There were really two problems with Hillary Clinton's claim earlier this week that she was ahead in the "popular vote."

First, the popular vote is not a relevant standard for determining who the nominee should be. PocketNines laid three very good talking points at Daily Kos a few days ago. I'll rephrase them here:

1. If popular vote were the standard, then we'd be counting Missouri four times as much as Minnesota, even though the two states are about the same size. The reason? Missouri holds a primary, Minnesota holds a caucus.
2. If popular vote were the standard, the candidates would have executed different strategies, likely spending less time in places like Iowa and New Hampshire and more time in larger states like California.
3. If popular vote were the standard, then states with closed primaries would be counted less than states with wide open primaries because more people vote in wide open primaries than in closed primaries.

Second, Hillary Clinton's claim to be ahead in the popular vote is false, even when you count both Michigan and Florida, and exclude Michigan's uncommitted vote, 73% of which was cast in support of Barack Obama.

In seeming contradiction to that, there is a number at RealClearPolitics that purports to show Clinton ahead by 11,561 votes when you count everything except for the uncommitted vote in Michigan. The problem with this number is that it doesn't actually count everything -- it leaves out the Washington primary and the Texas caucuses. It's true that in the case of the Washington primary, no delegates were at stake -- but then again, no delegates were at stake in either the Michigan or Florida primaries.

When you include the Washington and Texas vote totals, giving you the most comprehensive possible measurement, Obama takes a 16.21 million to 16.09 million lead over Clinton.

Clinton might reject this approach as "double-dipping." Well, if that's true then one solution is to exclude the Washington primary results since no delegates were at stake. If you do that, then you need to exclude both the Michigan and Florida votes because no delegates were at stake there either. Next, you you must do a weighted average of the Texas primary and caucus results, since delegates were at stake in each contest. Using that metric, Obama has received 14.5 million votes and Clinton has received 13.8 million votes.

You can also come up with a dozen other ways to count the vote, some of which Clinton no doubt leads. (For example, she is certainly leading in the states where she's won.) Which brings me to the final, and perhaps most important, talking point about why you can't use the popular vote as a metric in the Democratic nomination process:

Nobody can agree on how to count it.