One of the City Council candidates, John
Pitkin, approached me at a Danehy Park event in
September. His literature had the usual "Vote #1"
on it. I wondered, "Why don’t any of you ask
for the #2 vote? Asking for #1 sounds so greedy."
He replied disdainfully, "You have to get
number-one votes to win," losing my vote on the
spot. I think I understand Cambridge’s
"proportional representation" voting method — now
with the votes in, I can demonstrate that Pitkin
was wrong. "P.R." is a good voting system, even
if most people don’t quite understand the details,
and every vote counts, even number-twos and
beyond.
Here’s a summary of how P.R. works. First, you
tally up the number of valid votes, and set a
"quota" of how many votes are needed to get
elected — this year, that was 1,713 votes. Any
candidate who got 1,713 or more number-one votes
was elected on the "First Count." That applied
to three candidates, Mayor Galluccio, incumbent
Henrietta Davis and newcomer Brian Murphy. Their
votes over 1,713 are then distributed to the
number-two selection — 1,520 "surplus" votes
from the three immediate winners this year. Next,
any candidates who don’t have 50 votes are
declared "defeated," and their votes are
distributed to the next-ranked person on their
ballots. This year, 31 votes from the "under-50"
candidates were distributed. If the distributed
votes result in a candidate exceeding their quota,
they are declared elected, and their further votes
are distributed to the next-ranked candidate. Then
the candidate with the lowest number of votes is
declared defeated, and those votes distributed.
The process repeats until there are nine winners.
That took 14 "counts" this year, which means 14
rounds of distributing votes from winners’
surpluses or defeated candidate’s votes.
That’s a pretty complicated system, but
nowadays it’s all done by computer, so we no
longer have the long delays from people doing the
tallying. P.R. has been in place since 1941, and
it survived five referendum votes in the 1950s and
1960s, so we’re stuck with it. The main argument
against P.R. — that it required a long time to
count ballots — has become moot thanks to
computerization. The other argument against it —
that the "lottery" discounts number-two votes —
is demonstrably false.
How do the number-two and number-three votes
get counted? If you voted for Galluccio number-one
and then David Maher number-two, Maher would
immediately get your vote since Galluccio won on
the first count — Maher got 379 votes that way.
Actually, Maher would only have a chance of
getting your number-two vote, since there’s a
random selection of which of Galluccio’s votes are
counted as "surplus" — that’s why people
complain about the "lottery" component. Sure,
it’s possible that one candidate will benefit
disproportionately, but not likely, and "random"
means "fair." If you voted for Galluccio
number-one, Davis number-two, and Maher number-three,
then Maher would still get your vote, since Davis
had already met her quota on the first count.
How do the votes past number-three get counted?
In each "count," the candidate with the lowest
number of votes is declared defeated. This year,
the first over-50 candidate to be defeated was
James Williamson, and his 62 votes were then
distributed. On ballots where the number-two vote
was one of the three immediate winners, the vote
went to the number-three or number-four candidate
in sequence. As more candidates were elected or
defeated, the votes get distributed to the
number-five and six candidates. Ethridge King was
defeated on the 12th count, and when those votes
were distributed among the three remaining
candidates, two made it over quota and one, John
Pitkin, was defeated. If Pitkin had more
number-six votes, when the number-one through five
votes had gone to candidates already elected or
defeated, he could’ve won. In fact, more number-12
votes would have been enough to push him over the
edge — all that mattered in the 13th count was
whether Pitkin or Maher (the last candidate
elected) was ranked higher on the selected
ballots.
What about "bulleting?" People think it’s a
mark of true support for a candidate to vote only
for that candidate as number-one, and then for no
other candidates. That’s pointless under
proportional representation. Your number-one candidate
gets your vote no matter what, until she is
elected or defeated. If your ballot has no other
votes when your candidate’s fate is decided, it is
declared "exhausted" and the next random ballot
takes its place. Your additional votes can only
help other candidates — they can’t hurt your
number-one choice.
Back in Danehy Park, John Pitkin meant that he
needed a minimum number of number-one votes to
stay in the running. You do have to avoid the
"under-50" defeat, but those 50 can be either
number-one or two votes from popular candidates
who win on the first count. Pitkin got 48 votes as
number-two with Galluccio as number-one — all he
needed was his own number-one vote and his wife’s,
to avoid going "under-50." Not every number-two
vote gets counted — but some do, and even some
number-14 votes get counted, in the 14th count.
Pitkin beat Maher on number-one votes (1,091 to
1,017) but lost because Maher got more number-twos
and lower.
What about asking for number-two votes?
Candidates benefit from votes in any position on
the ballot, if they survive several rounds of
counting. Hence the number-two position and all
the rest down the line certainly do count. So go
out next time and vote all the way down past
number nine!
Jesse Gordon holds a Master's in Public Policy from Harvard's Kennedy School of Government.
.gif)
.gif)
.gif)
