Climate change and political jiu jitsu

In 2010, Maine elected strident climate change denier Paul LePage governor with 37.6% of the vote. LePage narrowly defeated Eliot Cutler, a wealthy lawyer and moderate progressive who secured 35.9% of the vote running as an independent. The Democratic nominee garnered a mere 18.8%.

As governor, LePage has acquired a reputation as a particularly extremist brand of Republican. Citizens for Responsibility and Ethics in Washington (CREW) summarizes his record as follows.

Augusta is a dangerous place for anyone who gets in the way of Gov. LePage’s ALEC-written agenda. 

The first-term governor packed his administration with lobbyists and used his office to promote their environmental-deregulation agenda, and allegedly went so far as to fire a state employee who testified in favor of policies the administration opposed. 

Gov. LePage also attempted to gut his state’s open records act, and is under investigation by the federal government for trying to bully employees of the state Department of Labor into deciding more cases in favor of business.

In 2014, with his approval ratings underwater, LePage ran for re-election. And once again Eliot Cutler threw his hat in the ring as an independent. But moderate and progressive voters, weary from four years of crazy LePage, urged Cutler supporters to throw their weight instead behind Democratic nominee Mike Michaud. This was a particularly ironic strategy, given that Cutler had previously trounced the Democratic challenger. Nevertheless, that strategic thinking apparently had a great effect, as Cutler was reduced to a mere 8.43% of the vote. Unfortunately, that wasn't enough. LePage defeated the Democrat by nearly 5%. At the time of this writing (July 2015), LePage is on track to break Maine's record for overridden vetoes, and may face impeachment.

But something very interesting happened in this second election. Volunteers in partnership with the non-profit Center for Election Science conducted exit polling in three Maine cities, asking not only who the voters had actually supported, but who they would have supported had there been no limit on the number of votes they could cast. This system is known as Approval Voting, and the results were profound.

The following charts reveal the complete reversal that took place with Approval Voting. The purple bar represents the Republican, the pink represents the Democrat, and the orange represents the independent.

The impact of tactical voting is staggering, particularly with regard to independent Eliot Cutler (orange). However, it is not that surprising when we look at pairwise head-to-head match-ups based on the preference rankings we also elicited from voters.

Cutler simply dominates both of his major party rivals. This convincingly shows that his scant support in the official election was a result of tactical voting. Simply put, voters didn't vote for Cutler because they didn't think he could win. This is the quintessential self-fulfilling prophecy. But with Approval Voting, voters have absolutely zero incentive to avoid supporting their favorite candidate. If they don't think their favorite is viable, they can support any number of additional candidates. In this case, a left-leaning Cutler supporter could vote for Cutler and the Democrat, or vice versa.

While these two elections may seem like mere anecdotes, I believe they exemplify a broader problem that is pervasive within our political system. For instance, consider this headline from today.

What this headline tells us is that a good fraction of voters are considering who they'll vote for based in large part on their perception of electability. Just think about that for a moment. Voters are going to decide who will get to be essentially the most powerful person in the world, based not simply on who they actually support, but who they think others will support in the general election. If that's not an insane way to choose a leader, I don't know what is.

Moreover, think about the factors that determine our perceptions around electability. One of the biggest is fundraising success. Hence our vote-for-one election system massively magnifies the influence of money. Thus a switch to Approval Voting may also be the greatest antidote to the problem of "money in politics". Commonly discussed strategies of repealing Citizens United via Constitutional amendment seem comparatively far fetched.

Speaking of money in politics, Democratic stalwart and billionaire investor Tom Steyer recently spent $57 million to influence seven key races, including the 2014 Maine gubernatorial race. But for a fraction of that amount, he could have funded ballot initiatives to get Approval Voting adopted in a number of small US cities, where it could establish precedent and plausibly spread to bigger cities, and then on to higher levels of government. In relatively short order, it's conceivable that the US government could be radically transformed. Less two-party dominated, less polarized, less incumbent-favoring, and less controlled by wealth.

The challenge for voting reformers is that the topic is dry and mathematical, and emotionally removed from the kinds of charged issues of our day. Police brutality on a small scale can incite weeks of mass protests. But try to get anyone to march in the streets for unlimited votes on their ballot, and you just hear crickets. This kind of wonky esoteric policy change may require concentrated wealth from a particularly analytical mind. There are only a few individuals with the resources and the passion. When will they pick up a copy of Gaming the Vote and push for a fix to the core of our democracy?

Democracy for Hong Kong

On Wednesday April 22nd, 2015, Hong Kong's local government unveiled a new election reform package outlining a process by which candidates for Chief Executive will be pre-screened by Beijing loyalists. The initial proposal set off widely publicized protests last year, and this latest update offers only minor changes from that. This has led to backlash from pro-democracy lawmakers and activists who desire free elections without filtering by mainland China, which will virtually guarantee the exclusion of any pro-democracy candidates.

The proposal is set for a vote requiring two thirds of the 70 local legislators to support it for passage. While the 27 members who are overtly pro-democracy (the so called "pan-democrats") could veto the bill, that could be a risky move, as they might not subsequently be able to negotiate anything better. A recent article in the International Business Times describes the hard line stance of senior Communist Party officials saying, "Beijing will not make any concessions over Hong Kong’s electoral reform that deviate from its own principles". The article continues:

The result of the package’s rejection would be the continuation of the existing system, where a nominating committee effectively appoints the chief executive, and the continuation of dysfunctional government in one of Asia’s key financial centers, which experts says is a losing proposition for all parties... 

Other analysts offered guarded support of the proposed reforms, indicating that if pan-democrats do not take the deal currently on offer, that could be the end of the political-reform process in the region.

“Even though the package is quite conservative, it is worth getting through, because if we lose this opportunity, we really don’t know when we will have the next round of negotiations with the central authorities,” said James Sung, a political analyst and lecturer at the City University Hong Kong.

Opportunities to negotiate a compromise that would allow Hong Kong to continue its democratic development in a manner acceptable to Beijing seem slim.

The inflexible attitude of mainland China leaves reformers with very little (read almost zero) room for negotiation. Thinking tactically and acknowledging practical realities, any remotely viable counter proposal must accept the general framework of the current package as well the basic premise that the Communist Party will have substantial influence over the process. But in order to leave room for democratic progress, it must create a subtle "lever" which can be the subject of future negotiation.

What chess moves do we have at our disposal? As a co-founder of The Center for Election Science, I've spent nearly a decade studying electoral system design with a focus on game theory. Here's my three-step plan. The third step is the "lever" I mentioned.

1. Remove the current 2-3 candidate limit

The current proposal allows the 1200-member nominating committee to select only 2-3 nominees using Approval Voting. Approval Voting simply means that committee members may vote for as many candidates as they wish. There is also the requirement that any eligible nominee must be approved by at least half of the committee members.

Instead, allow any and all candidates who meet the 50% threshold to run in the general election. This should not seem threatening to Beijing, since the committee is a group of hand-picked Communist Party loyalists. But this move creates at least some choice, and more importantly sets the stage for future maneuverings which I'll get to shortly.

2. Institute Approval Voting for the general election

Though the current reform package calls for Approval Voting in the nomination round, it unfortunately specifies Plurality Voting (aka First-past-the-post) for the general election. Political scientists have known for decades that Plurality is the worst voting method ever invented. For instance see this analysis of 18 different voting methods by the Voting Power and Procedures department at the London School of Economics. Plurality Voting came in dead last, whereas Approval Voting was their top choice.

Why? Plurality Voting can lead to the election of polarizing and unrepresentative leaders whenever there are more than two candidates, as the current package allows for. This is known as "vote splitting" or the "spoiler" effect, and it would be a particularly serious concern if the 2-3 candidate limit were rescinded. Approval Voting is the simplest solution to support an election with multiple candidates. It uses ordinary ballots and counting procedures. The candidate with the most votes still wins. But Approval Voting is mathematically proven not to harm voters who support their sincere favorite candidate. This allows voters to choose based on merit rather than on electability. It also elects candidates with the broadest consensus, leading to political stability.

Best of all, Approval Voting is already part of the reform proposal. In their Consultation Report and Proposals document, the Hong Kong Special Administrative Region Government (HKSAR) briefly alludes to the benefits of Approval Voting.

there are also views which suggest that the “voting on each person seeking nominations” should be adopted (i.e., each member could support all persons seeking nomination, or support only some of such persons), so as to enable such persons to have more opportunity to seek NC members’ nomination on a fairer basis; and that members could consider each person seeking nomination more freely according to the merits of each person.

Chief Secretary Carrie Lam clarified in her speech, "Each NC member may vote for all persons seeking nomination, or vote for only some of such persons."

Even Beijing-friendly LegCo member Regina Ip endorsed Approval Voting, writing the following:

In past elections, the number of candidates an Election Committee member could vote for was limited by the number of votes he or she had - a maximum of one. But under the new rules, with far more votes at his or her disposal, a nominating committee member would have far more choices, and much greater freedom to nominate a convincing candidate, irrespective of the candidate's political background. 

This method of voting, which political scientists classify as "approval voting", is adopted by the UN General Assembly in electing its secretary general, after nomination by the Security Council. As the UN experience shows, "approval voting" has the advantage of fostering the selection of a consensus candidate, a person who attracts the least objection and is most capable of accommodating diverse interests and factions.

3. Lower the 50% threshold

Here is the "lever" I spoke of above. In negotiations over any complex piece of legislation, it is dramatically easier for policy to evolve if key negotiating points are numeric. For instance, a parking fee of $10 can potentially be negotiated down to $5. And from there it can potentially be negotiated down to zero dollars. Or "free on Sundays".

If the pan-democrats could manage to pass points 1 and 2, then they could come back in the future and push for that 50% threshold to be lowered, perhaps to 45%. And then a few years later down to 40%, and so on. Giving a percentage point here or there is not a particularly worrisome proposition for mainland China, especially given that the nominating committee is their puppet. But it leaves open a path to gradual improvement. And this plan leaves nearly all of Beijing's package intact, creating the superficial appearance that Hong Kong as conceded, which has obvious political implications given their need to appear in control of Hong Kong.

Lastly, however flawed or insufficient this proposal may be, it seems unlikely that there is any alternative that stands much hope of passage. Give Beijing their nomination process with a few nonthreatening adjustments though, and you may have something.

Six reasons to support IRV = more IRV falsehoods

The following is a response I wrote in response to this op-ed by Maine representative Dick Woodbury.

In his recent op-ed, Dick Woodbury says discusses six reasons it’s time for instant runoff voting in Maine. Unfortunately almost every claim is false.

Consider the 2009 IRV mayoral race in Burlington, VT. The Progressive candidate won, despite a majority of voters favoring the Democrat to the Progressive. But remove the Republican spoiler from consideration and re-tally the ballots, and the Democrat wins. So much for Mr. Woodbury's claim that "there is no such thing as a spoiler candidate" with IRV.

Republican voters who preferred the Democrat to the Progressive were punished for supporting their favorite candidate. Doing so caused the Progressive to win. Had even a small number of them strategically ranked the Democrat in first place, then the Democrat would have won, giving them their second choice instead of their third choice. So much for Mr. Woodbury's claim that "voters can cast their vote for a preferred candidate without the strategic dilemma of potentially helping a candidate they oppose."

Finally, we're told that IRV is "exactly like an actual run-off election." Tell that to San Francisco supervisor Malia Cohen, who was elected in the 20th round in her 2010 IRV race. She had the third highest first-place votes, meaning that she would not have even made it to the second and final round of a traditional delayed runoff.

The ranked voting proposal does have some merit. It likely would have elected independent Eliot Cutler in the 2014 gubernatorial race, who was preferred by sizable majorities against his two major party rivals. However, Mr. Woodbury unfortunately undermines his case by demonstrating serious misunderstandings of the the very reform he's advocating.

Clay Shentrup
Berkeley, CA
Co-founder, The Center for Election Science

The Netherlands should pursue election reform for the USA

I was just listening to a book on Audible, called The Agile City. The author mentions that almost 65% of the Netherlands' GDP is at risk from climate change. I.e. it is imperative for the country to decrease CO₂ emissions.

The United States emits over 16% of the world's CO2, and has the highest emissions per capita of any country (aside from a handful of countries with minuscule populations). This is second only to China (24.65% of total output, although China produces a mere one-third of the USA's per capita output). The per capita output for the European Union is only a bit higher than that of China, and the EU's entire output is still far below the USA's at 11.04%.

Thus it would be to their great benefit if the Netherlands' could somehow cause the USA to adopt policies which would reduce those emissions. I mean, this should arguably be their top priority, since we're talking about a huge fraction of their GDP being at risk from climate change. But how could a small European nation convince the world's "greatest superpower" to change such policies?

Easy. Fund election reform efforts by American political activists.

I'll give you an example. This past November 2014, a climate-change-denying Republican governor won re-election in the state of Maine. He was deeply unpopular, but his opposition was split between two climate realists, allowing him to win anyway. This vote splitting was possible because of the USA's use of the horrendous plurality voting system, where voters are limited to voting for only a single candidate.

But a multi-site exit poll indicates that the finish order would have been completely reversed had Maine used approval voting. Independent candidate Eliot Cutler (who was massively preferred head-to-head to both major party rivals) would have won instead. Cutler seems objectively the better representative, based on data from numerous opinion polls.

Approval voting simply means that voters can select as many candidates as they wish. This addresses the vote splitting issue, and makes it always safe to vote for your sincere favorite candidate. But that same exit poll revealed that even the overly complex and error-prone instant runoff voting system (IRV) would have gotten the winner right. Plurality voting is the only system I'm aware of that would have produced such a terrible result. But this is the system used in nearly all US elections.

Better voting systems can also reduce the impact of money in politics, which in turn reduces the influence of powerful interests who would very much like to thwart any policies which might reduce greenhouse gas emissions.

There are groups working to reform US democracy right now. But they suffer from widespread apathy, and general ignorance on the relevance of electoral systems on political outcomes. For a mere few million dollars, low lying countries such as the Netherlands could help American election activists establish a beach head and remove a massive obstacle to climate change policy reform. This is a bang-for-the-buck bargain that is simply unrivaled by anything they could achieve via e.g. UN bureaucracy. Countries such as the Netherlands ignore this opportunity at their peril.

How not to be wrong on voting methods

On page 419, Jordan Ellenberg in his 2014 book  How Not To Be Wrong, the power of mathematical thinking,  writes:

One voting system to which Arrow's Theorem doesn't apply is "approval voting," in which you don't have to declare all your preferences, you just vote for as many of the people on the ballot as you want, and the candidate who gets the most votes wins.  Most mathematicians I know consider approval voting or its variant to be superior to both plurality voting and IRV;  it has been used to elect popes, secretaries-generals of the United Nation, and the officials of the American Mathematical Society, but never yet government officials in the United States.

Mock 2014 Maine Gubernatorial election, using Approval Voting

A recent article informs us that, "Maine voters are looking at the three-way 2014 governor race strategically." This is because there are three major candidates, and independent Eliot Cutler is considered by many to be a potential "spoiler".

But there's a simple solution to this problem. Rather than urge Cutler to drop out of the race, Maine voters could push for a switch to Approval Voting. This simple alternative system is identical to the present system, except that it allows voters to select as many candidates as they wish. The candidate with the most votes still wins. You never have to fear a vote for your favorite candidate, because you can still support a more electable "lesser evil" compromise candidate if you wish.

Try it out for yourself, by voting below in a mock version of the election, using Approval Voting.

Vote for one OR MORE - the candidate with the most votes wins
	Eliot Cutler
	Paul LePage
	Mike Michaud
	Lee Schulteis
pollcode.com free polls

Score Voting vs. Approval Voting

A common criticism of Score Voting is that it is strategically identical to Approval Voting. For instance, one member of the German Pirate Party recently responded on Twitter to a Score Voting election using a -3 to +3 scale:

Best strategy for assessment choice by the way: Favorites to +3, all other -3. Otherwise, voting power is wasted.

The basic idea is that it's always strategically optimal to use the maximum and minimum scores (we often refer to this as "polarization"), which seems like an argument for using Approval Voting instead of Score Voting. There are a number of reasons why this argument is flawed.

First, it's not always true. There are some instances in which honesty is a better strategy than polarization. Having said that, I acknowledge that those are exceptions to the rule, which I point out purely to be thorough. So let's continue with the premise that polarization really is the best tactic, since it's approximately true. In that case, we must look at the pros and cons of using Score Voting instead of Approval Voting.

One major advantage of Score Voting is that it's more expressive, which many voters enjoy. For example, in the 2000 USA presidential election, polls show that about 10% of the people who preferred the Green Party voted for their nominee, Ralph Nader. They knew Nader had virtually no chance to win, and they could have made the tactical decision to vote for their favorite of the two major party candidates. But expressing their opinion was just so important to them, that they chose to "throw their votes away" on the Green Party.

This principle is actually at work within the very act of voting! Because in the vast majority of elections, a voter's odds of affecting the outcome are so incredibly small that it is economically irrational for him to waste his time voting. This is concept is actually called the Paradox of Voting. So if our Pirate friend from Twitter votes, that is proof that he actually agrees with us.

But there is more benefit here than mere self expression. It turns out that when voters are honest, they tend to "donate" more utility (aka welfare/satisfaction/happiness) to the other voters than they sacrifice, resulting in a net welfare increase. You can see this clearly in Bayesian regret figures. The more voters are honest, the greater the average satisfaction.

Critics will often respond that this only addresses those voters who consciously chose to be honest. What about those voters who simply didn't know any better, and were then unfortunately victimized by the tactical voters? There are a number of counter-arguments to this concern:

• Those voters presumably are very few in number, meaning we should be skeptical about giving up the significant benefits of Score Voting simply to protect them.
• One can make the argument that any voter so naive as to be unaware of strategy would actually be better off casting a sincere Score Voting ballot than an Approval Voting ballot. This is based on analysis of the effectiveness of these respective behaviors. I.e. observe that the values in column C tend to be greater than those in column D in the first table here.
• It turns out that if there's enough sincere voting, then even the "naively honest fools" who vote sincerely do better with Score Voting than with Approval Voting. See table two from this page on the "Shentrup-Smith Experiment".

Summary

The bottom line is that if even one voter decides to sincerely use the intermediate scores, then Score Voting produces better average voter satisfaction that Approval Voting. And if you prevent sincere voters from using intermediate scores in order to protect the naive ones, you possibly hurt them, and you definitely hurt the voters who wanted to be expressive. Is it really worth it, just to make the voting procedure a little simpler? Score Voting is a little more complex than Approval Voting, but it is still much simpler than any ranked voting system. 

Approval Voting is a fantastic system, but a rational voter will want to take Score Voting instead, if he can get it.

Intepersonal comparisons of utility

I occasionally run into someone expressing skepticism about the validity of Warren Smith's Bayesian regret calculations, due to the fact that they employ a Utilitarian or "Benthamite" social welfare function (i.e. the social welfare for is the sum of the utilities of the members of the group). The particular criticism I refer to is that individual utilities cannot be summed because they are not "interpersonally comparable".

Let's assume, for the sake of argument, that utilities are not interpersonally comparable. In that case, the uniquely best social welfare function is the one which maximizes the expected utility of a randomly chosen voter, in his own personal units of utility. That is, the best option is the one that maximizes the average of all voters' utilities. But note that this is mathematically identical to simply summing the utilities, for any given group of voters! And therefore the complaint about not being able to perform interpersonal comparisons of utility is academic, and doesn't invalidate the social choice function Smith employed for his BR calculations.

Regarding the aforementioned Twitter conversation, this also some bearing on the applicability of "labels" to the scores listed on a Score Voting ballot. That is, some folks would like to have something like:

0 - terrible, 1 - poor, 2 - mediocre, 3 - good, 4 - great

The most obvious problem, to anyone acquainted with basic economic theory, is that this introduces error. For instance, suppose that my subjective connotation with these words is such that

u(great) - u(good) = u(good) - u(terrible)

And suppose that my utilities for some set of options are such that

u(x) - u(y) = u(y) - u(z)

Without labels, I would vote (assuming I'm a sincere voter)

x = 4, y = 2, z = 0

But if I vote using the labels, that would be

x = 4, y = 3, z = 0

This would produce a lower expected utility for me, making me worse off. Of course, if I'm a tactical voter, then it doesn't matter anyway. But why add more text than is absolutely necessary to a ballot, if it reduces the expected utility of voters? Why make the voters worse off?

There is a potential counter to this. But it appears to require that you accept the validity of interpersonal utility comparisons. Here's how it works. Suppose that Bob and Alice have the following utilities, on a universal scale.

u(x) = 8, u(y) = 6, u(z) = 0
u(x) = 8, u(y) = 6, u(z) = 4

Their sincere votes (on a 0-4 scale) would look like

x = 4, y = 3, z = 0
x = 4, y = 2, z = 0

The scaling distortion makes it appear as though they have different opinions on y, but the same opinions on z—which is precisely backwards. But suppose those labels, like "great" and "mediocre", have highly similar mappings to actual utility values for both Bob and Alice. In that case, the labels could prevent more loss than they caused—by nullifying some component of the normalization loss.

The irony here is that our friend on Twitter supported the use of labels, but opposed the validity of interpersonal utility comparisons!

The caveat is, as I said at the beginning, that a welfare function that maximizes the sum of all individuals' utilities also maximizes the expected utility of any individual voter, even if we disallow interpersonal comparisons of utility. So my aforementioned defense of labels is still equally valid, even if we don't permit interpersonal utility comparisons.

@selylidne can thank me for offering at least some evidence in support of his own argument.

CU Boulder upgrades to Approval Voting; could go even further

The recent student government elections at CU Boulder used an improved voting system called Approval Voting. Approval Voting means you can vote for as many candidates as you want, rather than being limited to the number of winners. This is a big improvement over the traditional system called Plurality Voting. Approval Voting ensures you can never be punished by voting for your favorite candidate, and solves the spoiler effect.

Approval Voting is a great way to elect in a single-winner election, e.g. for mayor, governor, senator, etc. But for multi-winner races (e.g. the CUSG Representative-at-large election, with five winners) Plurality Voting and Approval Voting can both produce lop-sided victories.

Consider, for example, the election of the Arts and Science Student Government board members. The Unite party took all five seats, despite earning only 54% of the votes. The Inspire party got almost 45% of the votes, but won zero seats. An arguably more fair outcome would have been three seats for Unite, and two seats for Inspire.

The tables were turned in the University of Colorado Engineering Council elections. Eight of the nine Representative-at-large seats went to Inspire, and only one went to Unite. But Inspire won 67% of the vote. Unite won 33% of the vote, but earned only 11% of the seats. A more fair result would have been three seats for Unite, and six seats for Inspire.

Inspire also swept the UCEC senator elections. There were four candidates (two Unite, two Inspire) running for two seats. Unite won 46% of the votes, but zero seats. Again, this was not a representative outcome. Looking at the results, it seems to me that the winners should have been Inspire's Benjamin Zatz, along with Unite's Rachel Sobke.

Arden Rose, UNITE 348 21%
Rachel Sobke, UNITE 407 25%
Jon Fearer, INSPIRE 430 26% Elected
Benjamin Zatz, INSPIRE 459 28% Elected

So what's the solution? Proportional Approval Voting!

It turns out that a Danish statistician figured out the solution to this problem back in 1890! Imagine if we elect winners one round at a time. Then after each new winner is selected, we "divide" each ballot by w, where w is "the number of winners that ballot has approved, plus one".

As an example, imagine you approved Benjamin Zatz and Jon Fearer:

• In the first round, your ballot is just divided by 1, which has no effect. Your votes count at full strength.
• Let's say Benjamin Zatz is the first winner, which he would be based on the actual election results.
• In the second round, your ballot is divided by 2 (because you voted for Zatz, who won in the previous round). Now your vote for Fearer counts as only 1/2 a vote. This means that all the people who didn't vote for Zatz (i.e. Unite voters) now have the power to elect one of their candidates.
• This process continues as seats are awarded, providing fair representation to any faction which has enough support to deserve a seat.

Proportional Approval Voting would have produced the kind of fair outcomes we just considered above. The total seats awarded apparently wouldn't have been much different. But the allocation would have allowed Inspire to have a voice in ASSG, and Unite to have a voice in UCEC.

But isn't this complicated?

Not really. Consider that Australia's senate has been elected via the much more complicated Single Transferable Vote system since 1938, far before modern computers or even electric calculators. It was all tallied by hand from paper ballots. Thus Australia has a truly representative democracy. Doesn't CUSG deserve that too?

Here I demonstrate how Proportional Approval Voting can even be tallied using a simple Google Docs spreadsheet.

 

In summation, Approval Voting is a big step forward for CUSG elections. But it can be even better in the multi-winner races. Proportional Approval Voting is worth considering.

Maine's Rep. Diane Russell's mistaken views on Instant Runoff Voting

As a founder of The Center for Election Science, I spend a lot of time doing education on electoral issues. Unfortunately, a significant amount of that time is spent fact-checking well intentioned but misinformed proponents of Instant Runoff Voting. The subject of this post will be several Twitter posts by Maine representative Diane Russell.

Some background. Today I came across a Twitter post by FairVote, urging people to support Instant Runoff Voting (IRV) for Maine gubernatorial elections, by signing this online petition. The the idea is not without merit. Maine's current governor, Paul LePage, is one of the least popular governors in the USA. A news article grimly states:

LePage’s job approval rating is significantly underwater at -16, with 39% approving and 55% disapproving. 18% of Republicans and 54% of independents disapprove of the job that LePage is doing as governor.

This isn't surprising given that Maine voters never wanted LePage in the first place. In the 2010 gubernatorial race, LePage walked away with a mere 38.33% of the vote. His win was a result of vote splitting (aka the "spoiler" effect) between his rivals. Eliot Cutler—a wealthy independent who served as Associate Director for Natural Resources, Energy and Science in the Office of Management and Budget in the Carter administration, and was the principal White House official for energy—narrowly trailed with 36.49%. Democrat Libby Mitchell came away with 19.12%. It's safe to say that if Mitchell had dropped out, her supporters would have heavily favored Cutler. And probably vice versa.

IRV (which also goes by the misnomer "Ranked Choice Voting", despite the fact that there are numerous other ranked voting methods besides IRV) would have avoided this mess. Democrat Libby Mitchell would have been eliminated, and her 19.12% would have almost surely sent Cutler to victory—much like a traditional runoff, but without the second round election.

Now it appears that Maine may be in for a repeat in the 2014 gubernatorial race. From the previously cited article:

PPP’s latest poll finds LePage ahead by 4 to 7 points in every potential three-way contest while losing every head-to-head contest by 8 to 21 points. If LePage and independent Eliot Cutler were the only two candidates, Cutler would lead LePage 49% to 41%. But in every three-way scenario, Cutler’s strength as an independent candidate gives LePage the lead over their possible Democratic opponents.

Given these facts, Diane's Russell's support for IRV seems outwardly sensible. But a closer look at the details reveals some concerning misconceptions. The petition's description says:

Maine politicians should be elected with a majority vote. Please join me in supporting Rep. Diane Russell’s bill to elect the Governor and Legislature with Ranked Choice Voting (AKA Instant Runoff Voting). Her bill encourages voter choice while avoiding “spoiler” candidates.

This statement claims that IRV prevents the spoiler effect, and that it ensures politicians are elected with a majority. As anyone who's spend an afternoon studying election methods on Wikipedia can tell you, both of these claims are completely false. Here's a simple four-candidate electoral scenario which demonstrates this.

% of voters - their ranking
 35% W > Y > Z > X
 17% X > Y > Z > W
 32% Y > Z > X > W
 16% Z > X > Y > W

Instant Runoff Voting selects candidate X as the winner, beating W in the final round, 65% to 35%. But note that:

1. A huge 67% majority of voters prefer Y to X (the 1st and 3rd rows). And Y received nearly twice as many first-place votes as X, 32% vs. 17%.
2. An even larger 83% super-majority of voters prefer Z to X (and Z gets just slightly fewer first-place votes than X).
3. Y is preferred to Z by an 84% super-majority.
4. Y is preferred to W by a 65% super-majority.

Y is the indisputable "majority winner", but IRV elects X. So the very first sentence of the petition description was misleading, to say the least.

Further, W is a spoiler. If W would drop out of the race, then Y would win instead, even with no change in voter preferences. So the final sentence in the petition description was also false.

This situation is similar to what occurred in the 2009 IRV mayoral race in Burlington, VT. The Progressive won, but a large majority of the voters preferred the Democrat to the Progressive, and to all other challengers. And the Republican served as a spoiler—if he had dropped out, with all voters preferences unchanged, then the Democrat would have won. (Note that Burlington voters repealed IRV by a significant majority after that, their second, IRV election.)

In the ensuing discussion on Twitter, Russell responded to a comment from Maine blogger Gerald Weinard:

Ranked Voting worked well in Portland and has a lot of support. Majority leaders super important.

Here we see Russell once again making the false implication that IRV ensures majority winners. It's concerning enough for an elected official not to understand a substantial reform that she's vocally pursuing. But more so because Russell has worked with the IRV advocacy organization FairVote, which has been involved in electoral reform since 1993, and presumably should understand IRV by now.

This rhetoric continued over the course of several more posts. In response to a criticism about IRV's being less than intuitive, Russell replied:

I've been ranking ice cream choices since I was a kid. Also pretty intuitive.

This comment demonstrates a lack of appreciation for the deeper issue of whether voters understand the tabulation process. A usability study in the San Francisco area (SF has used IRV since 2004) produced a couple of relevant conclusions:

• Some participants remarked that because they didn’t understand how the votes were counted that they didn’t trust the system.
• Many participants theorized that ranked choice operates on a weighted or point system. A few participants suggested that it was for breaking ties, but they could not describe how the second place votes were tallied. It was not uncommon for participants to talk themselves into corners as they tried to describe how the counting was done. Most ended with “I don’t know.”
• Very few people accurately described how ranked choice votes are counted, even in zip codes with high education and socio-economic levels; people in poorer neighborhoods had even more difficulty describing how their ranked choices were counted.
• It’s unclear that reading the instructions for ranked choice voting would help people vote as they intend, because the instructions are only about how to mark the ballot, not about how the votes are counted or what the consequences of ranking are.

I work as a software engineer in San Francisco. And my informal polling of numerous co-workers has been consistent with these observations. Here's a brief instant message conversation I had with one particularly smart software engineer who claimed to be a voter. I don't think anyone can read that and seriously claim that IRV is intuitive.

I replied to Russell, linking her to the above demonstration that IRV does not guarantee a majority winner, and is susceptible to the spoiler problem:

I linked you to mathematical proof that IRV doesn't guarantee majority winners. Response?

Her reply was simply:

Answer: http://www.pressherald.com/opinion/brennan-ranked-choice-voting-both-winners_2011-11-12.html

That's a link to an article in the Portland Press Herald, which touts the virtues of IRV and repeats the same inaccurate talking points:

Brennan, the candidate who proved to have the broadest appeal, ended up the winner. He will go into office next month with a majority of voters at his back.

I pointed out to Russell that her article was actually not an answer to my criticism:

That article doesn't refute the mathematical fact that IRV *does not* guarantee "majority winners".

Her well reasoned rebuttal?

Again, you have no vested interest in Maine. Why do you care? Oh wait... this is all you do. FOR YEARS.

It's hard to know what to say to that kind of response. It's an extreme example of both ad hominem and red herring fallacies, diverting attention from her mistaken claims by personally attacking me. Nevertheless, I feel compelled to address this.

The Center for Election Science is a non-partisan 501(c)3 non-profit, devoted to the objective scientific study of electoral systems, and the advocacy of systems that we believe will improve human welfare. We are something like the FactCheck.org of electoral reform. To that end, we strive to make the public better informed about election methods. That includes the occasional response to people who make the sort of false and misleading claims discussed herein.

Rather than attacking myself or my colleagues at the CES, I would ask Representative Russell to consider reading up on the subject of voting methods, and trying to be more careful with her statements. Election science will be critical to solving issues like the Maine gubernatorial elections, so it's important that we get the facts right.

Clay Shentrup

Co-founder, The Center for Election Science

The economic consequences of an economically irrational decision-making system

Amidst the hype surrounding the "fiscal cliff", some calm voices have emerged to challenge the fear mongering. One of them is Stephanie Kelton, an assistant professor of macroeconomics, finance, and money and banking at UMKC. Kelton has pointed out that the cliff is in large part a manufactured crisis. She suggests we start to address it by (among other things) repealing the Budget Control Act of 2011, whose key provisions established the cliff in the first place. She further suggests that the government should spend, and generally avoid tax increases, in order to increase the GDP (the oft-ignored denominator in debt-to-GDP ratio)—basically the opposite of what a lot of the mainstream pundits are advising.

I now want to take a step back and look at this issue from a very different perspective. The gravest danger in any calamity like this is not that we fail to fix it, but that we fail to detect underlying causes (i.e. calamity generators). Join me for a bit of abstract thinking...

Let's assume that Kelton is correct, and society is generally convinced that we need to implement the policies she's advocating. How do we do that? Well, primarily by electing people to public office who we believe will implement those policies. But imagine we're faced with an election for some highly influential position, and the preferences look like this:

33% Keltonite 1 > Keltonite 2 > Boehnerite
32% Keltonite 2 > Keltonite 1 > Boehnerite
35% Boehnerite > others

In this election, a massive 65% majority of the voters wants Keltonite policies. But if we use the ordinary vote-for-one "Plurality Voting" system, the Boeherite candidate wins with 35% of the votes. The lives of hundreds of millions of people are then significantly worse because of an obvious defect of the voting system.

But this is preventable! There are numerous other ways to vote. And voting systems can be objectively measured for "economic efficiency" using a metric called "Bayesian regret", which effectively measures average voter satisfaction. Here are some sample Bayesian regret figures from William Poundstone's book Gaming the Vote (2008).

Observe that a change from Plurality Voting to Range Voting (aka Score Voting) or Approval Voting virtually doubles the improvement gotten by the invention of democracy in the first place (the performance of Plurality Voting minus the performance of random non-democratic selection).

And these systems are very simple. Approval Voting just uses an ordinary ballot, but lets voters vote for as many candidates as they wish. The candidate with the most votes still wins. Score Voting is like Yelp reviews—you rate the candidates and the one with the most points wins. Not as simple, but more expressive.

So the question is, why aren't these economists who want to save the world talking about voting reform? If they're interested, I recommend they check out The Center for Election Science.

Candidate strategies with Instant Runoff Voting

In a recent email to her supporters, San Francisco District 5 supervisor candidate Hope Johnson outlines a strategy for the upcoming Instant Runoff Voting election.

Johnson and Resignato are launching The People's Ticket to give voters a better choice than the big-money downtown and progressive machine candidates who are constantly at odds with each other.  They are asking D5 constituents to vote for them in either the first and second or first and third positions, in any order, on November 6’s ranked choice ballot.

A friend asks, "Does this make sense, from a strategic voting perspective?" The answer is, it's complicated. Say you have these hypothetical preferences:

35% X > Johnson > Resignato
32% Johnson > X > Resignato
33% Resignato > X > Johnson

Johnson has only 32% of the first-place rankings, so she's the first eliminated, which causes 32% of the votes to be transferred to X in the next round. X then trounces Resignato 67% to 33%.

But say Johnson and Resignato successfully convince their supporters to adopt the strategy. Then it would look like this:

35% X > Johnson > Resignato
32% Johnson > Resignato > X
33% Resignato > Johnson > X

Johnson is still eliminated first (indeed, this strategy cannot possibly affect the election prior to the elimination of the first of the two colluders). But now Resignato trounces X 65% to 35%. So the strategy worked.

But things aren't so simple! The strategy helped one of the candidates, but it also caused 32% of voters to get their 3rd choice instead of their 2nd. So from their point of view, it was the opposite of strategic. While the Johnson supporters certainly want the Resignato supporters to help Johnson, they don't want to help Resignato, particularly within the privacy of the voting booth. And vice versa. This is the classic prisoners' dilemma.

History has disproved the notion that voters care more about loyalty to their favorite candidates than about getting what's best for themselves. Consider that exit polls in the 2000 U.S. presidential election showed that 90% of Nader-favoring voters claimed to have voted for someone other than Nader. Obviously a Nader supporter who preferred Gore to Bush was maximizing his expected value by voting for Gore instead of Nader. Likewise, we expect that Johnson and Resignato would have a hard time convincing people to vote against their own best interests.

There's also some interesting game theory related to which partner you should pick, if engaging in this strategy. Notice that if X were to team up with Johnson in the last example, that would lift X one ranking up (above Resignato) on 32% of the ballots. The same deal with Resignato would create a one-rank lift on 33% of ballots, but above Johnson. Which strategy makes sense? In this case, it doesn't matter, because X loses head-to-head against any rival. But in general, the strategy is to team up with someone whom you think will be ranked higher than your greatest rival, by a lot of voters. But you also want to pick someone whom you believe will be eliminated before you, otherwise you don't benefit from the strategy.

I'll leave the more complicated analysis to my math Ph.D. associates at The Center for Election Science.

What if we used Score Voting in San Francisco?

San Francisco currently uses Instant Runoff Voting, restricted to ranking up to three candidates. But imagine we used Score Voting (aka Range Voting), where you could rate as many candidates as you wanted to. Here's what that would look like if we used a simplified 0 to 2 point scale.

Bad candidates get a 0. Good candidates get a 2. Average candidates get a 1. Simple. Expressive (even more expressive than IRV, because you can vote for as many candidates as you want, not just three). And way less ballot space than the current Instant Runoff Voting ballot.

Other benefits:

Lower ballot spoilage
IRV results in about seven times as many spoiled ballots as Plurality Voting, on average. This is especially common among voters of lower socioeconomic status. That's one reason that many of IRV's detractors argue it's too complicated. But Score Voting experimentally results in fewer spoiled ballots. This is because there are actually fewer ways to spoil your ballot with Score Voting, since it's fine to give the same rating to multiple candidates (whereas you cannot give two candidates the same ranking with IRV, nor can you vote for more than one candidate with Plurality Voting). This is explained in much greater detail by Warren Smith (a Princeton math Ph.D. who studies voting systems) here.

People intuitively understand Score Voting
The most basic sanity check for a voting system is to have random individuals fill out a ballot, and then ask them how they intuitively think the result is tabulated. If their intuition tends to be right, then you know you have a voting system that will be easily understood, even without an expensive voter education campaign.

Instant Runoff Voting fails this sanity check, because it is not very intuitive. In my experience, polling many intelligent residents who claim to vote, the vast majority of them do not understand how IRV works. They tend to assume it's a "weighting" system, where candidates get more points for being ranked closer to first. Here's a conversation I had with a very intelligent software engineer I used to work with, who has voted in IRV elections in San Francisco.

San Francisco has tried to remedy this problem with a massive public education campaign. See this example, and another, and another. The problem is that this education not only costs money, but it isn't even very effective. Just poll some random friends or co-workers who vote in San Francisco elections. Show them the same example I showed my former co-worker, and see whether they get it right. Or even just ask them to correctly explain the tabulation procedure. You will likely find that many of them get it wrong.

Some people have suggested that it doesn't matter how well people understand the tabulation system, as long as they understand how to correctly fill out a ballot. But the problem is, if people assume IRV is a  weighting system, then they'll be more likely to strategically exaggerate. For instance, say a voter prefers Joanna Rees over John Avalos over Ed Lee over Tony Hall (ignoring the other candidates for simplicity). That voter may choose to strategically push Avalos and Lee apart, ranking Avalos in first place and Lee in last place, because he or she assumes that will help Avalos (the "lesser evil" frontrunner) defeat Lee (the "greater evil" frontrunner). It actually turns out that this tactic is partially correct, but that's not even very relevant. The relevant issue is that many voters will just do it "naively", based on their intuition of how the system works.

Score Voting simply doesn't have this problem. Virtually everyone is familiar with the concept of ratings. We've seen average ratings for movies, restaurants, and even products on Amazon. If you hand someone a Score Voting ballot like the example pictured above, and ask them to explain how they assume the tabulation works, they will just intuitively get it right most of the time. That is, people will better understand Score Voting the first time they use it than they will understand IRV after having already used it for years. I'm especially confident of this after having conducted polls, like this one in Beaumont, Texas, for the 2006 Texas gubernatorial election.

Voting machine simplicity
Unlike IRV, Score Voting can be counted on ordinary "dumb totaling" Plurality Voting machines, without any need for complicated (and potentially fraud-prone) electronic voting machines. We think the Department of Elections will really appreciate this.

Precinct subtotals
Score Voting can be subtotaled in precincts, unlike IRV. So you can just add up the subtotals at each precinct to get the final result. You don't have to transport the ballots to City Hall like you do with IRV. That's not only cheaper and easier, but it's better for election integrity. Moving ballots raises chain-of-custody concerns.

Summation in a single "round"
Score Voting doesn't have multiple "rounds of elimination" like IRV does. So the results are very clear. You just see a single sum of points for each candidate. You don't have this multiple rounds/columns business like you do with IRV.

Score Voting has none of the bizarre paradoxes of IRV
IRV has a crazy number of bizarre paradoxes. You can get a worse result by voting than not voting, or get a better result by staying at home instead of voting. A candidate can change from winner to loser by gaining support, or go from loser to winner by losing support. In some cases it's possible to reverse all the ballots (as if trying to elect the worst candidate) and still get the same result — meaning IRV thinks the best and worst candidates are the same person! IRV can also punish you (give you a worse result) for voting for your favorite candidate. Score Voting has none of these problems.

Better average voter satisfaction
This is perhaps the most noteworthy, albeit also the most esoteric, benefit of Score Voting. The following graph, from page 239 of William Poundstone's book Gaming the Vote, shows a graph of Bayesian Regret values for a host of different voting methods. Score Voting (called "Range Voting" in the graph) far exceeds Instant Runoff Voting.

The methodology behind Bayesian Regret is a bit complicated, but here it is if you care to understand it better. Bayesian Regret is sort of the opposite of "average voter satisfaction", so less is better (like a golf score). It's framed this way because a zero Bayesian Regret would be "perfection", and that's a convenient benchmark.

The further each bar is to the left, the better that system performs. The width of each bar represents its change in performance in response to a change in the ratio of strategic-to-honest voters. For example, Borda performs better than Condorcet or Approval Voting, if all voters are honest. But it behaves much worse than Approval Voting if there is a lot of strategic behavior.

While Bayesian Regret is indeed a very esoteric and mathematical concept, I believe it is the best (really the only) measure of ultimate voting method performance, or "democratic-ness". It precisely measures how well a voting system satisfies the will of the voters.

Here's an even more comprehensive list of benefits from Warren D. Smith, the Princeton math Ph.D. behind The Center for Range Voting.

What about getting a majority winner?

Score Voting occasionally faces criticism due to the (extremely unlikely) possibility of failing to elect a candidate who is the favorite of a majority of voters. We address that criticism in great detail here.

In a nutshell, IRV can do even more bizarre things. In the 2009 IRV mayoral race in Burlington, Vermont, the Progressive won, even though the Democrat was preferred to the Progressive by a 54% to 46% majority. And in the 2010 San Francisco District 10 supervisor race, Malia Cohen won in the 20th round with 4321 votes from 18308 voters who voted in that race. That's a 23.6% "majority".

IRV can even elect X in cases where Y was preferred to X by a majority and got more first-place votes than X. And to consider worst case scenarios for a moment, here's a hypothetical IRV election with over a million voters, in which the winner is the favorite of only two voters, and would lose in a head-to-head matchup against 19 of the other 20 candidates.

This is not meant as an indictment of IRV, but the moral here is to judge voting systems based on their average/typical behavior, rather than on the most bizarre and improbable worst case scenarios you can dream up.

What about having a runoff?

Some people like having a runoff in cases where no candidate got a majority in the first round. They see a runoff as a way to both ensure a majority winner, and allow voters to become better informed about their choices. Detractors of runoffs argue that they allow special interests to dictate election outcomes, and tend to attract older/whiter/wealthier voters, and decrease turnout in general.

As for the part about ensuring a majority winner even one no candidate was the favorite of a majority of voters, that actually turns out to be something of a mathematical impossibility.

With regard to the other pros and cons mentioned above, that's a very complex and controversial topic which we won't delve into here. The one case where there is pretty broad agreement from both the pro-IRV and anti-IRV advocates is that runoffs make sense in the mayoral race.

So we propose a simple way of adding a runoff requirement to our Score Voting proposal, for the mayoral race only. If no candidate receives at least as many points as there are voters, then a runoff is held between the top two finishers. So if 30,000 voters vote in the mayoral race, then a candidate must receive at least 30,000 points to win outright.

Why do we use as many points as there are voters? The derivation is pretty straightforward. We're just requiring 50% of the maximum number of points, which is the closest possible analog to "getting a majority". But it just conveniently/coincidentally happens that with a 0-2 scale, that number is the same as the number of voters. To demonstrate: number of voters × 2 (points per voter) ÷ 2 = number of voters.

A recent puff piece at BeyondChron.org casts "dissenters" to the Instant Runoff Voting system as representing business interests. While I don't expect them to publish it, here was my response.

This article does not give an accurate account of what went on that day.

I am not a "member of the business community", but rather a proponent of election system reform. I co-founded the Center for Election Science, a not-for-profit organization which researches voting systems and related issues. Our president, a Princeton math Ph.D. named Warren Smith, created this solution to prevent Gerrymandering for example:
http://brokenliving.blogspot.com/2012/01/ending-gerrymandering-with-math.html

Along with my associate, Dominic Paris, I explained that the IRV proponents have made a number of highly misleading and even outright false arguments.

For instance, attributing the increase in the board's diversity to the use of IRV, even though 38 of the 40 races to which IRV applied would have had the same outcome under the rules of Plurality Voting. (The two exceptions were Malia Cohen in D10, and Mark Farrell in D2.)
See http://www.electology.org/irv-diversity-sf

Or arguing that IRV has increased voter turnout, which is not supported by a thorough statistical analysis of the historical data.
http://www.electology.org/irv-turnout-sf

Another claim was that IRV allows you to safely cast your sincere preferences in order. This is also false. In the 2009 IRV mayoral race is Burlington, Vermont, the Republican voters who preferred the Democrat to the Progressive got a worse result for being sincere. If they had insincerely ranked the Democrat in first place, then they would have gotten their second choice instead of their third choice, because that would have caused the Democrat to win instead of the Progressive. The Republican was a "spoiler" in that race, and they "threw away their vote" on the Republican. Moreover, it is the generally beneficial strategy with IRV to top-rank your favorite AMONG THE PERCEIVED FRONTRUNNERS regardless of who your overall sincere favorite candidate is.
http://www.electology.org/irv-plurality

During this hearing, I handed a brief note to Judy Cox, explaining this. Rather than engaging me in a polite and honest discussion of this matter, and showing some sign of a careful open-minded consideration of the facts, she simply handed the note off to Steve Hill. I have no idea why. All I know is that I have seen a near-uniform tendency of IRV proponents to totally refuse any kind of deep and scientifically rigorous dialog on this complex subject.

IRV may be a bit better than the traditional runoff system, when all factors are combined. But it remains essentially the worst of the commonly discussed alternative voting systems, as you can see in this graph of Bayesian Regret values from page 239 of William Pounstone's book _Gaming the Vote_.
http://ScoreVoting.net/BayRegsFig.html

Score Voting and Approval Voting, for instance, are simpler and superior in essentially every way.
http://www.electology.org/approval-voting-vs-irv
http://ScoreVoting.net/CFERlet.html

But rather than embrace them as viable alternatives (which, to be fair, would benefit from more political vetting), IRV proponents such as Steve Hill have shot them down with a laundry list of absurd criticisms which are totally out of line with empirical data and game theoretical analysis. E.g. the "bullet voting" criticism:
http://www.electology.org/bullet-voting
http://www.electology.org/pirate-elections-germany

I urge you to publish fact-based criticisms such as mine, in the name of maximizing understanding of this issue and leading voters to make the most informed decision possible.

Clay Shentrup
Noe Valley

the secret world is big

According to Peter Galison: "Removing knowledge",
Critical Inquiry 31 (Autumn 2004)
online here
www.fas.harvard.edu/~hsdept/bios/docs/Removing%20Knowledge.pdf
...

over 250 million secret pages are added to US govt archives annually (250 = +330 -80 where the 330 are classified and the 80 declassified) which is about 5X the amount of UNclassified knowledge added annually to conventional publicly accessible repositories of knowledge.

The US govt employs about 4M people with security clearances ("black" employees) versus only about 1.8M "white" employees.