Fareed Zakaria:

…tens of millions of independents, the vast middle ground where elections are won and lost in America…

Matt Bai:

Even more consequential, though, is the fast-growing swath of voters who can summon no affinity for either party. As in other aspects of modern American life, brand allegiance in politics is at an all-time low; more than a third of Americans (and more than half of all Massachusetts voters) identify themselves as independents rather than as members of the blue team or the red.

Jon Bernstein beat me to this, but he has other fish to fry with Bai. Plus, I want to yell.



How many DAMN TIMES must this be said before this MOST BASIC OF FINDINGS —first explicated at length almost 20 YEARS AGO!—sinks into the heads of pundits.

I will keep linking to this post as long as it takes. To repeat: true, honest-to-God independents are about 10% of the American population. Declining support for Obama among independents accounts for less than a fifth of Obama’s overall decline in support.


  1. Ben Bishin January 25, 2010 at 4:29 pm #

    Well, I understand the frustration, but it wasn’t first explicated at length there. The central findings appeared here much earlier:

    Petrocik, John R. (1974). An analysis of intransitivities in the index of party identification.Political Methodology 1: 31–47.

    Although admittedly this isn’t the easiest volume to find (they are not online so far as I know) its pretty widely cited given the source.

  2. Adam Berinsky January 25, 2010 at 4:52 pm #

    And they are an even smaller proportion of the voting public — about 5% by my (outdated) calculation of NES data a few years back.

  3. Jim January 25, 2010 at 8:24 pm #

    Thank you, John. Keep fighting the good fight.

  4. Joel January 25, 2010 at 8:38 pm #

    i followed the link(s) on a couple of the comments from the original post to see some of the criticisms of this argument from die-hard indie-types.

    i actually found their responses fairly compelling (one goes, in brief: independent leaners act like partisans because they have no credible alternative than to vote for one of the two parties).

    i would be interested in the response to that. is it available in the literature?

  5. Jim Gimpel January 25, 2010 at 8:56 pm #

    “Declining support for Obama among independents accounts for less than a fifth of Obama’s overall decline in support.”

    Okay, so what/who accounts for the other four-fifths?

  6. Mike January 25, 2010 at 9:39 pm #

    I would venture to guess that the term independent has been loosely applied to include those falling into the political moderate category.

  7. John Sides January 25, 2010 at 10:22 pm #

    Ben: Thanks for the cite. I can amend my yelling thusly: “…first explicated at length almost 40 YEARS AGO”!

    Joel: I read those comments. Here’s a little data. In 1992-96, independents who lean Democratic were 9 points more likely to support Perot than were weak Democrats. But independents who lean Republican were only 1 point more likely to do so than weak Republicans. So there is equivocal evidence here that “independent leaners” are more likely to select this particular independent candidate. However, I realize that this is a limited analysis.

    Perhaps a better indicator is a measure of people’s open-ended likes and dislikes of the two major parties. Because these items are open-ended, nothing is forcing independents to say anything, and certainly not anything nice. For 2004, I computed a measure of “net party affect” — [(Dem likes – Dem dislikes) – (Rep likes – Rep dislikes)]. By this measure independent leaners are almost identical to weak partisans. The means are:

    Strong Dems: 3.5
    Weak Dems: 1.8
    Inds leaning Dem: 1.7
    Inds: .5
    Inds leaningn Rep: -1.0
    Weak Reps: -1.1
    Strong Reps: -2.8

    Jim: In the post I link to above, I calculated that 50% of the decline was due to trends among Republicans and 34% due to trends among Democrats.

  8. Jim Shoch January 25, 2010 at 10:44 pm #

    What do you make of Gallup’s recent (January 6) finding that during 2009, the share of Republican leaners grew from 31% to 40%, while the share of Democratic leaners dropped from 47% to 38%?

  9. Jim Shoch January 25, 2010 at 10:50 pm #

    What I meant to ask a moment ago was:

    What do you make of the recent (January 6) Gallup finding that during 2009, the share of independents who said they “leaned Republican” grew from 31% to 40%, while those who leaned Democratic dropped from 47% to 38%?

  10. Jim Gimpel January 25, 2010 at 11:00 pm #

    Okay, so if the movement is the result of partisans, they are likely weak or leaning partisans, not strong partisans. The standard understanding is that strong partisans are those whose behavior is stable across many types of events.

    But if all the journalists are doing is blocking weak partisans into an enlarged ‘independent’ camp, then I don’t see how it makes such a great deal of difference. Everyone is just playing around with words.

    Again, according to John’s post, a huge proportion of the movement is occurring among some type of *partisan*.

    But partisans are not supposed to move around so much, are they? Obviously these fickle folks aren’t *strong* partisans, or they wouldn’t be moving so much. To most observers, in fact, they seem quite independent-like. So why not just call them ‘independents’ –because they sure don’t seem like the stable, predictable, partisans the 20-year-old literature discusses?

    Maybe weak partisans are not that much like strong partisans after all? Not if two-or-three-fifths of the fluctuation is the result of their waffling. Sounds like the journalists aren’t so stupid, but maybe we are. Maybe political science is just too wedded to its own conventional wisdom? Either that, or we are just fooling around with labels and none of this really matters.

  11. John Sides January 25, 2010 at 11:16 pm #

    Mike: You’re correct, and this is also bad practice. There are independents who are not moderate and moderates who are not independent. Party identification and ideology are not the same thing.

    Jim S.: To the extent that there is any individual-level change in party identification, it tends to be movement in and out of the independent category, typically toward the party that is “advantaged” by the economy and other structural factors. Some argue, however, that Gallup’s question wording exaggerates this movement by prefacing the question with “In politics as of this moment” (or something very similar).

    Jim G.: I don’t think anyone would claim that partisans are never supposed to change their opinion of the president. The question is, do independent leaners look and behave more like partisans or more like true independents? My brief analysis of Obama approval suggests that independent leaners behave much like weak partisans.

    Even more importantly, when it comes time to vote for a President, independent leaners are basically as loyal as weak partisans.

    This is important because journalists often portray the entire pool of “independents” (the so-called one-third) as “up for grabs” in elections. (That is Zakaria’s implication.) That’s really not the case. Independent leaners behave like predictable partisans in many respects. The pool of people that really matches what journalists mean by “independent” is much smaller.

  12. Jim Gimpel January 26, 2010 at 8:40 am #

    Political science thinks it has done the world a huge favor by defining ‘independents’ strictly and narrowly, but what we have done in return is badly muddy the concept ‘partisan.’

    Partisan now means not only the rock- solid motivated reasoners who never waver, but this large mish-mash of people who barely have any attachment, but are not *pure* enough to be called independent in the narrow sense.

    I don’t see the big advantage of that exchange, especially if these weak partisans really do mimick independents in their self-reports and behavior. And I can also see why journalists would ignore this debate.

  13. Jim Gimpel January 26, 2010 at 11:00 am #

    Bottom line:

    Just not convinced it makes much difference. You block the wafflers in with partisans, you broaden and probably weaken the concept of partisanship. Throw ’em in with independents, and our concept of independence is broadened and less clear.

    So take your choice. Political science has decided to cheapen partisanship.

  14. Matt Jarvis January 26, 2010 at 11:37 am #

    I’m not sure political science made the choice. Down the line, if you compare all 7 points on the scale as if it were a true 7-point scale, the behaviors really fall into 5 types (multiply the partisans by the 2 parties): Strong partisans, nonpartisans, and the conglomeration of the weak and the leaners. Weak partisans and leaners behave very similarly, from turnout to vote choice to affect and so on. And, as John mentions, most of the partisan “movement” occurs within those groups (weak->lean, lean->weak) with some also occuring at the indie->lean level and vice versa. All of which suggests that, for these weak/lean types, that they are kinda answering the first branching question a la Zaller. Of course, the indies and the strong partisans aren’t, but these folks are kinda sampling their answer to that first question.

    Now, I’m less familiar with the folks who’ve looked at what happens when you actually present the respondent with the 7 point scale instead of the branching questions. Can anyone shed light on this segment of the literature?

  15. Marc January 26, 2010 at 3:50 pm #

    Thank you Jim G for such a convincing argument.

    Perhaps we need some research to determine whether extreme partisans are more likely to see the world as strongly partisan.

  16. Three Oranges January 27, 2010 at 11:31 am #

    Citing a 20-year-old book, written before Perot even, to describe the current political scene is a bit ridiculous, don’t you think?

  17. John Sides January 27, 2010 at 11:48 am #

    Three Oranges: If you’d read carefully, you’d see that my data (and my comments in this thread) speak to events through the 2008 election, and, with regard to Obama approval, through 2009.

  18. Navin Johnson January 27, 2010 at 11:56 am #

    Good thing the makeup of the Rs and Ds hasn’t changed over the last 20 years, otherwise your assertion would be meaningless. Oh, um, wait

  19. John Sides January 27, 2010 at 12:17 pm #

    Navin: I don’t know which assertion you’re referring to.


  1. Third parties don’t work « binckesblog - August 4, 2011

    […] “squishy middle,” composed of not-really-but-it-sounds-nice independent voters, doesn’t exist. To believe that they exist defies human nature. We all create our own […]