The Median Voter Theorem has its origins in Harold Hotelling’s spatial model of firm strategies in a competitive marketplace. To help us understand the nuts and bolts of Hotelling’s model, we will take the scenario where two parties, in our case two ice cream vendors, compete to make the most profit from sale of ice creams on a hypothetical neighborhood block. Let us assume for simplicity that all the houses on this hypothetical block are equally spaced and are of equal size with equal number of residents living in each house. We will assume that these two ice cream vendors sell the same product at the same low price. We will also assume that it is a hot sunny summer day and everyone on this hypothetical block will buy ice cream from one of these two vendors. Since both vendors sell the same product at the same price, the residents’ decision to buy ice cream from one of these two vendors will depend on how close the ice cream truck is to their home. The customers’ decision rule will be to buy ice cream from the vendor who is closest to their home. We will also assume that these ice cream vendors are mobile. They can push the truck from one end of the street to the other. They are also able to move their truck around the other as well. The goal of both vendor is to make the most profit from the sale of their ice cream. Let us assume that Joe will start his business by placing himself at the west end of the block. Having no competitors, he has the advantage of having all the customers on that block coming to buy ice cream from his truck. Now let us introduce Joe’s competition, Ted, who places his ice cream business on the east side of the block. The two vendors now split the customers down the middle. Joe takes every customer from the center to the west end of the block while Ted takes every customer from the center to the east. What is the winning strategy for one of these two vendors? Holding all else constant, if Ted pushes his truck towards Joe, he can have some of Joe’s customers on the west side of the street to come and buy ice cream from Ted’s truck. Having said this, is there an optimal strategy for Ted if all else is held constant? Clearly, the best strategy is for Ted to situate his truck right next to Joe so that all the customers to the east side of Joe’s truck will come to buy ice cream from Ted. Now, let us loosen one of our conditions about Joe staying put. If we allow Joe to react to Ted, is there something that Joe can do to better his chances of winning this competition of ice cream sales against Ted? The answer is “yesâ€? he can. Holding Ted’s place constant, Joe can simply situate his cart to the right side of Ted. Ted, on the other hand, can move his cart to the right side of Joe; and Joe can do the same. They will continue. But will they ever stop? And when they do, where do they stop? The answer is yes they will stop. They stop in the middle of this block. The reason for this is that it serves neither Ted nor Joe any better to move away from the center of the block. That is, their marginal gain in sale of ice cream from an incremental move away from the center of the block is less than zero. They actually lose sales if either vendor moves away from the center of the block. The actual theoretical basis for the convergence of these two players to the middle of the street is based on a game theoretic concept of “iterative deletion of dominated strategies”. That is, for each successive deletion of the dominated strategies from either end of the street, the strategy that leads the players to the middle of the street allows the players to be better off no matter where the other player is positioned. Now let’s take a look at how the spatial model of strategic firm behavior translates into party strategy in a competitive electoral democracy. Assume now that instead of ice cream vendors, we have parties. And instead of selling ice creams, each party espouses a particular ideology, which can be mapped on a uni-dimensional ideological space running from extreme liberal to extreme conservative. We will assume also that the voting population is equally distributed along this ideological plane. We will also assume that each voter’s decision to vote for either one of these two parties will depend on how close each party’s ideological stance is relative to the voter’s ideology. If the electoral rule is such that the party with the most number of votes is the winner and we assume that voter participation rate is 100%, what is the party’s winning strategy? Similar to the ice cream vendors, each party’s winning strategy will be to situate themselves right next to each other at the middle of the ideological spectrum. The middle will be referred to as the median. We will also change one of the assumptions about the voting population without losing generality of our model. Instead of assuming that the voting population is uniformly distributed across the ideological spectrum, we will assume a normal distribution. That is, the distribution of the voting population is single peaked and symmetric around the median. This is not an unrealistic assumption to make since we are suggesting that there are less ideological extremists than there are moderates. The winning strategy still remains the same for each party. Each party will seek to seize the median voter.

in other words, both parties are same s**t, and there's no real choice

tnx a lot!

@iwantcoolname No. This is one dimensional binary response and only two parties. Real politics is multidimensional with multinomial responses and multiple parties. Oh yeah, and it requires EVERYONE to vote. So you couldn't be further from the truth.

@nels00 can't really see why, sorry.

One of the most powerful things about the Median Voter Theorem is that it works for ANY distribution of voters on a single dimension. You video is great, and I'm going to use it in class, but in showing only uniformly & normally distributed populations of voters, you might lead people to the incorrect conclusion that the equilibrium is always at the "middle" of the line. That is not true in general. The location of the median VOTER determines the EQ, no matter where that voter is on the line.

Understanding these concepts reveals the weaknesses of democracy.

Professor Anagnoson brought me here.

haha same here

Thank you!

Good. But I felt a hunger at the end for a critical view of the theorem.

It's a mathematically proven theorem, so there's nothing to be critical about, really. :Now as for how well the assumptions match reality and how the theorem is applied to analysis of the real world, that's a different story. 🙂

I felt a hunger for ice cream.

fuck ted

fantastic explanation

How is the median defined? For instance, if we have a set of policies and we live in a majoritarian democracy, wouldn't the more popular and favorable policies win out within this model? That doesn't happen. I'm not saying the theorem is incorrect — it sounds very logical — but I don't believe that's how American governance works. There are checks and balances put on majoritarian democracy. I believe the Investment Theory is more accurate in describing how party competition works.

great demonstration

yo ted and joe must be twins

This explains why Neoliberals and Neoconservatives (almost mirror images of each other ideologically with exception of a couple social issues) are the dominate ideologies in the USA. Interesting.

Great video but am i the only one who thinks the narrator sounds like Dwight schrute?.