Being an incumbent politician seems like a pretty good gig. Members of Congress usually win re-election more than 90 percent of the time. Even in famously tough "anti-incumbent" years like 2010, 87 percent of those who sought re-election won it. And, of course, we're on our third consecutive two-term presidency right now. Incumbents seem to get all the breaks—they have an easier time raising money than their challengers, they're better known, they're more experienced, etc. Is it even worth it to try to unseat an incumbent?
In a thoughtful post a few weeks ago, Jonathan Bernstein pushed back a bit against the idea that sitting presidents never get defeated. Yes, it may seem that way lately, but it wasn't always thus.
If you know nothing else about an election, bet on the incumbent to win.
Here are the numbers: Since 1900, there have been 29 presidential elections. In 19 of those, an incumbent was seeking election or re-election. Of those 19, five have been denied re-election: Taft (1912), Hoover (1932), Ford (1976), Carter (1980), and Bush (1992). That means the sitting president won roughly three-fourths of the time. We might also add to that list Truman, who declined to run in 1952 out of legitimate concerns that he would lose, and Johnson, who stepped down in 1968 for the same reasons. In both cases, their party's nominees went on to lose the election. Throw in those two cases, and you have sitting presidents winning two times out of three. Those are good odds—if you know nothing else about an election, bet on the incumbent to win—but they're hardly iron-clad guarantees of victory.
Funny thing is, just two decades ago, we were talking about how vulnerable sitting presidents were. Thomas Patterson's 1994 book Out of Order argued that changes in the media and the party nomination systems had produced presidential nominees who were increasingly disconnected from their Washington power bases and were ever more vulnerable to re-election challenges. Coming on the heels of three incumbent defeats in less than 20 years, this seemed like a serious problem at the time.
So is it still helpful to be an incumbent, even in an era when voters distrust politicians so much? Well, yes—incumbents still tend to do better than non-incumbents. But why? John Zaller makes an interesting argument about Congressional incumbents in his paper "Politicians as Prize-Fighters": there's nothing about the office that makes incumbents much safer. Sure, they have an easier time raising money and have greater visibility than challengers, but those assets don't automatically translate into votes. Importantly, re-election margins don't grow with time; a 10-term member of Congress is no safer than a two-term one. Rather, there's a selection effect occurring. The ones that win are simply better politicians than the ones that don't. And they stay in office a long time because it usually takes a long time for a similarly gifted politician from the opposing party to rise to challenge them.
So generally, yes, it's good to be an incumbent, but mainly because it's good to be the sort of person who ends up being the incumbent in the first place.