When Interagency Conflict Is a Good Thing

One rationale for reorganizing government agencies and programs is to reduce conflict and increase coordination. The trend towards increased collaboration is happening against a backdrop of historically stove-piped programs and institutional conflict—between agencies as well as within them. Coordination is viewed as good and conflict as bad. But can interagency conflict be good?

A thought-provoking 2017 California Law Review article by Daniel Farber and Anne Joseph O’Connell documents various types of adversarial relationships that exist between agencies as well as the various mechanisms of conflict resolution.

They write: “We are reluctant to join the celebration of agency coordination, at least not without substantial qualifications. Coordination is not always desirable, or may not be desirable immediately. Conflict among and within agencies can provide substantial political, social welfare, and legitimacy benefits.”  

They note that “agencies have different institutional cultures, political allies, or policy priorities that lead to clashes” and that conflict “can be most constructive when they bring differing expertise, information bases, constituencies, and values into policy decisions.”

Their article reminded me of an interesting interagency regulatory conflict that I observed back in the early 1980s while working at the Government Accountability Office. The Occupational Safety and Health Administration and the Food and Drug Administration independently imposed conflicting regulatory requirements on chocolate manufacturers. OSHA insisted the manufacturers install sound baffling to reduce ear-damaging noise from machinery, while FDA insisted on stainless steel machinery to ensure the chocolates were not contaminated with foreign material. Eventually, the newly-created Office of Information and Regulatory Affairs within the White House Office of Management and Budget was brought in to referee the conflict.

Three Types of Interagency Conflicts  

Farber and O’Connell created a typology of three primary forms of interagency conflicts that will sound familiar to many. Interestingly, they conclude that in many of these cases conflict is “a feature not a bug” in program design. They describe three kinds of conflict relationships:

Hierarchical Conflicts. In this set of relationships, one agency is formally subordinate to another, or one agency can effectively overturn another’s decisions in at least a subset of cases. Examples include:

• OMB’s Office of Information and Regulatory Affairs has the authority to veto proposed agency regulations on cost-benefit grounds.
• The Justice Department’s Office of Legal Counsel provides authoritative legal advice to federal agencies, which agencies cannot ignore.
• Justice represents almost all executive branch agencies in court, at every level of litigation.  Agencies typically cannot represent themselves in court cases.
• Inspectors general have formal access to information within their agencies and have a dual reporting authority to both their agency heads and to Congress.
• The U.S. Postal Service must submit rate increases to the Postal Rate Commission for approval.

The authors say these types of design choices centralize control, where “the agent brings expertise, and the principal brings control.” They say “For democratic legitimacy, the mechanism of conflict resolution provides clear accountability.”

Conflict in Advisory and Monitoring Relationships. In this set of relationships, one agency has decisional authority but another can exert some degree of influence on a decision. This can create internal checks and balances by generating information to be shared. For examples:

• The National Environmental Policy Act, where “agencies must obtain comments from any other agency with relevant environmental expertise or authority for certain projects.” This built-in conflict came to public attention in 2016 when the Environmental Protection Agency disputed the Federal Energy Regulatory Commission’s environmental impact statement in the approval of a pipeline project, and resulted in an intervention by the White House Council on Environmental Quality.
• The Office of Advocacy in the Small Business Administration often raises regulatory cost-benefit issues of other agencies’ proposed regulations that may affect small businesses to OMB’s Office of Information and Regulatory Affairs.
• The Advisory Council on Historic Preservation also has statutory authority to advise, where “agencies must give it an opportunity to comment on specific projects.”

Farber and O’Connell say such relationships work to prevent groupthink and may render administrative decisions more legitimate in the eyes of stakeholders.

Conflicts Between Agencies with Equal Power. In this set of relationships, two agencies have equivalent and overlapping authority, much like the example cited earlier with OSHA, FDA, and the hapless chocolate manufacturers. This results in some degree of competition between agencies. This approach can result in redundancies but has both benefits and costs. For example:

• A classic case is the overlapping jurisdictions in food safety responsibilities between the Food and Drug Administration and the Department of Agriculture.
• Another is the tension between the FDA and the Drug Enforcement Agency over drug approvals.
• The overlap between independent agencies is even more challenging, such as the regulation of futures contracts between the Security and Exchange Commission and the Commodity Futures Trading Commission.

The authors say that Congress is aware of these conflicting jurisdictions and that “this design often results in less accountability but more participation.” They also note that “symmetrical conflict within an agency allows each party to always retain some authority,” and that as a result, interest groups have more difficulty in “capturing” a particular policy area.

Mechanisms for Resolving Conflicts

The authors believe that there are “potentially healthy benefits of conflict among administrative actors.” But there also needs to be mechanisms for resolving conflicts. They describe four such mechanisms:

• Negotiation and Mediation. This approach is largely agency-driven, not top-down. It’s typically done via the development of memoranda of understanding. For example, the FDA has over 100 MOUs, largely with other federal agencies, regarding information sharing and the division of regulatory or enforcement responsibilities. The Interior Department uses alternative dispute resolution, where a third-party facilitator is brought in to broker interagency agreements on contentious issues such as the interactions between pesticide registration and the Endangered Species Act.

• Adjudication of Administrative Disputes. This approach is largely top-down, often driven by the White House or Justice’s Office of Legal Counsel. These include issues such as drug control, AIDS, climate change . . . and regulations affecting chocolate manufacturers.

• Voting and Consensus. This approach is typically imposed by Congress. For example, a two-thirds majority of the members of the Financial Stability Oversight Council have the authority to veto regulations proposed by the Consumer Financial Protection Bureau. And in the case of financial credit swaps, there is a statutory requirement that the SEC, CFTC and bank regulators coordinate, using consensus-driven joint rulemaking procedures.

• Court Cases. In rare cases, the courts wind up serving as the primary dispute-resolution mechanism. Usually, the Justice Department will resolve a legal conflict, but some independent regulatory commissions have been given statutory authority to control their litigation in lower courts, such as the Federal Communications Commission and the Nuclear Regulatory Commission.

So, as OMB and Congress assess the President’s recently-released reform and reorganization proposals, it may well be worth pondering whether some degree of duplication, overlap, and conflict—with appropriate resolution mechanisms—is actually worth keeping in the design of selected agencies and programs. Such conflict can induce a form of institutional checks and balances—a feature, not a bug.