The STV proportional representation voting method

In the last post, I described the need for proportional representation to establish a more democratic and fair system for electing our representatives. I originally began thinking of the topic when I recently was browsing in the library and came across a short book entitled "Proportional Representation: The Key to Democracy" by George H. Hallett, Jr. The book was originally published in 1937, and is a short read at just a touch over 100 pages (on small size paper, so could be easily read in an afternoon).

What caught my interest was that there were several books in the library around that time period advocating for PR, and while casually flipping through the book, I saw references to the fact that cities like Cincinnati were using PR at the time. Is that still true today?, I thought. (Incidentally, Cincinnati used PR from 1925 until 1957; there have been attempts to bring it back since then that have lost by very close margins). FairVote actually provides an excellent brief history of proportional representation, which helps answer those questions. I'll provide some of those responses below, which actually echo most of Hallet's arguments for PR.

So I spent the last post talking about the high-levels of why we needed PR. Today, I thought I'd go through Hallet's small book and discuss exactly what STV PR is, and list the advantages and arguments for PR that it lists. I think all of them are good arguments just as applicable today as back then, and considering the book's age, I feel like it is worthwhile documenting older texts before they are lost to history.

First, the single transferrable vote (STV) method of proportional representation is a very straightforward name for the process. Voters are presented with the entire list of all candidates, and voters choose the candidate that they feel best represents them. As we will see, STV does away with the need for primaries, so voters can select from all available candidates and not just the single nominees of each party. However, unlikely plurality voting, voters can choose to rank the candidates in order from who is the first choice for their vote down to last choice. Voters are not required to rank all candidates; voters can pick only their top two favorites, or four, or any possibility. Voters therefore only vote for candidates they find acceptable, and are not required to pick any candidate they do not want.

When the votes are counted to determine the winners, we follow slightly different practices. A few alternatives exist, all roughly the same, so I'll explain the one that seems most common. The city sets a threshold for the number of votes needed to win a seat on the legislative body (city or county council, etc.). It might be say 1500 votes, or more, depending on the population of the city; the board of elections would set this number in advance. This means that ANY candidate that receives at least 1500 votes is on the council; the number of seats is therefore dynamic and depends on how many people turn out to vote. As the population for the city grows and more voters turn out, the council will automatically be expanded to ensure proper representation of all groups in the city. If population goes down and there are less candidates, the size of the council adjusts downward as well. (An alternative is to set a constant number of members on the council, and adjust the winning threshold based on the voter turnout, but I personally like the variable council size method better).

The tallying of votes is where the process is most significantly different. The votes are counted and sorted based on their rankings. First, everyone's first choice is counted and tallied. If there are any candidates that receive over the threshold (say, over 1500) votes, those candidates are declared immediate winners. Other candidates may not have received the full 1500, but are reasonably close. Other can be mathematically shown they will never win 1500 votes (because perhaps less than 1500 people chose that candidate in their rankings!). The candidates that cannot win are eliminated, and their votes are transferred to the second choice candidates. Simultaneously, candidates that received more than 1500 votes don't need more votes (they've already won! doesn't do any good to have more), so in order to try to balance out the minority votes from losing candidates, we also transfer the votes of the excess votes (over 1500) to the second choice candidates. In this way, both the majority and minority are represented and have their votes counted toward a candidate; no vote is wasted! We continue this process of transferring votes until there are no more candidates to transfer votes from. Then, ALL candidates over the threshold are winners. While slightly complicated in vote counting, the process is fairly intuitive and transparent (since the method of transferring votes can be calculated much like one balances a checkbook, anyone can confirm the arithmetic is correct) and ensures a council that represents the views of its citizens proportionally, ensuring everyone has a fair representation and a voice on the council.

This is only a brief overview of the method. We'll discuss how and why this works more as we review Hallet's arguments for PR below.

Specifically, Hallet points out the following advantages of STV PR:

• Effective voting, meaning that nearly everyone that votes will have at least someone from their list chosen to represent them. Even if it is not their top choice, nearly everyone will have their number two or three elected, and so most voters will feel comfortable with the results.
• Unanimous constituencies, meaning each representative is elected from a block of voters that agree on some issue and not simply voters that live in the same ward (as we do with our current single-member districts). Representatives can focus on issues that get them elected and not "straddling the line" in a district (in other words, the typical "I'm a centrist" that stands for nothing so as not to offend and get elected, as we hear today).
• Minority representation, meaning voters group themselves into minority "parties" based on important issues and can elect a representative even with a small group. They get some small representation rather than never having a voice in government under single-member districts and a two-party system. These minority groups can be "single issue" parties that feel shut out by the major parties, for example, by focusing things like fracking or single payer healthcare.
• Majority rule, meaning that while minority groups are represented, no minority group can force through their agenda without building a majority coalition with the other elected representatives that were elected on perhaps other issues. Most decisions made by the group will represent the will of the majority, whereas under single-member districting and two-party rule it is easy for a larger minority to take complete control despite up to half of voters completely disagreeing with the proposal. See Pennsylvania, where a majority of voters voted for Democratic candidates and yet Republicans control the state assembly with a 60% supermajority.
• New freedom in voting, meaning voters can worry less about "tactical voting" and vote for the candidate(s) and platform(s) they truly believe in. There is no concern about splitting the vote; one can rest safely knowing that if the first choice cannot win, the vote will be transferred to a second preferred choice rather than the last hated choice.
• A check to machine rule, meaning that the freedom to vote for any candidate typically means voters choose better candidates rather than settling on the establishment "political machine" candidates. It becomes harder for establishment candidates to win unless they truly represent the interests of the people, which leads to the next point.
• The transformation of machines, meaning that establishment political parties must bend their will to the will of the people and support popular policy or they won't be elected. There cannot be a lesser evil argument when your votes can be transferred. Establishment must adopt popular policy or will lose to another group that does.
• The gerrymander killed, since PR is applied to entire municipalities or to large permanent "super-districts", which completely kills the need for re-districting and hence partisan re-districting ("gerrymandering").
• A solution for reapportionment, making it easy to vary the number of representatives in the legislative body by simply requiring one representative for every set number of votes (for example, 75,000 votes). In this way, the number of representatives changes with the population automatically each election, and there is no need for reapportionment and re-districting after each census as we do now.
• Continuity, meaning that PR prevents sudden "landslides" that can push a minority into majority rule as under our current system. PR simply increases or decreases representation proportional to the votes, and popular leaders will remain elected, with only the unpopular ones losing their seats. Going from 49% of the vote to 51% means only an extra seat or two in PR, with each side retaining roughly equal control, while under plurality might mean the difference between total defeat or total victory.
• The development of leadership, allowing representatives to be independent leaders rather than "rubber stamp" the political party's goals. The party cannot threaten "losing endorsement" and ballot access, as happens under today's system. Anyone can be on the ballot regardless of party endorsement and still win, since voters can choose the best platform and not simply tactical voting.
• Development of interest, meaning that as the conversation changes from "lesser evil" to choosing the best platform, more voters will become interested in politics and elections and join the conversation. Many voters are today turned off by the angry discourse that results from a two-party system that pushes voting for a "lesser evil" that is still unacceptable. Why participate in a system where all of your options are ones you hate? With better options, we see more civic engagement, which is good for democracy.
• Reduction of fraud, as one would need to stuff a ballot with A LOT of votes under a PR system to get all of the candidates you want elected, and such a massive "landslide" would be immediately suspicious and investigated. There's less payoff for small fraud under PR, so less will even bother to attempt it.
• Elimination of primaries, since voters can simply pick their favorite candidates at once and have their votes transferred. There is no need to hold a separate, expensive election just for party nominees, especially when such primaries operate under party rules and can be manipulated. Parties are free to endorse as many candidates in the election as they wish, but they cannot prevent candidates from appearing on the general election ballot as they do with primaries.
• Cooperation and good feeling, since all of the above reasons contribute to campaigns focused on issues rather than simply attacking and trying to beat a particular candidate. Talking about issues wins votes as people find they agree with your platform, but there is little advantage to attacking an opponent when it might alienate voters and they'll simply pick some instead of you or the opponent.

Anticipating questions and retorts, Hallet's book provides a chapter on answers to common objections to PR:

• Does PR promote racial and religious blocs? While some voters may vote strictly along racial, religious, or other lines, most voters focus on issues important to them. Should some voters stick to small groups, even then the proportionality of PR means that they would not get a majority of elected officials and would not be able to pass legislation on their own if they are too radical. The book points out Cincinnati and other cities that used PR in the early 1900s held several successful elections and never saw evidence of this happening.
• Does PR deprive localities of representation? If we don't have districts, do the representatives really represent each neighborhood or locality? This is sort of the inverse question of promoting blocs as above, and has a similar answer. In order for each locality to NOT be represented, a different locality would have to form a voting bloc and win more seats. Since the seats are chosen proportionally, it is very difficult to imagine any scenario where voters in a locality would rank candidates from other localities that did not have their best interests in mind above candidates from their own area. In practice, this issue was not seen either, since voters focused on issues which tended to benefit their locality or the entire area as a whole.
• Does PR help extremists? Some worry that extremist groups may win seats in PR. For example, a Nazi may try to win a seat. But this fear is again unfounded because of proportionality. A small extremist minority is unlikely to have votes to win even one seat, and even if they did, would not have majority rule to pass any extremist agenda. Any such risk under PR is overblown, as it is much easier for an extremist party to take advantage of our current system's unfair districts and two-party primary process to take majority rule. Consider that the "Tea Party" of 2009 was a very small group of Republicans but was able to target certain districts and win a large amount of seats in Congress, and influence the entire Republican party, whereas such a win would not have been possible under PR since the Tea Party was clearly not endorsed by a majority of Americans (or even a majority of Republicans). Tea Party would have at best won a couple seats in PR, not a takeover of party goals as was what happened under our current system. You need majority backing in PR to win a majority, but that's not true under the current system.
• Does PR increase the bargaining power of minorities? Certainly minorities would have representation in PR and so a bargain or compromise would sometimes be needed to win a majority vote, depending on the proportions involved. But again the issue is overblown, as any "risk" of needing to bargain with a minority is much less than our current system which can give minority parties outright control to do completely as they wish.
• Does PR make legislation harder? Since PR focuses on issues that affect the community, in most circumstances the representatives would all be focused on the same or similar issues affecting the whole community, and can make compromises to ensure it happens. Our current system focuses on winning elections only, leading to two major parties that constantly flip flop, getting nothing done. PR actually would make the system work more efficiently for things we all agree on instead of focusing on the "horse race" of who is "winning".
• Does PR break down one-party government? Effectively, PR allows for a one-party (effectively: no parties), two-party, or multi-party government, so if you feel on one side of spectrum or other, PR is neutral. PR chooses the best candidates based on the issues, and so will reflect the will of the people. If there are many important issues, there will be many parties. History has shown some cities to be satisfied with two parties, while others like New York saw a vibrant multiparty system form due to the large amount of issues for a city of such large size. It is up to the voters that live in the area to determine how best to structure politics, and PR works with any choice.
• Is PR hard to understand? For voters, it is easy to understand that you simply number the candidates in order from favorite to least favorite. The counting process is a little more complex, but most voters already don't know the process of handling votes under today's system, so not much would change in terms of what voters need to know. PR does open up the ability to monitor and challenge elections since the votes are centrally counted and the arithmetic easily verified, which would actually make elections more transparent than they are today.
• Does the PR count take too much time? People are used to hearing the preliminary results immediately on election night. PR would take a bit longer to go through and transfer votes to second choice, etc., but election workers can do this over the course of only a couple days. We can allow a day or two to ensure a correct count when we're choosing leaders for the next 2-4 years. And if you want a preliminary estimate, exit polls might help.
• Does PR lend itself to manipulation? Since PR requires several vote counts, during which we transfer votes from losing to winning candidates, the ballots will be counted several times by many poll workers with many poll watchers monitoring the entirely process. Thus, fraud and manipulation are actually harder to pull off and the system is much more transparent. More than worrying about PR or voting method, if we are concerned about manipulation, we should focus on electronic voting machines currently in use in Pennsylvania. Such machines have no audit trail and may be easily manipulated. We should use paper ballots with PR for maximum transparency.
• Does PR cost too much? While PR would likely require hiring extra workers to count votes in a timely manner, the cost is much less than the costs of a primary and a potential run-off vote. PR replaces the need for these extra elections since votes are transferred automatically during the count. The net cost historically has been much lower in cities that moved to PR.
• Does PR make the ballot too long? In most cases the ballot won't be too long because there is no need to artificial raise the number of candidates to "split the vote" as is sometimes done in today's primaries and general elections. Even if there are many candidates, voters need not rank EVERY candidate, but only their favorites or the ones they know well. Even if the ballot has 30 candidates, voters can rank their top 5. In practice in cities that moved to PR, this was not an issue.
• Does PR make it harder to know the candidates? While there are potentially more candidates, most voters will find it easier to learn about candidates and their choices. Under today's system, most voters are not sure which district they are in, and media (newspapers, etc.) tend to devote their resources to top races like mayor or governor and less resources to every small district. The effect being that many voters don't have any idea about who the candidates are in their local races. If there are no districts, coverage must be more fair for all candidates rather than only the larger races and districts, if for no other reason than any candidate a voter hears about is someone that the voter can choose (since districts no longer restrict choice).
• Does PR make campaigning harder? While it is true the campaign area can be larger since there are no more districts, the flip-side is that candidates are no longer competing for a majority of votes, just the amount of votes necessary to win a seat. Therefore candidates can customize their campaign to reach the constituency they want: focus on a certain neighborhood or group to win the voters needed or a seat. Campaigning in every neighborhood in the whole city is therefore not strictly required to win. Since PR removes the need for primaries, this also frees up more time and money for the general election.
• Does PR decrease interest in elections? Since PR elections tend to focus on issues, more citizens tend to become interested. Historically, cities that used PR saw increased voter turnout.
• Does PR mean minority rule? This is an objection already seen and answered, basically. PR allows for some minority representation, but ultimately the majority of elected representatives set the agenda. Majority rule does not require that each representative strictly represent the majority, it requires a majority of representatives that decide a question to represent a majority of voters, via coalition and agreement.
• Does PR infringe the rights of the voter? The PR system presented allows for a single transferable vote (STV), such that you get a single vote that is transferred to your next choice should your top choice not be able to win a seat. This system is the most fair because if you give everyone multiple votes (say, 7 votes for 7 seats), the tendency will be that the largest group will elect its 7 candidates (Democrats will vote for the 7 Democrats on the ballot, for example) and the resulting election will not proportionally represent all voters. Each person choose a single representative that best represents them, regardless of party affiliation or district.
• Does PR leave elections to chance? This is an objection based on the method of counting votes described in Hallet's book. Specifically, the objection is to the idea that votes for a winning candidate over the threshold are also transferred to second choice candidates; essentially, how do you choose which votes are transferred and which stay with the first choice? I'll leave the full details to the book, but effectively PR's method of counting is so fair and accurate that randomly choosing ballots to transfer and which to not still accurately represents the whole (because of statistics over a large group of people). And if one is not convinced by mathematics, there are ways to more deterministically decide the question (although they are a bit more complex). The winners are not left to chance in either method.
• Is PR "un-American"? PR carries out the American principles of majority rule and equality of voting power better than any other system.
• Is PR unimportant? Some object to a campaign for PR, because there are "more important things" to focus on than PR and we just need to "elect the right people" to get it done. However, focusing on "electing the right people" to accomplish the will of the people is much more difficult under today's system. PR makes it easy to stick to the issues and elect the right people, so PR needs to be a top priority reform before we can focus on other issues.
• Has PR failed where it has been tried? Many cities have moved to use PR successfully, producing the most balanced and issues-focused budgets in recent history. Where the establishment has attempted to repeal PR, voters historically soundly rejected it. (There were cases of PR being removed due to changes in state governance, and racism as we'll see, but most voters that tried PR were extremely satisfied with the results). In a nutshell, nearly everyone that tries it prefers it and sees immediate results.

Hallet's book was published in 1937 as PR was becoming very popular; unfortunately, many cities and counties that adopted PR would see it phased out later in the 1950's and '60s. So I would like to add one more "modern" bullet point to the previous list of objections:

• If PR has obvious advantages, why did cities that used it in the past almost completely abandoned the method in the 1950s and '60s? As FairVote's brief history of PR discusses, PR worked almost too well. PR was incredibly successful at breaking down party machines and including minority representation in government, and Hallet's points above have proven to be true across the country in all municipalities that adopted PR. However, during the civil rights movement era and the "red scare", leaders from the two major parties -- angry at being unable to control elections and therefore government like they used to -- used racist appeals to fear of "Negro mayors" or fear of Communists winning elections under PR in order to convince the population to vote to repeal PR and move back to the older plurality systems. Unfortunately, the scare tactics were successful in most places, with white voters voting 2:1 to repeal PR due to those fears. So after decades of successful use of PR, many cities votes to scrap it, not for any problem with PR itself (as said above, previous repeal attempts were unsuccessful due to huge satisfaction with the system) but fear-mongering from the major parties. To me, this is actually all the more reason and evidence that PR is successful and a worthwhile goal, and that we need to fight to educate voters and establish PR in the US once again!

In light of the history of the PR movement and the points above, one final question comes to mind. If we want to implement PR today, how do we actually get it implemented around our current corrupt system that will not support PR for fear of losing power? If the Green Party would like to address concerns of "splitting the vote" and gerrymandering, what can it do aside from running more candidates in an unfair election system?

Hallet's book and FairVote's article actually provide historical evidence of one path to victory: utilize the voter referendum. What does that mean, though? I will post a follow-up blog post where I will discuss referenda at length, and I will review my research on whether this is feasible soon.