What Matters: August 2006
Range Voting: The Best Way to Select a Leader?
By Warren D. Smith '84
The CRV's shortest splitline algorithm eliminates gerrymandering. Basically: find the shortest line that splits the state into two parts with the right population ratios, then continue recursively. This produces better district maps for less cost than present methods and is completely unbiased. Above: The blue map shows Massachusetts as gerrymandered to have 100 percent Democratic congressmen (courtesy Adam Carr, based on US census data), versus the districts that would have been produced by the CRV's method.
The world is approaching major crises, such as the end of cheap oil, the exhaustion of important fossil water reserves, climate change, overpopulation, and nuclear and bioweapon proliferation. So it is more important than ever that the world make the right decisions. But what is the decision-making algorithm for the world? With the US now the sole superpower, the closest simple approximation to the answer unfortunately is our country's appalling voting system. It features:
- 98 percent year-ahead predictability of all important races (the largest among democratic nations) and uncontested state house races about a third of the time;
- Rampant gerrymandering seemingly worse than most democracies;
- An elected partisan official in charge of supervising elections in 33 out of 50 states (and in other states he is appointed by the governor, which may be even worse);
- Massive two-party domination (over 99.5 percent) causing opposing views to be almost completely blocked out of both power and the media;
- An electoral college that inspires candidates to cater to only a small number of swing states while ignoring the rest;
- Highly restrictive ballot-access laws in many states designed to ensure two-party (or one-party) domination.
How can we change this system? What should we change it to? There are many obvious improvements, such as eliminating gerrymandering, not allowing partisan top election officials, and doing away with the Electoral College, but unfortunately, these changes are unlikely.
In 2005, I founded the Center for Range Voting, which is currently suggesting a new voting solution. With range voting (RV), voters rate every candidate on an 0-9 scale (e.g. Gore=9, Nader=9, Browne=5, Bush=0, Phillips=X) or intentionally leave a candidate's score blank (the "X"). The candidate with the greatest average score wins. Blanks are not incorporated into the averaging, and hence do not affect outcomes.
Here are some properties of range voting:
- Voters can provide quantitative information about all (or any subset of) candidates, as opposed to qualitative information about just one.
- Votes tend to be comparatively honest. With plurality, voting for a third-party candidate is strategically stupid. For example, polls showed about 90 percent of Nader supporters voted for somebody else in 2000. But with range voting, there never is any incentive to betray your favorite candidate.
- Range voting is immune to vote splitting between similar candidates, which can cause them both to lose.
- Range voting can be used on every voting machine in the US right now without making modifications or reprogramming.
If RV is enacted, then many problems, such as two-party domination, gerrymandering, 98 percent predictability, and media blackout of minor-party views will cure themselves over time. With RV, your vote says and does more, making voting more worthwhile. More than two candidates can have a chance, making it more likely one will appeal to you. Since people vote more often when they don't know the outcome, when they think their vote can have an impact, and when there are appealing candidates, RV should yield higher turnouts.
Voting simulators give range voting high marks
Computer simulation studies give strong evidence that RV is the best among commonly proposed single-winner voting systems. A statistical yardstick called Bayesian Regret (BR) valuates voting systems by performing Monte Carlo computer simulations of an unlimited number of elections with artificial voters and candidates. The BR of a voting system depends not only on what voting system it is, but also on the number of voters and candidates, how the simulation makes each voter feel about each candidate, how ignorant voters are, how strategic they are versus how honest, etc.—there are many knobs on the side of the simulator you can turn. I performed a comprehensive study of all the most common voting system proposals, measuring the BR of each with 720 different knob settings. Range voting had a BR as good or better than every competitor at every one of the 720 settings. Six years later, I conducted a new simulation of three-candidate, left-middle-right scenarios and again RV clearly came out superior to all commonly proposed competitors.
Democrats and Republicans should enact RV in their primaries, especially the Iowa 2008 presidential-nomination caucuses. RV will tend to produce better presidential nominees versus plurality, benefiting that party's chances in the real election, without hurting them in any way. It will make them look like reformers and generate free publicity for them, voting reform, and the Iowa politicians who push for it. Iowan voters, as well, would appreciate RV's increased expressivity. And all it takes is a party's internal rule change, not even a new law. That means RV can really happen.
Visit the CRV Web site for more information, to post to the RV bulletin board, or to email CRV members.
About the Author
Warren D. Smith '84 received a double SB from MIT in physics and math then a PhD in applied math from Princeton. He has published numerous mathematical papers and is writing a book on mathematics and democracy.
What Matters is a guest opinion column written by a different MIT alumnus or alumna. The views expressed are entirely those of the author and do not necessarily represent the views of the Alumni Association or MIT. Interested in writing a column? Email whatmatters@mit.edu.
