The primary reason for drawing new political districts every 10 years, after the decennial census, is to make sure that the districts are balanced — that every representative elected by the people represents the same number of constituents. Complete balance is impossible, but the goal is to get as close as you can.

By that measure — the most important one — Humboldt County’s redistricting process so far has been a failure. At its Wednesday meeting, the county’s Redistricting Advisory Committee will look at the two final maps it has produced, and will forward one or both of them on to the Humboldt County Board of Supervisors.

Both of the final maps are, in fact, less balanced than if the redistricting had not happened at all. Despite the jostling population over the last 10 years, the supervisorial map drawn in 2010 is more balanced, today, than any of the maps on the table in 2020.

The county’s Redistricting Advisory Committee does not know this, because it has never been presented with a map based on the status quo, and it hasn’t asked for such a map to be produced. What’s more, it hasn’t been much interested in the question of population balance, at least in its deliberations thus far.

So we’ll get to all that in a minute, along with a deeper look at the two maps recently produced by the committee. First, a quick note about the shambolic county process to date, with some advice about how to best proceed if you have strong feelings about any of this.

First of all: As mentioned Saturday, the county solicits feedback on these draft maps via a cumbersome questionnaire on its “Open Humboldt” website, which you can find via a link on the its Redistricting web page. At this point in the process, the feedback received via that method seems to have received little to no attention from county staff or members of the Redistricting Advisory Committee.

For instance, when the county released the first set of draft maps — we’re on round two now — it published a different questionnaire. One respondent to that questionnaire very reasonably asked for the mapmakers to break down the Native American population in each proposed district, as they had done for the Black, Asian and Latinx populations. This is important because country redistricting is governed, in part, by the California Voting Rights Act, which seeks to ensure that minority voters are not being disenfranchised through gerrymandering.

However, despite the request, and despite the fact that Humboldt County is home to California’s two largest tribes and has the largest Native American population in the state — it is our third-largest ethnic group — the current round of maps does not, in fact, bother to break out the Native American population in that way, though such data is eminently available to the mapmakers, as it is for the less populous Black and Asian groups in our county.

It’s unclear if anyone involved actually read that respondent’s request. If someone had — either members of the Redistricting Advisory Committee, county staff, or representatives of the various partners hired by the county to work on this — they didn’t only do the respondent and the Native American community a disservice; they potentially left a tiny door cracked open for legal action under the California Voting Rights Act.

Another example will follow, but this is all by way of saying that if past history is anything to go by, people who have thoughts on county redistricting should probably not rely on being heard through the questionnaire — or not only through the questionnaire. The Redistricting Advisory Committee’s next meeting is scheduled for Wednesday at 10 a.m. A link to the meeting should be available on this page. Here are instructions to submit your thoughts to the committee via email.

Next week, at its Nov. 2 meeting, the Humboldt County Board of Supervisors will receive recommendations from the advisory committee and take the issue up on its own. The board will have the final approval over any new map, and it’s required by law to take action by Dec. 15.

Now let’s take a closer look at the advisory committee’s two current draft maps, and let’s also compare them to the current map, which the committee is not considering at all.

Draft Map A2

This map represents the latest iteration of the most eyebrow-raising idea to come out of the Redistricting Advisory Committee — the idea that the Third District, which represents Arcata, should reach down across the Samoa Peninsula to take in a chunk on the far side of Eureka.

The current iteration is the most radical yet. As you can see, if you zoom in, not only is the northwestern corner of Eureka swallowed by the Arcata district — nearly the entirety of downtown Eureka, including Old Town, is subsumed into Arcata as well. 

That’s far from the only oddity of this map. The Arcata Bottoms would no longer be in the same district as Arcata. And the First District, represented in blue in the above map, would cut an insane path across the county from the northeast to the southwest. It would no longer include Willow Creek, but it would include Titlow Hill and everything else south of Highway 299 to the Trinity border. And from there it would snake along to Blue Lake, to Kneeland and Indianola and Cutten, over to Ferndale and all the way down to Shelter Cove.

Given the transparently odd geographies outlined above — odd to anyone with eyes, but particularly odd to people actually familiar with Humboldt County — it’s hard to object to anyone who wants to call Draft Plan A2, here, that dirty word: A gerrymander, a map drawn for political ends. That would particularly be the case for supporters of Natalie Arroyo, the popular Eureka City Councilmember who had already announced her intention to run for the Eureka seat on the Board of Supervisors next year, but who is (conveniently?) drawn out of the Eureka seat entirely, by a margin of one block.

But if it’s a gerrymander, what is its aim? And how is it being accomplished?

The latter question is easier to answer. Before any maps were drawn, the Redistricting Advisory Committee took public testimony on “communities of interest” — places that feel like places, to the people that live in them. When drawing lines, redistricters are supposed to take these communities of interest into account, and to attempt to keep them whole if possible.

During this testimony, a handful of people, almost all anonymous, said that they wanted to see the West Side of Eureka, in which some of the anonymous commenters purported to live, joined up with Arcata. They had two talking points: Renters and public transportation. The West Side of Eureka formed a natural community of interest with Arcata, they said, because both places are home to a lot of people who rent. The talking points fell down a little bit when it came to mass transportation: Sometimes the commenters said that both the West Side and Arcata were in dire need of mass transit options, and sometimes they said that the West Side and Arcata were natural partners because of the superabundance of wonderful mass transit options. (See this document — a record of “community of interest” testimony — for the amusing breakdown of this talking point.)

Still others fully admitted that they did not live in the area in question, but nevertheless held that it formed a natural “community of interest” because it kept both of the urban centers in question out of their own backyards. A writer who identified himself only with by the first name of “Rob” and a handscrawled email address of either “” or “” said that he lived in the Freshwater/Kneeland area, but wished to submit a “community of interest” map for the Eureka/Arcata area rather than his own: 

(Rob McBeth of O&M Industries, the longtime Humboldt County conservative activist, lives in the Freshwater area.)

But what was the goal of the gerrymander, if gerrymander it was? It might not be simply to bounce Arroyo out of contention. Rather, by packing as many left-leaning city-dwellers into the same district as is physically possible, more conservative rural voters would have more of a sway in county politics going forward. 

As should be apparent by now, for whatever reason — perhaps the general paucity of “community of interest” testimony — the Redistricting Advisory Committee took all this at face value and ran with it, and the Eureka-Arcata union has now made it through two rounds of map revisions and looks likely to be passed onto the Board of Supervisors.

This is despite the fact that public opinion on the first round of draft maps, as registered through the county questionnaire system mentioned above, was near-universally against this plan — 60% of the respondents said that the map “didn’t represent their community well, as opposed to 22% who said it represented their community “well” or “adequately.” But as stated above, at its last meeting the committee referred to the questionnaire results glancingly or not at all.

Is there anything good to be said about this map, community-of-interest-wise? Sure. The Samoa Peninsula is a coherent political unit, which at least one person asked for. Rio Dell and Scotia are reunited — perhaps the most glaring fault of the current map — and the southern part of Eureka is united with the eastern part of Eureka. But the main thrust of it — that downtown and northwestern Eureka form a natural community of interest with Arcata, or that Titlow Hill forms one with Shelter Cove and Cutten — should be recognized as patently absurd.

And while the populations of each of the districts is not as balanced as it could have been, it is not as bad as it might otherwise be. See the chart below. The total population of the county, according to the 2020 Census and after prisoners and other institutionalized people have been returned to their home counties, is 138,810. That means that ideally, each district has 27,362 residents.

Here’s how that shakes out in Draft Plan A2:

Pop.Diff.Diff. %
Largest deviation: 4.7%
Average deviation: 2.84%

The maximum deviation allowed by law is 10 percent. So this is within the range, but you can do both better and worse.

Draft Map C2

One way to do worse, in regards to population balance, can be seen in the latest iteration of the “C” series of maps — which are, themselves, meant to be based on the county’s current map, with not-so-radical adjustments.

C2, though, ends up being pretty radical as far as population balance goes. If the county adopts this map, the new Fifth District, centered around McKinleyville, will have almost 3,000 more people than the new First District:

Pop. Diff. Diff. %
1st 26,447 -915 -3.3%
2nd 27,421 +59 +0.2%
3rd 26,706 -656 -2.4%
4th 26,941 -421 -1.5%
5th 29,295 +1,933 +7.1%
Largest deviation: 7.1%
Average deviation: 2.9%

Why is this? Well, again, it all traces back to the “community of interest” testimony, which seems to be this Redistricting Advisory Committee’s sole concern. During the course of that testimony, the committee heard from a few people in the Blue Lake, Glendale and Fieldbrook areas who said that they wished for some combination of those three locales to be in the same district.

So at the last meeting the advisory committee asked the mapmakers to incorporate those people’s suggestions into the next version of Map C. The mapmakers — a Sacramento-based contracting firm called Redistricting Partners — followed those orders, and so incorporated all three of those areas into the Fifth District. (Blue Lake currently belongs to the Third.)

This ended up lopsiding the map powerfully, population-wise. Again: The whole point of redistricting after a census — or at least 90 percent of the point — is to bring population into balance. This map sacrifices that in order to please a handful of people in Blue Lake, Fieldbrook and Glendale, who may or may not continue to be pleased after finding out that their political voice has been so diluted, that each resident of Ferndale gets 10 percent more of a supervisor than they do.

Status Quo

So look at the population charts above — the ones corresponding to each of the draft maps — and then look at the chart below, which corresponds to the current supervisorial districts, as represented in the preceding map. That map is the status quo.

Pop. Diff. Diff. %
1st 27,116 -246 -0.9%
2nd 26,817 -545 -2.0%
3rd 28,296 +934 +3.4%
4th 26,407 -955 -3.5%
5th 28,174 +812 +3.0%
Largest deviation: 3.5%
Average deviation: 2.56%

The current supervisorial map is more balanced than either Draft Map A2 or Draft Map C2. Its largest deviation from the ideal is quite a bit smaller than either of the draft maps. Its average deviation from the ideal is also smaller than either of the draft maps.

So why draw new maps at all?

Well, you can do it to balance the population further — not the case and seemingly not the goal at the Humboldt County level. If the advisory committee, or eventually the Board of Supervisors, wished to rebalance the population, it could go a long ways to that simply by giving 3.5 percent of the population of Arcata’s Third District to Eureka’s Fourth District — the Manila end of the peninsula, say, or maybe even more sensibly, the southeast corner of Eureka near Redwood Acres, Mitchell Heights and the like.

You can also do it to redress ganky lines drawn in the past. The current map, despite being more balanced than either of the draft maps, has its own oddities, though none so odd as those offered to us in Draft Map A2. As mentioned before, Rio Dell and Scotia are currently in different districts, and to compensate for that fact the First has grown a weird protuberance down the Highway 101 corridor. That could be cleaned up, and the draft maps do clean it up — though either at the cost very dramatic population imbalance, or else through the introduction of much more dramatic crimes against communities of interest.

It didn’t have to be this way, and maybe it doesn’t. As I said before, no one is looking at the current map. I had to reconstruct the 2010 map myself. It’s not the exact same map, given that the Census Bureau alters the building blocks every 10 years — imagine being handed a very slightly different mix of 7,138 Legos to recreate the same Death Star you made in 2010 — but it’s mathematically as close as you can get. If anyone wants to check my work, here’s a 27-megabyte block-level GeoJSON file that you can plug into your favorite GIS.