UP@NIGHT

There may be good reasons for Democratic superdelegates to hold off on making a decision between Senators Clinton and Obama, but the national popular vote is not one of them. Over fifty years ago Jean-Paul Sartre warned us about something he called “bad faith.”  We are in “bad faith” when we are free to make a decision but convince ourselves that there is something preventing us from making this decision. For example, those who seek advice can be in bad faith. They say that they cannot decide until they get some good advice, knowing in advance what the advice will be.

How does this relate to the so-called popular vote? Well, if we can believe a lot of pollsters and journalists, the SUPERdelegates really want to know what the national popular vote is going to be before they can make up their minds. Until they know, they cannot choose between Clinton and Obama. But for anyone who has seen some of the (often well intentioned) attempts to calculate the national popular vote, it should be obvious that no such total will be available. There is no evil plot afoot. The simple reality is that states have chosen very different ways to select delegates. The first great divide is between caucus states and non-caucus states. And then there are the different ways in which the caucus states choose to select delegates. But although many have spent many hours focusing on the latter, these differences are really trivial. The bottom line is that any attempt to determine a national popular vote runs into the apples and oranges problem. Caucus states and primaries are different animals, and if you attempt to combine them into a national popular vote, you will short-change the caucus states. Why? Because every statistical model that seeks to create a national popular vote from these apples and oranges will be suspect and subject to abuse. The caucus system simply involves many fewer participants. One can complain that it is less democratic, although no candidate did so before Iowa. But the Democratic Party did not warn the citizens of caucus states that their systems would mean reduced representation, and this is just what it would mean if pollsters create statistical Rube Goldberg devices for calculating a national popular vote.

Pollsters and journalists are free to go through all of the statistical contortions that their patience will allow. They are free to create formulas, and then more formulas. However, they should know this: they are supporting the bad faith of some of the superdelegates. They are enabling people who have a responsibility to make a decision avoid a decision. They are giving them an excuse. They are telling superdelegates that there may be an Oz-like “metric” that can help them out of their alleged indecision. I say, let them fish or cut bait.