Monday, December 26, 2016

How Democrats could have won the Thurston County Commission

[EDIT 12/26/16 at 3:50 p.m.] I added an explanation of the chart and added a link to Steve Salmi's post at Green Pages that I meant to include.

Okay, so it's true that there was more voter participation by south county voters in the last county commission election.

But, I've have had a hard time reconciling the data I see in this chart below with the map I put together in that linked post above.

This chart ranks independent and Republican returns across a partisan spectrum in Thurston County. The most conservative precincts towards the left, the more liberal on the right. What is shows is that independent returns tracked well with Republicans and a consistent number of otherwise Democratic voters across nearly every precinct switch independent. They had the same slope, just one was a bit higher.

The map included in the post I linked to shows greater participation by south county voters than in recent elections. The story is that that there was increased participation driven by the rural policies of the current commission. The chart above shows voter confusion across the board. That no matter where you landed on the partisan scale of Thurston County, more people voted for conservative independents than for down-ballot Republicans.

The pro south county argument would be that a Democratic voter in rural Thurston County would vote independent because they were tired of how the county was treating rural residents. Well, sure, okay. But, that doesn't explain the behavior of typically Democratic voters in the cities, were the plight of the rural landowner is less well expressed.

So, if you erase what I call voter confusion from the north county cities (Lacey, Olympia, Tumwater) by pasting the partisan down ballot results over the county commission results, you get a 1,600 vote margin for Democratic county commission candidates. Here's my spreadsheet (column O in "if cities stayed pat").

This sort of contradicts what Steve Salmi is talking about at Green Pages:

What the data suggests is that urban Democrats will not win a county commission seat unless they are competitive in unincorporated parts of the county — which even in a high-turnout, Democratic wave election like 2008 represent the majority of votes. In 2016 both of the Democratic candidates got clobbered in that realm. To make matters worse, as I discuss here, neither Hulse nor kindred spirit Jim Cooper did very well in Lacey or Tumwater. Interestingly, these two cities saw their proportion of total votes jump 2.5 percent over 2008 while Olympia went down by .1 percent. 
This is why I suspect that in 2016 something more was going on than high turnout in the south county. Steve Klein may be at least partially right — Donald Trump had coattails. However, the most important single factor may have been that Gary Edwards and John Hutchings ran as independents, which appealed to swing voters throughout the county.
What I'm saying is that if Democrats maintain a more typical partisan lead in the three northern cities, then they'd be able to overcome even a very energized rural vote.

So, if the Democratic candidates had been able to sew the story that independents really were Republicans in sheeps' clothing, then they would have been able to tighten the results. That said, I know they did try to do that. In low information races, it's hard to create an effective narrative sometimes. But, maybe in two years, with a strong majority, the conservatives will shed their sheep clothing and progressives will be able to make a better case.

2 comments:

James Haynes said...

Steve Klein's argument that Trump provided coattails for Edwards and Hutchings is a prima facia absurdity: Edwards and Hutch both polled about fifteen percentage points higher than Trump.

S. Salmi said...

James, let’s flesh out Steve Klein’s Trump argument and see where it goes. Trump generated 16.6 percent fewer votes in unincorporated precincts than Edwards. Add in Johnson’s votes and the margin shrinks to 10.7 percent.

Klein’s analysis may be partially colored by Yelm’s unusual voting patterns. In that city Trump beat Clinton 55.4 to 32.6 percent. Countywide Trump received 37.7 percent of the vote, which was less than half a percent lower than the statewide total.

Edwards’ support clearly reached beyond that of the Republican and Libertarian presidential candidates. In addition, he racked up 60.2 percent of the vote in unincorporated parts of the county at the same time a liberal Democratic governor hit 49.4 percent. However, let’s remember that in 2008 Romero won unincorporated Thurston with 55.2 percent of the vote.

If Trump had an impact it could have been in drawing to the polls certain demographics particularly favorable to Edwards (and Hutchings). The Pew Research Center noted that Trump’s national demographics were similar in many respects to those of Romney (2012) and McCain (2008) but that he drew an exceptionally large proportion of white men without a college degree.

We can only guess about the demographic composition of the Thurston electorate because there was no exit polling. However, noodling around with census data combined with returns from previous elections might offer partial insights.