The challenge grows when the unreliable partner is indispensable — someone you cannot compel but must do business with.
Canadian Prime Minister Mark Carney is facing exactly this situation in his relationship with United States President Donald Trump. The U.S., Canada’s main trading partner, is indispensable but the rules of the game have become much less predictable. Even though the the Canada-U.S.-Mexico (CUSMA) trade deal is already in force, Trump imposed new tariffs on Canadian goods.
Negotiations on those new tariffs broke down even though both sides appeared to be close to a deal. According to Canadian negotiators, the Trump administration presented additional conditions at the very end of the talks. The Canadian government suspended the talks rather than accept the proposed terms.
When trust disappears
This raises a direct question for Canada: why enter into an agreement if there’s a sense the other party could go back on its word at any moment?
This situation also arises in other sectors and situations. How can a company manage a dispute when it depends on a dominant supplier? How can a parent negotiate a curfew with a defiant teenager?
When faced with such unreliable negotiators, our instinctive reaction tends to be a moral one: “He lacks credibility,” “It’s hard to trust him,” or “He’s negotiating in bad faith.” These judgments may be fair, but they’re of little help. You can’t force anyone to become trustworthy.
The useful question then becomes: Why would this person keep their word if the cost of breaking it is practically zero? Instead of wondering why the other person isn’t keeping their word, we can ask how to create incentives that will motivate them to do so.
Research in game theory and self-enforcing agreements shows that without trust, a partner’s promise matters far less than the threat of real consequences if they break it — especially when no outside authority exists to enforce the deal.
Here are three says to manage the type of situation Carney and Canada are facing:
1) Reduce dependence
The BATNA (Best Alternative to a Negotiated Agreement) was popularized in 1981 by Americans Roger Fisher, a Harvard law professor and pioneer in negotiation research, and William Ury, an anthropologist and negotiation specialist, in Getting to Yes.
Put simply, this means having a Plan B during negotiations. The better the alternative is, the more power you have in the negotiation, since you can reject a bad offer without much risk.
Take this concrete example that will sound familiar: A company that depends on a single large neighbouring market feels the full brunt of every tax or policy change for lack of other markets. The company’s partner has little to gain by moderating its stance. But that changes if the company also develops markets in Europe or Asia: the partner knows it could lose some of that business to a foreign competitor.
This is precisely the point of BATNA: we first seek to address our own vulnerabilities, not to harm our counterpart. Once you have a genuine alternative, the other party’s failure to comply will eventually cost them something, without you having to say so outright. You can simply look elsewhere, and they risk losing you if they go too far.
2) Limit risks
The second tactic addresses risk. Research has shown that co-operation can emerge and be sustained in repeated interactions, even in the absence of an external authority. In particular, repetition allows the parties to adjust their behaviour based on what they’ve come to know about each other.
The principle of “starting small” extends this logic. When a partner’s reliability is uncertain, you begin with a limited commitment and gradually increase it as the partner’s reliability is confirmed. So instead of one large agreement, break it into stages. Access to subsequent stages depends on compliance with the previous ones.
Say two countries are negotiating the removal of tariffs. They proceed in stages: steel first, then automobiles, then other sectors. Each country must fulfil its commitments from the previous stage to move on to the next. If one of the countries quickly goes back on its word, the other loses only access to the subsequent stages, which are often the most advantageous.
The risk changes. The defector isn’t necessarily punished, but their gains are limited. Their calculation becomes less attractive: it’s no longer “I can get a big advantage, then back out,” but “if I back out now, I’m giving up the even greater gains that would come later.”
This isn’t as much of a deterrent as an actual penalty, but it reduces the incentive to breach the agreement and, above all, protects the other party.
3) Attaching a cost to the breach
The third tactic goes a step further: it directly links future benefits to compliance with current commitments. It’s rooted in the literature on self-enforcing agreements. You may choose to honour an agreement not out of fear of legal sanctions, but because breaking it today results in the loss of future benefits from the relationship.
In practice, adhering to the agreement grants access to a tangible benefit — a preferential rate, privileged access or favourable terms — which is automatically renewed at short intervals, such as monthly or quarterly. If there is a breach, those benefits are simply not extended.
The key is that the benefit is linked to, but separate from, the core agreement. This makes the tactic proportionate and reversible: the perk can be withdrawn after a breach and restored once compliance resumes. Two conditions still matter: the benefit must outweigh the gains from cheating, and breaches must be observable before renewal.
Nevertheless, this consequence must be credible. Threatening a trading partner with a measure you have no interest in enforcing is likely to have little effect. You need a legitimate consequence that’s known in advance, but also realistic enough and inexpensive enough to implement so that the partner wants to uphold their end of the deal.
Consequences, not punishments
These three measures don’t work in the same way: the first gives us a better way out, the second limits what we stand to lose, and the third makes future benefits contingent on fulfilling commitments. What unites them? Each ensures that failing to honour a commitment ultimately leads to concrete consequences.
When dealing with a dominant party, their effects are real but marginal. They do not reverse the balance of power. They simply alter the cost-benefit analysis of a breach. Introducing a real consequence doesn’t eliminate the power imbalance, but it makes it less overwhelming.
Nor does this approach transform an unreliable partner into a loyal ally. Rather, it allows negotiations to continue when trust, legal enforcement or viable alternatives are lacking. In short, these are survival strategies for a relationship that cannot be left or repaired.
When dealing with a partner whose word is no longer reliable, the goal isn’t a better promise. It’s building a relationship in which keeping promises is the best option.
Jean Poitras, Professeur titulaire en gestion de conflits, HEC Montréal
This article is republished from The Conversation under a Creative Commons license. Read the original article.