Daniel J.H. Greenwood

Home | Previous Page

Understanding Respondeat Superior

I. The rule.

Respondeat superior is the doctrine that states that an employer is responsible for an employee. More precisely, it states that a master is liable for torts of its servants committed in the course of their service.

Respondeat superior is NOT exculpatory and does not eliminate ordinary negligence doctrine. Thus, the master remains liable for any negligence of its own that can be proved without use of respondeat superior: negligent hiring or supervision, for example. Similarly, the servant remains liable to the tort victim for his or her own torts. Moreover, under standard agency law, the master usually is entitled to indemnification from the servant for any damages the master suffers from the servant's tort - that is, if the master pays damages under respondeat superior, it is entitled to sue the servant for repayment.

Respondeat superior relieves the tort victim of the need to prove that the master was directly negligent: it suffices to prove that the servant was negligent. Thus, if an employer reasonably creates a work situation in which employees can negligently injure people, respondeat superior means that the employer is liable without proof of employer negligence. For example, imagine a delivery service that reasonably trains employees and sends them out in reasonably well-maintained cars, but an employee carelessly runs a stop sign and hurts someone. Respondeat superior means that the employee's negligence will be deemed the employer's: the delivery company cannot take the (inconsistent) position that the employee is working for it (and her actions are theirs) when she makes deliveries properly but not when she runs stop lights. Because she ran the stop light while on its mission, the company is liable.

Respondeat superior can be thought of in two ways that usually, but not always, reach parallel conclusions.

First, it can be understood as a rule about organizations: the servant's act is the act of the organization, even if some organizational decisionmaker would prefer to reject it. This understanding makes the tort rule quite similar to the contract agency rule, which also holds masters liable for contracts made for them by their servants, even if they don't like the contract. The master, then, is liable because the servant's negligence IS the master's negligence.

On this view, respondeat superior is simply the intersection of ordinary agency law and ordinary tort law. A tort is no different than anything else the servant does. If an employee creates a product, the product normally belongs to the employer; if the employee enters into a contract, normally it is the employer that is bound; so too with torts. The employee's negligence, by agency, simply is the employer's, and ordinary tort law then applies. Respondeat superior is a rule of enterprise liability, seeking to explain which actions are the enterprise's and which are not.

Second, and less consistently, respondeat superior can be understood as a "no-fault" or vicarious liability exception to the usual negligence rule in torts. Ordinarily, one element of negligence is a finding that the tort-feasor acted unreasonably. Here, an exception is made: masters are liable for the torts of their servants without fault. This view sees the master and the servant as separate; the servant's negligence is not the master's.

Although the cases regularly invoke this "no fault" understanding, I find it quite confusing. Most simply, respondeat superior just isn't a no fault doctrine: unlike "no fault" products liability doctrines, respondeat superior always requires proof of all the elements of negligence. Calling it "vicarious" liability helps somewhat, but is still confusing: this terminology suggests that the master is being held liable for someone else's tort, but gives no guidance on when or why the usual rule that tort liability requires committing a tort is suspended.

Both "vicarious" and "no fault" terminology suggest that respondeat superior is somehow exceptional, thus further suggesting that the doctrine ought to be construed narrowly. I assume this rhetorical trick is the main reason why the argument does not die of its own frailty. But it has no principled explanation for when this "exception" should apply. If employers can be held liable "without fault" or "vicariously," why can't other tortfeasors? Why do we revert to the "fault" rule in frolic and detour contexts but not when employees defy direct employer orders? On this view, justifications of respondeat superior often depend heavily on external justifications that don't always make sense or seem to help in deciding actual cases.

In fact respondeat superior is no exception: tort cases overwhelmingly involve respondeat superior. Corporations act-tortiously or otherwise-through their agents. Most significant torts are committed by business organizations; it is virtually impossible to sue a corporation or other business enterprise without invoking respondeat superior in some form. Even the exceptional areas where tort liability without repondeat superior is possible are shrinking: automobile accidents are increasingly covered by no-fault, non-tort based, schemes and medical practices are increasingly institutionalized, so that the doctor is an employee rather than an independent contractor.

The basic doctrinal problems and complexities stem from these definitions.

A. Frolic and detour.

Most important, in a modern capitalist economy, servants are not slaves, which is to say they also have lives of their own when they are off duty and their actions have nothing to do with their master. Usually, this is easy to see and goes unremarked: employers are not liable for torts of their employees when they are not at work. One fairly arbitrary rule is that driving to and from work is ordinarily not considered part of the job: employers are not responsible for auto accidents before the time clock is punched or the employee shows up at the office door (unless, of course, the employee is paid for travel time). Similarly, even during the work day, when a servant clearly has taken a personal break and is not in service of the master, the master's liability ends under the "frolic and detour" exception.

At a certain point in the late 19th century, these exceptions threatened to swallow the rule, as employers contended that they had ordered their agents to act appropriately and thus any tort must be a "frolic and detour." Indeed, at one point the argument was seriously made that a corporation could not commit a tort, since torts are not among the permissible purposes for which a corporation may be formed. This position was rejected first in contract (where the basic rule is that if a reasonable third party would think that the agent had the authority to make the contract, the agent's act binds the master regardless of any private instructions the master may have given the servant) and then in tort (where the basic rule, analogously, is that if the tort was committed in the course of employment, the employer is responsible).

Today, the easiest way to think about the "frolic and detour" exception is to ask yourself whether the employee should be being paid - whether an hourly employee in a fairly normal workplace should have clocked out before engaging in this activity. Employers prevail when employee drivers detour to the other side of town to visit a friend, but not when they deviate from the prescribed route out of carelessness or to buy a pack of gum. For professionals, the lines may be more difficult to draw, but again the usual question the courts will ask is whether the servant was acting on behalf of the master at the time or in the place (taken moderately broadly) when and where the tort occurred.

B. Servant/independent contractor.

Agency law distinguishes between two basic types of agents: servants (including all employees) and independent contractors. Respondeat superior normally applies only to servants.

A servant is an agent who is subject to the control of the principal (called a master). The master provides the tools of the trade, determines when and where the servant works, sets up rules and standards for the workplace, provides the raw materials and owns whatever the servant makes. A servant typically has only one master at a time. Under standard agency law, the master may set rules unilaterally and the servant retains only the right to quit and any other rights she may be able to convince the master to agree to. Put crudely, and somewhat exaggeratedly, a servant sells his time, body and mind to the master, who then uses the servant as an instrument of production (and owns the product). All employees are servants. A factory worker is a paradigmatic servant: s/he follows orders, works to a schedule set by the master, has relatively little control over how the job is done.

In contrast, an independent contractor controls the workplace, the tools, and the manner of working. Typically an independent contractor has many principals, determines when to work for each one, supplies her own tools and materials, and retains the basic authority to determine how to do the job. Your doctor or your electrician are paradigmatic examples (with respect to you - if they work for organizations, they are almost certainly servants of the organization). An independent contractor sells her work product, not herself.

For this reason, most actions of an independent contractor are not imputed to the principal. Thus, for example, when an employee creates something, it simply is the employer's: if an associate is fired, the brief he just wrote belongs to his employer, even if he hasn't been paid for months. In contrast, when an independent contractor creates something, it belongs to the contractor until he is paid for it: if I hire an electrician to work on my house and don't pay him, he is entitled to place a lien - the lights he installed are still his. The tort rule follows the general agency rule.

Obviously, it will not always be easy to distinguish between servants and independent contractors. The free professions (law, medicine, architecture) were once paradigms of independent contractors, but today most lawyers and doctors are servants. Employers seeking to avoid employee taxes (e.g., social security, nanny tax, workers' compensation, income tax withholding), union contracts or workplace norms (e.g., norms or rules granting benefits to employees), or tort liability, or less cynically attempting to create entrepreneurial incentives, may attempt to re-characterize employees as independent contractors or give them greater autonomy on the job than standard Taylorism recommends. The basic test, however, follows the basic purpose of the distinction: if you have a boss, you are a servant; if you work for a market, you are an independent contractor. If the boss could control you, even if the firm chooses not to, you are a servant: however much autonomy the firm gives you, a law associate is still a servant.

C. Negligent/Intentional torts.

Respondeat superior applies to both negligent and intentional torts: if an employer orders the employee to assault a customer, the employer is unquestionably liable for the assault. However, intentional torts often seem more likely to be "frolics and detours."

Today, it would take some fairly unusual facts to persuade most courts that a negligent tort committed on the company premises and on company time was a "frolic and a detour." Even if the employee gained some measurable benefit (a higher bonus or easier work, for example) by a negligent action, the presumption is that the employee was working for the employer, that the employer created the working conditions and made decisions about the extent of autonomy to give the employee, and that were the employer dissatisfied with the choices employees were making, the employer could restructure the options open to employees. Thus, even if employees are making bad choices within the structure provided by the master, the master is still responsible for their choices.

In contrast, when an employee commits an intentional tort, courts are less likely to see that act as within the scope of employment. In effect, some courts seem to hold that in structuring workplaces, employers are entitled (as a matter of law, without regard to the actual facts) to assume that employees will not rob, rape, assault, slander their customers, and that if the employee does so, he is probably acting on his own time rather than on behalf of the employer.

Obviously, no court will treat this as more than a presumption: if the plaintiff demonstrates that the employer authorized the tort or that the tort was within the scope of employment for some other reason, the employer is liable. Similarly, if the employer was itself negligent, that always suffices: so if the employer (or its agent) was negligent in hiring the tort-feasor (for example, by hiring a past offender and placing him in a position to offend again without proper supervision), the employer is liable for negligence, even if it is not liable for the underlying intentional tort.

II. The context.

Agency law is the set of doctrines that govern the relationship between employees and employers as well as other forms of agents and principals. The basic rule is that an agent is acting for his/her principal. So, when a factory worker makes a product, the work-product belongs to the firm. When an agent signs a contract, it is the principal, not the agent, who is bound (think of an employee negotiating a sale in the firm's name). When an agent writes a document (think of an associate writing a brief), it is the employer, not the employee, who owns the writing.

Respondeat superior makes the perhaps superfluous point that torts are not distinctively different from other aspects of an agent's role. If an employee makes a bad contract for the corporation, the corporation is bound by it just as much as by a good one. Torts are treated the same way. Just as agents act for their principals when they create contracts or otherwise do their jobs as they promise, so too when they mess up.

The individualistic language of the law makes this doctrine more confusing than it need be: fundamentally, this is simply a statement of the obvious and commonsense proposition that when an enterprise commits a tort, the enterprise should be liable (in addition to any human who may be). Modern enterprises typically operate through employees. Indeed, most employers in our economy are corporations that generally act only through employees; the corporation itself can act only by resolution of its board of directors, something that happens only in extreme circumstances. Thus, if the enterprise were not liable for the actions of its employees, corporations would never be liable at all (unless the tort were committed pursuant to a vote of the board of directors).

III. The Justifications.

On the first view outlined above, respondeat superior is a logical corollary of agency law and tort law and does not require special justifications beyond the ordinary justifications for those two areas of law. Agency law is foundational to all modern economies: it is the basic set of rules that allows one person to act on behalf of another and, therefore, one person to employ another. Tort law (to oversimplify this course) has three basic functions: a criminal law-like function of condemning socially inappropriate behavior, a contract-like function of compensating people who have been injured by the actions of others, and a market-regulatory function of helping to ensure that private costs (and therefore prices) reflect social costs in order to harness markets to the public good.

The sound justifications for respondeat superior stem from the basic functions of the legal systems of which it is an intrinsic part. Firms can behave in socially inappropriate ways just as individuals can (and often far more so); tort law must be able to condemn anti-social firms for the same reasons it must be able to condemn anti-social individuals. Firms can injure others just as individuals can; tort law must force them to compensate those they've hurt just as it forces individuals to. And firms are far more important players in the capitalist market than individuals-if tort is necessary to correct market pricing failures, it is essential that it apply to firms. On this view, respondeat superior is a recognition of the reality of firms: that the enterprise is responsible for what its parts and members do. When an employee acts for the firm, her torts are the enterprise's torts.

In contrast, the second view, which sees respondeat superior as an exception and aberration to tort law, demands a special explanation. It is hard to find a convincing one in the literature or cases.

A. Compensation and cost spreading as special explanations for a Respondeat Superior exception

Respondeat superior is sometimes defended by invoking the desire to compensate victims or spread costs. Unquestionably, these are important (if subsidiary) goals of tort law, reflecting powerful moral and political instincts. Citizenship is a common enterprise; we have some degree of responsibility towards one another; decent societies do not allowed injured people to suffer without help.

But tort law generally takes an extremely limited view of the responsibility of citizens to help each other. In particular, as a matter of tort law the baseline is always lack of responsibility, not the opposite: unless the plaintiff proves a duty and its breach, the defendant has no tort obligation to help. Injury alone makes no claim in tort. Tort has no principle that those who are injured must be compensated or that those who are solvent must help-if we accepted those moral principles consistently, we would replace tort by a universal health insurance system financed by progressive taxation. Tort requires a finding that the defendant caused the injury, not just that the plaintiff is hurt and the defendant is solvent. In tort we are not our brothers' keepers, nor must we be Good Samaritans, citizens are not required to be patriots, fraternité is at most an aspiration not a legal obligation, and it is permissible to "act in the manner of Sodom,"(1) until some special relationship changes matters.

Tort law, instead, makes compensation a goal only after causation or fault has been established. The injured may have a claim, but in tort law the claim is only against those who have injured them. It is a private law rule, regulating the relations between individuals not individual and society. On the tort view, the first issue must always be responsibility of the tortfeasor, not victimhood or need of the tort victim.

Sometimes, commentators state that respondeat superior is justified by the compensation goal, because employers purportedly are more likely to be solvent or insured than employees and therefore victims are more likely to be compensated if employers are liable. But this reasoning proves either too much or too little. Too much, because if we accepted this argument, respondeat superior would be woefully inadequate: the moment you accept the premise that need alone demands a response, you must reject the fundamental tort principle of limited responsibility. If we are our brothers' keepers, we need a social insurance and safety net system, not tort.

Similarly, if the problem were lack of insurance, we could simply require insurance, either through a universal governmental program or (as in the automobile context) by statute mandating private insurance.

Too little, because no doctrine of respondeat superior that I am aware of depends on any facts about solvency or insurance. Of course, it isn't necessarily the case that employers are more solvent than employees. Indeed, many small businesses are deliberately organized as near-insolvent corporations, with the owner removing all profits as they are generated. Respondeat superior has never been extended to the owners of corporations: it is the corporation, not its shareholder, that is liable, even when it is the owner, not the corporation, who is solvent.

Cost spreading defenses often suffer from the same sloppy thinking. Of course, there is a great deal of moral power to the notion that we should all be our brothers' keepers: that when anyone is hurt everyone is and we ought to help out. But ultimately, this is a justification for a social security, workers compensation and health insurance system, not for tort. Tort never cost-spreads in a satisfactory fashion: no tort doctrine inquires directly into whether specific tortfeasors or victims are or could be insured, can include costs in prices or otherwise are able to transfer their costs to others.

B. Pricing, cost spreading and compensation in the general tort model

Arguments based on incentives or cost spreading can be made in a more careful way that seems much more persuasive, especially if we recognize that respondeat superior is not an exception to the general tort rules but rather simply the obvious way to apply them to organizations.

One of the most important justifications for tort law as a whole is that it corrects a serious market failure. Capitalist markets function by allowing individual consumers to make individual decisions about purchases based on their own individual assessments of price and quality. These individual decisions are aggregated by the market into a socially useful product, because producers who supply products that consumers want succeed, while those who don't, fail.

But the market system only works if the prices charged in the market reflect the social costs of the product. If the producer is able to produce a product without paying its full costs, the producer will be able to make a profit with prices that are "too low", in the sense that the social cost is higher than the price. Consumers will buy more of the product than they would if they had to pay the full cost, and society is worse off. The same is true, in reverse, if producers have to pay costs that are not associated with their product.

When a product causes an accident, the accident is one of the social costs of production just as much as is the energy, labor, technology and raw materials it uses. Coal mining accidents are part of the costs of using coal. So is pollution. But if coal producers do not pay for the accidents or pollution they cause, coal will be priced too cheaply, and consumers will use more of it relative to other energy sources or conservation than they would if they were paying its full costs. Tort helps to solve this problem by forcing tort-feasors to pay for the accidents they cause and thus making the accidents a cost of producing the product. Consumers, thus, will pay for the accidents caused by the product in its price, just as they pay for the raw materials, labor, energy and technology used in producing it. Paying the full price, they will make better choices between alternative products.

C. Incentives in the general tort view

Tort has a second major benefit on this pricing view: correcting incentives.

Absent tort liability, when a company has to decide whether to spend money to test a product for safety, develop a safer product, or provide safer working conditions, the cost of prevention reduces profits, while the accidents are someone else's problems. Given this set of rules, the market, as by an invisible hand, produces results that are to no one's advantage. Rational (amoral) profit-maximizers, or ordinary bureaucrats who believe that their job requires them to act as if they were rational profit maximizers will decide not to prevent the accidents. Consumers buy products that cause accidents, because they are cheaper. And the world is an uglier place for it.

Tort liability can help to correct the market incentives. If a producer is forced to pay for the accidents it causes, then it can save money by spending on safety (so long as the safety measures cost less than the accident). The profit motive is harnessed, then, to social benefit: what is good for the producer (minimizing the costs of accidents AND prevention) is good for all of us.

Of course, this rationale is only as good as the tort system is accurate. If corporations are not charged for injuries that they have caused, they will-in the pursuit of profit- produce more dangerous products at cheaper prices. Individual tort victims will be sacrificed to provide private profits and cheaper prices that are not socially beneficial, while retarding the progress of technical improvement and best use of the resources we have. Conversely, if they are overcharged, they will price products too high, and consumers will suffer for no reason.

D. Respondeat superior as a fulfillment of tort, not an exception

Respondeat superior is an essential aspect of this pricing and incentive function of tort. But the pricing function justifies the first view of respondeat superior, not the second. The key issue for pricing and incentive views of torts is to assure that the producer of products for sale pays for all-but no more than all-the costs (including injuries) of the product. The hard issue will be determining when an injury should be associated with the product. This is much the same issue as the one that the first view of respondeat superior focuses on: should the employee's act be seen as part of the process of production or as extraneous and unrelated? The reason to include the servant's negligence in the price of the product is that the negligence (and avoiding it!) are (or should be seen as) part of the cost of producing the product.

Respondeat superior simply stands for the proposition that torts committed by employees in the course of manufacture are part of the process of production. The firm created the circumstances that gave rise to the tort. The firm can change those circumstances-it can change the workplace, change the tools, reorganize the way the work is done, change the way it trains and pays those who do the work. Respondeat superior recognizes that the firm is responsible for the way its employees act and that their behavior-careless as well as careful-is part of the process of creating the firm's products. Their torts are its torts.

(Incidentally, this analysis also helps explain why respondeat superior does not extend to corporate shareholders, despite the implications of "take the bad with the good" analysis. In many firms it is the shareholders who are the primary beneficiaries of employee torts: they get the profits the firm makes because it did not spend money to prevent the injuries. One might think, then, that shareholders ought to pay for injuries caused by employee shortcuts. That is not the law, however. The law focuses not on who profits from the activity that gave rise to the injury, but rather on who committed it, and in this context, specifically on whether the employee was acting as the firm when she committed it. The boundary of a firm is often not obvious and is generally subject to a good deal of manipulation by the firm itself, its owners and its lawyers. This flexibility is highly useful for the economy as a whole, so courts are reluctant to impose rigid definitions of what is a firm or what is not. But one thing is clear: even without taking away from the firm's participants the right to set firm boundaries, courts can require they be consistent. Firms are deemed to be firms for better or worse: a firm shouldn't be allowed to say its employees are its agents when they do things that (after the fact) turn out to be useful for the firm, but reject them as outsiders when they do things that (after the fact) turn out to be expensive - if only because if firms were allowed to play that kind of "heads I win, tails you lose" game, we would all be the losers. But shareholders are consistently viewed as not part of a corporation; that is one of the most fundamental rules of corporate law. So there is nothing inconsistent about failing to hold the shareholders responsible for the firm's torts (or the firm's employees' torts) even when the shareholders are the most likely beneficiaries.)

-- Dec. '04

(1) The sin of Sodom, according to Jewish legal tradition, is to stand on one's legal rights even for pure spite, i.e., where it will injure someone else without benefitting the actor. Babylonian Talmud, Bava Batra 12b; see also, Ezekiel 16:49 ("only this was the sin of Sodom: arrogance! She had plenty of bread and untroubled tranquillity, yet she did not support the poor and the needy"); Mishnah Avot 5:10 ("There are four types of men. One who says, "Mine is mine, and yours is yours" -- this is the average person. Some say, this is the manner of Sodom. "Mine is yours, and yours is mine" -- the boor. "Mine is yours, and yours is yours" -- the hasid. "Mine is mine and yours is mine" -- the wicked."); Babylonian Talmud, Yevamot 44a ("One may not throw away the waters of one's well when others are in need of them").