Propositions maps and PVI correlations from the November 2014 election
David C. Latterman
Fall Line Analytics
Below are maps for November 2014 turnout, results of all the propositions, and charts of the propositions respective PVI correlations. A little commentary is provided, but most of the results are pretty straightforward. All of the proposition maps and PVI color patterns are set to the same scale.
As usual, turnout (around 53% overall) was dominated by Districts 7 and 8. There was some increased turnout in some of the eastern neighborhoods like Bernal and Potrero, but it is interesting to note that with most of the electoral activity on the eastern side of San Francisco this cycle (AD17, Supervisorial races), there doesn’t seem to be a big turnout advantage in the east side. This indicates that turnout patterns in San Francisco are extremely set year after year, and it takes extraordinary organizing efforts to elevate turnout in other parts of San Francisco.
A note about PVI correlations: these correlations show whether or not a proposition has an ideological overprint in its results. It doesn’t necessarily speak to how strongly a voting group feels about a measure – just that the results are controlled to some extent by ideology and its associated political geography. Many of the measures from this cycle do exhibit strong correlations with PVI.
Figure 1: November 2014 turnout
Figure 2: Transit bond (yes 72%)
Prop A performed similarly to other recent bonds, like the parks bond and streets bond. The PVI correlation was strongly positive, as is quite typical of current bond measures for street and other infrastructure improvements.
Figure 3: Transit set-aside (yes 61%)
This was another transit measure that performed well. Prop B also had a strong positive PVI correlation, though overall it did about 10 points lower than Prop A. The West Portal area notably downvoted this even more than other conservative neighborhoods, which I find a bit odd given their importance as a transit hub.
Figure 4: Children's Fund (yes 74%)
No children’s measure has lost in San Francisco in forty years, and it wasn’t going to happen here. The degree to which it passed still had a typical geographic/ideological overprint. But seriously Stonestown, what the heck?
Figure 5: Redevelopment benefits (yes 56%)
Benefit measures are often confusing to voters, so they typically vote along “do we or do we not like city employees” lines. This in turn is reflected in a very strong PVI correlation, which pertains to somewhat to how people feel about public sector unions. In fact, labor measures often go into the development of the PVI because it is so ideologically controlled. Prop D was no exception.
Figure 6: Sugar tax (yes 56%)
Prop E is the first measure we see that doesn’t correlate was strongly with PVI, indicating that support for this didn’t fall along typical political lines. Prop E did well in the central part of the city, and more poorly in D1, 4, 10, and 11. There is clearly a racial component to the results as well, since the outer neighborhoods where E did more poorly are more heavily minority. The split down D7 is particularly enlightening, where E did better in the whiter West of Twin Peaks area, but less well in more heavily Asian Balboa/Ingleside.
Figure 7: Pier 70 (yes 73%)
Prop F had strong support across the city, though it did a little less well overall in the outer neighborhoods, in much the same pattern as Prop E. Prop F supporters put a lot of money and effort into the measure, having to overcome recent negative sentiment about waterfront development. Nonetheless, Prop F fared very well, especially in the immediate area to the project in the Central Waterfront.
Figure 8: Housing tax (yes 46%)
Prop G was another more traditionally ideological measures with a very strong PVI correlation. It lost in most parts of the city, except in the most progressive of the city in D5 and D9 (and bits of D6, 8, 10).
Figure 9: City Fields (yes 45%)
Prop H was an attempt to overturn the approval of the Beach Chalet soccer fields. This was defeated by most demographics in the city, except in the areas adjacent to western Golden Gate Park, and bits of D10 and D11. Even in the more progressive parts of San Francisco, this didn’t do that well, causing a scattershot correlation with PVI.
Figure 10: Recreational fields (yes 55%)
Prop I was the inverse to Prop H, representing more generic “support” for building recreation facilities. A minor difference with Prop H is that Prop I did slightly worse in progressive areas, perhaps indicating resistance to something the Mayor placed on the ballot.
Figure 11: Minimum wage (yes 77%)
While Prop J unsurprisingly correlation very well with PVI, it still performed very well throughout the city. Even the neighborhoods that didn’t feel as strongly about Prop L voted for it at over a 50% clip.
Figure 12: Affordable housing (yes 66%)
Although Prop K correlated well with PVI, there are a couple interesting anomalies. D10 supported this very strongly, as did D6 (the electoral home of the sponsor). Parts of D7 voted strongly against this, even more strongly than their conservatism would indicate. Maybe it’s a NIMBY thing.
Figure 13: Transit balance (yes 37%)
Finally, Prop L was soundly defeated by much the same margin by which the transit bond passed. Interestingly however, the PVI correlation wasn’t nearly as strong. This seems due to the fact that some of the southern neighborhoods, irrespective of politics, supported this a little more strongly.