The deluge

The last 48 hours or so have been an endless cascade of action, on both the national and local levels. There were the Montana and South Dakota primaries, two events I want to cover in detail shortly. The Democrats found themselves a nominee in Sen. Barack Obama. Somewhere, Keith Olbermann experienced a pleasure no mortal being can qualify.

This blog is not about the macro scale, though, and I'll leave commenting on those matters to the pundits. I do want to quickly run through what made for for Obama's 15-point Montana win and his 11-point South Dakota loss.

For clarity, I'm going to refer to counties mostly by their significant cities from now on. Typing "King County/Seattle" is just way too unwieldy.

What worked for Obama

1. Bohemian rhapsody. An under-remarked meme of the 2008 race is that income is not the best index of Obama's performance. He runs up truly impressive totals in areas where education rates outpace incomes -- college towns, hispter haunts, artist/hippie enclaves, and the like. Montana has more of these. Missoula (Obama +34) and Bozeman (Obama +43) showed his college muscle, while Park County (Obama +34) showed his dominance among boheme tourist villages. In heavily Republican areas with small enclaves, the Democratic electorate can be heavily these types of voters.
   
   
2. An open electorate. Hillary Clinton has never been popular in Montana, period. Independents in polls regularly gave her no consideration for the General. In Montana, Obama won Democrats easily (9 points), but did even better among the 31% of the electorate identifying as Independent (29 points). South Dakota was a closed primary, open only to Democratic party members.
   
   
3. Native Americans. Finally, the Crow and the Sioux can agree on something. Despite being a toss-up, maybe even Clinton-leaning electorate in the past, Native Americans went big for Barack Obama in Montana and South Dakota. "Third-tier races" (almost exclusively Indians in these two states) went to Obama by 12 in South Dakota while he lost whites by the same margin. In Montana, where the exit poll sample was too small to use, he dominated all the reservation counties. That included a 57-point blow-out on the Crow Reservation, where Obama was recently inducted honorarily as Barack Black Eagle.
   

What didn't

1. South Dakota swing voters. Unlike past Clinton-leaning electorates, South Dakota had nothing against Obama, really. In fact, his approvals were slightly higher than Clinton's. It was nothing like the vitriol seen in the Kentucky and West Virginia primaries. Clinton's approvals were also remarkably high. Simply, South Dakotan Democrats liked both candidates, but perhaps because they're anti-establishment or attention-starved, they gave their votes to the heavily-campaigning underdog Clinton at a strong clip.
   
   
2. Blue-collar whites. The mainstream media's "non-college educated whites are terrible for Obama" mantra is overly simplistic and badly defined, but it's true. Fundamentally, this is not a good group for him in the primary. Things were, unsurprisingly, better in Montana than South Dakota. He won a lot of these areas in Montana, and showed reasonably in Butte-Anaconda. In South Dakota, he got nearly universally demolished.

There is no number three because, really, at this point, what's the use of explaining how the party nominee messed up a decent showing at the end of a winning season? Obama does have some very significant weaknesses in the General, of course. But I'm not a big fan of extrapolating primaries to the general election, and we have pundits to make those mistakes for us.

Upcoming coverage: Certified Oregon primary results, Washington state filing week, General election look-ahead and oh-so-much-more. Finals season, too. The deluge has just started.

Montana and South Dakota tonight

Early exit polls: Montana favorables look like Indiana, South Dakota favorables look like Indiana for Obama and West Kentuckinia for Clinton. Could ARG be right?

More later.

South Dakota pacesetting

American Research Group (not the greatest track record in the world) has Clinton up 60%-34% in South Dakota and Obama up 48%-44% in Montana.

I'm skeptical of both, mostly the former. The Clinton folks have publicly trumpeted that South Dakota is "tightening," which may mean they have a lead. But 26 points? I'd call that just short of impossible. We shall see tomorrow. Unfortunately for Obama, based mostly on the campaigns' implications and one old, obscuro Obama +10 poll, the media has latched onto the "Obama favored in South Dakota" meme. A loss could be a minor embarrassment.

Will it matter? Not really. It's not even the most important thing for perception on Tuesday/Wednesday, which we can expect to be major days for Obama shaping his unification strategy.

Update: Poblano throws in his (model's) two cents: Obama +5.

Our own private Idaho

I figure not much will happen for the rest of the year in Idaho, so I can use that headline straight out of the gate. You can probably tell by now that blog titles are not my forte.

Idaho, to my surprise too, had a presidential primary last night. The GOP results are predictably boring; the ridiculously low turnout Democratic election was more interesting. The results are to the left.

It's finals season, so time is limited, but we're seeing typical patterns here. Owyhee County (the big one at bottom-left) is the most amusing; Obama won with 36% thanks to a 29% vote for "somebody else," which was actually competitive with Clinton.

Obama's best performance was easily, and predictably, in Sun Valley-based Blaine County, where he won 75%-23%.

In an indication of how little voters cared, only 26% of ballots were Democratic. Obama and Clinton both placed behind Ron Paul. While Kerry barely outperformed that statistic (he got 30%), it's pretty clear that Democratic turnout wasn't stellar. Democratic turnout has never been lower than Kerry's percentage vs. Bush before, or even close -- generally, it ran much, much above.

More coverage, including the enthralling GOP results (which actually counted for delegates), possible later.

Update: Minor correction to Teton County shading. Also, please note that results are unofficial.

Montana pacesetting

SurveyUSA has released our first poll of the Montana primary, conducted by Mason-Dixon. Obama leads Clinton by a margin of 52% to 35%, with 13% undecided. According to the Billings Gazette, specific findings include:

• Obama leads men by 25 points and women by 12. This is a slightly smaller gender gap than Oregon, where Obama won men about 2-to-1 and women narrowly.
  
  
• Voters under 50 favor Obama by 26 points, while those older than 50 favor him by 10 points. Again, this is a smaller gap than recent contests.
  

The survey also split Montana into the West and the East. The geographic split is surprisingly stronger than I anticipated

• In the 18 counties of "Western Montana," Obama leads by a margin of 37 points. Included in this area are Anaconda, Bozeman, Butte, Helena, Kalispell, Missoula and Ravalli County.
  
  
• In the 38 counties of "Eastern Montana," Clinton leads by a magin of 7 points. Included in this area are Billings and Great Falls.

A full map is at left. One of the "no man's land" counties is in the West (which one wasn't specified) and the rest are in the East. Click for a larger view.

This result is interesting. At risk of reading too much into sub-samples, this may indicate that Obama has some degree of weakness in the rural areas but is absolutely obliterating in the major cities of the West. It may point to a decent enough performance in Anaconda-Butte, too, an area that seems tailor-made for Hillary Clinton.

Traditionally, Obama seems to do better in areas based on farming agriculture and worse in light manufacturing areas and mining areas. There are exceptions to this rule, and we have limited non-caucus results from the interior West. But as this poll shows, it's a region that is expected to clearly favor Obama.

The Montana Presidential Primary is June 3rd.

Oregon postgame, Pt. 1

It's been three days since the Oregon primary, and a few things are clear.

First of all, Hillary Clinton got whupped and got whupped pretty bad. She underperformed in every group she needed to do well among. Her showing in Multnomah County (just under a third of the vote) wasn't exactly a shock, but it was clearly indicative of an under-performance in the sprawly, working-class residential areas. This pattern was duplicated statewide, and reflected her biggest problem of the night. Her "base foundation" in Oregon, and the Pacific Northwest in general, is not very strong.

Clinton and Obama were essentially tied among working-class whites (the media definition: whites without college degrees), and she got obliterated among the educated. A good microcosm of this phenomenon was Benton County, which contains Oregon State University in Corvallis, and in bedroom/declining logging community Philomath. Obama broke 70% in Corvallis in a walk, won Philomath over 2-to-1 and managed to perform solidly in the portions of lower middle class Albany contained within the county.

The media heads often pointed out that Oregon is much more liberal than Kentucky. Exit polls indicated about 57 of Oregon voters were "liberal," compared to 37 in Kentucky. But there is a marked cultural difference even in these voter subsets. In Kentucky, Clinton won "liberals" by 24 points; in Oregon, she lost them by 22. Oregon "conservatives" went to Obama by 12 points; in Kentucky, Clinton swept Obama in this group by 55 points. Oregon is, put simply, un-Kentucky.

If there was one sign of Obama's demographic-transcending success in Oregon, it was in tiny Curry County, located in the southwest of the state along the Pacific Ocean. In 2000, Curry County's per-capita income fell 27% below national average, and it is 39% older than national average. Blue-collar industries dominate the county's economic landscape and most of the affluent voters are retirees. It is an area where Hillary Clinton needed to carry handily to have a hope in Oregon. The New York Senator lost it by four points.

Conventional wisdom: Obama +10 in Oregon

Why it will be closer

1. The polls say so. Some of them, at least. American Research Group has Obama up only 5; Suffolk University says 4.
   
   
2. Clinton will do better among returned ballots. It's well-established that people lie about having cast ballots. Clinton tends to do better among early voters in Oregon polls. Thus, she should do better among actual ballot-casters.
   
   
3. Turnout in the Portland area is sluggish. Despite a late-game surge in Clackamas County, turnout in the Multnomah Metro could be higher.
   
   
4. Obama voters are getting complacent. A lackluster finish in West Virginia may indicate apathy, and increased "we'll show 'em" turnout from Clinton, playing into #2.
   

Why it will be a landslide

1. The polls say so. Of course, the polls say everything all of the time. In this corner we have PPP (+19) and SurveyUSA (+13)
   
   
2. Obama has great late momentum. There's a contested race in Portland. He's shuttling college students in Eugene. Some polls already show him up big among early voters, and he can only improve with the batch left.
   
   
3. Turnout in liberal areas is strong. Lane and Benton Counties are both beating expectations. No liberal areas are coming in meaningfully under-par. Suburban vote could be disinterested GOPers or non-affiliateds, of whom there are more in the collar counties.
   
   
4. Poblano says so. Poblano's often-brilliant regression analysis shows Obama up 13. I'm skeptical of the CD break-down, but Oregon is a state that bucks conventional wisdom and OR-2 is one weird congressional district.
   

My expectations
Tune in tomorrow.

Sunday night Oregon round-up

Disappointing quiet on the Northwestern front.

If I were down in Portland, I would have been totally bummed to have missed the Decemberists, which surely increased the already astronomical rates of trendy black-rimmed glasses at Obama rallies.

Poblano concludes that Oregonians are latte-drinking lesbians at atypical rates here.

Otherwise, the excitement begins tomorrow.

Where the (new) voters are

New Oregon reg statistics: Here

Between March and April 2008, excluding Yamhill County (which I'm going to assume wasn't updated since no numbers changed), Oregon added 66,496 new Democrats and lost 8,574 Republicans. In the two-party race, Democrats now hold a registration advantage of about 200,000, or about 56.5% of major party registrations. In 2004, that number was 52.0%.

Since last month, Democratic registrations rose statewide 8.4%. Only four counties broke the statewide clip -- Multnomah (+11.1%), Benton (+11.0%), Lane (+9.4%) and Washington (+9.3%). Those are, respectively, Portland, Corvallis/Oregon State University, Eugene/University of Oregon and the western Portland 'burbs. All four should be favorable to Obama.

Lagging behind the state total, but still with strong adds, were a lot of counties with electorates that Obama wants to tap -- Hood River (resort towns), Lincoln (resort towns), Clackamas (suburbs), Jackson (university), Marion (university and state employees), Deschutes (university) and Polk (university). The lone interruption in an obvious pattern comes from Douglas County, which Obama and Clinton have both targeted. It seems pretty clear to me, though, that the new registrations are coming from the Obama side of things.

On the other end of the list are mostly Dem-declining Eastern Oregon counties. Despite the primary fervor, Grant County lost Democrats and Sherman County added more Republicans than Dems. Even an active primary season can't stop the old bleeding.

This is a hell of a lot of people

A Barack Obama rally in Portland, Oregon, has attracted 75,000 people, Ben Smith of Politico reports. This is not the Obama people's estimate, by the way; it's the Portland Fire Department's. I'm not sure if this is the largest outside rally of the presidential primary so far, but as Smith somewhat confusingly postulates, "even if junkies are used to Obamamania, we're only just entering the reality of the general election."

For comparison's sake, a crowd of 75,000 registered Oregonian Democrats would be about 9 percent of the primary electorate.

Edit: This doesn't include a line, apparently multitudes of blocks in length, that was denied entry because the park was at capacity.

Update:

Oregon turnout update

Oregon has posted a new cumulative count of ballots returned through Friday. Clackamas turnout is still remarkably laggy, but a few counties (Coos, Douglas and Lincoln especially) are beginning to pull away. More than anything, turnout is beginning to match traditional rates.

We have a point of comparison here, although a mediocre one, in the 2006 Primary. That was the last contested, partisan primary in the state of Oregon. It's a shame we don't have information about turnout by political party for 2008, because the gap between Democratic and Republican turnout is expired to be higher than it was in 2006. Thus, Multnomah County (Portland) having a higher percentage of the overall vote share wouldn't be a surprise, because they have an unusually high proportion of registered Democrats.

The best we can do is compare the overall turnout (all parties plus non-partisans), since that's all we have for 2008, and determine which counties are constituting higher and lower percentages of this primary's vote than in 2006. This isn't ideal because Oregon is a closed primary, but it's better than nothing.

Analysis: Voteshare change
The overall percentage share of the vote has, unsurprisingly, fallen in heavily Republican counties and risen in heavily Democratic counties. This supports a model showing higher Democratic turnout relative to 2006.

The biggest falls are mostly found in small, staunchly Republican counties in Eastern Washington. However, there are a few exceptions -- Linn County (Albany) recorded a double-digit percentage fall, as did Yamhill County (McMinnville). One is working-class and quite Republican; the other is exurban/small town and quite Republican.

The biggest gains are a mixed bag by candidate -- double-digit gains were recorded in Douglas (Roseburg), Jackson (Medford/Ashland), Jefferson (Madras) and Lane (Eugene). Somewhat worryingly for Obama, his Portland metro turnout seems a little weaker -- Multnomah County is up slightly, but not surprisingly so considering it's overwhelmingly Democratic. Clackamas County is essentially stable, and Washington County's share, despite a Dem registration drive, is down.

Caveats
In the 2006 General Election, the Friday prior to the General, 635,281 ballots had been received. In the end, 1.4 million ballots were cast. Thus far, we have seen 582,998 ballots cast. Overall turnout currently stands at 29%, and if this election is like 2006 in ballot return patterns, final turnout should be 60%-65%.

Even accepting the idea that these numbers show much of anything now, they're going to be essentially irrelevant by Monday. With a lot of late advertisement-based reminders, and with more competitive races (including a Portland-specific one) than in 2006, I wouldn't be surprised by turnout approaching 2/3. Both Clinton and Obama are also showing unusual interest in making use of the final-day ballot drop boxes, so that may inflate final-day turnout some too.

Conclusions
1. Turnout is not looking as strong in either Oregon or Kentucky as in other states, as the New York Times reports and these numbers validate. Because Oregon is all vote-by-mail, that's slightly less significant than it is in Kentucky. Turnout rates in elections of low interest are always going to be much higher than Oregon in Kentucky, because the ballot comes to the voters' houses.

2. Portland-area counties are probably not turning out as well as Obama would like, but the difference isn't anywhere near devastating, and is likely made up by increased Democratic registrations in places like Eugene.

3. Plenty more ballots are left to be cast in these final few days.

4. As I mentioned in another post, I still think that Obama would need somewhat higher turnout to push his numbers into the mid-double digits some polls are seeing. The "already voted" crosstabs on Monday's polls are worth holding out for before making a prediction.

5. These last days will matter...some. Even if 80% of these late voters have made up their minds, they're still more than half the electorate, and getting 2/3 of 80% of 55% is still worth about 3%. Three percentage points is something, especially when you're playing the media cycle game. But things are still going to be vastly more anticlimactic than in other states.

Personally checking Oregon's bearings

Too awkward to resist.

Anyway.

Why Oregon matters #1: It isn't Kentucky
Oregon is a significant state mostly by virtue of being juxtaposed to Kentucky. If the utter evisceration that Barack Obama received in the West Virginia primary is any indication, the coalfields of Kentucky are fertile ground for Hillary Clinton. Take a look at the southeast of the Mountain State. Obama floats around 10 percent everywhere except McDowell County, which has some lingering black population. Eastern Kentucky is unlikely to be markedly different. Indeed, the most recent poll out of the Bluegrass State had Clinton crushing 84%-15% in this region.

The rest of the state is only marginally better, thanks mostly to the existence of Louisville. Obama needs an Oregon victory, and a solid one, to avoid limping out of Kentucky on the day he is likely to clinch pledged delegates under the scenario he claims.

Why Oregon matters #2: It's an opportunity for Wisconsin, Pt. II
Despite the media's occasional mislabeling of Oregon as "affluent" or "well-educated," it isn't. Portland is more working-class than Seattle, and than it is given credit for. Compared to national average, Oregon as a whole has lower income, higher unemployment and average education rates. It's also whiter than national average. Sure, we can conveniently ignore that Hillary Clinton is just not a good cultural fit to the Pacific Northwest, but so will the media. And that gives Obama a decent talking point.

The expectations game
We have four May polls out of Oregon: one each from Rasmussen (May 1st, Obama +12); the Portland Tribune (May 10th, Obama +20); SurveyUSA (May 11th, Obama +11); and Public Policy Polling (also May 11th, Obama +14). Correspondingly, the media mantra has been: Oregon is likely Obama's, and likely Obama's solidly. A Clinton upset would be very bad indeed. But there's no indication of any such thing.

The only vaguely worrying news for Obama fans is that the SurveyUSA poll has him only up 1% among those who have already returned their ballots (43% of the sample). First of all, a lot of those voters are lying; days after the poll, only about 25% of Democratic ballots have been returned. Secondly, Obama does better among the 78% voters who have made up their mind (Obama +21) than the 19% who haven't (Obama +15). How that jives with an overall Obama +11, I do not know. We do learn two lessons here, the first being not to hold these numbers in high reverence, and the second being that voters are filthy liars.

SUSA's poll is a little weird in a few other ways too. Obama leads by 9 in the Portland metro, but by 18 in the rest of the state. SUSA uses an incredibly weird definition of "Portland metro": not only do they manage to include Salem in it, but they include Wheeler County, which likely deeply resents being associated with the PDX. It's hard to gauge where Clinton is performing strongly enough to keep this area to Obama +9 while the rest of the state is such slaughter. With only a third of the state living outside this definition of the "Portland metro," it could be a subsample oddity.

In any case, though, Obama is running strongly statewide.

Turnout forecast
The Secretary of State's office reports that Democratic turnout is thus far 27%, with an ambitious target of 70%. Unfortunately, Oregon does not keep partisan turnout information by county. All that can be said is that, as of Tuesday, combined turnout (including Republicans and unaffiliated voters) was 22%.

Significant counties running above average turnout are Clatsop, Coos, Klamath and Lincoln.

Running close to state average were Benton, Columbia, Jackson, Lane, Linn, Marion, Multnomah, Polk, Umatilla, Washington and Yamhill.

Turnout was slightly more sluggish in Clackamas, Deschutes and Josephine counties.

At this juncture, with such a small separation between counties and with the date so early, not much can be made from this other than that no clear patterns have emerged.

More soon.

That giant sucking sound you hear

...is the total inactivity of this blog.

The primary season has been in full swing, and has certainly had some very interesting phenomena to analyze. It's had twists, turns and exciting, Bondian intrigue involving slow vote counting in northwestern Indiana. It's been riveting.

And what have you heard from here? Silence.

Don't fear! We still have a few primaries left, including three (count 'em!) in states where I have creepy knowledge--Oregon, Montana and South Dakota.

So, hang on and prepare for extensive Oregon coverage starting today. It's going to be Asperger's-y, and it's going to be intense.

2000 GOP Primary Flashback Special

With the Washington 2008 primary getting closer and closer, I thought it would be worth pulling out some maps of the 2000 primary. It gives an illuminating look at the different kinds of Republican voting blocs across Washington.

The race was between George W. Bush and John McCain. Bush barely beat McCain, receiving 48.3% of the vote to McCain's 48.0%, carrying 21 of Washington's 39 counties.

Essentially, here's what happened:

• Culturally conservative, but unexceptionally religious blue-collar workers favored McCain. This accounts for northeastern Washington, the parts of southwestern Washington he carried, and much of the coast.
• Older folks favored McCain, too. This helped him on the Olympic Peninsula and along the coast.
• Liberal areas have lots of moderate Republicans and independents. They also leaned McCain.
• Military helped McCain in Island County.
  
• I had someone ask me why a liberal county like Whatcom didn't vote McCain. Same reason that it went Bush in 2000. The Dutch reform areas around Lynden account for a big chunk of the GOP vote, and are extremely conservative.
• The results in Clallam and Spokane Counties are a little weird. I wish I had precinct break-downs for them. Clallam certainly trends older, but Spokane is an oddity.
  

Interestingly, the suburban areas (e.g., Snohomish County) weren't nearly as McCain-friendly as one might expect.

Now, voting only among Republicans:

Bush landslided here, 57.8% to 38.9%. He carried two counties, both very liberal ones. This is a fairly uninteresting map after the last one. Places with moderate GOP-leaning independents tend to have moderate GOP members, too.

And finally, among independents, where McCain dominated, 61.1% to 34.4%:

It's hard to apply this outcome to 2008. Michigan was one of the few states to give McCain a victory in 2000, yet in 2008 it delivered him a decent loss to Mitt Romney. Romney promised downtrodden workers there the world, and may do the same if Washington ends up competitive and useful. So, who knows? The blue-collar workers who supported McCain could flip. So could the affluent suburbanites of the Eastside. Or, maybe, we'll see a map not dis-similar to the above (although I suspect that McCain will win this time).

And what of Mike Huckabee? He certainly might woo those Dutch Calvinists, at least. He certainly did in western Iowa. That is, if Huck is still around then. Three weeks is an eternity in politics.

Edit: I know that Camano Island is unintentionally included with Snohomish County in the above maps. It should be with Island County. Long day. Sorry for the error.

Seattle neighborhood results, Pt. 4

Rainier Valley
Prosecutor: 58% Sherman (D)
Assessor: 86% Noble (D)
Simple School Majority: 65% Yes
I-960: 70% No
Sound Transit: 54% No
Notes: This working-class area of South Seattle is quickly developing into noticeable subneighborhoods. The increasingly affluent Mt. Baker neighborhood's school results alone show this; it cast 74% for simple majority, while Rainier View cast only a narrow 50% win. All-around, Mt. Baker and Columbia City (both gentrifying) have developed into solidly liberal areas. The others remain liberal, but not particularly ideological, and will give decent showings to anti-tax positions.

Seward Park
Prosecutor: 57% Sherman (D)
Assessor: 85% Noble (D)
Simple School Majority: 69% Yes
I-960: 71% No
Sound Transit: 53% No
Notes: South Seattle's oldest solidly middle-class area delivers an entirely typical result.

University District
Prosecutor: 70% Sherman (D)
Assessor: 87% Noble (D)
Simple School Majority: 77% Yes
I-960: 76% No
Sound Transit: 56% Yes
Notes: If it's news to anyone, the U-District is liberal. Really liberal.

West Seattle
Prosecutor: 50% Satterberg (R)
Assessor: 81% Noble (D)
Simple School Majority: 61% Yes
I-960: 64% No
Sound Transit: 59% No
Notes: A number of things conspire to moderate South Seattle. First off, there are a lot of affluent voters and older condominium-dwellers around Alki and Duwamish Head. Then, in south West Seattle, the voters become more suburban and far-separated from The City. Satterberg's victory came through small but comfortable wins in Alki, Fauntleroy and Arbor Heights. Sherman's only solid victory was Fairmount Park, which is easily the most liberal West Seattle district. The Arbor Heights area also delivered the worst defeat for Sound Transit Prop. 1, which received under a third of the vote there.

Woodland
Prosecutor: 68% Sherman (D)
Assessor: 88% Noble (D)
Simple School Majority: 77% Yes
I-960: 79% No
Sound Transit: 50% No
Notes: Woodland combines the ultraliberal suburbs of Fremont, Phinney Ridge and Wallingford in with Green Lake, which looks conservative only in comparison. Results here are fairly static, with Green Lake predictably voting slightly less to the left than the other three. Sound Transit Prop. 1's narrow failure here shows how poorly the measure performed; that Fremont was responsible for its failure is beyond my explanation.

That's all, and may be all for the results reports.

Seattle neighborhood results, Pt. 3

Interbay
Prosecutor: 68% Sherman (D)
Assessor: 80% Noble (D)
Simple School Majority: 68% Yes
I-960: 59% No
Sound Transit: 59% No
Notes: This is a working-class area lying between Queen Anne and Magnolia. It's also tiny, casting around 250 votes. Clearly, also not fond of taxes. But making suppositions based on the results of two precincts is generally pointless.

Lake City
Prosecutor: 56% Sherman (D)
Assessor: 83% Noble (D)
Simple School Majority: 68% Yes
I-960: 69% No
Sound Transit: 54% No
Notes: Residential North Seattle tends to vote fairly uniformly. Satterberg managed to come within shooting distance in Cedar Park and Matthews Beach (both wealthy and near the water), but still fell short.


Magnolia
Prosecutor: 57% Satterberg (R)
Assessor: 79% Noble (D)
Simple School Majority: 61% Yes
I-960: 63% No
Sound Transit: 59% No
Notes: Wealthy and rather fond of the status quo, Magnolia is typically more Republican than the city as a whole, and this election was no exception. The wealthy cliff homes around Briarcliff gave Dan Satterberg his best showing, with 63 percent of the vote districtwide.

Northgate
Prosecutor: 58% Sherman (D)
Assessor: 82% Noble (D)
Simple School Majority: 66% Yes
I-960: 68% No
Sound Transit: 51% No
Notes: A diverse range of neighborhoods, overall trending middle-class but with scattered portions on both ends of the middle-class spectrum, Northgate is one of the better cross-sections of residential Seattle. North College Park (also known as Licton Springs) and Maple Leaf help to keep this area staunchly liberal, with Pinehurst (around Jackson Park) and its very suburban vote almost competitive in the Prosecutor's race.

Queen Anne
Prosecutor: 58% Sherman (D)
Assessor: 85% Noble (D)
Simple School Majority: 71% Yes
I-960: 71% No
Sound Transit: 51% Yes
Notes: Dividing Queen Anne into four districts is mostly only to assure moderately-sized maps. Other than lower Queen Anne's more numerous multi-unit housing (it borders Belltown), they're all similar in culture and voting patterns.

Seattle neighborhood results, Pt. 2

Delridge
Prosecutor: 58% Sherman (D)
Assessor: 79% Noble (D)
Simple School Majority: 61% Yes
I-960: 65% No
Sound Transit: 58% No
Notes: Delridge – or, as it was once condescendingly put to me, "the West Seattle that West Seattle doesn't want you to know about" – is a working-class area sandwiched between the Duwamish River industrial area, Boeing Field, and the more middle-class areas to the west. Accordingly, it's Democratic without having extreme ideological fidelity to liberal causes. It is, however, liberal. This is Seattle, after all.


Downtown
Prosecutor: 59% Sherman (D)
Assessor: 82% Noble (D)
Simple School Majority: 71% Yes
I-960: 69% No
Sound Transit: 58% Yes
Notes: Downtown is really a vague grouping of dissimilar neighborhoods, ranging from Belltown in the north to the International District in the south, and over to First Hill in the west. First Hill is the most liberal, generally, while the small Pioneer Square vote veers unpredictably, but tends to be the least ideologically liberal. Pike Market's vote for Satterberg made for a surprising dash of blue in the otherwise solid red, here.

Delridge
Prosecutor: 58% Sherman (D)
Assessor: 79% Noble (D)
Simple School Majority: 61% Yes
I-960: 65% No
Sound Transit: 58% No
Notes: Delridge – or, as I had a friend put it, "the West Seattle that West Seattle doesn't want you to know about" – is a working-class area sandwiched between the Duwamish River industrial area, Boeing Field, and the more middle-class areas to the west. Accordingly, it's Democratic without having extreme fidelity to liberal causes.

Duwamish
Prosecutor: 65% Sherman (D)
Assessor: 81% Noble (D)
Simple School Majority: 61% Yes
I-960: 66% No
Sound Transit: 53% No
Notes: Duwamish is a small district consisting of two tiny sub-districts, Georgetown and South Park. Both lie near the Duwamish River in an industrial land long dominated by Boeing and the working class. Georgetown finds its roots as a neighborhood for aerospace workers, while South Park has seen waves of immigrants, and is now a primarily Hispanic area. Both have low education rates, although Georgetown is newly hip and is seeing some gentrification. Both also clocked in at under 350 votes. They share a typical working-class voting pattern for Seattle: heavily Democrat, but with some staunch Republicans. Georgetown gentrification has made it slightly more liberal, but also less partisan Democratic. These are matters of inches, though.

East Points
Prosecutor: 56% Sherman (D)
Assessor: 85% Noble (D)
Simple School Majority: 73% Yes
I-960: 73% No
Sound Transit: 51% No
Notes: East Points is an indefensibly lame name for the points east of Interstate 5 and south of Northgate that don't fall within the University District. It includes the university-fueled neighborhoods of Bryant, Ravenna and Roosevelt, the traditional middle-class Wedgewood, and the tony areas around Laurelhurst and Windermere. All in all, it is an educated, well-off area. Satterberg carried Laurelhurst with nearly 60 percent, and with smaller, solid margins in View Ridge and Windermere. But a slaughter in Ravenna and solid losses in the other non-wealthy areas amounted to a moderate loss. On other matters, agreement was common; this area likes it school funding, and doesn't care for its Tim Eyman.

Seattle neighborhood results, Pt. 1

The following results are determined by matching the precinct boundaries imperfectly with the Seattle Neighborhood Map Atlas. Some places, they're fairly exact; others, a few streets are shaved inappropriately into other neighborhoods. That's just how the precincts are drawn, though, and this is a relatively accurate representation.

Ballard
Prosecutor: 61% Sherman (D)
Assessor: 83% Noble (D)
Simple School Majority: 71% Yes
I-960: 73% No
Sound Transit: 56% No
Notes: Nothing especially fascinating here in the sub-district results. West Woodland was both Sherman's best area and Noble's worst. Satterberg came in within a few points of winning Sunset Hill.

Beacon Hill/Industrial District
Prosecutor: 67% Sherman (D)
Assessor: 84% Noble (D)
Simple School Majority: 64% Yes
I-960: 68% No
Sound Transit: 54% No
Notes: Beacon Hill sort of melts from voting like downtown in the north to South Seattle in the south, and the results came in accordingly, with the north being slightly more liberal, in a mostly-meaningless way.

Capitol Hill
Prosecutor: 66% Sherman (D)
Assessor: 89% Noble (D)
Simple School Majority: 77% Yes
I-960: 75% No
Sound Transit: 52% Yes
Notes: These numbers seem conservative for Capitol Hill, until you consider that Madison Park (75% Noble, 38% Sherman) is included, along with the lesser liberal strongholds of Montlake and Portage Bay. Broadway and Stevens (the 15th Avenue area) are still as GOP-unfriendly as ever. Satterberg didn't even hit a quarter of the vote at Broadway, although his over-30% performance in Stevens was almost-impressive (he didn't do so badly in the affluent northern areas).

Cascade
Prosecutor: 64% Sherman (D)
Assessor: 85% Noble (D)
Simple School Majority: 71% Yes
I-960: 70% No
Sound Transit: 54% Yes
Notes: This area, perhaps better known as Lake Union (East- and Westlake plus South Lake Union/Vulcanville), is one of Seattle's smallest districts. The condominiums of Westlake were, as usual, the least liberal.

Central Area
Prosecutor: 72% Sherman (D)
Assessor: 90% Noble (D)
Simple School Majority: 74% Yes
I-960: 77% No
Sound Transit: 53% Yes
Notes: Seattle calls this the "Central Area" instead of "Central District" for a reason. In addition to the traditionally black (albeit rapidly gentrifying) Central District (the Mann and Minor sub-districts, here), this includes the similar Rainier Valley border district of Atlantic, plus the well-to-do communities of Harrison/Denny-Blaine, Leschi and Madrona. The latter three (especially the first) pulled down Sherman's numbers a bit, and Atlantic votes a bit like South Seattle's working-class areas, with a few staunch Republicans. Beyond that, there's not much to remark here, other than that gentrification sure isn't hurting Madrona's liberal heritage in the slightest.

More to come.

2007 King County Map - Sound Transit Prop. 1

I'll end the series with the weirdest of all of the maps. It isn't surprising that the "roads and rails" map is a little odd. But I was expecting something a bit more polarized. Obviously, urban central Seattle was this area's best performance, but I wouldn't have put money on the measure doing better on Mercer Island than in near North Seattle. Similarly, there's more green on the Eastside than I expected and a lot more red in west Seattle (that is, all parts of Seattle west of the median) than I'd have wagered. That's an impressive flunking in Ballard. It's explainable, but a little surprising.

Note: There are four precincts that just had a small percentage of their voters voting on the Sound Transit issue. I have no shapefile of the Sound Transit district area, so I just covered these in as a whole. If it's not clear where these are, the district is fairly rounded. Overall, it doesn't make much of a difference in the map.

King County Map - Simple Majority for Schools

This is my favorite map of the year thus far. I have never before seen a map where the suburbs and the exurbs disagree so profoundly. Things start off interestingly when Mercer Island looks like part of Seattle (which, arguably, it has for a while). Save for a few apartment-heavy areas, the Eastside dug this stuff up. Just look at the Sammamish Plateau (especially the Klahanie development in the southeast) and Issaquah area. Then, see the solid thumping this measure got in southeastern King County. The urban/rural divide even drags down margins on Vashon Island. I can't entirely venture to explain why the divide was so profound, but it makes a half-decent substitute for a density map.

Addendum
I received a question about the precinct in the southeast, Chinook, which obviously stands out. My original response was that this precinct primarily serves the King County portion of the rural village of Greenwater (the Census district is in Pierce County, but some of the village clearly spills over into King). It seemed a logical assumption, considering the name referred to Chinook Pass. But it looks like a solid chunk of the precinct's population lives in an unincorporated area northeast of Enumclaw proper. This makes it hard to venture whether it was Greenwater or Enumclaw influence that caused this to pass, and probably indicates that I've been overstating the Republicanness of the King County portion of Greenwater.

It also means my trip there to determine why the King County portion is politically inversive to the Pierce County portion was wasted on the beautiful scenery and small-town atmosphere. Damn.

2007 King County Map - County Assessor

Democratic incumbent Scott Noble scored electoral manslaughter here, beating like-named Republican Jim Nobles, scoring 71.47 percent of the vote. This is a moderate outperformance of John Kerry, and the map comes out accordingly.

That's not to say that this map is without intriguing results. There are two Seattle precincts to have voted Bush in 2004 - the ultra-exclusive Broadmoor Country Club being the biggest victory. It's the precinct that sticks out of Seattle in this map too. Bush won it with 57 percent, a severe underperformace for any Republican candidate. But Nobles (the Republican) barely cracked 42 percent here, which may be the worst performance for a Republican in Broadmoor in contemporary political history, if not ever.

The other precinct to vote Bush - a tiny, block-long pair of condominium towers in Madison Park (those two tall buildings on the waterfront, for locals) gave Bush 51 percent. Nobles managed less than 29 percent, which, again, may be the worst performance for a Republican in the precinct's history. Even Will Baker performed better, if memory serves.

But the swings weren't all so hot for the Democrats. Excluding incorporated areas, there's a good degree of blue east of the Highway 18 corridor. Exurban southeastern King County is fast becoming the only portion of the county where the Republicans can rely on regular victories. New subdivisions have helped; one in Covington nearly delivered a 2-to-1 margin to Nobles (the Republican).

Also bad news for Democrats is laggy performance around Ravensdale and Cumberland. This was competitive territory as recently as 2000, but Bush managed healthy margins in 2004, and even Nobles won most of these precincts this year. But as a sign of health for the King County GOP, this isn't much of one. These areas are still sparsely-populated. It's in the subdivisions (as in the aforementioned one) where the Republicans are seeing signs of life, and that's pretty much it.

2007 King County Map - I-960

Here we have Tim Eyman's I-960, which mandates supermajority vote (either via initiative or in the legislature) for any tax increase. Another map that isn't particularly shocking. But this is indicative that affluence does not necessarily mean anti-tax sentiment. Just look at the Gold Coast - this measure failed in Clyde Hill and Yarrow Point. There was also a rather solid thumping on Mercer Island. Inversely, this performed stronger in lower middle class liberal areas than a comparable partisan showing*. Not a surprise - Washington's lower middle class is not fond of taxes, regardless of politics.

* - Forgive the neologism. What I mean is, if a Republican candidate were to win the state of Washington by two percentage points (as I-960 did), they would assuredly not perform as well as yes on I-960 in the aforementioned areas.

2007 King County Map - County Prosecutor

Sorry for the lack of updates lately, but there hasn't been much to add. Now that results are certified and released, I can begin to upload some maps. The first is the King County Prosecutor's race, where Dan Satterberg (R) defeated Bill Sherman (D), 54.37% to 45.63%.

No huge surprises in this map. Only two incorporated areas voted Sherman: Seattle (of course), and tiny Skykomish, the little dot in the northeast.

Other than Seattle, only the hardcore Democratic areas to the south of town (White Center, Skyway, Riverton and Boulevard Park) didn't stray. Vashon Island also voted Sherman, although Satterberg won two precincts - quite impressive for a Republcian. Other than that, there are a few scattered Democratic areas, all reflexively Democratic precincts. The smallest swings came around the Algona-Pacific area, where John Kerry underperformed in 2004, and voters have a history of delivering more Democratic results downticket.

One day in America

Unfortunately, Census restrictions mean that day is actually some time in early 2000, but Time Magazine's web supplement to their somewhat schmaltzy Thanksgiving issue article of the same name is still pretty cool. Revelations abound (New Hampshirites drink Nevadans under the table? Housekeepers/butlers are the second-happiest occupation?), but my favorite graphic is the night vs. day population comparison. A map of eight American cities is available, and while L.A. is inexplicably not included, Seattle made the cut.

There aren't really any great surprises. And, over seven years out-of-date, it doesn't do Belltown, Pike Market, or South Lake Union any justice. I could keep complaining (low image resolution, color scheme clearly meant for Manhattan densities), but I'll just link instead. Click on "City Population Shift" to access.

It's a good waste of 45 seconds.

Simple majority declares victory

I don't generally post press releases, but the Approve EHJR-4204 (there is technically an "E" there) folks were nice enough to add me to their mailing list out of nowhere. I'll reward this (is this a blogging faux pas or standard practice? I really don't know) by engaging my copy-paste reflexes. I won't post it in full because, damn it, I have integrity.

Seattle -- Today, members of the Simple Majority Coalition to Approve EHJR 4204 gathered to declare victory for kids and schools. Yesterday, EHJR 4204 gained its first lead since voting started. Current trends show that 4204’s lead should increase after tonight’s posting.

Voter support for the Simple Majority has increased each day as county elections departments count late-arriving mail-in ballots. Supporters worked hard to encourage pro-education voters to cast their ballots. They expected the election would be close.

“The apparent passage of Simple Majority for school levies is a great victory for Washington’s 1 million public school students, whether they attend school in Seattle, Yakima or Spokane,” said Mary Lindquist, President of the Washington Education Association.

...although Yakima's voters apparently beg to disagree.

Everett Station site chosen for UW-Snoho

The UW's consultant has recommended the most urban of potential sites, near Everett Station, for the university's Snohomish County branch, says the Seattle Times. Apparently, Lake Stevens and Marysville are none too happy. I don't entirely blame them; they've spent millions preparing their sites for show. But, along with the Smokey Point location, I'm glad they lost out. As nice as the sites may be, Everett needs a center like this. We've seen what urban university branches can do to deprived downtowns (Tacoma). The Everett Station location will hopefully achieve the same symbiosis.

Site locations in an industrial area along the Snohomish River, and near Everett Community College, both have effectively been abandoned as possibilities (both are within the city of Everett, and there's no reason for the city to offer competition).

Nearly 1,000 new condominiums and apartments have been sprouting up around the Hoyt Avenue corridor in downtown Everett (here's a map from the Herald). With these and the $400 million Port Gardner Wharf, this is promising to be the most exciting (or, maybe, only exciting) time for Everett development in recent memory.

Don't count HJR-4204 out just yet

Darryl over at HorsesAss provides a mathematically sound analysis of why HJR-4204 (simple majority for schools) not only could pass, but probably will. All of this is not due to King County's slow ballot-counting, but rather to a late-term shift in results. Eric Earling at Sound Politics adds (to the chagrin of SP commentators ready to blame King County shenanigans):

[T]he race is closing [in] and the Snohomish County results themselves reveal in microcosm why the measure's results are changing from first reports. The intial [sic] count of ballots on Tuesday evening showed 4204 failing in Snohomish county by a margin of 48.98% to 51.02%. As of yesterday afternoon's tally, the measure is now ahead in the county 51.25% to 48.75% - meaning the measure has been passing 53.48% to 46.52% in the later ballots. This appears to be holding true in other counties, as well.

As we saw in 2006, late ballots seem to be benefiting liberal causes more than conservative causes. This has not always been the case, and there are a number of explanations for it - I may devote a post to that when election season is done with.

We should have a better idea of the fate of HJR-4204 tomorrow afternoon, although it will likely be a good while before the race can be called.

I-960 in Tacoma

Tacoma is handy for studying socioeconomics and politics in Washington state, because it can easily be broken into about six different neighborhoods:

• North Tacoma: Solidly middle class, residents trend older, World War I-era suburb
• West Tacoma/Ruston: Solidly middle to upper-middle class, few families with kids, World War II-era suburb
• Downtown/Dome District/Hilltop: Lots of subsidized housing, more diverse than other areas, urban
• Central Tacoma: Lower middle class with many renting families
• South Tacoma: Similar to Central Tacoma, although with higher income variance (separated maybe for geography)
• Northeast Tacoma/Browns Point: Solidly middle to upper-middle class, lots of families with kids, newer developments

The I-960 results (preliminary only, remember) reflect these descriptions as they do elsewhere in the state:

• North Tacoma: 42.6%
  
• West Tacoma/Ruston: 50.1%
  
• Downtown/Dome District/Hilltop: 42.0%
  
• Central Tacoma: 53.1%
• South Tacoma: 50.3%
  
• Northeast Tacoma/Browns Point: 54.9%

So, what does this tell us?

• Age of development has become about as important as income
  
• More urban and inner-ring suburbs, regardless of income levels, tend to oppose economically conservative propositions.
• Anti-tax does not equate to economically conservative.
• Washington's upper middle class is a lot more friendly to taxation than its lower middle class, regardless of location.
  

These are fundamentally obvious points, but it helps to explain shifts on the east side of Lake Washington. Whether Pierce County will see the same movement is uncertain, but conventional wisdom would indicate that it may take quite a while - Tacoma is no Seattle.

(Election result trivia: Preliminary results show I-690 is failing in Tacoma, as well as only two other Pierce County towns - Fife and, significantly to the points above, Fircrest.)

Pierce update: Prop. 1 up in 27th, and more

I was mistaken in saying prematurely that Sound Transit Prop. 1 would probably only pass the Seattle-based 43rd district. Early results indicate that it is passing with about 50.6 percent in the 27th district, which is based in Tacoma.

It looks these results owe primarily to a strong showing around downtown, a small but cozy m