Thanks to election technology expert Noel Runyan, I can explain another reason why the U.S. election systems market is under-served by today’s for-profit vendors of election technology. (And to read up on several other reasons, see Noel’s congressional testimony.)

The issue is simplicity. Election systems vendors have no
motivation to design systems for simplicity, and in fact have the usual
competitive motivation to add features and functions to outshine the offerings
of competing vendors. Yet this motivation is 180 degrees off course from
serving the public good, because simple products are better, in this sense: the
more complex a product is to administer and use, the more complex the polling
place procedures, and the more likely that hard-pressed volunteer poll workers
will make errors that could lead to perceived election irregularities.

Here’s one example: in New York state, a typical poll-worker manual
is over 160 pages of background and detailed instructions. Though county
officials make good efforts at training, and though most poll-workers are
diligent, there just a whole lot of stuff to get right, really more than can be
covered in training. And in most cases, there’s not a lot of accountability for
lapses in correct procedure by individuals in an almost-unpaid volunteer
workforce.

It’s not an uncommon situation for a voting machine to have
displayed some irregular behavior, and for the first attempted explanation to
be “operator error” by poll workers. Often, this is not the case, but it
illustrates the completely backwards mindset around features and complexity. Noel
Runyan:

I completely agree with the need to simplify polling place
procedures and poll-worker interfaces to the voting equipment. In my own
experience during our last six elections, inadequate knowledge prevented the
poll-workers from getting the "accessible" voting system working one
third of the time. I don’t blame the poll-workers for this, as I feel
that the systems, including polling procedures, were far too complicated for
typical poll-workers with reasonable amounts of training. Piling huge demands
on the poll-workers to make up for design flaws in the voting equipment and
then blaming most of the failures on poll-workers is just plain wrong.
Over complexity in the poll-worker human interface to voting systems is a
vendor system’s design flaw, not a poll worker flaw.

In other words, vendors effectively drive changes to voting
procedures, rather than voting procedures being requirements for voting system
products. The result is, basically, just a whole lot more to get wrong.

Why is that so terrible? Because our democracy depends in no
small part on confidence in election results, and that confidence is hard to
maintain as it becomes increasingly frequent that polling places are not
correctly run “by the book” because the book is (in the case of one poll-worker
I’ve worked with) just to darn heavy to lift! Complexity leads to too much
heavy lifting – there’s your metaphor of the day for voting technology misfit.

— EJS

(PS: Of course, it is possible to build voting systems that
meet the customers’ requirements, but otherwise minimize complexity; but today’s
vendors have no reason to go back to the drawing board. But OSDV does; that’s
part of the basic mission.)