What to Make of John Judis' 'Republican Advantage'
John Judis’ recent piece describing what he calls an emerging Republican advantage has created a feeling of whiplash among analysts. After all, how does one go from confidently predicting a Democratic realignment to predicting a Republican one in a few short years?
I’ll confess that this was my initial reaction. But upon further reflection, Judis isn’t really arguing for a pro-Republican realignment, at least not in the way he argued for a realignment toward Democrats in the aughts. If you watch his discussion here with The Crystal Ball’s Kyle Kondik, Huffington Post’s Mark Blumenthal, and professor Tom Schaller, he is making two claims. First, he argues that Republicans have neutralized any emerging Democratic advantage by performing well among working-class whites and middle-class voters. Second, he says that all other things being equal, Republicans are now more likely to win a trifecta -- control of the House, Senate and presidency -- than the Democrats. This makes it easier for Republicans to advance their agenda, and gives them an advantage.
This second argument doesn’t really rely upon the sorts of things that tripped up the “Emerging Democratic Majority”: long-term, straight-line projections of demographics and cyclical theories of American politics. Instead it looks more at structural factors in the House, Senate and presidency, and how they interact with the political coalitions as they are currently constituted. Viewed through this lens, I think he might be on to something.
Consider: Most analysts believe that, due to gerrymandering, the Voting Rights Act, and the clustering of liberals in urban areas, Republicans have an edge in the House of Representatives. In fact, a fair number believe that Republicans have something of a lock on that body. I don’t think the latter claim is justified, but it’s impossible to ignore the fact that, using Cook PVI as a measurement for base partisanship (that is, controlling for national effects by subtracting the president’s national vote share in the past two elections from the district’s vote for president in the past two elections), there are only 186 districts with Democratic leans.
The median congressional district has a PVI of R+2. This means that the easiest winning House coalition for the Democrats involves sweeping all of the Democratic-leaning seats, evenly matched seats, and R+1 seats, then winning 12 of the 14 R+2 seats. If they lose any of those seats, they have to make it up with a win in a district that is R+3 or greater. To put this in perspective, Republicans control half of the D+2 seats, 75 percent of the D+1 seats, half of the “even” seats, 75 percent of the R+1 seats, and all of the R+2 seats. Democrats hold only five seats that are R+3 or greater.
So let’s say that, under current conditions, Republicans will win the House 80 percent of the time. I tend to think the presidency is a 50-50 shot, but let’s go ahead and give the demographics-as-destiny theorists the benefit of the doubt here. After all, even the most ardent proponents of the emerging Democratic majority will concede that recessions, wars and scandals will result in what political scientists called “deviating” elections. These happen with some regularity, but let’s settle on Democrats winning the presidency 70 percent of the time.
Finally, we should look at the Senate. Let’s engage in this thought experiment: Suppose that, over time, Republicans should win all of the GOP-leaning states (in terms of PVI), Democrats should win all of the Democratic-leaning states, and the parties should split the even-PVI states. That means that, over time, Republicans should win 53 Senate seats.
But, you say, this is too strict: Republicans can win D+1 states and Democrats can win R+1 states. Our revised assumption actually hurts Democrats further. If we assume that R+1/D+1 states should also be split evenly, then Republicans should win 56 Senate seats over time. If our cutoff for competitiveness is R+2/D+2, Republicans should win 58 seats. If our cutoff is R+6/D+6, Republicans should win a filibuster-proof 60-seat majority over time.
What the self-inflicted GOP wounds of the past few cycles conceal is that Republicans presently have something of a natural Senate majority. There are many more solidly Republican states than solidly Democratic states: 23 states are R+5 or more but only 12 states are D+5 or more. So again, over time, Republicans are going to be favored to win the Senate, as they can practically capture a majority without winning a single swing seat. Let’s again be generous to Democrats and say that Republicans win 60 percent of the time.
Using these proportions, Democrats win trifectas a little less than 6 percent of the time, while Republicans win trifectas a little more than 14 percent of the time. Remember, though, we were being generous to Democrats in our assumptions. Assume the presidency is a 50-50 shot, and that Republicans will hold the Senate 70 percent of the time (about where I estimated things using a Monte Carlo simulation in 2013), and Republicans should win the trifecta 28 percent of the time, while Democrats should win the trifecta just 3 percent of the time.
It’s true that these things aren’t actually random: Years where Republicans do well in House races are also likely to be years where they also do well in Senate and presidential races (and vice versa). But at best, that just improves both Republican and Democratic chances of winning trifectas.
In reality, it probably hurts Democrats more, because an advantage in the presidency is bad for trifectas. Jay Cost recently explained the tendency of Americans to “auto-correct” by voting for the other party in off-year elections. Because of this, the tendency will be toward what we saw between 1952 and 1988.
During this time, the GOP won seven of 10 presidential elections. When they took control of the presidency, the party was actually in decent shape: 192 House seats and 53 Senate seats in 1980, 192 House seats and 43 Senate seats in 1968, and 221 House seats and 48 Senate seats (a majority, at the time) in 1952.
But after two terms in office, the party was in terrible shape, as they were pummeled over and over again in off-year elections. Democrats may have lost 70 percent of the presidential elections during these time periods, but when they did win the presidency, they were well positioned to enact major reforms: 258 House seats and 57 Senate seats in 1992; 292 House seats and 61 Senate seats in 1976; 262 House seats and 64 Senate seats in 1960.
This is what winning the presidency 70 percent of the time actually looks like: Midterm losses mean that the party digs itself a hole that is very difficult to come out of, and when the other side manages to win, it is well positioned to enact its agenda.
So, in this narrow sense, Judis has a point. Of course, we should remember that the Democratic coalition of today will not be the Democratic coalition of 10 years from now, so any Republican advantage may well prove fleeting. But if we limit ourselves to the near future, current structural forces probably make the advantage real.