So there's a bounce.

Post-convention polling bounces used to be taken for granted, but over the last three cycles we've learned not to do that. For one thing, they don't always happen—see John Kerry in 2004—and even when they do occur, they've been shrinking since 2000.

All of which is why I've tried to stay genuinely open-minded over the last couple weeks. It wasn't a sure thing that either Trump or Clinton would get a bounce, or how big it might be if one (or both of them) did.

In the three days after the close of the Republican convention in Cleveland, it was clear that Trump did indeed bounce. His bounce was modest—somewhere between +3 and +4 points—but it was real and solid enough to move him ahead of Clinton in the RealClear poll of polls. But then either Trump's bounce began to dissipate or Clinton's convention bounce took off early (or a little bit of both). As I write this, it appears that Clinton got a significantly larger boost from the Democratic convention.

The net effect, for now, is basically that Clinton is in a slightly better position than she was three weeks ago with a lead over Trump that amounts to +5 or +7.

And if you look at the polling history of this race, that feels a lot like equilibrium. Clinton's lead has been between +4 and +6 since June in the two-way matchup. In a four-way matchup with Gary Johnson and Jill Stein, her lead is only a touch smaller, between +3 and +5. All of which suggests that that's where this election lives, all things being equal.

Of course, things aren't always equal. Anything can happen and history, while a helpful guide, isn't a straight-line prediction. (If you care about the history of post-convention bounces, the rule of thumb is this: The state of the race 30 days after the convention ismore or less how things look on Election Night.)

But for the sake of argument, let's set that aside and pretend that what we have here is a true equilibrium for the race. If the environment really is Clinton +5 (or so), it suggests three interesting things about the road going forward.

1) It means that the race is close enough that maybe the equilibrium isn't what it looks like. Remember how the Brexit vote looked like a narrow, but solid, win for "Remain"? It turned out that polls were simply wrong all along.

Complaints about "bad polls" are the last refuge of a loser. You may recall that late into 2012 Republicans kept complaining that the all the polls showing Mitt Romney losing to Obama were skewed. Their specific complaint was that the polls were consistently over-sampling Democratic voters. (They weren't.)

Now, maybe sampling errors were behind the Brexit mistake. Or maybe the problem was more fundamental. A friend of mine who follows European politics for a living maintained the entire time that the "Leave" forces were likely to win the Brexit vote. He held this line even on election day. His operating theory was that the "Leave" position was essentially un-pollable. It relied on such a strange coalition of voters that pollsters had no idea how to accurately sample them.

And it's possible that Trump is the same way.

Personally, I doubt it. The close look at Trump's primary coalition showed that his support really did come from conventional Republican voters—just not the Republicans who usually vote in the primaries. But I wouldn't dismiss this possibility entirely. And if true, it means that the race is close enough that a systemic polling failure would make a Trump victory conceivable.

2) A Clinton +5 lead is close enough that the eventual outcome in November is susceptible to an external shock. Let's say something happens the last week in October. Maybe there's a terror attack. Maybe someone gets shot at a political rally. Maybe "Wikileaks" releases a bunch of emails from the DNC or the Clinton Foundation or Hillary Clinton's private email server.

Or maybe there's an honest-to-goodness international incident. If you're the leader of a foreign country with a strong preference for one candidate over another, this election represents a once-in-a-lifetime opportunity for you to influence events and get the American president you desire. Maybe you invade a neighbor, with regular or irregular forces. Maybe you encourage a non-state actor to do something to stir the pot.

Of course, a shock to the system doesn't have to be premeditated or nefarious. The world is an unpredictable place. No one planned the fall of Lehman Brothers and that was probably the most decisive external event in modern politics.

A lead of +5 is simply not secure against disruption.

3) Items #1 and #2 on our list are the long-shot plays, the dogs least likely to bark. The most important consequence of a race that naturally tends to Clinton +5 is that it stands to inflict the maximum amount of long-term damage on the Republican party.

Imagine if you will a world in which it's September 15 and Clinton leads Trump by 9 points. If that were to happen, you'd see Republican officeholders abandoning ship. The RNC would shift all of its spending to down-ballot races. The next time Trump cozied up to white supremacists or insulted a Gold Star family or talked about how badly he was beating Kelly Ayotte in New Hampshire (in case you missed it, this is a real thing), Republicans from Paul Ryan to John McCain and everyone in between would use it as an excuse to jump off the train. Not that it would matter. The Senate would be lost. Maybe the House, too.

Yet it would be very helpful in the long run, because the Republican party would have finally turned away from Trump, and done so before his actual defeat. This would go a long way toward removing Trump's taint, repairing the brand, and rebuilding the party. It would also restore some of the party's honor—if you care about that sort of thing.

But with the race hovering in the 5-point range, it will have the effect of freezing Republicans in place. They'll understand that Trump is very likely to lose, but they'll have just enough hope that they'll force themselves to keep their arms over the flame for a few more weeks. They'll fool themselves (or be intimidated) into throwing good money after bad.

And then, when Trump loses, they'll have the worst of all worlds: They'll have ridden the Trump train all the way to the end of the line. Which means that the entire party (with a few notable exceptions) will have his golden-haired carcass hanging around their necks. The Republican party will be like France trying to rebuild after the fall of the Vichy government. And there will be no way to succeed until all of the little Marshall Pétains have been replaced by Republicans who never capitulated.

And if you think I'm being hyperbolic about what the legacy of Trumpism could do to the party, go and read the transcript from his Sunday interview with George Stephanopoulos. It's one thing to make excuses for a man like this during the heat of an election.

It's another to have to pay the bill after the votes are counted.