Here is some brand new analysis from Washington state results that might shed light on the efficacy of the top two primary, which many are promoting as a good thing for CA. It is especially directed at whether the top two would elect more moderates -- or more extremists? This evidence below suggests it's a bit of a crapshoot, the top two primary could as easily elect more extremists as elect more moderates.
In taking a look at official WA state election results at http://vote.wa.gov/Elections/WEI/Results.aspx? for last year's primary, you can see there are basically four categories of results for the 98 house races and 25 senate races.
In the first category, which has by far the vast majority of races, one candidate (usually an incumbent) is either uncontested or is so far in the lead with anywhere from 53 percent to over 70 percent of the vote and a huge enough lead that it's obvious they will win in the general (November) election as well. That includes 24 races uncontested in the primary, and 3 with only token write-in opposition. The practical impact in those races is no different from what we have now in CA, as I outlined recently in my Los Angeles Times oped.
In the second category, there were FIVE primaries that were extremely close with only two candidates (one D, one R) in both the primary and general elections. Not a huge number of close races for 123 races.
In the third category, there were TEN primaries with 3-5 candidates, where the front runner had between 40-50 percent of the vote, and in 7 of those races the expected thing happens -- the front runner has a fairly solid lead and wins by picking up votes from the supporters of those eliminated, since most of those districts have a pretty clear partisan tilt. In two of those races the second place candidate comes from behind and wins, but that's completely expected since the third place candidate is from the same party in a district that is tilted toward their party. And in one race, a clear swing district in which the frontrunning Dem has 50.4 percent to 37.5 percent for the frontrunning Rep, with another Rep in third with 12 percent, the Rep eventually wins with 51.6 percent (beating a conservative Dem). A little sizzle there, but that's only one race in a swing district. And in all of the races, these results also would not be any different than what we have now in CA.
It's in the fourth category where things perhaps are most interesting. These races illustrate why a split field may result in EXTREME candidates advancing to the top two, not moderates. In this category, there are 5 primary races -- out of 123 races -- where there are multicandidate fields of 4-7 candidates where the front runner has less than 40 percent in a pretty split field.
For example, if you look at District 7, Position 1 you see there were 5 candidates, ALL REPUBLICANS, and the top two in the primary had 26.7 percent and 26.4 percent. In such a situation, with the top two having such a low percent of the vote, it's very possible that one or both of those candidates may not be moderate at all, they could be extremist. Yet one of them will win in the general election, since it was a heavily GOP district.
In the District 40 Senate race, there are 7 candidates, 6 Dems and 1 Rep. The Rep finishes first with 37.8 percent and a Dem is second with 28.3 percent. That Dem could easily be non-moderate, could be a far left Dem with such a low percent of the vote. But in the Nov election the Dem wins with 58.6 percent of the vote, since most Dem voters of course fall in line and vote for the "brand," i.e. the Democrat.
Here's another, District 14, Position 1: seven candidates, 6 Reps and 1 Dem, Dem finishes first with 30 percent with the highest Rep having 22 percent -- with such a low percent, that Rep could easily be an extreme candidate. Then that Rep went on to win in November, since it's a Rep district, with 53 percent of the vote.
District 8, Position 1: 5 candidates, one Dem and 4 Reps, Dem finishes first with 38.2 percent, the top Rep has only 19percent -- extremist or moderate? Then the Rep wins in Nov with 53percent.
This kind of "split vote" dynamic is reminiscent of the strategy Tom DeLay followed for years, providing enough money and firepower to help his extreme Rep candidates win a low plurality victory in the primary over more moderate Reps. And then that candidate would easily win the November election in a heavily Rep district.
It's also a reminder of how David Duke, a former top Ku Klux Klan leader, got into the top two in the 1991 Louisiana race for governor with only 32 percent of the vote. It's also how Jean-Marie Le Pen, a far right politician in France, got into the top two in a recent French presidential election with only 18 percent of the vote. Split votes are a common occurence in plurality elections, and the top two uses a plurality election to determine the top two finishers.
So when I look at these results, I don't see that very much was gained in WA from use of the top two. First, most races are still vastly noncompetitive and predictable, as they were before WA had the top two (and like CA has now). And second, in the five races where you have some real voter "choice" going on, with multiple candidates and a wide open field, the final results are a crapshoot with candidates getting into the top two with low plurality vote totals that can just as easily be reached by extreme candidates as by moderates.