Elections Commission
2009 2008 2007 2006 2005 2004 2003 2002 Minutes of
SPECIAL MEETING
San Francisco Elections Commission
and Supervisor Jake McGoldrick
Regarding Proposal to
Change San Francisco Election Cycles
Held at City Hall, Room 082
Monday, June 26, 2006
(These minutes were approved at the regular July 19, 2006 Commission Meeting)
CALL TO ORDER. President Matthews called the meeting to order at 10:06 am
ROLL CALL. PRESENT: Commissioners Gerard Gleason, Richard P. Matthews, Arnold Townsend, Jennifer Meek, Supervisor Jake McGoldrick, Betty Chan, aide to the Supervisor, Shawna Johnson, analyst with the Office of the Legislative Analyst, and Deputy City Attorney Miguel Márquez.
Presentation by Supervisor Jake McGoldrick’s Office regarding the proposed Charter amendment related to moving elections that occur in odd-numbered years: Supervisor McGoldrick said that “we have a lot of elections and we have a lot of burnout instead of high turnout”. He said, there have been complaints for decades, that there have been 24 to 25 elections in San Francisco in ten years. His plan is to consolidate some of these elections, especially the Citywide elections.
He stated that his goal is to get more people to vote for the offices that “are incredibly important in San Francisco. (Those) are: Mayor, District Attorney, City Attorney, Sheriff, and City Treasurer“. The Supervisor referred to the handout his office prepared which is attached to these minutes. The table, which lists the dates of elections beginning November 7, 1995, and ending November 8, 2005, and the voter turnout numbers for these elections was discussed. The numbers of voters for the even-year elections were higher than those for the odd-year elections.
Commissioner Gleason asked the Supervisor about the election reported on one of the charts in his proposal. The 11/4/2003 Municipal Election had FOURTEEN Ballot Measures. Under the new proposal, those measures would have had to wait until the 2004 election. The Commissioner asked that if that would be the case today, would the Board wait until an even year to propose Ballot Measures. Betty Chan responded that the Board of Supervisors could, under the
proposed Charter amendment, initiate a Special Election, as can members of the public by petition.
Commissioner Gleason then added, that in all likelihood, the Department of Elections would be operating an election that odd-numbered year. He reminded
Special Meeting – June 26, 2006
the Supervisor that even if there is only ONE ballot measure, the Department has to open precincts.
Supervisor McGoldrick replied that he would think that the Board of Supervisors would practice restraint regarding ballot arguments and the need for Special Elections to put those arguments to public vote.
Commissioner Townsend said that the problem is whether the City will see a benefit in “streamlining” the process of eliminating odd-year elections, or will there be just more Special Elections called, not reducing expenses. The secret to increasing turnout, the Commissioner said, was the hard work of campaigning and addressing the needs of people who don’t vote. He said that this is something, in his opinion, the City hasn’t done well over the last twenty years.
Commissioner Gleason asked for clarification from Supervisor McGoldrick. He asked if what is being proposed is that there not be Municipal Elections.
Supervisor McGoldrick replied that this statement is correct. He stated that during the time between elections, there would be more time for voter outreach, education, registering voters, and cleaning up the voter rolls.
Betty Chan referred to the chart on page 6 which shows the proposed savings if elections where held only on even-numbered years. She added that the savings could then be used to pay for staff for voter outreach instead of gearing up for an election.
Commissioner Meek reminded the Commissioners and Supervisor that this would only apply if there were no Special Elections held in that year. The Commissioner said that the likelihood of having Special Elections in San Francisco is very high.
Commissioner Matthews said that Supervisors will probably exercise restraint regarding putting forth Ballot Measures; however, the electorate may not.
Commissioner Townsend said that he understands the move to try to save money; however, when it comes to cutting jobs, saving money does not always happen. It looks good on the reports, but one must ask was there a strain put on the budget elsewhere, like in human services costs because this employee has lost her/his job. Saving money by cutting employment, for him, is never the best way to save money.
Supervisor McGoldrick replied that he feels that San Francisco is an "incredibly compassionate City”, and that the City always tries to move people around, transfer them, rather than lay them off.
Commissioner Matthews said that a huge part of the Elections Department having been able to have successful elections for the past four years has been the same people coming back to work the elections year after year, being laid off
Special Meeting – June 26, 2006
for a few months between elections, but coming back and bringing their expertise
with them. He asked how can the City guarantee that the quality of elections will continue if the Department is dormant for 12 to 18 months at a time. He speculated that the Department may not be able to get those people to come back after such a long layover. He said that, although there has not been a study of this situation, he doesn’t feel that staff will be coming back.
Commissioner Gleason agreed, and said that the Department would lose institutional memory. He said that there is a dedicated, regular group of people who come back to work at every election.
Supervisor McGoldrick said that the City has five to six more years to put this proposal, should it pass, into action. This change would “kick in in the year 2012.”
Shawna Johnson explained page 7, which compares ballot length to voter turnout. Her office’s analysis showed that voters did not show signs of “voter fatigue” in elections where the number of ballot measures was high.
Commissioner Matthews asked Ms. Johnson if her office had done any studies of Special Elections and her reply was that they have not.
Betty Chan stated that it might not be attractive for people to request a Special Election if there were no candidate(s). Perhaps proponents time their initiatives for periods when they can get more voters.
Commissioners Townsend and Matthews both responded that sometimes those who push for the Special Election do so because they don’t expect a high voter turnout.
Betty Chan said that the Ethics Commission, under this proposal would not save money but would save time and be able to catch up on other work, like conducting their audits (currently they are doing their 2004 election audit).
Commissioner Matthews said that the Elections Commission would probably not weigh in on the policy piece of the Supervisor’s proposal, and that the Elections Commission’s concern is, because it’s Chartered to do so, to look after the Department and make sure that the quality of work can remain what we have come to expect it to be over these last four years. Other Commissioners might differ on this, however.
Commissioner Gleason brought to Supervisor McGoldrick’s attention that the Elections Commission was responsible for bringing forth the proposal to amend the Elections Code to move the Public Defender and Assessor/Recorder races from the Primary to the General Election. The Director presented it to the Rules Committee of the Board. The Commission took a back seat, because our involvement is only overseeing elections. When someone wants to put something on the ballot, the Commission has no involvement.
Special Meeting – June 26, 2006
Commissioner Meek said that her concern was keeping the Department going, keeping the temporary positions and somehow making them permanent. We need to keep that “flow” going. If people are dropped, she said, then others
have to be re-trained at an expense greater than keeping the original staff. How does it effect the Department when it must run a Special Election? What are the effects going to be on the Department. What Supervisor McGoldrick proposes may be great for the voters, but what happens to the Department?
Betty Chan said that she had not discussed this aspect with the Director of Elections, but that she would and will get back to the Commission with her findings. Ms. Chan also stated that the Controller’s numbers show that a portion of the money that the City could save would be put towards the larger General Election.
Supervisor McGoldrick said that his office is looking into doing a socio-economic report of the effect of this proposal.
ADJOURNED at 11:11 am.
ATTACHMENTS:
LEGISLATIVE ANALYST MEMORANDUM
To: Supervisor McGoldrick, Chair, City Operations and Neighborhood Services
From: Shauna Johnson and Keith V. Lucas, Office of the Legislative Analyst
Date: June 22, 2006
Re: Elections and Voter Turnout (OLA No. 045-06)
Summary of Requested Action
The sponsor requested research on elections and voter turnout in the City and County of San Francisco from 1995 to 2005. Specifically, we were asked to provide the following information for the requested election cycles: election costs, voter turnout, number of ballot measures (propositions and referendums), and offices that are up for election in even and odd years. Additional information was requested regarding special elections, voter turnout trends, and the impact of long ballots on voter response and turnout.
Findings
Types of Elections
Since 1995, there have been 24 election cycles in the City and County of San Francisco. The City holds five types of elections - general, primary, municipal, run-off, and special elections. General elections occur every two years (during even numbered years) and can include federal, state, and local offices. A primary election is a nominating election in which political parties choose a candidate. Primary elections generally occur during the same year as general elections and the winning candidates from each party will go up against each other in the general election. Municipal elections also occur every two years (during odd numbered years), but only include local offices. Special elections are those that are not regularly scheduled, and are typically held between general elections to fill a vacancy. Run-off elections occur during municipal election cycles when no candidate for an office obtains the majority of the votes. This results in a run-off election between the two highest scoring contenders.
In March 2002, San Francisco voters passed Ranked-Choice Voting (Proposition A). Ranked-choice voting, also known as “instant run-off voting,” eliminates the need for separate run-off elections by allowing voters to rank up to three candidates for each office. With ranked-choice voting, if a candidate receives a majority of the first-choice votes cast for that office, that candidate will be elected. However, if no candidate receives a majority of the first-choice votes cast, an elimination process will begin. The candidate who received the fewest first-choice votes will be eliminated and each vote cast for that candidate will be transferred to the voter’s next-ranked choice among the remaining candidates. This elimination process continues until one candidate receives a majority and is declared the winner. Ranked-choice voting was implemented for the first time in the November 2004 general election. Attachment 1 outlines each election and includes information on the type of election, type of office, and the election outcome.
Elections occur in both even and odd-numbered years, and San Francisco has had at least one election cycle every year since 1995. In general, local municipal elections occur in odd-numbered years and state and federal elections occur in even-numbered years. Local elections held in odd years consist of the offices of Mayor, City Attorney, City Treasurer, District Attorney, and Sheriff. The offices of Mayor, District Attorney, and Sheriff are held together, and the offices of City Attorney and City Treasurer are held together in alternating, odd years. The Mayor is elected every four years, in odd-numbered years that precede the U.S. presidential election (see Attachment 2 for detailed information on the past Mayors of San Francisco). Local ections held in alternating even years include the offices of Board of Supervisors, Assessor-Recorder, Public Defender, Board of Education, and Community College Board. Table 1 shows the offices that are up for election during even- and odd-numbered years.
Table 1: Offices up For Election in Even- and Odd-Numbered Years
Even Years
|
Odd Years
|
President
|
Mayor
|
U.S. Senator
|
City Attorney
|
U.S. Representative
|
City Treasurer
|
Governor
|
District Attorney
|
Lieutenant Governor
|
Sheriff
|
Secretary of State
|
|
State Controller
|
|
State Board of Equalization
|
|
Attorney General
|
|
Insurance Commissioner
|
|
State Superintendent of Public Instruction
|
|
State Senator
|
|
State Assembly
|
|
State Judicial Offices
|
|
Board of Supervisors
|
|
Assessor-Recorder
|
|
Public Defender
|
|
Community College Board
|
|
Board of Education
|
|
Voter Turnout
Among other things, the type of race, the position, and the number of candidates are important factors that affect voter turnout. While voter turnout varies based on the types of elections, the number of registered voters has remained consistent over the past decade. Table 2 is a summary of the voter turnout from 1995 to 2005.
Table 2: Summary of Voter Turnout from 1995-2005
Date
|
Registration
|
Turnout
|
% Turnout
|
Precinct
|
Absentee
|
Election Type
|
11/8/2005
|
428,481
|
229,714
|
53.61%
|
136,897
|
92,817
|
Special Statewide
|
11/2/2004
|
486,937
|
361,822
|
74.31%
|
226,354
|
135,468
|
General
|
3/2/2004
|
448,948
|
190,828
|
42.51%
|
120,104
|
70,724
|
Primary
|
12/9/2003
|
466,127
|
253,872
|
54.46%
|
162,673
|
91,199
|
Municipal Run-off
|
11/4/2003
|
459,213
|
209,723
|
45.67%
|
140,010
|
69,713
|
Municipal
|
10/7/2003
|
454,929
|
269,143
|
59.16%
|
187,647
|
81,496
|
Special Statewide
|
12/10/2002*
|
93,537
|
35,897
|
38.38%
|
19,079
|
16,818
|
Municipal Run-off
|
11/5/2002
|
449,508
|
225,102
|
50.08%
|
159,645
|
65,457
|
General
|
3/5/2002
|
440,016
|
150,249
|
34.15%
|
98,669
|
51,580
|
Primary
|
12/10/2001
|
453,961
|
75,267
|
16.58%
|
44,294
|
30,973
|
Municipal Run-off
|
11/6/2001
|
452,543
|
134,024
|
29.62%
|
96,876
|
37,148
|
Municipal
|
12/12/2000
|
398,234
|
129,629
|
32.55%
|
81,670
|
47,959
|
Municipal Run-off
|
11/7/2000
|
486,636
|
324,031
|
66.59%
|
241,234
|
82,797
|
General
|
3/7/2000
|
468,198
|
210,229
|
44.90%
|
155,955
|
54,274
|
Primary
|
12/14/1999
|
467,295
|
228,247
|
48.84%
|
154,357
|
73,890
|
Municipal Run-off
|
11/2/1999
|
453,151
|
203,674
|
44.95%
|
151,447
|
52,227
|
Municipal
|
11/3/98
|
448,888
|
250,719
|
58.85%
|
183,839
|
66,880
|
General
|
6/2/1998
|
430,794
|
199,157
|
46.23%
|
137,812
|
61,345
|
Primary
|
11/4/1997
|
413,355
|
125,697
|
30.41%
|
87,352
|
38,345
|
Municipal
|
6/3/1997
|
411,230
|
178,948
|
43.52%
|
129,156
|
49,792
|
Special Municipal
|
11/5/1996
|
481,902
|
298,648
|
61.97%
|
223,339
|
75,309
|
General
|
3/26/1996
|
440,254
|
178,165
|
40.47%
|
126,643
|
51,522
|
Primary
|
12/12/1995
|
425,969
|
198,326
|
46.56%
|
130,573
|
67,753
|
Municipal Run-off
|
11/7/1995
|
417,883
|
216,735
|
51.86%
|
152,017
|
64,718
|
Municipal
|
*Note: This election was the result of a Supervisorial run-off (for Districts 4 and 8), with a 35,897 turnout reflecting the 93,537 voters registered in the two districts.
Voter turnout rates have fluctuated, ranging from as low as 16.58 percent in the December 2001 run-off election to as high as 74.31 percent in the November 2004 general election. Presidential elections typically had the highest turnout rates, and non-mayoral municipal elections the lowest. December runoff elections, which are no longer necessary since voters approved ranked choice voting in 2002, had even lower turnout levels than non-mayoral November elections.
The OLA compared voter turnout rates for November elections (both general and primary) held in odd and even years. Since 1995, five November elections were held in even years and six were held in odd years. On average, the voter turnout rate during November even-year elections (63.36%) was higher than the turnout rate in November odd-year elections (42.86%).
Special Elections
The San Francisco Elections Code governs special municipal elections. A special election may occur in one of several ways:
The Board of Supervisors also has the authority to consolidate a special election with a general municipal or statewide election. Since 1995 the City has had three special elections, two statewide (November 2005 and October 2003) and one municipal (June 1997).
Trends in Voter Turnout
- Larger cities tend to have lower turnout rates.
- Cities with residents of higher socioeconomic status tended to have significantly higher turnout.
- Race and ethnicity appear to be related to turnout. Past research suggests that African Americans participate at rates roughly equivalent to whites. However, the presence of large Asian American or Hispanic populations is associated with lower overall turnout.
Ballot Measures
From 1995-2005, 184 local ballot measures have come before the San Francisco voters. Table 3 is a summary of the ballot measure results from 1995 to 2005. Of the 24 elections held during this period, 17 elections contained local ballot measures. None of the six run-off elections (held in December of 1995, 1999, 2000, 2001, 2002, and 2003) contained ballot measures. Generally, November elections contained more measures than March or June elections. There was not a significant difference between the average number of ballot measures in even years (11) compared to odd years (10). For more detailed information regarding local ballot measures see Attachment 3.
Table 3: Local Ballot Measure Results from 1995-2005
Election Date
|
Election Type
|
# of Ballot Measures
|
% Pass
|
% Fail
|
11/8/2005
|
Special Statewide
|
9
|
78%
|
22%
|
11/2/2004
|
General
|
16
|
50%
|
44%
|
3/2/2004
|
Primary
|
11
|
91%
|
9%
|
11/4/2003
|
Municipal
|
14
|
93%
|
7%
|
11/5/2002
|
General
|
20
|
65%
|
35%
|
3/5/2002
|
Primary
|
7
|
86%
|
14%
|
11/6/2001
|
Municipal
|
9
|
78%
|
22%
|
11/7/2000
|
General
|
18
|
56%
|
44%
|
3/7/2000
|
Primary
|
6
|
67%
|
33%
|
11/2/1999
|
Municipal
|
11
|
91%
|
9%
|
11/3/1998
|
General
|
10
|
100%
|
0%
|
6/2/1998
|
Primary
|
12
|
64%
|
33%
|
11/4/1997
|
Municipal
|
8
|
75%
|
25%
|
6/3/1997
|
Special Municipal
|
6
|
67%
|
33%
|
11/5/1996
|
General
|
10
|
60%
|
40%
|
3/26/1996
|
Primary
|
2
|
100%
|
0%
|
11/7/1995
|
Municipal
|
15
|
53%
|
47%
|
Total
|
|
184
|
58%
|
42%
|
Election Costs
Over the past 10 years, an annual average of approximately $8.6 million has been appropriated to the Department of Elections. From 1995 to 1999 election appropriations were fairly stable. However, in 1999 election appropriations more than doubled from the previous year. Appropriations nearly doubled again in 2000 and have subsequently fluctuated from 2001 to 2005. During the 2003-04 fiscal year, appropriations were significantly higher compared to other years. According to the Department of Elections (DoE) appropriations were high during this time because four elections were held that fiscal year. For the 2004-05 fiscal year, $2,451,000 was appropriated specifically to implement instant run-off voting. Information related to the specific cost per election was not readily available from DoE.
Table 4 is a summary of the DoE budget appropriations for the past decade and includes the average appropriation per election. The appropriations reported in Table 4 represent balances recorded in the DoE FAMIS system and do not include allocations to individual elections.
Table 4: Department of Elections Appropriations from 1995-2005
Fiscal Year
|
Appropriation
|
Number of Elections
|
Average Appropriation Per Election
|
2004-2005
|
$10,253,326
|
2
|
$5,126,663
|
2003-2004
|
$19,098,436
|
4
|
$4,774,609
|
2002-2003
|
$11,277,032
|
2
|
$5,613,516
|
2001-2002
|
$14,335,884
|
3
|
$4,778,628
|
2000-2001
|
$11,633,045
|
2
|
$5,816,522
|
1999-2000
|
$6,195,688
|
3
|
$2,065,229
|
1998-1999
|
$2,708,189
|
1
|
$2,708,189
|
1997-1998
|
$3,722,756
|
2
|
$1,861,378
|
1996-1997
|
$3,519,737
|
2
|
$1,759,868
|
1995-1996
|
$3,871,143
|
3
|
$1,290,381
|
Elections Model
The Department of Elections estimated the City’s election costs over the next ten years under the current structure for odd and even year elections, including primary, general, and combined (general and municipal) elections. They also estimated the election costs under a hypothetical scenario in which odd year municipal elections were eliminated by consolidating them into even year, combined elections. The cost factors for this model include:
Variable Costs
|
Fixed Costs
|
Temporary salaries
|
Poll-workers
|
Temporary overtime
|
Polling place rentals
|
Temporary benefits
|
Sheriff and Department of Parking and Traffic staff
|
The cost of ballot cards
|
Materials and supplies
|
Voter information pamphlet (VIP) printing
|
|
VIP layout and translation costs
|
|
This model does not account for the change in ongoing costs such as labor and inflation rates. Based on the consolidated model, DoE estimated that the City could save at least $26.2 million over a period of 10 years.
Effect of Ballot Length on Ballot Completion
The OLA conducted statistical analysis to determine the effect of ballot length on voter response (see Figures 1-3). We looked at the three longest ballots (determined by the number of items) from 1995-2005. The figures show the percentage of ballot holders who voted for each ballot item in the San Francisco elections of November 5, 2002, November 7, 2000, and November 3, 1998. The number of items on the three ballots ranged from approximately 35 to 60, and items were generally placed in the following order: Federal positions, State positions, Local positions, and propositions.
Based on statistical analysis, a study of the figures, and an understanding of the ballots, we make the following observations:
- The proposition portion of the ballot, which appears at the end of the three ballots examined, does not show any statistically significant trends, negatively or positively. Additionally, the average proposition response percentage is on par with the beginning portion of the ballot. We thus conclude that voters did not show any signs of “voter fatigue” in these three elections (for the given metric of response percentage of ballot holders).
- Judicial portions of the ballots show significant drops in response percentage, from averages between 80 percent and 95 percent before these sections 40 percent and 50 percent. Voters returned to higher response percentages after these sections.
Figure 1: Comparison of Ballot Length to Vote Response, November 2002 Election
Figure 2: Comparison of Ballot Length to Vote Response, November 2000 Election
Figure 3: Comparison of Ballot Length to Vote Response, November 1998 Election
Effect of Ballot Length on Voter Turnout
The following figure shows the effect of ballot length on voter turnout (the percentage of registered voters who went to the polls) for San Francisco elections from November 1995 through November 2005.
Figure 4: Comparison of Ballot Length to Voter Turnout
Based on statistical analysis, a study of the figure, and an understanding of the ballots, we make the following observations:
- Run-off elections have the shortest ballots (1 or 2 items) and the widest variation in turnout, ranging from 17 percent to 54 percent. For both of these reasons, we do not believe that run-off elections should be included in an analysis to determine the effect of ballot length on voter turnout.
- Excluding run-off elections, ballot length from November 1995 through November 2005 does not show a statistically significant effect on voter turnout, negatively or positively.
- Election type may be a significant factor in determining voter turnout. The top four elections in terms of turnout from November 1995 through November 2005 were the 2003 recall and the three presidential elections.
Conclusion
Based on the study of San Francisco elections cycles from November 1995 to November 2005, the OLA found the following:
- Voter turnout in San Francisco varies greatly across different types of elections. Presidential elections typically had the highest turnout rates, and non-mayoral municipal elections the lowest. December runoff elections, which are no longer necessary since voters approved ranked choice voting in 2002, had even lower turnout levels than non-mayoral November elections.
- Election appropriation allocations have more than tripled on an annual basis from 1995 to 2005 and have averaged $8.6 million over that time period. The switch from punch cards to optical scan cards during the 2001-02 fiscal year largely explains this increase.
- Of the 184 ballot measures placed before the voters from 1995 to 2005, 58 percent passed and 42 percent failed. Generally, November elections contained more measures than March or June elections. There was not a significant difference between the average number of ballot measures in even years (11) compared to odd years (10).
- Based on a study of the three longest ballots from 1995 to 2005, we determined that there are no signs of drop off in voting activity towards the end of the ballot. The average response percentages for items listed toward the end of the ballot were similar with those of the beginning portion.
- Based on a study of ballot length and voter turnout, we determined that ballot length does not have a statistically significant effect on voter turnout.