My post yesterday argued for eliminating the Securities and Exchange Commission. I wrote:
Eliminating the SEC would make investors bear full responsibility for monitoring corporate behavior. This occurs to a substantial degree already, since the SEC cannot effectively monitor all the firms subject to its regulations. But eliminating the SEC would spur additional private monitoring and strengthen investor incentives to engage in due diligence.
Greg Mankiw has responded with skepticism, arguing that my logic also suggests eliminating police. Greg wrote:
Eliminating the police would make people bear full responsibility for protecting their own safety. This occurs to a substantial degree already, since the police cannot effectively monitor everyone who might commit a crime. But eliminating the police would spur additional private monitoring and strengthen individual incentives to engage in due diligence.
Greg is right that the structure of my argument applies to government efforts to prevent ordinary crime as much as to government efforts to prevent white collar crime like CEO fraud or embezzlement. Thus, consistency might seem to suggest drawing the same conclusion about both types of government policies.
Many libertarians respond by opposing both government attempts to deter white collar crime and government efforts to deter ordinary crime. And they use the kinds of arguments that I suggested for white collar crime and that Greg has mimicked and applied to ordinary crime.
I have two responses. To begin, the libertarian position on government and ordinary crime is not as bizarre as most people assume. The scope for private crime deterrence is substantial (locks, alarm systems, guard dogs, guns, private security personnel, gated communities, and so on). The evidence on whether police deter crime is mixed at best. And some current police activities increase rather than decrease crime (prohibitions against vice).
That having been said, I am not convinced that eliminating police is the right policy. But that does not change my view on the SEC because ordinary crime and white collar crime differ in at least one key respect.
In the case of ordinary crime, the victim has not entered into a relationship with the perpetrator of the crime. In the case of white collar crime, the victims are the shareholders, and they willingly purchased the stock of the company in question. That is, shareholders entered into a contract with the corporate executives; the shareholders gave the corporation money in exchange for shares, and the executives agreed, at least implicitly, to maximize the value of those shares.
Thus when CEO’s commit fraud or embezzlement, the shareholders have grounds for complaint, but they are not victims in the same sense as victims of a mugging or a robbery. Shareholders have more than ample opportunity to design CEO contracts that make fraud difficult and that punish CEO’s for transgressions (loss of pension, stock options, etc). Shareholders can hire private auditors to watch over CEOs.
Under this “contractual” approach to disciplining CEOs, government still plays a role in limiting CEO misconduct by providing courts that resolve contractual disputes between shareholders and CEOs. But they do not criminalize CEO misconduct, and they do not independently pursue white collar “crime.”
Reasonable people can debate whether the private, contractual approach to disciplining CEO misconduct will be sufficient without the threat of criminal penalties. But at a minimum, the moral claim of a murder victim is different than that of someone who bought Enron stock. The later knowingly assumed the risk; the former (usually) did not.
My bottom line, therefore, is that all government attempts to criminalize private actions have a range of costs and possibly some benefits. The balance differs dramatically depending on the private action in question.
At one end of the spectrum, criminalizing murder is almost certainly the right approach, but even this has drawbacks (the direct costs of police, prosecutors and prisons, plus any crowding out of private efforts to avoid getting murdered in the first place).
At the other end of the spectrum, criminalizing victimless crimes like drug use, prostitution, or gambling has at best modest benefits in reducing an undesirable activity, and it generates substantial costs.
So the question of what to criminalize should take into account a range of issues, including whether the “victims” knowingly assumed the risks in question, whether criminalization itself has substantial costs, and whether private actions can reasonably address at least most of the negatives associated with the action in question. My assessment is that for ordinary crime, the case for criminalization is probably convincing. For white collar crime, the costs of criminalization likely outweight the benefits.