Termination for Breach: What Standard to Use?

It’s commonplace for contracts to provide that a party may terminate for “material breach” by the other party. I’m not keen on that standard: as longtime readers will know, material is not only vague, it’s also ambiguous—drafters use material to mean either “of dealbreaker-level importance” or just “significant” (in other words, “nontrivial”). (If I’m using defined terms, I’d use the defined term Significant for the latter meaning; if I’m not using defined terms, I’d use nontrivial or trivial, depending on the context, as it’s clearer what level of importance is being expressed.)

For example, if in a contract Acme states that it’s in compliance with its contracts, except for breaches that aren’t material, one suspects that the drafter had in mind that in this context, material means “nontrivial.” But in a provision that permits Acme to terminate on material breach by WidgetCo of its obligations under the contract, the drafter might well have intended that Acme could terminate only if a breach by WidgetCo brings the entire transaction into question. Because of this ambiguity, I prefer not using the word material in this context.

Here’s how I’d express the “dealbreaker” meaning of material for purposes of one component of a section specifying grounds for termination:

WidgetCo breaches any of its obligations under this agreement and a reasonable person in Acme’s position would want to terminate this agreement because of those one or more breaches, whether considered individually or in the aggregate;

The problem with this standard is that if it goes to litigation, you likely would have to revisit all sorts of intangibles. For example, what did Acme expect to get out of the transaction? And a court might well be reluctant to hold that there had been a “dealbreaker” breach unless the breach is relatively dramatic in nature. That’s why in some contexts the following standard might be appropriate:

WidgetCo breaches any of its obligations under this agreement, except for breaches that individually or in the aggregate are trivial;

Under this standard, if WidgetCo screws up in a way that creates a problem for Acme that is big enough that Acme would be unreasonable to ignore it, then Acme may terminate the contract. The justification for this lower threshold would be that Acme has the bargaining power. Acme is telling WidgetCo, If you mess up, even in a way that’s less than existential, we can get rid of you.

What do you think?

By the way, you could omit any qualification—Acme may terminate for any breach, no matter how trivial. Although that would likely bump up against the implied obligation of good faith, I’d prefer not to leave it to that, whichever side I’m representing. Of course, you could mitigate the effect of termination for any breach by giving the breaching party the opportunity to “cure” any breach. But as I noted in this 2013 post, the notion of cure has problems of its own.

  • Vance_Koven

    With a statement that long and convoluted, I think you’d need to have it in a definition of “Material Breach” that leaves you free to use the word “material” in other contexts (and also allows you to apply the standard to both parties). I’m also uncomfortable with the “want to terminate” language in first statement since it basically hands Acme a blank check. I’m not sure this is much better, but my thoughts run along the lines of “breaches one or more of its obligations under the agreement such that the breach or breaches, individually or in the aggregate, deprive [the other party] of the principal benefits of the agreement.” I think this way of putting it also has the advantage of making the concepts underlying the definition correspond to the way common law courts have always regarded the concept of material breach, independent of any particular contract language.

  • AWrightBurkeMPhil

    I can’t really crack this one in the abstract, only in particulars.

    If you are going to license one or more breaches, as long as they collectively don’t hurt the nonbreaching party “too much,” you’ve got unavoidable and maybe insoluble-in-the-abstract qualitative and quantitative problems. How much, how many, what kind constitute “too much”?

    Maybe a way out is to use the “vague standard plus a cap” method that you recommend in other situations, as is “promptly but in no event more than 2 hours after such an explosion.”

    But as soon as you try to define a boundary between trivial and nontrivial, you lose hope:

    “Any breach for which the cure costs more than $10 per widget per month is nontrivial.”

    “No breach is trivial if it alone or together with other breaches leaves the Plane unready for the Big Race by the Deadline.”

    “No breach is nontrivial unless a reasonable person in Purchaser’s place would reasonably consider it nontrivial.”

    “The Purchaser may terminate only for breaches that individually or collectively deny the Purchaser one or more principal benefits to Purchaser of this agreement.” (Now we have to define the difference between “principal” and “nonprincipal” benefits.”)

    All the above stabs at clarification kick the can down the road or define one “black box” word in terms of another.

    Defining performance more precisely may be better than defining permissible breaches.

    For example, instead of saying, “a breach is trivial if it results in Quality Control’s rejecting no more than 3 cars per 100 cars off the assembly line,” say “‘Adequate Quality’ means that Quality Control rejects no more than 3 cars per 100 off the assembly line.”

    The latter way, there’s no breach at all unless rejections exceed the limit. Either way anoints Quality Control as the decision maker, moving the decision making from the jury box to the assembly line, or so one hopes.

    This problem is both knotty and basic. It highlights the generic problem that the “reasonable person” standard is by convention “objective” but also inherently vague.

    Perhaps the takeaway is “don’t use vagueness unless you know you can live with the resulting uncertainty.”

  • Uninterested Observer

    The easier thing for me is to identify up front what breaches are material for each side (you don’t pay – that’s material breach; I violate your IP rights – that’s material breach). It doesn’t necessarily take the material breach concept out completely, but it does provide some certainty that certain breaches are material breaches. Beyond that, trying to define “material” in a negotiation is usually a waste of time for me.

  • Chris Lemens


    I mostly write services agreement that don’t require the purchaser to buy any set amount. That is, they are indefinite quantity contracts. Given that they are indefinite quantity contracts, the term does not matter much to me — there is no value to be booked on signature. Additionally, our corporate culture is devoted to customer service. In that context, we generally don’t care much about termination for breach: we usually negotiate the termination for convenience provision to allow the customer to walk whenever would be reasonable.

    If I did care more about termination for breach, then I have several observations. First, Vance’s rendition seems pretty good as a general standard. And I agree with Wright that deal-specificity can overcome a lot of problems. The concern I have with all of these is that there can be breaches that are material, but are not about the things motivating the deal. For example, if you perform services that have personal information involved, you are probably focusing on service, price, quality, etc. If there is a data security lapse resulting in bad publicity, you probably want to walk away. But it might be hard to show that data security was one of the principal benefits of the deal.

    Uninterested Observer’s approach appears to be adding a statement to each section (or a list similar to the list of termination-surviving provisions found in many agreements) identifying the sections that are material. The problems there are (a) it’s a lot of work for something that will probably never be used, (b) even identifying sections probably lumps important and unimportant together, and (c) you still have the problem of the breach of a provision that appears to be immaterial but the impact of the breach is so significant that any reasonable person would want to terminate.

    For me, all that boils down to being comfortable with the vagueness of simply saying material breach and leaving it at that. If I really cared about locking in the benefits of an agreement so that very few provisions would be the keys to termination, then I would use Uninterested Observer’s approach. Specifically, I would structure the agreement so that the specific sentences that are the triggers for termination would be in a unit that I could reference separately (e.g., a subsection, or an item in an enumerated list). Then my termination provision would refer to those specifically. Thanks goodness that I don’t run into this very often.