No voting system is perfect, but it seems that electronic manipulation of votes is harder to track than manipulation of paper ballots. That's just my gut feeling, though. I may be wrong.
It made me ask myself, How long did it take for electronic hacking take off after computers came into common use?
It just seems that we have just seen the tip of the iceberg regarding electronic voting manipulation and no doubt more is coming down the pike. In some cases, only circumstantial evidence such as a mismatch between the voting-age population count and the number of electronic votes tallied can give a clue that something has gone wrong.
Well, that and seeing that an Access DB is being used. I couldn't believe that. At the very least, I would want to see:
(1) SQL or Oracle used without access given to poll workers,
(2) some process of encrypting the data,
(3) writing the results to two places for bookkeeping, and only one person with access to both places (for accountability),
(4) a list of everyone who is authorized to access the data and voting equipment.
This list is off the top of my head and I am sure subject to much criticism, but I bet it would be a lot more secure than an Access DB. I haven't specified any usable detail for security, but I would design something around not giving access to anyone who is a poll worker, and instead having an independent, randomly assigned database specialist to assist with any problems using the system. I'd also opt for somewhat cheaper machines and have more of them, so that in case of a problem the whole voting line would not be delayed. Because I would want each reported problem that required tinkering with the machine or getting database help to be logged for security purposes and disclosure.
Probably pie in the sky and full of holes and prohibitively expensive, but again, I think it's a better mindset than relying on Access for such a mission-critical system.
Then there is the issue of which voting system is used to begin with. Even assuming complete data integrity, some have argued that certain voting systems are inherently unfair. For example, Gaming the Vote by William Poundstone. I haven't read the book, and I see from the Amazon reviews that there are some criticisms of it. But it's just another thing to think about regarding voting, even if it is not a data issue as such. Namely, how do you best capture the will of a people as to which candidate most of them want?
Just my two cents,