It’s now three weeks after the 2014 midterm elections, and a good time to reflect on how serious analysts differ on what happened and what it means. I did a bit of that at Washington Monthly:

I’m no big number-cruncher, and don’t have access to voter files or other data more sophisticated than exit polls, but my general take (articulated here and here) has been that the big GOP victory was the product of a number of things that happened to coincide in one cycle: a strongly pro-GOP midterm turnout pattern, a strongly pro-GOP “map” (at least for the Senate), a second-term midterm “drag” on the party controlling the White House, and negative perceptions of the economy that also hurt the party controlling the White House. I’ve conceded that individual candidates and campaigns may have won or cost a few contests, and it’s possible voter suppression (in the broadest sense of the term) may have mattered in a few places. I haven’t really come to grips with the idea that an entirely different Democratic message could have turned things around, but it’s possible, though very hard to demonstrate.
In any event, this week we’ve seen Sean Trende of RealClearPolitics demonstrate convincingly that the election wasn’t all about turnout demographics, and Harry Enten of FiveThirtyEight demonstrate convincingly that it wasn’t all about the map.
What these analyses suggest to me is that the real fault line in 2014 interpretation could wind up being between those who think the factors driving the results are cyclical–whether it’s turnout, the map, the stage of the presidency, or the economy, or more likely a combination of them–or non-cyclical. Sean Trende, for example, clearly thinks Obama’s unpopularity was the crucial factor in 2014, and will probably sink Democrats in 2016 as well, despite better turnout patterns, etc. It’s really hard to prove or disprove the transitive nature of approval ratings for two-term presidents to their wannabe same-party successors, because the sample set is so small. But I’m still betting 2014 was mostly a “cyclical” election, just like the last three. That does not mean Democrats are guaranteed victory, by any stretch of the imagination, but does mean the winds should shift and give them a shorter and straighter path.

Some “analysts,” of course, who are engaged in Republican triumphalist spin, simply assume without bothering to demonstrate anything that 2014 was part of a GOP march to power that will culminate in a great gettin’-up morning in 2016. And some Democrats may too casually dismiss without scrutiny 2014 GOP gains in “Democratic” demographic groups–probably a sign of differential turnout patterns but possibly something more–or worry a bit too little about the cumulative effect of Republican gains at the state level. In the end, we probably won’t completely understand 2014 until we are looking at a fresh set of numbers two years from now.