There are Democrats in rural areas in Oregon. In some small towns and rural areas Democratic voters outnumber Republican voters. In some areas they even elect Democrats to represent them in Salem and Washington, D.C.
On the other hand several urban and suburban areas lean Republican and elect Republicans to represent them. Even in Republican areas, though, there are still people who vote Democratic. And it seems that when a Democrat outperforms the baseline, they seem to do so more in Republican areas and rural/small town areas than they do in urban and more Democratic areas. Below I will explain.
In 2012 Kate Brown ran two points behind President Obama in their respective re-election campaigns in Oregon. In state house districts held by Democrats, she ran over 3 points behind him. Yet in Republican-held districts, she ran nearly a half a point ahead of him. In urban and suburban state house districts she ran nearly four points behind him. Yet in districts predominantly in small towns and rural areas, she ran nearly a full point ahead of him. Using presidential performance as a baseline, and gubernatorial as a baseline for a midterm election, other Democrats perform better in more rural and Republican areas than in more Democratic or urban areas relative to the baseline.
Classifying districts based on whether they are urban or rural is subjective. I classified districts as urban if their votes were predominantly cast in precincts within the Metro boundaries, or within the city limits of other cities with at least 50,000 people. Most urban/suburban districts are held by Democrats, and most rural/small town districts by Republicans, but Democrats hold 7 seats classified as rural/small town by my standard, and Republicans hold 6 urban/suburban seats. In the map to the left Democratic urban/suburban seats are dark blue, Democratic rural/small town seats are light blue, Republican urban/suburban seats are pink, and Republican rural/small town seats are red. Republicans hold urban/suburban districts in Medford, Albany, southeast Salem, the southern end of the Portland metro area (Sherwood, Tigard, Tualatin, West Linn), and Bend. Democrats hold rural/small town districts based in Ashland (with some Medford precincts), Coos Bay/North Bend and Florence, Lincoln City and Newport, Creswell and Harrisburg (with some Eugene-Springfield precincts), Tillamook and Astoria, Woodburn (with some Salem precincts), and Columbia County (with some Portland precincts).
Rural Democratic districts |
5 |
9 |
10 |
11 |
22 |
31 |
32 |
Urban Republican districts |
6 |
15 |
19 |
26 |
37 |
54 |
|
I don’t think ignoring the difference between suburban and urban, or small town and rural, or even small town and urban, is easy to do without substantial justification, but this is not the place for a full discussion of that. Suffice it to say here that while a city of 50,000 people may not be terribly significant in the politics of a state like California or Texas, in Oregon that city will anchor at least one legislative district, and so that makes it significant in that sense if none other. And because a standard of 50,000 makes a sharp distinction between places that have 100,000 or more in their metro areas (Eugene-Springfield, Corvallis-Albany, Medford, Bend) and much smaller communities, it seems like a strong standard for comparison in the state.
I did not stop with comparing the 2012 presidential election to the 2012 Secretary of State election. My analysis also extends to individual precinct results, though that will not be included here today. Included are comparing the 2012 Attorney General and Treasurer races to the presidential election of that year, and also the 2014 U.S. Senate election and 2014 gubernatorial election.
Consider the performance statewide of the candidates to how they did in rural versus urban districts, and Democratic-held versus Republican-held districts.
Candidate |
Obama |
Brown |
Wheeler |
Rosenblum |
Candidate |
Kitzhaber |
Merkley |
2012 and 2014 performance by percentage
Statewide |
56.3 |
54.3 |
61 |
58.9 |
Statewide |
53.1 |
60.2 |
Rural |
44.1 |
44.9 |
49.9 |
48 |
Rural |
41.7 |
50 |
Urban |
65.3 |
61.3 |
69.3 |
67.1 |
Urban |
61.9 |
67.9 |
Republican |
42.1 |
42.6 |
48 |
46 |
Republican |
39.8 |
48 |
Democratic |
65.9 |
62.7 |
70.3 |
68.1 |
Democratic |
62.7 |
68.9 |
But more specifically, how those numbers stacked up to the baseline of the main contest interesting voters, assuming the baseline in 2012 is President Obama’s performance, and in 2014 it is Governor Kitzhaber’s. Positive numbers are performances better than the President or Governor Kitzhaber, negative numbers are worse.
Candidate |
Merkley |
Brown |
Wheeler |
Rosenblum |
Relation to Presidential/Gubernatorial percentage
Rural |
8.3 |
0.9 |
5.8 |
3.9 |
Urban |
6.1 |
-4 |
4 |
1.7 |
Republican |
8.2 |
0.5 |
5.9 |
3.8 |
Democratic |
6.2 |
-3.3 |
4.3 |
2.2 |
The differences are not drastic, but it also appears that in general Democrats were able to perform better relative in rural areas relative to urban ones, than in Republican areas as compared to Democratic ones. The differences in those can be seen in the chart below. For example, Jeff Merkley outperformed John Kitzhaber by 2.2 better in rural districts relative to urban districts, but did only about 2 points better in Republican districts compared to Democratic ones. Kate Brown’s performance relative to the president’s was 3.8 points better in Republican districts as compared to Democratic ones, but about 4.9 points better in rural as compared to urban ones.
Although consistent so far, the differences across just these 6 races in two elections is not sufficient for me to be confident in a claim that the urban/rural divide consistently offers a greater contrast than the partisan divide. I will need further study, and I would prefer to use a more precise measure, that being precincts rather than districts. In Oregon many precinct lines fall along city limits and do not cross them, in particular for larger cities. I have begun a long term analysis of precincts for this study and other purposes which I may publish at a later date.
But in the interest of finding any more basis for a conclusion today, let us compare the data for the four distinct groups rather than just putting them in one of two groups. Let’s separate Democratic urban districts from Democratic rural ones, and Republican urban ones from rural Republican ones.
At this point the numbers become inconsistent in some ways, but in others are very consistent. For example, The four non-baseline Democratic candidates all had their strongest performances relative to their respective baselines in rural Republican districts, and all had their weakest in urban Democratic districts. But the relationship between their performances in urban Republican and rural Democratic districts are not as consistent. Ted Wheeler, in his 2012 re-election for Treasurer, did performed better relative to President Obama in urban Republican districts than rural Democratic ones, yet Jeff Merkley, Kate Brown, and Ellen Rosenblum performed better relative to the baseline in rural Democratic districts than urban Republican ones.
|
Merkley |
Brown |
Wheeler |
Rosenblum |
Comparing performances across all 4 categories
Rural Rep |
8.5 |
1.5 |
6.1 |
4 |
Urban Rep |
7.1 |
-2.5 |
5 |
2.7 |
rural Dem |
7.7 |
-0.7 |
4.75 |
3.1 |
Urban Dem |
5.8 |
-4.3 |
3.8 |
1.5 |
A good illustration can be seen by comparing two neighboring state house districts, district 24 based in McMinnville on the one hand, and district 26 based in Sherwood and Wilsonville on the other. The 24th qualifies as a rural district, while the 26th is suburban. They are similar districts in partisanship: in 2014 Kitzhaber received 46.8% of the two party vote in the 24th, and 46.4% in the 26th, while in 2012 President Obama received 49% in the 24th and 50.3% in the 26th. While Jeff Merkley outperformed John Kitzhaber by 5.4 points in the 26th, he outperformed him by 7.4 points in the 24th. In 2012 Kate Brown got just about a half point less than President Obama in the 24th, but nearly 4 points worse in the 26th. Attorney General Rosenblum did nearly a point better than the President in the 26th, but more than 3 points better in the 24th. Treasurer Wheeler did about 4 points better than the president in the 26th, but 6 points better in the 24th. These Democrats consistently did better in the more rural 24th than in the suburban 26th, despite that the President narrowly won the 26th and narrowly lost the 24th.
What does all of this mean? It appears that when a Democratic candidate does better than the baseline, they tend to pick up more or lose fewer votes in rural and small town areas than urban and suburban ones. Is this a phenomenon limited to incumbents, as all of the candidates I’ve considered above are? I cannot say for sure at this point, but looking at legislative results it does not appear to be so. Several non-incumbent Democrats running in rural districts outperformed the top of the ticket in 2012 and 2014, and in fact also in 2008 and 2010, while they were less likely to do so in urban districts if they were not incumbents, but fleshing that data out will have to come later. Is it a factor of rural voters crossing over to vote for inoffensive moderate Democrats, but not liberal ones at the top of the ticket? Well, again, not really. Kate Brown had the biggest divergence between urban and rural performance compared to President Obama, but she has generally been perceived as a liberal Portland Democrat. In 2014 as well, John Kitzhaber was actually the more moderate Democrat of the two statewide campaigns, and Jeff Merkley did sigificantly better than he among rural voters. Is this a matter of fewer Democrats compared to Republicans undervoting in downballot races, compared to urban voters? That could be a portion of it, but in many situations the amount of overperformance is greater than that factor could have been. I have been pursuing and intend to pursue this further in the future to see if this notion holds up, with more and more specific data, as I said above, and when I am able I will expand upon this.
In the meantime, does this mean that I think we should focus more on targeting and turning out persuadable rural voters, particularly those in Republican districts? Not necessarily. In reality of the 60 districts in the state house, 28, nearly half, are urban/suburban districts held by Democrats. Of the rural/small town Democratic districts as I’ve classified them, another four contain parts of Portland, Salem, Eugene, and Medford, some of the biggest cities in the state. The amount of resources it would take to target rural voters is substantial because they are less densely located compared to heavily Democratic cities such as Corvallis, Eugene, and Portland, and even less Democratic cities like Bend and Salem. Statewide candidates, aside potentially from Ron Wyden and Jeff Merkley, who have no term limits and can develop a longer-term relationship with rural voters, have little incentive to commit resources to targeting sparse potential voters. Moreover, it is unlikely that moving less than a large number of rural voters would change much in many districts. Currently of the “rural” house districts occupied by Republicans, only perhaps the 24th in McMinnville and the 52nd in Sandy and Hood River are close enough where moving a bit more rural/small town voters could change the outcome. Rather than making a prescription based on this slim evidence, I’m just providing a little illumination of something I’ve noted in my research.