Machine Error The Case for Paper Ballots
Machine Error The Case for Paper Ballots
Hal Plotkin, Special to SF Gate
Thursday, December 21, 2000
,/p>
URL: http://www.sfgate.com/cgi-bin/article.cgi?file=/technology/archive/2000/12/21/evoting.dtl
Now that the jokes about hanging chads are beginning to subside, many are looking to new voting technologies in the hope they will help us clean up the sloppy election procedures that surfaced so alarmingly in Florida.
Unfortunately, one of the most talked about of these potential reforms, electronic voting terminals, has the potential to undermine what little public confidence remains in our beleaguered electoral system.
Without doubt, better technology would certainly improve our system of voting. But as the Florida experience demonstrated, we also need to make sure election results can be verified in an open, reasonably fast and totally transparent way. Anything less will put public faith in our democracy, and perhaps our democracy itself, at risk.
That’s why I’m convinced we must reject the electronic voting terminals that have recently been proposed unless they are coupled with a system that creates paper backups of each ballot cast. The paper backups might never be needed. But their very existence would safeguard the honesty and credibility of the process while discouraging electronic tampering.
Voting officials from both major political parties are pushing for fully electronic voting systems. In fact, we’re beginning to see the start of a potentially dangerous stampede to the new technology despite a very fundamental flaw.
First, the good news:
The ATM-like voting systems now under consideration, which are already in operation in some communities, are easy to use and can even be programmed to ask voters to review and approve their selections before each vote is finalized.
Backers of this new system favor it, in part, because it does away with paper ballots, which are cumbersome to count by hand and which can too easily be mismarked. The ATM-like voting process, by contrast, is clearly much easier to use and far more reliable than present alternatives.
Unfortunately, though, election results from the ATM-voting machines now under consideration can only be verified by electronic means. After the voting is complete, you push a button and you get the numbers. Push the button again and, presumably, you get the exact same numbers.
The lack of paper ballot backups is the fatal flaw. The digital elite may disagree, but it’s just too easy to cast doubt on election returns without them. What’s more, the absence of a paper audit trail serves as an enormously powerful invitation to mischief.
Twenty years of covering Silicon Valley as a reporter may have turned me into a cynic. But I’m sure I’m not alone when I say I’m not ready to put my unqualified trust in any machine when it comes to something as sacred as counting my vote. At least not when a kid in Bulgaria can still penetrate our defense department networks using little more than a PC and a telephone.
The new electronic voting machines can certainly be a part of the solution. But they are not, by themselves, the answer. Instead, here’s what it will take to put together a credible and believable electronic voting system:
In addition to storing and forwarding results electronically, the new ATM-like voting machines should also print out a hard-copy ballot facsimile after each ATM ballot is cast. The hard-copy facsimile would be the property of the election officials and would be used only in the event the electronic results were brought into question. We could also consider giving voters a copy of their voted ballot facsimile for their own records.
That way we’d get the verifiability that only a review of paper ballots can provide without their many curses, such as the improper markings we’ve all heard so much about. The printing devices could easily be equipped with optical scanners and redundant systems that continually test ballot facsimile printing quality and clarity.
The other alternative now being talked about, the entirely electronic system, would, no doubt, be cheaper.
But such a system would do little to restore or protect public confidence in an already shaky system.
For the record, the computer security experts pushing ATM-voting scoff at the security concerns that bother me. In particular, most of them see little need for something as prehistoric as paper ballot backups.
They point out, for example, that the best electronic voting systems already have redundant encrypted data storage modules which archive results independently of one another. Backers of ATM-voting have also proposed the extension of laws that require the public registration of source code from all electronic voting machines, which, in theory, would make tampering with machine tabulations more difficult.
Pretty universally, the experts say its safe to trust the machines.
It reminds me of the exasperating comments made by former Secretary of State James Baker during his successful drive to halt the hand-counting of disputed ballots in Florida.
In making his case, Baker claimed that machines don’t make mistakes or, if they do, they make them in a random fashion that favors no one candidate.
As anyone who has ever used a computer or a cash register already knows, Baker was only half right.
Machines certainly don’t make deliberate mistakes. But the people programming machines can and do make mistakes. They can also commit fraud. And the fewer people involved in any process, the more likely fraud becomes.
Many Americans were not pleased when five Republican U.S. Supreme Court appointees effectively decided our most recent presidential election. I wonder how we might feel if our next election is decided in secret by a handful of unappointed computer programmers?
To be sure, the chances are quite slim that an electronic election could ever be stolen, given the many safeguards that are contemplated.
But perception is just as important, maybe more so, than reality. Fewer people will trust reported election results, whether they are accurate or not, if there is no way they can be independently verified. That may cost more. But the dollars involved would be a small price to pay in exchange for a more credible system of self-government.
After all, up until recently one of the beauties of the American system of voting was its utter transparency. Under proper supervision citizens and candidates could always inspect ballots after an election. It would be a blunder of historic proportions to throw that basic and fundamental right away under the guise of improving things.
Our last election left perhaps half the country convinced that the election results didn’t reflect the will of the voters. But as perilous as the last few weeks have been, the danger we now face is even greater.
In the rush to repair our now-discredited election system we might be about to do it permanent damage.
