SEC Cases and Actions: A Different Way to Measure Securities Enforcement

Figure 1 illustrates that in most of its cases, the SEC charged one defendant, and that most cases involving multiple defendants were prosecuted in more than one action. As illustrated in Figure 2, Multi-Action Cases account for nearly half of the SEC’s enforcement actions in our sample.

Finally, as the defendant-level data in Figure 3 illustrates, nearly half of all defendants in our sample were charged in a Multi-Defendant Case.

Figures 1, 2, and 3 illustrate how sorting data at different levels impacts overall counting. In Part III we apply these concepts to the SEC’s data and consider how cases versus actions impact the enforcement counts that the SEC reported.

C. Multi-Action Cases

At this point, one might ask what makes Multi-Action Cases distinct or otherwise noteworthy, so as to merit categorization, beyond the fact that they have received scant attention in the literature. Put differently, why does the SEC prosecute certain cases across multiple enforcement actions? And does it matter that the SEC chooses to do so in certain cases?

In cases implicating multiple defendants, the SEC has a choice: consolidate all defendants into a single enforcement action (a Multi-Defendant, Single Action Case) or charge defendants separately (a Multi-Action Case). It is theoretically more efficient, as a matter of litigation resource management, to consolidate charges against multiple defendants into a single legal proceeding. And yet, as the prior section shows, Multi-Action Cases constitute a sizeable portion of the SEC’s caseload.

The choice to charge defendants separately necessarily impacts the total number of enforcement actions the SEC reported. Consider, for example, the SEC’s Multi-Action Case involving Manitex International, Inc., a manufacturer of forklifts and other heavy machinery. In September 2020, the SEC alleged that “certain of Manitex’s officers and employees engaged in two distinct, fraudulent schemes involving the use of related party entities to engage in fraudulent accounting practices.” Instead of one consolidated action against all defendants, the SEC filed four settled enforcement actions, all on the same day, against, Manitex, its CEO, its CFO, and a senior manager.

Professor Velikonja raised, without testing, two possible explanations for Multi-Action Cases like the Manitex example. The first rests on the fact that the SEC is incentivized to split up cases into multiple actions to increase the overall number of enforcement actions it brings. That is, the SEC might opportunistically slice and dice cases involving multiple defendants into multiple actions. The second alternative explanation is that Multi-Action Cases simply arise in situations where some defendants settle their charges but others do not.

SEED data suggests that the second explanation is more likely, though our descriptive statistics point to a broader explanation, one not limited to settlement. Namely, Multi-Action Cases may facilitate the resolution of cases where parties have divergent priorities. These divergent priorities are shown not only by defendant settlement decisions, but additional factors, including: (1) the nature of the defendant, (2) the defendant’s decision to cooperate (or not cooperate) with the SEC, and (3) the SEC’s (and maybe the defendant’s) preferred enforcement venue.

1. Multi-Action Cases: Slicing and Dicing

While some Multi-Action Cases do appear to be the kind of opportunistic slicing and dicing of cases into multiple actions that Velikonja identified, these cases represent only a small subset of Multi-Action Cases in our sample. In part, this is because not all Multi-Action Cases were filed contemporaneously; nearly half of the Multi-Action Cases on SEED were prosecuted successively over time. Consider, for example, the case of Hertz, in which the SEC charged the car rental company and its executives for accounting issues that overstated the company’s income in public filings. On December 31, 2018, the SEC initiated administrative proceedings against Hertz and announced that its investigation was “continuing.” Nearly one year later, in December 2019, the SEC initiated administrative proceedings against Hertz’s comptroller and Senior VP, and in its press release, the SEC also noted that its investigation was ongoing. Finally, in August 2020, the SEC filed charges in the District of New Jersey against Hertz’s CEO and former Chairman, concluding its investigation into the matter. It is common for the SEC to announce charges in a press release and note, as it did in the Hertz case, that the investigation is ongoing. The SEC’s Enforcement Manual, the public document that describes the SEC’s internal enforcement procedures, describes this type of staged enforcement as follows:

The Division may continue to investigate and issue investigative subpoenas pursuant to a formal order of investigation while simultaneously litigating a related civil action if there is an independent, good-faith basis for the continued investigation. An independent, good-faith basis may include the possible involvement of additional persons or entities in the violations alleged in the complaint, or additional potential violations by one or more of the defendants in the litigation.

On SEED, during the Relevant Time Frame, 100 of the 189 Multi-Action Cases were comprised of actions all filed within a one-month period. Eighty-nine of the 189 Multi-Action Cases were comprised of enforcement actions filed successively over time, in some cases spanning months or years between the first and last action filed in the case. The subset of Multi-Action Cases in which actions were filed contemporaneously (or near-contemporaneously) do not suggest a systematic practice of opportunistic slicing and dicing. The SEC consolidated defendants in far more of its cases where all defendants settled concurrently than it broke them up.

Based on those Multi-Action Cases filed contemporaneously—we choose an arbitrary cut off date of one month—we find that the SEC “netted” 63 total actions from these cases. In other words, if the SEC had consolidated all Multi-Action Cases filed contemporaneously into a single action (that is, not slice and dice them), the SEC would have filed 63 fewer actions. To arrive at the net number of 63, we focused on the 17 percent of SEED cases (126/729) in which the SEC charged multiple defendants and all defendants in the case settled charges concurrently with filing. In 43 of these 126 cases (34 percent), the SEC filed multiple actions even though all defendants settled concurrently with filing. These 43 cases generated a total of 106 discrete actions which, in theory, could have been consolidated into 43 Multi-Defendants, Single Action Cases, and, consequently, the SEC would have filed 63 fewer actions. Conversely, in 83 of the total 126 cases (66 percent), the SEC consolidated all charges into a Multi-Defendant, Single Action Case. In these 83 actions, the SEC named a total of 215 defendants. In theory, the SEC could have filed 215 discrete settled actions, but instead elected to consolidate the charges into 83 Multi-Defendant, Single Action Cases.

Are sixty-three additional actions, which appear to be the product of filing decisions and not any other reason, a lot or a little? This is a subjective determination; however, our data does not suggest any evidence of overt opportunism, at least in terms of when the SEC chose to consolidate Multi-Action Cases. If the SEC were slicing and dicing cases to boost its enforcement numbers, enforcement staff might be more inclined to do so near the end of the SEC’s fiscal year. While SEED data shows a September Swell across all case types, and thus more apparently sliced and diced cases in September, it does not show a greater proportion of Multi-Action Cases deconsolidated into multiple enforcement actions in September compared to other months. Instead, SEED data suggests that Velikonja’s second hypothesis concerning defendant settlement patterns better explains Multi-Action Cases. Velikonja observed anecdotally that the SEC often splits up cases into separately filed enforcement actions, which might facilitate the resolution of cases where some defendants settle, and others do not.

SEED data provides support for this hypothesis. In cases involving multiple defendants, the SEC overwhelmingly charged defendants in separate actions when some defendants settled and others did not. Of the 134 Multi-Defendant, Single Action Cases on SEED (in which a total of 409 defendants were named) only two cases featured at least one defendant that settled concurrently and one that did not. Defendants in these cases nearly always either settled the same day charges were filed (83/134), or did not settle concurrently with filing (49/134). Conversely, in 56 percent of Multi-Action Cases on SEED (105/189), at least one defendant settled the same day charges were filed, and one did not. In fifty-eight Multi-Action Cases, all defendants settled concurrently, and in twenty-six, no defendants settled concurrently.

The difference in settlement patterns between Multi-Action Cases and Multi-Defendant, Single Action Cases is striking and thus worth illustrating. Figures 4 and 5 suggest that when all defendants in a case acted in concert—at least regarding whether to fight or settle charges—the SEC tended to consolidate them in a single action, generating a Multi-Defendant, Single Action Case. Conversely, when defendants were not aligned in their decision to fight or settle charges, the SEC appeared to prefer prosecuting the defendants in separate actions, generating Multi-Action Cases.

Some of this may be exclusively driven by timing. That is, Multi-Action Cases may simply reflect the staged and dynamic nature of the SEC’s investigations. As noted earlier, about half of the SEC’s Multi-Action Case prosecutions unfolded over time, often spanning months or even years. On average, a Multi-Action Case lasted one year from the first action filing to the last; some in our dataset extended up to five years from the first action.

Actions against auditors in our dataset, which comprise a small portion of Multi-Action Cases, illustrate, to some degree, the effect of timing. For example, in 2013, the SEC charged a China-based distributor of wholesale chickens and its CEO with making false statements between 2009 and 2011; in 2014, a year later, the SEC filed charges against an accounting firm for failing to properly audit the poultry-supplier. Here, it is plausible that the SEC’s investigation of the company, prompted by suspicious public filings, led the agency’s investigators to examine the auditors who reviewed the company financials.

However, the SEC has discretion over when it decides to bring the initial charges. It can, and in fact routinely does, wait until it has all the evidence against all parties to bring charges. Consider the SEC’s case against Satyam Computer Services, a different Multi-Action Case involving auditors. In 2009, Satyam’s chairman published a letter confessing widespread fraud and misstatements of the company’s financial performance, yet the SEC waited until 2011 to file charges, charging the company, its executives, and its auditors on the same day. Thus, while timing may certainly play a role in when the SEC’s investigators identify evidence of wrongdoing, the SEC can be strategic in when it files charges.

Moreover, the overwhelming difference in settlement patterns between Multi-Action Cases and Multi-Defendant, Single Action Cases suggests that charging defendants in separate enforcement proceedings somehow facilitated the enforcement of multiple defendants with different priorities or incentives, at least as evidenced by their decisions to settle or fight charges. However, our case-level data suggests that case-filing decisions may not be exclusively driven by settlement, and that additional, related factors reflect different priorities of multiple defendants, evidenced in Multi-Action Case data. The first factor concerns the nature of the defendant, namely whether the defendant is an individual or a firm, which is intimately connected with settlement patterns. Most individuals charged by the SEC were prosecuted in enforcement actions related to other actions. SEED tracks charges against 664 individuals during the Relevant Time Frame, of which 480 (72 percent) were charged in a Multi-Action Case and 184 (28 percent) were charged in a Multi-Defendant, Single Action Case. In Multi-Action Cases involving individuals, the SEC tended to charge individual defendants and firm defendants in separate enforcement actions. In 73 percent of Multi-Action Cases on SEED, the SEC prosecuted all implicated individuals and all implicated firms in separate enforcement actions.

In Multi-Action Cases, firms nearly always settled SEC charges concurrently with filing (70 percent), while most individuals did not settle concurrently (56 percent). The reasons for this are well-documented in the securities enforcement literature and apparent in SEED data. The SEC charges more individuals than firms with more severe charges. The severity of the charges increases the stakes of an enforcement action. Individuals thus tend to face penalties, like industry bars, with stakes relatively more significant than the financial penalties most often applied to firms. These factors in turn affect a defendant’s decision to fight or settle charges. Unsurprisingly, individuals are much more likely to fight their charges; firms, conversely, are much more likely to settle their charges, routinely concurrently with filing.

2. Multi-Action Cases: Cooperation

A second, related feature of Multi-Action Cases, which also suggests these cases facilitate the prosecution of multi-defendant cases, involves cooperation. Cooperation has long been recognized as a critical element in corporate enforcement. Among scholars of corporate crime and punishment, there is a growing appreciation for the practical difficulties in prosecuting large corporations, specifically the significant resources required to detect and prosecute fraud in large organizations. Cooperation significantly reduces these enforcement costs. Accordingly, enforcers have increasingly relied on cooperation, remediation efforts, and even self-reporting by alleged wrongdoers to improve their enforcement efforts. Throughout its history, the SEC has leveraged firms’ voluntary disclosure to assist its enforcement efforts. In 2001, the SEC issued a report on an investigation (the Seaboard Report) that detailed factors the agency considered in weighing possible sanctions, notably cooperation. In a recent article on the SEC’s cooperation efforts, practitioners observed that the “considerations laid out in the Seaboard Report became a new playbook for entities considering the steps to be taken when faced with actual or potential securities law violations.” Since the Seaboard Report, the SEC has taken further steps to formalize its cooperation program. SEED data suggests that cooperation is a crucial feature of SEC enforcement: the SEC credited at least one defendant in a case with cooperation in 62 percent of cases initiated during the Relevant Time Frame.

There are striking differences in cooperation patterns between Multi-Action Cases and Multi-Defendant, Single Action Cases. As illustrated in Figure 6, Multi-Action Cases feature the cooperation of one—but not all—defendants far more than in concurrent Multi-Defendants, Single Action Cases.

The patterns illustrated in Figures 6 and 7 are consistent with the observations on defendant type and settlement patterns. They are also consistent with observations in the securities enforcement literature, where it is well established that firms cooperate with the SEC more than individuals. During the Relevant Time Frame, 533 defendants out of a total 1,616 tracked on SEED cooperated with the SEC—roughly 33 percent of all defendants. Firms accounted for 93 percent of these cooperating defendants; individuals accounted for seven percent. Moreover, on SEED, cooperating defendants settled their charges concurrently with filing 92 percent of the time.

One possible explanation, which we simply introduce and propose as an object for further study, is that Multi-Action Cases afford the SEC some flexibility in managing cases against multiple defendants with different priorities. This flexibility might facilitate a divide and conquer approach. Under such an approach, the SEC flips one defendant (often the corporate defendant) against the other defendants (typically individuals) in the case. Multi-Action Cases facilitate this by allowing the SEC to leverage the cooperation of a defendant, not only during its investigation of the matter but also during the litigation of a parallel case. This feature of SEC enforcement—ongoing cooperation with the SEC after settlement in connection with a related enforcement action—has not, to our knowledge, been highlighted in the literature, but the SEC routinely secures this kind of cooperation in undertakings with defendants in their settlements.

As part of a settlement, defendants routinely agree to various undertakings, either in the consent decree or in the final enforcement document. Consider one example from the final order resolving the SEC’s administrative action filed in 2011 against FTN Financial Securities. In that order, FTN agreed to “cooperate fully” with any SEC “investigations, litigations or other proceedings relating to or arising from the matters described in the Order.” Such cooperation included an agreement to produce non-privileged documents, to cause FTN’s officers, directors, and employees to be interviewed by SEC staff, and even to “[u]se its best efforts to cause its officers, employees, and directors to appear and testify without service of a notice or subpoena in such investigations, depositions, hearings or trials . . . .”

Undertakings vary in form and substance. However, a standard cooperation undertaking that appears in many SEC documents is reproduced below:

In connection with this action and any related judicial or administrative proceeding or investigation commenced by the Commission or to which the Commission is a party, Defendant (i) agrees to appear and be interviewed by Commission staff at such times and places as the staff requests upon reasonable notice; (ii) will accept service by mail or facsimile transmission of notices or subpoenas issued by the Commission for documents or testimony at depositions, hearings, or trials, or in connection with any related investigation by Commission staff; (iii) appoints Defendant’s undersigned attorney as agent to receive service of such notices and subpoenas; (iv) with respect to such notices and subpoenas, waives the territorial limits on service contained in Rule 45 of the [FRCP] and any applicable local rules, provided that the party requesting the testimony reimburses Defendant’s travel lodging, and subsistence expenses at then-prevailing U.S. Government per diem rates; and (v) consents to personal jurisdiction over Defendant in any United States District Court for purposes of enforcing any such subpoena.

Other undertakings have been more explicit about the nature of the cooperation, and some undertakings expressly conditioned the resolution on the defendant’s prior and continued cooperation. In our study of all multi-defendant cases, the SEC secured undertakings to this effect from 249 defendants, of which 214 were part of Multi-Action Cases. The SEC’s required undertakings necessitate further study but point to how the SEC has used cooperation—both during its investigations and during its enforcement of related defendants—to efficiently prosecute cases involving parties.

One final, distinctive feature of Multi-Action Cases that also points to the divergent priorities of multiple defendants concerns the SEC’s choice of venue.

3. Multi-Action Cases: Venue

Perhaps no aspect of SEC enforcement is more hotly contested today than where the SEC prosecutes its cases—that is, the venue in which it charges defendants. The SEC can impose sanctions on alleged wrongdoers in two venues: it can initiate an in-house proceeding to decide the matter, or file a civil suit in federal court. In its early years, the SEC had limited flexibility, at least relative to today, to choose one forum over another. Over time, however, the SEC’s discretion to choose between forums expanded. In a recent article, Professors Choi and Pritchard describe the progressive expansion of the SEC’s administrative authority. The shift to administrative proceedings began in the 1980s when, along with expanding the types of penalties available to the SEC, Congress also granted the SEC the power to initiate a new form, for the SEC at least, of administrative proceedings to impose a cease and desist order. Like an injunction, a cease and desist order enables the SEC to pursue certain kinds of flexible relief through the administrative process. The scope of the SEC’s cease and desist authority expanded in the wake of Enron when Congress empowered the SEC to impose bars on officers and directors in administrative proceedings. The Dodd-Frank Act further expanded the SEC’s cease-and-desist authority, enabling the SEC to impose civil monetary penalties in the administrative forum. Therefore, today, the SEC has the power to pursue most of its enforcement remedies—including civil monetary penalties, disgorgement, and industry suspensions and bars—in administrative proceedings, without the need to file a suit in civil court.

This expansion of the SEC’s authority to choose where it enforces the law has met resistance. Criticisms of this expansion can be categorized roughly into two types. First is a series of what might be classified as doctrinal challenges, as summarized in a recent article by Professor Platt, that, broadly speaking, claim that the SEC’s power to seek this broad range of remedies in either venue exceeds the agency’s statutory authority—in other words, is unconstitutional. These arguments include claims of equal protection, among other constitutional arguments. The second criticism—the one more closely related to this Article—is a behavioral or empirical critique. It suggests that the SEC is, in practice, abusing its new-found discretion to choose where it enforces cases. Specifically, a study from the Wall Street Journal found that the SEC wins around 90 percent of the cases it adjudicates internally.

Lost in the growing body of commentary on the SEC’s choice of venue is the fact that the decision about venue is not necessarily binary—that is, in many cases, the SEC avails itself of both venues, charging some defendants in civil court and others in administrative proceedings. One such example is the SEC’s prosecution of the Galleon insider trading case. In 2009, the SEC charged hedge fund manager Raj Rajaratnam for operating a scheme that netted his hedge fund, Galleon Management, LP, over one million dollars in profits. The SEC also charged various other individuals and organizations for their participation in the scheme. In particular, in 2011, the SEC filed charges against an individual involved in the same scheme, Rajat Gupta, for illegally tipping Raj Rajaratnam, but unlike the twenty-seven defendants that preceded him who were charged in civil court, the SEC routed Gupta’s case to an administrative proceeding. Gupta sued the SEC on equal protection grounds, and Judge Rakoff seemed sympathetic to this argument in his decision, noting that the SEC charged Gupta with a similar set of violations stemming from the same underlying conduct as other defendants, but unlike those other defendants, in Gupta’s case it “decided it preferred its home turf.” Rakoff noted that “the SEC’s interest in taking these and other extraordinary steps was to selectively prejudice Mr. Gupta,” and he criticized what he called a “seeming exercise in forum shopping.”

More recent cases in which the SEC has availed itself of both forums include the agency’s prosecution of Herbalife for FCPA violations, in which the SEC filed settled charges with Herbalife in an administrative proceeding and also filed charges against a Herbalife managing director in civil court. And in its recent 2021 case alleging accounting fraud at Heinz, the SEC filed charges against Heinz and an executive in an administrative proceeding, and on the same date it filed charges against a different executive in civil court.

These cases, where the SEC avails itself of both venues (hereinafter Dual Venue Multi-Action Cases), make up a significant percentage of the SEC’s output. On SEED, nearly 24 percent of all SEC enforcement actions initiated during the Relevant Time Frame were part of a Dual Venue Multi-Action Case. These actions make cases that account for roughly 13 percent of all cases initiated by the SEC, in which the SEC named 432 defendants.

Dual Venue Multi-Action Cases exhibit patterns similar to those involving defendant type, settlement, and cooperation in Multi-Action Cases. In these cases, most individuals are charged in civil court (64 percent), while a minority are charged in administrative proceedings (36 percent). Conversely, most firms in Dual Venue Multi-Action Cases (68 percent) are charged in administrative proceedings, while a minority (32 percent) are charged in civil court.

Similar patterns appear in terms of settlement. Of the 93 Dual Venue Multi-Action Cases in our sample, defendants that settled charges concurrently with filing almost overwhelming were charged in administrative proceedings. Only five percent of defendants in Dual-Venue Multi-Action Cases charged in administrative proceedings did not settle concurrently. Conversely, 99 percent charged in civil court in these cases did not settle concurrently. These percentages are consistent with other cases—i.e., non-Dual Venue Multi-Action Cases—tracked on SEED.

Dual Venue Multi-Action Cases suggest that venue may be a carrot to induce defendant cooperation. In recent years, some scholars have suggested that certain defendants might actually prefer the administrative forum. For example, one study suggested that the SEC routs more politically connected defendants to administrative proceedings. On whether defendants might actually prefer settling in administrative proceedings, Professor Velikonja wrote:

Settled actions filed in administrative proceedings receive no formal external scrutiny. Settlement negotiations remain confidential until after the Commission approves the settlement in a closed hearing. Despite a handful of high-profile settlement rejections, judicial oversight was never terribly effective at policing settlement practices. Most judges perceived their review duties as limited and rubber-stamped most settlements. But sometimes, judges pushed back. As a result, both the SEC and defendants perceived in-court settlements as costly and unpredictable with ambiguous benefit to them. So it should come as no surprise that settlements are no longer filed in court.

Our data is consistent with the above. In most Multi-Action Cases, the SEC charged a firm and individuals connected with the same scheme, and the firm settled concurrently while the individuals did not. This is also true in Dual-Venue Multi-Action Cases. In twenty-eight of these ninety-three cases, all firm(s) settled charges concurrently, and no individuals settled charges concurrently. This number rises to fifty cases if we add the twenty-two cases where all firm(s) settled charges concurrently, and at least one, but not all, individuals, settled concurrently.

These patterns also appear to be connected to cooperation. As noted in the prior section, rarely did all defendants cooperate, and cooperating defendants were nearly always firms instead of individuals. In Dual-Venue Multi-Action Cases, defendants that cooperated were almost always charged in an administrative proceeding (89 percent), almost always settled the same day charges were filed (90 percent), and were prosecuted separately from non-cooperating defendants in most cases (in 90 percent of the 60 cases of cooperation). Another interpretation is that the SEC has gone out of its way to rout individuals, or more serious cases, to civil court—to either preserve due process for these defendants if not to protect itself against accusations of opportunistic venue choices, though more study is necessary to better understand these venue decisions.

Taken together, these features of Multi-Action Cases—cooperation, defendant type, and venue—suggest that these cases do not arise merely by procedural accident; they may be a procedural tool used by the SEC. These phenomena are worth exploring further, as they shed new light on aspects of the SEC’s enforcement program. More importantly for our present purposes, however, these cases clarify how not all cases are alike, and thus, accounting for these differences in calculating enforcement data is important. As we show in the following section, the SEC’s case filing decisions can impact how the agency’s enforcement data appears.

III. Reevaluating Action-Level Statistics with Cases

As noted in Part I, the SEC does not systematically publish information that identifies which of its enforcement actions are related—that is, are part of the same case. As such, the SEC’s enforcement data obscures important dimensions of each enforcement matter and of the aggregate enforcement landscape.

The purpose of this part is to show how case-level data reveals aspects of SEC’s enforcement that is unobservable in action-level data. We argue that scholars should account for, and the SEC should consider publishing, case-level data to complement action-level data.

A. Total Action Counts

We begin with perhaps the most significant metric published by the SEC: the overall number of enforcement actions initiated each year. The most notable trend concerning this figure in the last decade has been a marked increase in the overall number of enforcement actions the SEC initiated beginning around 2014, as SEED data illustrates in Figure 10.

Figure 11 layers in the total number of defendants charged and cases initiated in the same period, alongside the overall number of enforcement actions in our sample. It illustrates how the actions reported by the SEC sort into cases and shows how the number of cases initiated each year by the SEC is (unsurprisingly) lower than the number of actions. As described in Part II, many (26 percent of) SEC cases are comprised of multiple, related enforcement actions.

Figure 11 clarifies the case/action distinction and its relationship to the action data that the SEC reported. Consider, for example, FY 2020 in Figure 11: the SEC filed eighty-two actions tracked on SEED. That year, however, the SEC initiated fifty-seven unique matters, or cases. In theory, the SEC could have charged all defendants implicated in these matters in a single, consolidated enforcement action. Had the SEC done so, its action count for FY 2020 would equal the total number of cases initiated that year: fifty-seven. Conversely and more granularly, the SEC’s eighty-two actions in FY 2020 named a total of 105 defendants. Compare the FY 2020 results with FY 2013—in FY 2013, the SEC charged more defendants even though the SEC initiated fewer actions than it did in FY 2020.

Regardless of whether cases or defendants charged are more representative of enforcement intensity than actions, Figure 11 clarifies how two SEC enforcement choices impact the overall number of actions that the agency reports each year. Namely, the number of actions the SEC reported is affected by decisions about how many defendants to charge in a case, and then, for multi-defendant cases, whether to charge defendants in multiple, separate actions or in a single, consolidated action.

The most striking trend in our case-level data is the sharp increase in cases against only a single defendant, relative to other case types. As illustrated in Figure 12, these Single Defendant, Single Action Cases are especially well-represented in recent years, particularly those years in which the SEC broke enforcement records, as it touted in FY 2015.

The increase in Single Defendant, Single Action Cases since FY 2014 corresponds to observations in prior studies.

Single Action, Single Defendant Cases differ, in some respects significantly, from other types of SEC cases in terms of the resources required to prosecute such cases. First, Single Defendant, Single Action Cases rarely involved allegations of serious wrongdoing—that is, violations requiring a showing of scienter to prosecute. Prosecuting scienter-based violations is more difficult and more resource-intensive than prosecuting negligence-based or strict liability violations, as the latter only require proof of negligence or harmful intent. In our dataset, 85 percent of Single Defendant, Single Action Cases do not include any alleged violation requiring scienter. Conversely, 46 percent of defendants in multi-defendant cases (Multi-Defendant, Single Action Cases and Multi-Action Cases) on SEED are charged with scienter. Due to the SEED criterion, Single Defendant, Single Action Cases only capture firm defendants, while individuals are part of the multi-defendant cases (we elaborate on this later). Even excluding individuals in multi-defendant cases, and thus only comparing firms charged in Single Defendant, Single Action Cases to firms charged in multi-defendant cases, there is still a large distinction. Only 15 percent of Single Defendant, Single Action Cases against firms involved scienter; conversely, 35 percent of firm defendants in multi-defendant cases were charged with a scienter-based violation.

Second, most defendants in Single Defendant, Single Action Cases cooperated with the SEC. In Single Defendant, Single Action Cases, 69 percent of defendants cooperated with the SEC’s investigation. On SEED, the percentage of cases with at least one cooperator is lower in multi-defendant cases: 44 percent of multi-defendant cases feature at least one cooperating defendant, and only 22 percent of defendants in these cases cooperated (typically one defendant, a firm, cooperates, while individuals rarely are credited with cooperation). As scholars have pointed out, cooperating defendants significantly reduce the resources required to prosecute a matter.

Third, Single Defendant, Single Action Cases settled concurrently at a far higher rate than multi-defendant cases. Eighty-eight percent of defendants in Single Defendant, Single Action Cases settled on the same day charges were filed. Conversely, only 53 percent of defendants in multi-defendant cases (often the firm) settled charges concurrently when they were filed, while the remainder either settled later or litigated. In concurrently settled actions, the SEC does not need to expand enforcement staff to resolve the matter.

Fourth, the small number of litigated Single Defendant, Action Cases required less SEC enforcement attorneys to prosecute and resolved far more quickly than other case types. The number of enforcement attorneys tracked is based on what the SEC reported in press releases or enforcement documents—i.e., complaints filed in civil court.

These aspects, taken together, suggest that since FY 2014 the SEC has increased its focus on lower-cost cases, relative to other case types; these cases are the primary driver of the increase in output of SEC enforcement actions since FY 2014. We summarize the differences between Single Defendant, Single Action Cases and the other two case types in Table 1.

Table 1 Comparison of SEC Enforcement Case Characteristics by Case Type

Notable spikes of these one-off enforcement actions, in FY 2015 and FY 2019, are related to sweeps or issue-specific enforcement initiatives. These SEC enforcement sweeps are typically organized around a single substantive legal issue to encourage self-reporting. For example, in 2018, the SEC launched its Share Class Selection Disclosure Initiative (SCSD Initiative) to address recurring issues concerning mutual fund fees charged by investment advisers without adequate disclosure. As the SEC noted in its press release announcing the SCSD Initiative, it intended to “recommend that the Commission accept favorable settlement terms for investment advisers that self-report to the Division possible securities law violations relating to their failure to make necessary disclosures concerning mutual fund share class selection.” In those actions initiated as part of the SCSD Initiative, the SEC rarely expanded the scope of its investigation or pursued other potential wrongdoers beyond the firm that self-reported the violation. Accordingly, these enforcement sweeps generated a significant number of one-off cases—i.e., Single-Defendant, Single Action Cases. Since 2019, the SEC’s focus on these enforcement initiatives has expanded to include a well-publicized “off-channel” communications sweep pursuing recordkeeping violations.

To reiterate, we are primarily concerned with the increase in the number of Single-Action, Single Defendant Cases relative to actions in other SEC case types. This ratio is impossible to calculate without case-level data. While Single-Action, Single Defendant Cases are recorded in the same way as the SEC’s current action-level data, there is no way to identify which of the SEC’s published actions are connected to other actions. With case-level data, we are able to show an increase in Single-Action, Single-Defendant Cases—lower-cost cases relative to other case types.

The SEC’s increased focus on lower-cost cases has implications for the penalties the SEC collected, which we analyze in the next section.

B. Total Penalties

Behind the total number of enforcement actions, perhaps the second most important data reported by the SEC is penalty amounts. The SEC touts the aggregate dollar amount of penalties imposed and disgorgement recouped; in press releases, at times, it likes to show how penalties remain high relative to previous years, or even eclipse prior dollar amounts. For example, from FY 2010 to FY 2021, the SEC reported an increase in overall penalties, compared with the preceding year, in eight out of the eleven enforcement reports.

While SEED tracks only a subset of these actions (and thus penalty amounts), the shifting mix of case types documented above also appears to have impacted the types of cases from which the SEC draws its penalties. Unsurprisingly, different case types generated different penalty amounts. Single Defendant, Single Action Cases (the lowest-cost cases initiated by the SEC, which typically did not involve serious wrongdoing) tended to generate lower penalty amounts than other cases. Table 2 compares penalties in Single Defendant, Single Action Cases by defendant type with the other two case types.

Table 2 Monetary Penalties and Disgorgement in SEC Enforcement Cases by Case Type

And yet, as Table 2 shows, the penalty amounts in Single Defendant, Single Action Cases, while lower, are still substantial. Accordingly, in the aggregate and as illustrated in Figure 13, these cases contributed greater penalties assessed by the SEC, relative to other case types.

As such, the mix of cases initiated by the SEC, at least those tracked on SEED, appears to have shifted over time to include more one-off Single Defendant, Single Action Cases. This shift is also reflected in the types of cases from which the SEC generated its penalties.

This trend does not appear to be slowing. In its most recent annual report for FY 2023 (not captured by the data published in this Article), the SEC reported nearly $5 billion in penalties, by a wide margin the largest penalty amount reported by the SEC in its history. Based on the SEC’s own press releases, $1.6 billion of these penalties (nearly one-third) were secured in a major sweep targeting off-channel communications, a sweep comprised of actions that we expect (though cannot show at this time) will be one-off enforcement actions.

To sum up, case-level data contextualizes the SEC’s total number of actions reported and how these actions contribute to the penalty amounts the SEC has reported. This data allows us to better understand how the SEC allocates its resources. More study is needed to better understand SEC enforcement staff choices, and, in particular, the relative deterrent impact of smaller, one-shot cases versus more sprawling, resource intensive cases against multiple defendants.

The above discussion of the rise in Single-Defendant, Single Action Cases and the impact of these cases on action counts and penalty amounts clarifies the limits of action-level data. The case-types the SEC has initiated are not static year over year, nor are the SEC’s choices to consolidate Multi-Action Cases. Accordingly, action-level data is, at best, irregular and inconsistent, and, at worst, manipulatable. Case-level data, however, can address these issues and provide a complementary metric by which to assess the intensity of wrongdoing or SEC enforcement.

C. Individual Accountability

Beyond the total number of enforcement actions, perhaps the second most important data the SEC reported is the number of individuals subject to enforcement. In this section, we consider SEED data on enforcement against individuals to show how case-level data reframes the SEC’s action-level statistics.

Since the 2008 financial crisis, a recurrent critique of the SEC has focused on its failure to prosecute individual misconduct. Commentators contend that the SEC and other white-collar enforcement agencies focus too much on corporate prosecutions instead of targeting executives. Writing in 2020, Judge Jed Rakoff crystallized this critique: “High-level corporate executives, with only the rarest of exceptions, have become effectively immune from any meaningful prosecution for crimes committed on behalf of their companies.” Versions of this criticism have been made in different forms, based on different theories. Some commentators, for example, attribute the lack of individuals charged to prosecutorial reluctance; other commentators have theorized that political capture or revolving doors between the SEC and industry explain the lack of individual prosecutions. Most recently, Professor John Coffee argued that the lack of individual prosecutions results from agency resource constraints. Regardless of the theory, the concerns are the same: if the SEC does not hold individuals personally accountable, future wrongdoing will not be deterred.

The SEC has increasingly focused on the number of individuals that it prosecutes—and, important for our purposes, reports the number in terms of actions. For example, in its FY 2022 report, the SEC announced: “In fiscal year 2022, more than two-thirds of the SEC’s stand-alone enforcement actions involved at least one individual defendant or respondent.” The SEC began reporting on the percentage of actions involving at least one individual in its report for FY 2017; previously, the agency simply highlighted datapoints that evidenced the agency’s tough-on-individuals position. The percentage of actions in which the SEC named at least one individual has slowly trended down in the past five years. The SEC reported that 80 percent of the actions it initiated in FY 2017 named at least one individual; this measure dropped to 70 percent in FY 2018; 69 percent in FY 2019; 72 percent in FY 2020; and finally 70 percent in FY 2021.

However, even as the numbers have slowly declined, the SEC’s individual accountability metrics (again, measured in terms of actions) indicate that the SEC names a relatively large number of individuals in its enforcement proceedings—certainly more than the harshest criticisms of the agency suggest. The metrics also suggest that although the percentage of actions involving individuals has gradually declined since 2017, the SEC appears to be fairly consistent year to year in prosecuting individuals.

The SEC’s numbers also appear consistent with the few studies that have analyzed SEC enforcement of individuals. The most recent academic study of SEC enforcement of individuals, by Klausner and Hegland, analyzed SEC enforcements involving misstatements and omissions by publicly held companies from 2000 through 2014 and concluded that “commentators who believe that executives are rarely punished are incorrect.” Klausner and Hegland—who, notably, rely on case-level, as opposed to action-level, data—found that the SEC charged individuals in 88 percent of the cases in their dataset. Moreover, high-level executives faced significant penalties and were named in more cases than lower-level executives. Klausner’s and Hegland’s conclusions are consistent with preceding studies of SEC enforcement of firms versus individuals that have generally found that the SEC charges a significant number of individuals, consistent with the SEC’s own reporting.

Our case-level data, however, suggests a different story, particularly when analyzed in terms of cases, at least in the years following Klausner’s and Hegland’s study. As shown in Figure 14, the SEC charged individuals in 53 percent of actions tracked during the Relevant Time Frame. Figure 15 shows that, when analyzed at the case-level, the relative number of cases involving individuals is much lower: the SEC named an individual in only 37 percent of cases initiated during the Relevant Time Frame. The difference in the case- and action-level tracking is stark.

Here, we rush to qualify our data in two crucial ways. First, SEED only tracks a subset of actions that the SEC reports—i.e., those actions against public companies and subsidiaries, and related enforcement actions arising from the same underlying violation. This difference limits one-to-one comparisons between SEED data and the SEC’s data; it also explains why the percentage of actions naming an individual is lower on SEED than in the SEC’s reports. Second, and more importantly, SEED only tracks enforcement actions against individuals who are related to an enforcement action against a public company or subsidiary. In other words, SEED does not currently track enforcement actions that only name individuals and are not related to other enforcement actions naming public companies or subsidiaries. Thus, SEED would not track a case where the SEC charges the CEO of a public company but not the company itself. In one study, these isolated cases against individuals comprised 25 percent of all SEC cases in the study’s dataset.

Notwithstanding the differences in datasets, two preliminary conclusions can be drawn. The first, and most fundamental, concerns the use of actions to measure individual enforcement: if the SEC’s data were reported by cases, we expect the percentage of cases involving individual(s) would be different (likely lower) than the percentage of actions involving individual(s) the SEC reports.

By reporting actions as opposed to cases, the SEC’s data makes it impossible to understand how many discrete enforcement matters generated individual prosecutions. Based on SEED, cases involving individuals are particularly sensitive to this limitation: when the SEC prosecuted individuals, it overwhelmingly did so in multiple, related actions. Most actions in our dataset naming an individual (465 of 555 total actions involving individuals, or 84 percent) were connected to other actions and comprise Multi-Action Cases. Only 90 of the 555 were one-off actions (i.e., Multi Defendant, Single Action Cases) in which the SEC charged individuals alongside the public company or subsidiary in the same action.

One paradigmatic example of a Multi-Action Case involving individuals is the SEC’s case against Ameriprise Financial Services, a financial institution comprised of a network of investment professionals across the United States. In 2018, the SEC filed a settled enforcement action against Ameriprise, alleging that Ameriprise failed to properly supervise its investment representatives, certain of whom had misappropriated client funds, forged client signatures, and disbursed client funds without client consent (among other alleged violations). A week after filing the Ameriprise action, the SEC filed a second, related enforcement action against one of the representatives at issue; a few weeks later, the SEC filed two more enforcement actions, each naming a representative.

As the Ameriprise case illustrates, measurements of individual enforcement based on actions limit year to year comparisons and are subject to manipulation because they are affected by the SEC’s decisions to consolidate (or not consolidate) multiple defendants into one enforcement action. Cases do not suffer this limitation: they capture the entire universe of defendants charged in connection with a given violation and cannot be manipulated by the case filing decisions of SEC enforcement staff.

Second, though more speculatively, the percentage of SEC cases involving individuals appears to be lower today—potentially much lower—than previous research and the SEC’s action-level data suggest. Compared to Klausner’s and Hegland’s study, which found that the SEC charged individuals in 88 percent of the cases of their dataset, SEED shows a much lower percentage of cases involving individuals. The differences between SEED data and the statistics in Klausner’s and Hegland’s study are explained by differences in datasets, which analyze different time-periods and track different types of securities violations. Most importantly, SEED does not capture isolated cases against individuals, which comprise 25 percent of the cases in Klausner’s and Hegland’s dataset. And yet, if we assume the SEC brought an additional hypothetical 181 cases (25 percent of all cases on SEED) against individuals not tracked in our dataset, the number of cases involving individuals on SEED would only rise to 50 percent: far lower than the 88 percent previously reported by Klausner and Hegland and lower than the action-level percentages published by the SEC.

One possible explanation is that the SEC has brought fewer and fewer cases involving individuals since 2014. Klausner’s and Hegland’s study published SEC enforcement from 2000 to 2014; we publish SEED data from FY 2010–FY 2021. Figures 16 and 17 show the relative decline, again only based on SEED data with the limitations described above, in the past decade.

When analyzed at the case level, the decline is more pronounced, relative to cases in which no individuals are charged, as Figure 17 illustrates.

As Figure 17 shows, the SEC’s reporting on individual accountability, at the action-level, does not provide clarity about the actual percentage of matters that result in charges against individuals. For example, when the SEC reported that two-thirds of its matters involved individuals, this figure depends on a series of variables, including the relative number of cases against a single defendant, and the SEC’s choices to consolidate multi-defendant cases into a single action. If the SEC were to report its individual enforcement data in terms of cases, it would help clarify and make more consistent this reporting.

Conclusion

In this Article, we propose a distinction between “case” and “action”—terms frequently used interchangeably in the securities enforcement literature. In its reports, the SEC has also used the terms variably. At times, it appeared to use “case,” as we do, to refer to the clusters of enforcements tied to a particular matter. In other times, it used the term interchangeably with “action,” as it did in its FY 2023 enforcement report. As we clarify in this Article, defining these terms distinctly and shifting the analytical vantage from actions to cases as the basic unit of measurement used to analyze SEC enforcement matters reframes and recontextualizes certain of the SEC’s data. It also provides a more complete and consistent picture of the enforcement landscape relative to action-level data, which does not capture the relationships among enforcement actions.

More study is necessary to understand how aspects of Multi-Action Cases—namely, settlement, cooperation, and venue—empower the SEC with options to pursue sprawling cases against multiple defendants in ways that are cost effective. For example, does the SEC use a divide and conquer strategy to harness firm cooperation against individual defendants? If so, how does the nature of the defendant and the available venue affect the SEC’s calculus? How should we understand those cases where the SEC prosecutes multiple defendants in different venues? As various commentators have pointed out, the SEC must use its limited resources judicially to maximize deterrence of securities law violations. A clearer empirical picture of the SEC’s activity that considers cases—and the different types of cases—would clarify the agency’s decisions and better inform policy decisions that impact the SEC’s enforcement program.

The authors thank Professor Stephen Choi for his guidance on this project and very helpful comments as well as the participants at the New York University Lawyering Scholarship Colloquium. For their dedicated work on the SEED database, the authors thank the team of SEED Graduate Student Research Fellows, among them Jerry Chai, Ryan Huck, Arya Dhakal, Jaskiran Kaur, Risa Das, Rilwan Shittu, and Pilar Laitano. We are grateful to the reviewers of The Business Lawyer for helpful comments that improved this article.