Will Moving to the Center Really Benefit the GOP?

In a recent blogpost, Matthew Yglesias suggested that moving to the right might not hurt the GOP electorally. But statistical evidence shows that ideology can still make a difference.

Matthew Yglesias writes in a recent blogpost that, setting aside the merits of the issues, Republicans should not worry about potential electoral losses arising from a move to the right.  Yglesias argues:

It's not at all clear... that the heart of [David Frum's] criticism--that Republicans need to moderate in order to become electorally viable--is really true. The empirical evidence to me suggests that our default view about the relationship between ideology and electability ought to be one of nihilism--any challenger can win provided the economy is doing poorly, and any incumbent can get re-elected provided things are going alright.

Yes, but not completely.

There is definitely some evidence that moderate candidates do better. Steven Rosenstone discussed this in his classic 1984 book, Forecasting Presidential Elections, and others have looked into this as well.  For example, my 2008 paper, "Should the Democrats move to the left on economic policy?".

We also have some graphs in chapter 9 of Red State, Blue State, one showing the (estimated) benefits of moderation in congressional elections, and another graph for presidential elections.  The short story is that moderation can get you something like 2 percentage points of the vote (or, if you want to look at it another way, extremism can lose you something like 2 percentage points).

This is all short-term analysis; it neither captures the long-term gains from being able to implement desired policy (if you happen to push extremist candidates and win the election) or long-term loss of credibility from being outside the mainstream.  Nor does it consider asymmetries such as the potential appeal to rich donors of right-wing low-tax policies, or potential feedback effects such as left-wing policies that could lock in political advantages for unions.  (I'm using "right-wing" and "left-wing" to place positions in the U.S. political context, not as value judgments.)

All that is worth studying.  My point here is that, even within the political science world in which "the fundamentals" are what win elections, there still is room for ideology to make a difference.  Not enough to turn around a landslide election, but enough to swing a close one or to make a potential landslide closer than it could've been.