The power to influence money's value is a public power exercised through private decisions. It enters the price of a loan, the prospects of a business, the real burden of a debt and the terms on which a worker can find employment. The preceding plate traced the connections without making the institution omnipotent: much depends on expectations, contracts and events beyond its command. Yet uncertainty about the effect makes responsibility more demanding. Someone must choose the aim, authorize the instruments and decide how the judgment will be examined. Central bank independence addresses one part of this arrangement. It protects a sphere of decision from immediate political direction. Its democratic justification depends upon the purpose for which that protection was granted and the means by which the institution remains answerable for using it. The word independence describes a relation within government, whose terms must be specified.
The distinction between goals and instruments begins that specification. A legislature can establish the ends of policy while delegating the choice of interest rates and other authorized means. The Federal Reserve Act, for example, directs the Board and the Federal Open Market Committee to promote maximum employment, stable prices and moderate long-term interest rates. Those statutory ends do not themselves supply a rate for tomorrow's meeting. They require interpretation and judgments about feasible paths. Stanley Fischer distinguishes this instrument independence from the freedom to determine operational goals, such as the numerical meaning attached to price stability. Different arrangements can place these decisions with different institutions. A broad instruction written in law can therefore coexist with substantial discretion over both its implementation and its practical interpretation. To ask whether a bank is independent is only the beginning. The more revealing questions are which decisions are delegated, under which purposes, and with what obligation to explain the choices.1
A longer horizon supplies an important reason for delegation. Policy may impose costs before its benefits arrive, while an election rewards an earlier appearance of success. The argument of The Temptation identifies a further difficulty: under its stated incentives, an announcement made before wages are chosen can be tempting to abandon afterward. Expectations of that abandonment can defeat the promise in advance. Insulation from immediate pressure can help sustain a commitment, especially when the pressure is to pursue an employment gain that inflation cannot permanently deliver. The institutional conclusion remains conditional. Giving an official a longer term does not remove every incentive to revise a plan, confer correct knowledge of the economy, or ensure that the official pursues the public's purposes. A credible arrangement requires a reason to expect the delegated authority to act as promised. Legal distance from the government of the day can contribute to that reason; the mandate and the conduct of the institution must contribute too.2
The weight placed on competing costs remains a social choice even when the forecast is agreed. Consider an illustrative decision between two feasible policies. One leaves inflation one percentage point above its objective and output two percent below its reference level; the other leaves inflation two points above and output one percent below. Suppose the criterion is the squared inflation deviation plus a weight λ on the squared output deviation. The first policy costs 1 + 4λ; the second costs 4 + λ. At λ = one-half the first costs three and the second four and a half. At λ = two the ranking reverses, to nine and six. At λ = one they tie. No disagreement about the forecast is needed to produce disagreement about the decision. Nor does this small example establish which weight is right. It shows why expertise about consequences cannot, by itself, determine how heavily different consequences should count.3
Accountability makes the exercise of delegated judgment available for examination. A mandate gives outsiders a purpose against which to compare conduct; explanations must then connect the instruments chosen to that purpose. Published forecasts, reasons for revisions, records of disagreement and answers to elected representatives can expose the institution's understanding of the alternatives. Their value depends on whether they permit a decision to be challenged, rather than merely announce that a decision occurred. A missed target can reflect a poor judgment or an unforeseen disturbance, and a favorable result can conceal a poor judgment rescued by events. The record should allow those possibilities to be distinguished. Equally, repeated surprise cannot become an indefinite excuse for refusing to examine the model. Independence can protect the decision while scrutiny tests the reasons offered for it. The balance is an institutional design problem, requiring both room to act and an intelligible account of the action.4
The boundary becomes especially consequential when monetary operations carry fiscal risks or select among particular recipients. An asset purchase can influence financial conditions while also changing the public sector's exposure to gains and losses. Emergency lending raises questions about collateral, repayment and who bears a shortfall. The fact that a central bank conducts an operation does not settle the proper allocation of authority over every consequence. Bernanke's defense of independence expressly distinguishes monetary decisions from other functions and calls for clarity at the fiscal boundary. That distinction also prevents an easy confusion about distribution. Ordinary monetary policy can affect borrowers, savers, workers and asset owners differently without becoming a program of targeted transfers. Its distributional consequences still belong in the public account of what it does. The relevant questions concern the authority granted, the risks accepted and the reasons for choosing this action over another—not whether an economically consequential instrument can be made to affect everyone identically.5
The mandate closes the volume by locating monetary authority within the society whose money it governs. Voters can authorize institutions that pursue durable purposes without directing every instrument setting. The delegation must still have content: ends that can be identified, powers whose scope can be understood, and reasons that can be examined. Insulation earns its justification through that arrangement. It cannot turn an official's preference into a public purpose merely by sheltering the official from pressure. Nor does accountability require every judgment to be remade at the instant it becomes unpopular. The difficult task is to preserve a commitment while keeping its interpretation and use open to serious judgment. The value of the common measure reaches every contract, but it does not decide how the product measured by it is divided. Having followed the authority over money's worth, the series turns to that division: the part received by labor, the part received by ownership, and the institutions through which each is claimed. The next volume is The Share.