Sunday, August 1, 2010

Why is Tucson Liberal?

As we know, Arizona's two large metro areas occupy very different political territory. That's not exactly shocking to anyone. What's strange, I think, is the lack of a general understanding as to why that's the case. Explanations I've heard for Tucson's liberalism range from "more Latinos" or "socially liberal whites" to "it's just Tucson". I spent some time looking at Census data to see if I could begin to come up with a data-driven explanation.

Phoenix vs. Tucson
I started by just comparing the core cities. Tucson looks significantly more hospitable to Democrats than Phoenix in one demographic area - women. Tucson is 50.5% female (48.6% in Phoenix) and more of those women are single (34.7% to 30.9%). In a couple of other categories, Tucson actually looks like less friendly territory. Phoenix is both younger (median age 31.9 to 33.2) and more Latino (42.1% to 39.5%). Pulling back to the county level doesn't make much difference. Maricopa makes up some of the gender gap and widens the age gap (33.9 median age to 37.0 in Pima). The Phoenix suburbs pull down Maricopa's Latino numbers, allowing Pima to take a 32.7% to 30.3% edge. Big picture, though, the two counties look far too similar to be so different politically.

So what's the answer? I think it's two-fold.

First, Latinos in Tucson/Pima score much better on assimilation. Non-citizens are only 8.2% of Pima County (10.9% in Tucson). Compare that to 12.4% in Maricopa (18.5% in Phoenix). 8.2% is also the figure for Pima residents (11% in Tucson) who primarily speak Spanish and are NOT fluent in English. In Maricopa, that's 11.6% (18.1% in Phoenix). So while Pima may only have a slight edge in Latinos as a percentage of the population, they probably have a larger edge in Latinos who are likely voters.

Secondly, Tucson/Pima has a much larger lower-middle class than Phoenix/Maricopa. Median household income is $37,396 to $49,933 and $46,229 to $56,555 in the counties. Roughly 44% of Maricopa families (37% in Phoenix) have incomes above $75,000. In Pima, that's more like 37% (27% in Tucson). In 2008, Obama won AZ voters under $50,000 by 3 points. He lost the over $50,000 crowd by 15 points (both figures are nearly identical to the 2004 results). In 2008, that disparity was almost entirely evidenced among non-whites. Non-whites over $50,000 gave Obama only a 2-point edge, while non-whites under $50,000 chose him by an overwhelming 52-point margin. Unfortunately, CNN didn't publish exit poll results by both race and income until 2008, so we can't compare to past elections.

One last note on economics: Pima has a significantly higher proportion of its workforce in government (18.2% to 11.4%) and in the "education/healthcare/social work" fields (23.2% to 17.4%). I've never seen party preference by occupation, but I'm going to hazard a guess that both of those are big boons to Pima Democrats.

There's never going to be ONE explanation for why Tucson/Pima is more liberal than Phoenix/Maricopa. A number of minor factors - slightly higher proportions of women, postgrads, and Latinos - all likely contribute to Tucson's liberalism. Making an educated guess here, I'm suggesting that three of the major factors driving Tucson to the left are: 1) a more efficacious Latino population, 2) a significantly larger under-$50k family population, especially among non-whites, and 3) a larger share of their workforce in government and social services.


ワイリー said...

I had heard some casual remarks that Tucson was becoming less and less liberal. Do you think there's truth to that?

-- wy

Alton Brooks Parker said...

I looked for evidence of this in two places. First, I compared Pima County's presidential results to the national results from 1988-2004 (2008 is spoiled by home state effects for McCain). A positive score indicates that the Democratic margin (%) was larger in Pima County than nationwide.

Here are the scores:

1988: +5.9
1992: +5.7
1996: +4.3
2000: +7.5
2004: +8.4

I also looked at the voter registration figures for Tucson proper. Currently, it's 44%Ds, 29%Is, 26%Rs. When you just look at new registrants (since 1/1/04), it's 37%Ds, 40%Is, 21%Rs. Finally, new 2008 voters were 41%Ds, 35%Is, 22%Rs.

So registration figures don't really reveal a trend either way (except for the statewide Indy surge), but presidential results show Pima County moving slightly to the left.