New census numbers provide clarity on City Council redistricting

Back in March SCC Insight published a preview of what the City Council redistricting process will entail, including some predictions of where district boundaries might change based upon 2019 population estimates.  But earlier this month the official 2020 census numbers were published, bringing more certainty to where we can expect changes.

Here are the final 2020 census numbers, with comparisons to ten years ago and the 2019 population estimates (courtesy of the City of Seattle):

Council District 2010 Census 2019  Estimate 2020 Census
1 86,785 99,700 99,698
2 87,268 101,900 99,163
3 86,558 113,200 106,845
4 87,677 109,000 106,103
5 86,950 98,300 99,102
6 86,771 101,100 102,753
7 86,651 123,900 123,351
Total 608,660 747,100 737,015

The first take-away: assuming that the 2019 population estimates were reasonably accurate (they were done by the US Census Bureau but through a different process than the official decennial census), Seattle’s population shrank by 10,000 people last year. It’s hard for anyone to know exactly the full effects of both the pandemic and the contentious census process under the Trump administration, but the official count is definitely a bit less than what most people assumed it would be going into the 2020 census process. For context, other urban areas around the country also lost population. At the same time, King County as a whole gained about 17,000 people, increasing from an estimated 2,252,782 in 2019 to 2,269,675 in the official 2020 census.

Districts 2, 3 and 4 came in lower than expected — especially District 3, which lost 6,500 people. District 6 increased by about 1,600, and the other districts stayed about the same.

The rules for redistricting require that the district boundaries must be adjusted so that the difference in population between the largest and smallest districts is no larger than 1% of the entire population (using the official 2020 census numbers). According to the census, the total population of Seattle is 737,015. That means all seven districts will need to be within 7,370 people of each other. The average district size is 105,288; the absolute maximum size for a district, then, is 111,605 (if the other six are all just under the average), and the absolute smallest is 98,971, but realistically they will all likely need to be more tightly clustered — say within 0.5% of the average. That would put the acceptable range for a district’s population at 101,603 to 108,973.

There are two impactful changes from when we looked at this in March based on the 2019 numbers. Officially, District 6 no longer needs to grow: it is now just a bit over the minimum. And District 3 no longer needs to shrink; losing 6,500 people last year dropped it below the maximum.  But Districts 1, 2 and 5 do still need to grow, and likewise District 7 still needs to shrink — by a lot.

But because of the other rules guiding redistricting, namely that each district must be contiguous, neighborhoods should be kept intact, and existing boundaries must be maintained whenever possible, most of the analysis from March still holds true — at least for the south end.  District 1 must grow, but it only touches one other district: District 2. So some portion of District 2 will need to be carved out and moved over. The last time we looked at this, moving both Georgetown and SODO over together wouldn’t be enough to bring D1 into the acceptable range, so the only real choice was to move Pioneer Square or the CID over; but with the revised numbers, moving Georgetown and SODO (together about 3,700 people) would now work, and might be the simplest solution. But that still puts District 2 farther below the minimum.

Fixing District 2 is a bit trickier than last time, because District 3 is not overflowing with people anymore. Carving off the remainder of Mount Baker (which is currently split between D2 and D3) and Leschi would balance out District 2 by adding about 9,000 people, but that now leaves District 3 under-populated.

That, too, can be fixed, since it’s adjacent to District 7: the district most in need of adjustment, as it needs to transfer out between 14,300 and 21,700 people to neighboring districts. District 3 can’t take on nearly that many, nor can District 6; any solution to D7’s size problem will by necessity involve multiple districts. Moving Magnolia up into D6 (over 13,000 people), and then Greenwood from D6 to D5 (around 8,000), gets us most of the way there; then there are several possibilities for slicing off a piece of District 7’s easternmost edge to move about 6,000 more into District 3. That leaves all seven districts within the acceptable population range.

District 4 might be the only one that comes out of the redistricting process unchanged.

One should take this as an example of the kind of changes that will need to happen, not as the only solution or the definitive one. The shuffle between Districts 1 through 3 seems inevitable in the south end of the city, forced by the fact that District 1 only touches one other district. Likewise, the fact that District 7 needs to shrink so much and District 5 needs to grow, while Districts 4 and 6 are between them and already at acceptable sizes, means another complex shift is unavoidable. But there are more choices available now for exactly which pieces of the map are traded around than there were six months ago given the population data we had at the time. The redistricting commission can also move smaller groups of blocks along the borders between districts to fine-tune the numbers even further if they need to.

We will hear more about the redistricting process in the months to come as the Redistricting Commission gets to work.

I hope you found this article valuable. If you did, please take a moment to make a contribution to support my ongoing work. Thanks!


    1. And Sawant lives in Leschi — if it moves into D2, then she would no longer live in D3 and wouldn’t qualify to be its Councilmember.

  1. The excess for 7 can be handled by splitting it between 1 and 2. They each get about 8,000, while 7 loses 16,000. This can be done by first connecting 1 with 7 (which means moving at least part of SoDo from 2 to 1). Then give 1 and 2 panhandles made up of the southern part of 7 (e. g. everything below Denny). There are other ways to finesse this — like having one grab more of 2, and 2 grab more of 7 — but that’s the basic idea.

    5 would grow by extending its southeast border (to 80th or 75th).

    This would leave 3 and 6 alone. In the future, it allows either 1 or 2 to grow while 7 shrinks. I think this would minimize disruption, while making the districts more flexible in the future.

    1. The redistricting rules make it pretty clear that “panhandles” are strongly discouraged. I looked at the “have 1 grab more of 2, and 2 grab more of 7” but it ends up with a very distorted District 2 that almost certainly wouldn’t be allowed under the rules. The southern part of D7 isn’t densely populated; you would have to move a large area out in order to significantly reduce D7’s population back into the allowed range (further distorting the shapes of D1 and/or D2).

  2. Panhandles are discouraged? That’s nuts.

    This means that District 3 will change for no good reason. It happens to be very close to average, but will shrink only because of the arbitrary way they drew up these districts, along with silly rules to accompany them. It is as if they wanted as much disruption as possible.

    1. Well, regular shapes are preferred to discourage gerrymandering oddly-shaped districts. The rules also require minimizing boundary changes when redistributing. So the districting commission has some work to do in balancing those rules.

Comments are closed.