Why 2012 Postmortems Overstate Republican Woes

Why 2012 Postmortems Overstate Republican Woes

By Sean Trende - April 4, 2013

Since the 2012 elections, there’s been an extended debate over the strength of the current Democratic majority and the future viability of the Republican Party. I’ve long argued that these postmortems are overwrought, and nothing about the results altered that. After all, the demographic changes in that election were more attributable to a surprisingly large number of white voters staying home (fewer whites voted last year than in 2004, despite steady growth in absolute numbers) than to any rapid growth in minority votes. The GOP performed about as well in November as we would have expected given the state of the economy; in fact, the exit polls suggest that the GOP actually fared quite well on key policy questions, demographic changes notwithstanding.

But the debate lurches forward, and has spilled into academia. The Washington Post recently featured a well-argued column by John Sides asserting that Republicans do not need a reboot (he continued this argument on his must-read political science blog, “The Monkey Cage”). This prompted a partial rejoinder from Eric Schickler, which was picked up by Jonathan Chait at New York Magazine. The gist of the argument is summarized by Chait: “The Democrats forged a national majority beginning in 1932. That majority came apart beginning in the mid-sixties. . . . Over the last couple of election cycles, the environment itself has changed. Racial and cultural divisions no longer naturally cut in the GOP’s favor. Another, very simple way to tell the story is this graph of partisan preferences.”

This is the standard-issue realignment narrative -- on which the Emerging Democratic Majority theory rests heavily -- with the added twist of partisan self-identification. The referenced chart is reproduced below. It shows partisan preferences from the late 1930s through the present day. The basic takeaway is that Democrats opened up an increasingly large margin in partisan preference from 1939 through the mid-1960s, which didn’t really close until the mid-1980s. That gap has re-opened in the mid-2000s, and hasn’t shrunk since:

There are two problems with this line of argument. The first is that the chart of partisan preferences doesn’t really do as much for Democrats as Schickler and Chait might hope. The share of the population self-identifying as Democrats has remained more or less constant since the mid-1980s, when Southern Democrats finally decided they’d had enough of the party.

The only salient change has been a diminution in the share of the electorate self-identifying as Republicans, and an increase in Independents. And therein lies the rub. You see, the vast majority of Independents aren’t all that independent. They are “closet partisans,” regardless of how they might label themselves. These Republicans-turned-Independents have almost entirely driven the growing gap in partisan identification, but have actually tended to hold on to their Republican voting habits. This is why the Romney-Ryan ticket was able to win self-identified Independents by five points while still losing the election; this is also what gave birth to the horribly mistaken notions that the pre-election polls were under-sampling Republicans, and that the Republican edge with Independents foreshadowed a big GOP win.

Second, and perhaps more importantly, this entire line of argument rests on a narrative of realignment that simply doesn’t hold up to close scrutiny. Yes, Democrats opened up a massive lead in party identification during this time. But that didn’t translate into the electoral majorities we would expect to see if party identification were as salient as claimed, and the periods of Democratic national dominance were fleetingly small.

By 1938, the GOP had clawed its way back to near-parity in the popular vote for Congress, and in 1942 it actually won the popular vote for the House by almost five points, although the party failed to take a majority of the seats in the chamber (the last time this would happen until 2012*). For a 20-year stretch, from 1938 through 1958, the GOP was quite competitive in battles for the House and Senate. The Democrats once again opened up a big lead in the popular vote after the recession in 1958, which disappeared in the mid-1960s, opened again in the post-Watergate years, and then shrunk in the 1980s.

In terms of the presidency, Democrats fared better, but Franklin Roosevelt’s coalition was actually never replicated after his passing. Harry Truman’s 1948 map more closely resembles Woodrow Wilson’s than FDR’s. The incredibly close race of 1960 -- a majority of states were decided by margins of less than seven points -- was truly sui generis, driven by a Catholic/Protestant split that hasn’t been seen since. Lyndon Johnson’s 1964 win, with its strength in the Northeast and weakness in the South and (relatively speaking) the Plains states, is more properly seen as foreshadowing the Clinton/Obama maps than as the last gasp of the New Deal coalition.

To put this all together, consider the following chart. It takes an average of (1) the Republican share of the popular vote for the House and (2) the Republican share of the popular vote for the presidency. For midterm elections, an average of the popular vote for the preceding and succeeding presidential election is used (e.g., for 1982 I took the average of the 1980 and 1984 presidential elections, and averaged it with the 1982 House vote). To smooth out noise, the data are presented on a three-cycle rolling average (e.g., for 1936, the data presented are an average of the 1932, 1934, and 1936 results). I’ve displayed the results for 1932 through 1984, the time of Democratic dominance in party self-identification:

As you can see, Republicans move to an equal position with Democrats pretty early on. They lose it after the recession of 1958, but regain it quickly. Otherwise, they are more or less at parity with the Democrats. In other words, this edge in partisan identification didn’t translate to massive wins for Democrats then. It is unclear why, assuming the recent Democratic edge continues, we should expect it to translate to wins today. In fact, the claimed Democratic “natural majority” from 1932 to 1968 looks pretty shaky overall viewed in this light.

1 | 2 | Next Page››

Sean Trende is senior elections analyst for RealClearPolitics. He is a co-author of the 2014 Almanac of American Politics and author of The Lost Majority. He can be reached at Follow him on Twitter @SeanTrende.

The Tired Party
Rich Lowry · November 7, 2014
Did We Vote for War?
Pat Buchanan · November 18, 2014
When Republicans Win, Politics Are "Dysfunctional"
Dennis Prager · November 11, 2014
The Other GOP Wave: State Legislatures
David Byler · November 11, 2014

Sean Trende

Author Archive

Follow Real Clear Politics

Latest On Twitter