Proportional Representation in Cambridge, Massachusetts

Background Information

  • Cambridge is a city of more than 100,000 people. Its City Council consists of nine members elected at-large, and the School Committee consists of six members elected at-large.  Both city elections use the choice voting method of proportional representation (also called “single transferable vote”, preference voting and “the Hare system”) since 1941.   
  • Any cohesive group that numbers more that 10% of the voters may elect at least one member of the city council. Any group that numbers more than one seventh of the voters (14%) may elect at least one member of the School Committee.
  • Proportional representation is deeply rooted in Cambridge political traditions. There were five referenda to repeal choice voting that all failed: with the system now fully accepted, there have been no repeal efforts in decades.
  • PR has survived legal challenges, most recently in 1996. The Supreme Court of Massachusetts deemed choice voting to be constitutionally valid. Note though, that the legislature has repealed the Plan E form of PR government, with Cambridge as an exception grand-fathered into the legislation.

Turnout and voter registration

In his article Preference Voting and Voter Turnout , George Pillsbury compared voter turnout in Cambridge to turnout in three other Massachusetts cities with similar demographics and types of municipal elections, but using winner-take-all elections (either at-large or mixed).

Average Turnout Over 3 Decades (% of registered voters)

Decline in Registered Voters Between the years 1961 and 1991
 1961 1991 Decline 
Cambridge 49,387 44,794 9,3%
Somerville 47,328 39,546 16,4%
Medford 35,232 31,698 10,0%
Worcester 95,062 69,583 26,8%

The results show that Cambridge had the “least decline” in both voter turnout and voter registration over these 3 decades when voter turnout has fallen precipitously in all Massachusetts and US elections. This happened despite the fact that all other 3 cities had strong mayors compared to Cambridge’s Plan E city manager form of government – turnout is usually higher in local elections with mayoral contests.

Recent Voter Turnout Decline

For the 2007 Cambridge elections, everyone knew there would be a big drop. One contributing factor was that two long-time major vote-getters, Anthony Galluccio and Michael Sullivan, had been elected to other offices and were out of the running. Normally this draws out new candidates who bring out new voters, but the timing of the departures was late enough that few new, credible candidates emerged. Open seats usually bring greater turnout, but not this time. The lack of any significant issues also led to a decline in turnout.

Voter turnout in municipal elections and primary elections has plummeted in many places in Massachusetts in recent years. The drop in Cambridge wasn't nearly as bad as the drop in Boston, and cities like Worcester saw drops that were well below even the most pessimistic predictions. This can be explained by the fact that people are likely not to see as much relevance in local elections now, especially now that rent control has gone and access to City jobs is less of an issue as higher-income people buy up condominiums in Cambridge. Voter turnout generally rises when the stakes are higher and they were a lot lower in 2005. How the current economic situation influences voter turnout is yet to be seen.

African-American representation

Historical overview

African-Americans have been able to elect representatives to both bodies in almost every election in the 1960’s and 1970’s even at a time when their share of the city’s population was less than 10% and below the threshold:  African-Americans often have held a higher percentage of political seats than their proportion of the total population.
Once African-Americans crossed over 10% of the voting age population in 1980 (it is 12%), they have always had representation on both bodies, sometimes with two representatives on a body (1971: two city councilors, 1993 & 1997: two school committee members, 2001: two city councilors & two school committee members).

2001 election pattern   

African-American voters in Cambridge tend to vote with their first choice along racial lines (as to other groups, such as Italian-Americans). They almost always rank candidates outside their choice farther down the ballot, but also take race into consideration with lower rankings along with political “slate” endorsements.

In 2007, for example, the 1st choice voters of winning African Americans Denise Simmons and Ken Reeves’ most often put the other candidate as their 2nd choice even though they ran on different slates. Denise Simmons was the most frequent 2nd choice candidates of Ken Reeves electors (30.3%) and inversely, Ken Reeves was the most frequent 2nd choice of Denis Simmons electors (19.2%). Yet, they ran on different slates.

Precincts with high African-American populations also gave the most support to African-American candidates. In 2007 (as well as in 2001, 2003 and 2005) Ward 2/Precinct 1 overwhelmingly gave its 1st choice votes to the African-American candidates for City Council and School Committee. For instance, for the City Council in 2007, the leading candidate in this precinct was Denise Simmons (25% of first choice votes), followed by another African-American candidate, Ken Reeves (21% of first-choice votes). This first choice pattern appeared throughout the city.

This cohesive voting allowed Simmons and Reeves both to be elected in 2007 (as in 2005). When Larry Ward and Kevin Moore, two others African-American candidates, were eliminated, most of their votes moved to Simmons (who gained 120 votes thanks to this transfer pattern) and Reeve (95 votes) helping to elect them.

Comparative data

Representation of the African-American community in City Councils in 2008

  Type of election 
 Share of African-American council membersProportion of African Americans in the population (2000 Census)
CambridgeProportional Representation 22% 11%
SomervilleAt-large winner-take-all + wards 9% 6,5%
MedfordAt-large-winner-take-all 0% 6,1%
WorcesterAt-large winner-take-all + wards 0% 6,9%

PR voting vs. Winner-take-all at large voting.

In choice voting, in 2001 African Americans Harding and Price were both elected to the school committee. Under a simulated winner-take-all election, Price would have lost. In 1999, Ken Reeves (the only African-American candidate) won a city council seat with a margin of 314, due in part to being the next choice of losing candidates. Under a simulated winner-take-all election, a change in 23 votes would have caused his defeat.

Women’s representation

Proportional representation has allowed women to achieve much greater representation than in other methods of election: Between 1997 and 2001, the City Council and School Committee had female representation between 1/3 and 2/3 of each body.

Representation of women in 4 Massachusetts City Councils in 2008
   % of women in the City Council

Efficient elections


In 1997 the Election Commission computerized the City of Cambridge PR elections using a precinct-based optical scanning system and specially designed software. The PR count, which used to be performed manually during the course of a week by a staff of over a hundred, is now completed in a matter of minutes through the electronic sorting, counting, and transfer of votes.
Unofficial results are available on election night. These results are “unofficial because all ballots have not been counted. The tabulation does not include ballots with write-ins or ballots marked in a way that cannot be read by the scanners. These are auxiliary ballots that must be processed manually and added to the computer totals. They are added on the day after elections. Only then are the results declared to be officials.

To learn more about the unfolding of a PR election in Cambridge, check the Cambridge Board of Elections official website :

Error rate

In Cambridge, the elections have an average of 2% error rate.  This figure includes both incorrectly marked ballots and blank ballots where the voter may have only participated in a higher-level election.  In the future, to further reduce this error rate, Cambridge could allow error correction for the voters.

There were very few invalid ballots.

    % of invalid ballots
City Council election
2005 School Committee election
2007 City Council election 0,64%
2007 School Committee election
  Average 3,16%

Variation in invalid ballots

Cambridge elections hava experienced a wide variation in invalid ballots. There are two main reasons for this. Beginning in 2005, the firmware in Cambridge ballot scanners was (finally) changed to reject incorrectly cast ballots. Overvoted ballots giving the same rank to more than one candidate (a "vertical overvote") were kicked back out and returned to the voter with a list of possible errors. This had been the most common error in invalid ballots over the years. It's important to point out that a ballot with two #1 votes but with correct subsequent rankings would not previously been classified as invalid - due to the presence of at least one unambiguous valid choice. With the new ballot rejection procedure, an overvote at any ranking would be rejected even if there were other valid rankings. A request was to reject ballots with skipped rankings (often due to an insufficiently filled-in bubble on the MarkSense ballot), but was not implemented. If the voter wants to cast his incorrectly filled-in ballot anyway, there is an override option. Use of this override varied somewhat from precinct to precinct and there were still some overvotes, but they dropped to a small fraction of what they had in previous years.

The other reason for variation over the years is that the Election Commission has not always been consistent in how they handle blank ballots, either ballots left blank intentionally or inadvertently, or unvoted ballots that were simply placed in with other ballots by a warden at one or more precincts. It's also very common for a voter to submit a voted ballot for City Council and pass in a blank ballot for School Committee. These are separate cards.


It's certainly fair to say that #1 rankings are far more significant than later rankings because most ballots will never be transferred in a typical election. Everyone needs to understand this, and the political posters and bumper stickers in Cambridge attest to this fact by the appending of "#1" to any candidate's name. A #2 or subsequent ranking generally matters only when a) the #1 ranked-candidate is likely to lose, and (to a much lesser extent) if the #1 ranked candidate will have a very large surplus. This reality is reflected in the behavior of the candidates. Candidates who have a good chance of winning will generally follow around and speak highly of candidates who are in any way similar to themselves and who have little chance of winning. Those are the only #2 votes (and #3 votes, etc.) that might possibly find their way toward the quota of the winning candidate.

The main reason why voters should vote deep occurs when there are organized slates. Some slate candidates may win easily, but whether or not your slate wins a majority of seats may well be determined in the later rounds when many of the #1, #2 and similarly ranked candidates have already been elected or eliminated. That was the big advantage in years past in Cambridge for the slate endorsed by the Cambridge Civic Association (CCA). They always encouraged voters to "vote the whole slate, in any order". Naturally, they couldn't prescribe any specific order as this would not go over so well with slate candidates not at the top of the list. In fact, the CCA always printed up equal numbers of slate cards with each of their endorsed candidates printed at the top of the list (and often cyclic alphabetically after that).

The CCA effectively went out of business almost ten years ago. They continued to formally endorse candidates for a while, but they did no promotion. They made no endorsements from 2005 onward. There really are no significant organized slates in Cambridge now, though many long-time voters still remember who is affiliated with what groups and continue to cast their votes accordingly (like the smile on the Cheshire Cat). If there's any grouping going on, it tends to now be among School Committee candidates. There was, for example, a group of three School Committee candidates in 2007 who were organized into what some called "the Walser slate" the purpose of which was to assure the election of a candidate to succeed Nancy Walser. It's fair to say that this slate was effective in electing Nancy Tauber (though it could well have been Gail Lemily-Wiggins) as she vaulted past Richard Harding in the deciding round.

Recent Articles
October 19th 2009
A better election system
Lowell Sun

Election expert Doug Amy explains how choice voting can "inject new blood" into the elections of Lowell (MA), and give voters a greater incentive to participate.

October 16th 2009
Haven't Detroit voters spoken enough?
Livingston Daily

In Detroit, there have been three mayors in the past two years and the current one has come under scrutiny. Perhaps a system like instant runoff voting will help bring political stability to motor city.

August 21st 2009
Black candidate for Euclid school board to test new voting system
Cleveland Plain Dealer

Limited voting, a form of proportional voting, will be used in Euclid (OH), in the hopes of allowing better representation of minorities.

July 2nd 2009
Reforming Albany
New York Times

FairVote's Rob Richie responds in a letter to the editor making the case for proportional voting systems to bring substantive reform to New York's legislature.