Does the Third Party Matter?

Just read this article, The Power of the Protest Vote, which talks about how the third party in swing states might affect the outcome of general election, especially when the race is in a tie, like the 2000 Gore vs. Bush.

It reminded me of an interesting question which was tested once in a course I TAed for, ECON200 Introductory Microeconomics. The question is as follows:

In the US, there have always been only two major political parties. Although third-party candidates such as Ralph Hader have some, perhaps double-digit, support in the general population, these candidates never get elected. Apply marginal analysis to this question and explain why the US operates with only two major parties.

I remember at that time we gathered in the office and tried to figure out a perfect answer to students, but none of us was convinced by any interpretation we came up. Basically, I think, because the electoral system of the US is that the candidate who wins more popular votes in a state gets all electoral votes of the state (winners take all), this all-or-nothing makes a third-place candidate's marginal benefit of extra effort equal to zero. In other words, even if he finally moves to the second place in the election, he gets nothing. His cost is definitely greater than his zero marginal benefit. The fundamental principle of economics says that an action is rational only when the marginal benefit is higher the marginal cost. Therefore the only way for this third-party candidate to exert his influence is to attach himself to one of the two leading candidates. This is also what the article above is about.


(I am not pretty sure if my explanation makes sense. Any comment is welcomed.)