Regulatory System Design: Some Solutions to the Delegation Problem

Abstract: We present solutions to each of the major delegation problems that arise when elected officials delegate rulemaking authority to government agencies. These problems include principal-agent issues, monopoly provision, information asymmetry, and tragedy of the commons. Rather than presenting our solutions to these problems as incremental changes to an existing system, we discuss how a regulatory system built from scratch might avoid these problems. Following our problem-specific solutions, we present a detailed structure of a regulatory process that alleviates many delegation problems simultaneously. This structure better aligns the incentives of regulators with those of legislators and with the well-being of the public. We intend the solutions and process structure presented here not to serve as a collection of proposed changes but as guideposts for those hoping to make any part of the regulatory system better attuned to the needs of the populace.

JEL codes: K23, L51, Q58

Keywords: benefit-cost analysis, principal-agent, RegData, regulation, regulatory accumulation, regulatory budget, regulatory impact analysis, regulatory reform

Jerry Ellig (1962-2021) was a dedicated and accomplished scholar dedicated to making a difference in people’s lives. Jerry believed in the need for a “bridge” connecting the academy and the policy world and he traversed the bridge many times – from university to government and back again throughout his career. A mentor to many, he is remembered as “relentlessly cheerful” and “generous in advice and time.” 

This special study is a summary of many of Jerry’s works and proposed solutions and is meant to carry on his legacy as a great bridge builder. 

If you would like to learn more about Jerry Ellig, you can read his full bio here.

Mercatus AI Assistant
Ask questions about this research.
GPT Logo
Mercatus AI Research Assistant
Ask questions about this research. Mercatus Chatbot AI More Details
Suggested Prompts:
Ask us anything. We use OpenAI's ChatGPT 4o base model to answer any question about Mercatus research.