Regulatory Diffusion.

• Author: Nou, Jennifer

Table of Contents Introduction I. Diffusion Trends A. Measuring Diffusion 1. Dataset 2. Identifying text reuse B. Analysis 1. Scope 2. Leaders and followers II. Explaining Diffusion A. Why Diffusion B. Channels of Diffusion III. Implications A. Evaluating Diffusion B. Interpreting Similar Regulations Conclusion Appendix A: Detailed Description of Similarity Measure Appendix B: Detailed Description of Weighted Leadership Scores Appendix C: Additional Tables and Figures Introduction

Regulatory texts, like other forms of law, diffuse. They migrate across administrative agencies in substantially the same form. Agencies, in other words, borrow rules from one another. The broader phenomenon of legal diffusion is usually studied as a comparative, international matter. (1) Laws are frequently "transplant[ed]" to foreign arenas. (2) Developing countries, this literature suggests, borrow laws from more developed countries to signal legitimacy, garner prestige, or preserve resources. (3) Sometimes regulations are mimicked to promote harmonization between member states, such as within the European Union. (4) Other times, regulatory standards in one country are adopted abroad because they are perceived as more normatively desirable. (5)

Regulatory diffusion, it turns out, also occurs closer to home. American administrative agencies borrow the texts of rules from one another. Such behavior is not merely a cost-saving resort to templates, which are commonly used when agencies draft responses to executive orders. (6) Rather, agencies issuing separate regulations reuse substantive and procedural regulatory provisions. One provision is promulgated by an agency, only to be later adopted in substantially the same form by another. (7) The same provision can be mimicked over decades by dozens of disparate agencies. (8)

Sometimes this practice is required by Congress or the President. (9) Many times, however, it is not: Agencies, in their discretion, choose to import regulatory language from another agency's rulemaking. Consider some examples:

* In 1980, the Department of Education (ED) promulgated Title IX regulations regarding nondiscrimination on the basis of sex. (10) Two decades later, twenty-one agencies decided to substantially copy most of ED's final regulations on Title IX. (11) In their "common rule," they cited several reasons for doing so, including "the history of public participation in the development and congressional approval of ED's regulations"; ED's "leadership role" in regulatory enforcement; the public's preexisting familiarity with the regulations; and an interest in maintaining regulatory consistency. (12)

* In 2001, President George W. Bush issued Executive Order 13,199 calling for greater inclusion of faith-based organizations in federal grants. (13) The challenge was how to do so in light of competing Establishment Clause concerns. Over the following years, a number of agencies promulgated rules allowing federal grant funding, but not for "[inherently religious activities such as, religious worship, instruction, or proselytization." (14)

* In 2010, however, President Barack Obama's Advisory Council on FaithBased and Neighborhood Partnerships recommended that agencies replace the term "inherently religious" because the term was confusing and did not sufficiently indicate the boundaries of government subsidies. (15) That same year, President Obama issued Executive Order 13,559, which used the phrase "explicitly religious activities" though it did not specifically require any changes to agency rules. (16) Six years later, nine large federal agencies jointly changed their funding restrictions to cover "explicitly religious activities," which included "proselytization" and "religious instruction." (17) Other agencies followed the next year, even after President Obama left office. (18)

* In 2015, the Department of Commerce proposed revisions to its Export Administration Regulations (EAR) to "enhance clarity and consistency" with the terms of the International Traffic in Arms Regulation (ITAR), which is administered by the Department of State. (19) The proposed rule imported the definitions of a number of terms from ITAR. (20) Because the two sets of regulations were issued pursuant to different statutes and administered by different agencies, the proposed rule noted that "each set of regulations" had "evolved separately over decades without much coordination between the two agencies regarding their structure and content." (21) As a result, they often used "different words, or the same words differently, to accomplish similar regulatory objectives." (22) So to "facilitate enhanced compliance while reducing unnecessary regulatory burdens," the proposed EAR copied the ITAR definitions rather than rework definitions in concert with the other agency. (23)

In this manner, regulations diffuse across different agencies, at different times, and through different channels. While one might expect this dynamic to occur only between agencies sharing statutory authorities, it occurs between agencies authorized under different statutes as well.

This phenomenon raises a host of questions, both descriptive and normative: What motivates the borrowing of regulatory texts? Do drafters simply seek to save time and resources? Or are these efforts to standardize regulatory language and reduce compliance costs? Through what mechanisms does diffusion occur--interagency working groups, staff migration, centralized coordination? Are some agencies more influential leaders in drafting regulatory texts than others? More broadly, what networks exist across agencies when writing rules? And what are the implications, if any, for regulatory process and interpretation?

These inquiries lie at the intersection of a number of rich literatures. Comparative scholars, as mentioned, have long examined how and why laws are "transplant[ed]" across different jurisdictions. (24) Some of this work studies the migration of legal rules without regard to their form, whether case law, statutes, or otherwise. (25) Social scientists have also long studied the broader notion of "policy diffusion," the "spread of [policy] innovations from one government to another." (26) Work in this area has mainly focused on policy choices adopted by legislative bodies and how they spread, "while ignoring the equally important decisions made by executive agencies." (27) Relatedly, some analyses have focused more specifically on the extent to which state statutory texts are "reuse[d]" or copied verbatim by other state legislatures. (28) Contract law scholars, for their part, have long studied the use of boilerplate and drafting templates, asking how copied contractual terms proliferate and through what channels. (29)

These themes have yet to be explored in the context of U.S. administrative agencies. One reason may be the previous regime of strong judicial deference to agency regulatory interpretation. (30) As a result, there was little need to consider how and why agencies draft regulations. But that is no longer the case. (31) We suspect that another reason is the until-recent lack of access to machine readable versions of the Federal Register, (32) Earlier work on agency behavior used data drawn almost exclusively from the Unified Agenda (33)--a semiannual publication of agencies' planned activities. (34) The Unified Agenda essentially consists of self-reported data on planned regulatory activities, which some research now suggests is incomplete. (35) Agencies, however, are legally required to publish their final rules in the Federal Register. (36) Data drawn from the Federal Register thus provides the most comprehensive portrait of agency rulemaking available.

At the same time, administrative law scholars have increasingly studied the ways in which agencies relate to each other. Their work has documented that these interactions can be demanded and designed by the President or Congress. (37) Other times, these interactions are "interagency," (38) that is, decisions by agencies themselves to coordinate or clash with other agencies. (39) Much of this nuanced scholarship explores case studies and examples. In addition, it often focuses on formal legal mechanisms, such as statutes or executive orders, interagency memoranda, and jointly issued policy statements and guidelines. (40)

While these approaches have yielded rich insights, they have granted less attention to patterns of agency interaction, informal relationships, and the ways in which interagency connections are structured. They underappreciate, in other words, the importance of agency networks. (41) Without an understanding of these connections, it may be more difficult to see how agencies exercise power by virtue of their positions within networks and how cohesive those networks are. For example, some agencies are highly influential rule drafters: When they write a regulation, other agencies follow. (42) Similarly, some agencies are central network players and interact with a great number of other agencies. (43) Our analysis begins to shed light on some of these associations and related dynamics.

More broadly, this Article considers the phenomenon of regulatory diffusion as both a positive and normative matter. One of its primary aims is to help establish a research agenda on regulatory diffusion informed by our descriptive work. To that end, Part I uses text and network analysis to evaluate twenty years of data drawn from the Federal Register. These tools have long been staples of social-science and public-administration research, (44) but they have yet to be deployed robustly in the administrative law literature. (45) Our findings suggest that the share of reused texts has increased over time, from less than 3% in 2000 to more than 10% by 2020--more than a threefold increase. Strikingly, as of 2020, one out of every ten new regulatory paragraphs was borrowed from a previously published regulation.

In...