For what it’s worth, though, I believe we should consider returning to paper ballots entirely.

When paper ballots are used, voters can check their own paper ballots, and poll watchers from opposing political parties can see if votes were properly counted. However, both these simple safeguards disappear (at least on the first count) when voters use computers.

Computers can make both fraud and error more difficult to detect. While fraud is always possible with paper ballots, there is at least a chance that honest poll-watchers will see the cheating immediately. With computer voting, we will have to rely, at best, on a relatively small number of computer experts. What if other experts have out-maneuvered them?

If an election was expected to be close, wrong results may not look particularly suspicious.

Moreover, various commonly proposed solutions (voter-verified paper trails, random audits) fail to address a serious problem. Once the media announce the winner of any major election, the results may be difficult to overturn.

There is another advantage to old-fashioned methods. If ordinary people cheat with paper ballots, presumably the cheating will come from various sources and will not necessarily favor a particular party.

By contrast, with computer voting, a few dishonest people at a voting machine company could cause widespread but subtle fraud favoring one party in an election. In short, I believe it is dangerous to rely on any machines that are complex enough to be programmed for fraud by their manufacturers (especially with so few companies in the game). And it seems unwise to resign ourselves to a system in which we will be relying so heavily on recounts that may never occur.

Barbara Bazyn, Chelsea