In short, termination standards from the nonunion market, responding largely to the exit of young workers, may not
reflect the aggregate costs and benefits as well as the union collective--choice mechanism, which reflects the preferences of
average workers. On the other hand, also unwarranted is the bolder statement that the collective choice outcome is clearly
a superior way of toting up the costs and benefits. As with many analyses, showing the flaws of a market mechanism does
not, by itself, demonstrate the superiority of other solutions. In this case the problems with union median--voter solutions
are well known. n88
Ultimately, then, an appeal to empiricism cannot provide a universal answer to whether at will or just cause is the
optimal standard. Each standard has costs and benefits, and the balance may shift depending on the situation. Those who
debate the issue too frequently ask whether at will or just cause should be the presumption in all employment relationships.
A universal answer cannot be found. The next section attempts to break down the inquiry into parts of the life cycle, and
to explain the inherent logic of the common law solution.
III. Legal Supervision of Opportunistic Firings
It is time to examine the employment termination cases. In explaining the common law approach, I will first show
that courts have attempted to police opportunistic firings by employers. I will then attempt to fit their efforts into the
A. Ad Hoc Judicial Scrutiny of Employer Opportunism
A party to a long--term relationship is most vulnerable to opportunism when the party has substantially performed
while the other side has not. n89 The other side, having received most of its benefit, has an [*33] incentive to terminate
the relationship to avoid paying out its side of the bargain. Contract doctrine generally imposes a duty of good faith upon
parties, in part to protect against such opportunistic behavior. Section 1--203 of the Uniform Commercial Code (U.C.C.),
for example, mandates good faith in commercial transactions. n90 While good faith defies simple definition, n91 for
our purposes an appropriate meaning is that a party to the contract cannot deprive the other party of the benefit of the
Courts have hesitated in imposing a generalized good--faith standard upon at--will employment relationships. The
courts fear that a nebulous legal standard will make the delicate relationship too rigid or legalistic. n93 Nevertheless,
some courts have made good--faith inroads [*34] on at--will employment. Although the courts rarely articulate it in
these terms, they tend to make these inroads when investment asymmetries make the employee particularly vulnerable to
A leading example of a good--faith inroad is Fortune v. National Cash Register Co., n94 in which the court imposed
a good--faith limitation on an employer's ability to fire a salesperson about to receive a commission. Orville Fortune wasa
salesperson under a written at--will contract that entitled him to a bonus for equipment orders placed in his territory and
an additional bonus when the equipment was installed in his territory. One day after Fortune signed a $5 million order,
National Cash Register (NCR) fired him, depriving him of the installation part of the commission. n95 The Massachusetts
Supreme Judicial Court, in upholding a jury verdict for Fortune, recognized that NCR had paid Fortune all the bonus to
which he was entitled under the express contract. The court found NCR had breached, however, an implied covenant to
act in good faith. n96
Richard Epstein has attacked the Fortune decision as "wrong in principle." n97 He points to the important fact that
NCR did not keep the disputed commission for itself, but rather paid it to an installations employee. n98 Epstein argues,
therefore, that "the case is not simply one where a strategically timed firing allowed the company to deprive a dismissed
employee of the benefits due him upon completion of per [*35] formance." n99 Rather, "the contractual provisions
concerning commissions represent a rough effort to match payment with performance where the labor of more than one
individual was necessary to close the sale." n100
We should heed Epstein's warning against applying the good--faith principle without carefully considering the
incentive and bonding devices the parties have created. Perhaps the Fortune decision is wrong in application. By paying
all of the commission to some employee or other, the company avoided the temptation of an opportunistic termination.
But the Fortune case is not wrong in principle. The valid principle, for which Fortune is usually read to stand, is that courts
should scrutinize opportunistic firings in which the employee has largely performed his side of the bargain but has yet to
reap his reward. Salespeople on commission are classic examples of persons who can find themselves in that situation.
Another classic example of a good--faith analysis occurs when an employer fires an employee shortly before he
qualifies for a pension. Again, the usual self--interest check on an employer's decision to fire a productive employee is
missing. An employer can save itself considerable money by exercising its at--will discretion just before a pension vests.
A routine practice of firing all employees before their pensions vest might have severe reputation costs. But the occasional
firing for this reason might be less damaging. The issue arose in Ingersoll--Rand Co. v. McClendon, n101 in which an
employee with nine and three--quarters years of service was fired four months before his pension was to vest. Such a firing
clearly violates the good--faith notion that an employer must give the employee the benefit of the bargained--for pension.
The Texas Supreme Court allowed a common law tort action for wrongful discharge in violation of public policy. The
U.S. Supreme Court did not disagree with the underlying rationale but reversed the Texas Supreme Court. It held that
the employee must bring this claim under section 510 of the Employee Retirement Income Security Act (ERISA), which
clearly prohibits an employer from discharging a worker "for the purpose of interfering with the attainment of any right
to which the worker may become entitled under the pension plan." n102 [*36]
Employer opportunism can occasionally surface when a worker quits rather than is fired. Again, the law must
decide whether to regulate opportunism. An example is Jordan v. Duff and Phelps, Inc., n103 a well--known case due
to the disagreement between two law--and--economics--oriented judges about the application of the good--faith duty to
employment at will. The majority opinion by Judge Easterbrook emphasized that courts should scrutinize situations in
which opportunism is a danger. Jordan was an at--will employee of a close corporation who acquired one percent of the
company's stock as part of his compensation plan. When Jordan resigned, he sold his stock for book value, as required
under the plan. The company's chairman accepted his resignation without telling him of a possible merger that would
increase his share's value from $23,000 to over $600,000. When Jordan learned of the merger, he filed a 10b--5 action
alleging fraud in the purchase of corporate securities. n104 The majority opinion by Judge Easterbrook reversed a
summary judgment for the employer, distinguishing between an employer who is "thoughtless, nasty, and mistaken" from
one who engages in "avowedly opportunistic conduct." n105 While the employer perhaps could have fired Jordan for
any reason, n106 reasoned Judge Easterbrook, it could not cash out the stock option on the eve of its appreciation. n107
Judge Posner, in dissent, noted that "the possibility that corporations will exploit their junior executives ... may well be
the least urgent problem facing our nation." n108 Judge Posner argued that business executives would rather rely on their
employer's good will and interest in reputation, and on their own bargaining power, than "pay for contract rights that are
difficult and costly to enforce." n109
Although leading examples of recent trends, the cases discussed so far do not illustrate the general hesitancy of courts
to adopt an across-- [*37] the--board good--faith standard for termination of employment. This hesitancy is clearest in
New York, where the highest court has declared that any good--faith requirement is inconsistent with employment at
will. In Murphy v. American Home Products Corp., n110 a company fired an accountant after he told top management
that corporate officers had illegally manipulated the accounts of secret pension reserves. Conceding he was an at--will
employee, Murphy complained that the employer breached the covenant of good faith by firing him simply for doing his
job - which was, after all, to disclose accounting improprieties. The court of appeals rejected his claim, reasoning that the
good--faith covenant was inconsistent with the employer's unfettered right to terminate employment at any time. n111
The Murphy holding is consistent with the asymmetric--investment rationale for policing terminations. Admittedly,
Murphy was between a rock and a hard place: be fired for being a poor internal auditor or be fired for being a vigorous
internal auditor. The firing seems arbitrary and unfair, but it is not an example of employer opportunism. n112 The
company will suffer if it wrongly chooses to smooth internal politics at the cost of chilling vigorous internal audits.
Society suffers no injury other than that borne by the company. n113 Because the dangers of contract opportunism and
third--party effects are absent, courts do not need to police the situation. To intervene in a situation like Murphy would
leave no room for employment at will.
Even in Massachusetts, from which the seminal Fortune n114 case on good faith originated, courts have been
careful to limit good--faith protection to situations of asymmetric investments. An employee has no [*38] claim for a
"bad faith" termination unless the employer intended "to benefit financially at the employee's expense, such as for the
purpose of retaining for itself sales commissions or pension benefits which would otherwise be due to the employee."
n115 For example, in one case employees sued when the employer changed its evaluation policy to a "Bell Curve," thus
assuring that some employees would receive low ratings and be constructively discharged. n116 The Supreme Judicial
Court found no violation of the good--faith standard, reasoning that the employees were merely complaining they had lost
the opportunity for "future compensation for future services" n117 rather than being denied compensation "specifically
related to a particular past service." n118 Only the latter termination is a form of opportunism, whereby the employer is
firing an employee who has largely performed his side of the bargain without receiving benefits.
B. Regulating Opportunism Over the Life Cycle
These cases represent classic but ad hoc examples of courts protecting employees against opportunistic terminations.
In Fortune an employee was fired before a commission came due. In Ingersoll--Rand an employee was fired before a
pension vested. In Jordan an employer repurchased a terminated employee's stock just before it enormously increased
in value. One can easily justify legal intervention on good--faith grounds. Employees can invest in the relationship more
freely, making the relationship more valuable to both sides, when courts are available to ensure that employers will
not exploit their investments. Likewise, Murphy illustrates the limits of a good--faith analysis. Without the potential for
opportunism, courts should be reluctant to intercede in the relationship.
Unfortunately, an ad hoc "opportunism" test is unsatisfying for several reasons. Courts may have difficulty identifying
opportunism when they see it. Even if courts can identify opportunism after the fact, an ad hoc test gives limited
prospective guidance to employers and employees. Finally, the amorphous nature of an ad hoc opportu [*39] nism test
may mean it becomes so broadly applied that it is indistinguishable from a just--cause requirement for all terminations.
This would weaken the deterrence of employee shirking - the prime rationale for employment at will.
If we recall the life cycle of the career employee, we can identify a more systematic pattern of legal intervention. Over
the life cycle of a career employee, a sequence of possibilities for opportunism exists. A career employee is particularly
vulnerable to opportunism at the beginning and end of his career. By contrast, employers are especially vulnerable to
opportunism at the employee's midcareer. The cases suggest that courts are sensitive to this life cycle. Courts are most
likely to scrutinize firings at the beginning and end of the life cycle. Courts do not get involved during midcareer unless
they see an obvious case of particular opportunism, such as a firing before a pension vests or a sales commission is due.
1. Beginning--Career Opportunism
Employees face a risk of opportunistic termination at the beginning of the life cycle. The risk arises because employees
commit irretrievable investments to the relationship before the employer does. n119 Usually the beginning--career cases
involve employees who have moved to take a job or quit another job in reliance on a job offer.
Grouse v. Group Health Plan, Inc. n120 provides an example of a court protecting a beginning--career employee
from opportunistic termination. A drugstore pharmacist resigned with two weeks notice in reliance on a job offer from a
health clinic. When he called to begin work at the clinic, however, the employer told him that it had filled the position.
The pharmacist was unemployed for some time. The court allowed the employee to recover under a promissory estoppel
theory. n121 It determined that the employee had reasonably relied on the job offer and that justice required that the
court hold the employer to its promise. The employer had argued it would be incongruous to [*40] provide a remedy
for firing someone the day before work begins when the employee, being at will, would have no remedy for being fired
the day after work began. The court agreed that such a result would be incongruous, but it resolvedthis contradiction by
suggesting that an employee might also have a reliance claim if the employer fired him shortly after beginning work.
Other courts have used a theory of additional consideration to give employees who moved to a new job a reasonable
time to recoup their investment before being arbitrarily dismissed. While many courts hold that merely working is
insufficient consideration to make a just--cause promise enforceable, some additional detriment to the employee or benefit
to the employer may lead to an enforceable promise. In Veno v. Meredith, n122 for example, a newspaper fired an
editor eight years after he quit a prior job and moved from Newark to Pennsylvania to accept a position. The court, citing
Corbin on Contracts, n123 declared that an employer could not arbitrarily discharge an employee for a reasonable
time commensurate with the hardship the employee had endured. n124 The court upheld a directed verdict against the
employee, however, declaring that after eight years the reasonable length of time "had surely passed." n125 In denying
the employee's claim, the court distinguished a prior case that upheld an employee's verdict for breach of a "permanent"
employment contract when he was fired three days into a job after moving from New York to Philadelphia. n126
Some courts have held that relocating or leaving secure jobs is evidence that the parties must have agreed on a fixed--
term contract rather than at--will employment. In Lanier v. Alenco, n127 a worker "with a wife and four children left a
secure and well--paying position with General Electric, a position that he had held for eleven years," to [*41] take a new
job. n128 The court found it "unlikely" that the worker would have left without substantial assurances of job security,
and it held that this fact corroborated evidence of a fixed one--year contract. n129
Similarly, in Miller v. Community Discount Centers, Inc., n130 a worker left his family in Toledo and moved to
Chicago after he received a letter from his new employer stating he had "a rewarding and satisfying career ahead of him"
and confirming the employer would pay one--half of his moving expenses immediately and the balance after one year.
n131 The employer dismissed the worker after three months. The court upheld his claim for breach of a definite one--year
contract, finding it "inconceivable that a man of plaintiff's age would leave his home to come to Chicago for the mere
possibility that he would have a permanent position." n132
We see, then, that courts sometimes allow claims by beginning--career employees who are arbitrarily fired after
moving or quitting a prior job. n133 Some courts use a promissory estoppel or reliance theory, some find an implied
contract for a reasonable time to allow the employee to recoup his expenses, and some simply use the decision to move or
to quit as evidence of an actual definite--term agreement. Regardless of the theory for recovery, one can explain these cases
as attempts to regulate opportunistic firings early in the life cycle. Employers have not yet invested in the relationship and
thus are not hurt if they arbitrarily dismiss the new employee. This means that the relationship is not self--enforcing, as it
is when both parties have incurred sunk costs.
Nevertheless, protection for beginning--career employees is far from universal. Many or even most courts refuse to
find that reliance on an at--will job offer is reasonable. n134 In these cases, an employee [*42] quits another job or
moves to a new job at his own risk.
The ambivalence of courts in this area is understandable. For at least three reasons, the opportunistic termination
rationale for protecting employees is weaker in these beginning--career cases than it is later in the life cycle. First, very
often the employer also makes substantial investments early in the relationship. Recruiting and training new employees
can be a major cost to many firms. As Paul Weiler describes recruiting costs:
The recruiting process itself imposes significant costs on the firm; not merely on the personnel department, which must
do the initial advertising and screening, but also on the operating divisions, which must interview and judge the suitability
of candidates. The magnitude of these costs can vary widely, depending on the nature of the job, the skills required, the
number of applicants, and so on, but on occasion they can be substantial indeed. n135
One study estimated that a typical firm spends 160 hours in hiring and training a new worker in the first three months
on the job, and that these costs are nearly thirty percent of the value of an experienced coworker during the three--month
period. n136 If the employer as well as the employee sustain heavy early costs, the risk of opportunistic termination is
smaller. An employer that arbitrarily or unjustifiably fires an employee hurts itself as well, for it wastes the expenses of
Second, even if recruiting costs are insignificant - as they will be in many cases - so that arbitrarily firing the
employee does not penalize the employer, the employer gains nothing from firing a person early in his career. Thus, while
employees often suffer no penalty from an arbitrary beginning--career firing, they gain no benefit from them either. This
fact distinguishes beginning--career from late--career firings, in which the employer can gain from firing employees whom
it pays more than their current output. [*43]
A final problem with job protection for new employees is that employers often need a probationary period to sort
out hiring mistakes, wherein they can fire employees without explanation or extensive documentation of their reasons.
Relevant here is the fact that competitive firms, largely responding to the entry and exitof early--career employees, virtually
always contract for at--will dismissal. n137 Even the Model Employment Termination Act, which calls for general good--
cause protection for employees, refuses to protect employees with less than a year of service. n138
In sum, in many situations employer opportunism against beginning employees is either a trivial threat or outweighed
by legitimate needs to maintain employer flexibility. In these situations, courts do not scrutinize the sudden termination.
Still, the potential for opportunistic employer offers is real. An employer engages in opportunistic behavior when it hires
a better person before training anyone but after the first job applicant has relied on the offer. Courts protect beginning
employees from such opportunism.
2. Late--Career Opportunism
Late--career employees face the greatest danger of opportunistic firings. At the end of their life cycle, they often earn
more than their current productivity. If they do, the employer has a financial incentive to terminate them, even if it violates
an implicit promise to allow the employee to reap the rewards of hard work earlier in his career.
The Age Discrimination in Employment Act provides one check against late--career opportunism. By prohibiting
employers from firing workers above the age of forty because of their age, the ADEA protects older workers from
discharges based upon stereotypes that lead employers to underestimate their productivity. "We need new blood" and
"Doe is slowing down a notch" are classic statements that create age discrimination lawsuits.
However, the ADEA may offer only limited protection against the central concern of the life--cycle model -
opportunistic firings when salary and forthcoming benefits outweigh current productivity. In Hazen Paper Co. v. Biggins,
n139 the Supreme Court unanimously held that an employer did not violate the ADEA when it fired a sixty--two-- [*44]
year--old employee just before his tenth anniversary with the company in order to keep his pension benefits from vesting.
Although the Court agreed that the firing was actionable under section 510 of ERISA, it held "that an employer does
not violate the ADEA just by interfering with an older employee's pension benefits that would have vested by virtue of
the employee's years of service." n140 Justice O'Connor, writing for the Court, emphasized that age is not the same as
years of service. n141 "The ADEA only requires an employer to ignore an employee's age; it does not specify further
characteristics that an employer must also ignore." n142 Under the life--cycle model, the employee becomes vulnerable
to opportunism with years of service, not with age. The ADEA is therefore of little help.
A recent Second Circuit case, consistent with but not cited in Hazen Paper, shows even more clearly the limitations
of the ADEA in protecting workers from late--career opportunistic firings. In Bay v. Times Mirror Magazines, Inc., n143
a fifty--four--year--old publisher who made nearly $200,000 per year brought an ADEA claim after being fired from his
position at Field and Stream. The employee thought he had a "smoking gun" when he produced an internal memorandum
from the company's chairman stating that the employee's salary alone mandated his dismissal. The employee asserted the
memo established his "high salary was a critical factor in the decision," and that the high salary was "a direct function
of his longevity, experience, seniority or periodic salary raises." n144 The Second Circuit, without disputing these
assertions, affirmed summary judgment for the employer. Conceding that "high salary and age may be related," the court
declared that "nothing in the ADEA ... prohibits an employer from making employment decisions that relate an employee's
salary to contemporaneous market conditions ... and concluding that a particular employee's salary is too high." n145
Other cases interpret the ADEA more expansively. In Metz v. Transit Mix, Inc., n146 for example, the Seventh
Circuit reversed a trial court judgment for an employer who had replaced a fifty--four--year--old, highly paid manager with
a younger, cheaper colleague. The court emphasized that an employer cannot assess the costs of employ [*45] ing an
older worker when deciding whom to terminate because pay is a " "proxy' for age." n147 Under this interpretation,
which received an enigmatic citation from the Supreme Court in Hazen, n148 the ADEA protects older workers fired
because their salary exceeds current productivity. Because of the close connection between a worker's age and time he
spends with the company, the ADEA indirectly protects late--career employees as well.
Greater protection may come from common law courts, which in recent years have begun policing opportunistic
firings of late--career employees. n149 The leading case is Pugh v. See's Candies, Inc., n150 in which a thirty--two--
year employee was abruptly fired after working his way up from dishwasher to corporate vice president. The employee
never had a clear agreement about job security, although managers had given him encouraging evaluations over the years.
n151 The court held that this career pattern, including the length of service and the policies and practices of the company,
could establish an implied--in--fact promise against arbitrary dismissal. n152 Pugh epitomizes the effi [*46] ciency--
wage story and its end--game dangers. Pugh committed himself to a single firm, worked hard to gain promotions to the
promised easy life, but then was terminated. Pugh also demonstrates the effect of the arrival of new management on job
security. In Pugh, new management arrived a year before Pugh's firing. Such major corporate changes may diminish the
reputational check on firings of late--career employees.
Length of service is the key element that motivates courts to scrutinize a late--career firing. Most opinions, like Pugh,
also examine oral statements and the company's general procedures. But a Montana case, Flanigan v. Prudential Federal
Savings & Loan Association, n153 starkly illustrates the centrality of longevity. In that case, a bank fired Mildred
Flanigan without notice or a hearing after twenty--eight years of service. The Montana Supreme Court affirmed a nearly
$1.5 million judgment for Flanigan, declaring that her "28 years of employment by Prudential gave her a secure and
objective basis for believing that, if her work was satisfactorily performed, her employment would continue." n154
In its decision, the court quoted extensively from a California appellate case, Cleary v. American Airlines, Inc., n155
which also emphasized longevityof service as a key element of a bad faith claim. In Cleary, the court upheld a claim
brought by an employee dismissed without cause after eighteen years of service. The court declared that "termination of
employment without legal cause after such a period of time offends the implied--in--law covenant of good faith and fair
dealing contained in all contracts including employment contracts." n156
Occasionally, an employee faces the danger of beginning--career and late--career opportunism at the same time. This
situation occurs when a long--time employee agrees to a job transfer. A prominent example is Foley v. Community Oil
Co., n157 in which a thirty--year employee was fired three years after accepting a job transfer to another state. The
court's explicit rationale in finding the employee had stated a claim tracks the life--cycle theory. The court first noted that
"uprooting and moving a family" could give rise to a contractual claim. n158 The court then declared that "longevity of
service can also give rise to an implied contract right." n159 Tracking the lock--in problem with [*47] which we have
wrestled, the court explained that "the employee, in providing long--term employment to a single employer substantially
diminishes his economic mobility." n160
In sum, one can explain these cases as attempts to monitor and enforce the implicit life--cycle employment contract.
Late in an employee's career, the usual checks against opportunistic firings unravel. Courts enter to monitor the bargain.
The bargain does not give late--career employees complete job security. They can be dismissed for cause, because
otherwise the shirking problems would be immense, but the employer does not prove cause simply by proving that salary
exceeds current productivity. That is the typical life--cycle pattern that both sides to career employment anticipate and, ex
ante, it is in the interests of both sides.
3. Midcareer Shirking
Once the employer has begun to make substantial, asset--specific investments in an employee, the risk of arbitrary
firing diminishes. The greater danger of opportunistic behavior - at least, behavior that an appropriate dismissal standard
could limit - comes from the employee's side. Because the employer does not want to repeat recruiting and training
costs with another employee, the incumbent employee has an opportunity to shirk without fear of dismissal. Shirking at
midcareer can occur even if the employer has the right to dismiss at will, but the shirking problem can be exacerbated if
the employer must also surmount the hurdle of proving just cause.
This is not to say that the employer cannot exploit the midcareer employee. Indeed, as I emphasized above, n161
being trapped by investments in firm--specific capital and in community roots can make a midcareer employee ripe
for exploitation. But the exploitation will not take the form of firing because the employer is making money from the
relationship. Rather than fire a midcareer employee, an employer may pay him less than would be called for under a fair
division of the gains from the long--term relationship or make his workload or working conditions more onerous. Just
cause cannot protect the midcareer employee from these abuses. Better, then, for the law to focus on something it can
handle, which is deterrence of shirking by midcareer employees.
The courts seem to have intuited this fact by refusing, in general, to create contract protections against arbitrary
terminations for mid [*48] career workers. Midcareer employees have made the fewest contributions to the doctrinal
erosion of at--will employment. n162
The most detailed data on wrongful termination plaintiffs come from a Rand study n163 of 120 California jury trials
n164 in the early 1980s. Over half the plaintiffs in this sample were early--career employees, with five or fewer years of
job tenure. n165 Nearly a quarter of the plaintiffs had over fifteen years of tenure, n166 with the remaining quarter of
the plaintiffs being midcareer employees with six to fifteen years tenure. n167 This sample of jury trials suggests that
midcareer employees are a minority of wrongful termination plaintiffs, and many of these may be bringing public policy
and other tort claims not inconsistent with the life--cycle model. n168
The leading cases also suggest to some extent the courts' reluctance to protect midcareer employees. One case that
typifies this hesitation is Rowe v. Montgomery Ward & Co. n169 The case is especially significant because Michigan
courts have been leaders in eroding the at--will presumption. In Rowe, a salesperson with eight--years tenure was fired for
leaving the store one day without explanation. The employee sued, claiming she had been orally told she would have a job
as long as she met her sales quota, and pointing out that when she was hired she had signed a "Rules of Personal Conduct"
that enumerated only four reasons - all involving theft, dishonesty, or immorality - [*49] for immediate discharge. The
Michigan Supreme Court denied her claim, n170 emphasizing that the words of assurance were "more akin to stating
a policy" n171 than offering an express contract, and noting that nothing in Montgomery Ward's "Rules of Personal
Conduct" suggested that the enumerated conduct was the only basis for dismissal. n172 The Rowe court distinguished
Toussaint v. Blue Cross & Blue Shield, n173 a leading employee--rights case, by emphasizing that Toussaint involved
actual negotiations over job security by high--level employees. n174 Rowe evoked a spirited dissent by the author of
Toussaint, who insisted that an enforceable promise was equally present for the low--level salesperson. n175
The Court in Rowe did not rely heavily on her moderate job tenure but simply disallowed her claim because it found
the circumstances were insufficient to infer an implied--in--fact contract. n176 One can only speculate that the court
would have viewed Rowe's claim more sympathetically if she had thirty--years tenure rather than eight. Under the life--
cycle model, the case would be dramatically different if Rowe had been a late--career employee.
The most prominent counterexample to my claim about midcareer employees not receiving protection is Foley v.
Interactive Data Corp, n177 in which the California Supreme Court upheld an implied--in--fact contract claim as well
as an implied--in--law breach of good--faith claim brought by an employee who served the company just six years and
nine months before he was terminated. The Foley case is remarkable for at least two reasons. First, and ironically, most
commentators view the decision as a dramatic cutback on employee rights because the court refused to grant tort damages
for employees claiming a breach of the covenant of good faith. Second, and more pertinent to our analysis, Foley pushes
the limits for defining a "late--career" employee. Basing an employee's just--cause claim on less than seven years of service
cannot realistically be viewed as an attempt to deter opportunistic firing of late--career employees whose seniority--based
earnings outrun their current productivity. n178 Indeed, the court recognized [*50] that short tenure might weaken
Foley's implied contract claim. n179 It emphasized, however, three additional factors: (1) Foley alleged (rather vague)
oral assurances of job security and consistent promotions and salary increases; n180 (2) Foley alleged breach of written
"Termination Guidelines" that suggested self--imposed limitations on the employer's right to terminate employees; n181
and, (3) unlike Pugh, Foley had "supplied the company valuable and separate consideration" by signing a promise not
to compete against the employer for one year after termination. n182 Many employees can allege the first two factors.
The third factor is less common, although far from unique. Perhaps these other elements explain why Foley is not
consistent with the general claim that courts rely on longevity of service and scrutinize only late--career terminations for
opportunism. More realistically, Foley probably reflects a general move in California toward a good--faith standard for all
In summary, my argument is that the general pattern of good--faith and implied--contract cases reflects an intuitive
understanding by the courts that employees are subject to opportunistic discharge at the end, and less consistently at
the beginning, of the life cycle. Courts are reluctant, however, to give general protection against arbitrary dismissal to
midcareer employees. The economic self--interest of employers should keep such dismissals in check. The greater concern
is with employee shirking.
To clarify the distinction I draw between scrutinizing opportunistic late--career terminations and scrutinizing all
employment decisions under a just--cause standard, let me return to the facts of Murphy v. American Home Products.
n184 Murphy, like the California Foley, was an internal whistleblower who reported to upper management wrongdoing by
immediate supervisors. n185 While bucking the corporate hier [*51] archy often gets employees into trouble, courts are
hesitant to referee the resulting turf fights. n186 Internal whistleblowing resembles too closely legitimate, but practically
unverifiable, concerns that an employee is not a team player or that an employee creates difficulty in the office. Legal
limning of these situations is probably not worth the costs.
Thus, I would have greater sympathy for Murphy if he claimed that he was fired after twenty--three years of service
n187 because he was no longer pulling his weight or earning his salary. Unless the parties have clearly agreed to at--will
dismissals throughout the life cycle, such a firing smacks of an employer opportunistically firing an employee who has
committed the best years of his life and should reap, based on the norms of seniority, the benefits of a career commitment
to the employer. In fact, Murphy did claim age discrimination - a claim that was still being litigated a decade after his
discharge. n188 But a legitimate defense might be that opportunism had nothing to do with the termination; he was fired
because he attempted to buck the corporate system. Of course, when an employer fires an older worker allegedly for such
a reason, the proof problems in sorting out the real reason for discharge are enormous. This dilemma quite likely will
make employers wary of firing older workers. My point, in short, is that just--cause protection should be limited to an
inquiry into whether the employee was fired in breach of the life--cycle commitment to pay seniority--based wages and
benefits or for other opportunistic reasons.
IV. Default Rules
"Aren't we done now?" the dear reader might ask. Not quite. We have seen that a life--cycle rule may be the best way
for employer and employee to minimize the dangers of opportunism on each side of the employment relationship. One
might still argue, however, that parties seeking this arrangement should put an explicit life--cycle rule in the contract.
If they do not, courts should presume the contract is at will [*52] or just cause. The final step in my argument, then,
explains why the life--cycle rule should be the default rule for courts, even though the parties could choose any rule by
A. Minimizing Transaction Costs
The traditional law--and--economics literature on default rules suggests that courts do and should choose rules that
minimize transaction costs. Two sometimes conflicting tests come from this approach. First and most prominent is the
"mimic the market" or "would have wanted" test, whereby courts supply the default contract term that most parties would
put in the contract were they bargaining without costs and with full information. This test saves most parties from the
costs of acquiring information and bargaining over the term. The trick in applying this test is to determine which rule
maximizes joint gains by helping one party more than it hurts the other. In general, a complex default rule is likely to
"mimic the market" better than a simple rule, in that parties who were bargaining costlessly would probably agree to
share risk and minimize opportunism on both side. The second test, the "improve bargaining that occurs" approach, urges
a default rule that lowers costs for those who must bargain rather than allows most parties to avoid bargaining. n189
A simple, clear default rule may be easier for the parties to bargain around, and it may thus lower transaction costs for
parties who will actually bargain over the particular term. n190
Luckily, in our situation both tests point in the same direction. n191 [*53] First, under the would--have--wanted
default test, our previous analysis suggests that a life--cycle termination standard would be optimal for most career
employees; by minimizing opportunism on both sides, it allows for the most productive relationship. Because it maximizes
the overall gains to the relationship, most parties bargaining under low transaction costs would opt for the life--cycle rule.
Second, the lower--bargaining--costs approach also favors a life--cycle default rule because it is easier for parties to
bargain away from than toward the life--cycle rule. At--will clauses are easy to compose. n192 Just--cause clauses are also
straightforward to write, although the phrase "just cause" is a rich and complex term of art in labor arbitration. n193 In
Professor Rose's marvelous terminology, these are "crystal" rules. n194 By contrast, a life--cycle rule would be hard to
draft because it would be difficult to specify at the outset of a relationship exactly when the relative vulnerability switches
from employer to employee. Rose would call the life cycle a "muddy" rule. n195 The parties cannot easily articulate
at the time of initial hire the proper governing structure for their future relationship. They may prefer to rely on courts'
often bumbling and instinctive judgment about relative vulnerabilities. Under the life--cycle default, parties can simply
say nothing too explicit in the contract and count on courts to apply the life--cycle approach. In short, the parties can
easily draft away from a life--cycle default if they choose, but they cannot easily draft away from an at--will or just--cause
presumption toward a life--cycle rule.
Indeed, the ambiguity in the timing of a life--cycle default rule may be itself desirable. Suppose a contract explicitly
called for at--will em [*54] ployment for the first fifteen years of the relationship but just cause thereafter. The employer
would have a strategic incentive to review an employee at fourteen years and eleven months and terminate if forecasts of
future performance were not rosy, even if current performance were adequate. The costs of termination to an employee
after nearly fifteen years are tremendous. n196 But under the more malleable life--cycle default, the strategic doomsday
evaporates. An employer, unsure when the at--will standard changes to just cause, is more likely to act in good faith.
Under either approach, then, a life--cycle default rule is optimal. To reiterate, under a life--cycle default, the court
presumes that midcareer employees have an at--will relationship with their employer because that contractual structure best
deters opportunistic behavior by the parties. Late--career employees, by contrast, are ripe for opportunistic termination,
and so courts require good cause for terminating such employees. The optimal standard for new employees is more
nuanced. On the one hand, employees who have quit other jobs or who have significant moving expenses perhaps should
have a reliance damages default. On the other hand, employers often need a probationary period to sort out poor matches.
Perhaps a default probationary period should be presumed, absent unusual reliance expenditures by the beginning
B. Relational Contract Default Rules
Recently, the literature on default rules has examined long--term relational contracts. The debate centers on whether
simple or complex rules are better able to deter opportunistic behavior in the relationship. Some have argued that complex
default rules that consider the particular circumstances of the parties better control the strategic behavior problems of
relational contracts. n197 Certainly the life--cycle default, which is premised on the fact that both employer and employee
can exploit the other's sunk costs, attempts to control strategic behavior on both sides. In this way it is a complex default
In contrast, Robert Scott has noted that most default rules in commercial relational contracts tend to be simple,
categorical, and winner-- [*55] take--all. n198 Scott cautions that these prevailing default rules may have normative force,
so that concocting a complex default may be a misguided attempt to control strategic behavior. n199 As he emphasizes,
legal rules "are both a threat and a temptation." n200 Legal attempts to prevent opportunism by one side invite evasive
responses from the other side. Certainly this is true for employment relations. Preventing employer opportunism by a
just--cause standard invites increased employee shirking. Scott concludes by emphasizing that legal sanctions are not the
only control on opportunism. n201 Social forces of reciprocity and honesty, particularly when benefits accrue to a good
reputation, are powerful deterrents. Rather than legally enshrining these social norms, which may destroy the informality
that makes them so effective, the optimal structure may rely on clear, harsh, legal defaults combined with social sanctions
against failure to cooperate.
Scott's argument against complex, contextualized defaults is powerful in the commercial context in which he uses
it, but its lessons may justify a life--cycle default here. First, the life--cycle default may not fall on the complex side of
Scott's spectrum. Although parties could not easily agree on the tipping points, the life--cycle default does call for a
categorical legal winner at every point in the relationship. In that sense, the default is clearer than a default rule that calls
for sharing. Second, to the extent a life--cycle default is complex, its complexity arises from the common law, albeit still
in embryonic form; Scott objects to complex default rules that the common law has ignored or rejected. Thus, like Scott,
I am using the common law to provide both a positive description and a normative base for a life--cycle default rule.
Finally, Scott's argument focuses on commercial relationships of indefinite or permanent length. The opportunism can
come at any time by either side, usually in response to exogenous shocks to the relationship. While random shocks can
also disturb employment relationships, the inevitable life cycle of the employment relationship presents clear end--game
and beginning--game problems, in which the employer's potential for opportunism becomes predictable and one--sided.
In midcycle, by contrast, the self--interest of the employer in not firing productive workers provides a nonlegal check on
arbitrary firings, so the law should focus on the possible opportunism by employees. Given this predictable cycle, the
legal default rule for employment ter [*56] minations can focus more precisely on opportunistic threats than can a default
rule in the usual commercial relationship.
V. The Variety of Employment Relationships
A. Who Are the Life--Cycle Workers?
At the beginning of this article, I noted the great variety of employment relationships. In subsequent sections, I
concentrated on the career employment relationship with two defining characteristics: (1) both sides invest heavily in
the relationship in ways that will be lost if the relationship is severed prematurely; n202 and (2) contracting problems
prevent easy monitoring of work performance or verification of poor performance to outsiders. It is time to consider what
workers fit the life--cycle model. Table 1 provides the framework. n203
Types of Employment Relationships
and Optimal Termination Standards
We must determine whether just cause or at will is the preferable standard for a particular type of job. The employer's
perspective depends on which row of the table applies. When verification is easy (top row), a just--cause presumption does
not harm employers. An employer can simply show a court or arbitrator it has just cause for firing a shirking worker. By
contrast, when verification is hard (bottom row), a just--cause standard makes employers vulnerable to opportunism by
shirking workers because the employer cannot verify to a court or arbitrator the reasons it suspects shirking. [*57]
The employee's perspective depends on which column applies. When workers have made only general investments
(left column), their skills are transferable to another firm. Even if they are terminated from one firm, they are not hurt
greatly because other firms are willing to pay them comparable wages. Because terminations are less costly to employees
in this column, at--will protection is adequate. By contrast, when workers have made firm--specific investments, which
include entering an individual firm's career ladder (right column), they suffer greatly from termination.
These observations lead to easy conclusions for the off--diagonal boxes (2) and (3). The employer and employee
perspectives clash more directly in boxes (1) and (4). In box (1), neither party is vulnerable, so the discharge standard is
less important. I have spent the bulk of the article addressing box (4) and will simply reiterate that a life--cycle rule best
accommodates the mutual vulnerability.
The more challenging task is pigeonholing a worker or job in a particular box. One dichotomy is between general
investments and specific investments. Jobs in which both sides have made only general investments are nearly an empty
set among workers with more than a few years' experience. Importantly, investments are broader than job skills. Many,
perhaps most, workers have skills that many firms value and are thus general. These general skills make them less
vulnerable to opportunism. But most workers develop specific ties to their workplace - familiar faces and routines - that
they will lose if they leave. Often, workers enter a career job ladder with a particular firm, assuming that hard work will
lead to promotions and future rewards. If one defines workers with specific investments as workers who will suffer from
job loss, most experienced workers fall into this category. n204
It is harder to draw the line that separates jobs in which acceptable job performance is easy to monitor and verify from
jobs in which monitoring and verification is hard. Perhaps common law courts can draw the line - as they draw so many
others - on an intuitive, case--by--case basis. This approach would be acceptable if the courts kept the function of the line
in mind. The line is supposed to separate jobs in which a just--cause requirement will not create severe shirking problems
from jobs in which an employer cannot easily prove objectionable employee behavior to a court and so employers cannot
credibly threaten to fire shirking employees.
Some have suggested that high--level as opposed to low--level jobs [*58] present such a dichotomy. An employer
may only have an unprovable sense that a high--level worker is not performing adequately. Such jobs have many intangible
characteristics, such as the ability to motivate and work with others. By contrast, low--level jobs usually involve the
performance of tangible and verifiable tasks. Even such a staunch and eloquent advocate of just cause as Professor St.
Antoine has suggested that high--level employees should be exempt from a comprehensive just--cause scheme. n205 He
has suggested that we should draw an indirect line excluding from just--cause protection employees entitled to a pension
above a certain amount or employees who have fixed--term contracts of two years or more. n206
One must be careful of class myopia in making such proposals. Most jobs, even the most unskilled or menial, require
complex mental states to perform them well. Boredom, frustration, and low morale can impair performance on any job.
While sometimes this fact manifests itself in objective, verifiable ways, many times it does not. Certainly job status and
job complexity do not correlate well. Nevertheless, a duality exists between jobs requiring simple, repetitive tasks and
those requiring complex, varied tasks. If the tasks are simple and repetitive, "successful job performance is relatively
easy to define and measure in ways that virtually every reasonable person would consider fair, accurate, thorough, and
Courts and commentators have thoroughly debated the analogous issue in employment discrimination law; few tests
for job applicants have survived a challenge of bias. If objective tests cannot be found to evaluate job applicants, we
should not expect employers to be able to provide objective evidence of cause when they fire a worker for not performing
adequately. Mark Kelman has noted this tension. n208 In advocating a near ban on employment testing for applicants
because objective criteria for hiring cannot be found, he recognizes that employers must be given greater leeway in
dismissing workers who do not live up to the admittedly subjective standards of the job. n209 In [*59] other words, the
law cannot simultaneously squeeze both the hiring and firing decisions.
I have noted already that unionized employees are universally governed by a just--cause standard while nonunionized
private employees generally are not. n210 One explanation for this difference is that unionized jobs are typically more
routine or repetitive, thus making objective measures of performance easier and lowering the burden of showing cause.
On the other hand, unions may demand regulation and rules precisely to facilitate objective evaluation of workers. n211
If objective evaluation is feasible, the problem of unverifiable shirking is reduced and the just--cause requirement becomes
less burdensome for employers.
Finally, in considering what jobs fit the life--cycle model, we must recognize that the model may apply differently
to female workers than to male workers because the career--employment pattern differs. For many women, the end
of the life cycle is not the period of greatest vulnerability. Rather, their time of concern is the middle period, when
many women leave the workforce temporarily to have children. A common employment pattern is for women to "prove
themselves" by working hard early in their career in exchange for implicit promises that the employer will give them
greater flexibility in early child--rearing years. Having performed their end of the bargain, women face the danger that
an employer will behave opportunistically by dismissing them rather than giving them the promised flexibility. The law
regulates these opportunistic firings under Title VII and other antidiscrimination laws. Although our general analysis
against opportunistic firings applies to this situation, the common law rules we have analyzed apply more often and more
cleanly to "traditional," that is male, life--cycle patterns. n212 Certainly, men have brought most of the leading life--cycle
wrongful discharge cases.
B. The Rise (and Fall?) of Career Employment
If legal intervention at the end of the life cycle is so wise, a critical [*60] reader might ask, why did it take courts
so long to get it? Or, alternatively, one might ask whether this sudden judicial rush to protect some workers suggests
an unwise departure from the wisdom of the past. In response, I would emphasize that career employment is relatively
recent in our economy, becoming common only after World War II. Thus, only in the last decade or two have employers
and workers played out the end game of career employment. Therefore, courts have only recently had the opportunity to
respond to opportunistic behavior at the end of the life cycle.
Certainly, career employment was less prominent fifty or seventy--five years ago. Henry Ford introduced his five--
dollar--per--day pay in 1914 in large part to counter the phenomenal turnover in his River Rouge factory, which exceeded
2000% per year. n213 Much of the rise in career employment can be attributed to the growth of firm size and the
increasing costs of employee turnover and lack of discipline. n214 Immigration and reverse migration before the 1920s
delayed a sense of community and roots among workers, diminishing a desire for job security. n215 Not until after World
War II did pensions - a key bonding feature of career employment - become prevalent. In short, until the last few decades
few workers spent their lives in a single career employment.
Because of the recent rise in career employment, I need not dispute Richard Epstein's claim that at will was the
optimal rule to regulate the employment relationship for much of this century. n216 Whether it was or not, times have
changed, and the common law has changed with it. With the rise of career employment has come the life--cycle doctrine
in employment law.
Some commentators suggest that career employment is becoming a thing of the past. n217 One bit of evidence for
this claim is the decline in [*61] pension coverage in the 1980s. n218 If life--cycle contracts decline in importance, one
might expect parties to call on courts less frequently to enforce perceived opportunism. As career employment ebbs, so
too may lawsuits ebb whose underlying theories rest on a breach of a long--term relationship.
The argument of this article has a classic form - it puts court decisions in an area of law into a framework and thereby
declares them to have some coherence. In this case, the declaration is that courts, with their embryonic life--cycle doctrine,
are reacting wisely to the issues litigants present to them. The life--cycle framework that courts have developed provides
the parties in a career employment relationship a legal structure that checks opportunistic behavior. Its fundamental
premise is that both employer and employees can act opportunistically. Consequently, a life--cycle analysis does not
categorically condemn or celebrate employment at will. It supports, in broad outline, the contract law inroads that have
been made on the at--will doctrine, particularly at the beginning and the end of an employee's career, and it explains the
continued vitality of the at--will rule for midcareer employees. The current position of the courts is superior to a dogmatic
insistence on the old at--will regime, which creates an excessive risk of opportunistic terminations for long--term, and
sometimes beginning--career workers. Moreover, the current hesitant, intermediate position may also be superior to a
general just--cause standard, which would lead to excessive shirking by midcareer workers.
The life--cycle framework therefore makes coherent the seemingly schizophrenic behavior by courts in employment
termination cases. Within the framework, courts will protect employees when the danger of employer opportunism is
high, but they will retain the at--will presumption when the employer is more vulnerable. One can thus argue that the
courts are reacting appropriately to the employment--termination cases they encounter.
This coherence in the common law is internal to the system. In particular, it assumes that common law litigation is
the chosen method of resolving these disputes, and that the courts largely do not consider the systemic costs of litigation.
It may be preferable, all things considered, to opt for an administrative or arbitration system that requires just cause for
all employment terminations. But it is unfair, in arguing [*62] for such a change, to portray the current common law
as hopeless chaos. Far from being chaotic, the current common law provides optimal rules for regulating employment