Mapping American Four-Quadrant Political Ideology

[RI]

I've long desired to make a series of detailed maps breaking down the American electorate into the common four-quadrant ideological spectrum used here on Atlas Forum and elsewhere on the internet, but the data and my ability haven't been there to date. Thanks to the VOTER survey and MRP, that has changed or at least made an attempt possible.

The VOTER survey has a large sample of 8,000 respondents asked a barrage of ideological questions. It was previously used by Lee Drutman to create the following chart, which depicts the American electorate spread predominately across three of the four quadrants (sorry, libertarians):



My first goal was to replicate Drutman's methodology and use MRP to project this distribution spatially (his methodology does have some issues, I discovered, but I ignored them to replicate his work). I then matched up the survey response categories with equivalent questions from the 2013-17 ACS PUMS sample, added a few of my own variables, and projected on to Public Use Microdata Areas, a geographic unit which, like congressional districts, tries to maintain rough population equality. This was the only substate geography available for the PUMS data, but I think it works just fine.

The result is this map of the leading ideology by PUMA, where red is liberal, blue is conservative, and green is communitarian (libertarian didn't lead in an PUMA):

http://maps.rynerohla.com/wp-content/uploads/2022/03/ideology-overall.png

The methodology doesn't work perfectly everywhere, but we do see some interesting patterns: the large divide among rural areas between the conservative west and communitarian east, the eastern conservative suburbs and exurbs, and, of course, liberal cities and college towns.

The following map shows which of the four quadrants has the highest z-score in each PUMA, that is, which quadrant is most prevalent relative to their overall average and standard deviation. The scale is no longer in percent of the population, but a relative scale:

http://maps.rynerohla.com/wp-content/uploads/2022/03/ideology-by-zscore.png

Here, libertarianism shows up, predominately in suburbs (especially DC, Chicago, the PNW) and in urban cores across the Sunbelt and West. Liberalism is far more constrained to the Acela Corridor, majority black urban PUMAs, and wealthy parts of the West. Communitarianism dominates in the Rust Belt, rural Northeast, and across the lowland rural South (both black and white areas), but conservativism reigns in ancestrally Republican parts of upland Appalachia, southern suburbs, retirement locales in Florida, and beyond the 100th Meridian.

Here are maps of the distributions of each ideology individually:

Conservative:
http://maps.rynerohla.com/wp-content/uploads/2022/03/conservative.png

Liberal:
http://maps.rynerohla.com/wp-content/uploads/2022/03/liberal.png

Communitarian:
http://maps.rynerohla.com/wp-content/uploads/2022/03/communitarian.png

Libertarian:
http://maps.rynerohla.com/wp-content/uploads/2022/03/libertarian.png

This is just one projection based on one survey, meaning it isn't a definitive mapping, but it reveals a number of interesting patterns within the American electorate.

[Kamala's side hoe]

My first goal was to replicate Drutman's methodology and use MRP to project this distribution spatially (his methodology does have some issues, I discovered, but I ignored them to replicate his work). I then matched up the survey response categories with equivalent questions from the 2013-17 ACS PUMS sample, added a few of my own variables, and projected on to Public Use Microdata Areas, a geographic unit which, like congressional districts, tries to maintain rough population equality. This was the only substate geography available for the PUMS data, but I think it works just fine.

Nice job! I'd like to hear more about these variables and how they correspond to policy preferences/ideology. It looks like your Communitarian enclaves within urbanized California correspond to areas with an Asian plurality/majority- specifically East San Jose, the San Gabriel Valley, NW Orange County, and possibly the SE corner of San Francisco.

The methodology doesn't work perfectly everywhere, but we do see some interesting patterns: the large divide among rural areas between the conservative west and communitarian east, the eastern conservative suburbs and exurbs, and, of course, liberal cities and college towns.

Here, libertarianism shows up, predominately in suburbs (especially DC, Chicago, the PNW) and in urban cores across the Sunbelt and West. Liberalism is far more constrained to the Acela Corridor, majority black urban PUMAs, and wealthy parts of the West. Communitarianism dominates in the Rust Belt, rural Northeast, and across the lowland rural South (both black and white areas), but conservativism reigns in ancestrally Republican parts of upland Appalachia, southern suburbs, retirement locales in Florida, and beyond the 100th Meridian.

The contrast between Left-Liberal and Libertarian strongholds in the Western US is pretty striking. You'd think areas with higher concentrations of wealthier, highly educated whites would be more receptive to Libertarianism than less well off, less educated, and/or less white urban areas.

[Southern Senator North Carolina Yankee]

The even split of Republicans tells you all you need to know of why Donald Trump got nominated while running to the left on issues like Social Security.

The present alignment is based on culture/Religion and that necessarily means that there are a large number of people in the GOP who are not represented by the Paul Ryan agenda.

The fact that big money continues to press this and that every populist movement ever has been coopted by big money and its mouth piece think tanks/pressure groups, is all you need to know to understand why the GOP is forever in a state of rebellion.

The GOP is not a limited Gov't party and it frankly never has been, it is a cultural conservative and semi-nationalist party with about half its members who prefer smaller gov't and half want a bigger government.

I am reminded of back in 2008 when Blaine Luerkmeyer and Brett Guthrie won, both were not the favored candidate of the Club For Growth. Their preferred candidate in both instances lost to more populist conservatives in these more rural parts of Missouri and Kentucky. It was the tea party that reversed this trend and led to the nominations of more fiscally conservative candidates and yet that was largely (at least at first) a suburban phenomenon, whereas the GOP has become more rural and working class since 2008 so in that regards these manifestations of populism make sense considering who comprises the GOP these days, as illustrated in the OP.

[darklordoftech]

What does “communitarian” mean?

[ElectionsGuy]

Fascinating. Great stuff, and confirms what I've suspected for a long time.

[RI]

My first goal was to replicate Drutman's methodology and use MRP to project this distribution spatially (his methodology does have some issues, I discovered, but I ignored them to replicate his work). I then matched up the survey response categories with equivalent questions from the 2013-17 ACS PUMS sample, added a few of my own variables, and projected on to Public Use Microdata Areas, a geographic unit which, like congressional districts, tries to maintain rough population equality. This was the only substate geography available for the PUMS data, but I think it works just fine.

Nice job! I'd like to hear more about these variables and how they correspond to policy preferences/ideology. It looks like your Communitarian enclaves within urbanized California correspond to areas with an Asian plurality/majority- specifically East San Jose, the San Gabriel Valley, NW Orange County, and possibly the SE corner of San Francisco.

The biggest variable I included but which the ACS doesn't have is a measure of the distribution of churches of various denominations at the PUMA/zipcode level. This was especially important to getting Mormon areas to reasonable levels.

Yes, Asians see the highest proportion of communitarians of any racial group (except possibly Middle Easterners, who you might classify as Asians regardless). Asians also have the highest share of libertarians of any race and few true conservatives. I suspect there's a large education/income or national-of-origin divide within the Asian population, although I don't have data on the latter.

The methodology doesn't work perfectly everywhere, but we do see some interesting patterns: the large divide among rural areas between the conservative west and communitarian east, the eastern conservative suburbs and exurbs, and, of course, liberal cities and college towns.

Here, libertarianism shows up, predominately in suburbs (especially DC, Chicago, the PNW) and in urban cores across the Sunbelt and West. Liberalism is far more constrained to the Acela Corridor, majority black urban PUMAs, and wealthy parts of the West. Communitarianism dominates in the Rust Belt, rural Northeast, and across the lowland rural South (both black and white areas), but conservativism reigns in ancestrally Republican parts of upland Appalachia, southern suburbs, retirement locales in Florida, and beyond the 100th Meridian.

The contrast between Left-Liberal and Libertarian strongholds in the Western US is pretty striking. You'd think areas with higher concentrations of wealthier, highly educated whites would be more receptive to Libertarianism than less well off, less educated, and/or less white urban areas.

Here's who libertarians are in the sample: they're the youngest group, skew wealthy, skew educated, skew urban, skew male, skew less religious, skew single. They work in information, finance, and science/tech. Here's the thing though: most of those are also true for the liberal quadrant, and the liberal quadrant is far larger and far more polarized than the libertarian quadrant (libertarians have remarkably similar demographics to liberals, save for being less black and far more male). That means that the places where libertarianism stands out based on the above categories, liberalism will likely stand out even more. Thus, the areas which area as libertarian in the z-score map are those which are both more libertarian than average and which are above the libertarian average more than they are liberal above average.

What does “communitarian” mean?

Communitarian refers to the "economically left, socially right" quadrant. Some four-quadrant tests label them as "statist" or "authoritarian", but I find those labels quite pejorative, certainly not something many would ever label themselves as, and not particularly accurate to what those in that quadrant believe.

Those in the communitarian quadrant make up a bit of an odd bunch because there's no formalized ideological or partisan movement which represents them in the United States, and thus don't have a common guidepost to orient themselves toward. Those in this quadrant tend to be far poorer, more rural, less educated, more female, older. They are moderately religious, most likely to be divorced, work in retail, transportation, or accommodations if they work at all; they are disproportionately unemployed, disabled, retired, or stay-at-home moms. However, they are less white than conservatives; Asians, Native Americans, and blacks all have sizable communitarian subsets (Hispanics are actually the least likely to be communitarian in this sample, somehow). They are a great social underclass, and many are in some way disaffected from modern society, which has broadly passed them by without a second thought. They are the least likely to be in politics, in academia, online, or in any way driving the national discourse.

To the extent they have a cogent ideology, they place high value in things like fairness, trust, and community cohesion; these arise in no small part because these people tend to lack these things in their own lives. Based on the survey data I have, they tend to value entitlements and welfare programs far more than conservatives do, are open to government-oriented healthcare reforms, would like higher taxes on the rich and more redistribution, are very skeptical of free trade, but also think businesses should not face heavy regulations and that the government shouldn't get too heavily in debt. On social issues, they're pretty split down the middle on the traditional wedge issues like abortion or gay marriage but do not approve of the transgender movement and are highly supportive of the death penalty. The biggest social issues which define this group is that they're extremely patriotic, highly anti-illegal immigration, and hold more meritocratic views with respect to race. Interestingly, they're quite moderate when it comes to views on Islam and actually fall to the left on issues of sexism in society.

They are the group Trump most resonated with. According to the survey, communitarians voted:

2016: Trump 50, Clinton 41
2012: Obama 53, Romney 43
2008: Obama 55, McCain 40

2016 D Prim: Clinton 57, Sanders 38
2016 R Prim: Trump 65, Cruz 12, Kasich 11, Rubio 8
2012 R Prim: Romney 29, Paul 19, Gingrich 12, Santorum 12

[RI]

For completeness, here's how other quadrants voted:

Liberals
2016: Clinton 93, Stein 3, Trump 2, Johnson 1
2012: Obama 94, Romney 3
2008: Obama 94, McCain 4

2016 D Prim: Clinton 56, Sanders 44
2016 R Prim: Kasich 51, Trump 19, Rubio 10, Cruz 8
2012 R Prim: Huntsman 38, Romney 32, Paul 17, Santorum 5, Gingrich 4

Conservatives
2016: Trump 90, Clinton 3, Johnson 3, McMullin 1
2012: Romney 93, Obama 4
2008: McCain 86, Obama 8

2016 D Prim: Sanders 60, Clinton 27
2016 R Prim: Trump 45, Cruz 33, Rubio 10, Kasich 8
2012 R Prim: Romney 29, Gingrich 21, Santorum 20, Paul 11

Libertarians
2016: Clinton 41, Trump 31, Johnson 20
2012: Romney 53, Obama 37, "Other" 9
2008: McCain 48, Obama 45, "Other" 7

2016 D Prim: Clinton 63, Sanders 33
2016 R Prim: Kasich 34, Cruz 22, Rubio 22, Trump 18
2012 R Prim: Romney 32, Paul 18, Huntsman 14, Gingrich 14, Santorum 9

[DC Al Fine]

RI do you know what share the various quadrants were of the primary electorate(s)?

[RI]

Breaking down different groups by quadrant:

Group	Liberal	Conservative	Communitarian	Libertarian
By Sex
Men	31.8	43.0	22.2	3.0
Women	38.3	24.4	34.5	2.8
By Race
White	31.9	36.2	29.1	2.8
Black	61.1	7.5	27.6	3.9
Hispanic	37.5	34.3	24.3	3.9
Asian	49.6	15.1	31.1	4.2
Native	29.5	39.3	29.5	1.6
By Education
No HS Diploma	16.0	30.7	52.7	0.7
HS Graduate	20.9	32.3	45.4	1.4
Some College	34.5	33.7	28.6	2.8
Associate  Deg	30.6	35.6	31.7	2.2
Bachelor Deg	42.3	34.4	19.6	3.8
Graduate Deg	49.1	33.1	13.2	4.7
By Income
Under $30,000	33.3	22.3	42.8	1.6
$30,000-$60,000	32.2	31.7	33.6	2.5
$60,000-$100,000	34.9	37.5	24.2	3.4
Over $100,000	41.9	38.8	14.7	4.6
By Age
18-34	53.0	19.4	20.0	7.7
35-49	43.4	28.4	23.3	5.0
50-64	32.4	36.1	29.0	2.5
65+	32.6	34.5	30.8	2.1
By Marital Status
Single	48.1	22.4	25.7	3.7
Married	31.3	39.2	26.4	3.1
Divorced	34.7	28.1	35.4	1.9
Separated	38.9	23.6	36.8	0.9
Widowed	28.8	30.1	38.4	2.7
Domestic Partnership	51.8	18.2	29.3	0.8
By Employment
Employed Full-Time	37.1	36.1	23.5	3.4
Employed Part-Time	39.6	28.2	29.6	2.6
Self-Employed	38.5	41.5	17.2	2.9
Unemployed	35.0	27.7	34.0	3.3
Retired	31.3	36.4	30.0	2.3
Disabled, NILF	31.1	21.0	46.5	1.5
Homemaker	24.9	32.8	40.3	2.0
Student Full-Time	47.9	14.9	30.9	6.4
By Industry
Agriculture and Forestry	24.4	53.3	22.2	0.0
Mining, Oil, and Gas	35.0	46.7	15.0	3.3
Construction	19.9	57.0	21.2	2.0
Manufacturing	22.5	47.8	27.9	1.9
Wholesale	30.4	37.7	27.5	4.4
Retail	30.3	35.0	32.1	2.6
Transportation	18.6	45.9	33.7	1.7
Information	47.9	35.0	13.5	3.7
Finance and Insurance	37.7	37.7	18.3	6.2
Prof/Sci/Tech	42.7	34.1	18.0	5.2
Education	49.7	26.6	20.4	3.3
Healthcare	45.5	25.8	25.8	3.0
Arts/Entert/Recreation	52.4	27.7	18.1	1.8
Accommodation	32.3	31.5	33.9	2.4
Government	36.6	39.4	21.2	2.8
By Church Attendance
Weekly or More	21.7	46.6	28.6	3.1
Irregularly	32.0	36.6	28.4	3.1
Seldom	33.3	33.1	31.2	2.5
Never	50.4	20.8	25.8	3.0
By Religion
Catholic	26.3	40.1	31.3	2.3
---Weekly or more	19.0	50.2	27.4	3.4
---Less often	29.6	35.4	33.2	1.8
Orthodox	25.0	42.9	28.6	3.6
Mainline Protestant	29.5	36.9	29.7	4.0
---Weekly or more	29.3	41.3	25.9	3.6
---Less often	29.5	35.1	31.3	4.1
Evangelical Protestant	22.9	39.8	35.2	2.2
---Weekly or more	15.6	48.1	33.9	2.5
---Less often	28.4	33.5	36.2	1.9
Mormon	21.4	51.4	22.9	4.3
Jewish	49.8	24.2	21.2	4.8
Muslim	45.8	4.2	45.8	4.2
Buddhist	69.0	9.9	19.7	1.4
Agnostic	66.2	19.2	11.9	2.8
Atheist	73.5	11.7	11.9	2.9
Other No Religion	41.7	23.8	30.8	3.7
By Union Membership
Union Household	40.2	27.9	29.9	2.1
Non-Union Household	34.0	34.8	28.2	3.1

[RI]

RI do you know what share the various quadrants were of the primary electorate(s)?

2016 D:
Liberal 69.1, Communitarian 24.3, Conservative 4.7, Libertarian 1.8

2016 R:
Conservative 67.1, Communitarian 24.2, Liberal 5.3, Libertarian 3.4

2012 R:
Conservative 64.4, Communitarian 26.3, Liberal 5.5, Libertarian 3.8

[DC Al Fine]

Interesting thank you.

The description of communitarians and their lack of influence in politics is interesting. It's a good example of political or economic actors not being entirely self-interested. It's a wonder that some self interested politician (Trump sort of has I guess), hasn't exploited this for their electoral gain, given how underserved they are in so many places.

[Kamala's side hoe]

RI do you know what share the various quadrants were of the primary electorate(s)?

The results from the original survey in the article RI linked was: 44.6% liberal, 28.9% communitarian, 22.7% conservative, and 3.8% libertarian. RI’s average for the national electorate is closer to 35% liberal, 28% communitarian, 34% conservative, and 3% libertarian, assuming a 50-50 gender split.

The description of communitarians and their lack of influence in politics is interesting. It's a good example of political or economic actors not being entirely self-interested. It's a wonder that some self interested politician (Trump sort of has I guess), hasn't exploited this for their electoral gain, given how underserved they are in so many places.

RI’s description of social/cultural issues is likely most salient for white communitarians. 2016 exit polls suggest that Trump’s anti-illegal immigration campaign resonated with nonwhite communitarians to some extent.

[darklordoftech]

It’s interesting how well Jimmy Carter performed with the “communitarian” part of America in 1976.

[Nyvin]

This supports the idea that the American public does NOT support fiscal conservatism.   It's really just the donor class that pushes for it.

[RI]

RI do you know what share the various quadrants were of the primary electorate(s)?

The results from the original survey in the article RI linked was: 44.6% liberal, 28.9% communitarian, 22.7% conservative, and 3.8% libertarian. RI’s average for the national electorate is closer to 35% liberal, 28% communitarian, 34% conservative, and 3% libertarian, assuming a 50-50 gender split.

The original article's methodology was not very specific in how the measures were created, exactly which variables were included or not included, etc., so I did my best to replicate them, but differences were inevitable. My equivalent of the upper most chart looks like this:

http://rynerohla.com/wp-content/uploads/2019/08/matrix.png

My version is less extreme/polarized than Drutman's. It does appear slightly less skewed economically left, but not dramatically so. I did feel that Drutman's version weighted certain economic issues more heavily than they should have been. This probably explains a lot of the differences in 2016 vote; Drutman pushes some of my conservatives into the communitarian quadrant, which inflates Trump's share with them. His electorate is also a bit more socially liberal than mine as I included a question about the death penalty which he did not.

[darklordoftech]

I have trouble believing that New Jersey is more liberal than Vermont.

[100% pro-life no matter what]

Interesting.  Not sure that I totally buy the size of the commutarian group- and I'm not a huge fan of 2 axis charts.  Sometimes they have an agenda (e.g. the political compass trying to say that the Democrats are conservative or libertarian organizations trying to push people into the bottom right), but regardless, they are not super descriptive for an ideology.

The part I really do find interesting is that this suggests that conservatism as an ideology is actually rooted in the suburbs and is not inversly correlated to education at all.  I've been saying that for a long time.  I imagine that the same actually applies for the backbone of religious-based social conservative movements.  Even if it's not everybody in well-off suburbs, movements like these tend to have their most fervent support in suburbs.  For example, I guarantee you that evangelicals who actually don't drink, save themselves for marriage, avoid cursing, and the like come from a lot less rural and "white working class" backgrounds than a lot of people assume.

[Linus Van Pelt]

Interesting stuff.

Which demographic variable(s) lead(s) to the conservative/communitarian split between the non-metro west and the non-metro east? (Outside of the black areas of the south).

[RI]

Which demographic variable(s) lead(s) to the conservative/communitarian split between the non-metro west and the non-metro east? (Outside of the black areas of the south).

Good question; I'm not 100% sure. Within the model, I'd imagine its some combination of the large difference in population density, differing religious patterns (evangelicals are far more communitarian than the very conservative LDS, for example), fewer disabled people in the rural west, and higher education levels in the rural west.

In reality, I'd place the majority of the emphasis on the fact that rural areas in the west are far more sparse than in the east; there's quite a lot of rural population in the east which is organized around and in between small towns. In the rural west, there's often virtually nothing but a few large farms between towns. The former allows more easily for community identities to form while the latter promotes individualist attitudes.

[darklordoftech]

I feel like these maps overestimate social and cultural conservatism. I feel like Nevada should be yellow and Vermont should be red.

[RI]

I feel like these maps overestimate social and cultural conservatism. I feel like Nevada should be yellow and Vermont should be red.

Vermont perhaps, Nevada no. Remember, "social conservatism" is defined here by more than just wedge issues like abortion and gay marriage. It's more abstract than that. Second, VT's liberalism is a bit of an oddity as it's not really a demographic phenomenon (outside things like atheism or LGBT identity, which ACS data doesn't have, or ancestral characteristics which the VOTER survey doesn't have), which means it's difficult for this type of model to pick it up.

As far as NV goes, the state has a very low education rate and lots of other demographics (e.g. industrial composition, higher union membership, higher median age, higher divorce rate/lower single rate) which make it not so libertarian-friendly in the model, although there are parts of the state where this is less so.

[💥💥 brandon bro (he/him/his)]

Very interesting project although there are some results I am curious about. In particular some of those very sharp contrasts in the z-scores (looking in particular at that divide between Eastern Wisconsin and Western Wisconsin and Minnesota which is not exactly what I would expect).

Next steps? I opened the Drutman report on my phone back in 2017 when it came out and, checking just now, I still have a tab open in my browser which I never closed (I think I started it once when I was at work but never finished)

Interestingly, they're quite moderate when it comes to views on Islam and actually fall to the left on issues of sexism in society.

I found your explanation of the "communitarians" as basically white parochial/traditional folk to be pretty compelling but I am having trouble reconciling your explanation with the above claims. Do you have some sort of mechanism in mind which would create these views or is this some sort of polling aberration?

Interesting thank you.

The description of communitarians and their lack of influence in politics is interesting. It's a good example of political or economic actors not being entirely self-interested. It's a wonder that some self interested politician (Trump sort of has I guess), hasn't exploited this for their electoral gain, given how underserved they are in so many places.

The crazy thing about this too is that if you put a different Trump-like candidate (i.e. one with similar communitarian views but without the personality) in there in 2016 he/she likely wouldn't have won. The book "Identity Crisis" (the authors are Sides, Tesler, and someone else whose name I am forgetting) have a chapter in their book about the somewhat unique nature of the 2016 primary, e.g., the fractures within the party elite which meant there was a failure to rebut Trump with a unified message the base could easily digest. It seems in retrospect that there was always an obvious hole for a Trump-like entitlement-protector to exploit. The fact that the Tea Party started as a movement of deficit hawks but was almost immediately coopted by immigration hawks foreshadowed this.

We're seeing some more populist-right energy from young Republican politicians like Hawley and Cotton for what it's worth.

[RI]

Interestingly, they're quite moderate when it comes to views on Islam and actually fall to the left on issues of sexism in society.

I found your explanation of the "communitarians" as basically white parochial/traditional folk to be pretty compelling but I am having trouble reconciling your explanation with the above claims. Do you have some sort of mechanism in mind which would create these views or is this some sort of polling aberration?

They were fairly hasty generalizations on my part, so let's dig a little deeper.

1) On Islam: Turns out I incorrectly coded one of the Islam questions. I went through and checked all the questions, and that was the only one I coded incorrectly. When you fix that, you get more of the expected results: communitarians are a bit left of conservatives on Islam, but nothing notable compared to the gap between them and liberals/libertarians. It doesn't really make any differences in the overall distributions or anything.

Looking at the Islam attribute (scaled -1 as furthest left and 1 as furthest right based on the answers available; this is relative, not an absolute scale) goes:

Conservatives: 0.4224
Communitarians: 0.2551
Libertarians: -0.3262
Liberals: -0.4981

There is a bit of a racial divide within communitarians, who are far more black/Asian than conservatives:

White communitarians: 0.2893
Non-white communitarians: 0.0983

This divide doesn't show up as much within the other quadrants:

White conservatives: 0.4278
Non-white conservatives: 0.3884

White libertarians: -0.3275
Non-white libertarians: -0.3219

White liberals: -0.5074
Non-white liberals: -0.4731

2) On sexism: The key thing is that communitarians are the most female of any of the quadrants.

Communitarians: 62 Female, 38 Male
Liberals: 55 Female, 45 Male
Libertarians: 49 Female, 51 Female
Conservatives: 37 Female, 63 Male

Looking solely at the sexism attribute (scaled -1 as furthest left and 1 as furthest right based on the answers available; this is relative, not an absolute scale), we see a clear gender gap where women are further left than men, but even so, communitarians of all genders are further left than conservatives of all genders:

Conservative Men: 0.0698
Conservative Women: -0.0823
Communitarian Men: -0.1748
Communitarian Women: -0.3836
Libertarian Men: -0.4449
Libertarian Women: -0.5969
Liberal Men: -0.7402
Liberal Women: -0.8039

Breaking this down further, the two prompts were communitarians really set themselves apart from conservatives were "WHEN WOMEN DEMAND EQUALITY THESE DAYS,THEY ARE ACTUALLY SEEKING SPECIAL FAVORS" and "In the U.S. today, do men have more opportunities for achievement than women have, do women have more opportunities than men, or do they have equal opportunities?" In both cases, communitarians were far more likely than conservatives to select the more left-wing answer. Certainly in the latter case, this shows the "fair world" vs. "rigged world" divide between the two quadrants: communitarians think society is rigged, including against women. This answer also likely informs the former; if you think the game is rigged against women, demanding equality can't be seeking special favors, only restoring balance.

[catographer]

RI do you know what share the various quadrants were of the primary electorate(s)?

2016 D:
Liberal 69.1, Communitarian 24.3, Conservative 4.7, Libertarian 1.8

2016 R:
Conservative 67.1, Communitarian 24.2, Liberal 5.3, Libertarian 3.4

2012 R:
Conservative 64.4, Communitarian 26.3, Liberal 5.5, Libertarian 3.8

Thinking about the 2020 Democratic primary electorate, I’m guessing that Biden is gonna do better with communitarians, whereas Warren and Sanders will do better with liberals. Not sure where conservative and libertarian Democrats will go tho...

[RI]

Let's try tweaking the model a bit to see what falls out. What happens if we restrict our quadrant definitions to such that your social score is only defined by your views on culture wars issues (abortion, gay marriage, transgenderism, etc.) and your econ score is only defined by your views on the abstract size of government? This is the model which is most plausibly favorable to libertarians and least favorable to communitarians.

How does this change things? Quite a bit. The quadrant distribution becomes:

Conservative: 43.9%
Liberal: 32.1%
Libertarian: 16.0%
Communitarian: 7.9%

If someone landed on an axis, I used their trade score to break an economic tie and their immigration score to break a social tie.

Here's how these alternative groups voted:

Liberal
2016: Clinton 91, Stein 3, Trump 3, Johnson 1
2012: Obama 94, Romney 2
2008: Obama 94, McCain 3

2016 D Prim: Clinton 53, Sanders 47
2016 R Prim: Kasich 55, Trump 29, Cruz 8, Rubio 8
2012 R Prim: Paul 27, Romney 22, Huntsman 20, Gingrich 7, Santorum 3

Conservative
2016: Trump 85, Clinton 8, Johnson 3
2012: Romney 85, Obama 12
2008: McCain 79, Obama 15

2016 D Prim: Sanders 46, Clinton 44
2016 R Prim: Trump 51, Cruz 32, Rubio 10, Kasich 8
2012 R Prim: Romney 27, Gingrich 22, Santorum 19, Paul 11

Libertarian
2016: Clinton 55, Trump 31, Johnson 9, Stein 2
2012: Obama 58, Romney 36
2008: Obama 62, McCain 33

2016 D Prim: Clinton 59, Sanders 39
2016 R Prim: Trump 44, Kasich 27, Rubio 15, Cruz 14
2012 R Prim: Romney 42, Paul 19, Gingrich 10, Santorum 10, Huntsman 8

Communitarian
2016: Clinton 57, Trump 36, Johnson 2
2012: Obama 66, Romney 30
2008: Obama 65, McCain 32

2016 D Prim: Clinton 71, Sanders 26
2016 R Prim: Trump 65, Cruz 17, Kasich 11, Rubio 7
2012 R Prim: Romney 35, Gingrich 15, Paul 14, Santorum 12

This model is far less predictive than the first model, and some of the above numbers are a bit hard for me to swallow. Some of the demographics become very wonky, too: libertarians are now over 60% female, which seems preposterous, communitarians are over 20% black, blacks are 23% conservative and only 39% liberal, etc. As such, this best-case-for-libertarianism model seems unlikely to be the case, suggesting a size-of-government vs. culture wars dichotomy is not very reflective of actual American voting behavior.