Every Demographic is a Swing Demographic

by Chris Bowers, Wed Feb 16, 2005 at 11:45:37 AM EST

In order to present a crude snapshot of the current political landscape, here are how around fifty demographic groups stack up against each other in terms of partisan self-identification (source). The degree to which these groups shifted from the 2000 cycle is also indicated.

Democratic Self-Identification Margin Among Select Groups
			 2004	 Change from 2000
African-American Women	  +71	      +4
African-American Men	  +54	      -2
Latino Women		  +29	       0
Union Women		  +27	      +3
Urban Women		  +23	      -1
Women under $35K		  +21	       0
Northeast Women 		  +20	      +6
Women 18-29		  +15	      +6
Union Men			  +14	      +4
Women HS or less		  +14	      -2
Latino Men		  +13	      -7
Women 45-64		  +11	      -3
Men below $35K		  +10	       0
Women 65+			  +10	      -3
Midwest Women		   +9	      +1
Urban Men			   +7	      +2
West Women		   +7	      -2
Suburban Women		   +5	       0
South Women		   +5	      -6
Women 30-44		   +4	      -4
Rural Women		   +3	      -5
Men HS or less		   +2	      -5
Men 65+ 			   +1	      -1
Northeast Men		    0	      -3
White Women		   -1	      -3
Women $75K+		   -2	       0 
Men 18-29			   -3	      +5
Women Some College		   -3	      +2
Men 45-64			   -3	      -2
Midwest Men		   -5	       0
Union Married		   -5	      -3
Women $35K-75K		   -6	      -1
West Men			   -7	      +3
Men $35K-75K		   -7	      +1
South Men			   -7	      -4
Men Some College		   -8	      +2
Rural Men			   -8	      -4
Suburban Men		  -10	      -2
Women College or More	  -10	      -3
Men Some College or More      -11	      +4
Men 30-44			  -11	      -2
White Men			  -13	      -2
Men $75K+			  -15	      +1
White Born Again Women	  -22	     -10
White Born Again Men	  -29	      -7

The point of this table is to demonstrate that every demographic has the potential to be a swing demographic, not just those that are closely divided between Democrats and Republicans. To prove this, I need to no more than list the groups where the biggest shifts from 2000 to 2004 occurred:

Groups with largest swings, 2000-2004
			 Shift
White Born Again Women	  -10
White Born Again Men	   -7
Latino Men		   -7
Northeast Women 		   +6
Women 18-29		   +6
South Women		   -6
Men 18-29			   +5
Rural Women		   -5
Men HS or less		   -5

The degree to which these groups support one party or the other in 2000 shows no pattern at all. Some were closely divided, others were strongly pro-Republican, while still others were strongly pro-Democratic. Further, the way each of these groups broke shows no pattern. For example, northeastern women, who were already strongly pro-Democratic, became even more so, while Latino men, who were strongly pro-Democratic, became less so. In fact, the number one swing group in the 2004 election were white, born again / evangelical protestants, who were already the strongest pro-Republican demographic in 2000. This is something pundits have failed to realize. It was not just that white born-again / evangelicals voted in large numbers, but that their already strong support for Bush shifted even more dramatically in his favor--a shift larger than that experienced by any other demographic group.

The only relationship between these diverse groups is that they swung heavily in favor of one party or the other. It is important that we realize that swing groups are not defined by how much they support one party or the other, but by how much their support is prone to shift between parties. Republicans won in 2004 because six of the nine largest shifts, including the three largest, went in their favor.

In order to win in the future, we need to identify not where we are getting beat the worst, but where we have the potential to cause the largest shift. Considering long-term history of the south, which has experienced several of the most dramatic partisan index shifts of any demographic ever, it does not surprise me that many in the Democratic Party are still fixated on how to win in "the south." It is not a bad instinct, considering history, but as I have written in the past, I for one no longer believe that such dramatic shifts are possible in that region. From a regional standpoint, the West is actually doing a lot more swinging than the South anyway. Other than that, I am not sure what future swing groups will be. What are your ideas?

Tags: Demographics (all tags)

Comments

9 Comments

Sort by

permalink

labels, oh my!

so how many of these sub-demographics do we fit into? there's got to be some serious overlap there.

let's see, i would be: southern women, suburban women, women some college or more... sheesh, i think my head is spinning!

chris, as usual, i agree with you (get out of my head!). i think the west has potential to swing, but i admit i say this as someone who's never lived in the region so i may be way off-base with my guess. but just looking at the numbers from past posts, i think that the mountain west and southwest are ripe targes for the dems. and i do wish we'd do more to reach out to the growing latino population (thanks, simon rosenberg!). while i don't expect them to swing big, i definitely think their interests are more in line with what the dems stand for.

by annatopia 2005-02-16 11:51AM | 0 recs

permalink

Re: labels, oh my!

I fit: Men 30-44 (ugh), Urban Men, Northeast Men, Men under $35K, white men and union men.

It is kind of crazy, but at least it isn't as cloying as Greenberg's groups (Golden Girls? Young the the Restless? Excuse me while I gag).

by Chris Bowers 2005-02-16 12:16PM | 0 recs

permalink

If I could get my hands on the raw data...

...I could figure out how much "weight" to apply to each variable, and estimate the likelihood that a voter will consider him/herself Republican or Democratic given his/her constellation of demogaphic variables. For example, "female" would have a certain weight, "southern" would have a certain weight, "suburban" would have a certain weight, etc. These weights would be determined statistically via logistic regression or similar methodology.

by KTinOhio 2005-02-16 12:40PM | 0 recs

permalink

Re: If I could get my hands on the raw data...

now that would be interesting!

by annatopia 2005-02-16 01:45PM | 0 recs

permalink

Re: If I could get my hands on the raw data...

That would be--like a blogopshere data mining company. Why don't you ask NAES for the raw data?

by Chris Bowers 2005-02-16 04:08PM | 0 recs

permalink

Maybe security moms don't exist after all?

These data really shook up one of my core assumptions about 2004: that the so-called security moms really hurt us... that is, northeastern suburban women who, post 9-11, gravitated more toward the "strong daddy" posture of the GOP, and that seemed borne out by that some of the biggest percentage drops from 2000 to 2004 were in NJ, NY, CT, and RI. But now it turns out that northeastern women moved +6? Wow. So "security mom" seems pretty much a fiction. Or maybe (since southern women moved -6) they were just limited to the suburbs of Atlanta and Dallas, places which ironically seem a lot more out of the line of fire?

Anyway, looking at the movement of born-agains, I think it's clearer than ever that the main lesson we should take from 2004 is: if you're fighting a war of inches, focus on riling up the base and actually getting them to the polls (as the GOP did with the evangelicals) instead of worrying about the swing groups that may or may not even exist (like the "security moms")!

by Crazy Vaclav 2005-02-16 03:33PM | 0 recs

permalink

Women

I think it's interesting that Women 75k+ are more Democratic than Women 35k to 75k. Any ideas on why that is?

Both men and women in that income group tend to be Republican. Any way we could swing them our way?

As a man aged 30-44, it's interesting that we are much more pro-Republican than men of other age groups. Note that we were born between 1960 and 1974, so we came of age between roughly 1978 and 1992 (just in time to watch Carter screw up, not too late to see us win Gulf War I) -- the Reagan Generation.

I have no idea what happened to me. I must have been dropped on my head when I was a baby.

by tgeraghty 2005-02-16 05:34PM | 0 recs

permalink

Re: Women

If you think of it like this its more clear.

Republicans have a networth of $10,000 to 10 mill (I am picking numbers out of the sky).

Everyone with a lower networth or higher networth will tend democratic. The poor because Dems are the only ones who will help them, the rich because 10 mill and 9 mill isn't much difference once you are that rich you don't notice 5-10%.

Also the very rich actually understand that too little tax is a bad thing. That is why the Warren Buffets of the world actually want more tax.

by donkeykong 2005-02-17 02:37PM | 0 recs

permalink

Demographics

Just to make sure that I understand the list above--the number on the left means the percentage advantage/disadvantage for the party. So African-American women +71 would mean that that sub-group identifies Dem by something like 85.5 to 14.5, correct? I think that the voters who are more prone to swing from candidate to candidate and party to party are those focusing on a few or even one issue, particularly if the issue is not in the bedrock of the platform. These are the groups that need to be sought out and courted. I think that evangelicals are committed to the Repugs for now, but a real scandal could disaffect them greatly, however there is not much that the Democrats can do, other than wait and point out potential scandals. The Democratic party should look west--hold on to the coast, despite Arnold, and win control of Nevada, Colorado, New Mexico, and perhaps Arizona and pick up some high-profile wins in states like Montana.

by flatblade 2005-02-17 02:22AM | 0 recs