Trump Won, but Nothing Re-aligned

Donald Trump shocked the political world by winning the 2016 presidential election. There has been much talk of his victories in several Blue states as well as his improved margins among white working class voters. But did Trump really "change the map?" Has he found a new coalition? There is a straightforward way to test this.

Gerald Pomper (1967) first examined continuity and change in presidential elections via linear correlation of state-by-state election results in paired presidential elections.  Pomper was interested in identifying realignments. According to Pomper, realignment would be evident via a “change in the parties’ bases of support… the geographic distribution of each party’s vote would be different from the past: traditional strongholds would fall, while new areas of strength would become evident” (1967, p. 539). Such changes or shifts in support would be evident in statistical analyses of the correlation between sequential elections. In short, a break in continuity, as demonstrated by a low level of correlation with preceding elections, suggests realignment.  Pomper (1967, p. 540) compared the Democratic party’s share of the total vote by state from 1824 to 1964  and identified five “electoral cleavages” in American history, cleavages that correspond to prevailing scholarship on American party systems.  

Like Pomper, I paired successive elections by comparing the Democratic share of the total vote by state to generate correlation coefficients. The state-by-state comparison generates a Pearson correlation coefficient with a low or negative value indicating a disruption or cleavage point and a higher positive value indicating continuity or electoral stability.  Figure 1 presents the results. The changes taking place in the parties' coalitions in the late 1970s and early 1980s are evident as is the rather stable electoral period that emerged in 1996 - a period marked more narrow election victories and the emergence of the familiar Red and Blue states.

Donald Trump pieced together an impressive electoral coalition in the 2016 election, a coalition that included traditionally Democratic states like PA, MI, and WI - yet Figure 1 shows the 2016 election demonstrated a high degree of correlation with the election of 2012 – no fundamental changes took place in 2016. Rather Trump was able to achieve narrow victories in otherwise Democratic states. The results suggest that the present path to 270 electoral votes remains both stable and competitive.

Pomper. 1967. Classification of Presidential Elections. The Journal of Politics.

American Politics: Electoral Realignment or a Model of Stability?

As part of a larger research project, I've been working on updates to Gerald Pomper's 1968 study of electoral coalitions by studying the linear correlation of state by state election results in paired presidential elections. According to Pomper an electoral conversion or realignment would be evident via a “change in the parties’ bases of support… the geographic distribution of each party’s vote would be different from the past: traditional strongholds would fall, while new areas of strength would become evident.” Such changes or shifts in support would be evident in statistical analyses of the correlation between sequential elections. High levels of correlation would indicate stability, and low levels change.

Though some have openly wondered whether the 2008 election of Barack Obama represented a political realignment, I contend that the political realignment of the late 1960s has been maintained, and is in fact incredibly stable.

Using election data for all states and DC since 1928 I have calculated linear correlations for all successive elections in the last 8 decades.  As shown in the first graph below, the correlations reveal the very stable New Deal realignment as well as it's collapse in the 1960s and the emergence of a new and stable electoral pattern - a realignment. The only disruption of that new stability occurred in 1976 with the election of Jimmy Carter, likely a reflection of fallout from Richard Nixon and Carter's strength in the south, in the midst of the southern realignment toward the Republicans.


What is unmistakable is that 2008 fit comfortably into the electoral pattern that emerged at the end of the 1960s - 2008 was very much a normal election, its correlation with 2004 and the average of 1992-2004 were both quite high. This suggests that no realignment, defined by James Campbell as "a durable and substantial shift in the parties’ national electoral balance of power,” took place in 2008.

I wanted to be certain that I was not missing more subtle evidence of realignment by focusing too much on national trends, so I next divided the states into three groups - Reliably Democratic states, Reliably Republican, and Swing states. Reliably Republican or Democratic states were defined as having voted with their respective party's candidate in at least four of the last five elections, through 2004. The designations were based on an analysis by the Population Reference Bureau conducted prior to the 2008 election.


This resulted in 39 states (including DC) that were not considered to be competitive. The remaining 12 states were coded as swings.

As shown in the following graph, I found a high degree on stability in the Democratic states (light green line)through 2008, lesser but still high stability in the Republican states (red line) and in the swing states (blue line) I found high levels of correlation from one election to the next only when a president was re-elected, or succeeded in office by a member of the same party. Conversely, elections that witnessed the defeat of an incumbent or a shift in party control of the White House was marked by lower correlations - until 2008. Swings state results were highly correlated with 2004.





The swing state finding for 2008 was intriguing given that George W. Bush won 11 of the 12 swing states in 2004 and Barack Obama only 5 in 2008.  I suspect the high degree of correlation reflects the closeness of the results across those swing states in both elections - even though the partisan outcome shifted in nearly half.


It is also possible that the high degree of correlation indicates some changes taking place in swing states. Six swing states (AR, KY, LA, MO, TN and WV) have not gone Democratic since 1996, suggesting that they may no longer be swing states.


I recoded those states as reliably Republican and it reduced the correlation coefficient for 2008 among swing states from the high .845 to a much lower .416 (no graph). The Republican correlation coefficient in 2008 declined slightly from .917 to .795.


The results appear to suggest that the Southern realignment that began in the late 1960s is complete as 4 of the 6 swing states recoded as Republican are southern states. In all, it suggests that the battlefield of swing states may be decreasing.


It's too soon to tell whether Indiana, North Carolina, or Virginia need to be removed from the reliably Republican column and added to the swing state group.


A recent blog post by Professor Andrew Gelman noted that there has been a decline in state by state vote swings, noting "the red-blue map is much more stable from election to election than it used to be." My analysis suggests that much of this stability results from the culmination of the southern realignment, the Democratic electoral coalition has been very stable since 1968 and the hiccups in the Republican coalition have involved southern states in elections with southern Democratic candidates. In recent elections, southern states once considered swing states (because they still voted for southern Democratic presidential candidates) seemed to have completed their journey to the Republican column - witness the failure of southern Democrat Al Gore to carry a single southern state in 2000.


The overall results from my analyses thus far do support my contention that a relatively stable electoral pattern emerged in the late 1960s and there is little evidence of any break in that stability. America remains in the post New Deal electoral system. A highly competitive system in which Republicans and Democrats are near parity and neither party dominates the political scene. Though Republicans have won 7 of the 11 Presidential contests between 1968 and 2008, Democrats have have enjoyed majority control in the House of Representatives in 15 of 22 Congresses between 1968 and 2010. And unified party control of the House and the White House has only existed for 5 of 22 Congresses.


Everett Carll Ladd labeled the post New Deal era the "No Majority Realignment." It's a fitting and effective description of an era that emerged in the late 1960s, and it appears, remained intact through 2010.

Is it Really All about Trust or is the Party Over?

In Why Trust Matters: Declining Political Trust and the Demise of American Liberalism, Marc Hetherington argues that a decline in  political trust has undermined progressive policy enactment in America since the 1960s. He writes that America has not become more conservative, rather that Republicans have been able to exploit declining levels of trust to gain power.

In chapter 8, under the heading What the Left Should Learn he argues that liberals can reenergize public trust and reinvigorate progressive policy making if they only stopped the reflexive criticisms of the system and started to talk positively about government and the effect of policy. From page 146 and 147 "If progressives provide and alternative vision of government as one that takes care of older Americans, protects the environment, builds highways, and the like, Americans will trust that version of government... Recent Democrats have been too frightened to tell this story."

I cannot help but think that Democrats have been engaged in that exercise since 2006 and certainly since 2008. Democratic leaders like Speaker Pelosi and President Obama have an incredibly positive view of what government can do. President Obama spent his campaign and first 18 months in office trying to change public perception about government - yet the result thus far has been the opposite of what Hetherington predicted. Trust in government has fallen further, expressions of conservativism have risen, preferences for less government and fewer services are at recorded highs.

It seems that we are currently putting Hetherington's theories in Why Trust Matters to the test and thus far they are falling short.

The caveat may be that Democrats attained power by attacking government as helmed by Republicans - so they had to undermine confidence in Republicans to achieve victory, but may well have undermined confidence in government as well. This is a challenge for liberals, how do you simultaneously try to bolster confidence in government while criticizing government as led by Republicans. Can one expect the public to discern the difference between criticizing Republican governance specifically as opposed to American government in general?

This is of course the reality of governing and campaigning in an era of weak party attachment among the mass public. Since the 1960s the electorate has become much less attached to either political party. Data from the American National Election Study show that there has been a marked increase in the share of the electorate that is either Independent or Leaning Independent.

At the same time, the share of the electorate identifying as Strong Partisans has fallen from its highs in the 1950s and early1960s.

Why does the decline in partisanship matter? Because it effects how parties must campaign for office. In an era when partisanship was strong and very few Americans considered themselves to be Independent candidates for office could campaign based on voters positive attachment to their party. Voters were being asked to vote for the Democratic party, for the party's agenda. As relatively few Americans identified as Independent such partisan appeals worked. In the present era, however, neither the Democrats nor Republicans have a sufficient foundation of electoral support to campaign merely based on party attachment. Instead, they must campaign on a platform of rejection of the opposing party and its administration of government. To accomplish this it becomes necessary to undermine public confidence in opposition's governance and this typically, inevitably undermines public confidence in government.

Evidence of these efforts were seen in the Democratic criticisms of President Bush’s handling of the War in Iraq, in calls for a 9/11 Commission, and in response to the government’s response to hurricane Katrina. More recently, Republican criticisms of President Obama’s handling of the gulf oil spill serve the purpose of undermining confidence. Though these efforts at undermining confidence in the other party may offer short-term political gain, they also serve further to erode public confidence in government. Solutions to big problems, such as budget deficits, often seem too risky for either side in this “confidence game” to pursue.

Hetherington treated trust as an independent variable that undermined progressive policy making - in reality trust is a dependent variable resulting from a dramatic decline in mass partisan attachment since the 1960s. Trust is unlikely to rebound unless the American party system rebounds - at present there is little indication that such a rebound is likely to take place in the near future. Though Hetherington, in subsequent research, and other scholars like Alan Abramowitz (see his just released The Disappearing Center) argue that the public has become more partisan in recent years and more attached to party what these scholars have actually found is increased partisanship and polarization among political elites and they have mistakenly attributed that polarization to the mass electorate.

Rather more than one-third of the electorate operates as non-partisan free agents. So each party seeks immediate strategic advantage, but neither can undertake substantial policy reforms as they lack mass-based support. We entered an era that political scientist Walter Dean Burnham defines as the “politics of collision, coalition, and the permanent campaign.” This has created a system of government incapable of addressing the significant challenges before us, which then reinforces distrust and likely further undermines partisan attachment.

Breaking Waves of Dealignment

Realignment: the coming to power of a new coalition, replacing an old dominant coalition of the other party (or replacing a stalemate, as in the United States in 1896 or 1932). The concept of realignment or a critical election was first put to paper by political scientist V. O. Key in a 1955 article titled "A Theory of Critical Elections." According to Key and subsequent realignment adherents political parties, voter loyalty, and policymaking routinely shift in swift, dramatic sweeps - or critical elections. Realignment literally means that voters, en masse, switch allegiances from one party to another. In contrast, a dealignment is said to occur when voters abandon party loyalty to become independents or nonvoters.

Many believed that the one-two punch of the 2006 mid-terms and the 2008 presidential election was heralding a realignment in America with voters moving to the Democratic party. Indeed, the party reclaimed Congress and the White House and enjoyed a clear advantage in voter expressed party preference for the first time in 20 years. It appears that such realignment predictions were made in haste. Democrats no longer enjoy an advantage in voter preference and according to the latest NBC News/WSJ Poll only 38 percent of voters said their representative should be re-elected, while nearly half (49 percent) believe it’s time to give a new person a chance. According to Politico “That’s the lowest net re-elect number for Congress since November 2005 – and even worse than the polls taken right before the landslide election of 2006 that swept Democrats in control of Congress (39 re-elect/45 new person), and worse than those taken before the Republican revolution of 1994 (39/49).”

Translation? 2010 is shaping up to be another wave election – a term that non-political scientists and media commentators often use to describe a critical mid-term election. This presents a problem – critical elections were thought to usher in generational shifts in loyalty. But we have now had two wave elections in 12 years – 1994 and 2006 and may have a third on the way - only 4 short years after the last wave. All of this suggests that we have in fact dealigned in America. Voters are no longer loyal to any party, as such they willingly and easily switch allegiances from one election to the next. Writing for Politico back in August, Eamon Javers referred to this new generation of voters as “Fickle Kids” who change voting preference as often as they do cell phones. If this is true, if we have dealigned, if voters no longer have a true connection to either party, America is likely to enter a very unstable era in which party control of government will be very unpredictable and likely short-lived. It remains to be seen how the parties will react to this change.