Re: [OVC-discuss] Privately and independently -- how will OVC deal withthis?

From: Nancy Tobi <nancy_dot_tobi_at_gmail_dot_com>
Date: Sun Jun 28 2009 - 09:36:05 CDT

This October our friends at NIST are hosting a conference to discuss what
they like to call* "end to end voting systems"*. This is the vision
articulated in the "voluntary" (until Holt succeeds in legislating them)
voting guidelines put forth by the EAC.

The "end to end" approach underlies Holt's latest V VSG-sponsored hands-free
ballot box technology. It underlies his ballot readback technology, which
was in the 2005 V VSG. The end to end approach says - as articulated in the
EAC guidelines - that it is not enough to have multilingual ballots (as
mandated in the VRA), it is not enough to have "accessible" ballots, but
rather the *entire voting system *must meet technology-based accessibility

Holt's legislation is ensuring that the morphing of HAVA accessibility
standards meets this new definition: the end to end definition - or actually
ANY definition that the kooks at the EAC-TGDC happen to come up with.

9 adopting any voluntary guidance under subtitle B of title
10 III of the Help America Vote Act with respect to the ac11
cessibility of the paper ballot verification requirements for
12 individuals with disabilities, the Election Assistance Com13
mission shall include and apply the same accessibility
14 standards applicable under the voluntary guidance adopt15
ed for accessible voting systems under such subtitle.

My professional responsibilities include defining software requirements so
my company's technology product meets accessibility standards (as codified
in Section 508 of the Disabilities Act in federal law). This is a broad and
deep field. Nevertheless, I am unaware of any software producer who is
endeavoring to create an "end to end" product in order to meet accessibility
standards. There are many practical reasons for this, not the least of which
is that it is cost-prohibitive. But more importantly, it is also not

*We design sidewalks now with built in ramps to make them wheelchair
accessible. We do not design automated sidewalks to march over to any person
in a wheelchair, speak to them in their language of choice, and then pick
them up and carry them to the ramp.*

Such a design exceeds practical necessities and the legal bounds of
"reasonable accomodations" as defined in the ADA.

Anyone who has sat through the EAC-TGDC discussions knows that these people
are happily and irresponsibly designing pie in the sky technological
concepts that make them giggle deleriously at the possibilities before them
(ah yes, the deep taxpayer pockets), but which in fact have nothing to do
with voting rights, disability rights as commonly accepted and understood
both legally and practically, or with the realities of election system
management and administration.

On Sun, Jun 28, 2009 at 10:18 AM, Jim Tobias <> wrote:

> Hi Alan,
> I think you're right that the major problem is not having a working
> definition of "independent". Obviously, this is tricky ground -- that's a
> pretty loaded word. But if you look at other disability laws and
> regulations, similar concepts have been agreed upon and put into action.
> We've got phrases like "reasonable accommodation", "undue burden", "readily
> achievable", "access comparable to a person without a disability",
> "fundamental alteration", etc. For some of them it took years to hash out
> the specific meanings and applications, but that's an inevitable process
> anyway, so we should get started with it rather than avoid it or complain
> about it.
> Our best strategy would be to engage the disability communities, esp. those
> focused on accessibility, and with their help, begin to craft a good
> definition. Doing the technology design now, not knowing *how* independent
> the method must be for *which* users is a dead end. I personally have no
> doubt that many acceptable technological solutions can be made, esp. once
> they are put in a context of use (polling places, poll workers, etc.) But
> we are missing an important part of the "spec" right now; assuming how it
> all will play out is neither wise nor necessary.
> Let me start the disability context dialogue by making an argument that
> would be understandable to those communities. In accessible public
> transportation, a rider in a wheelchair is not expected to or even allowed
> to operate the lift him/herself. We know that it would be possible to
> design a system that did this. It would be expensive and might require as
> much cooperation from wheelchair companeis as it did from bus
> manufacturers.
> But the point is, in this context "independence" is not an absolute; it's
> balanced by other factors. There's a continuum of such bargains in every
> instance of accessible technology I'm aware of.
> As a practitioner of accessible design, I'm not sure that your methodology
> of categorization of disabilities will work as you think, and the
> demographic information you're looking for simply isn't there. Such a
> categorization scheme hasn't been necessary in any of the other
> accessibility arenas I'm aware of, and it is not needed to show commitment
> or to make progress. Beyond that, demographic and probability arguments
> are
> more likely to push us into the explosive territory of being seen as
> denying
> civil rights as any arguments we could make.
> You sort of make that argument in a way that some advocates might find
> objectionable:
> > Even if a machine would allow them to
> > vote privately and independently at the pollsite, they're
> > just not going to use it anyway.
> > Informal studies show that practically no one uses existing
> > accessible voting machines.
> This is exactly what happened with access to public transportation: at
> first, there was almost no ridership, mostly due to low awareness and low
> expectations on the disability side, and poor training and maintenance on
> the public transit side. Over the years, as people with disabilities began
> to experience and expect accessible buses, ridership continues to grow. At
> the same time, non-disabled riders use many of the features originally
> intended to serve disabled riders. There are other such examples of good
> technology meeting good policy, and usability meeting accessibility.
> Almost
> enough to tempt one into technological utopianism....
> I'd like to think that an open source project such as ours could do a much
> better job than a proprietary enterprise of both adding accessibility where
> we already know it's missing, and initiating a wide-ranging dialogue about
> where accessibility can be balanced against other policy factors.
> ***
> Jim Tobias
> Inclusive Technologies
> +1.908.907.2387 v/sms
> skype jimtobias
> > -----Original Message-----
> > From: Alan Dechert []
> > Sent: Saturday, June 27, 2009 2:17 PM
> > To: Open Voting Consortium discussion list
> > Subject: [OVC-discuss] Privately and independently -- how
> > will OVC deal withthis?
> >
> > I've thought about this a bit more, and I have arrived at a
> > few conclusions.
> > Eventually, we will have to design an add-on device (or
> > perhaps several different devices) for accessibility. In
> > thinking about the AutoCAST feature ES&S claims to have
> > developed for the AutoMARK, we could probably make something
> > that works with our system - maybe better than the AutoCAST.
> >
> >
> >
> > Depending on the nature of the voter's disabilities, the
> > device could plug into, say, three USB ports: one for the
> > barcode reader, one for the mechanical paper handling device,
> > and one for sip-and-puff input. A blind quadriplegic could
> > use it - so long as he or she could hear via headphones.
> > After review of the printed ballot, the voter could choose to
> > cast or reject the ballot. The ballot would then descend
> > through one chute or the other, into the ballot box or spoiled bin.
> >
> > But this brings up the problem we have with the Holt bill.
> > Who are we designing this for?
> >
> > Then again, the problem is not so much with the Holt bill,
> > but the HAVA bill itself. HAVA says the voter with
> > disabilities needs to be able to vote independently and with
> > privacy. It's mentioned several times in HAVA. Here is one example:
> >
> > . need to make voting equipment fully accessible for individuals
> > with disabilities, including the blind and visually impaired, the
> > need to ensure that such individuals can vote independently and
> > with privacy,
> >
> > This language is highly problematic. "Privacy" is not really
> > a big problem.
> > We can pretty easily understand and define what's meant by
> > that. The voter should be able to indicate selections and
> > cast the ballot without revealing those selections to anyone else.
> >
> > The big problem is what is meant by "independently." My
> > opposition to Holt is not because I don't care if people with
> > disabilities can vote privately and independently. It is
> > because "independently" is poorly defined.
> >
> > I believe that the litigation will be necessary to get this
> > definition.
> > Strategically, we should get the litigation going right away.
> > This could freeze Holt until we get a better definition of
> > "independently."
> >
> > We need to know two things:
> >
> > 1) For whom are we designing the accessible device? For
> > example, what if our system works for a blind quadriplegic,
> > but not a short and obese blind quadriplegic?
> >
> > 2) What is meant by independently? For example, does it
> > really mean the voter will not need assistance reviewing and
> > casting the ballot? If the voter rejects the ballot, must
> > the system allow restarting without assistance?
> >
> > Here's an outline of what needs study:
> >
> > We need to work up profiles of voters with a variety of
> > disabilities and combinations of disabilities. For the sake
> > of argument, let's say we develop 100 profiles such that all
> > of the potential voters with disabilities can be categorized
> > in one of these profiles. Disabilities taken in to
> > consideration would include such things as dexterity,
> > cognition, sight, hearing, mobility, as well as unusual size
> > and shape. These things could even include allergies (maybe
> > they have to wear a mask if they go out). For example,
> > profile 73 could be for a short and obese blind quadriplegic
> > with poor cognition (we may need a lot more than 100 profiles).
> >
> > Then we need to collect data on people that fit these
> > profiles. And we need to get some answers about these
> > people. One of the issues not dealt with in HAVA is that
> > people with disabilities tend to vote at home with whatever
> > assistance they need. Even if a machine would allow them to
> > vote privately and independently at the pollsite, they're
> > just not going to use it anyway.
> > Informal studies show that practically no one uses existing
> > accessible voting machines.
> >
> > - How many people fit this profile?
> > - If a pollsite voting machine that works for people in this
> > category, what percentage of them would go there and use it?
> > - Do people in this category tend to congregate? If so,
> > where? (for example, it might be that people in this
> > category can't live in MN. they all move to Ourtown FL. So,
> > there is no point setting up a machine in MN that meets this
> > specific disability when they all live in FL).
> > - Can people in this category be best served with a portable
> > voting machine brought to their home?
> >
> > Then we need to propose some thresholds. What if people in
> > category 73 make up .0002 percent of the population? Of
> > these, say, only one in 50 would go to the poll site to vote
> > if a machine was there to accommodate them.
> >
> > I don't have any specific numbers for thresholds, but some
> > such numbers must be used somewhere. Bottom line is that we
> > need something that says something like, "if there is a less
> > than one in five chance that a voter in this category will
> > come to the pollsite on Election Day, then the accessible
> > voting machine does not need to accommodate voters in this category."
> >
> > In other words, we need to define all the categories, figure
> > out where they live, and then figure out for each category
> > the odds that the voters in this category will go to the pollsite.
> >
> > Until we have this information, we can't really know what's
> > required to meet the independence requirement. In addition,
> > some other specifics will be needed in the independence
> > definition. For example, is it okay if the system requires
> > the voter to ask for assistance in order to print another
> > ballot in case the first printout is rejected?
> >
> > So, the nature of the lawsuit might be that we demand that
> > the government provide this information. The HAVA
> > accessibility requirement can't be enforced without it.
> >
> > Alan D.
> >
> >
> >
> > _______________________________________________
> > OVC-discuss mailing list
> >
> >
> > By sending email to the OVC-discuss list, you thereby agree
> > to release the content of your posts to the Public
> > Domain--with the exception of copyrighted material quoted
> > according to fair use, including publicly archiving at
> >
> _______________________________________________
> OVC-discuss mailing list
> By sending email to the OVC-discuss list, you thereby agree to release the
> content of your posts to the Public Domain--with the exception of
> copyrighted material quoted according to fair use, including publicly
> archiving at

OVC-discuss mailing list
By sending email to the OVC-discuss list, you thereby agree to release the content of your posts to the Public Domain--with the exception of copyrighted material quoted according to fair use, including publicly archiving at
= The content of this message, with the exception of any external
= quotations under fair use, are released to the Public Domain
Received on Tue Jun 30 23:17:17 2009

This archive was generated by hypermail 2.1.8 : Tue Jun 30 2009 - 23:17:20 CDT