This article in the Washington Post’s The Fix explains new political science research by Andrew Hall and Daniel Thompson, of Stanford. Their research shows that when one of the major parties nominates an extremist for U.S. House, that nominee is more likely to lose the general election. This is partly because the extremist candidate’s existence causes increased voter turnout among the supporters of the other major party.
What Donkey/Elephant gerrymander area winner is NOT an extremist ???
Average hack gets about 60-62 percent per area.
IE — very LOW percentage of *marginal* areas.
IE – lots of extremists in party hack primaries.
PR and AppV.
Only depends if opposition chooses a wrong (as means very broken) candidate (again) to beat easily beatable extremist candidate.
This seems to be commentary pretending to be research. How one defines “extreme” is loaded with bias. I imagine the tea party candidates were labeled as “extremist” to the right, but cannot think of any “extremist” to the left nominated by a major party. Other research shows that candidates that appeal to the base of a party have a much better shot of winning than “moderate” candidates.