Family Feud: Democratic Activists v. Democratic Voters - The Ideological Gulf that Thwarts a Sustained Majority

Family Feud: Democratic Activists v. Democratic Voters - The Ideological Gulf that Thwarts a Sustained Majority

Family Feud: Democratic Activists v. Democratic Voters - The Ideological Gulf that Thwarts a Sustained Majority

This report unravels a defining mystery of the modern political era. Since 1972, Democrats have held a lead in the number of Americans who identify with the party, but that hasn’t translated into sustained Congressional and White House dominance. In this report, we explain this quandary and its serious political implications, based on four findings:

  1. Since 1972, more voters have consistently identified as Democrats or Democratic leaners than Republicans.
  2. But the ideological divide between Democratic voters and activists has been far larger than the GOP’s.
  3. This ideological gulf coincides with less party loyalty from Democratic coalition voters.
  4. Democratic leaning Independents are a growing part of the coalition and cannot be counted on to be reliable Democratic voters.

These findings have significant electoral consequences for Democrats. Odds favor a re-emergent Democratic majority, but only if liberal party activists will cede control of the agenda and allow the party to move in the direction of its moderate, non-activist voters.

Finding #1

Since 1972, more voters have consistently identified as Democrats or Democratic leaners than Republicans.

In modern political history, since regular exit polls began in the 1970s, Democrats have held a non-stop lead in identification among the American electorate. According to data from the American National Election Studies (ANES)—which conducts national surveys of the American electorate over time for use in academic study—that lead has fluctuated but remained significant and steady across many decades

Figure 1: American Electorate by Party Preference

ANES asks a two-part question to identify party preference. First, they ask whether a voter identifies as a Democrat, Republican, or Independent. If the voter chooses one of the two parties, ANES will then ask whether the voter identifies strongly or weakly with that party. But if the voter identifies as Independent, ANES will ask whether that person tends to lean towards one party or the other. Thus, they identify seven categories of voters: strong Democrats, weak Democrats, Independents who lean towards the Democratic Party, pure Independents, Independents who lean towards the Republican Party, weak Republicans, and strong Republicans.1

Looking at the partisan advantage in a recent election, for example, in 2008 Democrats led Republicans by 14 percentage points when voters who said they leaned toward one party or the other (referred to as leaners hereafter) were included in the calculation. Even in years when Republicans won the presidency, including the Reagan landslide of 1984, Democrats enjoyed a voter preference advantage of nine points, with 48% of the electorate identifying as Democratic or leaning Democratic, compared to 39% identifying as Republican. Democrats enjoyed similar advantages during midterm elections even when they lost control of Congress or failed to retake control.

Finding #2

But the ideological divide between Democratic voters and activists has been far larger than the GOP’s.

So at first blush, Democrats seem to be in an enviable position, but the picture complicates from there. Part of the reason for Democrats’ electoral problem comes from an ideological disconnect between those who are activists and agenda setters for the party versus those who merely identify themselves as Democrats or Democratic leaners. In the 10 presidential election cycles dating back to 1972, Democratic activists (defined as those who attended a meeting or rally and donated to a campaign) rated themselves at an average of 3.06 on the 7-point ideological scale. Democratic non-activists came in at 3.77—indicating a 0.71 gulf between the active and non-active wings of the Party (with 1 representing extremely liberal, 7 extremely conservative, and a score of 4 representing moderate).

Republicans have a much smaller and less sustained gulf. Between 1972 and 2008, Republican activists rated themselves at 5.22 on the 7-point ideological scale, with non-activists very close at 4.89 (a .33 difference). In fact, the ideological gulf between Democratic activists and Democratic non-activists is more than twice that of their Republican counterparts dating back to 1972. Thus, Democratic activists are blue; the Democratic base is purple, and Republicans of all stripes are red

Table 1: Mean Ideological Scores for Activists and Non-Activists in Each PartySource: ANES Guide to Public Opinion and Electoral Behavior. * Denotes a statistically significant difference between party activists and non-activists. Party activists were defined as those who attended a campaign meeting or rally AND contributed money to a candidate or campaign. In 2000, though the differences between Democratic and Republican activists and non-activists were significant, the sample of Democratic and Republican party activists were extremely small—roughly one-third the size of the average sample size across the ten elections reviewed for Democratic party activists and one-fifth for Republican party activists.

Democrats tended to lose during years the gap between their activists and non-activists are larger—1972, 1980, 1984, and 2004. And for every presidential election between 1972 and 2008, there was a significant difference between Democratic party activists and non-activists. Democrats won only four of ten elections during that time. In two of those elections, 1992 and 1996, there were special circumstances—a 3rd party candidate that allowed Democrats to win with less than 50% of the vote. Among Republicans, there was a significant difference between Republican party activists and non-activists in only five election years—including two of the years when Democrats won the presidency and in 2000 when Democrats won the popular vote.

Figure 2 illustrates this problem in another stark way—a very tiny portion of the entire electorate identifies themselves as liberal or extremely liberal, while the vast majority describe themselves as moderate. Republicans have a larger ideological base from which to start (conservative and extremely conservative), although moderates are by far the largest chunk of American voters.

Figure 2: Ideological Divisions among General Electorate

This division has changed over the past few decades. As shown in Figure 3, the ideological distribution of non-activist Democrats and Republicans in 1972 was actually quite similar. Both parties were dominated by self-identified moderates—the Democratic distribution skewed slightly left and the Republican right. On a 7-point ideology identification scale, the mean score for non-activist Democrats was 3.88 compared to 4.59 for Republicans—a statistically significant difference.2 By 1996, the Republican electorate had shifted significantly to the right (mean value of 5.13) and Democrats had shifted just slightly to the left (mean value of 3.69). The distribution observed in 2008 was very similar to that of 1996 for both parties. By 1996, a clear divergence between the two party coalitions was evident, but it was driven almost entirely by the Republican Party’s move to the right—not a movement by the plurality of Democratic voters to the left.

Figure 3: Ideological Distribution of Democrats and Republicans

This is element one of the polarization story: party polarization among the broader, non-activist electorate has been driven by Republican voters shifting right, while the Democratic voting coalition essentially stood still. But there’s more to the story, and it’s told in Panel B of the figure. In 1972, Democratic Party activists differed significantly from non-activists, yet moderates were a sizable component of each parties’ activist base. Interestingly, on the Republican side, there was no significant difference between the mean ideological distribution for activists and non-activists in 1972. Among Democrats, however, the mean value of 2.89 for party activists was already significantly to the left of non-activists (3.88). This pattern held through 2008.

Though the Republican Party’s activist base has become more conservative over time, so has the party’s non-activist membership. In 2008, the mean score on the ideological scale for Republican activists was 5.50—not vastly different from the 5.16 value for non-activists. Yet among Democrats, activists sported a mean value of 2.85 as compared to 3.64 for non-activist—a significant difference. This means that since 1972, Republican Party members—activists and non-activists alike—have become more conservative, suggesting little disconnect between most voters who pull the lever for the party and its most committed members. Among Democrats, however, there remains a sizeable and significant divide between more liberal activist members and more moderate Democratic coalition voters.

That is element two of the polarization story. Republican activists and voters moved right together. Democratic activists have stayed left while their voting coalition remained more moderate. So there is clear polarization between parties, but also polarization within the Democratic Party. These changes in the party’s coalition offer further explanation for the Democratic Party’s electoral difficulties. The Democratic Party’s electoral coalition has shifted only slightly to the left but remains well anchored around a core group of moderates. Party activists, however, have remained decidedly left of center.

Since a substantial share of the Democratic Party’s coalition finds itself ideologically situated between the extremes of partisan activists on the left and the right, a Democratic Party agenda tailored to liberal party activists is more likely to alienate a much broader segment of the Democratic coalition than would a Republican Party agenda tailored to conservative activists.

In short, one reason Democrats lose is likely because the folks who set the agenda for the party are more out of step with most of party voters than are the folks who set the agenda for the Republican Party. For Republicans, there is strength and ideological cohesion on the right. Republicans win because there is little difference between party activists and voters. Among Democrats, however, strength comes not from the left but from the center—that’s where the party’s core coalition of voters is, and when the party strays left, many of those voters defect.

Finding #3

This ideological gulf coincides with less party loyalty from Democratic coalition voters.

The ideological gulf between Democratic voters and Democratic Party activists is a likely explanation for the higher likelihood of defections among Democrats. A review of data from an American National Elections Studies panel survey of the same folks in 2000, 2002, and 2004 demonstrates this predicament.

Though Democrats enjoy an overall party identification advantage over Republicans, Democrats are especially vulnerable to defections at the ballot box—especially among Independents who lean Democratic. Since 1970, leaners have increased from 18.5% of the Democratic Party coalition to fully one-third in 2008 (a nearly 15 point gain). Among Republicans, such leaners have increased from 25% in 1972 to 31.5% in 2008 (a gain of 6.5 points). Perhaps of greater concern to Democrats, during that same time there has been no increase in the share of strong Democrats while the share of strong Republicans grew by 6 percentage points—equal to the growth among Republican leaners. So even though Democrats have a larger coalition, it is a coalition increasingly reliant on Independents who lean in that direction, rather than self-identified Democrats.

In Table 2, data from the ANES panel study shows Democrats of all partisan strengths in 2000 were more likely to have left the Democratic Party by 2002 or 2004 than were their Republican counterparts, but the tendency to switch parties was far more pronounced among Independents who leaned Democratic. It is immediately clear that weak Democrats and Independents who lean towards the Democratic Party are not the same with regard to party loyalty over successive election.Table 2: Party Switching by Strength of Partisanship

Party switching is only one indication of party loyalty—voting is a far more telling indicator. Using the same panel survey to explore the voting patterns of Democrats and Republicans across three Congressional elections shows that Democrats, especially Independents who lean Democratic, are less loyal to party over time—sometimes dramatically so.

Finding #4

Democratic leaning Independents are a growing part of the coalition and can’t be counted on to be reliable Democratic voters.

Much political science literature and considerable political commentary tends to equate those who identify as weak Democrats or Republicans and those who identify as Independents who lean in a party’s direction,3 but recent research casts doubt on common perceptions about similarities between those two groups.4 Much of the past research on Independent voters considered partisanship in a single election and rarely followed the same voters across multiple elections. When voters are followed across elections, the focus tends to be on presidential elections and often only two sequential elections are considered. With few exceptions, most elections since 1968 have featured an incumbent president seeking re-election or a vice president seeking a promotion. As such, candidate preference or loyalty may be confounding party preference or loyalty.

But Table 3 below shows that strength of partisanship matters when we follow voters over multiple elections. Weak Democrats and Democratic leaners had roughly the same party loyalty as Republicans in the 2000 election—confirming that weak Democrats and Independents who lean Democratic behave much the same in a given election year—but the larger question really pertains to the size and stability of a governing coalition over time.

Table 3: Relation of Strength of Party Identification to Partisan Regularity in Voting for the House of Representatives (2000, 2002, and 2004)

Source: ANES Guide to Public Opinion and Electoral Behavior.Partisan vote choice was determined by calculating only the two party vote shares for each election. Respondents who indicated that they had not voted or did not indicate for whom they voted were excluded.

Voters who identified as weak Democrats and Democratic leaners in 2000 were more likely to defect and vote Republican in 2002 or 2004, but Independents who leaned Democratic were significantly more likely to do so, with 46% voting Republican in the 2002 midterm and nearly 40% voting Republican in the 2004 election. Given the propensity of leaners to switch their party affiliation (Table 2), many of the Democratic leaners from 2000 who voted Republican in 2002 or 2004 were likely no longer identified as Democrats in those elections. So anyone studying partisan attachment or loyalty in any one of those elections would have missed the larger picture—a picture of diminished loyalty and significant differences between weak party-affiliated voters and those who identify as Independent leaning toward one party.

Unfortunately for the Democratic Party, Independents who lean Democratic are the fastest growing segment of Democratic voters, consisting of one-third of party identifiers. And as the data above demonstrates, these folks are less attached to the party over time and less likely to vote for the party over successive Congressional elections. This may provide another explanation for the Democrat’s inability to translate their partisan identification advantage into consistent electoral victory.

Although many pundits will try to argue that Independent leaners are just partisans in disguise, we have seen above that they are significantly more likely to switch parties or vote for candidates from different parties when studied over time. And Independents who lean towards one party also view the world in a distinct way. A review of the attitudes and preferences of the attitudes of leaners compared with weak and strong partisans across eight attitudinal measures regularly included in the ANES survey since 1972 reveals leaners do differ from their weak and strong partisan counterparts in important ways. For example, Independents who lean Democratic, as compared to their weak and strong Democratic counterparts, are less supportive of government intervention in the economy, more likely to believe that the government has gotten too involved in things people should do for themselves, and express higher levels of support for cutting Social Security spending. With regard to parties and election outcomes, leaners are less likely to see major differences between the two major parties and were less concerned with which party won the 2008 presidential election.

These policy differences do not support the argument sometimes made by political pollsters that leaners are indistinguishable from weak Democrats and Republicans. Independent leaners often held positions more similar to pure Independents than to their weak partisan counterparts or occupied a midpoint between weak partisans and pure Independents. A comparison of the positions of these leaners in 2008 with their weak partisan counterparts to similarly identified partisans in 1992 shows that on nearly every measure reviewed, the gap between Independent leaners and weak partisans has widened during the 16-year span—a span supposedly was marked by partisan retrenchment. Among Democrats, the gap between Democratic leaners and weak Democrats has increased for every measure considered.5


There may be more money and passion among activists on the left, but there aren’t enough voters there to secure consistent electoral victory for Democrats. The true wealth of voters in the Democratic coalition resides in the vital political center and that’s where the Democratic Party will find the path to sustained electoral dominance.

But Republicans should not look to this report as good news for the GOP. At present, Republicans are able to win because so many Democratic voters occupy that middle ground between the extremes of the activist elements of the two parties. But the GOP has been moving right faster than the overall electorate and in the opposite direction of even non-activist Democrats. If the GOP continues to trek to the right, they will reach a point where moderate Democrats no longer view the GOP as an acceptable alternative in elections.6

Those voters in the center may decide to support their own party or just stay home—either way the GOP would be left with a smaller coalition of voters. Perhaps more dangerous for the GOP, should the Democratic Party choose to moderate its agenda and actively seek to close the gap between activists and non-activists, then the party could likely unify its coalition and consistently attract the broad middle of America’s electorate.7 Should that happen, the GOP could be sunk and return to the near minor party status it occupied between 1932 and 1964.

The data on party coalition ideology suggests that Democrats can move to the center and win, but if Republicans move to the center, they risk alienating a substantial portion of their electoral coalition. So long as America remains a moderate nation, odds favor a re-emergent Democratic majority—but only if Democrats actively work to make it happen. The real question for Democrats is whether liberal party activists will cede control of the agenda and allow the party to move in the direction of its moderate, non-activist voters.

  • Public Opinion185


  1. Excludes respondents who answered “Don’t know or haven’t thought much about this.”

  2. American National Election Studies (ANES) Guide to Public Opinion and Electoral Behavior, University of Michigan, Center for Political Studies. Accessed January 2012. Available at:

  3. See, for instance: Alan I. Abramowitz, “Setting the Record Straight: Correcting Myths About Independent Voters,” Larry J. Sabato’s Crystal Ball, July 7, 2011. Accessed November 5, 2011. Available at:; See also Mark S. Mellman, “Myth of the Independent Voter,” The Hill, November 10, 2011. Accessed February 7, 2012. Available at:

  4. See, for instance: Todd Eberly, “The ‘Myth’ of the Myth of the Independent Voter,” The FreeStater Blog, January 13, 2012. Available at:

  5. Drew Kurlowski, Disappearing Intransitivities in the Party Identification Scale, Presented at the 2011 Midwestern Political Science Association Conference, Chicago, IL.

  6. To be certain, if one were to consider every question included in the ANES survey, there would be areas of convergence and divergence among weak and Independent leaners, but these questions were selected as they speak to core ideological differences between the parties.

  7. Look no farther than the recent ballot initiative in Ohio in which voters overwhelmingly rejected a Republican-backed law to curtail the collective bargaining rights of public employees. Of course, in an equal warning shot to Democrats, those same voters endorsed a measure that would ban state officials from enforcing the individual mandate central to President Obama’s Patient Protection and Affordable Care Act.