Respondeat Superior and Coasian Cost Shifting
One of Coase’s key points is that when transaction costs are low, the legal rule may not matter because parties will negotiate around it. In those circumstances, since even cheap negotiation costs something, there is something to be said for setting the rule where the smallest number of parties will need to negotiate around it. In other circumstances, one party may have a bargaining advantage over the other, and it may make sense to set the default rule in a way that protects the weaker party. That way at least we aren’t adding to the power of the already powerful.
Respondeat superior offers an interesting empirical window into how Coase’s theory works in the real world, where transaction costs are always critical and bargaining power is often markedly unequal.
Employees and their employers are in constant communication in a bilateral relationship, making negotiation about as easy as it gets. Transaction costs here are relatively low.
On the other hand, power is quite unequal: typically, employers are larger and more diversified and therefore better able to walk from a negotiation than their employees, especially if they are negotiating individually. The result should be that employers are likely to obtain most of the surplus created by cooperation and trade.
The default rules about employee torts strongly favor employers. Under the basic legal rules, if an employee (servant) is negligent, he or she is liable, both to the tort victim and to the employer (master) (if the employer is damaged), even when the servant is functioning within a system created by the master. As a general rule, for example, even when a corporate master is sufficiently large that various forms of human error –including unreasonable mistakes, egregious error and even deliberate wrongdoing –become both foreseeable and statistically predictable, courts do not require employers to plan for such mistakes. A master is free to put an employee in a situation where a simple mistake will have grave consequences, to not install a fail-safe system, and then to contend that the employee acted unreasonably; most courts will take the circumstances as given rather than question whether the master should have changed them. Instead of asking whether the master created reasonable circumstances, typically courts ask only whether the employee acted reasonably under the circumstances.
When the error hurts an outsider, the doctrine of respondeat superior holds that the master is liable for the servant’s torts. This locution itself is oddly tilted in favor of employers: it suggests that the servant’s torts somehow aren’t the master’s, that when an employee injures someone, somehow that isn’t the company acting.
Companies act largely through their employees, and the employer/employee relationship is governed by agency law. In ordinary agency law, when an agent acts for a principal, the principal is bound and the agent is not. For example, when a lawyer negotiates a plea, it is the client who goes to jail, not the lawyer. This is true even if the principal later decides it wishes it had made a better deal: if a company hires someone to sell goods, and the salesman doesn’t get as good a deal as the company wishes, it does not have the option of ordering the salesman to perform instead of it.
If respondeat superior doctrine simply followed ordinary agency law, the rule would be that a tort “by” the servant is a tort by the master, because the servant is acting on behalf of the master and the master is acting through the servant. As to third parties, the master would be liable and the servant would not be.
Tort, however, is different – more tilted towards the employer. When an employee makes an especially good deal for the employer, the deal is the employer’s; the employee has no claim to it just because he or she did all the work. But when an employee is careless and hurts someone, the employee – the servant – is liable to the third party along with the employer – the master.
Between master and servant, respondeat superior is closer to ordinary agency law, which is also quite pro-master. Agents are required to put aside their own interests and act on behalf of their principals. If they don’t, the master may sue for breach of fiduciary duties of care and loyalty. The duties go only one way, however. Masters have no legal duty to look out for their servants; they may exploit them in any way that contract law permits. Similarly, under respondeat superior, the master may sue the servant for any damage the master suffers, including payments to the third party. The employer need not prove that it acted reasonably in putting the employee in a position to do damage in the first place.
As noted above, it is relatively easy for employers and employees to negotiate around this set of rules. For example, employees could demand higher wages (or employers could demand lower ones) to compensate for the risks they must assume or other costs of the rule. Or, one side could demand that the other indemnify them for any damages the law makes them pay.
In fact, this is precisely what we see, in practice if not in form. The usual plaintiff’s lawyer practice is to sue employers and not employees when companies injure someone, presumably because juries are less likely to empathize with the firm (that is, because juries are likely to see the legal rule as incorrect) and because the firm is more likely to be able to pay to judgment. Under the law, employers are entitled to turn around and demand indemnification from their employee, but they rarely do. Moreover, employees who are high up in the company – upper level executives – and have the bargaining power to obtain contracts (as well as the wealth to be worth suing) nearly always have indemnification clauses in their contracts providing that if they are sued, the company will pay the judgment.
Thus the practice in the world precisely reverses the legal rule. The law says that employees are primarily responsible and employers are secondarily responsible for corporate torts that can be attributed to a specific individual’s acts. Practice is that tort victims and tort feasors alike agree that employers should be primarily, and indeed solely, responsible under all but the most extraordinary circumstances. The law benefits the more powerful party and the practice reverses it.
It is not hard to understand why employers and employees come to the result they do. Employers, especially large ones, should be able to predict negligence, to treat it as a statistical fact rather than a bolt of lightening or gamble. Since they can predict it, they can price it, adjust for it, think about it, counteract it. Employees typically can do one thing only: attempt to be more careful. Employers have many more options, starting by changing the procedures. “Human error” can nearly always be solved by engineering changes, and only the employer can do that. But even if they can’t prevent the injury, the employer can smooth out its costs, eliminate the risk, spread the costs so that they are not a catastrophe for an individual (the victim or the careless employee) but a minor annoyance for many.
Employers thus are better insurers – they live in a statistical world. And they are better cost avoiders – they control more of their world and have more ability to change it. And they are better cost shifters – they can easily spread the costs of unavoidable accidents to customers, shareholders or employees, while employees can only shift to their employers (and in the case of individual employees, only with serious bargaining difficulties). If employees and employers agree (implicitly or explicitly) to have employers absorb these costs that the law puts on employees, they should create a net social gain that they can then divide among themselves.
Coase suggests that all other things being equal, the law ought to set as a default the rule that most negotiations would reach, since otherwise it is creating needless transaction costs. The main reason not to do that would be to redress some negotiating imbalance – for example, that one side can more easily impose costs on others, so that free negotiation will lead to a socially harmful result.
Coasian analysis, like law and economics generally, is often thought to be overwhelmingly rightwing: pandering to the interests of the rich against those of ordinary people, defending the status quo even when it is obviously unfair. Here, however, it clearly demonstrates the opposite: that the law of respondeat superior inefficiently favors the powerful.
Here, the normal practice is to make enterprises alone responsible for organizational torts, even when it is possible to see them as the fault of an identifiable individual employee. Accordingly, reversing the law would minimize transaction costs: the rule ought to be that masters alone are responsible for the torts of their servants, unless they negotiate an indemnification agreement between them to the contrary.
Moreover, reversing the rule would put the burden of change on the party that is better able to demand change: the enterprise. If in some unusual circumstance it makes more sense to impose costs on employees, we should expect to see quick and easy negotiation of that by contract.
-- Dec. '04