By Peter Lewin, a professor of mine to whom I owe very much
A Statement of Basic Principles
Regulation refers to a set of activities, government by regulatory laws, known as regulations, some of them deriving from Congress and implemented by the Executive, the President, and some deriving directly from the Executive (executive orders) – but both are implemented by the Executive. The people, human beings like us, who carry out these activities of regulation are called regulators. They are a large group.
In every case, without exception, the implementation of regulation is subject to four basic, and closely related, problems. Each one by itself is formidable. As we go through the various types of regulation, we need to keep them in mind:
Costs and Benefits
Associated with each set of regulations is a set of stated, or understood, objectives. There are presumed benefits that follow from the achievement of these benefits (though it is important to ask to whom these benefits will go). There may also be benefits that are not obvious and may be unintended. Much less emphasized, however, there are always costs associated with the implementation of regulations. Some of these costs are stated, most are not, and many are unintended. The costs are almost always underestimated and are frequently greater than the benefits.
We are leaving aside here the question of how to value the costs and benefits. Strictly speaking there is no objective way to do this because value is subjective and cannot simply be added up across people. We follow the usual practice of assuming that market prices and costs can be used.
Type I / Type II errors
All regulation entails actual or potential coercion, intrusion into the private decisions of individuals. As such there are only two justifications for regulation of private activities:
The implementation of any regulatory policy is fraught with uncertainty. Often the scientific principles that underlie the regulation are unproven. Knowledge is continually changing. As a result, there are always errors. No single regulatory policy is exempt from the committing of errors. We may divide these errors into two categories.
Type I: Failure to achieve the prime objective of any set of regulations can be thought of as a Type I error. It is the error we, as a matter of policy, want most to avoid. So, for example, in the case of medicines, it is the error of allowing an unsafe, or unproven, drug to be used. In the criminal justice system, a Type I error is the error of convicting an innocent person.
Type II: In order to avoid these errors, regulatory policy implicitly commits other errors – like the error of failing to approve a safe drug for use by people who might benefit from it. These implicit, often invisible, errors can be called Type II errors. In the criminal justice system, a Type II error is the error of failing to convict a guilty person.
It is important to note that the more Type I errors we avoid, the more Type II errors we will commit, and vice versa. See if you can explain this to yourself using examples. Also remember, Type I and II errors are not given to us from the nature of the situation. They are categories that we, the policy-makers, impose on the situation to guide our actions. They reflect our values and motivations.
Knowledge and Incentive Problems
Consider again the first justification for regulation listed above - that people are too stupid and uninformed to be allowed, in every case, to make their own decisions, and mistakes. Certainly this, in itself, is not a self-evident principle, and it is certainly not consistently applied – you can think of many exceptions where people are left to make their own, good or bad, decisions. And we should wonder how it is that the regulators are in a better position, and can be trusted to make the right decisions for those being regulated.
In the last few decades, a new approach to this question has arisen under the umbrella of a branch of economics known as “Behavioral Economics”. Behavioral Economics investigates cases of individual actions that are said to be mistaken in the sense that they are “irrationally” not in the individual actor’s own interests. A huge literature documents cases where individual actions can be shown to be inconsistent with a given set of preferences, or inconsistent with what economists identify as required norms of rationality, taken from the discipline of “Neoclassical Economics”. This has given rise to numerous policy prescriptions identified as the “New Paternalism” – policies designed to encourage (nudge) individuals to act in a way that is more in tune with their “true” interests.
This raises the question of the nature of the regulators themselves. Who are they? What motivates them? Are they superhuman, super-knowledgeable, or are they just like the rest of us? In this course we shall assume that regulators are just human beings, like us, with their own particular self-interests and limited knowledge. This means that every single regulatory action is subject to two kinds of difficult problems, incentive and knowledge problems.
Transferring a set of objectives from the private to the public sector does not make it more easily achievable. In fact, the opposite is most often true.
Burden of Proof
Finally, considering all of the above, there is the question of how we decide upon regulatory policy in the first place. Merely identifying a potential need for any regulation is insufficient. One must consider all of the costs and benefits, including the costs of errors that will be committed, and the knowledge and incentive problems that are involved in implementing it. These will of course be uncertain. We cannot know ahead of time what kinds of errors will be committed, how damaging they will be, and so on. We cannot know how easy or difficult it will be to incentivize those employed to implement the policy or how difficult it will be for them to know what they need to know to do so.
Accordingly, whether or not the policy is adopted, and further implemented or maintained, depends crucially on where we put the burden of proof to justify it. This is itself a Type I/II situation. Who should have the burden of proof, those who propose a particular regulation because of certain benefits that they believe will follow its implementation, or those who oppose it because of the intrusive nature of all regulation and the costs they believe are associated with this particular one. We cannot avoid this question, and where we place the burden of proof, will very much determine how many and what kind of regulations we end up with.