Political Geography: Advantage Hillary

Over the last week or so, I have been looking at the New Hampshire primary and Iowa caucus results to see if there were any interesting voting patterns in terms of how different areas of each state voted.  Perhaps the most noticeable pattern that emerged is how in both New Hampshire and Iowa the geographic areas which are often described as "red state" areas had a relatively high correlation with voting for Hillary Clinton in the Democratic primary.  Many, many people on this site and others keep repeating that Hillary "just can’t win" in "red" America.  And yet, that is exactly what she’s doing.

The nice thing about looking at these results from New Hampshire and Iowa is that in both states large numbers of independents and other non-Democrats participated in the Democratic primary/caucus.  This is particularly true for New Hampshire, where 46-48% of those who voted in the Democratic primary were independents, Republicans or unregistered voters.  Only 52-54% of New Hampshire Democratic primary voters were actual Democrats.  In Iowa, the proportion of independents and non-Democrats was smaller, though still very significant.

Furthermore, in both states the number of persons who participated in the Democratic primary/caucus was significantly larger than the number who participated in the Republican primary/caucus.  This is true despite the fact that both states have in recent history been swing states with voting registration split evenly between the two major parties.  In Iowa, there were literally twice as many persons participating in the Democratic caucus as there were those who took part in the Republican caucus.  In New Hampshire, the numbers were not as lopsided, although still clearly leaning towards the Democratic side.

Therefore, the persons who voted in each primary were not just partisan Democrats, but a relatively representative population sample of the area in which they voted.  

Looking at exit/entrance polls from Iowa and New Hampshire, you can see that independents voted at a higher rate for Barack Obama than for Clinton.  This is particularly true in Iowa.  In New Hampshire, more independents voted for Obama than for Hillary, but the difference was not as great – approximately 50,000 independents voted for Obama, and approximately 40,000 voted for Hillary.  I did a post on this a few days ago -  http://www.dailykos.com/...
(As I noted in that post, it’s interesting to note that more independents in New Hampshire voted for Hillary than for John McCain.)   However, these numbers do not provide us with the full picture of things.  For one thing, they do not distinguish between Democratic- or progressive-leaning independents and Republican- or conservative-leaning independents.  One needs to further study other factors in the exit/entrance polls and to look at these results through a geospatial analysis of voting patterns.  

NEW HAMPSHIRE:

I did a number of maps for New Hampshire – voting by county, congressional district, and town (please scroll down to see maps).   MAP 1 compares the 2000 and 2004 Democratic primary results to the results from this year.  As you can see there’s a decent correlation between "Howard Dean" areas voting for Obama and areas that went for John Kerry in the 2004 primary voting for Hillary.  There’s an even higher correlation, though, between "Bill Bradley" areas voting for Obama, and "Al Gore" areas voting for Hillary.

Looking at MAP 2 and MAP 3 you can see that "red" areas of New Hampshire (those that voted for George Bush in the 2004 general election) tended to go for Hillary in the primary last Tuesday.  Out of the 4 counties that voted for Bush in the general, 3 voted for Hillary in the primary.  Out of the 6 counties that voted for Kerry in the general, 4 voted for Obama in the primary.  Likewise, the 1st Congressional District voted for Bush in the general and Hillary in the primary; the 2nd Congressional District voted for Kerry in the general and Obama in the primary.  

The voting maps by town are perhaps more interesting.  You can get town by town results here: http://www.cnn.com/...

Overall, 127 towns and voting areas voted for Obama and 102 voted for Hillary (8 were tied or voted for other candidates; there are 234 towns in New Hampshire and 3 other voting areas, but for the purpose of this diary I will refer to all 237 entities as "towns").   In 2004 Kerry carried the state, but Bush still managed to win 131 towns.  Out of these 131 "red towns" Hillary carried 72 and Obama 59 (see MAP 4).  Out of the 98 "blue towns" Obama carried 68 and Hillary only 30.  However, these statistics still do not give us a true picture of the red/blue divide in New Hampshire because the population size of the towns varies tremendously.  

I correlated the statistics from MAP 4 and MAP 5 to get a good picture of partisan preferences for the different towns.  As you can see in MAP 4, the towns colored in blue are ones which Kerry won in the 2004 general election, and which Obama won in the 2008 primary.  These towns comprise approximately 22% of the population of New Hampshire, and include the state capital, Concord, as well as Portsmouth, Exeter, Keene, Lebanon, and the university towns of Hanover and Durham.   Many of these towns are in the Connecticut Valley, along the Vermont border.  Much of this area is rural, but ironically, in New Hampshire many rural areas tend to be more progressive than the urban/suburban areas.  

Towns colored in green are ones which Bush won in the 2004 general election, and which Obama won in the 2008 primary.  These towns comprise only about 12% of the population of the state and are mostly very small in character.  The largest town in this subgroup is Amherst, with a little over 10,000 population.  As you can see, when adjusted for population, towns which Obama carried lean Democratic over Republican by almost a 2:1 ratio.

Towns colored in yellow are ones which Kerry won in the 2004 general election, and which Hillary won in the 2008 primary.  These towns comprise approximately 21% of the population of New Hampshire, and include Rochester, Nashua, Dover, Hampton and Berlin.  Many are blue-collar industrial and post-industrial areas.  According to exit polling, Hillary won heavily in New Hampshire among those with no college degree (she also won among those with non-postgraduate college degrees, albeit more narrowly; Obama won only those with a postgraduate degree) and she won among union members.  Hillary won among the 58% of state voters who are "very worried" about the economy, while Obama won among the 40% who are only "somewhat worried."  As you can tell by now, the towns carried by Kerry in the 2004 general election, spilt almost 50/50 between Obama and Hillary (when adjusted for population).

Perhaps the most important (politically speaking in terms of our general election prospects) are the towns colored in red.  These towns are ones which Bush won in 2004, but which Hillary won in the 2008 primary.  These towns comprise approximately 43% of the population of New Hampshire.  They include Manchester and its suburbs and most towns in the Merrimack valley (Londonderry, Derry, Hudson, Salem, etc. as well as Laconia further north in the state.)  Many towns here are historical mill towns; blue collar areas where "Reagan Democrats" and "Reagan independents" predominate to this day.  Even as Kerry was carrying the state in 2004, these areas still voted Republican.  New Hampshire primary exit polling also confirms that Hillary carried these "Reagan Democrat" areas: http://www.cnn.com/...

Among Catholics (35% of the state’s population, and perhaps the most representative sample of "Reagan Democrats" and "Reagan independents") Hillary won with 44%, compared to 27% for Obama.  

It should be noted that among the areas that voted for Bush in the 2004 general, approximately 78% (when adjusted for population) went for Hillary in the primary.  The reason this is a very important net positive for Democrats -- and specifically for Hillary Clinton -- is because the already "blue" areas (those largely carried by Obama in the primary) will vote Democratic in November no matter who the nominee is, while the "red" areas are more likely to vote for the Democratic nominee IF that nominee happens to be Hillary, for whom they voted overwhelmingly in the primary (remember, almost half of the Democratic primary voters were not Democrats).  This is what I mean by "Advantage Hillary" in the title of this post.

By handily holding Democratic voters in the camp and expanding the Democratic universe to moderate- and conservative-leaning independents (not just to progressive-leaning independents) Hillary may be well positioned for the general election, if she indeed becomes our nominee.

New Hampshire maps:

IOWA:

You may think that, with Hillary coming in 3rd in Iowa, it would be hard to deduce anything positive for her from the Iowa results.  I was looking at the entrance polls though and one thing popped out right away – she actually won rural voters in the state:
http://www.cnn.com/...

Unlike New Hampshire, whose "rural" voters are kind of a political anomaly (much like rural voters in neighboring Vermont), rural voters in Iowa are more typical of the country, and in particular of "red state" areas.  You can see the Iowa caucus winners by county on MAP 6 (please scroll down to see maps).  MAP 7 provides perhaps a better gauge of the results as you can see 1st, 2nd and 3rd place rankings as well.  Despite coming in third in the state, Hillary actually managed to win the western 5th Congressional District.  This is really the only truly Republican district in the state (60% for Bush in 2004 – see MAP 10).

Like the New Hampshire data, I tried to correlate the Democratic caucus numbers (MAPS 6 and 7) with the 2004 general election numbers (MAP 8).  The result is MAP 9.  As you can see, the state was quite polarized when it voted on January 3.  The heavily Democratic areas in eastern Iowa (areas in dark blue on the map -- Cedar Rapids, Iowa City, Davenport, Dubuque, Waterloo) and the urban/suburban parts of central Iowa (Des Moines, Ames/Story Co.) voted very heavily for Obama over Clinton and Edwards.   On the other hand areas in western, and parts of northern and southern Iowa were more evenly split, with a clear advantage for Hillary in some areas, and an advantage for Edwards in others.  In many of  these counties, Obama finished in 3rd place (or in the case of Ringgold Co., in 4th place).  You can get actual results here:  http://www.cnn.com/...
or here: http://data.desmoinesregister.com/...
MAP 11 provides another interesting way to look at the results.  The map kind of speaks for itself as I added handy stats, which you can interpret on your own.

What’s quite interesting also in the Iowa results is that out of the 29 Iowa counties that voted for Edwards, in only 10 did Obama come in 2nd and Hillary 3rd, while in 18 Hillary came in 2nd and Obama was 3rd or 4th (in one county which Edwards won, Hillary and Obama tied for second place).  This means that, contrary to popular opinion, most Edwards voters may have gone to Hillary and not to Obama if Edwards was not running or withdrew.   In either case, it looks like in Iowa Hillary and Edwards split much of the rural vote, while Obama won big in the more Democratic and "progressive-independent" areas, with Hillary probably doing even better than Edwards among the "red state" voters which may prove important in this and other states in November.

Iowa maps:

•  post a tip jar already, it's a great diary

  very enlightening, thank you

  by Jim J on Sat Jan 12, 2008 at 05:42:54 PM PDT

•  Even though I back Obama

  I gotta rec this diary. I love political maps, and your analysis is deserving.

  by Nathaniel Ament Stone on Sat Jan 12, 2008 at 05:43:19 PM PDT

•  Looking at the Democratic primary results is a

  rather odd way to determine electability in those states.

  Polls in those states (since every candidate in both parties has already campaigned there heavily and the voters know them) are likely more meaningful.

  For instance, Iowa

  which shows Obama doing 20 points better in margin than Hillary against Giuliani, Romney, Huckabee and McCain.

  by DemocraticLuntz on Sat Jan 12, 2008 at 05:43:32 PM PDT

  •  after what happened with NH polls ...

    please don't tell me polls are more reliable than actual voting data.

    by silver spring on Sat Jan 12, 2008 at 06:13:39 PM PDT

•  Can you really

  compare Democratic Primary votes, which are a very small percentage of the electorate to general election votes which are a much broader and much larger statistical sample?

  What your studies seem to say is that conservative Democrats in rural districs who vote in primaries supported Hillary. That makes sense. That really says nothing about how well she would do in the broader electorate. It also says that people who are low income low information voters in primaries supported Hillary. That makes sense too. Says nothing, however, about how she would fare with the broader electorate.

  by brooklynbadboy on Sat Jan 12, 2008 at 05:47:14 PM PDT

  •  I would totally agree with you, except ...

    in the New Hampshire "Democratic" primary almost half of the voters were independents and non-Democrats.

    in Iowa, likewise, huge numbers of independents participated in the "Democratic" caucus -- whereby the Dem. caucus had literally twice as many people as the Rep. caucus (even though the state is almost 50/50 in terms of voter registration and historical voting patterns).

    in both states, RECORD numbers participated -- Democrats and independents/non-Democrats alike.

    by silver spring on Sat Jan 12, 2008 at 06:03:22 PM PDT

  •  In some respects it does ...

    Turnout was high all around, but it will be especially crucial in the fall to turn out working class Dems and Dem-leaning indies (quite a different group from the upscale indies the media loves), especially in pinkish states.

    The vote Obama mobilizes is in bluer areas, and in the fall in bluer states - where a big chunk of the upper middle class is pretty bluish. But pushing up our margin in upscale CA burbs isn't the most critical electoral battleground.

    by al Fubar on Sat Jan 12, 2008 at 06:33:44 PM PDT

•  Man, that's a lot of work

  for a truly meaningless conclusion.  Winning in  a Dem caucus and primary (albeit open ones}has absolutely nothing to do with winning against a Republican in a general election.

  Nice try though.

  Florida Kossacks Rock

  Blog Florida Blue

  by gatordem on Sat Jan 12, 2008 at 05:51:59 PM PDT

  •  Except

    You can draw a comparison with the primary and the last GE tallys, too.

    by justinb on Sat Jan 12, 2008 at 06:08:42 PM PDT

•  That's a whole lotta work for ... what?

  The hole in your donut is that Democrats in rural areas aren't representative of swing voters.  Those who voted for Hillary are likely to vote for most Democrats anyway so turning them out doesn't yield any more than turning out more urban Democrats, for example.

  On the other hand, Obama does actually draw swing voters (think more moderate independents and Republicans from more urban/suburban areas of largely rural states).

  I think you're committing a "compositional" fallacy.

  by chicago jeff on Sat Jan 12, 2008 at 06:00:58 PM PDT

  •  based on the patterns I saw ...

    it seems that Obama draws more "progressive-leaning" independents while Hillary draws more "moderate-leaning" and "conservative-leaning" independents.

    This is based not just on the geographical data, but on exit/entrance polling as well.  The two correlate to each other very closely.  

    In Iowa, for example, Hillary won the rural vote, even though she came in 3rd overall.  The more conservative, rural voters we're talking about are not just Democrats but independents as well (who cares about rural Republicans; they will not vote  Dem. no matter who our nominee is).

    In New Hampshire, it is the moderate/conservative Democrats AND independents she attracted in urban and suburban areas.

    by silver spring on Sat Jan 12, 2008 at 06:11:36 PM PDT

    •  I'll take your word on that one...

      ...but it's insane! My sense is that Obama should be getting the more moderate/conservative ones, because he's reaching out to them, whereas Clinton should be getting the "progressive" ones if by progressive you mean liberal, since she's much more the traditional liberal of the two.  And I say that as a Clinton supporter, so it's a "statement against interest" as the lawyers say--there are, after all, probably more moderate/conservative independents than progressive ones.

      If there's one thing I have learned in years of formal and informal study of people, it's that they don't know their own minds! At least Americans don't seem to.  

      by Rich in PA on Sat Jan 12, 2008 at 06:15:52 PM PDT

      •  This is in the media-verse ...

        Working class Dems tend to be more conservative on social and defense type issues and overall tone, but they are programmatic liberals. Hillary's policy riffs hand out the program candy. Upper middle class people, which is most of DKos and certainly the MSM, tune this stuff right out because it is boring.

        Think of choosing a teacher for your kid. One is young, the kids love him, and he gives great lecture. The other is kind of schoolmarmish, and she makes the kids work their ass off, but they come out knowing their basics.

        If you're an upper middle class parent you probably opt for the first teacher - your kids are doing okay, and the young teacher can excite and stimulate them to push farther. If you're working class you probably opt for the schoolmarm, because you see too many kids falling by the wayside, and hers usually don't.

        Analogies R Us - worth what you paid!

        by al Fubar on Sat Jan 12, 2008 at 06:44:13 PM PDT

      •  my gut feeling would also be...

        that Obama would attract the moderate/conservative independents ... but just don't see that in the results, but in fact, the opposite.  This is particularly clear in NH (the Iowa data was harder to gauge, as there were 3 strong candidates).
        Across New Hampshire, Democrats equal to only about 30% of the vote and Republicans equal to only about 30%.  Independents are the largest voting block at approximately 40% of the vote, and more so in many areas.  These numbers hold true for the general election, and independents therefore determine the political nature of an area.  Normally, a geographic area is therefore either "red-leaning" in the general election if that area’s independents lean "red" or "blue-leaning" in the general if that area’s independents lean blue.  As go the independents, goes the area.  There are exceptions, ofcourse.  An area that is 50% blue (instead of the "standard" 30%) will go blue even if 90% of independents go red.  However, as New Hampshire is rather monolithic demographically (especially in terms of race), the discrepancies are not that great throughout the state (like they would be in parts of the south for example where 90% of African-Americans would vote Democratic and 75% of the whites would vote Republican and where there would be great differences within short geographic distances in how people vote).  So in most cases, the independents will determine the political nature of an area in NH.  These same independents voted in the Democratic primary in New Hampshire (to the tune of almost 50% of the "Democratic" primary) and it looks like the independents in "red-leaning" areas (which are "red-leaning" largely because of these same independents) have now overwhelmingly voted for Hillary in the primary.  In other words, Hillary has attracted the support of "red-leaning" independents, who, if she becomes the nominee, are now more likely to stick with her in the general.  Granted, many or most may not; but the ones who do may constitute her margin of victory.  The other thing about NH is that there are still many "Reagan-Democrats" especially around Manchester and the Merrimack valley – areas which have voted Republican in the general.  These   Democrats constitute the largest voting block in some areas.  Based on the data, they are now attracted to Hillary also (as the areas in which they form a huge proportion of the Democratic vote have overwhelmingly voted for Hillary), and like the independents discussed above, may decide to stick with her in the general.

        (so much for gut feelings, btw)

        by silver spring on Sat Jan 12, 2008 at 08:48:19 PM PDT

•  Ethical Geography.

  Advantage Edwards and Obama

  by Morgan Sandlin on Sat Jan 12, 2008 at 06:02:30 PM PDT

  •  what brought you to that conclusion?

    i'm giving you the benefit of the doubt and assuming that you're responding to something written in the diary and that you're not just being a complete douchebag.

    by CC Music Factory on Sat Jan 12, 2008 at 06:18:40 PM PDT

•  facts won't sway many Obama supporters --

  well-informed and reasonable doesn't have the same zing as fired up and ready to go.

  by CC Music Factory on Sat Jan 12, 2008 at 06:21:35 PM PDT

  •  or fairy tale

    by dcrolg on Sat Jan 12, 2008 at 07:17:24 PM PDT