This fails on its talk about individual incumbents being blamed for scandals. The stimulus described two scandalous situations that vary in terms of which parties receive blame, so (A)’s talk of comeuppance for individual incumbents is a part v. whole mismatch.

(B)’s principle should make people more willing to vote out incumbents when all parties are to blame, but the stimulus describes the opposite phenomenon – incumbents keep their jobs.

Put differently, (B) would give people reason to doubt themselves when they blame only one party for a scandal, and that doubt should make them less willing to punish the allegedly-responsible party. But we see the opposite in the stimulus.

(C) fails on its talk about individual incumbents being blamed for scandals, and on its general attitude that scandal-ridden politicians should get voted out. (C) suggests that in the “all parties are to blame” case, the incumbents should lose their jobs.

(D) fails on its talk about individual incumbents being rightly blamed for scandals and the characteristics of individual challengers being relevant in the decision of whether to replace them.

The stimulus distinguishes between two situations, both of which involve one or more parties being collectively blamed for scandal. The stimulus does discuss a situation in which individual incumbents tend to keep their seats (when all parties are to blame for scandal), but in that situation the characteristics of individual challengers aren’t a difference-maker in their outcomes.

“When a whole party is to blame, punish the whole party if you can.”

(E) takes a step back and points to the stimulus’ narrow domain from the outside: did you notice that the whole stimulus just discusses situations in which parties and not individuals are to blame for scandals? The stimulus is about what happens when we don’t have the option to chuck out a single scandal-ridden politician.

In situations like the stimulus, anyway, where we can only apportion blame party-wide, (E) says we should punish that whole party “whenever possible.” This threads the needle between the blame-all-parties situation and the blame-one-party situation, because when all parties are responsible, it’s impossible to punish them all by voting in their competitors.