is the bottom line of the analysis of William Jacobson, the Legal Insurrectionist. This point is bolstered in the comments by a note on how Rossi lost the Washington Senate race because he blew off the Palin supporters, and failed to connect with the 12% in the top two primarywho voted for his Palin-endorsed opponent in that primary, Clint Didier.
But the more important underlying point is that electoral reform is the Achilles' heel of the Republican party. This is because they rely too much on an unholy fusion of tea-party slash religious right enthusiasm and big money pragmatism. With most electoral reforms, it becomes harder for a smaller portion of our population to dominate an election. This and how Minnesota is in the middle of a recount for its gubernatorial election made me think again about how to improve on "top two primary" rules for statewide elections.
Now, progressives don't typically like "top two primary" election rules, because it effectively excludes third party candidates (who typically don't win anyways) from the general election ballot and may result in two candidates from the same party being on the general election ballot. It also tends to nominate more "centrist" candidates that have appeal for independents and moderates from the other party. The "centrism" of the two finalists is also tainted by lower turnouts for primaries.
But this does not faze me. I am a pragmatic progressive who thinks that "top two primary" is fixable. I also am a believer in Local Third (LT) Parties that specialize in contesting local winnable elections and otherwise vote strategically together in order to move, rather than capture, the political center. Thus, I believe that it's not a good use of third party time and energy to run for a statewide office and that the use of "top two primary" rules for statewide elections (but only for statewide elections) could be a blessing in disguise for third party supporters.
I also understand that the possibility of a party getting two candidates nominated in an "open primary" depends on whether the potential voters strongly prefer one of the major parties. But in any statewide campaign, the percentage difference between the supporters of the two major parties is inevitably less pronounced. It's only in local elections that one of the two major parties has a good chance of getting two of their candidates on the final ballot. And I would not support the use of any sort of "top two primary" in local elections. One of my electoral reforms mantras is that one election rule does not fit all elections. Thus, FairVote's Instant Runoff Voting (IRV) is fine for municipal elections but not necessarily statewide elections. And, a "top two primary" is fine for statewide elections but definitely not for local state elections.
What is more problematic are the differences between primary and general election voters. This makes the two finalists less likely to be truly worthy of the label "centrist". An additional step to promote "centrism" would be to make voters give their approval to two of the candidates in the "open primary". Thus, if I were a strong Democrat/Republican I would vote for the two Democrat/Republican candidates I like the best. And if I were a tea-party enthusiast, I would likely have to vote for a candidate that would not get the full support of Sarah Palin. And, if I were an independent or moderate Democrat/Republican, I would likely vote for my favorite Democrat and my favorite Republican candidate, which would cement the tendency of the open primary to nominate two centrist candidates for the general election.
However, what if the "open primary" happened at the same time as the general election? The idea would be to collapse a "top two primary" into an instant runoff election so that both reflect the greater and more varied turnout in general elections. This could be done easily, without the need for any changes in voting machinery, if we limited the number of candidates to five. The five candidates could be selected by their incumbency or their ability to get distinct signatures or new voter registrations from potential voters. Voters would then pick from twenty options (twenty-one with a none-of-the-above option), a first ranked and second ranked choice from the five candidates. Then, in the first stage, the rankings would be ignored. This would make their two choices become "This election rule would make it so that we get moderate Democrats or Republicans elected to statewide offices. The five candidates could be selected on the basis of incumbency or their ability to gather large numbers of distinct valid signatures from potential voters. The parties would get the right to endorse up to two of the five candidates and a candidate could only accept one party-endorsement, which would then be listed on the ballot with their name. This would eliminate primaries and (much of) the spoiler effect, just like IRV3. Also similar to IRV3, the winner would have to have a majority of votes from the voters who ranked the finalists. approval votes". The number of approval votes for each candidate would be tabulated. The two candidates who get the most approval votes would then become the finalists. In the second stage, the ranking given by voters to their two favorite candidates would possibly be used to determine which of the two finalists is their favorite. The finalist who is preferred by the most voters would then be elected.
So let's say Tom Emmer(radically conservative Republican), Mark Dayton(liberal Democrat) and Tom Horner (A "closeted" moderate Republican) were on the ballot, along with one lesser known radical Democrat candidates who only gets votes from strong Democrat voters' and an independent candidate who only gets independent votes. There are 100 voters: 30 strong Democrats, 25 strong Republicans, 15 weak Republicans and 15 weak Democrats and 15 Independents. The strong Democrats vote for Dayton and then the other Democrat. The weak Democrats read the Minneapolis Star Tribune's endorsement of Horn and decide to vote for Horn and then Dayton. The strong Republicans vote for Emmer and then Horner. The weak Republicans vote for Horner and then Emmer. The Independents vote for Horner and then the other Independent candidate. The tallies for the first round of the election would be: 40/200 Approval Votes for Emmer, 70/200 for Horner, 45/200 for Dayton and 30/200 for the Democrat and 15/200 for the Independent. Thus, the two finalists would be Horner and Dayton. Horn would be the favorite of 25+15+15+15=70 voters. Dayton would be the favorite of 30 voters. Thus, there would be a landslide in favor of Horner, who the typically liberal Minneapolis Star Tribune believed to be the best candidate for governor.
If on the other hand voters only could vote once and Emmer, Dayton, Horner and the other independent were on the ballot. Strong Republicans would vote for Emmer. Strong Democrats would vote for Dayton. Moderate Republicans and Democrats might like Horner, but they wouldn't necessarily vote for him out of fear of wasting their ballot. Let's say only 1/3rd of the Moderates vote for Horner, with the other 2/3rds voting for their preferred party's candidate. Two thirds of Independents vote for Horner and the other 1/3rd vote for the other independent candidate. Then, the vote tallies would be 35 votes for Emmer, 40 for Dayton, 20 for Horner and 5 for the Independent. Thus, we would get the same outcome as is likely to happen in Minnesota after the recount.
So which is better? Well, it obviously depends on your political preference somewhat. However, if this safeguarded a state from ever getting a radically conservative governor then it would be worth it from a pragmatic progressive standpoint. And I think this would tend to take the edge off of recounts, since the differences between the two finalists would not be as dramatic as they are now between Dayton and Emmer and were between Coleman and Dayton. It also would stymie the radical right's ability to push for a Palinesque candidate and that would take some wind out of their sails, which would force a left-ward realignment within the Republican party.