Lucas Jackson / Reuters

Would he or wouldn’t he? In the end, Donald Trump signed the Republican National Committee’s pledge, but he did it his own way, in that inimitable, inch-high Sharpie scrawl, at once as angular and overstuffed as the man himself.

The document is Trump’s agreement (along with other Republican candidates) not to mount a campaign as an independent or with a third party in the event that he loses the GOP nomination.

Not that he’s too worried about that. “We’re leading in every single poll,” he noted, correctly, at a press conference Thursday. “A new poll came out today where we’re over 30 percent.” Moreover, he felt that the Republican Party had treated him fairly—at least since he shot to the top of the polls—so he decided to go along with the pledge.

“Frankly, I felt that the absolute best way to win and to beat the Democrats ... is if I win the nomination and go directly against whoever they happen to put up,” Trump said. “For that reason I have signed the pledge.”

The RNC’s pledge is the culmination of weeks of handwringing about what the Republican Party could do about its Trump dilemma. The clear vein of discontent into which he’s tapped showed the yuuuuge risk for the GOP: If Trump lost the nomination but decided to run as an independent, or representing a third party, he would likely attract a great number of voters who would otherwise line up in the R column, while winning over far fewer Ds. The winner in such a three-way race would almost certainly be the Democratic nominee. As Trump's campaign started to look durable, Republicans worried while liberals rubbed their hands in glee.

There was ample reason for the worry. After all, Trump has little real connection to the Republican Party, having only (re)joined it a few years ago. None of the tricks that party mandarins can rely on to rein in other candidates have worked on Trump. Worse, at the August 6 presidential debate, Trump was the only candidate who refused point-blank to rule out a run if he didn’t win the Republican nod, just in case he didn’t get the respect he thought he deserved. Several states began discussing making loyalty pledges a condition of getting on the ballot, followed by the national party taking up the charge. Here’s the text:

I _______ affirm that if I do not win the 2016 Republican nomination for president of the United States I will endorse the 2016 Republican presidential nominee regardless of who it is. I further pledge that I will not seek to run as an independent or write-in candidate nor will I seek or accept the nomination for president of any other party.

On the surface, it may appear that the GOP has outsmarted Trump, forcing him into a position where he had to show his fealty to the party, and which will ensure that he doesn’t wreck things when (as most political wise-people still expect) the Trump bubble bursts.

Perhaps, however, Trump is the one outsmarting the GOP. Say what you will about his policy incoherence or his inflated net worth, but wily contract negotiation is one of The Donald’s undisputed skills. Most obviously, there’s little evidence that the pledge is legally binding, and no sooner had it been reported that Trump would sign than people began speculating about how and when he’d break the pledge. (Trump maybe left himself some wiggle room: “I see no circumstances under which I would tear up that pledge.”) Moreover, assuming all the other Republicans also sign, it means that (at least insofar as anyone is bound by the pledge), they will all have to support Trump if he wins the nomination. Agreeing to a loyalty oath would have made no sense for Trump on August 6. But a month later, having shown that his bubble isn’t as fragile as everyone thought, it’s a no brainer: Now the GOP won’t be able to slough him off, either!

The pledge is open to criticism on its own terms. Bill Kristol quickly spotted how this could backfire with the conservative grassroots. One reason conservative voters have been attracted to Trump (and Ben Carson, and Carly Fiorina) is that they’re not traditional politicians beholden to a political machine that activists feel has betrayed them over and over again—on immigration, on abortion prohibitions, on spending, and so on. Forcing them into line with the party line will hardly make the GOP look open and accepting of the rank and file.

“If Trump wants to defy it later, all he has to do is say that he’s not going to let a bunch of establishment punks who’ve been bought and sold by lobbyists stop him from offering voters a real choice next fall,” writes Allahpundit. “He won’t lose a single vote for breaking his promise.”

Loyalty pledges, and candidate pledges more generally, aren't exactly new. The Virginia GOP made candidates sign one in 2012, and Democrats have done so in the past as well. But Republican candidates in particular seem to face an increasing raft of demands for signatures. Most obviously, there's Americans for Tax Reform’s pledge not to raise taxes. Candidates who break it can expect to face a barrage of attacks from Grover Norquist and Co. Those who don’t may struggle to gain traction in the GOP. And plenty of other groups have tried to emulate ATR—asking candidates to, among other things, oppose same-sex marriage, stay faithful to their spouses, and more.

Candidates often bristle at these demands, because they don’t want to lock themselves into anything, and because it makes it harder to go back on a campaign promise. But it makes sense that the pledges would flourish in the Republican Party, where voters often tell pollsters that they highly value elected officials sticking to their principles—as opposed to Democrats, who tend to put a higher value on compromise and cooperation.

It's tough to find many cases of candidates failing to win the nomination, running outside the party, and really affecting an outcome—though Trump’s run, as has been remarked unto tedium, breaks many molds. Republican John Anderson made a third-party run after dropping out of the GOP nomination race in 1980, but Ronald Reagan still won (and Anderson’s support may have come largely from would-be Jimmy Carter supporters). In 1968, George Wallace didn’t bother to try for the Democratic nomination before running as a third-party candidate. Anything farther back is getting into the ancient era of brokered presidencies.

Even if Trump wanted to run a third-party campaign, though, and even if he didn’t feel encumbered by the pledge, it’s unclear whether he’d be able to do so. As Tim Alberta reported earlier this week, managing to get on the ballot around the country outside of the auspices of a major party is a costly and, more prohibitively, logistically complicated thing to do. He’d need to get a jump on that as soon as possible. But it would be even harder thanks to “sore-loser laws,” which throw up obstacles to candidates who have run on a party ballot in a primary getting into a general election. (It would be entertaining if Trump became an unlikely hero not only to campaign-finance reformers but also to ballot-access reformers.)

Depending on how you look at it, these laws are either a good way to keep gadflies out of the process or else an undemocratic racket created by the Democratic and Republican parties to maintain their duopoly. Third-party advocates and political scientists have tended to take the latter view. Barry Burden, Bradley Jones, and Michael Kang calculated in 2014 that sore-loser laws are a major driver of polarization in today’s political system—perhaps as much as a tenth of the distance between the parties.

The clearest examples of how this happens have come from Tea Party challenges to Republican establishment candidates in recent years. Primary electorates tend to be far more ideological than general electorates, so that a very conservative candidate can beat a more moderate one in the primary, even if the general electorate would prefer the more moderate one. And since both time-honored custom and sore-loser laws prevent it, the moderate candidate, once defeated in the primary, has to slink away.

As Mickey Edwards noted in 2014, this happened in Texas’ 2012 U.S. Senate race. David Dewhurst, the Republican establishment pick, was very popular in the Lone Star State, but upstart Ted Cruz beat him in the primary, and then triumphed in the general election—because after all, Texans may have liked the moderate Republican best, but the conservative Republican still trumped the Democrat. The result was that Cruz went to Washington, where he has become one of the most conservative and polarizing figures in the Senate.

Now Cruz is one of Trump’s rivals for the Republican presidential nomination. Whether Trump is a “moderate” in comparison is hard to judge—his views on immigration are (to borrow a phrase) severely conservative, but his positions on other issues, when stated and comprehensible, are more eclectic. In general, however, loyalty pledges like the one the candidates are signing today are likely to have the same effect as sore-loser laws. What parties do is generally their business, but the pernicious and undemocratic results of loyalty pledges should give Reince Priebus and his fellow party leaders pause. The pledge seems unlikely to solve their Trump problem, and it could create all sorts of new headaches for American democracy down the line.

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.