On the surface, it seems hard to argue with the standards set by California's new Secretary of State Debra Bowen for the first-ever large-scale tests of the security and reliability of electronic voting machines used in most of this state's elections since 2004.

For both direct recording electronic voting systems (touch-screens whose votes are automatically recorded) and for vote tabulating devices (machines that scan paper ballots to count votes), Bowen demands andquot;features that secure ... against untraceable vote tampering.andquot;

That seems pretty simple and unassailable. Essentially, she wants no system in use for any upcoming election that might allow tampering with votes or vote totals.

But, predictably, one of the two interest groups with the highest stakes in keeping all current electronic voting systems in place has already objected vehemently.

That would be the county voting registrars who bought tens of millions of dollars worth of virtually untested voting machines over the last few years. They don't argue with the notion that there should be no possible tampering. Rather, they whine that they couldn't get new, paper-based voting and counting systems ready by next February's presidential primary election if tests find the machines they bought are hackable. They claim they'd have serious problems, even if testing is completed by August, as Bowen plans.

andquot;Time is not our friend here,andquot; complained Steve Weir, clerk-recorder of Contra Costa County, also president of the California Association of Clerks and Election Officials. andquot;If we're looking at a wholesale change in voting systems, we're sunk.andquot;

Who is sunk? What Weir didn't say is that if the voting systems he and his colleagues so eagerly bought while others were questioning them turn out to be unreliable, they will have wasted all that money they spent. And if that becomes obvious, how long can any of them expect to hang onto their high-paying jobs?

Democracy sinks

The basic reality here is that if these machines fail Bowen's test and are nevertheless kept in use, as Weir implies he would want to do, voters and democracy itself are sunk, not just registrars.

Ever since electronic voting became the vogue just after the hanging chads scandal in Florida discredited old-fashioned punch card voting systems, charges of election fixing and voting machine hacking have abounded. There were claims that Diebold Election Systems fixed its Ohio machines to bring that state in for President Bush in 2004. There were claims that machines were hacked last year to assure the election of Republican Brian Bilbray to a San Diego area congressional seat after incumbent Republican Rep. Duke Cunningham was convicted of political corruption. There were charges of irregularities in last year's voting in Arkansas.

And there was the fact the Finnish computer expert Harri Hursti was able to reverse votes cast on Diebold machines in a small-scale test conducted in Tallahassee, Fla. in 2005.

Of these, only Hursti's results are proven. But all these claims raise doubts about election integrity, questions often raised by Bowen during her two terms as a state senator and while she campaigned for secretary of state last year.

These doubts need to be resolved if the results of closely contested elections are ever again to be accepted as reliable and honest.

That's what Bowen's tests are about. Husti warned a Riverside County election panel last month that andquot;The level of expertise that you need (to hack voting machines) is scarily low. A high school student could breach most election systems.andquot; If he's right, these systems must go.

Yet, Weir and other registrars maintain the security in place during elections prevents any serious hacking problems. What security, ask doubters? They note that hundreds of voting machines have been carried home by precinct workers weeks before various elections. True, software is locked into these systems before they leave registrars' offices with andquot;tamper-evident seals.andquot; But skeptics maintain there is no such thing as a completely tamper-evident seal, one that can't be removed and then replaced after a machine has been reprogrammed without anyone being able to tell.

If Hursti's earlier hacking efforts actually do not reflect real-world conditions, that should become evident in Bowen's tests, which in part will pit bands of computer scientists against one another in exercises comparable to war games. No matter how determined they may be in their contests, these experts surely cannot be more motivated than hackers might be if they were paid to fix election results.

Right stance

If any of the several current election systems fails, Bowen vows to decertify it immediately, giving county officials at least five months to put paper ballots in place for next year's voting.

That's exactly as it should be. For if voters cannot be completely confident their ballots will be counted accurately, what's the point in voting, what's the point of conducting elections at all?

Which is why the testing must go on and with all results made public regardless of any whining of by voter registrars and other self-serving county officials.

Reach Thomas D. Elias, a long-time California political reporter: