Barry Nalebuff's Alternative Election Theory | Big Think

Barry Nalebuff's Alternative Election Theory

New videos DAILY

Join Big Think Edge for exclusive videos

----------------------------------------------------------------------------------

Barry Nalebuff on the problem of the third party candidate and what the French Revolution can tell us about campaign strategy.

----------------------------------------------------------------------------------

Barry Nalebuff:

Barry Nalebuff is the Milton Steinbach Professor of Management at the Yale School of Management. Professor Nalebuff has written on a wide variety of subjects ranging from strategy to pricing, bidding to bargaining, and innovation to incentives. He is an expert on game theory and has written extensively on its application for managers. His most recent book, The Art of Strategy, is an update of the best-selling Thinking Strategically, which explains the fundamentals of game theory using real world examples.

Professor Nalebuff's work on strategy focuses on the fundamental duality in business—the conflict between cooperating to create a pie and competing to divide it up—which he presents in Co-opetition. His book, Why Not?, focuses on providing a framework for problem solving and ingenuity. His work on product bundling was featured in the European Union's investigation of the proposed GE-Honeywell merger.

----------------------------------------------------------------------------------

TRANSCRIPT:

Barry Nalebuff: My own work in political science applies game theory to elections. And one of the things that we discovered there is that our [i.e. US] electoral system really doesn’t work when you have more than two candidates. We saw that if you go back, in terms of Ross Perot having a big influence in the [Bill] Clinton election.

We saw that inadvertently, with Pat Buchanan in the Florida election.

And we saw that in the primaries, where you might very well have imagined that either Mitt Romney, or others that were more in the conservative side of the Republican wing, might have won if their teams had been pooled, but there was one person who is a little bit separate, and in this case was [John] McCain who ended up getting the plurality.

And the problem is that, I have to vote not just for the person who I like the most, but also the one who I think may have the momentum. And it becomes a strategic question, as opposed to a preference question.

So how can you design electoral systems that allow people to do a better job expressing their true preferences, as opposed to their strategic preferences. And probably we have a chapter of this on the "Art of the Strategy" [book].

This is an old problem and it goes back actually to the French Revolution. A gentleman named Marquis de Condorcet was frustrated by the possibility that you get voting cycles, that you can have one person A who beats B, you’ve got B who beats C, and C who beats A. And it turns out it’s not that hard to come up with this, and so then you’ve got this challenge which is, who you end up picking or what policy you end up choosing, depends entirely on the order or the agenda, as opposed to what the will of the people are.

And therefore we hold up this national democracy as sort of the be all and end all. But actually that doesn’t tell you how to vote and how to express or aggregate people’s preferences.

Ken Arrow won a Nobel Prize for telling us that there is no perfect way of doing this; that each way has its problems. The way that I like to do it, my preference here is, that literally to have people vote on every possible pair. So if you’d like they could vote on Hillary [Clinton] versus [John] McCain, Hillary versus [Mitt] Romney, [Barack] Obama versus [Mike] Huckabee, Obama versus McCain and so on.

Then you can say, okay, this election was 70-30, this one was 65-35. Was there one person who could just beat everyone by a majority? It’s so great, that’s the one we want and that’s actually what it was now called a Condorcet winner.

But if there isn’t, is there someone who comes pretty darn close, always winning? So I don’t mind the fact that this person lost one election 51-49, as long as he beat everybody else by a healthy margin, and the person who they lost to at 51 lost to somebody else by more than 51.

And you say, well, okay. That’s really complicated. How could I get people to go and vote 26 different times or 100 times? And the answer is simply ask them to rank their candidates. So they’ll say this person is first, second, third and fourth, and having rank the candidates where a computer can vote for them.

Read the full transcript at