Applying PR: Case Studies
In this section, we’ll take a look at how PR might be implemented in real-world situations that are currently using some form of FPTP or plurality election.
Our first case is the County Board of Supervisors in San Mateo County, California. A second case is a pair of proposals for mixed-member list PR in California.
We’ll be adding more cases as time goes on.
San Mateo County is located on the San Francisco Peninsula, just south of San Francisco, at the north end of Silicon Valley. It has a population of about 700,000, and about 340,000 registered voters, roughly half Democrats and a quarter each Republicans and independents, with a sprinkling of smaller parties.
Typical of California’s 58 counties, San Mateo’s Board of Supervisors has five members, elected from five geographic districts. Less typically, the supervisors are elected at large to single-district seats—that is, all the voters in the county vote on each seat individually. Two or three supervisors are elected every two years to four-year terms. The election is held in conjunction with even-year June primaries, with a two-candidate runoff in November if no candidate receives a majority in June. In practice, runoffs, or indeed competitive elections at all, are rare.
In 2009, the county grand jury recommended that the county consider single-district elections, in which only the residents of a particular district would elect that district’s supervisor. A Charter Review Committee is considering this recommendation, among others, in 2010. The Lawyers’ Committee for Civil Rights has also asked the county to consider changing its at-large election mechanism, saying that they “believe that the county’s use of an at-large election system dilutes the vote of minority residents”. For reference, the US Census Bureau reports that the county’s population is about 45% white (non-Hispanic), 24% “Asian persons”, and 23% Hispanic, with other groups in the low single digits.
Proportional Representation in San Mateo County
Suppose that San Mateo County were to adopt an STV-based proportional election system that elected five supervisors every two (or four) years, using the Single Transferable Vote. (We’ll consider variations on this theme below.) In such a system, each voter ranks candidates, 1, 2, 3, 4, etc, in order of preference. With five seats to be filled, a candidate must reach a threshold of 16.7% (1/6) of the vote to be elected to a seat. (Refer to the STV PR section for more about how STV works.)
Benefits of PR
Under STV-PR, a minority group the size of San Mateo’s Asian or Hispanic population can easily win a seat on the board, if electing a member of their ethnic group is a priority. Alternatively, any sufficiently large group of voters (more than 1/6 of total voters), no matter what their affinity, can win a seat, by joining together over issues of their own choosing: development, jobs, education, coastal conservation, environmental issues generally, public transit—you name it.
Hispanic voters in San Mateo are so scattered that no amount of gerrymandering would be able to create a district in which they can elect their own supervisor. It might be possible, with creative (and possibly illegal) gerrymandering, to carve out a majority-Asian district, at the expense of disenfranchising non-Asians in that district and excluding the voices (and votes) of Asians in other districts from having a say in filling the designated “Asian seat”—always with the doubtful assumption that, for Asian voters, ethnicity trumps all other concerns.
Another group that often feels disenfranchised at the Board of Supervisors is the county’s Coastside voters. The San Mateo Coastside is semi-rural and mostly unincorporated, and so directly under the jurisdiction of the Board of Supervisors, without the buffer of a city council. But the Coastside makes up only about 5% of the county vote, and even under district elections would not have the power to elect a supervisor. Moreover, the Coastside is evenly split on what is often the most important issue of local elections: development and growth. So even that 5% is further diluted, with no more than 3% voting as a bloc.
PR offers an alternative to Coastside voters (and of course to other voters in similar situations) by letting them ally with like-minded voters across the county, and, if those voters achieve 1/6 of the vote, win a seat on the board.
This is the core benefit of proportional representation: voters themselves have the power to form voting coalitions of their own choosing, not determined by gerrymandered district lines, and to determine their own political representation. The resulting board represents a cross section of the views of the voters on their priority issues, as should be the case with representative government.
Variations on a PR Theme
STV PR is flexible; several variations are possible.
- The discussion above assumes five-seat elections every two (or four) years for five board seats. Voters and challengers might prefer a two-year cycle with two-year terms, giving themselves a more frequent crack at the board, while incumbents might prefer the security of a four-year incumbency. We observe that State Assembly and House representatives live with a two-year term and election cycle, so it’s clearly not out of the question to ask the same of county supervisors.
- Staggered terms. In the present system, two or three supervisors are elected every two years to four-year terms. This isn’t ideal, from a PR point of view, since a two-seat election offers little scope for proportionality. The usual argument in favor of this arrangement is “continuity”: we won’t have an entire board of inexperienced supervisors. This is probably a baseless concern, given the consistency with which incumbents are successfully re-elected to their seats. One way that STV could be better combined with staggered terms and five seats is to elect three seats every two years, giving the winner with the least voter support a two-year term and the other two winners four-year terms. This implies an election threshold of 25%, not ideal, but significantly better than at-large or district elections.
- More seats. Whether or not PR is employed, additional seats provide opportunities for additional representation. The usual objection is the expense of additional supervisors. On the other side are the arguments that supervisorial expense is a relatively small part of the county budget, that increasing the number of board seats needn’t increase staff levels proportionately, and that an improvement in the quality of county government outweighs the additional expense. For PR, adding seats has the benefit of lowering the election threshold, making it possible for smaller voter groups to earn representation on the board.
- Odd vs even seats. County boards typically have an odd number of seats, presumably to avoid tie votes. But there’s nothing sacred about having such a number, especially as the number of seats grows. With a 10-seat board, for example, it takes six votes to pass a proposal, just as it does with an 11-seat board; the only difference is that the larger board has a 10% bigger pool of votes in which to look for those 6 votes. A majority vote wins; anything less, including a tie, loses. Which is how it works with a five-seat board when not all five members are voting.
These options lead to several PR variations (not an exhaustive list).
- As discussed above, a five-member board, with all its members elected with STV every two (or four) years.
- A five-member board, with three members elected every two years, one to a two-year term and two to four-year terms.
- A ten-member board, with five members elected to four-year terms every two years.
- An eleven-member board, with six members elected every two years, one to a two-year term and five to four-year terms.
- Least desirable from the point of view of proportionality, but a distinct improvement over the status quo, is to keep the existing election cycle of two or three seats every two years, but elect the seats with STV.
Any of these variations yields substantially better board representation than the current majority-take-all at-large election of district seats, or than single-district elections. The resulting board is more broadly representative, and fewer voters end up disenfranchised, unable to contribute to the election of their choice of supervisor.
In 2008, Mark Paul and Micah Weinberg of the New America Foundation published Remapping a Nation without States: Personalized Full Representation for California’s 21st Century, a proposal for a mixed-member system for the California Legislature.
Here’s their summary:
In this new system, the seats in a 360-member unicameral legislature would be apportioned among California’s eight regions according to their respective populations. Within each region, half the seats would be elected by district, the other half by proportional representation. For example, the Central Coast would have 11 legislators, the San Diego Border region 36.
Californians would cast two votes for Legislature on Election Day, choosing: 1) an individual to represent their district, and 2) a party to represent their region. One vote for a neighbor to represent you, one vote for the party to lead your region.
Under this system, the individual who wins the plurality of votes in each district election would be elected to the Legislature, as under the current system. Since there are more districts, there is a more personal connection between legislators and those they represent.
The remaining regional seats would be filled in order from ranked party lists of candidates. Party lists help maximize ethnic and gender representation as well as increase skills and knowledge within the Legislature.
Each party would receive seats in proportion to the “party list” votes it received. A party would need to reach a threshold of five percent to win seats under this system. More parties mean more choice for the citizens of California’s regions.
Individual district candidates could be included on the regional party lists. If they win their districts, party list apportionment would skip to the next candidate on the list. This innovative system will improve representation on the regional level and reinvigorate California’s Legislature.
In 2010, Professor Allan Ides, Loyola Law School Los Angeles, published Approximating Democracy: A Proposal for Proportional Representation in the California Legislature, another proposal for a mixed-member California legislature.
The State of California elects a bicameral legislature through a first-past-the-post electoral system. Beginning in 2012, the election of California state legislators will be through an open primary, two-round system. Neither of these electoral systems provides for proportional representation in the legislature. Accordingly, neither system leads to a truly representative democracy. The author recommends that the state adopt a proportional representation electoral system — specifically, a mixed-member system — as a more democratic alternative plurality/majority systems now in place or projected to be implemented in 2012. In addition, the author explains why the state legislature should be unicameral and increased in size from the current 120 legislators to 320.
For comparison, see Germany’s mixed-member system.