Count my votes, please!

Been there before

The photo-finish in the race for the Democratic nomination for state controller in California has prompted observers to invoke the controversy over the vote-counting scandal in Florida's presidential election in 2000. I prefer, however, to invoke 1980. That's the year that an extended recount resulted in a reversal of the original results. The initial tally of votes from election day awarded a seat in the state assembly to Republican candidate Adrian Fondse. The Democratic candidate, Patrick Johnston, refused to concede and pursued a recount. The process took long enough that Fondse was sworn into office as an assembly member and took his seat in Sacramento, the recount hanging over his head like a dark cloud as Johnston narrowed the gap in incrementally released results.

Fondse, however, found a silver lining. He noted that he had done quite well on election day in the precincts yet to be recounted, so he was confident that his victory would be sustained. Despite his optimism, Fondse found himself trailing at the end of the recount and lost his assembly seat to Johnston in January. (I was in the assembly visitors gallery on that contentious day and observed the desperate last-ditch political maneuvers, including a motion by Fondse's own Republican colleagues to oust him from his seat—a motion sure to fail because it required a two-thirds vote. The Democrats instead insisted on a simple-majority motion to accept the recount results, resulting in Johnston's swearing-in as the winner of the election.)

What's the lesson we should learn from the Johnston-Fondse recount battle? Has John Pérez taken it into account in his decision to demand a recount in his razor-thin loss in the controller's race to nominee-apparent Betty Yee? Pérez had exercised his right under state law to cherry-pick the counties in which recounting is done. He chose those in which he had beaten Yee by the greatest margins. Yee's supporters immediately cried foul, clearly indicating their belief that Pérez has stacked the deck. Let's examine this assumption accepted by both camps.

Suppose there's a precinct in which Pérez enjoyed 100% support. No votes at all for Yee or anyone else. What could happen during a recount? Examine a ballot. It was counted for Pérez. If a mistake is found, Pérez loses that vote. It must then go to Yee (or one of the other candidates for controller). It should now be clear that Pérez's strategy of recounting only his strong precincts causes his own votes to receive greater scrutiny than others' votes. It's a “please double-check my votes” strategy. The degree of risk is directly proportional to the size of his original vote.

We can put an asterisk on this analysis, because nothing is ever simple when it comes to vote-counting controversies. In our imaginary 100%-Pérez precinct, suppose a new and uncounted ballot is discovered in a ballot box. Chances are that it's a vote for Pérez, given the nature of the precinct. If previously uncounted votes are turned up during the voting process, then Pérez has chosen the right strategy, having a decent expectation of turning up neglected votes in his favor and thus increasing his total. If not, Pérez and company are spending a lot of money to put his votes at risk.

An interesting choice, but all of the political commentators seem to endorse it as obviously advantageous to Pérez. It ain't necessarily so.