The article in Friday’s Fox and Hounds titled, “What if the Open Primary were Used for the Governor’s Race in June?” paints a very inaccurate picture of what we might expect were the “top two open primary” in effect for the June election.
The article reports that a survey of 794 high propensity voters would break down as follows in an open primary: Meg Whitman 29 percent, Jerry Brown 25 percent, Steve Poizner 21 percent. The author contends that Democratic voters would vote “strategically” by crossing over to help nominate Republican Poizner as the weakest candidate to face Democrat Brown in the fall. “We believe that Democrats and Liberals are expressing a sentiment not for any candidate in particular but against Meg Whitman. This interpretation might give ammunition to those who argue voters could cause mischief in an open primary system.”
This poll and these conclusions are both counter-intuitive and run against the history of voting in an open primary. First, almost all other polling has shown that Poizner has very low name ID. Why would we think that vast numbers of Democrats would throw away their votes on a candidate they have never heard of? If they wanted to vote against Whitman, as the poll implies, they would logically vote for Jerry Brown. Hardcore liberal Democrats would not go to the polls to vote for the most conservative candidate for governor.
This survey shows the two Republican candidates receiving twice as many votes as Jerry Brown, the sole Democratic candidate. That would require a huge crossover of Democrats voting for a candidate for governor they do not want, and such a phenomenon has never happened in an open primary state. Washington used an open primary for more than 70 years and there is no evidence of people there voting against their personal preferences. They used the open primary in their last race for governor and no abnormal voting occurred.
Fortunately we have hard evidence on how voters do behave in open primaries, because the 1998 governor’s primary was conducted under the former open primary law that was nearly identical to the open primary proposed in Proposition 14 on this June’s ballot.
In 1998, one Republican sought the governorship, then Attorney General Dan Lungren. On the same “open ballot” were three Democrats, Lieutenant Governor Gray Davis, Congresswoman Jane Harman, and businessman Al Checchi. A Democratic tide was building that year, and it was apparent in the two party turnout for the primary. Republican turnout was low for the primary – Lungren got only 34 percent of the total votes cast which closely matched both the GOP registration that year and the GOP turnout. But Lungren got 93 percent of those Republicans who voted. Davis, who won the Democratic nomination, actually outpolled Lungren, getting 35 percent of total votes cast, because Democrats turned out in much larger numbers. He went on to win in a landslide in November.
So the 1998 primary showed little evidence of crossover voting in the governor’s race, and no evidence of voters throwing their ballots away.
In that election, I participated in an exit poll of several Assembly districts to look at crossover voting (this project was sponsored by the Secretary of State’s office as part of the court case challenging the open primary then in effect.)
We found that voters did indeed crossover, but it was most prevalent in legislative races in one party districts where voters in the non-dominant party choose to participate in the opposite primary for the safe open seat. This type of crossing over is called “sincere crossing over” – voters wanted their ballots to count.
There was some evidence of crossing over in other statewide races, some Asian Democrats choose to vote for Republican Matt Fong running for US Senator that year. Independent voters seemed to crossover to find candidates most akin to their own philosophy. At no time did we find any evidence that voters crossed over to vote for the candidate most extreme from their own views. Voters did not go to the polls to vote for someone they did not know or did not like as a way somehow to affect the opposite party’s primary.
The idea that voters “raid” the other party to find the weakest candidate to vote for is unsupported by any empirical evidence or any exit polling. Voter are not stupid; they cast their ballots for the candidates they want to see elected.