Washington 2020: The Calm Before the Drizzle

[jimrtex]

More statistics about why this system sucks:

- In 92 out of the 124 races (74%) on the ballot, there is only one or two candidates on the ballot. So the primary is meaningless, but we're still going to spend money and time on it.

- There are 207 candidates for the Legislature this year, as opposed to 233 in 2006. So we get fewer choices.

How many races were there in 2006? There are going to be flaws in the first year of this system. People will get used to it and then elections will become interesting.

Actually there are 274 candidates this year, including 13 independent or 3rd party candidates.   In 2006, there was a single 3rd party candidate.

In 2006, there were 65 D v. R races where the primary was meaningless, and 21 D and 14 R races where both the primary and general were meaningless.  There were also 1 D v D primary, and 1 R v R primary where the primary was decisive.

In 2008, there are 57 D v. R races and 5 D vs. independent/3rd where the primary is redundant*, and
18 D and 9 R races where the primary and general are meaningless*.  In 2008, there are 2 D v D races, and 1 races with 3 D's only, and 1 race with 5 R's only.  In 2006, these would have been decided in the primary, but in 2008 they will be open to all voters without having to deliberately cross over.

So the number of races that were somewhat degenerate in form has dropped from 102/124 (82%) to 89/126 (72%).  Because there are an odd number of LD/senate seats, one election always has one more race, and there is also a special election for the remaining 2 years of the LD 34 senate seat.

* Washington has a formal system of write-in candidacies, and counting write-in votes, but does require a candidate to have 1% of the vote to advance to the general election.  A write-in candidate can actually declare his party preference, and if he does finish in the Top 2 and receive 1% of the vote, that preference will be shown on the general election ballot.  Given that many voters will skip a race with an unopposed candidate, someone with a minimally organized write-in campaign could secure a place on the general election ballot, which would make the general election an actual contest.

It is quite possible that there will be more participation in the primary by independent and non-partisan voters.  While in the past these voters could participate, they may have been more reluctant to interfere in what was formally a partisan primary.

There will always be many legislative seats that are uncontested because they are held by an incumbent.  The voters already elected them once, and unless the voters decide they made a mistake there is really no recent to expect a different result, and it costs money to run a political campaign, and even more for a successful one.  If a seat is open due to retirement or term limits, it will attract a bunch of people who see an opportunity.  But afterwards, the challengers may be more lackluster.

Washington's system of two representatives per LD, but separate positions may also reduce the number of contested elections.  It appears that it is fairly common practice for one party to contest only one position in a LD.  There are a certain share of voters who will deliberately split their vote.  If they are forced to vote for a D in one position, they pay pick the R in the other position to balance their vote.  If they had to choose between a D and R for both positions, they might still split their vote, but these ticket-splitter might not do so in a consistent fashion, which will simply result in a more dominant party sweeping the election.  In 2006, 14 of 49 LD's had that pattern.  In 2008, there are only 6 such candidates.

Of course it is possible that some of the additional candidates are attracted by the novelty of the system, and may not be serious candidates.  This may include candidates who filed with a preference for the Republican or Democratic party, but who did not have much connection to the formal party.

[jimrtex]

Washington has a fairly formal system for handling write-in votes.  Since most voters vote by mail, it is relatively easy to cast a write-in vote (it's also not like you are standing in a voting booth after standing in line for 15 minutes, and trying to figure how to do a write-in on a voting machine).  On the other hand, because many people vote early, they may not be aware that anyone is running as a write-in candidate.

In 1994, Linda Smith won the GOP nomination for Congress, and then went on to be elected.  The original GOP candidate had informally dropped out.  She also had been a state senator and been involved in several initiative petition drives.  She sent out mail to everyone in the district explaining how to do a write-in vote.   The election was done under the blanket primary system, where both Democratic and Republican nominations were on the same ballot.  In each race, voters could vote for any candidate, but the winners were determined on a party basis.  So a Democratic voter might decide to vote for a Republican candidate.  This could be either because they wanted that person to be elected, or because they wanted that candidate to be the Republican nominee, or because they wanted a candidate that they perceived as being weaker to be nominated, or perhaps simply because there was a choice on the Republican side, and only one candidate on the Democratic side who was sure to be nominated.  Since there wasn't a Democratic race for the nomination, some people may have voted for her because of the novelty of write-in voting.

In Washington, write-in candidates may declare they are running for office.  The total number of write-in votes is counted, but they are not counted for individuals unless there is a possibility of it changing the result, either by causing a write-in candidate to be nominated or elected; or by causing a result change among those who are on the ballot.  This would include overvotes, undervotes, and write-ins of the following form:

[X] John Smith
[X] Write-in John Smith

[ ] John Smith
[X] Write-in John Smith

[ ] John Smith
[ ] Write-in John Smith

all of which are valid votes for John Smith, but not machine-countable.

In Washington, originally the blanket primary was only used for the Republican and Democratic (or any other major parties) primaries.  Candidates of 3rd parties and independents would be direcly by party conventions or petition, and then have their name placed on the general election ballot.  Around 1970, a satirical party nominated a slate of candidates and received about 3% of the vote.  After that, the legislature required that nominees of 3rd parties and independents appear on the primary ballot, and that they receive 1% of the vote in order to appear on the general election ballot.  So on each office, a voter could participate in the Republican primary, the Democratic primary, or support the nomination of a 3rd party or independent candidate.  Since voters may tend to choose a race where they think there is meaningful choice, they may skip over the 3rd party nominees.  It might be hard to get supporters of a 3rd party to vote, simply to rubberstamp the nominations.

After the blanket primary was ruled unconstitutional, Washington switched to a Pick a Party Primary, where a voter receives a ballot similar to the the blanket party ballot with all (major) party candidates on the ballot.  But a voter had to mark on the ballot which party he was voting for.  Only votes for that party's candidates would then be counted.  In effect, a voter had to choose which party he was a member of, but that would then be a secret choice.  The Pick a Party Primary included a 1% provision, and this would be a challenge for the Libertarian Party, which briefly became a major party.  Since they had few contested nominations, voters might not pick their primary to vote in.  In 2004, they did have a contest gubernatorial nomination and this may have attracted enough voters to the Libertarian primary to secure a place on the general election for the statewide candidates.

When the Grange filed their initiative for the Top 2 primary, it was based on the law for the blanket primary, which was still being appealed before the 9th Circuit.  It retained a 1% provision for advancing to the general election.  It also specified that two candidates advance to the general election.  This was done for at least two reasons.  The Grange was trying to preserve the form of the blanket primary, as much as possible, and had generally produced two nominees, one Republican and one Democrat, for the previous 60+ years.  In addition, the US Supreme Court, in overturning the blanket primary in California, had suggested that a non-partisan primary in which the field was winnowed to 2 (or some other number) of candidates would be constitutional - since that would removed the unconstitutional feature of the blanket primary, where Democrats could participate in the nomination of the Republican candidates and vice versa.

The provision for two candidates advancing to the general election. along with write-in voting (and formal write-in candidates) means that there is the possibility in races where only one candidate filed for the primary ballot, the two candidates might be on the general election ballot.  The write-in challenger would need to get 1% of the votes in the race.  Since many voters will skip a race with a single candidate, especially if that candidate had expressed a preference for a party they did not care for, the 1% threshold will be an even smaller share of those who voted in the primary.  In a legislative race this might means 100 or 200 write-in votes could secure a place on the general election ballot for a write-in candidate who was somewhat organized.  It will also probably require hand counting of all ballots in those races where only a single candidate filed, since it is possible that 20 to 40% of the ballots skipped the races, and would have to be examined to make sure nobody wrote in a name.

The rules that the Secretary of State promulgated for the Top 2 primary do recognize the possibility of write-in candidates.  A declared write-in candidate may indicate a party preference, and if he advances to the general election, his preference will appear on the general election ballot.  In addition, a write-in vote for such a candidate is not required to include the party preference of a candidate to be valid.

Write-ins and write-in candidates are also possible in the general election.  However, a candidate that was eliminated in the primary, whether they were on the ballot or a declared write-in candidate, may not be a declared write-in candidate in the general election, and any casual write-in votes for such persons are void.

Washington also has true non-partisan elections.  These are used for judges and also the Superintendent of Public Instruction.  In these races, there is no party designation on the ballot, and any voter may vote.  Further, if a candidate receives a majority of the vote in the primary (held in the same election as the partisan primary), they are elected in the primary.  Otherwise the top two advance to the general election for a runoff.  Parties do endorse candidates in the non-partisan elections.

Hawaii does not have write-in voting.  This was challenged before the US Supreme Court, which essentially ruled that write-in voting is an essential part of voting in the US, but that Hawaii's electoral system provided a functional equivalent.  Elections in the United States were originally entirely write-in.  Voters would write the name of their favored candidate on a piece of paper and drop it in the election box (or in some cases vote in public or by voice).  Parties could still endorse candidates, but these were more like suggestions.  As a convenience to their partisans, parties would provide pre-printed ballots, though voters could cross out certain candidates or write an alternative name in.

In the early 1900s, the Australian ballot came into use, and quickly became universal.  An Australian ballot is printed by the election officials and includes all the candidates running for office.  It provides better ballot secrecy, makes it easier for a voter too choose among parties, and makes ballot stuffing harder.  But since write-ins were always a part of the election process they were permitted on the Australian ballots.  Later, the introduction of voting machines made it harder to cast a write-in vote, but it never became impossible.

In Hawaii, there is no write-in voting, but it has a late primary (September), has very minimal ballot access requirements for the primary, and all candidates including independents appear on the primary ballot, which is a pick a party format.  If an independent candidate receives 10% of the total vote, or if the leading independent candidate outpolls the nominee of any party, his name will be on the general election ballot.  Since very few voters participate in the 3rd party primaries, this latter standard is easy to surpass.

The Supreme Court ruled that in Hawaii since someone can get on the ballot with little effort, fairly close to the general election, that there was no reason to require write-in voting, especially given that they would require more effort by the voters.

[jimrtex]

Legislative races that will not be D v. R.

LD3 Se  D v I (I had 22%)
LD7 R1  R v R (5 R's in primary)
LD8 R2  R v R (2 R's in primary)
LD11 Se D v D (3 D's in primary)
LD12 R2 R v R (1 R in primary vs. WI R)
LD22 Se D v D (2 D's in primary)
LD27 R2 D v D (2 D's, 2 R's in primary, D vote was 70-75% for other positions)
LD36 R1 D v D (2 D's, 1 R in primary, D voter was 82% for other position)
LD37 R2 D v L  (D v. L in primary, L was 11%)
LD38 R1 D v I  (D v. I in primary, I was 41%)
LD41 R2 D v G (D v. G in primary, G was 19%; R for Senate had 37%)
LD46 R1 D v D (2 D's vs 1 R in primary, 82% D for other position)
LD49 R2 D v I  (D v. I in primary, I was 35%)

The positions where two candidates from the same major party, were either not contested by the other party, or were in districts that were overwhelmingly one-sided (75%+ for one party).

The 5 positions where it is D vs. non-R, were all positions where there was no R candidate.  Two of the independents appear to have credible support.

There are two posititions where a R filed as a write-in candidate and will apparently qualify for the general election ballot.

LD28 R2 D v R (R WI v unopposed D)
LD49 Se D v R (R WI v unopposed D)

There are 24 positions where there was a single unopposed candidate.  It is my interpretation of Washington statutes that all ballots not cast for the single candidate will have to be examined for write-ins, with the 2nd place finisher placed on the general election ballot.

LD1 R1  D v WI
LD1 R2  D v WI
LD7 R2  R v WI
LD9 Se  R v WI
LD12 Se R v WI
LD12 R1 R v WI
LD13 R1 R v WI
LD13 R2 R v WI
LD14 Se R v WI
LD16 Se R v WI
LD19 Se D v WI
LD19 R1 D v WI
LD19 R2 D v WI
LD22 R1 D v WI
LD24 Se D v WI
LD32 R2 D v WI
LD34 Se D v WI
LD34 R1 D v WI
LD34 R2 D v WI
LD37 R1 D v WI
LD38 R2 D v WI
LD40 R1 D v WI
LD41 R2 D v WI
LD43 R1 D v WI

[jimrtex]

Is the ballot counting for the primary going to last another two weeks?

Yes.

Voter Turnout

You will notice that for most larger counties, the next update is at the end of the day on Wednesday.  For smaller counties, it looks like they will wait until Friday.  One outlier is Columbia County which looks like they will wait for the two week period to be finished before they count again.

Voters in King County appear to be favoring non-partisan elections without partisan labels, over the alternative of non-partisan elections with partisan labels for county offices.

[jimrtex]

There are 24 positions where there was a single unopposed candidate.  It is my interpretation of Washington statutes that all ballots not cast for the single candidate will have to be examined for write-ins, with the 2nd place finisher placed on the general election ballot.

No. State law says that you must receive 1% of the primary vote in order to advance to the general election, thus avoiding a situation like this.

Washington law also says that if the number of undervotes and write-in votes is sufficient to change a result, that they be counted.  If there is a single candidate on the primary ballot, it is quite likely that the number of undervotes is greater than 1%.

[jimrtex]

No. State law says that you must receive 1% of the primary vote in order to advance to the general election, thus avoiding a situation like this.

Washington law also says that if the number of undervotes and write-in votes is sufficient to change a result, that they be counted.  If there is a single candidate on the primary ballot, it is quite likely that the number of undervotes is greater than 1%.

Can you rephrase that? I don't quite understand what you're referencing.

There is a fundamental right to cast a write-in vote.  If a machine-counted ballot does not indicate a choice, then the presumption should be that it is a possible write-in vote.  In a race where there is only a single candidate on the ballot, there are likely to be 20 to 30% of the ballots with no choice indicated. 

See for example LD1 where the uncontested House races had about 70% of the vote cast in the contested Senate race.   While many of about 8,000 voters will simply have skipped the House races, the presumption should be that every one of the ballots has a write-in vote.  Washington law requires possible write-in votes to be tallied if it would have an effect on the outcome.  If there is a person who receives 1% of the vote, then they would be placed on the general election ballot.  This would be a change in the outcome.  1% is only about 200 votes.

[jimrtex]

Why would an undervote be presumed to be a write-in?  The machine automatically distinguishes those from write-ins.

[X] Rossi
[  ] Write-in ______________

[X] Rossi
[X] Write-in Rossi                   

[  ] Rossi
[X] Write-in Rossi                   

[  ] Rossi
[  ] Write-in Rossi                   

[  ] Rossi
[  ] Write-in ______________
Rossi for Governor

Are all valid Rossi votes.  In a machine tally, only the first will be counted.  The 2nd will be counted as an overvote, the 3rd as a write-in vote.  The 4th and 5th will be counted as undervotes.

Washington law is explicit that if the total number of machine-tallied write-in votes, undervotes, and overvotes for an on-ballot candidate could change the outcome of an election, that the ballots be hand counted, since it is possible that all were marked in some way that was a valid vote.

In the case of a person who has not filed for the offfice, the following are valid votes (assuming they can figure out who Alcon is).

[  ] Rossi
[X] Write-in Alcon                     

[  ] Rossi
[  ] Write-in Alcon                     

[  ] Rossi
[  ] Write-in ______________
Alcon for Governor

None of these can be machine-tallied.  Again, Washington law says that they be hand counted if it is possible that they would change the outcome.  It doesn't matter that most undervotes are probably of the form:

[  ] Rossi
[  ] Write-in ______________

The law presumes the possibility that they are all valid votes.

[jimrtex]

I'd been told that the Tabulator machine separates out any ballots with extraneous marks

That's correct, but at that point they've already been through a hand check and any ballots with potential problems should have been dealt with. There are a very, very small amount that are rejected by the machines, but those are just dealt with by having a substitute ballot sent in its place.

King County reported roughly 2 to 3% write-ins in every legislative race where there was only candidate on the ballot.

[jimrtex]

I'd been told that the Tabulator machine separates out any ballots with extraneous marks

That's correct, but at that point they've already been through a hand check and any ballots with potential problems should have been dealt with. There are a very, very small amount that are rejected by the machines, but those are just dealt with by having a substitute ballot sent in its place.

King County reported roughly 2 to 3% write-ins in every legislative race where there was only candidate on the ballot.

Yes, however, they only count who those write-ins were for if someone has filed a Declaration of Write-In candidacy or if it is clear to election workers that a candidate may make the general election ballot.

RCW 29A.24.311 is clear that write-in votes are valid for undeclared candidates, so long as the office being sought can be determined from the ballot.

RCW 29A.60.021 (1) Says that a write-in vote for an undeclared write-in candidate in a general election are not valid if the candidate had lost the primary either as an on-ballot or a declared write-in candidate.  But this is a meaningless restriction, unless write-in votes for other undeclared write-in candidates are valid.

RCW 29A.60.021 (4) says that write-in votes are to be tallied if the number of undervotes and write-in votes is greater than that of an apparent qualified nominee.  While one could literally interpret this to mean that write-in votes don't have to be counted unless it appeared that there were two on-ballot canidates it would produce the following perverse outcome:

Candidate A and B are on the ballot.  Candidate C is not on the ballot.

Candidate A gets 96% of the vote.  Candidate B gets 1% of the vote.   There are 2% write-in votes.  These are examined and it is found that Candidate C has been nominated.

Candidate A gets 96% of the vote.  Candidate B gets 0.5% of the vote.   There are 3% write-in votes.  Since Candidate B does not have 1% of the vote he is not nominated, and therefore for the write-in votes are not tallied.

[jimrtex]

All ballots are looked at in person before being sent through a machine to avoid a situation like this. There are tables and tables of workers who look at ballots that have been filled out incorrectly and fill out substitute ballots that can be read by the counting machines.

In other words, there is no point whatsoever to machine counting.

Statewide Standards on What is a Vote (PDF file)

[jimrtex]

you can quote the RCW's until the cows come home, but I'm telling you right now that the way the counties are interpreting the laws is the way I'm describing them. And that's all that really matters.

It doesn't matter whether the counties interpret ballots in a manner that is consistent between counties and complies with the law?

[jimrtex]

Washington Legislature History

You might be interested in this.  Starting on page 157 (162 of the .PDF file) you'll find legislative district maps since statehood. 

There is also some stuff on the territorial legislature.  Early on the the population centers were Vancouver, Clarke County (sic until 1924); and Olympia, Thurston County.  The next boom was in Walla Walla.  Tacoma and then Seattle were somewhat later developers.

[jimrtex]

Awesome stuff Jim.  I'll look at it more in the morning.  I knew that Grays Harbor County used to be Chehalis County, but not about Clark-with-an-"e."  Growth redistricting should be cool, too.

I thought it pretty weird that Clarke was spelled with an 'e', especially since I guessed it was name for William Clark.  So I did a little bit of googling.  It was originally named Clark, when Oregon Territory created two counties north of the Columbia River.  The other was Lewis.  The 1850 Census for "Washington" is based on these two counties.

In 1853, an error by a clerk transformed Clark into Clarke County, and the name stuck until 1924.  I'm not sure if this was before or after Washington Territory was split off.

I have a census history of county population which shows the source counties for new counties, and it showed that all the counties from Whatcom to Thurston were formed from both Lewis and Clark counties.  So I spent some more time trying to find some old maps.

Incidentally, there is a musical group Lewis & Clarke, but they claim to be named after C.S. Lewis and Arthur C. Clarke.

I eventually found this animation:

Washington Counties 1844 — 1911

As you can see, Lewis was the western county, and Clark was the eastern county, but the boundary was not the Cascade crest, but a north/south line to its west.   Soon counties were created from Lewis County on either side of Puget Sound and the eastern boundary of Whatcom, Island, King, and Pierce counties was established on the crest.

The SoS office has a bunch of historical maps (high quality scans), as well as an online exhibit of redistricting, which it includes oral histories, newspaper accounts, etc.

[jimrtex]

You must be logged in to read this quote.

(Powell seems to be a secret Democrat or Democratic-leaner at least, but is squeaky-clean of affiliation on paper.)

How common is the term "Banjo Belt?"

[jimrtex]

How common is the term "Banjo Belt?"

You've totally lost me.

I was watching the committee discussion of the bill to split legislative districts into two house districts.  The sponsor Rep. Dunshee from LD 44, apparently previously represented LD 39, and was using that as an example of how the redistricting commission usually managed to satisfy the caucuses by generally not pairing incumbents (Dunshee was one of two exceptions in 2001), and gave as an example Rep.Kristiansen who lived south of him.

He made a reference to rural Whatcom County as being the banjo belt.  About 10 minutes later, another representative showed up, and complained about the insult to his constituents.  I think it was Rep.Overstreet.  I thought at first it was a joke, that he acting upset.  But it was apparent that he was demanding an apology for all the musically-inclined hillbillies of Whatcom County.

BTW, the district split bill is going nowhere.  Every representative was already calculating how it might effect him or her (one Seattle representative said that the district line would have to be drawn down an alley as another representative balcony overlooked her back yard).

[jimrtex]

I think it's a Deliverance reference.

Whatcom County is "Banjo Country"

It includes a clip from the movie.

[jimrtex]

Prop. 1 didn't do too badly in the second drop tonight, but Richard Conlin did...he may be toast.  Sawant needs about 52% of the remaining vote.  Amazing.

Also, it looks like Seattle's Prop. 1 (public financing) has a shot: http://slog.thestranger.com/slog/archives/2013/11/08/sawant-surges-in-latest-ballot-drop-seatac-prop-1-bleeding-slows

Have we ever seen late ballots this different?

Why are both Murray and Conlin dropping?   On election night Murray had 56% and has dropped down to 53%.  What is the correlating factor.

[jimrtex]

I've been looking through old State House results and noticed we used to use a system like the one currently used in New Jersey. Both State House members were elected in the same race and each voter got two votes. It seems 1964 was the last year this system was used. After we ditched it, we started "nesting" the State House districts within the State Senate districts (each State House district was half a State Senate district; 32-A and 32-B, for example). I'm not sure when this practice stopped.

Anybody know of an article or book that talks about why these changes were made? I tried Google but gave up pretty quick.

The Consitution literally says that representative districts may not be divided between senate districts, that there not be more than 99 representatives, and that the number of senators be between 1/3 and 1/2 the number of representatives.

After statehood, the legislature created districts that were largely based on counties.   Senatorial districts were either a county; two or three counties; or several districts within one county: King (6), Spokane(3+), Whitman (2), Whatcom(2), and Walla Walla (1+).  Spokane and Walla Walla were exceptional in that part of Spokane was paired with Stevens (this was before Pend Oreille and Ferry were created), and part of Walla Walla was paired with Franklin and Adams.  

Representative districts were either a whole county, or coincident with a senate district, and elected either 1, 2, or 3 representatives.   In the larger counties, the senate and representative districts were coincident, and usually 2 representatives were elected.   For example, King had 6 senate districts, 6 representative districts, with 5 electing 2 representatives, and 1 electing 3.  So it was immediately established in some parts of the state to have a senator and usually 2, but sometimes 3 representatives elected from the same territory (the numbering of the two sets of district numbers were independent, so while a senate district might be 27, the identical territory could be representative district 35.

Garfield, Asotin, and Columbia were a senate district, while Garfield, Asotin, and Columbia were each a representative district electing one representative.   Clarke (now Clark) and Skamania were a senate district, Clarke was a representative district electing two representatives, and Skamania elected one representative.  So in the more rural parts of the state, there was one senate district and several representative districts.   I wouldn't really call them subdistricts since the representative districts were for individual counties within a multi-county senate district, and likely had great differences in population.

When new counties were added (Chelan, Ferry, Benton, Grant, and Pend Orielle were created after statehood) they were assigned to a senate district and given a one-representative house district.

There was a new apportionment after the 1900 census, but the same format was mostly followed.  King (8), Pierce (5), Spokane (5), Whitman (2), Snohomish (2), Whatcom (2) and Walla Walla (1+) had multiple senate districts.

There were just minor tweaks until 1930, when reapportionment was accomplished via the initiative.   This initiative eliminated the difference between senate and representative districts by eliminating the single-county representative districts.   It created 4 new senate districts in King, and one in Yakima, and somewhat rearranged multi-county senate districts elsewhere.  By eliminating the one representative per county arrangement, and increasing the number of representatives to the constitutional maximum of 99, they found enough representatives to transfer to King, without radically changing things elsewhere.  After this redistricting there were 46 districts, with 39 electing two representatives at large, and 7 electing three representatives at large.

In 1956 another initiative was passed, but in 1957 the legislature overturned it (initiated legislation can be amended by a 2/3 super-majority in both houses.   The initiated version created 3 more senate districts, two in King and one in Clark, bringing the total to 49, the maximum permissible under the constitution.  49 districts elected 2 representatives, and one elected three representatives.  The districts were largely based on counties, but there was somewhat more splitting of counties, including counties like Benton, Yakima, Skagit, Snohomish having portions attached to other counties.

The version passed by the legislature kept the three new senate districts in King and Clark counties, but elsewhere for the most part restored the pre-1957 senate districts.  They then gave three representatives to 6 districts and 1 representative to 5 districts.  The other 38 districts had 2 representatives.  

To be continued.

[jimrtex]

continuing...

There was then an attempt to create a different redistricting system (which was defeated), and litigation (Thigpen v Meyers), and a 1962 attempt to apportion by the initiative, which was narrowly defeated (it won big in King, and some other large counties, but was thumped elsewhere.  Garfield voted 42 Yes to 1187 No.   The federal district court had deferred action until after the initiative failed, but then ruled that the legislative districts were unconstitutional.  The 1963 legislature did not act, and the district court enjoined holding elections on the existing lines.  This was stayed by the SCOTUS, pending its decision in Reynolds v Sims.  Seven days after Reynolds v Sims was announced, the SCOTUS rejected the appeal in Thigpen v Meyers.   The 1964 elections went ahead on the existing lines, but the district court enjoined the 1965 legislature from enacting any legislation, including a budget until they had redistricted.

Dan Evans was newly elected governor, while there was a Democratic legislature.  It took the legislature 47 days to pass a redistricting bill.  Slade Gorton who was in the legislature at the time is credited by some with getting a bill pass that was more favorable to Republicans, while the Democrats were more concerned with protecting individual incumbents by keeping them in separate districts, even it was not necessarily favorable to their election.

The 1965 legislation created 49 senate districts and 56 representative districts.  There were 14 single-member representative districts (2A-2B, 5A-5B, 8A-8B, 9A-9B, 11A-11B, 16A-16B, and 32A-32B).  These are described in the law as representative districts, and then the encompassing senate district, such as 2, is described as being constituted from representative districts 2A and 2B.

For example, 2A was Okanogan, while 2B was Ferry, Stevens, and Pend Oreille.   Senate district 2 was constituted from 2A and 2B.

In some cases these single member districts appear intended to ensure representation in an area.  For example, before 1965, District 10 was Columbia, Garfield, and Asotin, and had one senator and one representative; and District 11 was Walla Walla and had one senator and two representatives.  After 1965, 11 A was Columbia, Garfield, and Asotin, and part of Walla Walla, while 11B was most of Walla Walla (including the city of Walla Walla).   This avoided the risk of all 3 legislators coming from Walla Walla, which had a majority of the population.

Other splits included 8A part of Yakima, and 8B part of Benton; and 9A Whitman, and 9B Lincoln and Adams; and 16A part of Benton and 16B Franklin.

Districts 5 in Spokane and 32 in Seattle were also split.   5A and 5B appear gerrymandered with interleaved fingers, so I suspect it was intended to keep incumbents apart.  Or perhaps 3 representatives ended up in a senate district, and they wanted to force the two from an existing district to face off.

District 32 is really an anomaly, since the other 14 districts in King County were simple 1 senator, 2 representative districts.   If you looked carefully at the boundaries, and the incumbents, etc., it probably makes very good sense, but it certainly is an oddity, while the 5 splits in rural territory are obviously trying to maintain representation for distinct areas.

The 7 split districts were in a sense a throwback to the pre-1930 practice of having at least one representative district for every county, but now they were for distinct areas, and the numbering was coordinated.

The other 42 senate districts continued the practice of at-large elections for representatives.  41 elected 2 representatives and one in Whatcom, elected 3 representatives.   Previously, Whatcom had 2 senate districts electing 2 and 1 representatives (3 total).

In 1965, legislation was also passed providing for at large representatives to be elected by position.  This was separate from the redistricting, and I don't know why it was done.  The current statutes for election-by-position is general, and applies not only to the House of Representatives, but cities and school districts, etc (chartered cities and counties may be able to disregard this).   I found the 1965 law based on references in the current statutes.

It may have been that they wanted to provide more structure to the elections, where many representatives would be forced to run against each other.   Or it might have been simply following the practice for other at-large elections.  If OMOV was being applied to commissioner elections, it might be easier to switch to at-large elections where incumbents wouldn't have to be matched up since they could run for separate positions rather than to change the district lines.

Before 1965, districts 1 through 17 were east of the Cascades, with 17 including part of Clark, plus Skamania, and Klickitat.

Districts 1 and 10 were moved to King and Kitsap as a result of the 1965 redistricting; 2 and 11 to King and Pierce as the result of the 1970s redistricting; and 5 to King as a result of the the 1990s redistricting.   17 has been captured by Clark.  Otherwise, Washington has had a fairly continuous history of district numbering for over a century (with the King County districts in the 40s having been added later, along with 49 in Clark County).

The SOS elections data base shows the single member representative districts as being position elections (eg 2A is shown as 2 position 1, and 2B is 2 position 2), but if you look at the county returns they are correct.   32A and 32B had significant differences in number of votes cast.

The 1970s redistricting was done by a federal court (actually a special master) after the legislature failed to redistrict.   There were 49 senate districts, and all had two representatives.  Under modern OMOV requirements, having three representatives for some districts is impossible.   In the 1960s it might have been believed you could give an area entitled to 1.3 senators and 2.6 representatives: 1 senator and 3 representatives, where they somewhat compensated each other.

The 1980s redistricting was done by the legislature and they created four single member representative districts: 19A was Cowlitz and part of Wahkiakum; while 19B was Pacific, part of Grey Harbor, and part of Wahkiakum; while 39A and 39B were parts of Snohomish County.

The 19 split recognizes the difference between the coast and the lower Columbia, particularly Longview; but the OMOV interpretation was strict enough to require the division to occur in Wahkiakum which doesn't have much population.  One reason for not creating single-member representative districts is that it would probably require more splits of small counties.

The redistricting commission was provided for in 1983, and has done redistricting for the last 3 decades (1990s, 2000s, 2010s).  I don't know if the commission made an active decision to not have single member representative districts or not.  But there haven't been any created.

There have been bills introduced that would switch to having single member representative districts.   They have at best received committee hearings that can be characterized as cold (ie, if we don't ask any questions, we can go on to the next bill in 5 minutes).  One year, one representative asked about how the districts could be divided, noting that the other representative from her district lived across the alley and could look down into her backyard from his second story (whether he has binoculars, and she has a swimming pool, I do not know, but that is how I visualize it).

Sources:

Washington Redistricting History

Legislative History links

'History of the Legislature 1965-1982' has lots of interesting stuff, with the section on the 1865 including details of the process.

'Members of the Washington State Legislature' has the legislators each year by district, along with where the district was (I don't think this is always 100% accurate).  It also has an interesting section on redistricting history.   This includes the laws that actually implemented redistricting, which can be found here:

Index to Session Laws

The session laws include all the bills that became law each session of the legislature. 

Links to past voter Guides, not the most obvious format

I know that 1964, 1966, and 1972 have district maps, including details for the cities.

[jimrtex]

We just pulled ahead by 41 votes!

https://electionsdata.kingcounty.gov/election-results-nov/three/City%2C%20L-Z/Seattle/City%20of%20Seattle%20Council%20Position%20No.%202

Can someone run the math as to how many votes are left to be counted?

Just short of 160K votes cast in the council race of 186K Seattle ballots counted.  There are around 213K ballots that can be counted, with another 4K possible.  The number of unverified ballots dropped from around 7K to 4K with the last update on Tuesday.  The number of unverified had remained pretty constant until this last update.  I don't know whether that means they've started going back for a second look, or that they have made their final decision.

Sawant had 47% in the first 2/3 of ballots counted, and 55.6% in the remaining 1/3 (and this keeps climbing).

Sawant should put up another 2000 votes or so.

The council race has 14.6% undervote.  The mayors race around 4.0%, and the GMO initiative around 1.5% (countywide).   Were the state initiatives top of the ballot.  The mayoral undervote sounds kind of high, like it was buried.   Were the King County races above the city races also?

[jimrtex]

Huh, this is another really cool data set: http://www.sos.wa.gov/elections/absentee_stats.aspx?y=2000

ETA: Something is weird with the Pierce numbers for 2011 and 2012.

The number of "absentee ballots" cast in 2011 and 2012 is identical to that for 2010 - the last year when there was a meaningful difference in Pierce County.  Making the SOS sends out a spreadsheet populated with old numbers, and Pierce County is not updating.  Maybe the Auditor goes, "Huh?, why do I put the same number in both columns.", and someone over at the SOS office says, "Pierce County is always different because they still have in person voting".

[jimrtex]

King County is perplexingly awful on a lot of elections issues (drop-box availability, precinct results, integration with state databases) for no apparent reason.  I'm pretty sure they have the resources to do a decent job on accessibility.  They just seem to not care much.

They were forced to automate for practical reasons before other counties, and also can afford to have regular election staff.   Because they have such a large share of the population, there is probably conceit as well.

In Garfield County, the auditor probably personally counts the ballots, and anything the SOS does saves money for the county.

All states are required by federal law to have a statewide registration database.  In Texas, the smaller counties log in directly to the state database.  The larger counties which had already automated, do a batch download once per day.

The thing that is bothersome about King County is that they take their time counting the votes.  They had verified signatures on most ballots within a day of the election.  But they just run some through the scanners each day.   King County has 45,000 ballots left to count, about 80% of the statewide total.  Pierce County has 462, and Snohomish around 4,000.

They would probably claim that it either permits them to use their regular staff, or to do temporary hires for a couple of weeks instead of two days.  But it gives the outside appearance of manipulating the results (remember the Gregoire-Rossi race).