There Are No Permanent Majorities In America

There Are No Permanent Majorities In America

By Sean Trende - July 2, 2009

Over the past few months, Jay Cost and I have expended quite a few keystrokes identifying the problems realignment theory. But remember, our argument is not simply that there is no emerging Democratic majority. We argue that the concept of permanent majorities in American politics is at best problematic, period. Parties adapt to changing circumstances, and generally succeed in keeping the other party at bay. Given the American preference for divided government, it is nearly impossible for one party or the other to achieve a permanent majority.

And yet, the arguments for realignments keep coming. Now it is strategist Mike Murphy arguing that Republicans need to “build an ark.” The fact that both Republicans and Democrats are making the “permanent majority” argument is proof that this theory is not going away anytime soon. This means that the rebuttals need to keep coming as well! So at the risk of forcing you to hear such tunes as killed the cow, there are four additional points on permanent majorities that deserve amplification:

(1) Permanent Majorities are Almost Non-existent in American History. If a party attains “permanent majority” status, we would expect that party to dominate both Houses of Congress and the Presidency for an extended period of time, perhaps twenty to thirty years.

But such periods of dominance are rare. Consider the following chart. It measures Republican strength in the White House and House of Representatives by averaging the party’s percentage of the popular vote in the most recent Presidential election with the party’s percentage of seats won in the most recent House of Representatives election. It begins with the formation of the modern Republican party and continues to the present day. Senate seats are ignored for two reasons: (1) Senators were not elected by popular vote until the early 1900s (the exact date varies by state); and (2) only 1/3 of the Senate (give or take a seat) is up at any given time, so including Senate membership would necessarily serve as an artificial dampener on the partisan swings in the country (both toward Republicans and Democrats).


Notice how noisy the chart is. If there were stable, permanent majorities being formed at any time, we would expect to see long periods of time where Republicans are consistently well above the 50% threshold or consistently well below the 50% threshold. But the longest unbroken period of time we ever see for partisan control of Washington is the time of Republican dominance from just before the Civil War to just before Reconstruction ends. Of course, when most of your political opponents consider themselves a part of a different country and/or are not allowed to vote, it is easy to build a massive majority.

Even if we dampen the effect of outliers somewhat by charting a 3-year rolling average, we still see quite a bit of movement, which is inconsistent with permanent majorities:


Instead of a permanent Democratic majority from 1932-1968, and a Republican majority from 1968-2008, what we see is this:  Republicans get whacked after the depression, but the American public quickly pulls them back to parity. Then the recession of 1958 hits, and Republicans get knocked down again.  The American people pull them back to parity, then the Watergate scandal hits.  The American people pull the Republicans back to parity, then the recession of 1990 hits.  All of this is consistent with minority status driven by events and Republicans either holding the Presidency at inopportune moments or being incompetant, depending on your viewpoint.  This leads to our second point.

(2) The Dominant Alignment Narrative Does Not Match Reality. The dominant narrative of realignments is straightforward: Every 30 years or so, we have a realignment. Political scientists have labeled 1800, 1836, 1860, 1896, 1932, and, well, either 1968, 1980, and/or 1994 as realigning years. If we accept 1968 as the beginning of a pro-Republican realignment, 2008 has come along pretty much right on schedule.

But looking at our chart above, we can observe that years that saw substantial changes in the electorate fail to qualify as “realigning.” Moreover, years that are often described as realigning in fact often fail to produce lasting change.

What we see instead is an electoral history that is largely driven by events. In the midst of the supposed Republican alignment from 1860-1896, we see a lengthy period of slight Democratic dominance, beginning once Democrats could vote again. In the wake of the Panic of 1873, Democrats gained 94 House seats in a 293-seat House (the equivalent of 140 seats in today’s House). Democrats lost control of the House only in 1880 and in 1888, and won the popular Presidential vote in every year except 1880 (when Garfield won by only 1,900 votes).

So if you were Grover Cleveland sitting in 1892, you would have pretty good reason to believe that you had a pretty solid majority. Your party had won the popular vote three elections in a row (but for the Populists, you’d probably have had an outright majority in 1892), and you occupied 238 House seats in the 1890 elections (the equivalent of 312 seats in today’s House); this number fell only modestly in 1892. But events intervened, and one of the nastier recessions in American history hit. Some economists estimate that unemployment was 18.4% in 1894 , and the Democrats lost an astounding 125 seats that year (152 seats in today’s House). Republicans were actually lucky to have lost in 1892, as it enabled them to win in 1896.

By 1900 unemployment was back to 5%, and the public was in no mood to throw the Republicans out. This begins a period from 1894-1932 that most historians classify as a Republican alignment. But to do this you have to explain away the massive dip in Republican performance that occurs from 1910 through 1916. The Bull Moose/Republican split of 1912 did not help matters, but even in 1910 the Democrats were able to pick up 58 seats and control of the House, and they were able to retain solid control of the House after the 1914 midterms. Had Woodrow Wilson not botched the end of World War I (and hence the 1918 midterm elections) by calling for a League of Nations, had a sharp recession not appeared post-war, and had anarchists not been blowing up politicians, cities and Wall Street, the Democrats might not have had such a deep hole out of which to climb in the 1920s. In other words, even this supposedly dominant period for the Republicans was subject to and driven by events, the economy, and the parties’ governance.

We've already discussed the inconsistencies from 1932-2008.  But note how much our chart from 1968-2004 resembles the chart from 1942-1960, the supposed heyday of the New Deal alignment.

The bottom line is that permanent majorities are not predetermined, inevitable events. We can go through a series of thought experiments to illustrate this: What if the Populists had been stronger and Benjamin Harrison had been re-elected in 1892 and presided over the Panic of 1893.  What if Populists had been weaker and William Jennings Bryan hadn’t been the Democrats’ nominee in 1896, 1900, and 1908? What if TR had won at the Republican convention in 1912, and had led a reluctant America into World War I? What if Warren Harding hadn’t died, had stood for re-election and lost in 1924, and a Democrat had been re-elected in 1928? What if Herbert Hoover, widely viewed as an unbeatable nominee for either party, had remained a Democrat (he won the 1920 New Hampshire Democratic primary)?

What if the recession of 1981-82 had lasted six months longer? We’d probably be talking about how foolish Republicans were to believe that a conservative could win in our New Deal country, and we would view Ronald Reagan as an accidental President. What if Michael Dukakis had maintained his pre-convention lead, and then presided over the recession of 1990? Would Democrats have an iota of the credibility on the economy that they have today?

These types of questions are purely hypothetical and may seem beside the point, until you come to the thought experiment for today: What if we still have high unemployment and sluggish growth in 2010 and 2012? I think that suddenly this talk of permanent majorities will look pretty silly.

(3) Unified Government Occurs Only Rarely Americans are seemingly very reluctant to give true majorities to either party. Since FDR left the scene, Americans have given unified control to a party for four years or more only three times: 1949-52, 1961-66, and 2001-06 (we’ll forget Democratic control of the Senate from 2001-02, since we’re looking only at what voters actually voted for).

To illustrate this, let’s tweak our charts above a little bit. Instead of looking at House seats that Republicans won and the popular vote from the prior Presidential election, let’s look at the percentage of the popular vote in House elections Republicans won, averaged with their performance in the popular vote in the Presidential election (I only have data going back to 1942 here).


Note how little change there is! The American people have consistently hedged their bets, keeping both parties around 50% (with the exception of the early-to-mid 1960s). And remember, there is supposedly a realignment to a Republican majority that occurs somewhere in here!

Some may argue that the conservative Southern Democrats mask Republican strength. But if we count conservative Southern Democrats as Republicans due to their voting habits, the permanent Democratic majority of 1932-1968 was actually dead in 1938, when the conservative coalition came to dominate Congress (we would also have to then count the Jacob Javitses and Clifford Cases of the world as Democrats instead of Republicans).

I don’t know if this is part of a conscious effort on the part of Americans to achieve divided government, or if governing America is an impossible task, such that the President’s party always gets punished. There’s certainly a strong argument for the latter – after all, no President since Teddy Roosevelt has left office with more Congressmen of his own party than he had the day he entered office. Regardless of the cause, the fact that unified government is so rare should give further pause to those who expect our current arrangement in Washington to survive for an extended period.

(4) Our current realignment is actually very, very old. The standard narrative of our current alignment is that with the passage of the Civil Rights Act of 1964, racist Dixiecrats began to take over the Republican party, while Northeastern moderates began to flee it. But as one reader writes on his blog, , commenting on an earlier post of mine, “the shift in New England from Red to Blue is no recent phenomenon – in reality it began nearly a century ago and probably has little to do with Republicans being ‘too conservative.’” This is a spot-on point that is rarely made, and is true of the South as well.

The change in New England can be summarized by the following example. In 1908, Connecticut elected five Republicans: E. Stevens Henry, Nehemiah Sperry, Edwin Higgins, Ebenezer Hill, and John Tilson. Three of the five sported Yale degrees, and all were elected with 55% of the vote or more. In 1912, Connecticut elected five Democrats: Augustine Lonergan, Bryan F. Mahan, Thomas L. Reilly, Jeremiah Donovan, and William Kennedy. All five attended public schools, and obviously were of very different ethnic stock than Sperry, Hill, et al.

These Democrats were retired in 1914, but the change in New England continued albeit very, very slowly. The chart below shows the Republicans’ share of New England House seats from 1902-2008.


The change which culminated in the extinction of New England Congressional Republicans in 2008 is not a recent phenomenon. It was a decade-long process involving demographic change, ideological change, and the invention of the air conditioner (which lured many affluent, Republican voters to the South).

Similarly, the transformation of the South began long ago. Herbert Hoover nearly carried a majority of Southern states, and the Smith-Hoover election demonstrated that the South’s affection for the Democratic Party would not survive a transformation of the party into one dominated by urban, liberal, Northern, non-white-Protestant partisans. The South began electing Democrats who voted like Republicans in the 1930s and electing actual Republican congressmen in the very early 1950s.  It went for Eisenhower twice and was split on Kennedy (who didn’t put LBJ on his ticket because he loved him dearly). And none of this would not have happened had the South not begun to develop urban centers, suburbs and become much, much wealthier in the twentieth century.

The point is, there was no sudden switch in one election that made the Northeast Democratic or the South Republican. It was the result of very long-term changes, changes that would be difficult to observe until they were almost complete; V.O. Key spoke dismissively of Southern Republicans in his classic “Southern Politics,” and Kevin Phillips wrote “The Emerging Republican Majority” almost forty years after the South had begun to realign. A Northern Republican in the mid-1920s would breathe easy knowing that the Immigration Act of 1924 had frozen immigration levels, and that Catholic voters still generally failed to vote.

The point is, those who claim they can scry the pertinent long-term demographic trends that are occurring today are fooling themselves, and you. Perhaps the youth vote and the Hispanic vote and the college educated vote will propel the Democrats to a lengthy majority never before seen in American politics. Perhaps Obama will mess up the economy and lose the youth vote, perhaps Hispanic immigration will slow down with the recession, perhaps Hispanics will be re-defined as part of “white America,” (as Americans eventually did with ethnic immigrants from the 1910s and 20s), and perhaps college education will slow down in tough economic times.

Or perhaps the most important political and demographic trends of today are simply impossible for us to observe and to sort out right now.  More likely than not the Parties will simply do what they have always done: Fight strategic battles over swing voters, and adapt to changing environments.

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.

Sean Trende

Author Archive

Follow Real Clear Politics

Latest On Twitter