Pew's Political Typologies

Jeff Alworth

The Pew Research Center published the results of an ambitious study this week that seeks to go beyond the blunt definitions of red and blue America. Pew has identified nine political "typologies"--three groups within liberal and conservative camps, and another three in the middle.  This is the fourth in a series of studies since 1987 to type voters into "homogeneous groups based on values, political beliefs, and party affiliation."   It's a typically rich study for Pew, and typically frustrating.  Some of the ultimate findings you'll find revelatory ("Environmental protection now stands out as a major divide within the GOP's coalition".)  On the other hand, the political typologies seem as naive and unnuanced as those hired to represent opinion on FOX.  What to make of it?

First, let's have a look at the types Pew identifies, keeping in mind that there are serious methodological questions about how they arrived at these categories (we'll deal with that down below).  Fuller descriptions of the types are available here; descriptions with supporting statistical documentation here.

The Left

Liberals are opponents of an assertive foreign policy, strong supporters of environmental protection, and solid backers of government assistance to the poor. This affluent, well-educated, highly secular group is consistently liberal on social issues.

Conservative Democrats are quite religious, socially conservative and take more moderate positions on several key foreign policy questions. The group is older, and includes many blacks and Hispanics.

Disadvantaged Democrats ... are the least financially secure voting bloc. Members of this heavily female, poorly educated group are highly pessimistic about their opportunities in life, and also very mistrustful of both business and government.

The Political Middle

Upbeats [are] relatively moderate voters who have positive views of their financial situation, government performance, business, and the state of the nation in general. They voted for Bush by more than four-to-one last November. 
[are] much less affluent and educated than the Upbeats. They are deeply cynical about government and unsatisfied with their financial situation. Even so, Disaffecteds lean toward the Republican Party.

Bystanders largely consign themselves to the political sidelines. This category of mostly young people, few of whom voted in 2004, has been included in all four of the Center's political typologies.

The Right

Enterprisers have perhaps the most consistent ideological profile of any group in the typology. They are highly patriotic and strongly pro-business, oppose social welfare and overwhelmingly support an assertive foreign policy. 

Two other groups on the right are both highly religious and very conservative on moral issues. Social Conservatives agree with Enterprisers on most issues, but they tend to be critical of business and supportive of government regulation to protect the public good and the environment. They also express deep concerns about the growing number of immigrants in America.

Pro-Government Conservatives also are broadly religious and socially conservative, but they deviate from the party line in their backing for government involvement in a wide range of policy areas, such as government regulation and more generous assistance to the poor.

Did you find yourself in that group?  Here's where the rubber meets the road.  Although the categories appear descriptive, I'm afraid they're just a more sophisticated prescriptive formulation.  You become instantly aware of the problems when you take their online test to determine your own type.  Questions are divided into two statements, one of which you must either agree or strongly agree with.  What are the problems?  Take a look at this one (each time you go to the page, the unnumbered questions are shuffled into different orders): 

Statement 1: Poor people today have it easy because they can get government benefits without doing anything in return

Statement 2: Poor people have hard lives because government benefits don't go far enough to help them live decently

Care to agree with either of those?  I sure didn't.  And this exposes, to my mind, why we should ignore the typology and instead look at Pew's larger findings (a few of which are at the end of this post).  The methodology Pew employed to create the typologies drew from telephone interviews they conducted that "were those shown statistically to be most strongly related to the underlying dimension."  They then used cluster analyses to sort and filter out the nine groups. 

Statistically, it looks totally sound.  Where I question the study is in the lumpen questions it used to create the initial correlates.  This is the downfall of political surveys--it's very hard to measure political attitudes for people with very weak political beliefs based on very scant political knowledge.  To use the example I quoted above, no doubt "government benefits" are very strongly correlated with certain political attitudes.  But are these attitudes fixed and based on a polital philosophy, or are they merely the reflection of tenuously-held, shifting beliefs based on which party has political ascendency and control of the bully pulpit (to use one example)? 

But some of the findings are so striking that it's not really important whether you agree with the typologies.  For pols and hacks, these are hard to ignore, and here lies the real strength of the report.

Hat tip to Tom Maguire for pointing out the study.  Others commenting: Matt Yglesias (left), Daly Thoughts (right), Patrick Ruffini (right).  Draw your own conclusions about why the right-wing blogosphere is more fascinated about this than the left.

  • David Wright (unverified)


    I took the quiz, came out as an "Upbeat" -- one of those moderate groups that favors the GOP. Generally speaking, I think that's true. But looking at the "defining values" of the group, I'm a bit puzzled as to how I got placed there. I'm not satisfied with the direction the nation is heading, and I'm not supportive of Bush's "leadership" on economic matters versus social/foreign policy matters. Well, really I don't think much of his "leadership" on any matters...

    But then these are fairly arbitrary groupings after all, individuals are likely to see themselves in more than one group. I have an affinity for at least half the characteristics in both the "Enterprisers" and "Liberals" groupings (which may explain why I end up in the "mushy middle", actually...)

    So in light of that, it's kind of interesting to read this stuff in sort of the same way that it's kind of interesting to read one's horoscope. It's fun to look for yourself in these descriptions, but it doesn't really provide useful information for actually living your life.   ;-)

  • Gregor (unverified)

    Why is the right so fascinated with this? A quick glance at their categories gives me some guesses.

    Liberals - wealthy, well-educated and secular. And the right is not? Maybe they should say wealthy with a social conscious. Thoughtful to a fault. Concerned with nuances, even capable of understanding a nuance. one might even say they are not likely to give much credence to polls with obvious bias. And secular, as though we can't see that our role on this god given plant is to be stewards of the Earth and respect it because it was made with God's hands rather then trash it whenever a restraint would diminish the profit margin.

    Conservative Democrats? You know, the lost minorites that cling to the Democratic Party but are half way there. Perhps their Conservative side will bring them to love the Darkness.

    Disadvantaged Democrats? Well, why would anyone consider that a biased label? It's practically synonymous in the present state of the Union.

    Hmm? Why is the right fascinated with this study? Let's look at those middle three.

    Upbeats! They voted 4-1 for Dubya! Disaffected! But they lean the Right way! Bystanders! A big thank you for holding the door open to the White House. We love what we've done with the place.

    Now, let's get the Right right.

    Enterprisers! Not exploiters, or parasites. These folks have a "consistent ideological profile of any group in the typology. They are highly patriotic and strongly pro-business, oppose social welfare and overwhelmingly support an assertive foreign policy." I can;t say enough about these wonderful people. Oh, and they forgot to mention their patriotism is limited to a republican Administration. License is given to trash Bill Clinton at each and every opportunity. President's do not lie under oath, and this one refuses to testify under oath even if his pit bull is there at the table with him.

    Social Conservatives! Well, they agree with enterprisers, except they would regulate business. [Wait?!? Didn't they just say they agree?!?] They would protect the environment. Oh, and they would protect the public good. As who defines it? What about the gay immigrants?

    Finally, the Pro-government Conservatives. How noble. Again, whose government do they support? If you cherish the present state of the Union, you might be pro-government. But ten years ago, these two descriptions Pro-government and Conservatives were oil and water. Like a Conservative Democrat, you just kind of look at them and ask, so which one are you right now? Pschyzophrenia made simple.

    I'm fascinated by the study too. And I believe it must have been a relgious study because why would they call it a Pew study, and it was sanctioned by Gawd, because Pat Robertson told me so.

  • glenlivid (unverified)

    Seems like a huge waste of your time to me, Jeff. The survey is obviously flawed, and couple that with "interviewing" only 2,000 people and one has to ask, "Do you have to stick your finger in crap and smell it before you can say it's crap?".

  • BillO (unverified)

    As much as it's hard to categorize the complexity of Americans, I have to note that 2,000 folks is a statistically valid sample, if they're the right folks.

  • Gregor (unverified)

    Oh these were the Riech folks, I'm sure. Both giving and taking the survey.

  • Gregor (unverified)

    Oh these were the Reich folks, I'm sure. Both giving and taking the survey.

  • (Show?)

    Pew does great work, and I think this is a case wherein the usual methodology contains a hidden booby trap. I have no problem with Pew's methodology in general, and they were the first to document the import of religion in politics, they had the most accurate polling throughout the election season, and their numbers looked very close to the actuals after the election. In short, methodologically, Pew really rocks.

    And maybe it's me who's misinterpreting this. As someone who wants to use these kinds of studies to know how to approach changing minds and building coalitions, I went look for clues in this study. But maybe Pew doesn't care if I can't look at the way they've sliced the "disaffecteds" to see what that group really believes. But that's where my interest is.

  • glenlivid (unverified)

    "As much as it's hard to categorize the complexity of Americans, I have to note that 2,000 folks is a statistically valid sample, if they're the right folks."

    I think the survey is more important than the sample in this case. The choices are too simple. You shouldn’t be given a choice of "strongly agree" and "strongly disagree" to very divisive questions. Not to mention the fact that it reverses the original question with the same two choices, and makes the whole survey look like a Florida “butterfly ballot”. This is the first time I’ve seen a survey of this sort.

  • (Show?)

    Many surveys are of this sort. You can easily refuse to answer pairs and it will still categorize you as far as possible.

    Yes, of course issues are complicated. Surveys like this draw on correlated sets of responses. Picking at individual items is simplistic. Do you think a liberal would consistently respond "agree" to a set of questions? If you do, then the methodology is sound.

    I was a little surprised at my results. Like David, I thought I'd come out "upbeat", but apparently I'm too socially liberal (I came out "liberal"). I suspect I'm a lot more skeptical of government activity and more interventionist than many on this blog, but after teaching politics for 15 years, I just can't endorse statements that say all government is wasteful etc.

  • glenlivid (unverified)

    "And maybe it's me who's misinterpreting this. As someone who wants to use these kinds of studies to know how to approach changing minds and building coalitions, I went look for clues in this study. But maybe Pew doesn't care if I can't look at the way they've sliced the "disaffecteds" to see what that group really believes. But that's where my interest is."

    I think you have noble goals, but I have to say something is wrong with this survey. In the politically savage climate we live in, simple questions and small samples probably aren’t enough. Unless you can devise a more intricate survey and implement a bigger sample, you’ll just create another bunch of useless data.

    I personally feel there is something weird about the interpretations of the data as well. Labels like Enterprisers and Disaffecteds seem odd to me. Shouldn’t you just have a breakdown of data in a large variety of political views rather than putting labels on the data groups? In other words, something like this:

    43% of the people that aligned themselves as moderate republicans felt the following:

    50% attended church as many as 2 times per week

    20% attended church as many as 1 time a month

    15% attended church as many as 4 times a year

    25% attended church less than 4 times a year

    I guess I’m looking at a way of getting the most out of the survey. Get the most data, use the biggest sample, and don’t break the sample up with terms that sound like bad or good labels.

  • Gregor (unverified)


    My take is that there are surveys and there are agendas. This survey fits an agenda. Are the numbers valid and meaningful? Probably, but the answer was already in mind by the way they approached the questions.

    How do we change minds or build coalitions? Well, my ranting probably wasn't very useful toward attracting people to our cause, but I think we do need to express ourselves as forcefully and with as much conviction as we see from the other side.

    One study I recall reading reported that we see more then we hear, so when all is said and done, the drama relates the victor to the observers, and we concede far too much to these liars that proclaim their patriotism while destroying democracy.

  • Sally (unverified)

    The first few questions were easy to answer; after that many got tough. But I found no "typology" for the -- surely not uncommon? -- category of social liberal/fiscal conservative to which I would assign myself.

    I fell into the "disaffected" category, which would fit me only in part, and hardly including disinterested in current events or politics.

    The issues breakdowns were interesting, and not necessarily predictable. For example, "Enterprisers" overwhelmingly support the Bush tax cuts, but are almost evenly divided on raising the minimum wage. Enterprisers oppose gay marriage by 90 percent, while liberals favor it by 80 percent. Every other group stands opposed.

    I have never been affiliated with either political party, and am highly unlikely ever to do so. Aren't registered independents a near-plurality in Oregon? This site purports to be both "Blue" and "progressive," but it seems more straight DP to me. To which in my mind it is not at all safe to assume "progressive," though the assumption is largely automatic with insiders here.

  • Gregors Mommy (unverified)


    Don't hit your sister!

  • Gregor's Mommy (unverified)


    Don't hit your sister!

  • Gregor (unverified)

    Gonzo- Get a life.

  • Patience (unverified)

    As someone who wants to use these kinds of studies to know how to approach changing minds and building coalitions, I went look for clues in this study. But maybe Pew doesn't care if I can't look at the way they've sliced the "disaffecteds" to see what that group really believes.

    <h2>This isn't made clear in the survey report itself, but Pew does release the raw datasets for its surveys (usually after a 6-month delay), so eventually you will be able to drill down into the data for yourself and try to tease out what the "disaffected" believe.</h2>

connect with blueoregon