You didn't think that the Donald was going to pay a price for insulting John McCain, did you?

Trump went after McCain on July 19. He was sitting a half-point behind Jeb Bush in the RealClear average that day-15.0 to 15.5. After July 19, Trump's arrow-which was already pointing upwards-went vertical. Today he's at 18.2. The only person hurt by Trump insulting American POWs was Jeb!, whose number has dipped to 13.7.

Maybe next week Trump will go after Mitt Romney. Crazier things have happened. But remember: It's all fun and games until he mounts a third-party run.

A couple weeks ago I wrote that Trump "holds in his hands something like veto power over the Republican quest to win the White House." Since then he's gone back and forth publicly about his openness to an independent bid. But what does the math on that actually look like?

Let's go back to the year 2000, which is the last time an independent tipped the balance of an election. George W. Bush got 50.45 million votes; Al Gore got 50.99 million. Ralph Nader got 2.88 million_just 2.7 percent of the total. And we all know how that turned out. 

2000 was an extraordinary election in the sense that the votes were split quite evenly between the two parties. But it looks a little more ordinary if you assume that most of those Nader votes would have gone to Gore, which would have given the vice president a pretty comfortable margin of victory-about 3 million votes. It was also a pretty ordinary election in terms of turnout, with 54.2 percent of eligible voters participating. (Compared with, say, 1992, when 55.1 percent of eligible citizens voted.)

But the 2000 cycle began a period of increasing political intensity and polarization. (People tend to blame the Iraq war for these increases, but I suspect that it was actually a combination of Bush v. Gore, 9/11, and Iraq which created a catastrophic example of resonance in the body politic.) In 2004, voter turnout increased dramatically, to 60.1 percent, and Bush won 62 million votes-a tremendous increase over his 2000 total. 

In 2008, turnout increased again, to 61.6 percent. John McCain came close to Bush's 2004 vote total, with 59.9 million votes, but got blown out of the water by Barack Obama, who won 69.5 million votes. To put Obama's achievement in perspective, he went out and found 10 million more votes than Kerry, and 7 million more votes that Bush, just four years prior.

But after 2008, the political intensity began to wane. In 2012, turnout declined for the first time in a generation, dropping back to 58.2 percent. Obama took home 65.9 million votes_still a big number, but nearly 5 million fewer than he'd won in his first election. (Which, by the by, almost never happens in presidential politics. It was only the third time that a president won reelection with a declining share of the vote.)

It may be that, at least as regards voter participation, our politics is reverting to the norm after a generation of increasing political intensity. That would make sense. These fevers usually burn themselves out because normal people can't stay hyper-engaged with politics for forever. They have actual lives to lead.

So what happens if turnout stays around 58 percent for 2016, or maybe drops a bit further? 

It would mean that 63.4 million votes is enough to win the White House. The Republican nominee would have to find 3.5 million more votes than McCain got; 2.5 million more than Romney; and 1.5 million more than Bush 2004 (which was the most impressive Republican presidential campaign since 1980).

That's not impossible. The country is bigger than it was in 2004 and there are more voters. Hillary Clinton is a formidable, but beatable, candidate. And there are lots of reasons to believe that Democrats start in a small hole because of historical trends when one party seeks to hold the White House for a third term.

All of that said, 2016 seems likely to be a close-run thing. How many votes could go to a third-party candidate without putting the project in jeopardy? How many votes could make it impossible?

Ralph Nader took home 2.88 million votes with a campaign assembled from bailing wire and chewing gum. If Donald Trump decides to run, do Nader's numbers look like a ceiling to you, or a floor? How likely is it that the Republican nominee could get to 63.4 million while losing 2.88 million to Trump? If he (or she) pulled that off, it would mean that their natural total was actually 66 million-better than Obama got in 2012.

And what if Trump ran closer to Ross Perot's performance? In 1992, Perot won 19.7 million votes. How does a Republican candidate get to even 50 million votes-let alone 63.4 million-if Trump is lopping that many off the top?

I'm all in favor of the Trump show, I suppose. If nothing else, it gives me something to write about. But it might do him-and the Republic-a great deal of good if his supporters could get him to commit to not mounting a third party challenge sooner, rather than later.

Because on the off chance you need reminding-the foundations of our civilization aren't settled through elections. They're decided by Anthony Kennedy. And he, along with three of his robed confederates, will be over the age of 77 by January 2017.