For years, the United States has had an electoral college, and this has been a subject of contention particularly given the nature of recent close elections. After elections, we compare the popular vote against the electoral college, naively believing that the popular vote is a perfectly accurate reflection of popular opinion. As we know from every economic principle - incentives matter. This means that people base whether they vote on a combination of variables, including whether they believe their vote will matter in their state.
Compare the length of the voting lines in swing states to the lines in clearly Democratic or Republican states. Guess what.. they are longer. If I'm a liberal voter in Texas, I might not show up to the polls knowing that my one vote will make no difference in the outcome of the election. If I'm a conservative voter in California, there's no way my vote will change the winner-takes-all electoral vote in the state. Likewise, if I'm a liberal voter in Washington, D.C., I am sure that the majority will choose the liberal candidate whether I show up to the polls or not. So why vote?
The point is not that we should or should not have an electoral college. That's a discussion for another post. It is that we need to understand that ex post facto reasoning does not work when analyzing election results. Having an electoral college influences the popular vote and to assume that we can derive meaningful data about popular opinion from observing the popular vote is misguided.