[EM] Cost of Manual Counting vs. Machine Counting

[Abd ul-Rahman Lomax]

At 03:51 PM 5/25/2007, Jonathan Lundell wrote:
>I've used double-data-entry for some small internal elections in the
>Green Party, and it's been fairly effective. The additional cost is
>mainly recruiting extra volunteers. The extra time is minimal, since
>the two entry teams can work in parallel (or you can publish
>provisional results and then do the second count).
>
>Two details that complicate things a bit is that a) some counting
>rules are dependent on the order in which ballots are counted, so you
>have to make sure that the data entry proceeds in the same ballot
>entry in both cases, and b) if you find a discrepancy in the data
>entry, you need to be able to find the ballot in question, to resolve
>the discrepancy. In my small elections, I simply number the paper
>ballots as they're being entered the first time, and then pipeline
>them to the second data-entry group.
>
>I've never done it for more than about a hundred ballots, though.

My own analysis presumed plurality elections, with one race counted 
at a time, which allows counting through sorting, which is much 
faster than recording votes, generally. Though with proper 
procedures, you can make vote recording quite fast.

Time and motion people are good at this, generally. But election 
clerks are usually not time and motion people, and their jobs don't 
depend on being efficient. In one sense this is reasonable: counting 
ballots is not their usual job, its only something that they do 
occasionally. So why bother optimizing the procedures, it's faster 
just to count the ballots.

What's called for is an analysis and recommendation of procedures 
that could be widely used. Thus each clerk does not have to work out 
the details, run the experiments, maximize the accuracy while 
minimizing the effort.

Exact entry of ballots, where a computer record is created of each 
individual ballot content, is, in my view, highly desirable. But I 
was first comparing hand counting with machine counting in an 
environment where that isn't done currently.

It's inherently more expensive if done by hand. So this is a task 
which shouldn't be done by hand. I'll note that if paper ballots are 
used, they can be transported to wherever automated counting equipment is used.

However, the task of recognizing votes on paper ballots with scanned 
images of ballots, by computer, is, quite simply, not a difficult 
one. I would expect a system using standard computers and scanners -- 
and any fax is a scanner -- could be put together from equipment 
lying about, and the software could, again, be public source. I.e., 
free to the user, donations appreciated. (Public agencies could 
legally issue a grant to an organization to develop such software and 
make it available for free public use, but I don't even think this 
would be necessary. People would write the software. For free.)

So ballots could be counted at the polling stations, or at least the 
images could be collected there. Then the ballots are sealed and 
remain for confirmation if there is ever any suspicion that the 
images were altered. As readers may know, I'd make image alteration 
practically impossible to pull off successfully by allowing election 
observers to independently image the ballots, whether with a fax or 
with cameras, which would do the trick quite easily and with minimal 
security risk. Imaging and then sealing the ballots reduces ballot 
handling to an absolute minimum, thus greatly increasing security.