Mapping Enforcement Networks

Many modern problems—money laundering, investor fraud, cryptocurrency abuses—cannot be addressed effectively by solitary agencies. Instead, cross-border and borderless problems require a coordinated response from multiple entities in separate locations with different expertise and roles. They require, in other words, a network. But measuring disparate and dispersed coordination is difficult. This difficulty is particularly acute in the context of investigation and enforcement, which often take place behind closed doors. My forthcoming article, ‘Enforcement Networks’, develops a novel approach to quantifying how networks of domestic and international agencies coordinate in civil (non-criminal) enforcement. 

What do the Hampshire Constabulary, Texas Railroad Commission, City of Chicago, and the FBI have in common? One answer is that the US Securities and Exchange Commission (SEC) thanked each of them for assisting the agency in a securities enforcement action. The SEC announces enforcement activity publicly through its litigation releases and press releases. These often close with an acknowledgment of coordination: the SEC ‘appreciates the assistance of…’ or ‘thanks the following agencies for their cooperation and assistance…’ 

Taking advantage of the SEC’s routine practice, the study uses tools from computer and information science to construct a network from acknowledgements of assistance in the SEC’s public releases. The article reports the results for more than twenty years (1998-2018), mapping the SEC’s acknowledgements as a proxy for the underlying interactions: the enforcement network. 

The study allows definitive answers about the acknowledgments: whom the SEC thanked and when. And the information gleaned from the acknowledgments is new. Patterns of acknowledgment are not easily explained by particular categories of cases (eg, FCPA), by membership in coordinating organizations (eg, IOSCO), or by entry into more formal agreements (eg, MOUs). 

The information has some inherent limits, of course. By definition, for instance, it does not include unacknowledged coordination or enforcement activity that does not result in a public release. The study may pick up changes in the SEC’s acknowledgment practices as well as changes in the underlying interagency interactions. Nonetheless, such a study is a particularly useful indirect approach to an area where direct information is difficult to obtain.

What does the SEC’s network look like? 

• Large and diverse number of entities. The multi-decade study of acknowledgments reveals a network of almost 500 unique entities. They are diverse. The SEC thanks US state, US federal, and non-US entities; big city US Attorney’s General Offices; small town police departments; the FBI; self-regulatory organizations; and others. 
• Repeat players and one-shotters. The network is characterized by a mix of repeat players and one-time coordinators. US Attorney’s Offices, the FBI, and FINRA (and its predecessors) account for approximately half of the acknowledgments. However, concentrated interaction with the top repeat players is only one aspect of the network. Overall, single interactions with the SEC prevailed. Almost half of all entities were acknowledged only once, and seventy percent were acknowledged three or fewer times. 
• Domestic and international entities. The results depict an intertwined domestic and international network. Global coordination is part of the story. The SEC acknowledged entities from seventy-seven different international jurisdictions, with Canada, the United Kingdom, and Switzerland the most frequently acknowledged jurisdictions. But coordination between federal and state authorities within the US is also significant. The SEC acknowledged entities from forty-nine US states, as well as Puerto Rico, the District of Columbia, NASAA (the member organization for state securities regulators), and local police departments or other local authorities. 
• Increasing coordination. The results are also consistent with an increasingly interwoven system. Over time, for instance, the SEC has thanked an increasing number of entities per release. 

The enforcement networks study provides missing empirical underpinnings for theorizing agency coordination. It provides new information and a tool for quantifying interactions that cannot be detected through a study of formal structures and agreements alone. Although the approach has limitations, tracking the SEC’s acknowledgments identifies agency enforcement interactions that would otherwise be invisible to the public eye. It also models a tool for understanding agency networks more generally.

A version of this post also appears on the NYU Compliance and Enforcement Blog.

Verity Winship is Professor of Law at the University of Illinois College of Law.