Districts to Grow Into

The politics of housing and growth in San Francisco is something I bloviate about with some frequency. Usually, I try to respond to the events of the day in a timely fashion.

Today, however, I’m going to break new ground and write about a topic that is so tangential that you, dear reader, are going to correctly assume that I need to get a life.

Today’s topic is: Population Growth, the Census, and How the Board of Supervisors Districts are Drawn.

I’m going to argue that by ignoring growth patterns, the Redistricting Task Force created districts in a way that would expedite their obsolesce; a phenomenon which we can see now, just four years after the new map was finalized.

table 1

Figure 1: 2010 population by District before Redistricting

The History

Population changes identified during the 2010 census triggered an adjustment of the Board of Supervisors’ district boundaries. Each of the 11 districts is supposed have an equal population, which would have been 73,203 people each at the time of the last redistricting.

The census revealed that our districts were pretty far out of whack. District 6, most pressingly, had 20,000 residents to shed. In response, the city formed the Redistricting Task Force to adjust the district boundaries to bring them closer to equality.

As a result, several neighborhoods were moved from District 6 into Districts 9 (the North Mission), District 3 (several blocks north of Market St), and District 5 (areas west of Van Ness St).

The Rules

 Despite the fact that inequality is what prompted redistricting, the Task Force cannot simply just draw the most equal districts possible. In fact, equality in population is only one piece of criteria that governed the Task Force’s work. The city Charter considers districts with variations of +/- 1% from the statistical mean (732 people in 2010) to be equal enough.

Districts are allowed to exceed the 1% variation (up to +/- 5%) “if necessary to prevent dividing or diluting the voting power of minorities and/or to keep recognized neighborhoods intact.” There are federal elections laws and Charter requirements that require the the Task Force to use this higher threshold if necessary to preserve “Communities of Interest.”

Generally the pillars of redistricting (pg 3) are: Equality in population, Contiguity, Compactness, Preservation of Neighborhoods and Communities of Interest.

The Results

The Redistricting Task Force took full advantage of the 5% threshold in their final results. Seven out of 11 Districts had variations greater than 1%.  Here is the final map:


Figure 2: The Final Board of Supervisors Map after the 2012 Redistricting


Figure 3: The population by district before and after Redistricting

And a table of the changes:

The Redistricting Task Force, not wanting to take a sledgehammer to the whole map, made only the bare minimum changes so that as many communities and neighborhoods could stay intact and in their historic districts as possible (often at the behest of the public). The downside is that many of the boundaries push pretty close to that 5% threshold.

Plus-or-minus 5 percent is kind of a lot. Especially if one district is plus (like District 11)  and another is minus (like District 1). The result of the Redistricting Task Force  The residents of districts 11 , 9, and 8 are under represented compared to their friends in Districts 1, 2, and 3, but the deviation is within the tolerance set by Redistricting Task Force criteria.

These Districts would be fine if the city’s population grew evenly, but we know that hasn’t happened. Growth has largely been relegated to the Eastern neighborhoods. We know this is the case because the lop-sided growth in District 6 is the reason we had to redistrict in the first place. As we’ll see, those growth patterns are more pronounced now than they were in 2010.

Durable Districts

The issue is that these district boundaries were not made to last. They are not durable. By “durable,” I refer to whether or not these districts will ensure equal (ish) representation for the entire 10 years that they’ll be in place. If a district boundary, after 10 years of existence, still meets the original criteria then it is durable.

It’s entirely possible that you cannot reasonably create totally durable districts. It may be that growth patterns are such that there’s no way to draw districts today that will be valid 10 years, but that doesn’t mean you can’t design districts with the idea that they should be as durable as possible. At some point between 2002 (when the previous map was created) and the 2010 census (when the current map was created), the district boundaries stopped meeting the criteria set by the city’s charter and state and federal law. Maybe they fell out of tolerance after eight years. Maybe it was three.

The issue of how long these districts remain valid is important. These districts exists to ensure that voters are represented equally. We’ve set parameters around what counts as “equal” (contiguous, compact, about the same population, neighborhoods and communities are not split), but what good are these parameters if our districts only meet them for a year or two?

Our Districts Today

The Redistricting Task Force in 2012 was concerned with creating a map that met the criteria at the time of its creation, but it’s obvious that they didn’t really (or weren’t able to) care about how long those district boundaries remained in tolerance.

The city continues to grow unevenly, but thanks to the housing pipeline report have some idea how the city will grow and could use that information to inform how our district boundaries are drawn.


Figure 4: Current net units by district in the Housing Pipeline Report

Note: just because it’s in the development pipeline doesn’t mean its anywhere near being built (or will ever be built. 8 Washington was in my pipeline report as one of the largest projects in D3, but has since been yanked). The point of looking at the pipeline is not to look precisely at each project individually, but to understand overall what parts of the city are growing in proportion to others.

Here’s what housing growth looked like in San Francisco the last time I checked the housing pipeline report ( in Fall 2015). There were almost 55,000 net new units in the pipeline at that time. That number is now over 62,000. As you can see, the eastern neighborhoods continue to do the lion’s share of the growing. Development in Districts 6 and 10 account for over 75% of the total. District 7 comes in a distant third at 11% (largely from one project, Parkmerced, which accounts for over 5,000 of D7’s new units). Parkmerced, it should be noted, hasn’t yet opened (this is one of the limits of using the pipeline report, more on that later).

The San Francisco’s estimated 2014 population is 852,469; an increase of 47, 234 from the 2010 census.  If we use the growth patterns indicated by the pipeline report to allocate these 47K new San Franciscans across the 11 different Supervisorial Districts, we see that the districts created only four years ago are not be valid today. Indeed, they are already deeply unequal.


Figure 5: 2014 Population by district using the Housing Pipeline Report as a proxy for growth

If San Francisco’s population growth followed it’s housing development then six of the eleven districts are outside the +/- 5% threshold, and only three out of the eleven are within the +/-1% threshold.

The reason these districts became so lop-sided so quickly is because the Redistricting Task Force gave the benefit of being 5% below the target population to the slow growth districts. While they didn’t make the highest growth districts (Districts 6 and 10) 5% above the target population, that didn’t stop the populations of those districts from blowing right through the 5% thresholds.

In order to get District 6’s population within tolerance, the Task Force loaded up Districts 11, 8, 9 and (to a lesser extent) 5 with extra people. Those are almost all the districts that touch Districts 10 and 6. That means that we cannot just cleave off parts of Districts 6 and 10 and shunt them next door. Next door is already pretty full. That would really just be kicking the can down the road again.

By starting the decade with over representation in some of our slower growth districts (Ds 1, 2, and 3), we ensured that our faster growing districts would only be more under represented as the decade wore on.

An Alternate Universe 

Had the Task Force considered growth patterns when it was drawing its district boundaries and tried to make their districts as durable as possible, they may have created district populations that looked something like the tables below. On the left, we have the hypothetical populations of each district if I had been on the Task Force. On the right, we have those same districts in 2014, using the pipeline report to approximate population growth.


Accounting for projected growth requires that our growth centers, Districts 6 and 10, have the smallest populations possible (-5% from the statistical mean). In this demonstration I also gave District 7 a lower population because its also projected to grow more than some of the other districts. Districts 1 and 4 (the furthest West) need the larger populations because they experience the least growth.

The rest of the districts need slightly larger populations, basically just to account for how small Districts 10 and 6 need to be. Despite my best efforts, I wasn’t able to make still wasn’t able to make Districts 6 and 10 durable. This is because the current geography of the districts concentrates growth. The only way to make truly durable districts would have been to carve out some of the growth and put them in other districts.

Caveats, Provisos, Stipulations, Limitations

Changing the map changes growth rates: I played with numbers but not geographies. Increasing the population of a district may also impact it’s growth rate. For example, by expanding the boundaries of District 1 to include new populations, I might also have to include some new developments as well, which means the new District 1 would grow more quickly than projected. In a way this is ideal. By better distributing growth into several districts, you help mitigate the diluting effect it has on the voters when it all happens in one district. However, because growth is so incredibly concentrated along the East side of the city, we would need radically different district boundaries for this phenomenon to have an impact.

Communities of Interest: I didn’t yet consider the implication these hypothetical population distributions have on neighborhoods or communities of interest. I’ve only just started playing with ReDrawSF.

The Pipeline Report isn’t Perfect: I’ve already mentioned, that many projects in the pipeline report are a long way from being finished. Many of them may not be built. New projects are added all the time. District 7’s growth is contingent on one really large development. Districts 6 and 10 have a few very large projects as well. If something happens to derail those large projects then our projections will be off. The Task Force wouldn’t necessarily want to look at the whole pipeline, but they would want to see estimates of how many units would be completed each year. However, individual projects normally aren’t that important (except Parkmerced) because we’re really only trying to get a directional indicator of where growth is and isn’t occuring. For that reason, the pipeline report isn’t a terrible indicator of growth. It certainly shows us where growth isn’t happening.

Growth that is Significant and Concentrated is Impossible to Accommodate: As we’ve seen even, growth in San Francisco is so concentrated that it’s impossible to follow all the rules and create districts that will last for 10 years without really changing the current map. However, just because you can’t mitigate it entirely, doesn’t mean you should mitigate it as much as possible. And it certainly doesn’t give you license to exacerbate the impact, as the Task Force did in 2012.

Why Is This Important?

San Francisco, by design, has concentrated growth in it’s Eastern Neighborhoods. One of the unfortunate consequences of concentrated growth is that, as the decade wears on, the influence of those voters is gradually diminished relative to the voters in the Western neighborhoods. Because each Supervisor gets one vote, the voters in slow-growth districts influence increases as the city’s total population grows, but not in their districts.

Imagine the grid on the left is San Francisco in 2010 after redistricting. Each square represents a district with roughly the same number of people and each district gets one voting representative at the Board of Supervisors. With our concentrated growth, two districts grow much more rapidly than the others and, by 2014, those districts still only have one voting representative, but a much larger number of people. That’s the grid on the right.


It’s pretty obvious why district populations are supposed to be as close to equal as possible. The city Charter mandates near equality at the time the map is adopted, but it also goes to great pains to reinforce the idea that voters’ influence should not be diluted, especially that of minority voters.

Faced with the dramatic concentration of growth in San Francisco, Redistricting Task Force would never have been able to create districts that were completely durable without completely dismantling the existing map. However, by allowing the slow growth districts to begin the decade with over representation (though within the legal limits), they ensured that the voters in the faster growing neighborhoods would see their influence increasingly diluted as decade wore on. This is especially troubling, given that District 10 contains San Francisco’s largest, but shrinking, African American population.

Some will argue that accommodating growth runs, like an oncoming train, into the mandate to preserve communities of interest. I would argue the opposite. I think failing to accommodate growth puts certain communities above others. Districts 6 and 10 have diverse neighborhoods, whose communities should not have the quality of their representation systematically undermined.

Consider that if the next Task Force draws a map similar to what the previous one drew, then the residents of District 6 will have been under represented for about 30 years compared to their Western neighbors.

I am not arguing that growth should trump all other considerations. However, it shouldn’t be ignored, as it was in 2012. The next Task Force needs to strike a balance and take equal representation seriously. I believe the only way to do that is to consider future growth. Despite the risks and uncertainties that come with trying to anticipate growth,  to ignore it undermines the principle of “One Person, One Vote.”





Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s