Private Law Versus Public Enforcement

What is better when dealing with harm? Tort liability or regulation?

You are reading Economic Forces, a free weekly newsletter on economics, especially price theory, without the politics. Economic Forces arrives weekly in the inboxes of over 20,000 subscribers. You can support our newsletter by sharing this free post or becoming a paid subscriber:

---

One thing that we write about fairly frequently is harm. Sometimes our actions can have negative consequences on others. Since we do not bear the full cost of the consequences of our actions, we might fail to take proper care to prevent such consequences. This is especially true if we get all of the benefits associated with our actions, but only pay some of the costs. On this topic, what we tend to write the most about is what should be done about these consequences.

Last week, I re-ran a post from our early days on externalities. (Our readers love posts on externalities, but you know that because you are a reader.) The point of that post was to use some simple examples to think about when the proper response to the harm done by these externalities is a contractual solution, a tax, or regulation. An astute commenter replied to the post and asked about tort law. Rather than rely on government regulation, why not simply allow people to file a lawsuit against those imposing harm? This is a great question and thus I decided to write an entire post about it.

Let’s begin by thinking about tort liability and regulation as two alternatives. In reality, these might be complementary solutions, but let’s first focus on substitutability.

A typical way to think about tort liability and regulation is to think about choices along two different dimensions. Tort liability is a private solution. Regulation is a public solution. Tort liability is an ex post phenomenon. People bring lawsuits after the harm occurs, seeking compensation for damages. Regulation is typically an ex ante phenomenon in the sense that the rules are determined in advance in order to prevent the harm from occurring in the first place. Thus, we might give some thought to whether a particular type of harm is best dealt with ex ante or ex post, or by private actors or regulators.

When thinking about these issues in terms of ex ante or ex post remedies, Donald Wittman offered an important framing. Any attempt to deal with the potential for harm ex ante would require some monitoring of the inputs that ultimately create the harm. On the other hand, any attempt to deal with actual harm ex post requires some ability to monitor the output. This helps us to think about the choice between tort liability and regulation in terms of relative monitoring costs. If it is less costly to monitor inputs, then regulation would be preferable to tort liability. On the other hand, if it is particularly costly to monitor inputs relative to outputs, then tort liability would be the preferred solution.

Of course, in reality, there is no reason why it would necessarily be an either/or decision between tort liability and regulation. It could be some combination of the two. For example, consider automobile accidents. All else equal, the probability of that a car accident occurs is increasing in the speed with which someone is driving. At higher speeds, it is harder to stop or swerve to avoid hazards and it is harder to maintain control of the vehicle. One solution to this problem would simply be to handle everything ex ante by introducing speed limits and enforcing those limits. Doing so could limit car accidents. Another solution would be to allow people to speed, but allow the damage that results from car accidents to be settled through tort law and the legal system.

What we tend to observe is some combination of the two. There are a number of reasons for this. One is that a well-enforced speed limit will prevent some, but certainly not all accidents. Another reason is that not all car accidents are caused by high speeds. Speed limits thus can be thought of a regulation-induced minimum standard for care. However, since it doesn’t prevent the harm from happening, victims can still seek compensation for harm.

Nonetheless, the basic logic holds here in terms of monitoring costs. The total expected cost from speeding is the probability of getting caught speeding multiplied by the fine. However, there is an additional cost, which is the probability that speeding leads to an accident multiplied by the cost of the accident. Optimal policy would seek to adjust the total expected cost of speeding in order to get drivers to internalize the cost of speeding — and there are a number of margins on which to adjust.

What optimal policy looks like and how is is implemented would depend on relative monitoring costs. How hard is it to detect someone speeding and how much would it cost to do so? How hard is it to determine who is at fault in an accident such that the party that caused the accident is subject to those costs (either through fines or legal liabilities or both)? Answers to these questions will depend on factors like population density, the area of roads that need to be monitored, and technology. Thus, we should observe variations in policy based on those factors.

But monitoring costs aren’t the only costs that matter for determining whether to use tort liability or regulation. Steven Shavell outlined four basic criteria that need to be assessed when choosing between these alternatives.

The first is in regards to knowledge of private parties relative to the regulators about the risks associated with particular behavior. It is not unreasonable to think that private parties might have better information about the riskiness of their behavior than regulators. If so, tort liability has an obvious advantage. Nonetheless, if regulators have better information about the risks, then regulation would be preferable.

Furthermore, when thinking about the information that potential regulators have, it is important to consider what types of expertise and the scope of expertise that is necessary. One issue during the pandemic, for example, was that guidance from particular agencies was based on very narrow scope of expertise. An epidemiologist can make predictions about how school closings will affect the spread of a virus, but are unlikely to be able to assess the costs of closing down schools associated with learning loss. In situations that require a broad scope of knowledge, regulators are likely to either recommend inadequate rules or too restrictive rules. This is not necessarily because they lack expertise, but rather that their expertise is too narrow and thus fails to take into account all relevant margins.

The second criterion is whether the people who are potentially imposing harm are able to pay for the damages that they cause. Suppose that one engages in an activity that results in damages of $1 million, but has a net worth considerably below $1 million. In this case, the one imposing harm does not have the resources to compensate the victim. As a result, tort liability might not be a sufficient deterrent to behavior nor will it allow for adequate compensation for the victim. In scenarios like these, regulation might be preferable since it attempts to limit potentially harmful behavior ex ante.

A third criterion is whether people are likely to face a lawsuit in the event that they cause harm. Just because lawsuits are possible does not mean that they are likely. Think back to the example from my previous post in which there is a single factory polluter upstream from a number of fisheries. I noted that the likely solution in that scenario was regulation. This was based on two characteristics of the example. The first is that the reduction in production of the fisheries as a result of pollution makes this example ill-suited to a tax because the polluter could reduce its tax lability by polluting more. The second is that the threat of a lawsuit might be an insufficient deterrent for pollution. Since there are a large number of fisheries, there are coordination costs that potentially reduce the likelihood of a lawsuit. Even class action-type lawsuits tend to do more to punish wrongdoers than adequately compensate the particular victims, which might discourage victims from joining.

Shavell’s final criterion is the relative administrative costs between tort liability and regulation. In order to carry out regulation, one likely needs administrative agencies that are dedicated to monitoring and enforcement. This is costly. However, tort law also requires time, energy, and resources also. There are of course the private legal costs, but there are also public costs as well (e.g., judges must preside over trials, the public must serve on juries, etc.). As Shavell notes, however, the comparison in administrative costs itself depends on how the system of tort liability is designed. With regulation, the public costs are essentially the same regardless of whether harm occurs. However, to the extent to which legal liabilities discourages harm, they limit these public costs. Furthermore, to the extent to which tort liability results out of court settlements, this helps to minimize the public costs.

As Shavell notes, neither tort liability nor regulation dominate in all circumstances given these criteria. In fact, often times we observe the coexistence of regulation and tort liability. In short, we should expect society to deal with potential harms via regulation when the perpetrator either cannot afford to fully compensate the victims or faces a sufficiently low probability of a lawsuit. On the other hand, we should expect society to deal with potential harms via tort liability when private actors have more information about potential risk and when administrative costs are the primary concern.

Finally, as I just mentioned, tort liability and regulation need not be substitutes. As in our speeding vehicle example, it is possible that the solution is ultimately a combination of the two. Liability and regulation can be complements.

Susan Rose-Ackerman has specified conditions under which these potential alternatives can also serve as complements. In particular, she notes that tort liability can act as a stopgap when regulation proves inadequate. Alternatively, regulation might be used to create some minimum standard of care to prevent harm from being done. This minimum might be able to limit harm in most cases. However, in the cases that it doesn’t prevent, lawsuits can be used to seek restitution for damages ex post. This is true for our speeding example — and consistent with what we actually observe.

There is a broad lesson here. As with everything we write about, costs vary along a wide variety of margins. As a result, the choices that we make when it comes to tort liability or regulation or some combination of the two should ultimately come down to identifying the relevant margins on which to optimize.