Concept
Applying the Duty
Obligation of good faith and fair dealing
Applying the Duty
Restatement (Second) of Contracts § 205 states the duty of good faith and fair dealing in one sentence and attaches it to every contract, in both its performance and its enforcement.1 A rule that general does not tell a court what to do when one party says the other has breached it. Even agreement that good faith will be implied in every contract leaves the harder question open: what the duty requires of each party in this particular deal.2
Courts answer by putting the duty to a handful of recurring uses. The New Jersey appellate court in Seidenberg named three. The covenant may supply a term the writing omits, when the parties must have intended it because the contract needs it to work as a business arrangement. It may allow redress for bad-faith performance although the defendant violated no express term. And it may permit inquiry into how a party used discretion that the contract's own words gave it.3 Two further uses rest on the same ideas and have grown their own case law: the clause that makes performance depend on the other party's satisfaction, and the termination used to defeat benefits the other side has already earned. Each answers a different problem:
- Gap-filling terms: the writing is silent on a point the deal needs.
- Bad-faith performance: performance follows the words but defeats their purpose.
- Contractual discretion: the words grant a power but say nothing about how to use it.
- Satisfaction clauses: the power is to decide whether the other side's performance is acceptable.
- Terminating to defeat earned benefits: the power to end the relationship is used to keep the price of work already done.
The duty adds more in some uses than in others. At its most modest, it qualifies a right the text grants: a buyer's right under UCC § 2-601 to reject goods in whole or in part must be exercised in good faith.4 At its most ambitious, it creates an obligation the text never stated. One constraint runs through every use. The duty adjusts how a party performs and exercises its rights, and it cannot override an express term, a limit the New Jersey Supreme Court has repeated.5 Seidenberg adds that a court will not imply a term that conflicts with the parties' express terms.6 The limit follows from what the duty protects, the exchange the parties made, which the duty can guard but cannot rewrite.
Gap-Filling Terms
The first use is to supply a term the writing leaves out. Seidenberg confines that work to terms the parties must have intended because they are necessary to give the contract business efficacy, meaning that without them the agreement would not work as the commercial arrangement the parties plainly set out to make.7 The same cases bar any implied term that conflicts with an express term.8 Necessity makes a missing term eligible, and the writing's own words can still defeat it.
Necessity is the test because no drafter foresees every contingency, and silence on a point the arrangement needs can leave one side exposed to a result neither would have accepted at signing. Wood v. Lucy, Lady Duff-Gordon shows the pattern. Lucy gave Wood the exclusive right to license her name for dresses, hats, and fabrics, and her fashion opinions were then so much in vogue that manufacturers were expected to pay handsome fees for her endorsement.9 Wood promised only to account for his receipts and to pay her half the profits, duties that mean nothing if he does nothing. The court implied a promise to use reasonable efforts, and that implied promise supplied the consideration for Lucy's promises.10 The exclusive grant left Lucy no other way to profit from her name, and the parties plainly meant the agreement to produce profits. A term that would merely help one side would have the court improving the bargain, and the covenant protects the exchange the parties made.
Agreements of indefinite duration between a manufacturer and a distributor present a recurring gap: the contract says nothing about how to end the relationship. In Leibel v. Raynor Manufacturing Co., a Kentucky court faced an exclusive dealer-distributorship for garage doors within 50 miles of Lexington, under which the distributor had borrowed substantial sums for equipment, inventory, and working capital.11 The agreement fixed no duration, so either side could end it at will, and the court left that right in place. But it held that Article 2 of the UCC governed and that ending such a relationship requires reasonable notification, because a dealer compelled to carry a large inventory may suffer substantial damages if the relationship ends before the stock can be sold.12 The official comment to § 2-309 frames the point as good faith and sound commercial practice: ending a going relationship normally calls for notice that gives the other side reasonable time to find a substitute arrangement.13
The implied term leaves the right to end the relationship intact. An at-will distributorship still needs no good cause, and the constraint falls only on how the right is exercised. How much notice is reasonable depends on what the distributor has at stake: time to sell remaining inventory, to recoup investment made in reliance on the deal, and to find a substitute.14 A consignor with virtually no capital investment and other outlets for leftover stock needed only ninety days in Retail Associates, Inc. v. Macy's East, Inc.15 The length of notice tracks the reliance the deal invited.
The parties can also fill the gap themselves. An agreement that dispenses with notification, or limits the time for seeking a substitute arrangement, is valid unless putting it into operation would create an unconscionable state of affairs.16 Once the writing answers the question the implied term would have answered, nothing remains to fill.
The same logic keeps courts from implying terms a party merely wishes the writing contained. In Reiss v. Financial Performance Corporation, an issuer had sold warrants on three occasions, each letting the holder buy common shares at ten cents. Only one of the three included a provision adjusting the exercise price if the company later did a reverse stock split. After a reverse split, the issuer asked the court to imply the adjustment into the other two, so that their holders would not reap a windfall. The New York Court of Appeals held that the parties' rights came from their unambiguous contracts and refused to add the term.17 The adjustment may have been sensible, but its presence in one issuance and absence from the other two makes the omission look like a choice rather than an oversight. A silence the parties chose is not a gap, and a term the deal needs is not the same as a term one side now prefers.
Bad-Faith Performance
Because § 205 reaches performance as well as enforcement,19 a party can breach the duty without violating any express term. Seidenberg describes the second use of the covenant in just those terms: it allows redress for bad-faith performance of an agreement even when the defendant breached no express term.20 Gap-filling adds a promise the writing omits; this use leaves the writing alone and examines how a party carries out the promises it contains. Performance is where abuse hides, because the performing party controls details such as how much to buy, when to act, and how much effort to give. It can keep to the words of the agreement while draining the exchange of what the other side paid for.
In Sons of Thunder v. Borden, Inc., a vessel operator invested large sums in clam-fishing vessels on assurances that Borden would buy from it. Borden consistently failed to make the purchases and finally exercised an express right to end the contracts on 60 and 90 days' notice.21 The New Jersey Supreme Court held that Borden could have breached the covenant through that performance, because by never buying the required amount of clams it “destroyed Sons of Thunder's reasonable expectations and right to receive the fruits of the contract.”22 The sequence explains the result. Borden invited the investment, the vessels earned their keep only if purchases followed, and the money, once spent, was committed. Withholding the purchases removed the return the investment was made to earn. Seidenberg reads the decision as the best illustration of the covenant coming to the forefront even when the defendant acted in conformity with the express terms.23 The duty reaches performance because the operator suffered the same harm an express breach would have caused, and a rule that waited for a written term to be violated would let the party controlling performance defeat the bargain by doing nothing.
The same decision fixes the limit of the doctrine, and students most often go wrong here. The New Jersey Supreme Court agreed with Borden that the covenant could not override the express right to terminate, so Borden's later decision to end the contracts on the stated notice could not itself be attacked.24 Seidenberg repeats that the state's highest court has consistently held that the covenant cannot override an express term.25 A right to leave on notice is a term the parties bargained over: each knew the other could end the relationship after the stated period and priced the deal accordingly. A covenant that disabled the right would impose a longer commitment than either party made. What separates Borden's two acts is whether it was performing or exercising a right the contract gave it. Its failure to buy in the months before the notice was conduct in performance and could breach the covenant; the notice was the exercise of an express right and could not. An answer that treats the termination clause as disposing of the whole claim misses the first half, and an answer that treats the earlier bad-faith performance as invalidating the notice misses the second. Seidenberg also recognizes a distinct situation, bad faith serving as a pretext for exercising a right to terminate, which this chapter takes up under Terminating to Defeat Earned Benefits.26
Feld v. Henry S. Levy & Sons, Inc. draws the same line from the other direction. A bakery agreed to sell all its bread crumbs at 6 cents a bag under a one-year contract that renewed annually and could be cancelled by either party on six months' notice. When it decided the price was very uneconomical, it dismantled its crumb oven and stopped production without formal notice, and the New York Court of Appeals plainly regarded that conduct as lacking good faith. Crumbs were a minor part of a bakery that carried on as before, and earning less than expected, even earning nothing, did not justify stopping. The bakery's protection lay in its cancellation right, which the contract included so that either side could end its obligations in an orderly way. Beside Sons of Thunder, the case shows one rule from two sides: a contractual exit is protected when a party uses it, and a party that skips the exit and simply stops performing cannot rely on the fact that the exit existed.
Inattention can breach the duty too. In Bloor v. Falstaff Brewing Corp., a brewer that had bought the Ballantine brands owed a royalty on each barrel sold, and it let their volume fall because that course suited its overall profits. The court held that the brewer owed a good-faith effort to see that substantial sales were made.
Sons of Thunder stressed the operator's unequal bargaining power, but Seidenberg explains that disparate strength is sometimes prominent without being essential; it is one factor among many.27
Contractual Discretion
Many contracts give one party a decision that shapes what the other receives: whether to approve a submission, set a price, consent to a transfer, or settle a claim. Exercising a power the contract confers is part of performing the contract, so § 205 governs the exercise.29 That makes discretion the third main setting for the covenant. Where a contract gives one party a discretionary power affecting the other's rights, the holder must exercise it in good faith and with fair dealing.30 The measure is the other party's reasonable expectations at the time of contracting, and the limit applies unless the parties carefully negated it in express contract terms.31
Courts read a limit into nearly every grant of discretion because a grant promises only that the holder will decide. It cannot sensibly promise that the holder may decide however it pleases, because that would commit the holder to nothing and leave the other side's rights to its whim. In most commercial dealings, apparently unlimited discretion is therefore curtailed by an implied obligation to exercise it reasonably or in good faith, and that obligation keeps the holder's promise from being illusory.32 The limit has two parts. Because it leaves the decision with the party the contract chose, a court does not substitute its own judgment. Because it ties the power to the purpose for which it was granted, the holder cannot use the power to take back the fruits of the contract the other side was promised.33 When the words make unconstrained discretion unmistakable, a court honors them, so the first question in any discretion problem is what the grant actually says.34
Tymshare, Inc. v. Covell shows how courts read the grant itself. A sales-commission plan let the employer adjust quotas in its “sole discretion,” and then-Judge Scalia refused to assume that sales representatives had agreed to a clause allowing any change for any reason. He read the clause as discretion to decide whether the factors that would reasonably justify changing a quota existed, such as an unexpected volume of business from one customer or a poor sales year for the company. He refused to read it as a license to raise the quota from a simple desire to deprive an employee of the fairly agreed benefit of his labors.35 A clause that said outright the employer could change the quota solely to reduce earned commissions would leave little room for that argument, although unconscionability might still be raised.36 The argument rests on what a reasonable employee would have understood the clause to allow, so clearer words leave less to interpret.
Locke v. Warner Bros. separates a protected decision from a protected pretense. The studio's agreement gave it discretion to approve or reject each of the director's proposals, and the trial court granted summary judgment on the view that a studio's artistic and business choices are not proper subjects of judicial review. The appellate court held that the ruling missed a distinction: the studio may make a subjective creative decision that a court will not review for reasonableness, but its dissatisfaction must be bona fide.37 Testimony that studio executives had said the studio would not work with her, whatever she proposed, raised a triable question whether the studio was honestly dissatisfied with her projects or merely went through the motions of considering them.38 A trier of fact could also reasonably infer that the studio never intended to give her proposals a good-faith evaluation.39 Because the case reached the appellate court on summary judgment, the holding is that the evidence required a trial; the court did not decide that the studio breached. The studio answered that the covenant cannot create obligations the contract never contemplated, and the court rejected the point because she asked only that the studio's admitted discretion be exercised honestly.40 A studio that reads every proposal, concludes honestly that none deserves development, and declines has used its discretion as the contract allowed, however a court might rate its taste.41 The outward result, no film, is the same in both situations. What differs is whether a genuine evaluation produced it.
Dalton v. Educational Testing Service applies the same line to a testing contract. The test-taker had agreed to a registration bulletin that let the service cancel a score it had reason to doubt, after telling him why and offering several options, including the chance to submit more information. The service questioned his November SAT score because of a handwriting disparity, and he responded with statements from a service employee and two students who remembered him at the test, plus a document examiner's report. The trial court found that the service made no rudimentary effort to evaluate any of it.42 The New York Court of Appeals explained that where a contract contemplates the exercise of discretion, the covenant includes a promise not to act arbitrarily or irrationally in exercising it.43 Two consequences followed. If the service considers the relevant material and otherwise acts in good faith, the service, not the court, decides how much weight the material deserves, and a court intervenes only if the decision is arbitrary or irrational. If the service refuses to exercise its discretion at all by declining even to consider relevant material, the question becomes whether that refusal breached an express or implied term, and the court held that it did.44 The remedy was specific performance in the form of a duty to consider the information, and the court refused to order release of the score.45 The decision also marks a boundary: nothing in the bulletin required the service to investigate in the field, and an implied duty to do so would have conflicted with the express terms placing the burden on the test-taker.46 Dalton and Locke both turn on the difference between a judgment made and a judgment never made. In Dalton the trial court found that no evaluation occurred; in Locke the refusal to evaluate was an inference still for the jury.
Approval clauses in leases test the purpose limit. In Kendall v. Ernest Pestana, Inc., a lease barred assignment without the lessor's prior written consent and said nothing about reasonableness. The tenant proposed to assign to buyers with a stronger financial statement who were willing to be bound by the lease, and the lessor refused while demanding increased rent and more onerous terms.47 The California Supreme Court held that because the lessor kept discretionary power to approve or disapprove a proposed assignee, the power must be exercised in accordance with commercially reasonable standards.48 Relevant considerations include the assignee's financial responsibility, the suitability and legality of the proposed use, the need for alterations, and the nature of the occupancy. Denying consent out of personal taste, or in order to charge a higher rent than the lease provided, is not commercially reasonable.49 The reason is a purpose test. An approval clause exists to protect the lessor's interest in the preservation of the property and the performance of the lease covenants, and a wish for a better bargain has nothing to do with those interests.50 A lessor who withholds consent for leverage is trying to recapture an opportunity the lease, as signed, foreclosed, which is the pattern Professor Burton identifies as bad faith.51 The court described its rule as the minority position, and it matches the Restatement of Property rule that consent may not be unreasonably withheld unless a freely negotiated provision gives the landlord an absolute right.52 A franchise dispute follows the same logic. A franchisor whose approval was required for relocation refused both a relocation request and any communication with a prospective buyer while secretly wanting to replace the franchisee with a company-owned store, and the court held that these refusals frustrated the franchisee's reasonable expectations and breached the covenant.53
Crisci v. Security Insurance Co. supplies the setting where the yardstick is stated most explicitly. A liability insurer controls whether to settle a claim against its insured, and the California Supreme Court held that the covenant requires the insurer, in deciding whether to settle, to give the insured's interests at least as much consideration as its own. The test is whether a prudent insurer without policy limits would have accepted the offer, and liability does not depend on proof of dishonesty, fraud, or concealment, because the covenant includes a duty to accept reasonable settlements.54 The test removes a conflict built into the arrangement. The insurer's own payment stops at the policy limit, so refusing a settlement within that limit lets it gamble on the verdict while any excess falls on the insured. In Crisci the underlying claim was for $400,000 against a $10,000 policy, and verdicts of $100,000 and $1,000 left the insured owing $91,000 beyond the insurer's payment.55
The yardsticks in these cases differ, and they roughly track the kind of judgment the contract delegated. Locke asks for honest dissatisfaction, because creative taste has no outside measure. Kendall asks for commercial reasonableness, because lease approvals are commercial decisions with identifiable purposes. Dalton forbids the arbitrary or irrational, and Crisci applies a prudent-insurer measure. Courts have not settled on a single choice between an honesty standard and a reasonableness standard, so a discretion problem begins with the jurisdiction's cases on that kind of grant.56 Under any of them, a discretionary decision that leaves the other party worse off has no legal significance without bad motive or intention, which keeps ordinary self-interested choices outside the duty.57
The same implied limit explains why a contract that leaves particulars of performance to one party can still be enforced. A term is not too indefinite merely because one party specifies performance, since the exercise of that power carries implied limits such as reasonableness or good faith, and those limits give a court a standard to apply. How much definiteness a court demands varies with the subject matter; land contracts require more than employment or construction contracts.58 Compare a clause that defers agreement on a material term. In Joseph Martin Jr. Delicatessen, Inc. v. Schumacher, a lease option to renew at an “annual rental to be agreed upon” was an unenforceable agreement to agree, because the parties had not agreed on the rent and had merely postponed agreement to a later date.59 No one held power to fix the term, so the court had no standard to apply. In Arbitron, Inc. v. Tralyn Broadcasting, Inc., a licensor could raise a license fee if the licensee acquired more radio stations, but the contract stated no amount and no formula. The Second Circuit held the clause not too vague because the licensor had to set the increase in good faith. Enforceability and liability remain separate questions: the implied limit lets the contract stand, and whether the licensor used its power in good faith turns on what it actually did. Control over price drew the same scrutiny in Nanakuli, where the court of appeals reinstated a jury verdict that Shell breached its contract by failing to provide protection for Nanakuli in 1974.60
Courts disagree about how far the duty reaches into rights a contract states expressly, and lender cases show the split. One line of authority, represented by K.M.C., limited a lender's ability to terminate a financing agreement and demand repayment from a grocery business, especially since the lender also controlled the business's bank account. Many later decisions criticized K.M.C. for effectively overriding the contract's express terms.61 The Seventh Circuit in Reger Development, LLC v. National City Bank held that the covenant cannot limit express contract rights, and the official comment to UCC § 1-309 supports that view by saying the obligation of good faith “has no application to demand instruments . . . whose very nature permits call at any time with or without reason.”62 A second line involves home lenders that exercise an express right to buy hazard insurance and charge the cost to the borrower; borrowers alleging inflated prices and kickbacks have in some federal district courts survived motions to dismiss on a covenant claim.63 The dispute is over how much an express right gives way, and the next section takes up the answer: a party expressly given the right to engage in particular conduct cannot be held in breach for engaging in it, and implied terms are not read to vary express ones.64
Whether the parties may go further and contract around the duty depends on state law. In a dispute over an airline's frequent flyer program, the Supreme Court held a state-law implied-covenant claim preempted and explained that such rules escape preemption only if the law of the relevant State lets an airline contract around them. An airline whose agreement is governed by such a State's law can specify that the agreement does not incorporate the covenant. The provision may impose transaction costs and make the program less attractive, but the airline decides whether the benefit is worth the cost.65
Satisfaction Clauses
A satisfaction clause makes one party's duty to accept or pay depend on that party's own approval of the other's performance. Read literally, it looks like a blank check, because a buyer who must pay only if satisfied could always announce that it is not. That freedom would make the buyer's promise illusory: the buyer would be committed to nothing and would supply no consideration. Courts therefore refuse to read the clause as unchecked power. Unless the only plausible reading of the agreement is that the parties meant to give the judging party unrestrained discretion, a court implies a limit that keeps the judgment either honest or reasonable.67 Section 205 supplies the limit, because the judging party exercises its judgment in the course of performing.68
Which limit applies is the harder question. Everyone agrees the judging party must act honestly; the open issue is whether the law demands more, such as reasonableness. Courts in practice choose between two standards. Under the objective standard, the court asks whether a reasonable person in the obligor's position would have been satisfied. Under the subjective standard, the obligor's honest dissatisfaction is enough, however unreasonable it looks to outsiders.69
The Seventh Circuit drew the dividing line by the kind of performance being judged in Morin Building Products Co. v. Baystone Construction, Inc. The reasonable-person standard applies when the contract involves commercial quality, operative fitness, or mechanical utility that other knowledgeable people can judge. The good-faith standard, meaning honest dissatisfaction, applies when it involves personal aesthetics or fancy.70 Judge Posner noted that some cases accept honest dissatisfaction as the only test even for utilitarian work, but he observed that most follow § 228 of the Restatement (Second), which prefers an interpretation under which the condition is met if a reasonable person in the obligor's position would be satisfied, whenever it is practicable to decide that.71
The objective standard is the default for a reason that has nothing to do with protecting the weaker side. Courts read reasonableness into the contract to approximate what the parties would have provided had they foreseen the dispute, and the Morin court disclaimed any paternalistic purpose even though both parties were substantial commercial enterprises.72 A seller that has built or shipped something would ordinarily not want its payment left to the buyer's whim, and where a yardstick exists outside the buyer's head, the court can enforce that shared expectation. Where performance turns on personal taste, no such yardstick exists, and the parties are presumed to have wanted the buyer's own honest response. The choice of standard follows the parties' likely intentions in each setting.73 The presumption only approximates intent, so clear contrary language or circumstances can displace it.74 The objective default also echoes a wider preference in contract law for judging parties by externals instead of private states of mind, a preference Holmes defended on grounds of efficiency and fairness.75
Judge Posner marked the two ends of the range with examples. A steel-plant manager who rejects a shipment of pig iron because the pigs are not prettily shaped must still be reasonable, even under a clause saying acceptability rests strictly with the owner, since it would be fantastic to think the iron supplier meant to expose its contract rights to the buyer's agent's whim. A buyer who reserves the right to reject a commissioned portrait, by contrast, wants a portrait that pleases the buyer rather than a jury, even a jury of connoisseurs, so the only question is the buyer's good faith.76 Plumbing installed in a mountain cabin falls on the iron side of that line, and a portrait of the promisor's spouse falls on the other.
Morin itself sat between those poles. Baystone, the general contractor on an addition to a General Motors plant, hired Morin to supply and erect mill-finish aluminum walls. The contract made the work subject to the final approval of the owner's authorized agent, whose decisions on artistic effect and on the quality or fitness of materials were declared final. Viewed in bright sunlight from an acute angle, the siding did not look uniformly finished, General Motors' representative rejected it, and Baystone replaced it and withheld the $23,000 balance. The trial court told the jury to ask whether the owner, as a reasonable person, should have been satisfied, and Morin won. The only issue on appeal was whether that instruction was correct, so everything depended on which standard governed.
The court first asked what kind of performance was at stake. Appearance sounds like personal aesthetics, and the contract even used the words artistic effect, but the building was a factory, a structure not usually meant to be a thing of beauty, and the contract specified unpainted mill-finish aluminum, which in the trade does not have a uniform finish from sheet to sheet. An owner who cared about uniformity could have ordered painted siding. Whether Morin's siding achieved a reasonable uniformity amounting to satisfactory commercial quality could be judged objectively, so a reasonableness standard was practicable.77
That classification showed only what would be read into an ordinary satisfaction clause, and the touchstone is the parties' intent, so the court turned next to the words they actually used. The artistic-effect language was item 17 in a list of conditions in a general-purpose form, followed by a qualifier limiting it to matters within the terms of the contract documents, which suggested it applied only to contracts in which artistic effect was among the buyer's aims. The clause on quality and fitness was likewise qualified by a first-class-work requirement and had been incorporated from a different form contract. The court did not disparage form contracts, but it was left with more than a suspicion that these clauses were never meant to cover the aesthetics of a mill-finish factory wall.78
The court added that Morin might prevail even under the minority view, because that view presupposes a contract that conditions acceptance on the buyer's satisfaction in the very respect in which the buyer is dissatisfied.79 Its reasoning about price shows why the objective reading fits the deal. A perfectly uniform mill finish is difficult, perhaps impossible, to achieve, so a supplier who had agreed to meet whatever perfection the owner's agent demanded would have risked rejection and would have demanded a higher price, and the owner would have paid a premium for a freedom it could not have valued highly, since its objective was not aesthetic. The court affirmed, while stating that if the language or circumstances had shown that the parties meant the owner to reject work for failing a private aesthetic taste, the rejection would have been proper even if unreasonable.80
The same words call for different standards because the standard follows what the buyer is buying and what the circumstances show the parties meant. Mattei v. Hopper applies the same purpose-driven reading to the illusory-promise problem. A buyer promised to purchase land if the leases he obtained were satisfactory to him, and the California Supreme Court held that the promise was genuine because his judgment of satisfaction was subject to a good-faith obligation.81 Read against the buyer's purpose of building a shopping center, the clause worked as a financing contingency, so “satisfactory” pointed to whether a construction lender would accept the leases and not to the buyer's private taste.82
A different rule governs when the person whose satisfaction is required is an independent third party such as an architect or an engineer. Where no forfeiture is involved, as in an ordinary sale of goods, an express condition of third-party approval is enforced literally and is usually not excused even by the third party's death or incapacity.83 Courts also tolerate a subjective test more readily in this setting, on the assumption that a third party is less likely to be swayed by the obligor's selfish interests.84 In Hanscom v. Gregorie, buyers were freed from their purchase obligation when their building inspector found the building was not free from defects of a substantial nature, even though the trial court concluded the defects were not substantial, because the contract left that question to the honest judgment of the buyers' expert.85 The tolerance still rests on honesty. An evaluator who announces dissatisfaction at the obligor's urging, in collusion with it, is not honestly dissatisfied, so the condition is excused, and the obligor cannot use the third party as a proxy for the unchecked discretion the law will not allow.86
Choosing a standard sets the burden and leaves the outcome open. A plaintiff facing the honest-dissatisfaction test carries a difficult burden that it can still meet, as in Forman v. Benson, where a seller's rejection of a buyer's credit report was held pretextual because it was an attempt to renegotiate price and interest rate. A plaintiff who wins the reasonableness standard can still lose, as in Kennedy Associates, Inc. v. Fischer, where a mortgage lender could reasonably disapprove a property after the agreed inspection.87 The pretext principle in the first case, the one developed earlier through the portrait example, is the floor under the honest-dissatisfaction test, and it keeps that test from becoming a blank check.
Terminating to Defeat Earned Benefits
Section 205 attaches the duty to every contract, including an employment agreement.89 Applying it there is delicate, because employment of indefinite duration has traditionally been at will, with little redress for an employee who is arbitrarily treated unless a union or an explicit employment contract protects him.90 A duty that forbade bad reasons for firing would seem to swallow that presumption. Courts that apply the covenant to at-will employment have therefore left the freedom to end the relationship in place and have policed one use of it: ending the relationship to deprive the employee of compensation fairly earned or nearly earned. The leading decision is Fortune v. National Cash Register Co. A sales representative under a written at-will agreement was entitled to a substantial bonus commission only if he stayed employed, and he was dismissed the day after the employer received a $5 million order he had procured. The court upheld his contract claim on the jury's finding that the dismissal was in bad faith.91
Fortune accepted the employer's side of the at-will bargain. The court acknowledged that an employer may act on its own legitimate business interests, needs wide latitude in deciding whom to employ, and needs flexibility as circumstances change.92 What it required was narrower: where commissions are paid for work the employee performed, “the employer's decision to terminate its at will employee should be made in good faith.”93 A principal who ends the relationship while the agent is on the brink of completing a sale, in order to deprive the agent of all compensation, acts in bad faith, and the result is the same when the principal tries to defeat any portion of a commission that is due. Courts have used that rule to prevent employees from forfeiting benefits they had almost earned through substantial services.94
The covenant targets this device because salary and commissions differ in timing. An employee paid as the work is done loses nothing the work had already earned when the employer ends the relationship, even on a whim, because performance and payment occur together and neither side is left high and dry.95 A commission reverses that sequence: the employee sells first and collects later, and the gap gives the employer a chance to keep the benefit of the work and avoid its price simply by ending the relationship. A reading of the contract that allowed that result would make one party's performance the cause of the other's nonperformance, so the law implies the covenant where it is necessary for a party to receive the benefits promised for performance.96 Gram v. Liberty Mutual Insurance Co. states the same limit from the employee's side: an employer that discharges an employee must answer for the loss of compensation clearly related to the employee's past service, such as commissions that service generated.97
Geysen, decided by Connecticut's highest court, states both sides of the rule in one opinion. The employee managed the marketing of security services under a sales incentive plan that paid commissions only once work had been performed and invoiced and that ended commissions on termination except for amounts invoiced before the termination date.98 The company ended his employment by memorandum, citing an investigation into improper business activities, and he alleged that its reasons were false and a pretext for nonpayment of owed commissions.99 The court held that an employer does not act in bad faith solely by refusing commissions on sales invoiced after termination when an express contract term provides for that result.100 It also held that action or inaction meant to avoid the spirit of the bargain, or that shows a dishonest purpose, violates the covenant as it applies to commission provisions, and it recognized a claim where the termination was done with the intent to avoid paying commissions.101 The plan's cutoff therefore answers what an honest employer owes after a termination. It gives no license to use a termination as a device for avoiding commissions the employee was about to earn. The next section takes up why an express term of this kind otherwise controls.
The procedural posture limits what Geysen decided. The trial court had struck the covenant count, and the Connecticut court reversed because a motion to strike takes the complaint's facts as alleged, and the allegations, read in the employee's favor, stated a legally sufficient claim.102 The ruling let the claim go forward and left open whether the company actually ended his employment to avoid commissions, a question on which the company's stated reason, the investigation, would be weighed against his allegation of pretext. The court also explained that an employee who had not “earned” commissions under the plan's terms may still be “owed” them if the employer's breach of the covenant is what prevented him from earning them.103 That distinction lets the covenant reach commissions the plan itself would not yet count as earned, and it is the part of the doctrine most open to challenge.104
The claim cannot be built on the absence of good cause. An earlier Connecticut decision, Magnan v. Anaconda Industries, Inc., declined to transform the covenant into an implied condition that an employee may be dismissed only for good cause. It left open whether the covenant reaches a discharge motivated by an intent to deprive an employee of clearly identifiable compensation for past services, a question Geysen later answered yes.105 The bad faith that supports a claim means more than negligence and involves a dishonest purpose, the improper-motive strand taught earlier in this chapter.106 A termination that is unwise, unfair, or even whimsical breaches nothing so long as it does not target earned compensation, because the covenant does not convert at-will employment into just-cause employment.107
Proof of the purpose usually comes from the evidence that tests honesty in fact: whether the stated reason was the real one, and what the timing of the decision suggests. Delaware's approach illustrates the point, because a breach there is established by proving that the employer created false grounds and a fictitious basis for terminating an at-will employee, and not by the termination itself.108 The device need not be a firing, either. In Wilson v. Career Education Corp., the Seventh Circuit recognized that an employer may breach the covenant by using its discretion to end a bonus plan with the intent to deprive an employee of a commission already earned.109
Timing and purpose carry the analysis in both scenes. The first recruiter was dismissed when the work was done and the payment was close, the stated reason had no support, and the owner's remark about not wanting to pay supplies the dishonest purpose, so a factfinder can conclude that the termination was the device for keeping the price of completed work. The second recruiter lost only the prospect of future commissions that an at-will relationship never promised, and the agency honored the plan's cutoff as written. The usual mistakes run in opposite directions: treating at-will status as a complete defense to the first recruiter's claim, and treating the covenant as a demand that the agency justify the second recruiter's dismissal.
Jurisdictions differ on how far the covenant reaches into at-will employment, so the first step on an exam is to identify the local rule. Some courts hold that no independent good-faith claim exists when employment is terminable for no reason, as the Seventh Circuit reasoned in LaScola v. U.S. Sprint Communications, and a federal court applying New York law has described it as well settled there that no covenant attaches to at-will employment contracts.110 Even so, a New York appellate decision allowed an at-will employee to state a claim that a termination was specifically designed to cut off commissions that were coming due, which shows the cases are not uniform.111 Other courts apply the covenant to at-will employment where the employer's conduct was deceitful or fraudulent in some sense.112 Most states that accept the covenant do so where the employer acted with the purpose of denying the employee benefits already accrued through the employment, which is the commissions rule taught here, and Alaska and Wyoming go furthest by implying the covenant in all employment contracts.113 Alaska tests the employer's conduct for subjectively improper motive and for treatment a reasonable person would regard as unfair, while Wyoming allows tort damages only where a special relationship of trust and reliance exists.114
The covenant claim must be kept apart from wrongful discharge in violation of public policy, which is a narrower tort. A covenant claim asks whether the employer defeated the employee's reasonable expectations of compensation, and the termination is only the means by which that breach was carried out. Because the claim sounds in contract, an employee whose termination was a bad-faith interference with securing commissions recovers in contract.115 The public-policy tort requires a discharge for a reason that violates an important and clearly articulated public policy, and courts read that exception narrowly. The two counts in Geysen came out differently for that reason: the covenant count was reinstated, while the public-policy count stayed struck because the wage statutes protect earned wages and give no remedy for unearned wages that were not due.116 A worker fired for refusing to commit perjury at the employer's request presents the standard public-policy claim, and nothing in it depends on commissions.117 Most courts have also declined to carry the tort remedies available against insurers who breach the covenant over into employment, and the same caution marks ordinary commercial contracts, where tort remedies are reserved for special relationships.118