How Did Division I-A College Football Cities Vote in 2016?

[100% pro-life no matter what]

Sun Belt Conference:

Arkansas State University....

Jonesboro, Arkansas:

8,396 HRC (33.9%), Trump 14,823 (59.8% R)                                    + 25.7% R

Very interesting.... not sure if anyone has insights onto Jonesboro and Arkansas State University, but it does seem a bit odd that it voted much more heavily Republican, than even "Whiter" College Cities in the Cotton Belt and Appalachia....

A possibility (that may be impossible to confirm) would be that Arkansas St. had a higher percentage of the vote coming from undergraduate students compared to professors and academics (including grad students).  It is possible that groups on campus made more of an effort to get students to vote at their campus address (rather than their parents' house) than at some other universities.  It wouldn't surprise me if the undergrads at somewhere like Arkansas St. voted 60-75% Trump (depending on the racial demographics of the student body, which I'm not sure of), which offset the professors and academics in town.  Unfortunately, other than seeing how many votes were cast in the campus precinct (which can be very low and reflective of student bodies as a whole if a group of one party encourages it and the other doesn't, for example) or popular off-campus residence areas for students, it's impossible to confirm or deny this theory.

[100% pro-life no matter what]

Nashville can be counted in a couple different ways.  Officially, the city government and Davidson County are co-terminous.  However, there are a couple semi-independent municipalities that vote for the Nashville city government and their own government.  These tend to be quite Republican, so there is no doubt that Hillary won Nashville, but it is a question of whether or not we want to include them.  If we do (since they do vote for the mayor and city government), then we can just take Davidson County's result for Nashville.

[100% pro-life no matter what]

Big 12:

West Virginia University:

Morgantown, West Virginia:

4,580 HRC (55.4% D), 2,767 Trump (33.5 % R)                      +21.9% D

So this is really interesting... for a City of some 31k there were only 8.3k total votes, which seems incredibly low, even for a University town where students are less likely to turn out to vote, many vote absentee in other counties within their home-state, etc....

Also, I was a bit surprised by the size of the Clinton margin here, considering WVU pulls heavily from a student population with strong local ties to the Steel & Coal belt areas that swung heavily towards Trump in '16.... Would be interesting to run comparative numbers for '12/'16 to look at the 'Pub and 'Dem baseline numbers respectively to see how Morgantown is trending....

Actually, that jives pretty well with my theory that only a tiny percentage of the votes in college towns come from traditional students and that a lot more come from professors, administrators, and maybe some bohemians.  I think it would be a safe assumption that most WVU students voted for Trump, (a few probably did vote in Morgantown, but most probably didn't), but they don't make up the majority of the city's electorate.  The only places that we have seen vote for Trump by double-digits are in the Deep South or Mormon states, where (given racial and religious polarization in these places), it might make sense that even the professors/administrators wouldn't be as liberal.

[100% pro-life no matter what]

Big 12:

West Virginia University:

Morgantown, West Virginia:

4,580 HRC (55.4% D), 2,767 Trump (33.5 % R)                      +21.9% D

So this is really interesting... for a City of some 31k there were only 8.3k total votes, which seems incredibly low, even for a University town where students are less likely to turn out to vote, many vote absentee in other counties within their home-state, etc....

Also, I was a bit surprised by the size of the Clinton margin here, considering WVU pulls heavily from a student population with strong local ties to the Steel & Coal belt areas that swung heavily towards Trump in '16.... Would be interesting to run comparative numbers for '12/'16 to look at the 'Pub and 'Dem baseline numbers respectively to see how Morgantown is trending....

Actually, that jives pretty well with my theory that only a tiny percentage of the votes in college towns come from traditional students and that a lot more come from professors, administrators, and maybe some bohemians.  I think it would be a safe assumption that most WVU students voted for Trump, (a few probably did vote in Morgantown, but most probably didn't), but they don't make up the majority of the city's electorate.  The only places that we have seen vote for Trump by double-digits are in the Deep South or Mormon states, where (given racial and religious polarization in these places), it might make sense that even the professors/administrators wouldn't be as liberal.

(Bolded) I don't think that would be a safe assumption at all tbh...

Take a look at Precinct #10 in Monongalia County for example, which is directly located right next to campus.... 741 total Pres votes (57- 28 D).... precinct # 16 right next to campus 543 total votes (58- 29 D)....

We could also pull up Precinct #2 which essentially was 2:1 HRC, home to a large number of college students...

Generally in my personal experience, University Professors don't like to live in overwhelmingly student precincts since the inevitable parties and noise distracts them from their research projects, and obviously in a relatively inexpensive city to live in like Morgantown, their salary can afford better housing, with still only a 15 minute commute to work....

Additionally, the further one moves away from the WVU campus, the greater increases for the Republican vote share within the City, as well as cities/communities in "Metro Morgantown"....

Now, aside from the objective data from Morgantown, which is but one blip of a data point, here's the part of your theory that I believe to be generally fairly accurate:

"my theory that only a tiny percentage of the votes in college towns come from traditional students and that a lot more come from professors, administrators, and maybe some bohemians."

When I was going to college in Ohio in the early '90s, I didn't vote in Ohio. I voted via absentee ballot as a college student for elections in Oregon, since although I had a choice of which state I could register to vote in, I was much more interested in elections back home than elections in Ohio...  

This is not unusual for out-of-state students, and this pattern exists to the present day.

Even WVU undergrad students who grew up and have lived in West Virginia most of their lives, many still vote absentee ballots in other parts of the State....

Now, where college students are most likely to register to vote within the home county of their academic community, would be Grad students. After all, the most exploited of all of academia  slaving away in the "knowledge farms and factories" of America, to basically teach the 100/200 level classes for the professors, in order to gain a slight break on their student loan debt, in the hopes that one day they might be able to get hired on and eventually tenured, have greater roots in the community than undergrads.

Now, flip side to the coin, the vast array of workers are either direct hires or contract workers that do everything from the cube farms of the Universities (Registry, Financial Aid, Admins to dpts) to the maintenance, food service, custodial, and landscaping employees....

So, basically I agree with certain parts of your theory regarding the % of students that actually vote within their municipality, disagree with the concept that Administrators and Professors account for the bulk of the Democratic vote within these communities, and really the giant question mark are the huge number of workers in office and facility support jobs to keep a major University campus running....

In certain parts of the "Deep South" in University towns and cities, it appears that the workers that support the University tend to be much more heavily African-American, while the student population tends to be much more Anglo.... (Thinking of you Auburn, Alabama). In other cases, such as Morgantown, it appears that the "worker bees" tend to be heavily White and skew a bit Republican...

I confess as an individual who grew up in a City dominated by the University, the whole townie/gownie thang was real.... Back in the '80s student campus precincts were the most Republican parts of the City, other than the wealthy upper-income giant houses in the hills North of town...

This is an interesting subject that you raised, essentially the question of what is the actual share of the "student vote" in College/University towns vs those working directly/indirectly for these      "Non-Profit", public academic institutions, where the College/University is still one of the biggest employers, and plus if one gets hired on as a state employee, you are protected by the Union, have collective bargaining rights, actual decent employer covered medical insurance, PTO that you can bank and roll over, decent employer contribution to 401ks, and all that great stuff.

Once again, you raised many important questions, since for many of the communities on the list, the "Company" is the "University", for a decent chunk of individuals that live, work, and go to school in these communities....

There three reasons I don't trust campus precincts (it goes both ways- I doubt Trump won the student body at Pitt- despite the fact that he won the campus precinct):

1. Few students vote on campus particularly in places without on-campus polling places- and that small group may not be representative of the student body.  In my experience, the College Dems were the one group pushing their members to register on campus, so the campus precinct may be more Democratic than the student body as a whole.  Pitt may be a reverse example of that, but I kind of suspect that more places have the bias that I observed.

2. Particularly at smaller private colleges, there could be a large number of faculty members who vote on campus and distort the results due to several live-in faculty members.  Vanderbilt may be a good example of this effect.  There were only 135 votes in the campus precinct.  However, that includes some areas just off campus (and in a city, those could be any sorts of people).  Plus, there are ~30 faculty members (including spouses and perhaps more if there are some I don't know about) who live on campus, mostly due to a freshman residential program.  It seems likely that there were significantly under 100 students who actually voted on campus, and probably not even that many undergrads- which gets to my last point.

3. When I think of a student body, I am generally talking about the undergrads.  Undergrads everywhere are going to be significantly less liberal than grad students (not because of any inter-generational divide, but because people who pursue graduate education tend to be more of the academic mindset than simply going to college to improve job prospects, and we all know academia leans overwhelmingly liberal).  Grad students are usually going to live just off campus (which, for some schools, may overlap with certain key precincts to evaluate the student body) and probably vote at the campus address at a much higher propensity than undergrads, as they have been removed from their parents' home for significantly longer and likely have less of a connection to it.

An interesting test of this might be LSU.  They did a mock election in their student center that had no partisan sponsors a few weeks before the election, which Trump won by 15.  I'm not sure if that was just undergrads or grad students too, though.  But, we could try comparing that to LSU's campus precinct.
http://www.lsunow.com/daily/lsu-student-body-picks-trump-in-mock-election/article_b1873e34-8ab4-11e6-82da-4721c10b20d9.html

[100% pro-life no matter what]

We can see how 18-29s with a college degree voted by state (yes, this might vary a little- I do agree that the "State Name STATE" school is usually more GOP than the "University of State Name" school) in 2012 and potentially use this as a good starting point for undergraduates (above 49% probably means a Democratic win with third party support levels from 2012):

Utah- 32% Democrat
Wyoming- 32%
Louisiana- 35%
Alabama- 37%
Idaho- 40%
Oklahoma- 40%
West Virginia- 40%
Tennessee- 41%
Arkansas- 41%
Nebraska- 41%
South Carolina- 42%
Kansas- 42%
Mississippi- 43%
Kentucky- 43%
South Dakota- 44%
Texas- 45%
Indiana- 45%
Georgia- 45%
Alaska- 46%
North Dakota- 46%
Missouri- 49%
Montana- 50%
North Carolina- 50%
Ohio- 51%
Arizona- 52%
Virginia- 53%
Michigan- 56%
Wisconsin- 57%
Pennsylvania- 57%
Minnesota- 57%
Florida- 57%
Colorado- 57%
New Hampshire- 58%
New Mexico- 58%
Iowa- 59% (given the trends of Iowa from 2012-16 and the lack of an age gap in the state, this may have been somewhat different in '16)
Illinois- 59%
Delaware- 59%
Connecticut- 59%
Oregon- 61%
Nevada- 61%
New Jersey- 62%
Maine- 63% (again, this margin might be less in '16)
Maryland- 63%
Washington- 64%
Massachusetts- 66%
California- 66%
Rhode Island- 66%
New York- 68%
Vermont- 75%
Hawaii- 79%
DC- 85%
(Data gathered from this infographic: https://www.nytimes.com/interactive/2016/06/10/upshot/voting-habits.html)

While millennials as a whole swung Republican in 2016, it is likely that college-educated millennials did not due to Trump's unpopularity with the group.  From what we found in this thread, combined with these numbers, trends from 2012-16, and general understandings about the student bodies, here is what I would say for the Power 5 conference schools' undergraduate student bodies:

ACC: Trump: Clemson, Louisville, NC State (narrowly)
Clinton: Florida State, Miami, Georgia Tech (narrowly), Duke, UNC, Wake Forest (narrowly), Virginia, Virginia Tech (narrowly), Notre Dame (narrowly), Pittsburgh, Syracuse, Boston College

SEC: Trump: South Carolina (narrowly), Tennessee, Kentucky (semi-narrowly), Missouri (narrowly), Alabama, Auburn, Ole Miss, Mississippi State, LSU, Arkansas, Texas A&M
Clinton: Florida (semi-narrowly), Georgia (narrowly), Vanderbilt (semi-narrowly)

Big 12: Trump: Baylor, Texas Tech, TCU (narrowly), Oklahoma, Oklahoma State, Kansas State, West Virginia, Iowa State (narrowly)
Clinton: Texas, Kansas (semi-narrowly)

Big Ten: Trump: Purdue (narrowly), Nebraska (semi-narrowly)
Clinton: Maryland, Rutgers, Penn State (semi-narrowly), Ohio State (semi-narrowly), Michigan, Michigan State (narrowly), Indiana, Northwestern, Illinois, Wisconsin, Minnesota, Iowa

Pac-12: All Clinton (Utah narrowly- conservatives in Utah probably mostly go to BYU and Utah State)

Overall, I am giving Trump 24 student bodies and Clinton 41.  The vast majority of Trump's come from the SEC and Big 12.

[100% pro-life no matter what]

I am not always as familiar with G5 schools, but here would be my best guesses for their undergrads:

American: Trump: East Carolina (semi-narrowly), Houston (narrowly), Navy, Memphis (semi-narrowly), SMU (semi-narrowly), Tulsa
Clinton: Cincinnati (narrowly), South Florida, Central Florida (semi-narrowly), Temple, Connecticut, Tulane

C-USA: Trump: Marshall, MTSU, Western Kentucky, North Texas, Southern Miss, Louisiana Tech, UAB
Clinton: Florida Atlantic, Florida International, Old Dominion (semi-narrowly), Charlotte (narrowly), UTSA (I'm honestly not sure about the racial demographics there, but that will be important), Rice (narrowly), UTEP

MAC: Trump: Miami (OH) (semi-narrowly), Central Michigan (narrowly), Ball State, Bowling Green (narrowly)
Clinton: Akron, Ohio (narrowly), Buffalo, Kent State (semi-narrowly), Western Michigan (semi-narrowly), Eastern Michigan, Toledo (semi-narrowly), Northern Illinois

MWC: Trump: Air Force, Wyoming, Boise State (semi-narrowly)
McMullin: Utah State (narrowly)
Clinton: Colorado State (semi-narrowly), New Mexico, San Diego State, Fresno State, San Jose State, UNLV, Nevada, Hawaii

Sun Belt: Trump: Troy, Arkansas State, Louisiana-Monroe, Louisiana-Lafayette, Idaho (semi-narrowly), Georgia Southern, South Alabama, Coastal Carolina (semi-narrowly), Texas State (again, I'm not really sure of the racial demographics)
Clinton: Appalachian State (semi-narrowly), Georgia State, New Mexico State

Remaining Independents: Trump: Army
Clinton: UMass
McMullin: BYU

Of these, I have 33 going for Clinton, 30 for Trump, and 2 for McMullin.

My overall FBS estimates are 74 campuses for Clinton, 54 for Trump, and 2 for McMullin.  The overall popular vote is probably a bit worse for Trump than that sounds, as Trump isn't racking up the 90-10 margins from anywhere like Clinton is from places like Berkeley, Washington, and UCLA.  I would guess that the overall popular vote of FBS students is similar to the overall millennial popular vote because they are academically relatively elite (leaning towards Clinton), but also more affluent and more white than their generation as a whole (which could be good for Trump).

[100% pro-life no matter what]

Pretty amazing map if you look at it, especially considering that those graduating from the public universities with the highest status and prestige can pretty much roll out of college and get a job anywhere, and essentially represent the next generation of leaders in the worlds of business/commerce, math/science, political leadership, arts/culture, are voting so overwhelmingly against the party of Trump....

I still remember the days as a Gen Xer where the College Students were so much more conservative than the "townies" during the days of the "Reagan Youth", including working class precincts like mine....

That requires a fundamental assumption that the college town votes represent college students- particularly undergrads- themselves, rather than the feelings of career academics living in these communities.  Especially for campuses without on-campus voting, the number of undergrads voting on campus may be so tiny as to not be able to take any meaningful conclusions from (and often may not be representative, if the College Dems push voting from a campus address, for example).  Out of curiosity, do you have the numbers for LSU's campus precinct- to compare that with a relatively thorough mock election that they did?

What do you think of my list for undergrad students?

[100% pro-life no matter what]

The bottom line is it is impossible to know with certainty how undergrads voted.  Also, it is likely that at public colleges only the freshmen (maybe less likely to register on campus their freshman year) and a few students who are "in with the administration" like RAs and the like will live on campus.  Even still, it is impossible to square the precinct (13% Trump??) with the campus mock election (Trump +15 at a time when it did not look like he would win the election) and other general demographic trends (plus, I think we could all agree that there is no way a majority white group of students in the Deep South voted overwhelmingly Democratic).  It may be that there is selection bias or that others (a significant number of live-in faculty members?).  The bottom line is that the best we can do here is make educated guesses, and we have different starting points.

Also, didn't DDHQ or somewhere do an analysis of the neighborhoods of all FBS schools (this would account for a mix of undergrads, postgrads, faculty, staff, and academics) a month or so following the election?

Finally, what do you think about the issue of Nashville's boundaries, as I mentioned earlier?  Should we just take Davidson County, or should we exclude the semi-independent neighborhoods that vote for both their own mayors and the mayor of Nashville?

[100% pro-life no matter what]

The bottom line is it is impossible to know with certainty how undergrads voted.  Also, it is likely that at public colleges only the freshmen (maybe less likely to register on campus their freshman year) and a few students who are "in with the administration" like RAs and the like will live on campus.  Even still, it is impossible to square the precinct (13% Trump??) with the campus mock election (Trump +15 at a time when it did not look like he would win the election) and other general demographic trends (plus, I think we could all agree that there is no way a majority white group of students in the Deep South voted overwhelmingly Democratic).  It may be that there is selection bias or that others (a significant number of live-in faculty members?).  The bottom line is that the best we can do here is make educated guesses, and we have different starting points.

Also, didn't DDHQ or somewhere do an analysis of the neighborhoods of all FBS schools (this would account for a mix of undergrads, postgrads, faculty, staff, and academics) a month or so following the election?

Finally, what do you think about the issue of Nashville's boundaries, as I mentioned earlier?  Should we just take Davidson County, or should we exclude the semi-independent neighborhoods that vote for both their own mayors and the mayor of Nashville?

Sigh....

Extreme Republican--- Sometimes in the world of Political Science, it is less a question of trying to squeeze the foot into the shoe (Cinderella style metaphor) , versus trying to scientifically analyze the size of the shoe, and the size of the foot using height, width, and length dimensions....

Honestly, I do not care an elephants ass about some fake student election at LSU....

These are not official election results, as opposed to the actual election returns and results from overwhelmingly student precincts at LSU...

This is the 3rd time where once again you have made your spurious argument based solely upon your ideological paradigms and partisan belief structures from West Virginia to Jonesboro, to Baton Rouge, without yet providing a single shred of evidence to support your "theory".

I don't have an ax to grind on this, and don't even consider myself a Democrat.

What I am interested in and care about deeply, is transparency of election results, and objective discussions and analysis of the data, regardless of whatever "team" those results might benefit.

We don't have different starting points, since the math is the math, and the data is the data. Maybe what you refer to as different starting points assumes partisan and ideological bias?

Regarding Nashville, I would approach it in a classical style and say that precincts located within the City Limits of Nashville are the Votes from the City of Nashville....

I know there's some weird crap going around these days in various States, where essentially poorer and heavily minority Cities throughout the US are now essentially being annexed as part of "extra-jurisdictional" bodies as some type of fiscal power grab, but Nashville is Nashville, regardless of how the Unincorp areas and suburbs try to make it otherwise....

I think I made one exception to this rule regarding Athens, Georgia where the City became the County, but even there I threw some asterix's around it asking for a second opinion....

Anyways, hope you don't think I'm hitting at you at all----   at this point there is very little evidence to suggest any significant support for Trump in the overwhelming majority of I-A University campuses....

What might be interesting would be to see to what extent, how well Romney performed in '12 in many of these College/University towns....

I think we just disagree about whether campus precincts are worth much relative to other available data.  I am a current college student, and I only know a couple of people who actually vote using a campus address.  I think the other anecdotal data that I have presented is more valuable.

I may be willing to reconsider slightly if we see that there was a massive shift from Romney to Clinton, even in non-US News Top 50 universities, given that some of my data was from 2012.  Still, I think only a tiny fraction of college campus area votes come from actual undergrads.  I know you disagree, which is fine.

[100% pro-life no matter what]

For what it's worth, I do think that the precinct numbers are somewhat valuable for the "academic culture" of the reason, so I decided to check out the Tennessee schools' (sticking to precincts that contain campus):

MTSU (precinct 17-1, though it includes some off-campus area):
2016-PRES: 52-40 Clinton
2014-GOV: 60-29 Haslam
2012-PRES: 53-45 Obama
2012-2016 Swing: D+4

ETSU (combining precincts 25 and 29):
2016-PRES: 53-40 Trump
2014-GOV: 65-24 Haslam
2012-PRES: 55-42 Romney
2012-2016 Swing: R+0
Precinct 29 is overwhelmingly Republican, while Precinct 25 is competitive, but it seems that the campus spills into both, while Precinct 25 includes a good amount of downtown, while 29 has some suburban areas.

UT-Knoxville (precincts 10S and 10W):
2016-PRES: 57-33 Clinton (it is worth noting that the vote total is comically low for a large university)
2014-GOV: 43-41 Brown (Brown takes it by 5 votes out of only 228- precinct 10W was literally 1-1)
2012-PRES: 59-37 Clinton
2012-2016 Swing: D+2

Memphis (precinct 46-2, includes some neighborhood areas, low vote total in 2012, better in 2016):
2016-PRES: 50-44 Clinton (technically a plurality of 49.66%)
2014-GOV: 48-35 Haslam (that 3rd party vote is crazy)
2012-PRES: 51-44 Obama
2012-2016 Swing: R+1

UT-Chattanooga (Courthouse 1), includes significantly more than just campus:
2016-PRES: 70-21 Clinton
2014-GOV: 45-42 Haslam
2012-PRES: 72-24 Obama
2012-16 Swing: D+1

Austin Peay (unfortunately, I don't know if we can get good numbers there because it is a largely white university that is located in a broader precinct that appears to be majority-minority)

UT-Martin (Martin precincts 5 and 7, includes the entire college town):
2016-PRES: 67-29 Trump
2014-GOV: 69-23 Haslam
2012-PRES: 67-31 Romney
2012-16 Swing: R+2

So, based on these numbers (even when adding in less elite schools) not adding up to what we saw for college-educated 18-29 voters even in 2012 (in fact, the average swing was less than a point between 2012 and 2016), I am willing to conclude that campus precincts are not a fair representation of the undergraduate student body, but may represent more the views of the academic culture (grad students, professors, administrators) or may just represent which party registered students to vote on campus.

[100% pro-life no matter what]

I did check the campus precincts for Alabama and Auburn, and they both went for Trump by double digits and were more Republican than that downballot (AL gives data on straight-ticket voters, and they were roughly 2:1 GOP).  Maybe it's easier for more students to register on campus there compared to other states?  There is no reason that Alabama and Auburn would go for Trump comfortably while LSU wouldn't even give him one-in-seven votes.

[100% pro-life no matter what]

RI compiled data of all votes within a mile of campuses for DDHQ (imperfect, but another data point).  He posted the full list here in a Dropbox file, but it has expired, so I can just post what is in the public DDHQ article (for campuses in very urban or highly minority neighborhoods, it may be less accurate):

Temple: Clinton 93, Trump 5 (this one might reflect its neighborhood of Philly more than the campus)
Berkeley: Clinton 88, Trump 4
USC: Clinton 88, Trump 7 (Los Angeles)
Rutgers: Clinton 89, Trump 9 (NYC metro??)
Duke: Clinton 88, Trump 8 (Durham)
Northwestern: Clinton 87, Trump 8 (Chicago??)
Washington: Clinton 84, Trump 7 (Seattle)
Syracuse: Clinton 85, Trump 10

Arkansas St.: Trump 66, Clinton 27
Auburn: Trump 69, Clinton 28
BYU: McMullin 41, Trump 33, Clinton 20
Utah State: Trump 34, McMullin 31, Clinton 28
Troy: Trump 65, Clinton 29
Alabama: Trump 58, Clinton 35
SMU: Trump 56, Clinton 39

Idaho: Clinton 49, Trump 27
Utah: Clinton 68, Trump 14, McMullin 11 (Salt Lake City)
Wyoming: Clinton 46, Trump 37
New Mexico: Clinton 75, Trump 12 (Albuquerque)
Oregon State: Clinton 72, Trump 16
Washington State: Clinton 70, Trump 18

Additional Schools from "Liberal-Conservative Index" (Clinton+Stein vs. Trump+McMullin+Johnson):
FIU: CONS +26 (Cuban community of Miami)


FBS League Avg:
Pac-12: Clinton 77, Trump 13
Big Ten: Clinton 74, Trump 19
ACC: Clinton 73, Trump 20
MWC: Clinton 61, Trump 26
C-USA: Clinton 61, Trump 32
Big 12: Clinton 59, Trump 31
SEC: Clinton 56, Trump 36
Sun Belt: Clinton 50, Trump 43

For some reason, they excluded the MAC.  Also, this would represent all of the people related to the college town, not just the undergraduate student population.

[100% pro-life no matter what]

I did check the campus precincts for Alabama and Auburn, and they both went for Trump by double digits and were more Republican than that downballot (AL gives data on straight-ticket voters, and they were roughly 2:1 GOP).  Maybe it's easier for more students to register on campus there compared to other states?  There is no reason that Alabama and Auburn would go for Trump comfortably while LSU wouldn't even give him one-in-seven votes.

Would be interesting in seeing your data from Auburn and Tuscaloosa, in terms of which individual precincts you are referring to....

Unfortunately election results from Alabama are much less transparent than those from many other States within the region.

Trump won the UA Student Rec Center precinct in Tuscaloosa County 51-41 (excluding write-ins).  Shelby won it 61-39.  Straight ticket voters were 710-425 GOP.  Essentially, it largely looks like the Johnson voters were solidly GOP downballot.

As for Auburn, combining the Frank Brown, Dean Road, and National Guard precincts gets you 52-41 Trump and 59-41 Shelby.  Straight tickets are 3150 to 2392 GOP.  Since it includes several precincts, it probably includes more off-campus votes than at Alabama.  From everything I know, though, Auburn should be significantly more GOP than Alabama, since 50% of Alabama students are out-of-state, and those would probably be the most likely to register on campus since that would be the only way they could vote in the Alabama elections.

[100% pro-life no matter what]

I have obtained from RI the full list of all votes from within 1 mile of each campus.  This would include a mix of academics, university faculty/staff, grad students, and some undergrads (possibly some bohemians too).  Naturally, this isn't great for schools located in or near downtowns of big cities, but it can be a starting point for the "academic culture" of most schools.

[100% pro-life no matter what]

**: School located in large city or state capital- results may not simply be representative of people with a tie to the university

ACC:
Clemson: 48-42 Trump
Miami**: 58-38 Clinton
Notre Dame: 64-28 Clinton
Virginia Tech: 64-27 Clinton
NC State**: 73-21 Clinton
Louisville**: 73-20 Clinton
Pittsburgh**: 77-18 Clinton
Florida State**: 77-17 Clinton
Boston College**: 78-15 Clinton
Virginia: 79-14 Clinton
North Carolina: 82-12 Clinton
Georgia Tech**: 85-11 Clinton
Syracuse: 85-10 Clinton
Duke**: 88-8 Clinton

Big 12:
Texas Tech: 47-45 Trump
Oklahoma State: 45-44 Clinton
West Virginia: 51-39 Clinton
Kansas State: 56-33 Clinton
Iowa State: 58-31 Clinton
Oklahoma: 60-30 Clinton
Baylor**: 66-25 Clinton
Kansas: 75-18 Clinton
Texas**: 79-14 Clinton

Big Ten:
Purdue: 56-34 Clinton
Nebraska**: 61-30 Clinton
Penn State: 64-29 Clinton
Michigan State: 69-24 Clinton
Iowa**: 73-20 Clinton
Ohio State**: 75-18 Clinton
Indiana: 76-17 Clinton
Wisconsin**: 77-15 Clinton
Minnesota**: 77-14 Clinton
Rutgers**: 81-15 Clinton
Illinois: 79-13 Clinton
Maryland**: 80-13 Clinton
Michigan: 82-12 Clinton
Northwestern**: 87-7 Clinton


Pac-12:
Arizona State: 60-30 Clinton
Washington State: 69-18 Clinton
Utah**: 68-14 Clinton
Oregon State: 71-16 Clinton
Arizona**: 77-15 Clinton
UCLA**: 79-16 Clinton
Colorado**: 79-13 Clinton
Oregon: 80-9 Clinton
Stanford**: 83-11 Clinton
Washington**: 83-7 Clinton
USC**: 88-7 Clinton
Cal**: 88-4 Clinton (Trump only very narrowly edges Jill Stein)

SEC:
Auburn: 69-28 Trump
Alabama: 58-35 Trump
Ole Miss: 50-46 Trump
Texas A&M: 45-44 Trump
Mississippi State: 48-48 Clinton (0.19 margin)
Tennessee**: 56-35 Clinton
Kentucky**: 62-29 Clinton
Arkansas: 62-26 Clinton
Missouri: 65-27 Clinton
Georgia: 66-28 Clinton
South Carolina**: 66-26 Clinton
LSU**: 67-24 Clinton
Florida**: 70-23 Clinton
Vanderbilt**: 73-19 Clinton

So, Trump won 6 P5 campus neighborhoods (Auburn, Alabama, Clemson, Ole Miss, Texas Tech, and Texas A&M, in order of margin).  I still think grad students and professors/academics certainly lean more to the left compared to undergrads, so he likely won undergrads at far more schools.  I can provide G5 numbers later (I want to get back to watching games now).  I wonder what to make of Alabama and Auburn's neighborhoods being significantly more Republican than literally anywhere else.  Do they encourage a greater percentage of undergrads to vote at school in comparison to at home somehow, or is it something else. (I wouldn't expect Alabama/Auburn to be significantly different from Ole Miss/Mississippi State- if anything I would expect the MS schools to be more GOP because they attract less from out-of-state.  Do the campuses have significantly different racial demographics?)

[100% pro-life no matter what]

SEC- EAST

Vanderbilt University:

Nashville, Tennessee:

138,931 HRC (60.9% D), 75,853 Trump (33.3% R)             + 27.6% D

Ok---- Nashville was initially tricky because of whole Joint City-Davidson County mayoral jurisdiction thing, which I'm not used to....

So I defaulted to what is now considered to be the standard municipal boundaries of the City of Nashville, that essentially includes the entire County, with the exception of six municipalities: (Ridgetop, Goodlettsville, Berry Hill, Belle Meade, Oak Hill, and Forest Hills)....

I backed out the the results from these places from the total County vote total, as well as the 6.5k absentee votes not coded by precinct, and the 1.3k Write-In votes coded to minor party write-in candidates....

If rural and Uninc parts of Davidson County want to be part of the City of Nashville, so be it, and good for them. since it gives them access to resources that would otherwise would not be available, and also helps support the tax base of the City....

Obviously Nashville is a heck of a lot more than just Vanderbilt, and overall the Metro area is now some 1.6 Million people, and the largest in the State, with about 600k residing within the municipal boundaries of Nashville....

Ok, how did Vanderbilt Student Precincts Vote?

Vanderbilt University Campus:

Precinct 18-01:

113 HRC (83.7% D)-----    12 Trump (8.9% R)     +74.8% D

Ok---- apparently the vast majority of students don't live in on campus dorms....

So where is the student off-campus housing?

Precinct 21-11: (90% 18-21 Yrs) roughly located with the River to North, Ed Temple Blvd to the East, Albion St to the South and 39th Ave N to the West:

1,254 HRC (97.1% D), 15 Trump (1.2%), Others (1.7%)                   + 95.9% D

Ok---- so these numbers were a bit astounding and we look at Census Tract Tract 013602 and see that it is 77% African-American and 17% White... compare that against the overall undergraduate admission and State enrollment numbers....

https://admissions.vanderbilt.edu/profile/#undergradstudentpopulation

https://admissions.vanderbilt.edu/profile/#enrollmentbystate


Precinct 21-4: (80% 18-19 yrs)   Located with Hinkle Drive to the East, West End Avenue to the South, I-40 to the West, and Charlotte Pike to the North.

Appears to be about 85%+ White:

932 HRC (62.1% D), 412 Trump (27.4% R)                              + 34,7% D

Precinct 21-3: located with I-40 W to the North , 18th- 20th Ave to the East, Charlotte Pike to the South, and I-40 to the West....

Appears to be about 62% White and 19% African-American....

892 HRC (89.0% HRC), 62 Trump (6.2% R)                             + 82.8% D


So what I suspect that we are seeing is that many of the White College students at Vanderbilt come from out of State, and out of region, and vote absentee in their home states (As I did as an undergrad a few decades back going to College in other State)....

Still it's pretty clear that White Students at Vanderbilt voted to the Left of the City as a whole, and significantly more Democratic than both the White population within the City, Tennessee, and Whites within the United States of America....

Overall students at Vanderbilt most likely voted more Democratic than in any other SEC-EAST campus, other than the University of Georgia, or possibly the University of Florida....

Actually, over 95% of Vandy students live on campus.  The only explanation for precinct 18-01 is that very few actually vote at their precinct address (it's worth noting that TN requires a TN ID or a US Passport (which legally also means that you can't drive on an out-of-state license).  Plus, there are a significant number of live-in faculty members on Vandy's campus that would likely vote in precinct 18-01 as well (it also includes some areas to the west of campus).

Vandy is also the exact sort of campus that someone like Rubio would have done significantly better than Trump.  I would guess that Vandy was probably about 55-30 Clinton with a strong third party vote this year.  But, Rubio might have won undergraduates.

[100% pro-life no matter what]

As for Baylor, I think that may be a case of a low vote total for a private university clouding the results a bit.  Faculty, while likely more conservative at Baylor than other private universities, may live on campus at surprisingly high rates.  Especially if it includes a bit of off-campus, grad students may be a significant component too.

I have actually been thinking about why the University of Alabama looks more in line with what I would think possible than most other campuses.  Simply, undergraduate students are significantly more likely to vote on campus at Alabama because of a group known as 'The Machine' that tries to control local politics for the interests of the student body.  That probably makes the on-campus vote somewhat more representative of the undergrad population and, therefore, likely more Republican (using things like the breakdown of college-educated millennials in 2012, 2016 young voter exit polling (closer than 2012), and campus mock elections that differ wildly from the precincts (especially at LSU) as reasons for that hypothesis.

I would need more time to look at this theory, but I also wonder if, at most campuses, undergraduates who vote on campus lean somewhat more Democratic than those who do not.  At least at my campus, the College Dems are the only group that I have ever seen push for on-campus registration, and it is possible that Democrats might be more interested in the local races of Democratic-leaning college towns or big cities, while Republicans may be more interested in races in the more Republican-leaning areas of back home.

As for South Alabama, those numbers seem quite odd, especially relative to the other campus areas in Alabama.  Those numbers are so far Democratic that Hillary would even be having to win the white vote there to get those numbers, which I can't fathom at all (I would not expect South Alabama students to be demographically different from Alabama, Auburn, or Troy- if anything, they might be more in-state than Alabama or Auburn).

The DDHQ numbers for South Alabama (everything within one mile of campus) are:
Trump- 56.8%
Clinton- 39.8%

Obviously, that would include grad students and professors too (but they are more of the academic nature and every data shows that people with post-grad degrees are significantly more liberal than people with undergrad degrees, not to mention college professors).  I think, with that, it is safe to say that Trump won South Alabama undergrads.  Sometimes campus precincts tell a very misleading tale (I read somewhere that Trump apparently won Pitt's campus precinct, which I highly doubt was true), but I still do think this project is fun.

[100% pro-life no matter what]

NOVA Green, you might want to also look at which schools have an on-campus polling place, because those might have a higher proportion of students vote on campus.  I still have my questions about the accuracy of on-campus precincts (with the points I mentioned upthread), but this is a really interesting project.

Also, I think 2012 numbers would be really interesting, because, using my Facebook feed as a judge (which is like 80-85% Republican and mostly consists of ideological conservatives between 18 and 25 from well-off to wealthy families around the country who attend(ed) these sorts of major universities), probably a quarter of the conservatives didn't vote or voted third-party or write-in as a protest.  I think Rubio probably does 15 points better (perhaps even more) than Trump against Hillary nationwide with these groups.  But, my finding of no gap between 2012 and 2016 in campus precincts in Tennessee puts either this theory or campus precincts into question (but since few students in Tennessee seem to vote on campus, perhaps the Midwest would be a better test of this).