Wednesday, November 9, 2011

This is What Democracy Does Not Look Like in Cambridge, Massachusetts

We've just had another municipal election here in Cambridge, Massachusetts. Elsewhere it's one man, one vote. Here it's you may fill in as many choices as you please, only one vote per candidate, only one vote per column, and do not use red to mark ballot. We end up with ballots that look like an SAT test from hell.

Voting in the city of Cambridge is more complicated than it needs to be. Cambridge uses a system of proportional representation that means you are asked to rank all the candidates in the city council race against each other (same for school committee). If you want, you can vote for all of the candidates by ranking every last one of them in your order of preference #1, #2, #3, etc.

When it came time to count the ballots for the 9 seats on the city council, here were the #1 votes for each candidate:

Candidate#1 VotesResult
Leland Cheung1974Elected on 1st count
Timothy J. Toomey, Jr.1654Elected on 1st count
David P. Maher1636Elected on 1st count
Henrietta Davis1407Elected on 9th count
E. Denise Simmons1224Elected on 13th count
Marjorie C. Decker1092Elected on 14th count
Craig A. Kelley1075Elected on 13th count
Minka Y. vanBeuzekom1011Elected on 14th count
Kenneth E. Reeves979Elected on 14th count
Larry W. Ward816Defeated on 13th count
Sam Seidel768Defeated on 12th count
Matthew P. Nelson523Defeated on 11th count
Charles J. Marquardt488Defeated on 10th count
Thomas Stohlman, Jr.336Defeated on 9th count
James M. Williamson171Defeated on 8th count
Gary W. Mello128Defeated on 7th count
Jamake Pascual57Defeated on 5th count
Gregg J. Moree54Defeated on 6th count

As the results column hints, it doesn’t stop with #1 votes under the proportional system. The “extra” votes from winning candidates over what was needed to win and “throwaway” votes for losing candidates get reallocated based on the #2, #3, #4 etc. preferences. For example, 434 #1 votes for Leland Cheung were redistributed to the other candidates based on the #2 votes on those ballots. Supposedly, these were selected randomly from among all the ballots cast for him.

That means the 979 votes for Ken Reeves weren’t actually what got him elected. It was all the #2, #3, #4, … #18 choices that got reallocated to him during the 14 successive ballot counts that put him over the top. That worked out to 35 votes from Leland Cheung, 18 votes from Timothy Toomey, 13 votes from David Maher, 2 votes from Jamake Pascual, 4 votes from Gregg Moree, 9 votes from Gary Mello, 8 votes from James Williamson, 16 votes from Thomas Stohlman, 21 votes from Charles Marquardt, 53 votes from Matthew Nelson, 107 votes from Sam Seidel, and 275 votes from Larry Ward.

That’s perfectly democratic, right? And perfectly opaque. It seems especially egregious that Ken Reeves’s election required redistributing 275 votes from Larry Ward, as he’s the candidate with the next highest vote total that Ken Reeves defeated. How does the “winner” getting to count votes for the “loser” make the election democratic?

The ballot could be made a lot simpler with no change to this year’s results by eliminating all but #1 votes. The 9 candidates with the most votes win, period. That would mean the ballot columns for #2, #3, #4, … #18 etc. could simply be eliminated, which would be 94% simpler. Yes, there might be some years where the 9th place finisher would actually rank 10th or 11th and out of the money after all the #2, #3, #2, #3, #4, … #18 preferences are reallocated. But, really, should someone’s #18 preference be able to cancel out someone else’s #1 preference? There is, after all, another election in two years.

The real failure of the Cambridge election system is in participation. Only 15,393 ballots were cast in the city council race and only 14,939 were cast in the school committee race. This in a city with 105,162 residents. That means each of the nine city councilors theoretically represents 11,685 people, between 80% to 90% of whom either can’t or didn’t vote. Of course, that could be because the great majority of us are happy with the job our city government is doing. Still.

That's how it goes in the People's Republic of Cambridge. The election results are in, and once again the people haven’t spoken.


Anonymous said...

Proportional voting done right is actually fair -- but it's typically done where the person with the least votes gets eliminated, and that person's votes then get distributed to other candidates. That way, you can vote for a Nader without wasting your vote.

Left Bank of the Charles said...

Anonymous, you are confusing proportional voting with instant runoff (a misnomer if you've ever watched them count the Cambridge ballots).

You're assuming the Nader voters would fill in the Gore oval for #2. Maybe they would, but as each voter can conduct their own instant runoff (the polls say Nader won't win, so maybe I should vote for my #2 Gore), the evidence is they won't.

And if you think the American public would stand for 14 rounds of ballot counting in a Presidential election, you really don't remember the Florida recount.

UserGoogol said...

Single transferable vote is just instant runoff with multiple seats. So instead of having instant runoffs until one candidate gets 50% of the vote, you hold instant runoffs until nine candidates get 10% of the vote. (In general, to fill N seats, candidates need 1/(N+1) of the vote.)