Notice & Comment

Reasoning for Votes, by Roger G. Noll

*This is the eighth post in a symposium on Jed Stiglitz’s “The Reasoning State.” For other posts in the series, click here.

The Reasoning State is compelling because it manages to be both informative and enjoyable to read. The book explores why legislators delegate policy-making authority to agencies, subject to the requirement that agency officials adhere to procedural rules that require the use of “credible reasoning” to reach their decisions. In addressing this question, The Reasoning State adopts and extends the positive political theory (PPT) of policy formation, which begins with the assumption that elected officials enact statutes to serve their re-election objectives and, as a result, have policy preferences that are broadly in accord with those of their support constituencies. The book extends this theory by arguing that a commitment to a policy-making process that employs credible reasoning, combined with judicial review of whether this commitment was followed, serves the re-election interests of legislators because voters prefer such a process to a less structured process, such as legislative deliberation. 

A source of tension between PPT and traditional scholarship in administrative law is the latter’s focus on normative rationales for procedural requirements.[1] Normative theories of administrative law do not offer much of an explanation for why legislators enact statutes and courts adopt common law standards that establish normatively compelling procedures for policy decisions by agencies. To the extent that traditional normative theory seeks to explain the content of administrative law, its focus is on actions by judges. The normative theory of public law assumes that courts prefer administrative processes that exhibit a minimal degree of fairness and will punish agencies that do not comply with this preference by refusing to enforce their decisions. By behaving in this way, courts incentivize agencies to accept the court’s procedural norms and legislators to include normatively appealing procedures in statutes that give agencies discretion in implementing policy. Why courts proactively pursue procedural fairness as an end instead of a means, why legislators allow courts to impose procedures that legislators would not adopt on their own or sometimes enact statutes that mandate more demanding procedures than courts require, and why elected officials appoint judges who doggedly pursue procedural fairness rather than the policy preferences of those who appoint them, remain unexplained. 

As The Reasoning State explains in detail, PPT answers these questions by providing a positive theory of the content of administrative law. PPT regards administrative procedures as designed to cope with a principal-agent problem that arises when elected officials delegate policy-making authority to bureaucrats (civil servants and/or political appointees) who may pursue policy goals that differ from the goals of the legislative coalition that enacted the statute containing the delegation.[2]

PPT analyzes this principal-agent problem in a policy-making process that is a four-stage repeated sequential game.[3] In Stage 1, citizens vote based on their evaluation of candidates’ current and proposed policies, and in so doing establish an “electoral connection” between the policy preferences of citizens and elected officials. In Stage 2, a majority coalition of successful candidates enact statutes that do three things: (1) establish policy goals that often are broad and vague; (2) give discretion to agencies to fill in the details of policy; and (3) impose procedures that agencies must follow when exercising this discretion. In Stage 3, agencies develop and implement detailed policies. In Stage 4, courts review agency decisions for compliance with statutory goals and procedural norms (from both statutory and common law). Each stage is then repeated, giving voters, legislators, agencies, and courts opportunities to respond to previous decisions by others. 

PPT assumes that the relevant actors in each stage pursue their preferred policy outcomes, but that in so doing they take into account the possibility that other actors will amend or even overturn their decisions in subsequent stages. Because ex post policy correction is costly and, in any case, can be overturned, legislators (or courts) impose procedural requirements to reduce the extent to which policy decisions by an agency deviate from their preferred outcome. 

The Reasoning State extends PPT by postulating that voters evaluate candidates in part on the basis of whether policy decisions by the agencies that they create and oversee are derived from credible reasoning. That is, voters are assumed to be willing to accept a reduction in the extent to which policy outcomes are in accord with their preferences in return for an increase in the extent to which policy emanate from a transparent, accessible process in which decisions are based on evidence and are subject to judicial review for compliance with norms of procedural fairness. Moreover, voters regard a commitment to reasoned decision making as more credible if it is hard-wired by statute into agency procedures and judicial review rather than if the same or similar process for gathering information is used by the legislature to make policy itself. 

Professor Stiglitz’s explanation of the electoral connection to administrative law suggests that the reasoning state exists to solve the principal-agent problem between voters and legislators. That is, voters prefer delegation of policy making to agencies in the presence of procedural rules and judicial review because it protects them against legislators who, on their own, would adopt policies that deviate more from the preferences of voters than the policies that are adopted by the reasoning state. Thus, unlike traditional normative theory, the theory of administrative law advanced in The Reasoning State is consequentialist: voters prefer the administrative state to policy making by legislatures because it produces policy outcomes that are more in line with their preferences, and responding positively to these preferences is attractive to legislators because it advances their electoral prospects. 

The theory of administrative law set forth in The Reasoning State enriches understanding of the passage of the flagship statute of administrative law, the Administrative Procedure Act (APA) of 1946. The Reasoning State (pp. 78-86) argues that the passage of the APA cannot be explained solely by the enacting coalition (in this case, essentially a unanimous Congress) fine-tuning political control of agencies to which policy making power has been delegated. While this conclusion is valid, it does not establish the next step, which is that voters favored its passage because they perceived the administrative state to be fair and impartial as ends in themselves, as opposed to perceiving that procedurally imposed fairness and impartiality would lead to policies that are closer to their most preferred outcomes. 

Passage of the APA affected policy outcomes from New Deal agencies in several ways.[4] First, the APA allowed a broader array of organized interests to participate in the policy-making process of New Deal agencies. Second, the APA gave opponents of agency decisions grounds for judicial appeal if the agency failed to respond to the evidence and arguments submitted by these interests. Third, by imposing on agencies greater procedural complexity, new information processing requirements, and more demanding standards for justifying policy decisions, the APA slowed the pace and reduced the variance of future changes in policy, thereby making future policy less uncertain in the face of political uncertainty after the death of FDR. Fourth, the APA increased the chance that legislators would receive advance warning of changes in agency policies before they were decided, thereby giving a Congressional majority coalition more opportunity to stop an agency from deviating too far from its preferred outcome. 

Each of these features of the APA are consistent with the norms of fairness, transparency, and impartiality, but these normative features also were attractive to voters and their elected representatives because of their policy consequences.[5] Thus, in a deep sense The Reasoning State can be interpreted as bridging the gap between normative and positive theories of administrative law. To a die-hard consequentialist, the normatively attractive procedural requirements of the administrative state plausibly are politically compelling to voters, politicians, bureaucrats, and judges, directly or indirectly, because of their positive consequences. 

Roger G. Noll is Professor of Economics Emeritus at Stanford University. 

[1] See, for example, Cass R. Sunstein and Adrian Vermeule, “The Morality of Administrative Law,” Harvard Law Review 131,7 (2018): 1924-1978. 

[2] The severity of the principal-agent problem depends on the technical complexity of implementing a policy. For an excellent analysis of how greater technical complexity leads to less effective political control of policy outcomes, see Kathleen Bawn, “Political Control Versus Expertise: Congressional Choices about Administrative Procedures,” American Political Science Review 89.1 (1995): 62-73.

[3] For a more comprehensive statement of this theory, see McNollgast, “The Political Economy of Law,” in A. Mitchell Polinsky and Steven Shavell (eds.), Handbook of Law and Economics, Vol. 2 (2007), North-Holland Publishers: New York, 1651-1738.

[4] For more details about the policy effects and political rational of the APA after a decade of failure by the Conservative Coalition (Republicans plus Southern Democrats) in Congress to impose tighter administrative procedures on New Deal agencies, see McNollgast, “The Political Origins of the Administrative Procedure Act,” Journal of Law, Economics, and Organization 15.1 (1999): 180-217. 

[5] The appeal of the APA to Congress is examined in George B. Shepherd, “Fierce Compromise: The Administrative Procedure Act Emerges from New Deal Politics,” Northwestern University Law Review 90.4 (1996): 1557-1683.

Print Friendly, PDF & Email