Quick math quiz: What is the sum of 5 porpoises, 11 fire engines, the Eiffel Tower, 8 molecules of ammonia on the outermost ring of Saturn, and 13 kiwi fruits? In some sense, it's a collection of 38 things, but the collection is one nobody in possession of his or her wits could possibly take any interest in. The moral of the story is that just because basic arithmetic allows you to add two figures together doesn't mean you should, or that the sum will be meaningful if you do. On the contrary, if you're careless, unscrupulous, or both, the result of your work is liable to be a mereological abortion.
Which is why those folks pushing back against Hillary Clinton's preposterous claims to be leading the national primary popular vote have largely been following the wrong trail, as admirable as their efforts have been. Yes, it's true that to arrive at an aggregation of all the votes cast in all contests that puts Clinton ahead, you have to first assume that "the will of the people," nebulously defined, trumps all considerations of procedural fairness, only to then throw out hundreds of thousands of expressions of popular will in support of Barack Obama on procedural grounds. So that even if the Clinton math weren't a transparently cynical and breathtakingly mendacious display of disrespect for the intelligence of the American people (including Clinton supporters), it would still be flagrantly contradictory on its own terms.
But all that is beside the point because there is no such thing as a meaningful national popular vote in presidential primaries, at least as they are currently constituted -- and in perpetuating the myth that there is such a thing, the media and a surfeit of non-Clinton supporters have played an unwitting role in enabling Hillary Clinton's mendacity to gain a purchase in the national political conversation. Adding together the votes of all the primaries and caucuses and seeing who is ahead is like adding porpoises to fire engines to Eiffel Towers to ammonia molecules to kiwi fruits: the sum is numerical gibberish. The difference between the two cases is that it's easy to disguise national popular vote totals as the addition of apples to apples -- they're all just votes right? -- so that thus far, even those who have noticed that there is something suspicious about the math haven't been able to put their fingers on just what the problem is. While at the same time, the Clintonites have been staking out absurd comparisons of their esoteric arithmetic to bloody struggles for democratic and human rights, in an effort to smother any scrutiny of their popular vote claims. No wonder the Clinton math has gulled so many people; no wonder critics of the Clinton math have been emphasizing secondary points.
To see why the national popular vote in a primary is a case of porpoises plus fire engines rather than apples plus apples, it's helpful to see why the national popular vote in a general election really is a meaningful count. Beyond the emotional manipulation involved in Hillary Clinton's likening of her plight to Florida 2000, the analogy between the general election and the primary is what gets the whole deception off the ground. But a general election for president has two qualities that a presidential primary campaign lacks, and which make it reasonable to talk about the national popular vote in the former but not the latter: (1) all fifty state elections in a general election are synchronic, i.e., they take place at the same time; and (2) the fifty state elections are all open to the same pools of voters and governed by roughly the same procedure.
No two elections that lack either quality can be combined into a meaningful aggregate result. The reason that synchronicity is essential for aggregating elections is that the result of an election is not some Platonic form of the popular will, but simply an imperfect measure of popular preferences at a particular time. Consequently, elections held at different times are measures of categorically distinct things. Suppose the last presidential election had been held in September 2005, after Hurricane Katrina, rather than November 2004. Is there any question that the result in both the popular vote and the electoral college would have been different? Now suppose a group of states with half the American population had voted in November 2004, and the other group had voted in September 2005. In that case, the electoral college results may or may not have been different, but either way, referring to the national popular vote would be utterly senseless. There would be an overall popular vote for the November 04 group, and another for the September 05 group, but adding the two would be a pointless arithmetical exercise, not something anyone could claim with a straight face is a measure of "the will of the people."
Likewise, the results of two elections can't be added together meaningfully if the electorates participating in each contest are incommensurate, or if the elections themselves are incommensurate. Suppose that in the general election, 25 states allowed only men to vote, and 25 states allowed only women to vote; suppose a different 25 states had a minimum voting age of 18 and the other 25 a minimum age of 21; suppose a still different group of 25 states had a maximum voting age of 72. Suppose in half the states, voting meant casting a secret ballot in the ordinary way, and in the other half, voting meant casting a rock in one of two piles like the Athenians used to do. In that case, once again, by electoral college rules, the winner of the most electoral votes would win the presidency legitimately, but it would be risible to say that whoever happened to have received the most total votes had won "the will of the people."
These considerations are straightforward. Anybody arguing that if the winner of the most popular votes nationally in the foregoing scenarios wasn't the same person as the winner of the most electoral votes, then as Hillary Clinton puts it, "the will of the people isn't realized and our democracy is diminished," would and should be laughed off the stage. It's not that electoral college totals in such circumstances are a truer measure of popular will than popular votes. It's that when elections aren't held at the same time, for the same electorate, under the same voting procedures, there is no measure of the overall popular will, whereas electoral votes just happen to be what decides the presidency in our system.
That's why there is no meaningful national popular vote in presidential primaries. The constituent elections of the primaries are diachronic, scattered over five months, and so while it might be reasonable to talk about who had won the most total votes in the February 5 states, who won the overall popular vote between Indiana and North Carolina, etc., it's nonsense to draw any conclusions from the sum of all the votes cast from January 3 to June 3. But that's not all: Whereas any registered voter may participate in a general election by casting a secret ballot, a primary campaign is divided between ordinary elections and caucuses, which are utterly distinct and incommensurate procedures for measuring popular preferences, and so trying to combine them into a total is a priori nonsense. (The Clintonites' Alexandrian solution is to disregard caucuses entirely -- sometimes all the caucuses, sometimes just four -- because, as you know, they care so much about democracy.) Moreover, there is incommensurability between the ordinary primary elections themselves; some are open, some are closed, some are semi-open to varying degrees, meaning that when the overall vote totals are relatively close, asserting that "the will of the people" is for one candidate over the other devolves into nonsense as well.
All of which does not mean that the elected delegate count is a better proxy for "the will of the people" than the popular vote; rather, under the rules that govern presidential primaries, there is no useful or usable proxy for "the will of the people" at all, and the winner of the delegate majority is simply the winner of the primary campaign under those rules. Maybe future primaries should be made democratic, maybe not (I think not, and that the suggestion that they should involves a mistake about what primaries are), but either way, claiming that the delegate count should be overturned in deference to "the will of the people" is simply a shell-game that, as I think I've demonstrated conclusively, doesn't stand up to minimal scrutiny.
So yes, it's grating to hear Hillary Clinton and her supporters claim that she is in ahead in the popular vote, but she is only in a position to make that claim because non-Clinton supporters have preemptively and foolishly acquiesced to the idea that the sum of all the votes cast for each candidate in all the primaries and caucuses is a meaningful and interesting number in the first place. People who persist in arguing that the popular vote is "the will of the people" and must therefore be heeded are entitled to no more or less respect and attention than those who heed the sum of 5 porpoises plus 11 fire engines plus the Eiffel Tower plus 8 molecules of ammonia on the outermost ring of Saturn plus 13 kiwi fruits.