Reply #6: There is nothing wrong with them. [View All]

[Bill Bored]

Studies show they have as low an undervote rate at the top of the ticket (which is the only reliable way to measure undervotes) as any other method. They don't allow overvotes, even with write-ins. They can only be rigged one machine/one race at a time, manually. They do not allow vote switching. And you can't "program" more than one lever machine at a time which makes it very difficult and time consuming to rig an election, esp. without getting caught.

In NY State, the undervote rate is about 3/4%. This is comparable to any other method and much better than some. And there are ZERO overvotes.

Al Gore would have won FL 2000 if that election were held on lever machines.

Sometimes the counters can get stuck on 99, 199, etc. and you get a few unintended undervotes that way, but there is no evidence that this favors any particular party or candidate. And the machines can be inspected to make sure they work in the event of an election contest, even after the fact.

You also can't stuff or steal the paper ballot box, which is one reason why levers were invented in the first place.

They are a low risk alternative to electronic voting and don't be surprised if NY State keeps them beyond 2006 for everyone except disabled voters who have trouble using them.

I know they don't provide voter-verified paper records, but they are the only form of automated voting that is NOT computerized. Computerization is one of the primary reasons for requiring VVPRs and audits, but of course most states don't even bother with that!

Some will say that levers have high down-ticket undervote rates, but in my opinion, this is NOT a reliable way to measure undervotes. I've seen punch card results with lower down-ticket undervote rates than other voting systems, yet it's been well-established that punch cards have among the highest rates of undervotes at the TOP of the ticket.

When you look at the big picture of ALL the risks, including ballot definition programming errors (virtually impossible on lever machines without them being discovered), e-voting system failure rates of 9.2% allowed in any 15 hour election day (according to federal standards!) and the potential for scalable wholesale attacks on e-voting systems both before and after the election, thoughtful people will realize that lever machines have a much lower risk than any e-voting system. The only way to approach that lower level of risk with an e-voting system is to hand count a SIGNIFICANT number of paper ballots and compare them to e-vote tallies. This option does not even exist in about half the states and only a few states actually require it; those that do, have not even defined what a significant number might be.

We are long way from making e-voting as safe as lever machines.