Back in a bygone era of tech, someone had the bright idea to combine a television with a video cassette recorder, figuring that viewers would want to both record shows directly from their television, and play back videotapes through their television. The idea was sound; the execution was not. The manufacturers combined not-very-good screens with not-very-good VCRs. They ended up with not-very-good appliances.
Portland's measure 26-228, proposed by the Portland Charter Commission that was appointed for this purpose, would increase the city council from 5 to 12 and divide the city into districts, with three councilors elected from each district. Under the proposal the mayor would have the right to introduce ordinances and break tie votes, but would not be a member of the city council. The mayor's primary function would be to supervise the City Administrator, a new position.
Measure 26-228 would also change how the city counts votes. The mayor and city auditor would be elected by ranked-choice voting, a system used in some other American cities that allows a candidate in a crowded race who is no one's first choice but everyone's second or third choice a shot at being elected. The city councilors, however, would be chosen by a sort of hybrid method (think TV plus VCR here) in which voters would rank their candidates by preference, as with ranked-choice voting. The difference comes in how the ballots are counted. Let us suppose that 90,000 ballots are marked in a particular council district, with three persons to be elected. The three persons would each need 30,000 votes to be elected. Now imagine that the race has attracted seven candidates (call them A through G, like notes on a scale). If 42,000 people mark A as their first choice, then A will keep 30,000 of those votes and be elected. The other 12,000 votes of A will be reallocated to the second choices of those voters. After the excess A ballots are allocated, the lowest of the remaining 6 candidates, let's say it's candidate G, will be eliminated, and G's votes will be reallocated to the voter's second choice, if the second choice is not A, who has already been elected, or G because it's a transferred vote of someone who ranked A first and G second. I think those ballots get reallocated to the voters' third choices.
Then the process starts anew with the candidates B, C, D, E, and F. If one of them now (let's say it's B) has more than 30,000 votes then B is elected and the excess votes of B are now allocated among C, D, E, and F, who may have been ranked second, third, or fourth on those ballots. The lowest finisher, let's say it's now F, is eliminated, and F's votes are allocated among C, D, and E. The process may continue for one or two more rounds, depending on how the votes are cast.
One problem with devising a vote-counting method for a race with more than two candidates is that in 1951 Kenneth Arrow proved that it is impossible to design a vote-counting system that is assured to yield a winner and that satisfies five basic requirements of fairness: first, that it not produce a cyclical result in which the voters prefer A to B, B to C, C to D, and D to A; second, that the result does not always match the preferences of any particular individual (i.e., no dictatorship); third, that it can handle any possible combination of voting alternatives (e.g., people who vote for fewer than the maximum number of candidates, or who have odd orders of preference); fourth, that it respect head-to-head alternatives (e.g., if more than half of the voters rank A above B, then A will finish the count ahead of B); and fifth, that the system respects unanimous individual preferences (e.g., if everyone ranks A above B, then under no circumstances can B win and A lose). One person of the many who submitted public comments, James Kahan, mentioned Arrow's Impossibility Theorem. I don't know if the Charter Commission discussed it, and ranked-choice voting survived for the mayor's and auditor's seats but not for the city council.
What did survive for the city council is a kludge - a collection of ill-fit parts without a coherent design. The Charter Commission had a good idea when it proposed to remove the "commissioner" part of the city council's duties and to make them simply councilors to establish policy that the mayor, through the city administrator, would then implement. I'll agree that we are ready to elect councilors from districts to ensure that each part of the city is known to someone on the council, which is rarely the case. (The Laquedem memory goes back to a time when four of the five councilors lived in Irvington or Alameda.) I'm not persuaded that the council has any systemic problem that we can solve by tripling its size and converting the governing body from a council to a jury-sized debating forum. I recommend a NO vote on Measure 26-228. I look forward to the proposal that Commissioner Mapps intends to bring forth in the spring.