Criminal Subsidiaries

Each year, Department of Justice (DOJ) components resolve a handful of corporate criminal cases with subsidiary-only conviction (SOC) settlements. In SOC settlements, a subsidiary pleads guilty to offenses that its parent or siblings share liability for. From 2013 to 2022, SOC settlements occurred in at least 3.3% of all federal corporate criminal resolutions, including 5.6% of cases in which prosecutors sought an entity conviction. For parent companies and their other subsidiaries, isolating conviction to one entity protects the rest of the corporate group from criminal collateral consequences. This result can be referred to as criminal entity partitioning, a subset of the entity partitioning that serves as a core function of organizational law. For prosecutors, SOC settlements allow greater flexibility in balancing between the need to avoid social cost that could result from fully prosecuting a firm’s culpable constituents (e.g., the need to avoid the “corporate death penalty”) with the need to deter and punish corporate crime, prevent recidivism, and achieve other criminal-legal ends. In other words, SOC settlements allow prosecutors to obtain entity convictions when appropriate, while avoiding the regulatory and other collateral consequences associated with parent-level convictions.

In my forthcoming article, Criminal Subsidiaries, I introduce a theory of criminal entity partitioning in which subsidiarization allows the worst consequences of criminal enforcement to be quarantined within a firm. This criminal partitioning differs, however, from the better-known concept of asset partitioning, in which a firm ex ante allocates private assets and liabilities between its subsidiaries. That is because within a corporate group, entity borders tend to be more porous to criminal than to private liability. One implication of this theory is that subsidiarization lacks ex ante reliability for partitioning in the face of criminal enforcement, whereas it is reliable ex ante for isolating private liabilities. Instead, effective criminal entity partitioning must occur ex post via settlement between parent and prosecutor.

Beyond theorizing SOC settlements, my article offers a preliminary empirical investigation of their use. Using the Lexis news database, I identify over 200 SOC settlements dating back to 1974. Given the limitations of the search strategy (not all news publications are in the Lexis database, and not all corporate criminal resolutions are reported on by the press), this newspaper-based study likely undercounts SOC settlements. But it does permit tentative exploration of three key questions.

First, how common are SOC settlements? The answer is that they are not ubiquitous. But given that there have been at least 48 such federal settlements between 2013 and 2022, they are also not rare. More, their collective total fines ($12.3 billion, excluding non-fine financial penalties and amounts paid under parallel resolutions with other agencies) are noteworthy standing alone.

Second, how do SOC settlements compare to federal corporate criminal resolutions generally, including those in which prosecutors seek an entity conviction or agree to non- or deferred prosecution? To answer this question, I excerpted all cases from the Corporate Prosecution Registry—a database of corporate criminal prosecutions built and maintained by Brandon Garrett and Jon Ashley—with a “date” field between 2013 and 2022, inclusive. Using fine amounts as a proxy for the seriousness of underlying misconduct, SOC settlements are notably bigger—sometimes by orders of magnitude—than other corporate criminal resolutions.

                                * Brandon Garrett & Jon Ashley, Corporate Prosecution Registry

 A few points can be drawn from these descriptive data. First, SOC settlements represent a small, but meaningful, share of federal corporate criminal resolutions (3.3%) and a larger portion of resolutions in cases that prosecutors considered to be conviction appropriate (5.4%). Beyond collectively representing $12.3 billion in fines over ten years, SOC settlements’ mean fines were an order of magnitude greater than mean fines for all cases and conviction-appropriate cases and five times the mean fines for N/DPA cases. One note of caution about the second point is that DOJ practice appears to be inconsistent in how fines are allocated in SOC settlements between parents receiving N/DPAs and subsidiaries being sentenced following guilty pleas. The median fines are less susceptible to inconsistencies in allocating fines between parents and subsidiaries, however, and there again the SOC median ($93 million) is orders of magnitude greater than the other median fines.

Third, what kinds of offenses lead to SOC settlements and what kinds of companies enter them? I categorized the forty-eight 2013-2022 SOC settlements according to the primary offense to which a subsidiary pleaded guilty. It comes as no surprise that FCPA violations—which are apt to feature misconduct by one or more overseas subsidiaries—represent the largest share of SOC settlements, a third of cases in the dataset.

To illustrate what kinds of defendants enter SOC settlements, I categorized the dataset according to the parent’s industry. Note that financial services and healthcare—two industries that are particularly susceptible to the collateral consequences of parent-level conviction—accounted for 52.1% of all settlements in the dataset. This point suggests that participants in corporate criminal cases recognize SOC settlement as an option for balancing public enforcement interests with the avoidance of social cost.

Although SOC settlements could be criticized as allowing parents to avoid full accountability, I conclude that the practice enables valuable enforcement calibration between accountability for corporate wrongdoing and the avoidance of social cost. That is, limiting conviction to subsidiaries can reduce the salience of social cost in prosecutorial decision-making. More, this practice supplements, but does not replace, other enforcement practices embedded in typical corporate criminal settlements. And importantly, it is already (an albeit small) part of federal corporate enforcement practice. In light of these benefits and precedents, I argue for expanded use of SOC settlements. To reap the full benefits of this expansion, however, the DOJ and other agencies should implement guidelines that promote consistency, transparency, and efficacy around SOC settlements. In the article, I propose and justify four such guidelines: (1) presumption in favor of conviction when multiple entities within a corporate group share liability; (2) conviction only of subsidiaries that are economically meaningful to the parent; (3) conviction only of wholly owned subsidiaries; and (4) public disclosure regarding the reasons for and considerations embedded in SOC settlements.