The November 2018 Election in Ketchikan, Alaska: Maps and Analysis

I recently wrote about the 2018 general election as it took place across Alaska as a whole, and I judged that it wasn’t a “red wave” at all, but rather a slight recalibration of the state political scene.

The results in my hometown of Ketchikan confirm that overall conclusion, but I’d still like to go into detail to take a look at how my community and our close neighbors voted in the election.

Ketchikan is the main population center of Alaska House District 36, with approximately 75% of the district’s population living on Revillagigedo Island in the Ketchikan Gateway Borough. The other quarter of the district’s population lives in Wrangell (on Wrangell Island), Metlakatla (on Annette Island), Hydaburg (on Prince of Wales Island), or on the nearby parts of the Alaska mainland, notably in the small town of Hyder (population 87 in the 2010 census).

According to the latest (still unofficial) election report, voter turnout across District 36 was just over 50%. If you compare that to the last five general elections, that turnout number seems low: In 2016 it was nearly 62%, in 2014 it was nearly 56%, in 2012 just over 58%, in 2010 it was lower at 49%, and in 2008 it was almost 63%. However, these percentages belie the fact that the number of registered voters is now higher than it’s ever been! And yes, I did go back and look at all the general election records in Alaska state history in order to confirm that fact. Here are the results of that research:

available voter turnout information in all elections since statehood for Ketchikan’s electoral district

There’s clearly been a fair amount of ebb and flow in voter turnout over the decades, as well as the total number of voters in the electoral district. The communities included in the same district as Ketchikan shifted back and forth over the decades as well: Often we were linked with Metlakatla, almost always parts of Prince of Wales Island, sometimes Wrangell, and at one point even Petersburg. Turnout has been low in the last few elections, but not as bad as the early 2000s. Nevertheless, it’s clear that there are now more registered voters in the area than ever before, no doubt thanks to the automatic voter registration system initiated in 2016, which now registers adult Alaskans to vote when they apply for the Permanent Fund Dividend.

As for the results of the 2018 election, let’s start with how District 36’s relative leanings compared to the state as a whole:

For Alaska’s representative in Congress, District 36 voted 45.03% for independent challenger Alyse Galvin and 54.6% for longtime Republican incumbent Don Young. The state overall gave Galvin 46.5% and Young 53.08%—so District 36’s results for that race were just over a percentage point off the average.

In the gubernatorial election, District 36 voted 55.21% for Republican Mike Dunleavy, 39.56% for Democrat Mark Begich, and about 2.5% each for Bill Walker (the independent incumbent who suspended his campaign) and Billy Toien (the Libertarian candidate). Meanwhile, Alaska as a whole gave Dunleavy 51.45%, Begich 44.4%, and Walker and Toien each about 2%. Therefore, while District 36 was slightly more supportive of Young than Alaska as a whole, it was much more supportive of Dunleavy, and was in fact more supportive of Dunleavy than Young, unlike the state overall.

The only other statewide election was on Ballot Measure 1, (called the Stand for Salmon Initiative), which was resoundingly defeated 62.33% to 37.67% statewide. In District 36 the result was an extra few percentage points more lopsided, 64.89% to 34.11%.

All of these results make District 36 look somewhat more conservative and Republican-leaning than Alaska as a whole. However, at the same time, the district reelected independent state representative Dan Ortiz to a third term by a wide margin. Ortiz spent his first term in the legislative minority, declining to join the Republican majority coalition. After his reelection in 2016, Ortiz spent his second term in the new majority coalition made of 17 Democrats, 3 Republicans, and 2 independents—a coalition that mostly supported Governor Walker’s fiscal policy. (I discussed this coalition and its future in my last article.)

Given how unpopular Walker’s policies ultimately proved to be with many Alaska voters, and given Ortiz’s alliances with Democrats in the legislature, it may seem surprising that Ortiz was reelected in a landslide in a Republican-leaning district. Nevertheless, Ortiz received 59.77% of the vote, compared to Republican challenger Trevor Shaw’s 39.22%. The only result more of a landslide than Ortiz’s victory over Shaw was the defeat of Ballot Measure 1.

Part of the explanation, clearly, is that Trevor Shaw was a flawed candidate. Shaw had an extremely rocky tenure as a member and president of the Ketchikan Gateway Borough Board of Education, which ended with Shaw’s resignation in August after a petition had been filed to recall him from office. It is interesting to consider how a Republican candidate with less “baggage” than Shaw might have fared against Ortiz in this election, but no such candidate ran.

Now we should look at District 36 election results precinct by precinct. In addition to the seven voting precincts of the Ketchikan Gateway Borough that I described in a previous article, there are three other official precincts—Wrangell, Metlakatla, and Hydaburg. However, Hyder appears to be the only community whose votes were labelled as “Early Voting” in the data, so their preferences are distinguishable as well. (Those who voted early in Ketchikan, like me, had our votes designated “Absentee.”) With Hyder differentiated, there are effectively 11 precincts in District 36 that we can analyze, and I’ll refer to Hyder as a precinct from now on.

The U.S. House Race

In the Galvin-Young race, there were six precincts that favored Galvin over Young—Ketchikan №1, Ketchikan №2, Saxman, Hydaburg, Metlakatla, and Hyder. However, Galvin’s margin of victory in most of these areas was slim—only 9 votes (1.9%) in Metlakatla, for example. Young defeated her much more heavily in the other precincts while also winning the absentee and question ballots.

District 36 results of the Don Young-Alyse Galvin race for U.S. Congress

Nevertheless, Alyse Galvin fared exceptionally better against Don Young in District 36 than some previous challengers: In 2016, challenger Steve Lindbeck only won in Hydaburg (by eight votes) and Hyder (by one vote). In 2014, Forrest Dunbar did not win a single precinct in District 36, losing Hydaburg by one vote and tying with Young in Hyder (seven votes to seven). However, Galvin’s campaign did have a lot more money to spend than Lindbeck’s or especially Dunbar’s: According to this site, Dunbar’s campaign only raised about $240,000 in 2014, Lindbeck’s raised about $1.1 million in 2016, and Galvin raised nearly $1.5 million this year, even more than Young. Although ultimately unsuccessful, (like over 30 previous campaigns), Galvin’s campaign clearly inspired many Alaskans who want to replace Young to get out and vote.

The Gubernatorial Race

Turning to the gubernatorial race, Mark Begich fared worse against Mike Dunleavy than Alyse Galvin did against Don Young, winning only five precincts to Galvin’s six. (He lost in Ketchikan №2, and he only won Ketchikan №1 by a single vote.) Clearly, this indicates that there were a fair number of voters in the district who rejected Young for Galvin but still wanted Mike Dunleavy for governor. Perhaps these voters were mostly conservative-leaning voters with a dislike for Don Young and a desire for change in his position, or more centrist voters who had a bone to pick with Mark Begich. Indeed, I personally did not see a single Mark Begich yard sign displayed anywhere in Ketchikan in the lead-up to the election, and I do not believe he had any official campaign staff in town.

Interestingly, however, there were two places where Begich received more votes than Galvin—Metlakatla (250 to 239) and Hyder (29 to 28). These discrepancies indicate that there were at least a few voters out there who decided to support both Don Young (R) as congressman and Mark Begich (D) as governor. (Perhaps they placed a high value on political experience, which Young and Begich have in droves compared to Galvin and Dunleavy.)

District 36 results of the Mike Dunleavy-Mark Begich race for Alaska Governor

The State House Race

A precinct map showing only the results of the Dan Ortiz-Trevor Shaw race for state house would not be interesting: Ortiz won every precinct—his first time doing so in the three elections he has won. However, it would be interesting to make a map showing the culmination of Ortiz’s victories by precinct over his past three elections. Here it is:

In 2014, when Ortiz was first elected, it was an open race with no incumbent: Ortiz won Ketchikan №1, Ketchikan №2, Saxman, Metlakatla, and Hydaburg, while Republican Chere Klein won Ketchikan №3, North Tongass №1, North Tongass №2, South Tongass, Wrangell and Hyder. The race between Ortiz and Klein was extremely close, with Ortiz ultimately winning by only 104 votes (less than 1.5%). In 2016, Ortiz had a much more comfortable margin of support: He only lost Ketchikan №3, North Tongass №1, and Wrangell to challenger Bob Sivertsen, and won the race by 790 votes (10.1%). Then in 2018, he won a majority everywhere with the over-20-point victory against Trevor Shaw.

One tidbit I find particularly interesting is how Ortiz decisively won over the voters of Hyder: In 2014 they voted 8–6 in favor of Klein, but by 2016 it was 16–4 in favor of Ortiz, and then in 2018 it was 35–12 for Ortiz—a smaller win as a percentage, but a larger one in total votes. (I do wonder why turnout in Hyder seems to have increased so much from 2016–2018, though.)

Analyzing District 36 Voters

I also wanted to investigate if there was any interesting interplay between the gubernatorial race and the state house race. I found that there were five precincts in which Mike Dunleavy won the governor’s race and won a higher percentage of the vote than Dan Ortiz: These are the consistently more conservative precincts of the district—Ketchikan №3, North Tongass №1, North Tongass №2, South Tongass, and Wrangell. Then there were the five precincts where Mark Begich won the governor’s race, but Ortiz received a higher percentage of the vote than Begich: These are the more liberal (or merely less conservative) precincts—Ketchikan №1, Saxman, Metlakatla, Hydaburg, and Hyder. (There were no precincts in which Begich won a higher percentage of the vote than Ortiz.) Lastly, however, there was just one distrct where Dunleavy won the governor’s race but Ortiz received a higher percentage of the vote—Ketchikan №2. Here’s the map showing this information:

If you want to think about the aggregated results of these two elections for a moment, consider the different groups of voters that must have existed:

  1. Group RR voted Republican twice, for Mike Dunleavy and Trevor Shaw. (This might have been the largest single group, since most all of the 39.22% who voted for Trevor Shaw almost certainly voted for Dunleavy.)
  2. Group DI voted Democratic and independent, for Mark Begich and Dan Ortiz. (This might also have been the largest single group, since Begich received 39.56% of the vote in his race and most all of the Begich voters are likely to have voted for Ortiz.)
  3. Group RI supported Dunleavy for governor but wanted to keep Ortiz as their representative. This must have been the third-largest group of voters, and I imagine that most people in this group are likely characterized as relatively conservative or Republican-leaning but were ready to support Ortiz because of personal connections, Ortiz’s track record and experience, disapproval of Shaw, or some combination of the above.
  4. Group DR supported Democrat Mark Begich for governor but wanted a change of representative, voting for Republican Trevor Shaw. I believe this group must have been quite small, but I do know that it did exist: In an early, incomplete counting of the votes from Hyder, when only eight votes had been tallied, Begich was beating Dunleavy 5–3 while Ortiz was tied with Shaw 4–4. If there was at least one voter in Hyder who was in Group DR, there must have been a few elsewhere too!
  5. Lastly, there are a few other groups of voters that must have existed, because of the 5% who voted for either Bill Walker or Billy Toien for governor, and the 1% who wrote in a name instead of voting for Ortiz or Shaw. However, these groups are pretty small and I feel it’s ok to set them aside in this analysis.

In the map above, the five red precincts must have had Group RR as the plurality, giving Dunleavy a stronger win than Ortiz. However, there were still enough Group RI voters to give Ortiz the win over Shaw. Group DI was strongest in the five blue precincts, and there were still a number of RI voters around who gave Ortiz a stronger win than Begich.

In Ketchikan №2—the odd precinct out—there were 464 Dunleavy votes, 381 Begich votes, 553 Ortiz votes, and 303 Shaw votes. These numbers indicate a very large RI group. If all 381 Begich votes were DI, that would indicate there were around 172 RI votes (553–381). If all 303 Shaw votes were RR, that would indicate there were around 161 RI votes (464–303). With all the other variables involved (Walker, Toien, write-ins, and the possibility of DR voters) those numbers are pretty close. Approximately 165 RI voters out of approximately 870 votes in Ketchikan №2 means an RI group of around 19%, a few points higher than the 16-point spread across District 36 between Dunleavy’s win (55.21%) and Shaw’s loss (39.22%).

To make a long story short, I think this interplay would indicate that Ketchikan №2 has a strong claim to the title of “swing precinct” in District 36.

However, I did want to check my conclusion by making a table with the data for every precinct, comparing the percentages each candidate won in the gubernatorial and state house races. Here’s my complete data table:

data table for state house and gubernatorial races (figures that are above average for that candidate are highlighted)

I notice one major takeaway from the table above: The results of where Ortiz, Shaw, Begich, Dunleavy each received higher-than-average support all align pretty well with each other across the races and with the usual patterns of the more conservative and more liberal precincts. However, there is one exception where both Dunleavy and Ortiz did above average—South Tongass, which is Dan Ortiz’s home precinct.

Then I used that data to make a comparison between the margins of victory in each precinct—or rather the margins between Ortiz and Begich support and Dunleavy and Shaw support:

data table for relative levels of support for Dan Ortiz vs. Mark Begich and Mike Dunleavy vs. Trevor Shaw

By looking at the “Ortiz-Begich margin” and the “Dunleavy-Shaw margin,” I’m really trying to isolate these voters who are open enough to the center and center-left to support Ortiz, but not enough to support Mark Begich, and those who are open enough to conservative Republicans to support Mike Dunleavy, but not enough to support Trevor Shaw. I highlighted the spots where these margins exceeded the average, and they mostly line up with either other. The precincts and categories where these margins were highest were also almost all places where Mike Dunleavy got higher-than-average support—more conservative and Republican-leaning precincts, that is (Ketchikan №3, North Tongass №1 and №2, South Tongass, and Wrangell). Again, the only odd-one-out is Ketchikan №2, where support for Dunleavy was over 2 points below the average and Begich was nearly 3 above, while support for Shaw was over 4 points below average and Ortiz was 4 above.

The fact that these precincts where the margins are highest are almost all the more conservative districts really solidifies my conclusion that the RI group (Dunleavy-Ortiz voters) must be mostly made from relatively conservative or Republican-leaning voters—people who would generally vote Republican on a generic ballot, but who concluded Dan Ortiz is a good representative (or thought Trevor Shaw was a bad candidate). Probably the best example to look at is North Tongass №1, which has a solid claim to the title of most conservative precinct in District 36. One quarter of voters there were part of the RI group: They voted for Dunleavy, but couldn’t bring themselves to vote for Shaw—or they voted for Ortiz, but wouldn’t vote Begich. Support for Dunleavy in North Tongass №1 was the highest of any precinct, while Shaw’s support there was three points below his best performance (Wrangell). That might feel particularly disappointing for Shaw, since North Tongass №1 is his home precinct. In my mind, however, it just goes to show how powerful and significant Dan Ortiz’s personal connections are with so many people in Ketchikan—and those connections just aren’t quite as strong in Wrangell.

Ultimately, perhaps one of the most compelling takeaways from the 2018 election in Alaska District 36 was that, overall, the number of Mark Begich voters and Trevor Shaw voters was almost exactly the same, but a Venn diagram of those people would probably be two completely separate circles.

Meanwhile, a sizable and influential group of voters put themselves in a different circle, voting for both Mike Dunleavy and Dan Ortiz. These are the voters who gave Dunleavy and Ortiz their decisive wins, despite the candidates coming from different sides of the aisle. Now one will be Alaska’s new governor, and the other will continue his service in the legislature. It will be quite interesting to see if and how they can work together.

Please leave a comment below if you found these maps and analysis interesting, or if there are any points you disagree with. Thank you for reading!