Race and Cambridge Elections

Choice Voting In Action

History of African-American Success in Cambridge

New brochure on choice voting in Cambridge

Cambridge, Massachusetts ��� Choice Voting Fact Sheet

Background Information

• Cambridge has a 9-member City Council ��� The PR threshold is 10%.
• Cambridge has a 6-member School Committee ��� The PR threshold is 14%.
• Even with the thresholds above, African-Americans have been able to elect representatives to both bodies in almost every election in the 1960���s and 1970���s ��� with between 5-10% of the total population. Hence, African-Americans 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, they have always had representation on both bodies, sometimes with even two representatives on a body (1971: two city councilors, 1993 & 1997: two school committee members, 2001: two city councilors & two school committee members).
• PR 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.
• Choice Voting 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 Note though, that the legislature has repealed the Plan E form of PR government, with Cambridge made an exception. Other municipalities may not now switch to Choice Voting. There have also been 5 referenda to repeal Choice Voting, but they all failed.

Racially Cohesive Voting

• African-American voters in Cambridge tend to vote along racial lines (as to other groups, such as Italian-Americans). Despite the existence of political ���slate��� endorsements, African-American candidates on different slates will all receive support from African-American voters. In 2001, Denise Simmons and Ken Reeves��� 1st choice voters most often put the other candidate as their 2nd choice.
• Precincts with high African-American populations also gave the most support to African-American candidates. In 2001, Ward 2/Precinct 1 overwhelmingly gave its 1st choice votes to the African-American candidates for City Council and School Committee. This pattern appeared throughout the city. 
• This cohesive voting allowed Simmons and Reeves to both be elected in 2001, as Ethridge King, a third African-American candidate, transferred enough votes to the other two, to elect them.

Choice Voting vs. Traditional Voting 

• In Choice Voting, in 2001 Harding and Price (two African-Americans) were both elected to the school committee. Under a simulated winner-take-all election, Price would lose. In 1999, Ken Reeves (the only African-American candidate) won a city council seat with a margin of 314, due in part to transferred support from losing candidates. Under a simulated winner-take-all election, his lead shrinks to 45 votes, so that a change in 23 votes would cause him to lose.

History of African-American Success in Cambridge

The Cambridge 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. In city council elections, a cohesive voting-bloc of 10% of the total population can guarantee their choice of one city council member, and a voting-bloc of approximately 14% can guarantee their choice of one school committee member. In 1903, before proportional representation was adopted, the first African-American, James Lew, was elected to the school committee. After his retirement in 1908, the community wouldn���t have representation on the committee for more than half a century. From 1941-1957, the African-American population in Cambridge was less than 5% of the population, and never elected an African-American member of the city council or school committee. It wasn���t until 1959, after proportional representation was adopted and the African-American population neared 5.3%, that they were able to elect a school committee member, Gustave Solomons. Solomons was subsequently re-elected to five consecutive terms on the school committee.

In 1963, after decades without represenation, Cambridge elected an African-American council member, Thomas Coates. Note though, that 10% is needed to guarantee a city council seat, but the population was only around 5.3% at this point. Hence, while Coates was re-elected in 1965, no African-Americans were elected in 1967. Coates eventually won his seat back in 1969, but as you can see, electoral success can be precarious with such a small percentage of the population.

See Table A Below

By 1970, Cambridge���s African-American population reached 6.8%, and in the following year���s election, two African-Americans were elected to the city council. With only approximately 7% of the population then, African-Americans held 22% of the city���s council seats. Unfortunately, the victory was not a long-lived one, as only one of the candidates was re-elected in the next election. Saundra Graham, however, ended up serving nearly two decades and nine terms on the council. The school committee elections during this period, on the other hand, were not as stable for African-Americans. From this group, one member was elected in 1971 and re-elected in 1973, but African-Americans had no representation in the 1975 and 1977 elections. Finally, in 1979 Henrietta Attles was elected for two terms, followed by Fran Cooper for four terms.

Note that in 1980, the African-American population in Cambridge finally crossed 10%, the percentage of the population needed to guarantee a council representative. Ever since then, African-Americans have been able to elect members to the city council, and even on the school committee. Sometimes, the group has held more than one seat on the council or school committee, as happened in 1993, 1997, and 2001. The level and consistency of this electoral success would not likely be possible under a traditional at-large, winner-take-all system, because a majority 51% of the population could dominate all 9 seats. In addition, the Cambridge African-American population is geographically dispersed, making super-majority, single-member districts unfeasible.

It is important to keep in mind though, that Choice Voting is not a guaranteed quota system. A 10% citywide population of African-Americans does not automatically get a seat on the council, merely an opportunity. For example, the Asian-American voting-age population in Cambridge is 12.3%. This is a significant figure, in that this group has crossed the 10% threshold needed to elect a city council member, and is very close to the near 14% needed to elect a school committee member. In spite of this, this group has not yet elected an Asian-American candidate, nor run a candidate, even as the Asian-American percentage of the city population has increased from 3.8% in 1980, to 8.4% in 1990, and 12% in 2000.

According to some accounts however, there is an explanation for this. The Asian-American population is mainly driven by those who are students at Harvard and MIT, and most students never vote in local elections. This perception is confirmed by the fact that the Asian-American voting-age population is larger than the overall Asian-American population. Typically when minority groups relocate or expand in an area, the overall population percentages will be higher than the group���s corresponding voting-age population, as children will make up a large part of the group���s membership. The fact that the Asian-American population is largely a student population helps explain why the voting-age figures are larger. In contrast, African-Americans make up 11.9% of the overall population, but only 10.1% of the voting-age population, while Hispanics are 7.4% and 6.5%, respectively. The Hispanic community being so small, has had limited success in electing members to represent them. There have been two members elected to the school committee from that group, Susana Segat and Sara Garcia. Garcia however, was only a one-term member, and though she ran with the community���s support likely won due to the fact that she was endorsed as part of the CCA���s slate, as 14% is needed to win. Segat, on the other hand, has won numerous terms on the committee, as a CCA candidate, however some accounts indicate that the Cambridge Hispanic community does not feel that she has represented them. Segat tends to win with CCA support and by playing to different groups, especially labor unions.

Despite this, African-Americans have formed a cohesive voting block, as many of their families have been fully integrated into Cambridge civic life for several generations. Like other longtime residents and city natives, African-Americans generally vote in every election. Some accounts indicate that these voters tend to vote pretty solidly for black candidates for school committee, and largely so for City Council, but to a lesser extent. So, as you can see a high minority population is not in itself determinative of electoral success, but it certainly provides a significantly greater opportunity under a proportional representation system. Further dampening the strength and cohesiveness of minority voting groups is the fact that the voting-age population figures do not take into account whether members of the racial groups are citizens and how long they have been in Cambridge. The 1990 Census data indicates that 22.3% of the city���s population was foreign-born in that year.

Another interesting aspect of Cambridge���s political representation, is that it has afforded women nearly proportional electoral success. In 2001 both the school committee and the city council were 1/3 female, and in 1999 the school committee was 2/3 female, while the city council was 1/3 female. Lastly, in 1997 the school committee was 2/3 female, while the city council was nearly 45% female. Women in countries using PR generally comprise somewhere in the range of 25% - 35% of elected officials, as opposed to less than 5% here most U.S. cities. Hence, Cambridge���s elections reflect the international trend of providing women greater representation through PR.

Table A:
Cambridge African-American City Council
& School Committee Representation
(format: Year #C.C.-#S.C.)

< TD >

2001    2-2

1941    0-0	1971    2-1
1943    0-0	1973    1-1
1945    0-0	1975    1-0
1947    0-0	1977    1-0
1949    0-0	1979    1-1
1950 Afr-Amer population = 4.3%	1980 Afr-Amer population = 10.9%
1951    0-0	1981    1-1
1953    0-0	1983    1-1
1955    0-0	1985    1-1
1957    0-0	1987    1-1
1959    0-1	1989    1-1
1960 Afr-Amer population = 5.3%	1990 Afr-Amer population = 13.5%
1961    0-1	1991    1-1
1963    1-1	1993    1-2
1965    1-1	1995    1-1
1967    0-1	1997    1-2
1969    1-0	1999    1-1
1970 Afr-Amer population = 6.8%	2000 Afr-Amer population = 11.9%

Slate Analysis & Race - 2001 City Council Elections

In Cambridge, voters tend to follow along lines of voting for the progressive Cambridge Civic Association (CCA) slate endorsements or else voting along independent lines. The official CCA candidates in 2001 were Davis, Murphy, Pitkin, Simmons, King, and Pitkin. Decker was not a CCA candidate, but her voters were for the most part. The independent candidates were Galluccio, Maher, Reeves, Sullivan, and Toomey. Note that Reeves is generally seen as a liberal, though he no longer has the endorsement of the CCA.

Following the distribution of 2nd choice votes, several trends emerge. If a white CCA candidate was chosen as a 1st choice, an African-American CCA candidate never received the most or even the 2nd most 2nd choice votes. (insert table) If a white, independent, non-CCA candidate was chosen as a 1st choice, an African-American non-CCA candidate never received the most or even the 3rd most 2nd choice votes, and an African-American CCA candidate did even worse, garnering not even the 7th most 2nd choice votes.

Further evidence of racially polarized voting lies in the analysis of the two strongest African-American candidates, Denise Simmons and Ken Reeves (as determined by # of 1st choice votes). Simmons was a CCA slate candidate and Reeves was an independent, non-CCA candidate. Simmons voters broke the slate lines to give Reeves the largest percentage of their 2nd choice votes. Likewise, Reeves voters broke slate lines to give Simmons the largest number of 2nd choice votes. So essentially, those voting for white candidates first tended not to vote for African-American candidates of similar politics, but those voting for African-American candidates first tended to vote for other African-American candidates next, regardless of slate endorsements. For example, Simmons received between 12-13% of 2nd choice votes from those voting for white CCA candidates, but received 34.6% of Reeves' 2nd choice votes. Likewise, Reeves received a paltry 3-8% of 2nd choice votes from other non-CCA candidates, but received 20.8% of Simmons' 2nd choice votes. In short, racially cohesive voting was a factor in electing both Simmons and Reeves.

See Table B Below

Under the Choice Voting system, both candidates needed 10% of the vote, 1713 votes in this case, in order to be elected. By the 12th count, out of 14 counts, neither of them had crossed the threshold. However, in the 13th count, a CCA endorsed, weaker African-American candidate, Ethridge King, was eliminated, and transferred enough votes to Simmons to push her over the winning threshold, and transferred 120 of his 587 votes to Reeves - which allowed Reeves to overcome the last remaining challenger in the election, even though Reeves and King were not running on the same slate. In fact, had this been a traditional winner-take-all election, King could have jeopardized the election of Reeves. King received 378 votes, while the margin separating Reeves from election or defeat was only 124 votes. In a traditional election, a stronger King could've sapped more votes from black voters, thus causing Reeves to lose along with him. Choice Voting allowed African-American voters to vote for King without worrying about their vote actually working against them.

Lastly, Reeves was an independent, non-CCA candidate, yet his 1st choice supporters gave King, a CCA candidate, virtually the same percentage of 2nd choice votes as did those voting for CCA candidates first. Specifically, of the voters who chose a CCA-endorsed candidate as their 1st choice, 3.6%-6.1% supported fellow slate-member, King as their next choice. For those voters choosing a white, non-CCA candidate as their first choice, only between 0.4%-2.4% selected King as their 2nd choice. Reeves��� voters though, crossed political lines to vote for King 5.6% of the time, nearly the same percentage as the CCA candidates. Meanwhile, Simmons and Reeves, the two other African-American candidates, had the highest percentages of 2nd choice votes for King.

Anecdotal accounts also indicate that Simmons and Reeves have also translated their support from African-American voters into support for issues of importance to African-Americans. Simmons especially was known for consistent support of African-American educational opportunities during her long tenure on the school committee.

Crossover voting of the type described above has also been seen for other ethnic groups too. For example, in at least one early 1990���s election, the Cambridge Italian-American population gave much support in terms of 1st and 2nd choice votes to two Italian-American candidates, even though they were running on opposing slates.

Table B:
2001 African-American City Council Candidates
(All figures based upon top ten candidates &
includes Decker as CCA)

Denise Simmons (CCA Candidate):
Avg. % of 2nd Choice Votes received from White Ind. Candidates: 3.1%
Avg. % of 2nd Choice Votes received from White CCA Candidates: 12.95%
% of 2nd Choice Votes received from Reeves (Ind.): 34.6%

Kenneth Reeves (independent candidate):
Avg. % of 2nd Choice Votes received from White Ind. Candidates: 4.675%
Avg. % of 2nd Choice Votes Received from White CCA Candidates: 5.075%
% of 2nd Choice Votes received from Simmons (CCA): 20.8%

Ethridge King (CCA Candidate):
Avg. % of 2nd Choice Votes received from White Ind. Candidates: 1.225%
Avg. % of 2nd Choice Votes received from White CCA Candidates: 5.15%
% of 2nd Choice Votes received from Reeves (Ind.): 5.6%
% of 2nd Choice Votes received from Simmons (CCA): 6.1%

2001 School Committee Elections & Race

In the 2001 School Committee Elections, another interesting dynamic appears. Harding was the African-American candidate supported by a large African-American population, while Price was the African-American candidate supported by a large white, liberal population. The largely African-American Harding voters were the most racially cohesive voters, as the largest percentage of their votes were ���bullet��� votes. 24.2% of Harding voters only voted for Harding, and did not list any 2nd choice votes, despite the fact that Harding was part of the CCA slate, which contained five other candidates. Price, the other African-American candidate, still received the highest amount of 2nd choice votes from Harding voters as their first choice vote. The largely white Price voters, on the other hand, selected the white candidate, Turkel, the most often as their 2nd choice, though Harding still did well. In addition, those voters selecting white CCA candidates as their first choice never gave the most 2nd choice votes to an African-American CCA candidate. In fact, two white CCA candidates only gave between 2.9-4.6% of their 2nd choice votes to Harding. He actually usually did better than this with people running against his slate, as the non-CCA candidates gave Harding between 3.5-11.3% of their 2nd choice votes.

Precinct Analysis

Precinct and ward returns have shown that African-Americans in the city though, do tend to engage in cohesive voting. In the 2001 school committee elections, for example, African-American candidate Richard Harding was elected, with most of his core support of 1st choice votes coming from precincts with strong African-American populations. Ward 2 has a large African-American constituency, and gave 41% of its 1st choice votes to Harding, the largest percentage given to any candidate. The next highest candidate only received 10% of 1st choice votes in the ward. Similarly, in Wards 1, 7, 8, and 9, which have very low African-American populations, Harding���s share of 1st choice votes dropped to 3 or 4%. In fact, this racially cohesive voting spills over to the City Council races too. For example, Harding received the most 1st choice votes in six precincts in the city, and those same six precincts all gave either Simmons or Reeves the most 1st choice votes: W2/P1, W2/P2, W3/P4, W4/P4, W5/P4, and W6/P1. Moreover, in all but one of those precincts, Simmons and Reeves came in 1st and 2nd.

The election of Harding and Price also elucidates yet another point about Choice Voting: this method is not a quota system, as demonstrated by the fact that in 2001 African-Americans held 33% of the school committee seats, while only making up 10% of the voting-age population. Proportional representation, it should be noted, does not limit African-American representation to its percentage of the population.

In 1999, a similar dynamic existed in the school committee elections. There were three African-American candidates: Denise Simmons, Don Harding, and Alvin Thompson. Harding and Thompson were eventually eliminated, but voters for both candidates supported Simmons the most often as their 2nd choice, offering her between 20 and 30% of their second choice votes. Those votes helped propel Simmons to reelection.

Choice Voting vs. Traditional Winner-Take-All Elections

The Choice Voting system in Cambridge has many clear advantages for minority representation over the traditional winner-take-all method of election. By allowing for proportional representation, Choice Voting only requires that a minority population is at least 10% of the total population in order to guarantee a City Council seat or 14% for a school committee seat. Under a typical winner-take-all system, a 51% white-majority can dominate all nine seats of the council or all six seats of the committee.

In the 1999 City Council race, Ken Reeves was the only African-American running. Under the choice voting system, he was the 8th candidate elected, out of nine seats to be filled. He was 314 votes ahead of the candidate defeated in the final round, which is a sizeable margin at that point in the race. However, under a simulated traditional at-large, winner-take-all election (where all 9 ranks are weighted as equal votes) Reeves drops to 9th place and is only 45 votes away from being defeated. A change of 23 voters would cause Reeves to lose in that election. Under the traditional system, voters can hurt their favorite candidate by voting for others, but under Choice Voting, voters can safely rank their votes without endangering their favorites. Hence, in 1999 Reeves initially started with only 1,141 1st choice votes, but needed 1,713 to win. After eliminating weaker candidates and transferring their next best preferences, Reeves finally crossed the threshold needed to win. In traditional elections winner-take-all elections, you can hurt your candidates by voting for others, so there is incentive to vote for only one person. The zero-sum nature of the traditional system discourages coalition building and creates an adversarial process. Hence, even in if there is only one African-American candidate running in a nine-member race, winner-take-all makes it harder for he or she to win.

An even more telling example though lies in the 2001 school committee race. Here, there were two African-American candidates: Alan Price and Richard Harding. Under the Choice Voting method, Price and Harding were elected 5th and 6th, in the race for six seats. However, if we simulate a traditional winner-take-all election, where the rankings are weighted as equal votes, Harding comes in 6th place and is barely elected, with Price losing the election and coming in 7th. This reveals the negative elements of winner-take-all elections in two aspects. The higher threshold in that form of election makes it harder for minorities to win seats, and it could end up pitting African-American candidates against each other, as they must fight for the same voters. Under Choice Voting though, Harding and Price ran on the same slate and could help each other win.

Also, under Choice Voting, African-Americans, while only 10% of the population, can more safely field more than one candidate, as the election system encourages coalition-building. In 2001 Simmons and Reeves both needed crossover votes from other candidates and were able to build this support through similar ideals, slate endorsements, or other forms of networking, and thereby build on their African-American voting vase. Simmons, in fact, got a sizeable amount of support through defeated, white, non-CCA candidates. Likewise, under a Choice Voting system it is not in the candidates��� interest to run ���against��� the other candidates, as they must often rely upon receiving 2nd, 3rd, and 4th choice votes in order to get elected. As a result, negative campaigning is almost non-existent in Cambridge elections. The candidates are more respectful of each other, realizing that they might need the support of each other���s voters. In addition, voter turnout in Cambridge elections is significantly higher than in comparable cities, such as Lowell, and even in nearby Boston. Cambridge does not even have a mayoral election, which boosts turnout in Boston city elections, yet still does better. Cambridge���s turnout can be attributed at least in part to the proportional representation system, which allows minorities that would otherwise be unable to elect a member to do so.

In this situation too, supermajority single-member districts would not be feasible, as the African-American population is geographically dispersed, with small concentrations in the eastern part of the city and in the western part of the city. In fact, Choice Voting allows African-American voters (or any political group) of much smaller percentages than exist in supermajority districts to elect a candidate of their choice, despite their dispersion. For example, in 1999 school committee candidate Richard Harding was largely supported by African-Americans and was able to get over 50% of his 1st choice votes from only 8 of the city���s 42 precincts. While some may see this as a negative aspect in terms of creating de facto districts, this is mitigated by the fact that despite campaigns that are often based on targeting certain neighborhoods, candidates usually try to reach out to larger bases, in order to try and capture some of the voters��� next best preferences. Certain candidates then, will enjoy great citywide support, while others may be supported by smaller pockets of voters, but in the end greater opportunities for a diversity of representatives are created. This is certainly more representative than a district-based system, which will still leave 49% of voters in each district without a representative, or an at-large winner-take-all system which can do the same thing citywide. Instead, Choice Voting allows Cambridge voters to coalesce along various lines of interest.

Others still, will complain that Cambridge candidates are forced to spend a great deal of money on their elections, as they must run citywide. However, as explained above, many candidates choose to focus on different regions of the city, and in fact the high cost of Cambridge elections is explained by other factors. According to one Massachusetts voting expert, Cambridge elections are pretty costly because there is a lot of money there and a lot of donors. Apparently Cambridge has one of the larger political donor bases, so it would have expensive elections regardless of the method of election. A comparable town, like Lowell, Massachusetts, would have cheaper elections. This difference is due to tremendous disparities in socioeconomic status and local civic engagement, Cambridge obviously having a wealthier and more engaged citizenry. Money is not always decisive in Cambridge elections, as the candidates that do the best are not necessarily the ones who spend the most money. For example, in the 1997 school committee elections, Denise Simmons spent $15,204.93 and received 2,043 1st choice votes. This is a cost of $7.44 per vote. Meanwhile, Fred Fantini spent a total of $1,155.25 and received 1,781 1st choice votes, at a cost of $0.65 per vote. Simmons spent more than eleven times as much as Fantini did for her first choice votes. Granted, Fantini eventually lost the election, but Simmons came in third place out of the six winners. The first place winner, Alice Turkel received more 1st choice votes than Simmons, but Simmons spent nearly three times as much money per vote. The same patterns appeared in the city council races that year.

Conclusion

Choice Voting has clearly provided African-Americans with electoral opportunities unavailable under a winner-take-all system. Cambridge voters, to a large degree, engage in racially cohesive voting. While the introduction of coalition-building in Cambridge has eased the election of African-Americans and women, Choice Voting was instrumental in allowing these opportunities to form. There have been five referenda (in 1952, 1953, 1957, 1961, and 1965) on whether to repeal or retain the Choice Voting system. Each time the vote was to retain it. Choice Voting has also withstood legal challenges, as recently as 1996.