and we have a
website and people are more than welcome to drop us a line at our
MAG: Would there be any county hearings where people could go locally?
SoSB: At the county level, I'm not really certain that a Board of
Elections would go to the point to hold that hearing and with the
limited time we have to get recommendations actually adopted and
funded, it's unlikely that we would be able to make it to all eighty-
eight counties between now and when we think we need to have this
ready to roll which would be at the latest mid-April.
MAG: Now, will citizens be able to observe the central counting?
SoSB: That is our plan and right now they can and we intend to push
any proposal that would allow that to still occur.
MAG: What assurance will they have that the transfer of the ballots
from the precinct to the central locations will be completely secure?
SoSB: We will adopt chain of custody procedures and we will verify
that that is done. We have regional liaisons for our office who each
work with the county Boards of Elections that are assigned to them
and they would be checking and spot-checking to make sure that the
procedures were being followed.
MAG: Fine. Is there anything in summary that you would like to say to
both the citizens of Ohio and also to those in the United States
because we face a critical election in 2008.
SoSB: I ran for Secretary of State because I wanted to do this job. I
bring my judicial experience, my years of working in the Secretary of
State's office, my years as an election attorney and serving on the
board of elections. I bring that fully to bear and I resigned my
position as a judge so that I could actually run for the office and
it's important to me to preserve our democracy, and I and my very
dedicated staff will work day and night to do our very best to make
MAG: Fine. Thank you very much.
SoSB: Thank you Mary Ann.