Tuesday, February 27, 2018

Generational Partisanship (or Why Millennials May Not Save Us from Partisan Loyalty and Polarization)

As many may have realized by now (by reading some of my previous posts), I've become very interested in the generational dynamics underway in American politics, and particularly in North Carolina (through voter registration figures and analyses). But voter registration only tells part of the story--a companion aspect to our political environment is voting "behavior," especially about voters' party self-identification and supporting presidential candidates, when it can be a driving force of our partisanship and polarization.

In trying to better understand the partisanship and potential polarization coming from party identification in our nation, along with the generational dynamics that are happening (especially with the rising Millennial generation), I ran a series of analyses using survey data from the American National Elections Studies (ANES) to investigate the following questions:
  • Are Americans really moving into political independence with the "rise of the independent" or are we seeing another phenomenon, especially when it comes to voting behavior of these independents?
  • What have been the trends in recent presidential elections when it comes to partisan self-identification and presidential vote choices among the four generations in the American electorate?

First, some definitions of terms being used, starting with generational cohorts. In defining the four key generation cohorts in this piece, I used the corresponding birth years (and thus, the ages in each election year) to define members of the following generations:

  • Greatest/Silent Generation: those born before 1945
  • Baby Boomers: those born between 1945 and 1964
  • Generation X: those born between 1965 and 1980
  • Millennials: those born after 1981 (some are now saying this generation ends in 1998, with Generation Z starting to come of age and born after 1998; for this study, I simply use Millennials to capture those born after 1981 in the analyses)
In order to capture the generational analyses, I recoded the ANES survey respondents, based on their stated age in the year of the election, into one of these four generational cohorts. As Millennials were just coming of "political age" to participate in national elections in 2004, I use that year's ANES data and the subsequent presidential election years, focusing heavily on the three most recent elections.

First, to give a sense of the respondents and the generational cohorts in the four ANES studies, the respondents in each year's study are classified into the cohorts:

In 2004, Millennials were just beginning to hit double-digits for respondents within the survey, and thus their impact on the national electorate, with the Greatest/Silent generation a quarter of the respondents. Twelve years later, in 2016, Millennials were nearly a third of the respondents, while the oldest generation cohort slipped into single digits of the respondent pool.

Within the ANES surveys, a battery of questions are asked of respondents before the November general election and immediately after the election.  One of the pre-election questions posed is "Does the respondent think of self as a Democrat, a Republican, an Independent, or what?" This simple question captured an interesting phenomenon among the generational cohorts, especially the entering Millennial generation. The following charts show the individual generations and their initial party self-identification, from 2000 to 2016 (Millennials from 2004):

Among Millennials, political identified "independents" were a plurality of the respondents in 2004, 2008, and 2016, while self-identified Democrats were a plurality in 2012; Republican identification, however, has been less than a quarter of this generational respondents. Generation X sees a back-and-forth among self-identified Democrats and independents since 2004, with Republicans fluctuating in the twenties. Baby Boomers have seen pluralities among self-identified Democrats, but Republicans were a slim plurality in 2004; self-identified independent Baby Boomers have bounced around between second and third rankings, unlike the younger generations, where independents were consistently first or second in identification. Among the oldest generation, Greatest & Silent, Democratic self-identification held sway in 2000 to 2008, but Republicans tied and secured the first ranking in 2012 and 2016, respectively.

The next question in the ANES sought to find out the "strength" of the respondent's partisan self-identification: "strong" to "not very strong" among the respondents who identified as partisan, while the independents were asked if they leaned to one party over the other. This creates a seven-point party self-identification. The generation breakdowns in the recent presidential elections were:

Among Millennials in 2004, numbers of self-identified independents leaned to the Democratic Party, but over the next three elections, partisan identifiers became much more evident within the generation, with pure independents below 20 percent of the generation. This could demonstrate that growing numbers of Millennials were beginning to find their partisan "home." What is interesting is that Baby Boomers and those of the Greatest generations became "balanced" in strong partisan identification within both generations, while younger voters (Millennials and Gen Xers) tend to have a more Democratic-lean towards their cohorts in the 2016 study.

Next, one of the key questions within the post-election studies is "if the respondent voted in the election, who did you vote for?" In taking all the responses for 2008 to 2016 by party self-identification and strength, one sees a very clear pattern in presidential vote choices (the numbers below the categories reflect the partisan self-identification within the electorates of that year):

In the most recent presidential elections, those who identify as partisans have been "loyal" to their party's presidential candidates: not surprisingly, those who identify as "strong" in their partisan identification report voting for their party's presidential candidate at levels of 95 to 99 percent of the time in the past three election cycles. Those who are partisan, but "not very strong," in their identification range from 75 percent to 88 percent in support of their party's presidential nominee.

But what may be surprising to some, those who identify as "leaners" in their identification voted at slightly higher levels than their "not very strong" partisans: 78 to 89 percent voted for their party's presidential candidate. Only those who identify as "purely" independent (with no leanings to one party over the other) tend to be more in line with a "competitive" level of presidential vote choices.

This phenomenon has been apparent in American National Election Studies for some time, but since 1992, the pattern has clearly emerged in partisan loyalty, even among independent leaners. The next chart shows partisan self-identification (by strength) from 1952 to 2016, and the presidential vote choice for their party's candidate (independents are shown as voting for the Democratic presidential candidate):

While the above graph may look like a plate of colored spaghetti, a clear division appears to develop in 1992, so the next two charts show 1952-1988 and 1992-2016:

From 1952 to 1988, partisan loyalty, while strong at times, was much more dispersed; from 1992 to 2016, the clear strength of loyalty among strong partisans, with the second grouping of loyalty among "not very strong" and "leaning" partisans, shows a much tighter grouping.

What I would take away is that independent "leaners" aren't really all that independent in their vote choices: that these leaners demonstrate as much partisan loyalty as their fellow partisans. Therefore, I regrouped all respondents into three collapsed partisan identification groups: Democrats and Republicans, to include strong, not very strong, and lean to identifiers, in addition to a separate category of 'pure' independents. Now, in taking the past three presidential elections, we can see the collapsed partisan identification within the generations (again, the numbers below the generation cohorts represent the percentage of the electorate/respondents in that election):

So what we see is a pattern of partisan loyalty to the party's candidate that has emerged in the American electorate, and that party identification may be better defined by using partisan loyalty. But has this partisan loyalty extended within the generations as well?  The following charts show the collapsed partisan identification and the vote choices for that party's presidential candidate by generations for the 2008 to 2016 elections:

During the past three presidential elections, Millennials have shown an 80 to 98 percent level of party loyalty, with Gen Xers showing an 82 to 90 percent loyalty, Baby Boomers with an 88 to 92 percent loyalty, and the Greatest generation showing a loyalty of 86 to 96 percent. In the 2016 presidential election, I noted the levels of "party defections to a third party" candidate: Millennials tended to have the highest defectors (8 percent of Democrats and 11 percent of Republicans voted for a third party) while only 3 and 4 percent of Democratic and Republican Baby Boomers defected to a third party candidate.

So, if one takes the notion of party loyalty to demonstrate an example of the power of polarization (Democrats giving strong loyalty to their party's presidential candidates, while Republicans doing the same, and only a small number of "pure" independents giving a competitive level of voting among the two parties), then the era of party loyalty and polarization is only continuing with the newest generation entering the electorate. While some would put the great hope on the Millennial generation to bring about an independent revolt against the partisan loyalty of previous generations, this data tends to point that Millennials will be as partisan, and seemingly as polarized, as other generations. What seems to be evident with Millennials, and with Generation X to some degree, is that Democrats may see greater numbers and loyalty among these voters than Baby Boomers and the Greatest generation.

In a future post, I'll examine the notion of polarization through various policy issues, again through the lenses of partisanship and generations.