Campaigns & Elections has this essay by Sean J. Miller analyzing the California 30th U.S. House race. This is the race between two incumbent Democrats, Brad Sherman and Howard Berman, who were put into the same district by redistricting. They placed first and second in the June 2012 primary and carried on a very intense, personal fight in the general election.
The article illustrates what is wrong with the theory of top-two supporters, that when there are two members of the same party in the November election, and no one else, members of the other major party can determine the winner, and therefore the winner will be a “moderate.” Both congressmen generally agreed on the big national issues, but Congressman Berman was known for his ability to work with Republican members of Congress, so his advertising tried to make that point, and to attract Republican voters. But the more he talked about his endorsement from Republicans, such as Senator John McCain, the more that persuaded liberals and Democrats that Berman didn’t deserve their vote.
The Sherman-Berman race is not an example of how top-two changed the winner in California. Even without top-two, Sherman still would have won the race, but he would have done so in the June Democratic primary, and the amount of money spent on this race would have been less. Thanks to Nancy Hanks for the link.