It didn't take long in the wake of the 2012 elections for Democrats to point out an inconsistency: The party won the popular vote in House races by more than 1 million votes, but the Republicans still controlled more seats. This was fodder for all sorts of prognostication, focusing on redistricting and the "Big Sort" as possible rationales.
That same scenario didn't repeat itself this year, however. In fact, 2012 is one of only two times in the past 12 cycles that the winner of the House popular vote didn't also win more seats. The other was 1996, when Democrats barely won more of the popular vote. Both years followed strong shifts in control of the House.
Looking at the data another way, 2012 is the big dot on the graph below, the one point that's distinctly not in either the lower left quadrant (where Democrats win the popular vote and more House seats) or the upper right (where Republicans triumph in both).
Also note in the first graph that, since 1992, Democrats have received more of the popular House vote in four of six presidential cycles. Republicans have received more votes in five of the six midterm cycles. So 2016 seems to be setting up as a possible repeat of 2012: a presidential year following a dominant Republican performance. It could be a much better test of whether the Big Sort is providing a substantial long-term benefit to the GOP -- or if 2012 was an outlier.