“[W]hile many democratic societies have independent central banks, everyone leaves tax policy in the hands of elected officials.” Alan S. Blinder, Central Banking in Theory and Practice, page 59.
Through what administrative means should a democratic society in an advanced economy implement regulation? How can we promote the various interests of the people efficiently, while being faithful to the evolving wishes of the majority? Ultimately, these are questions of legitimacy. However effective a form of administration may be, its sustainability in a democracy depends on maintaining the trust of the people.
In practice, democratic governments opt for a variety of solutions to this challenge. Most enact detailed laws that define crimes and punishments (implemented by the executive and judicial branches) as well as create agencies that carry out the wishes of the executive branch, within the rule of law. These approaches earn their legitimacy by allocating power to elected officials who make the laws or directly oversee their agents.
Increasingly, however, governments have chosen to implement policy through agencies with varying degrees of independence from both the legislature and the executive. This approach leads to a deep and important question: Under what circumstances does it make sense in a democracy to delegate powers to the unelected officials of independent agencies (IA) who are shielded from political influence? How should those powers be allocated to ensure both legitimacy and sustainability?
These are the critical issues that Paul Tucker addresses in his ambitious and broad-ranging book, Unelected Power. The subtitle—“the quest for legitimacy in central banking and the regulatory state”—highlights central banks as the prime example of IAs (as the opening citation suggests). Given the new powers that central banks have accumulated over the decade since the Great Financial Crisis (GFC)—ranging from balance sheet flexibility to regulatory authority—and the heightened public sensitivity to their actions, Tucker’s focus could hardly be more timely.
In addition to suggesting areas where delegation has gone too far, Tucker highlights others—such as the maintenance of financial resilience (FR)—where agencies may be insufficiently shielded from political influence to ensure effective governance. His analysis raises important questions about the regulatory framework in the United States, where the structure remains absurdly balkanized, especially by comparison to the streamlined mechanisms in the United Kingdom.
In this post, we briefly discuss Tucker’s principles for delegating authority to an IA (see Appendix below). As illustrations, we consider how well three U.S. agencies conform to those principles, identifying areas where the delegation of authority might be narrowed or expanded to promote both legitimacy and effectiveness. A key premise—that we share with Tucker—is that better governance can help substitute where simple policy rules are insufficient for optimal decisions.
What is an IA? Tucker defines an Independent Agency as a “public agency that is free to set and deploy its instruments in pursuit of a public policy goal (or goals) insulated from short-term political considerations, influence, or direction” (page 11). In the language of central banking, IAs enjoy instrument independence, but not goal independence, with their goals set by legal mandate (see Debelle and Fischer). Within that remit, the agency’s decisions are irreversible by others in government. Furthermore, the law shields the IA from various forms of influence, such as the policy pressures that may accompany frequent legislative budget review or the turnover of personnel.
Principles of Delegation. With efficiency, democratic legitimacy and sustainability in mind, when does it make sense for a government to delegate significant authority to an IA while shielding it from day-to-day political influence? Tucker proposes seven conditions that he views as necessary and sufficient. These delegation criteria (DC; see Appendix below) constrain the nature of what should be delegated and the circumstances under which delegation is likely to be more successful than execution by a simple agent acting directly at the behest of the executive.
Tucker rejects a number of historical arguments for delegation as inconsistent with legitimacy and sustainability. Above all, because experts can and do advise elected officials who then make the decisions, the need for technical expertise (frequently associated with Wilsonian reforms and Roosevelt’s New Deal entities) is not a sufficient basis for independence. Instead, Tucker focuses on those matters where a credible commitment by an IA can achieve a better outcome than leaving decisions to elected officials with their invariably short horizons (DC3). Readers familiar with the arguments for central bank independence will immediately see this as an instance of overcoming the problem of time consistency (in which a future incentive and opportunity to renege on a policy commitment undermines current policy effectiveness).
To ensure that delegation goes only as far as needed to be effective in achieving the stated social objective, Tucker’s criteria impose a range of further limits:
- The mandate―that is the result of a broad public debate―should be specified in detail (DC1), preferably with one goal. Or, where multiple goals are mutually reinforcing and require commitment, there should be a clear hierarchy of importance (see Multiple-Mission Constraints in the Appendix).
- Society’s preferences with regard to a major social cost (DC2)—as well as the policy regime for addressing it (DC3)—should be reasonably stable.
- The instruments for achieving the goal(s) should be reliable (DC4), with an external community of experts available to assess the IA’s performance and propose improvements.
- Consistent with the opening quote, the IA’s decisions should not primarily address issues of distribution, which must be determined by elected officials, or significantly alter the balance of political power (DC5). Even the unintentional impact of IA decisions on distribution should be limited.
- For the legislature to hold the IA publicly accountable, its choices and outcomes must be monitorable (DC6).
- Finally, to overcome the risk of reneging, IA officials must value their reputations sufficiently to bind them to their mandates (DC7).
Moreover, the IA should be legitimacy-seeking: that is, it should welcome ongoing public review of its performance, and act transparently to foster that review. As in the case of judges, IA officials should exercise self-restraint, limiting their role in public issues outside of the IA’s remit. Finally, Tucker details five major design precepts for structuring an IA and organizing its activities that he sees as critical for effectiveness, legitimacy and sustainability (see Appendix).
Applying Tucker’s Delegation Criteria. Tucker’s delegation criteria and design precepts are a set of principles that governments can use to determine when and how to delegate administrative authority. These principles are sufficiently broad to be applicable across a wide range of government responsibilities, well beyond economics and finance. Focusing on our expertise, we illuminate the practicality of Tucker’s criteria by seeing how they apply to three U.S. financial-sector agencies that have very different remits and structures: the Federal Reserve System, the Securities and Exchange Commission (SEC), and the Consumer Financial Protection Bureau (CFPB) (see table below). To be useful, the criteria should help us judge whether these agencies have the proper degree of shielding from day-to-day political influences, or how they might be altered to make the policy framework more effective and sustainable.
Applying Tucker’s Delegation Criteria to the Federal Reserve, the SEC and CFPB
Starting with the Federal Reserve, the central bank has a mandate with multiple, non-hierarchical goals. The Fed has publicly interpreted its monetary policy mandate in a balanced way that nevertheless focuses on its ability to keep inflation low and stable (while noting that non-monetary influences dominate the long-run evolution of growth and employment). That interpretation helps make both the Fed’s choices and outcomes monitorable. With regard to financial resilience (FR), the Dodd-Frank Act broadened the Fed’s mandate substantially, but the central bank’s actions in this realm are inherently more difficult to monitor and assess. First, the law does not state the financial resilience goal in a way that quantifies social preferences (for example, as Tucker suggests, by limiting the expected frequency of a GFC-like crisis over a specified time period or, as we have proposed, utilizing a measure like GDP at Risk). Second, the information used in ensuring financial resilience tends to be institution-specific, and therefore proprietary, limiting the potential for transparency and accountability compared to monetary policy. Third, given the cross-border integration of global financial markets, Fed (and other U.S. regulatory) officials must work in closed settings with foreign regulators and international standard-setting agencies.
Both price stability and financial resilience are enduring elements of U.S. preferences. While the tools and framework for setting conventional monetary policy are constantly evolving, they do so within a reasonably circumscribed range. However, unconventional monetary policy tools are less well understood, and may have greater distributional consequences. Similarly, with the tools for securing financial resilience still in development, we remain far from being able to reliably anticipate the quantitative impact of macro-prudential policy actions.
In our view, despite this mismatch between current Fed practice and Tucker’s delegation criteria, the time consistency arguments for central bank independence remain compelling, both with regard to price stability and growth on the one hand and financial resilience on the other. Yet, as the design principles suggest, there can be little doubt that the threats to Fed legitimacy grew sharply with its role as lender of last resort (LOLR) in the crisis and its reliance on unconventional tools since that time.
Consequently, we can think of several changes that, consistent with Tucker’s delegation criteria, would go a long way to securing the long-run legitimacy of the Fed as an IA. These include: clarifying both its obligations and limits as LOLR in a way that make those policy actions more systematic (for example, no lending to knowingly insolvent organizations); emphasizing the importance of the evolving resolution regime for insolvent intermediaries to address the too-big-to-fail problem; establishing clear principles for the use and limits of unconventional monetary tools (such as large asset purchase programs), with greater restrictions on those that are likely to have more significant distributional consequences; and a better understanding of the relationship between key macroprudential tools (such as capital requirements, liquidity regulation and stress tests) and the ultimate goal of maintaining financial resilience.
Turning to the Securities and Exchange Commission, for most of its history the SEC focused on ensuring “fair, orderly and efficient” capital markets, much like Britain’s Financial Conduct Authority does today. Its principal tool was to require and enforce disclosure (including, for example, the profit-and-loss statements and balance sheets of corporations that issue public securities). Its mandate was broad and flexible, providing considerable leeway that the SEC exercised. For example, while Congress made financial fraud a crime in the 1934 Act, it was only through actions decades later that the SEC criminalized insider trading (Tucker, page 314).
The SEC’s historic focus on law enforcement did not pose a serious time consistency problem. Not coincidentally, it remains substantially less independent than the Fed. For example, unlike the Fed’s budget, which is insulated from legislative and executive oversight, annual review of the SEC’s budget provides Congress an avenue for influencing SEC policy choices (see, for example, page 25 here). In addition, the openly partisan alignment of individual Commissioners, combined with the “informal convention” of chair resignations at the advent of a new Presidential Administration (Tucker, page 142), provides ample means for the government to influence SEC rule-making.
Post-crisis, however, market regulators, including the SEC and the Commodity Futures Trading Commission (CFTC), have a substantial role in ensuring financial resilience, greatly increasing the importance of credible commitment (for example, in setting collateral and margin requirements, limiting the risks of shadow banking, overseeing central clearing parties that are too big to fail, coordinating with foreign regulators, and even in sharing sensitive market information with other key regulators). Tucker provides a glimpse of the time consistency problem for 21st century capital market regulators: “[I]t might suit politicians to allow exuberance in asset markets if that improves the feel-good factor and eases the supply of credit to voters and donors” in the short run (page 144).
However, it may be difficult to shield market regulators from day-to-day political influence when much, if not most, of their activity remains traditionally oriented toward the maintenance of fair and orderly markets. This raises the question of whether the U.S. regulatory framework should be re-structured—as the U.K. system was in 2012—in part to separate financial resilience mandates, with their need for credible commitment, from functions ensuring market fairness, efficiency, and the like.
Finally, we come to the CFPB. Created by the Dodd-Frank Act, the Bureau has had only one Senate-confirmed Director since it came into operation in 2011. However, in contrast with the matters that Tucker views as reasonably safe to delegate, experience shows that consumer protection is a matter of intense political sensitivity. (Tucker points to the CFBP as an example of over-delegation.)
Given the CFPB’s brief history, neither the policy regime nor the tools for implementing it can be thought of as settled. It also is not clear on precisely what grounds a legislative monitor should assess the Bureau’s performance. From this perspective, Dodd-Frank’s shift of consumer protection away from the Fed probably helps to secure the central bank’s independence. However, that only highlights the question of how independent the CFPB should be from political oversight. With regard to budget independence, it is at one end of the spectrum: aside from the Fed, it is the only agency whose budget is virtually self-determined and funded directly out of the central bank’s seignorage. Nevertheless, as we have seen, the 2016 election has led to sharp politically-driven changes at the CFPB, with the President’s Director of the Office of Management and Budget taking over as the Acting Director.
Conclusion. So, where does all this leave us? In the past, we have argued repeatedly for streamlining the U.S. financial regulatory apparatus as a necessary condition for ensuring resilience. Tucker’s principles could serve as a guide for such a restructuring. Unfortunately, we see little appetite for this from the legislative or executive branches. If anything, current efforts are more like to reduce, rather than enhance, financial resilience (see, for example, here and here).
We can only hope that a future U.S. government will recognize the need for change and seize on Paul Tucker’s principles and design precepts to promote a more effective, legitimate and sustainable financial regulatory system.
APPENDIX: THE PRINCIPLES FOR DELEGATION TO INDEPENDENT AGENCIES INSULATED FROM DAY-TO-DAY POLITICS
A public policy regime should be entrusted to an independent agency (IA) insulated from the day-to-day politics of both elected branches of government only after wide public debate and only if
1. The goal can be specified.
2. Society’s preferences are reasonably stable and concern a major social cost.
3. There is a problem of credibly committing to a settled policy regime.
4. The policy instruments are confidently expected to work, and there exists a relevant community of experts outside the IA.
5. The IA will not have to make big choices on distributional trade-offs or society’s values or that materially shift the distribution of political power.
6. The legislature has the capacity, through its committee system, properly to oversee each IA’s stewardship and, separately, whether the regime is working adequately.
7. The society is capable of bestowing the esteem or prestige that can help bind the IA’s policy makers to the mast of the regime’s goal.
1. Elected legislators should provide a statement of purpose, objectives, and powers, and a delineation of the regime’s boundaries (Purposes-Powers). In particular,
a. The objective or standard must be capable of being monitored.
b. If there are multiple objectives, they should be lexicographic.
c. Any statutorily mandated objective, standard, or prescribed instrument rule must be understood by legislators and broadly comprehensible to the public.
d. An IA’s rule making should not interfere with individual liberal rights more than necessary to achieve the legislated purpose and objective (proportionality).
e. An IA should not be able to create or frame criminal laws or bring criminal prosecutions.
f. Its sanctions should not include ruinous fines.
g. Subject to the above, the agency’s policy makers must be in control of the delegated policy instruments, and so must have statutory job security and not be subject to frequent political budgetary approvals (necessary conditions for insulation). Thus,
i. any powers of override by elected politicians should be explicit in statute and, if exercised, should require transparency to the legislature and the public.
ii. any provision for the IA to seek special approval for uses of powers outside emergencies (see DP5) should apply only where a political judgment is needed on whether a statutory regime’s purpose can best be pursued by setting aside constraints framed as monitorable standards; it should be exercisable only in specific cases, and its use should be disclosed as soon as is safe to do so.
2. There should be clear prescriptions of who will exercise the delegated powers and of the procedures to be employed in delegating and exercising them (Procedures). In particular,
a. The policy-making body should be deliberative, with a voting committee of equal members (one person, one vote).
b. Terms of office should be long and staggered; and appointments should be subject to a dual key, so that a single elected politician cannot appoint a committee of allies.
c. Undue concentrations of power within agencies should be avoided.
d. The agency’s processes must help deliver the values of the rule of law in agency rule making, adjudications, and other actions.
e. Within a rule-writing agency, the structure for determining (adjudicating) individual cases should have degrees of separation; and each distinct phase of policy making should have its own integrity.
3. The IA should publish principles for how it plans to exercise discretion within its boundaries (Operating Principles ).
4. There should be sufficient transparency to enable the stewardship of the delegated policy maker and, separately, the design of the regime itself to be monitored and debated by elected representatives (Transparency-Accountability ). In particular,
a. An independent agency should contribute to those public debates with information and research on how it evaluates the effectiveness of its instruments, and on the social costs of the ills it is mandated to mitigate.
b. An IA should publish data that enables ex post evaluation of its cost-benefit analysis and other forecasts, and more generally to enable independent research.
c. An IA involved in international policy standards should do what it can to ensure that plans and questions are properly exposed domestically.
5. There should be provisions determining what process is to be followed when the boundaries of the regime are reached during a disaster, including how democratic accountability works then (Emergencies). To be subject to the following constraints:
a. An IA should seek to consult key elected representatives before taking measures that, formally, are within its legal powers but were never remotely contemplated by legislators and the public.
b. Any decision to extend an IA’s powers should be taken by the legislature or by the elected executive using delegated powers.
c. It should be completely clear whether the IA’s powers are being reset for a period with independence maintained or are being both extended and subjected to case-by-case political review.
d. Where independence is maintained, no substantive extension or reset of powers granted by elected representatives shall be inconsistent with the purpose and objective of the regime or with any prior limits on the IA’s powers intended to be absolutely binding in all circumstances, and so should not compromise the integrity and political insulation of the IA’s core mission.
e. There shall be no de facto suspension of independence during a disaster or its aftermath without a formal legal measure transparent to the legislature and the public.
An IA should articulate extensive contingency plans for extraordinary measures that it could deploy using its standard set of powers and capabilities.
1. An independent agency should be given multiple missions only if (a) they are intrinsically connected, (b) each faces a problem of credible commitment and meets the other Delegation Criteria, and (c) it is judged that combining them under one roof will deliver materially better results.
2. Each mission should have its own objectives and constraints, consistent with the Design Precepts.
3. Each mission should be the responsibility of a distinct policy body within the agency, with a majority of members of each body serving on only that body and a minority serving on all of them.
4. Each of those policy committees should be fully informed on the debates and deliberations, as well as the actions, of the others.
Beyond the parameters of the formal regime, an ethic of self-restraint should be encouraged and fostered. Consistent with political insulation in their delegated field, IA policy makers should refrain from participating in the many other and broader issues confronting their societies.
Note: This appendix is excerpted from pages 569-572 of Paul Tucker, UNELECTED POWER: The Quest for Legitimacy in Central Banking and the Regulatory State, Copyright © 2018 by Paul M. W. Tucker. Reprinted by permission of Princeton University Press.
Disclosure and acknowledgement: We are deeply grateful to Paul Tucker for the many ideas that he has shared and discussed with us over the years.