Aaron Hamlin
12 min readAug 9, 2016


Looking Closer at 10 Popular Reforms: Commentary on Freakonomics

There are many electoral reforms we hear about all the time. But are they as effective as they seem? Recently, Freakonomics surveyed ten people in politics and asked them about their preferred reform.

Let’s take a look at what they had to say. (Due to length, only the main issues are addressed.)

1. Olympia Snowe: Top-two primary

“A closed primary [needs to end]. … Today there are very few of what you would call centrist, moderate candidates on either side of the political aisle.”

Snowe’s idea is a top-two primary. This is when you have all the parties’ candidates run in one big primary together, allowing voters to vote for any one candidate they want. The two candidates with the most votes go onto the general election. Snowe’s thinking is that this will lead towards more moderate winners.

It’s dubious that this approach elects more moderate winners. Freakonomics itself pointed out that the evidence for this claim is pretty weak. Here’s another article that also show doubts. We’ve pointed out before why this might be the case. When you have crowded primaries and force people to choose only one candidate, this has a tendency to squeeze out the middle — even when two people are elected. (See figure.)

Another criticism of top-two is that it tends to entirely remove third parties from the general election where their ideas would otherwise get more attention. Overall, it seems better to simply encourage more candidates in the general election and use a method like approval voting which favors moderates. Most countries besides the US tend to let parties figure out their own nominations anyway.

2. Howard Dean: Ranked choice voting/ Instant runoff voting (IRV)

“There’s a system called ranked-choice voting, where you don’t get just your vote for the top choice that you have, you also get to vote on all the other choices. … What that does is create as the winner, the person who is best respected and best liked overall in the electorate.”

Dean has the right idea in moving away from our choose-one, plurality voting method. It’s awful. And IRV is an improvement over plurality, albeit with some complexity. IRV suffers from some anomalies that can cause a more consensus candidate to lose, as was shown in the Burlington, VT election. The same strange outcome would have likely occurred in the 1991 Louisiana gubernatorial election as well, which used a runoff. We were able to see, however, that IRV could have saved moderate candidate Eliot Cutler (just barely) in Maine’s 2014 gubernatorial election. This used rough exit poll data, but this indicates that IRV may sometimes pull through in elections where it initially appears poor. IRV does a more consistent job at mitigating the spoiler effect when the third-party candidate has very little support.

Dean has the right idea of moving away from plurality voting. And while IRV is an improvement over plurality, it makes an odd choice given alternatives like approval voting which is much simpler and tends to perform better.

3. Rob Richie: Use three-member single transferrable vote (STV)

“That degree of opening up the system does some really interesting things. One, it makes the general election matter. And it very methodically and reliably represents the left, center, and right of every district.”

Richie’s overall criticism here makes a lot of sense. His gripe is against winner-take-all (a.k.a. single-member districts). Winner-take-all makes gerrymandering super easy and leads to highly disproportionate results. Also, if you didn’t vote for the winner in your district, then you effectively don’t have anyone to go to since they’re unlikely to share your viewpoint.

A multi-winner proportional system gets right at the problem. Voters are more likely to elect someone who represents them, results are more proportional, and gerrymandering becomes difficult to accomplish. (Note that a multi-winner bloc system — which most at-large elections use — doesn’t accomplish any of this.)

The drawback with this particular proposal is the three-member districts. The number of people you elect at-large is important because it directly influences the threshold required for election. Three is likely too low because it creates an election threshold as high as 25%. Five-member districts at minimum are generally recommended instead. Three-member districts will deal generally well with proportionality of major parties and help with gerrymandering. But independents and third-party supporters will likely continue to be ignored. For example, Illinois had three-member districts for a hundred years up until 1980, and they didn’t elect any third parties despite using a semi-proportional voting method (cumulative voting).

Looking at STV in particular, it’s not bad. It does offer proportionality. Though, it can behave strangely electing its first winner, because in that stage it’s essentially using IRV (STV is different than IRV overall). Also, when there are many candidates, it can get daunting for voters. Ranking 12+ candidates is tough. STV’s algorithm is challenging as well, though that can be an obstacle in many sensible proportional methods. As an alternative, there are ways to make approval voting a proportional method, allowing for simplicity and a more sensible person elected for the first seat.

One hangup with using a proportional method at the House level is that federal law currently forbids it. House seats are required to be single-member districts (2 U.S. Code § 2c). While that law would have to be changed, there’s nothing to prevent proportional representation in state legislatures.

4. Joaquin Castro: Remove straight-ticket voting

“Straight-ticket voting means that you can go into a ballot booth and without looking at any of the individual candidates or races on the ballot, at the top of the ballot, you can simply mark that you want to vote for all the Democrats or all the Republicans.”

The idea behind Castro’s recommendation is that it would encourage voters to evaluate each candidate independently. It’s hard to know whether people would vote the same regardless, and this data doesn’t appear immediately available.

If someone is to go and look at candidates independent of their party, then they would have already had to take steps beforehand to research candidates before arriving at the voting booth. That would make the straight-ticket voting component largely irrelevant. Likely one of the largest obstacles to getting people to evaluate all the candidates is the fear of a wasted vote. Otherwise, there’s little incentive for most people to look beyond the major-party candidates. If we want to alleviate that wasted vote fear, then we must use a voting method that allows voters to be honest about the candidates they like the most — a method like approval voting. A straight-ticket voting ban won’t solve that problem.

Castro may have just meant that we should encourage voters to look between Democrats and Republicans within each race. But the idea that we should limit voters to focus only on major-party choices is a much greater problem in itself. An election where we only entertain two choices is not a competitive election.

5. Norman Ornstein: Implement mandatory voting

“In my various trips to Australia and my discussions with politicians of all political stripes there, they will tell you that when you know that your base is going to be turning out to the polls, and when the other side’s base you know is going to be turning out to the polls, your focus turns to the persuadable voters in the middle.”

Voter turnout is more of a measure of how strong a democracy is perceived rather than how strong a democracy actually is. According to Pippa Norris in Electoral Engineering, compulsory voting does have an effect on voter turnout, particularly when there’s a penalty. But there are easier and more legitimate ways to increase turnout.

One way is making it easier to register to vote. Some US states have started using opt-out voter registration rather than opt-in. This means voters are automatically registered unless they indicate otherwise. And if we wanted people to vote, we could also either make election day a holiday or maybe not put it in the middle of the work week. (Ornstein has written about this himself.) Also, countries that use proportional representation also tend to have higher voter turnout, presumably because their vote has a much higher chance of actually electing someone that represents them.

As we’ve mentioned in previous articles, the voting method largely affects the partisanship of the winners. Want more moderates? Don’t use plurality voting, which splits the vote in the middle. Also, while you’re using a better voting method that favors the center, make sure that you allow competition during the general election and not just the two major parties.

Overall, if you want more people to vote, make it easier for them to do so, and give them a meaningful way to vote in elections that are actually competitive.

6. Eric Posner: Enact quadratic voting

“As people have known for thousands of years, democracy is prone to various problems, like tyranny of the majority. And one interpretation of tyranny of the majority is that a majority of people who don’t really care about a particular outcome or a particular candidate, let’s say, but nonetheless favors that candidate, could outvote a large minority of people, who passionately care.”

There are some creative voting methods out there, but this one has to be around the top of the list. Quadratic voting lets the voter indicate where they want their vote to matter. This reweighting makes the election outcome proportional for multi-winner elections. The effect on single-winner elections could be something similar to a Borda count, but it’s unclear without digging in more deeply. Even more interesting is that quadratic voting’s proportionality occurs not just within a particular election contest, but between multiple contests. That is highly unique for a voting method.

The fact that this method occurs between contests creates some serious logistics concerns. Essentially, a locality couldn’t use this without the entire state using it as well. That’s the only way you’d get consistent tabulation. Also, there’s the whole issue of auditing an election using this method. Good luck with that.

Perhaps it’s better to just to use a proportional voting method when you can if your goal is to make sure minority voices are heard. And if you want to focus on maximizing utility and factoring in degree of preference, score voting is something to consider. Plus it’s much easier.

7. Bruce Ackerman: Allow donation vouchers for voters ($50/voter)

“The problem isn’t that candidates are spending too much money letting us know where they stand. It’s that the rich are dominating the conversation.”

Providing donation vouchers to voters is an interesting approach. This route gets around potential First Amendment issues associated with contribution limitations. As referenced in the Freakonomics episode, campaign vouchers will be proposed as initiatives in both Washington and South Dakota this year. If they pass, we’ll have to stay tuned on how they work in practice, which should be interesting.

8. Sanford Levinson: Allow Constitutional amendments through referendum

“I believe that the United States Constitution would be a far better Constitution if, in fact, the national electorate had the same opportunity [to amend it].”

It’s interesting that Levinson started with pushing for a referendum for the US Constitution without also mentioning federal law in general. There are 14 states that allow referenda for state law and 16 for the state Constitution. (This only counts referenda that go directly to the voters.) The idea of a national referendum is far from foreign. Many other countries do it. And according to Gallup, two out of three Americans support the idea.

There’s a certain viewpoint that may question whether we even need national referenda. After all, isn’t that what we elect people for?

On the other side, there are issues where elected officials have a clear conflict of interest, such as election laws. Also, because our voting method is so bad, we have a hard time kicking people out. Despite an approval rate that hovers around the low teens, the Congressional incumbency rate remains as high as ever. When our Congress behaves so dysfunctionally, it may merit having the voters step in.

It’s also worth noting that referenda do not have to limit voters to two options. Often, there are more than two courses of action. Approval voting would be a good way to do this. Our advisor Dr. Steven Brams looked at this more closely in Nature.

9. Kathleen Hall Jamison: No audience in debates

“The reason I want to get rid of that idea, and I want to get rid of that concept, as quickly as possible is that we know that the audience response in the auditorium to the debate content affects the audience at home, and as a result can bias not only what people learn, but their evaluation of the candidates. … There’s no reason to think that audience is representative.”

Jamison has an interesting point here. If the debates are intended to be an objective source of information from candidates, then showing us a partisan audience’s response could shape public opinions.

There’s also a bigger issue here which didn’t come up. And that is that the Commission on Presidential Debates is quite partisan itself, its founders coming from the Democratic and Republican Party. Consequently, it uses its power to exclude voices outside the two parties. This gives Americans a distorted view of their options and shuts out alternative political ideas. If we’re to start anywhere with reforming the debates, it should be there.

10. Karl Rove: Keep the Electoral College

“The Electoral College pushes us towards a two-party system and that thereby promotes stability by providing a barrier against multi-candidate races and the kind of disasters that we see democracies in Western Europe and elsewhere … The Electoral College prevents, for example, presidents with a deeply minority vote.”

Rove goes wrong in so many ways talking about the Electoral College. Firstly, the Electoral College is not what gives us a two-party system. It’s a winner-take-all Congress that uses plurality voting to elect members into single-member districts. That’s Duverger’s Law, which we’ve written about before. Regardless, it’s hard to think about lack of competition within our elections being anything but bad.

Rove also mentions the idea of majority, which is a fair enough mistake. This is not understood well among the public. The fact is that no voting method can guarantee a majority when more than two candidates run. Even methods that elect the candidate who can beat everyone head-to-head don’t guarantee a majority, because that beat-all candidate doesn’t always exist. Some voting methods whittle down the candidate pool until there are just two in order to get a majority, but that’s coming about a majority artificially. The method is just using an arbitrary rule to get down to two candidates using procedures like runoffs which can still involve vote splitting.

Rove’s solution to get around this majority issue is to limit the options to two, but you have to consider that a small minority of the population give us those two options. This is the same strategy as the major parties when they use alarming ballot access laws to elbow third parties out of the election. Excluding competition does indeed prevent the spoiler effect, but using a more sensible voting method to begin with is far more democratic..

There’s a Supreme Court opinion that touches on a somewhat similar issue of whether we should think of representation being in terms of space versus people and how voting power should be decided. What’s clear is that if the Electoral College didn’t come from the Constitution, it would have to be overturned because it violates the equality of one person, one vote. (Swing states aside, votes from some states have significantly more weight.) Chief Justice Warren gave the opinion here in Reynolds vs Simms which outlined why this equality matters:

“Legislators represent people, not trees or acres. Legislators are elected by voters, not farms or cities or economic interests. As long as ours is a representative form of government, and our legislatures are those instruments of government elected directly by and directly representative of the people, the right to elect legislators in a free and unimpaired fashion is a bedrock of our political system.”

Originally published at electology.org on August 9, 2016.