San Francisco Demographics and Voting Trends

1
San Francisco Demographics and Voting Trends

• David Latterman
• October, 2004

2
Agenda

• Description and methodology of demographic
  dataset
• Correlations to Rich DeLeons PVI
• Correlations to recent elections
• Turnout
• Identity Politics
• A look to the future

3
New SF demographic dataset

• Compiled from 2000 census
• Use blocks, block groups, and tracts to convert
  data to SF voting precincts
• Blocks are standard unit
• Much useful data come from block groups, too,
  though these are harder to convert
• Larger tract data is best for district analysis

4
Conversion techniques to SF precincts
5
Data from various census units

• Blocks
• Race
• Age
• Gender
• Housing tenure (home ownership vs. renting)

• Block Groups
• Family data
• Education
• Employment
• Income
• Immigration status (especially useful in SF)

• Tracts
• LGBT
• (Same Sex Householders)

6
Example of block conversion - Asian/Pacific
Islander -
7
Example of block group conversion - Median HH
Income -
8
Using precinct data for analysis

• Different than poll data reflects how people
  actually voted
• Issues with data
• n 570 citywide
• n much less for districts (45-60)
• Ecological Fallacy
• Difficult to establish which characteristics are
  important in heterogeneous parts of SF

9
Correlations to PVI

• PVI high score means a more liberal voting
  population low score means a more conservative
  voting population
• Some stronger correlations to certain
  characteristics
• Varies within districts (this is important!)

• More liberal
• Hispanic
• Younger voters

• More conservative
• Asian
• Older voters
• Higher income
• Higher homeowner percentage

10
San Francisco PVI Map
11
Good correlations with district variation
(inverse slope)
voters over 50 vs. PVI
owned housing units vs. PVI
12
Good correlations with district variation
(positive slope)
voters 25-29 vs. PVI
Correlations are consistently stronger with
demographics that vote more conservatively!
13
Not a good correlation, but with some district
variation
Born in US vs. PVI
14
Not a good correlation with some district and
neighborhood variation (1)
Hispanic voters vs. PVI
The Mission
15
Not a good correlation with some district and
neighborhood variation (2)
female voters vs. PVI

• The Castro
• The Tenderloin
• SOMA

Female Becomes a proxy for LGBT voters
Thanks Rich!
16
Correlations to recent elections

• Turnout
• Identity Politics
• Potential future trends

17
Turnout absentee and voting booth- is
definitely affected by politics
Dec 03 Absentee Turnout
18
Dec 03 Election Day Turnout
19
Similarly, TO is affected by demographic
characteristics
Using average turnout of last three elections
20
Identity politics is alive and well in San
Francisco

• Look at African-Americans, Asians, Hispanic, LGBT
• Variation between and within districts

21
African-American Voting Trends
Precinct gt15 census pop
22
Asian-American Voting Trends
Precinct gt30 census pop
23
Hispanic Voting Trends
Precinct gt25 census pop
24
Racial politics in 2004 DCCC and beyond

• Look at Asian, Af-Am, LGBT, Latino correlations
• Eye toward the future

25
DCCC 04 Asian (1)
Hsieh and Jung were the top two vote-getters in
AD12
Hsieh
Jung
26
DCCC 04 Asian (2)
Ow (AD13) and Yee (AD12) didnt win
Ow
Yee
27
DCCC 04 African-American
28
DCCC 04 - Hispanics who did not win
Garcia AD13
Trevino AD13
Ramos AD12
29
LGBT winners in AD13
Haaland
Wiener
30
Looking ahead

• SF will see increased proportions of Hispanics
  and Asians, from many countries.
• They will have a huge impact on the electorate,
  especially as immigrants become naturalized and
  are able to vote
• If, in general, Asians are more moderate than
  Latinos

31
SF Demographics Trends
32
Ethnicities vs. naturalized immigrants
33
Naturalization vs. PVI
34
For the future / other uses

• Subdivide racial groups to nation of origin
• Use data in policy debates (look for
  commonalities between Progs/Mods for compromises)
• City agencies like HHS, MUNI, PUC

35
(No Transcript)