The Case of Cambridge, Massachusetts
George Pillsbury
In national elections, countries
employing proportional voting methods have significantly higher voter turnout than
countries with winner-take-all election systems. This has led leading political scientists
like Seymour Martin Lipset and Walter Dean Burnham to the conclusion that proportional
voting systems encourage greater voter participation. Turnout is higher because of voters'
increased choice and ability to elect candidates of choice.
To study how this proposition translates
to local elections, I compared voter turnout in Cambridge, Massachusetts -- the only city
in the United States to use preference voting for all municipal elections -- to turnout in
three other Massachusetts cities with similar demographics and types of municipal
elections: Somerville, Medford and Worcester. The results support the idea that
proportional voting tends to encourage greater voter participation.
The main result showed that Cambridge had
the "least decline" in both voter turnout and voter registration during a time
when voter turnout has fallen precipitously in all Massachusetts and U.S. elections.
Cambridge ended the study about 10
percentage points above the other 3 cities despite the fact that all other 3 cities had
strong mayors compared to Cambridge's Plan E city manager form of government and that
turnout is ordinarily higher in municipal elections with mayoral contests.
During the early part of the study, voter
turnout was similar in all four cities -- in part due to strong Democratic, ethnic,
patronage political participation structures in all four cities typical of Massachusetts
politics at the time. In the 1970s, as political parties and traditional
"machine" politics declined, turnout fell sharply in the other 3 cities
(although less quickly in one city, Somerville, because of some strongly contested mayoral
races in the 1970s).
Background on Cities
Form of Government: For most of the period studied, Cambridge, Medford and
Worcester used a city manager/city council form of government; Medford and Worcester
changed to a stronger mayor in 1987. Somerville had a strong mayor and a city council
elected from districts and at-large.
Voter Turnout in
4 Massachusetts Cities
(Percent of Registered Voters)
Chart 1
Average Turnout Over 3 Decades
1961-1969 | 1971-1979 | 1981-1993 | |
Cambridge | 67% | 59% | 54% |
Somerville | 68% | 64% | 46% |
Medford | 59% | 55% | 44% |
Worcester | 64% | 51% | 45% |
Chart 2
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% |
Demographics: Cambridge and Somerville are the most similar of the four cities.
Medford is slighter higher income and Worcester has become, during the period, slightly
lower income.
Impact of Elected Mayor: Elections for a strong mayor generally produce more
turnout than an election for a city council, especially in a competitive mayoral election.
The figures in this study support this. In Somerville the turnout averaged 69% in the 5
elections where new mayors were elected between 1961-1981. Worcester and Medford both had
a more than 10% boost in turnout in 1987, the first year they both had a popularly elected
mayor, though turnout declined afterwards.
The main result showed that Cambridge had the "least decline" in both voter turnout and voter registration during a time when voter turnout has fallen precipitously in all Massachusetts and U.S. elections. |
Political History: Cambridge, Somerville and Worcester all had strong ethnic
political patronage machines in the 1950s through the 1970s. During the 1970s, like
elsewhere, these machines went into decline. At least one factor in this decline was
simply the changing demographics, as each city received an increasing number of racial
minorities and new immigrants. These patronage and ethnic based political networks likely
played a positive role in voter turnout in the 1950s and 1960s.
Cambridge vs. Somerville
Somerville's strong mayor elections
helped produce marginally higher turnouts than Cambridge's weak mayor elections in the
early part of the study from 1955-1981: Somerville's average was about 67% and Cambridge's
about 64%. Somerville's turnout was especially high when a new mayor was elected, as noted
above.
Cambridge moved ahead of Somerville in the
1981-1993 period with an average turnout of 54% compared to a 46% average for Somerville.
The question, then, is what factors made Cambridge's turnout competitive with Somerville
in the early part
of the study and move well ahead in the last 7 elections -- even though Somerville has a
turnout
advantage in its strong mayor form of government. Cambridge's preference voting system
almost certainly was a factor.
Cambridge vs. Medford and Worcester
Since
1960, Cambridge's turnouts have been higher than those of Medford and Worcester, two
cities with similar demographics and forms of government. The gap has been increasing.
The sustained and quantitatively higher
difference gives clearer evidence that Cambridge's preference voting system gives it an
advantage. The difference cannot be explained in terms of demographic factors because
Medford voters have remained wealthier. Also, as Medford's voting in even-year federal
elections is more consistently similar to that of Cambridge, greater voter apathy cannot
be considered a factor either.
This analysis is only preliminary and in
part demonstrates how difficult it is to isolate one factor in the decline of voter
turnout since the 1960s. However, the evidence certainly points to preference voting
providing more of an incentive for voters to participate than plurality elections.
George Pillsbury is a Board member of
The Center for Voting and Democracy. He is a 1994 graduate of the John F. Kennedy School
of Government, where this article was written.