February 3, 2013
Rules versus Discretion
Why I Favor Rules Over Discretion-Becker
Posner shows there is no absolute distinction between rules and discretion (and standards) because most sensible rules allow for discretion in extraordinary cases. Yet there is a difference under less than extraordinary circumstances, and rules are better in the great majority of circumstances when they are feasible.
One important advantage of regulatory rules partly stems from the nature of human behavior. When the great majority of businessmen and consumers are behaving in a particular way, such as the frenzied borrowing and lending on housing and other assets that preceded the financial collapse, it is extremely difficult for regulators to stick their necks out and shut down the party. As a result, SEC and other bank regulators took no extensive actions to rein in banks even though their asset to capital ratios had increased into dangerous levels before the financial bubble burst. Both Fed heads Alan Greenspan and Ben Bernanke in the early stage of the financial crisis said it posed little threat to the real economy (of course, Bernanke later radically changed his view).
Even when regulators can resist getting caught up in the enthusiasms of the businesses and consumers that they are regulating, they are still subject to enormous political pressures to go along with prevailing behavior. Many companies and industries have full time lobbyists who spend their time cultivating and trying to influence the officials who are regulating their clients. Consumers are not as well organized, but they have votes, and that makes regulators reluctant to take actions that reduce the availability of credit or goods and services that consumers want. To take one of many examples, Congressmen and lobbyists put tremendous pressure on Fannie Mae and Freddie Mac to continue to take on risky subprime loans up to the collapse of the mortgage market.
The pressure from interest groups on government officials has long been noted. Simon Newcomb, a 19th century American astronomer and economist, wrote in 1885, "One cent per year out of each inhabitant would make an annual income of $500,000. By expending a fraction of {their} profit, the proposers of policy A could make the country respond with appeals in their favor…Thus year after year every man in public life would hear what would seem to be the unanimous voice of public opinion on the side opposed to the public interests"(Principles of Political Economy).
Rules set the tone for regulatory actions even though they must allow for discretion under unusual circumstances. If bank regulators had a rule requiring that the ratio of bank capital to bank assets had to be above a specified threshold, they very likely would have been forced to take actions that limited bank lending prior to the financial crash, even when they had misgivings about these actions or when subject to political pressures to let the financial party continue. This is why I favor such capital requirements, and even steeper ones for the banks that are "too big to fail".
If the Fed had been following a rule that made its intervention dependent on some combination of inflation, output growth, and unemployment- as in Taylor rules- it probably would not have intervened before the crisis hit since all these measures were doing well. But a Fed following such a rule would have intervened quickly after unemployment began to rise sharply and output growth slowed dramatically.
The Dodd-Frank Wall Street Reform and Consumer Protection Act does require capital requirement rules for banks, and introduces other rules as well. However, it offsets these good rules with various new and dubious discretionary powers of regulators. These include the creation of a consumer protection agency with potentially considerable power over consumer actions. It also created the Financial Stability Oversight Council, a panel drawn from the Fed, SEC, and other government agencies, that monitors Wall Street's largest companies and other market participants in order to respond to what they assess is excessive growth in the economy's degree of systemic risk.
I am not arguing that rules alone are always sufficient to guide regulators and other government officials. Clearly, at times rules have to be supplemented to take account of circumstances that were not adequately covered by the rules. Also rules have to be adjusted periodically over time as new institutions and technologies develop, and as those being regulated devise ways to get around the rules. In addition, regulators subject to rules may manipulate the implementation of any rules. Nevertheless, we are likely to get far better policies if we begin with a widespread set of rules that guide officials, and then introduce discretion as supplements to this rule-based system of regulations, rather than supplementing a basically discretionary approach with some rules.
Rules versus Discretion in Financial and Other Regulation—Posner
There is an academic literatureon the choice between rules and discretion in regulation. Milton Friedman famously advocated that the Federal Reserve be required to increase the money supply by a fixed annual rate, rather than the Fed's being allowed to exercise its traditional discretion over the money supply. But the choice between rules and discretion is not limited to the money supply, or indeed to financial regulation.
It helps to distinguish not just between rules and discretion, but between rules, standards, and discretion. A posted speed limit is an example of a rule; liability for negligent operation of a vehicle is an example of a standard; and the President's power of appointment of the members of the White House staff exemplifies the operation of discretion. In fact, though, the administration of rules and especially of standards typically involves a degree, often a large one, of discretion. Police officers don't ticket all speeders, but only those who drive "too far" above the speed limit; and legal standards such as negligence tend to be vague and so operate to grant considerable discretion to judges and juries. So the real issue in choosing among rules, standards, and discretion is how tightly to rein in the exercise of discretion by persons exercising government authority (including pro tem. government officers such as jurors).
The fact that discretion operates even in the enforcement of precise rules (such as speed limits) is a clue to the limitation of rules. Literal enforcement of rules is unworkable because rulemakers can't anticipate the full range of circumstances to which the rules they promulgate will apply. For the essence of a rule is to abstract, in the interest of clarity and precision, from all relevant circumstances, a few circumstances to be determinative. Clarity and precision are important values. "65 m.p.h. speed limit" conveys more information (though not complete information, because the de facto speed limit may be considerably higher), more clearly and concisely, than "don't drive faster than conditions permit" does.
The choice among rules, standards, and discretion is especially controversial in the law. Rules limit judicial discretion and provide crisp guidance to persons subject to them, while when judges are exercising discretion, as when they "interpret" extremely vague constitutional or statutory provisions, they seem to be stepping out of the proper judicial role and acting as legislators. But really they have no choice. The first duty of a judge is to decide. He can't refuse to decide a case on the ground that he lacks guidance in authoritative legal materials that are sufficiently clear to dictate the decision.
The situation with respect to the role of discretion is similar when we switch focus from the judiciary to the regulatory process, as illustrated by the Federal Reserve. The Fed's principal power is to alter short-term interest rates (though often with an effect on long-term rates) by buying and selling federal securities. If it sells, the cash it receives for the sale reduces the amount of money in the economy and so reduces interest rates, which ration the demand for money rise. If the Fed buys securities, it pours cash into the economy and interest rates fall. Lower interest rates stimulate borrowing and hence spending, unless the rates get so low that no one wants to lend. Also, the larger the money supply, the cheaper the currency is relative to foreign currency, which makes debt held by foreigners cheaper to repay and stimulates exports. Which is one example of the broader point that the larger the money supply is, the greater the risk of inflation (because the ratio of money to goods and services is higher). So creditors, who suffer from inflation, don't want the Federal Reserve to increase the money supply, while debtors do. W the economy is depressed, cheap money encourages debtors to spend, by reducing their liabilities, and money borrowed because interest rates are low is largely spent, thus increasing consumption, production, and employment.
Obviously there are huge political and economic stakes in actions by the Federal Reserve, and the Fed is little constrained by rules or standards (though I'll give an example of where it is constrained), and so exercises very broad discretion. Its mandate requires it to pay attention to both inflation and employment; increasing the money supply can stimulate both, and reducing it depress both, and the economic consequences are very different.
The Fed's discretion leads to a concern, fed by the traditional hostility that many people feel toward banking in any form (banking is one of the few businesses which refuse to sell to people willing to pay its price, if they are not good credit risks, though lessors, such as rental car companies, unlike sellers, are often picky about whom they rent to, lest the rented property be damaged), that the Fed's decisions on increasing or reducing the money supply are influenced by politics. If the Fed refuses to be influenced, the President or Congress may try to curtail its discretion and by doing so reduce its power and maybe harm the country in the long run; politicians typically have a short-run perspective because of their limited terms of office. Milton Friedman's concern, however, was more with the unpredictability and doubful competence of the Fed's money-supply decisions.
Having learned from our recent, unhappy economic experience that depressions are not a thing of the past, I find it hard to accept the proposition that the Fed should be subjected to tight rules. Had it not been for the Fed's massive purchases, which greatly increased the money supply, of mortgage-backed securities during the financial crisis that began in the fall of 2008, the banking industry might well have collapsed and credit been completely frozen, precipitating a 1930s-style deep depression. Of course this means that any Friedmanite rigid rule governing Fed actions would quickly have been lifted by Congress.
Let me give an example that illustrates both the difference between a rule and a standard and the hazards of trying to limit the exercise of discretion by the Federal Reserve. It is widely believed (including by me) that the failure of the Fed, in September 2008, to bail out Lehman Brothers precipitated the world-wide financial crisis that ensued immediately. Fed Chairman Bernanke has said that the law forbade the Fed to bail out Lehman Brothers. I am dubious. Section 13(3) of the Federal Reserve Act authorizes the Fed to lend money to a nonbank (Lehman was a nonbank bank, which is to say a financial company that provided banking services but was not regulated as a bank) in "unusual and exigent circumstances," provided that the loan is "secured to the satisfaction of the Federal reserve bank." Lehman did not have good security for the large loan it would have needed in order to survive, but in the emergency circumstances created by a collapsing global financial system the Fed could have declared itself "satisfied" with whatever security Lehman could have offered. For section 13(3) establishes a standard rather than a rule, and standards (even rules, as I said) usually don't extinguish discretion. But they cramp its exercise. It would have been better had the statute just said that the Fed could lend to a nonbank if unusual circumstances justified such a loan—a looser standard, allowing more play for discretion.
Finally, rules do not really eliminate or even seriously constrain discretion, but rather merely shift the way in which, or the level at which, discretion is exercised. A regulatory agency, having issued a rule, will usually have discretion to amend or rescind it; and if Congress imposes a rule on an agency, Congress always has discretion to amend or rescind the rule.