Michael A. Perino is a Lecturer and Co-Director of the Roberts Program in Law, Business, and Corporate Governance at Stanford Law School. Mr. Perino's primary areas of scholarly interest are securities regulation and litigation, civil procedure, class actions, and complex litigation. Prior to coming to Stanford, Mr. Perino received his LL.M. degree from Columbia Law School, where he was valedictorian, a James Kent Scholar, and the recipient of the Walter Gellhorn Prize for outstanding proficiency in legal studies. Mr. Perino was also formerly associated with the New York law firm of Cadwalader, Wickersham & Taft, where he handled primarily securities actions and other complex federal litigation. He is a member of the New York bar.

Mr. Perino has written numerous articles on securities fraud and class action litigation, including: (1) Securities Litigation Reform: The First Year's Experience (A Statistical and Legal Analysis of Securities Fraud Litigation Under the Private Securities Litigation Reform Act), Working Paper No. 140, John M. Olin Program in Law and Economics Working Paper Series, Stanford Law School (Feb. 1997) (with Joseph A. Grundfest); (2) Class Action Chaos? The Theory of the Core and an Analysis of Opt Out Rights in Mass Tort Class Actions, 46 Emory L.J. ___ (forthcoming 1997); (3) The Pentium Papers: A Case Study of Institutional Investor Activism in Litigation, 38 Arizona L. Rev. 559 (1996) (with Joseph A. Grundfest); and (4) A Strong Inference of Fraud? An Early Interpretation of the 1995 Private Securities Litigation Reform Act, 1 Securities Reform Act Litigation Reporter 397 (1996).

Mr. Perino is also one of the principal developers of Stanford Law School's Securities Class Action Clearinghouse which has been nominated by the Smithsonian Institution for the 1997 Computerworld-Smithsonian Award as one of the five most important applications of information technology created by an educational institution in 1997.

I. Introduction

In the eighteen months since Congress passed the Private Securities Litigation Reform Act of 1995¹ (the "Reform Act" or the "Act") there has been an outpouring of analysis and speculation concerning how the Act has affected or might in the future alter securities litigation.² This written testimony reviews and updates the empirical literature analyzing what is perhaps the most basic question regarding the impact of the Reform Act: Has the Act caused a decline in the number of companies sued in securities fraud class actions?

The short answer to that question is that it is too early in the Act's life history to draw any reliable conclusions about long-term effects on filing rates. Nonetheless, an analysis of litigation activity in 1996 and 1997 reveals at least two significant points about the volume of securities class action litigation activity following passage of the Reform Act. First, the overall number of companies sued in securities class actions appears to be roughly equivalent to the number sued prior to the Act. Second, the relative stability of the aggregate litigation rate masks a significant shift of activity from federal to state court in an apparent attempt by plaintiffs to avoid the procedural and substantive hurdles Congress created when it passed the Reform Act. The data are also consistent with the hypothesis that the Reform Act and the Supreme Court's recent decision in Matsushita v. Epstein³ may have created strategic incentives to file parallel state and federal actions against a company in order to gain advantage in discovery and settlement.

II. Measuring Litigation Rates

This testimony uses the number of issuers sued rather than the number of class action complaints filed to measure changes in litigation activity. This statistic is the most reliable indication of litigation activity because in many class action securities cases a single issuer is sued in multiple complaints filed by different named plaintiffs represented by different plaintiffs' law firms.⁴ These complaints are often consolidated and litigated as a single proceeding. The number of issuers sued is, therefore, a superior predictor of the volume of post-consolidation litigation.

The Reform Act's effect on overall litigation rates, moreover, can be measured only by aggregating post-Reform Act litigation activity at the federal and state level and comparing that litigation rate with aggregate federal and state activity prior to the Act. In particular, recent experience suggests that plaintiffs may attempt to avoid the heightened pleading requirement and other provisions of the Reform Act by seeking alternative remedies through state court proceedings. If these alternative remedies are equally effective, we should observe a substitution effect that shifts litigation from the federal forum to state court without reducing the aggregate incidence of filings.

The aggregate litigation rate must also distinguish between issuers sued in state court only -- which provide strong indicia of substitute litigation -- and issuers sued in parallel federal and state court proceedings, a pattern which may be characteristic of litigation strategies unrelated to a pure substitution effect. In addition, issuers can be sued in state court on corporate law derivative claims which are not subject to the provisions of the Reform Act. The state court data must therefore be carefully filtered to identify only those claims and causes of action which are either substitutes for or parallels of claims that could otherwise be asserted as federal class action securities fraud claims.

III. Empirical Studies of 1996 Securities Class Action Litigation

Three empirical studies have examined securities fraud class action filings in 1996: (i) a study I co-authored with Joseph A. Grundfest⁵ (the "Grundfest-Perino Study"); (ii) a study conducted by National Economic Research Associates⁶ (the "NERA Study"); and (iii) a report prepared by the Securities and Exchange Commission⁷ (the "SEC Report"). All three studies observe similar trends, although the exact data reported in each differs slightly due to methodological differences and the difficulties inherent in collecting data on class actions filed in federal and state courts.⁸ All three studies also document an important unintended consequence of the Act. Although there is no evidence that the Reform Act has caused a significant decline in overall filing rates, plaintiffs' attorneys - in what appears to be a strategic response primarily meant to avoid the Reform Act's heightened pleading requirement and other procedural provisions - are increasingly suing issuers solely in state court proceedings under state blue sky or common law theories. Moreover, there is a significant increase in the number of companies sued in parallel federal and state proceedings, which may indicate attempts to avoid the Act's discovery restrictions.

A. Overall Filing Rates

None of the three studies of 1996 litigation activity find any significant overall decrease in securities class action filings in the wake of the Reform Act. The Grundfest-Perino Study was the first to analyze all of the class action litigation activity occurring in the first year under the Reform Act. Table 1 reports the number of issuers sued in 1996. These figures have been revised to include cases uncovered in additional research conducted after the study was published. A total of 150 issuers were sued in securities fraud class actions in 1996.⁹ Of these, 110 issuers were sued in federal court. These figures were likely depressed by certain one time effects, in particular an increase in filings prior to the effective date of the Reform Act and a "learning curve effect" that may have slowed filings as plaintiffs' attorneys determined how to best plead their cases under the new statutory regime.¹⁰ These twin effects appear to have reduced significantly the amount of litigation activity in the first quarter of the year.¹¹ Absent these effects, the study predicted that 163 issuers would have been sued.¹² While the estimated and observed litigation rates represent a decline of about 7% to 15% from the average of 176 issuers sued per year in the period from 1991 to 1995, there was a great deal of variance in the pre-Reform Act filings, and the number of issuers sued in 1996 was not materially different from the number sued in 1991, 1993, and 1995.¹³ As a result, the authors concluded that it was too early to draw firm conclusions with respect to the Reform Act's effect on overall litigation rates.¹⁴ The SEC Report contains filing figures that are not materially different from those found in the Grundfest-Perino Study.¹⁵

Slightly different results are contained in the NERA Study, which finds for example no net decline in federal class action lawsuits in 1996.¹⁶ NERA also identifies slightly more filings than the other studies. In part, the differences in these figures may reflect differences in what each report is counting ("federal filings" versus "issuers sued"). The Grundfest-Perino Study also seems to have taken a more conservative approach with respect to whether a case should be included within its database than the NERA Study. Ultimately, the only way to reconcile these different figures is to analyze each study's underlying data. Unfortunately, unlike the Grundfest-Perino Study, the NERA Study does not list every filing included in its 1996 count, thus an accurate comparative analysis is not possible. The empirical studies agree, however, that there is no evidence of a net increase in federal class action lawsuits in 1996 and that overall filing rates appear to have been little changed in 1996.

B. The Shift to State Court Litigation

All three studies find a significant increase in state court class action filings as a result of the Reform Act, although there are again differences in the exact number of actions identified. The studies agree, however, that this increase in state court litigation is one of the most significant developments in securities litigation in the wake of the Reform Act.¹⁷

The Grundfest-Perino Study documents that the relative stability in overall filing activity masked a significant shift in litigation activity from federal to state court. Table 1 demonstrates that a total of 70 issuers were sued in state court proceedings in 1996.¹⁸ Forty issuers, or approximately 26.7% of class action claims, were sued in state court proceedings without parallel federal claims filed in 1996. Additionally, 27.3% of federal class action securities fraud cases in 1996 had pending parallel state claims. These figures likely undercount the level of state court activity due to the difficulty in compiling data on state class action activity, although it remains unclear just how much of an undercounting problem exists.

The NERA Study reaches similar conclusions with respect to state court filings, but it likely overcounts the relevant amount of state court activity. The NERA Study reported seventy-eight state court filings through October 1996.¹⁹ The discrepancies between the figures these two studies report are likely due to methodological differences. The NERA Study includes all cases compiled by Securities Class Action Alert ("SCAA").²⁰ But SCAA reports two very different kinds of state court cases, the increase in only one of which is likely tied to passage of the Reform Act. The filing of class actions alleging misrepresentations or omissions in the purchase and sale of securities in state court is likely evidence of a substitution effect. These are the only cases included in the Grundfest-Perino Study of state court filings. Companies have traditionally been sued in state court class or derivative actions alleging breaches of officers' and directors' duties of loyalty, care, or candor. SCAA reports these kinds of state court filings as well.²¹ Because the NERA state court filing figures appear to include these cases, which would have likely been filed in state court regardless of whether the Reform Act was passed, their data should not be used as a measure of the substitution effect to state court.

The Grundfest-Perino Study attributed the increase in state court litigation to two phenomena. First, a substitution effect was thought to be present whereby plaintiffs' counsel file state court actions when the underlying facts appear not to be sufficient to satisfy the heightened pleading standard found in the Reform Act or where plaintiffs are attempting to avoid other procedural or substantive hurdles of the Act.²² In other words, instead of eliminating these cases, as proponents of the Reform Act had hoped, the Act appears to have shifted the relative profitability of class action litigation for this subset of cases in favor of state court. Second, the Grundfest-Perino Study and the SEC Report suggest that one explanation for the increased use of parallel proceedings is that plaintiffs are attempting to circumvent the Reform Act's discovery stay provisions by seeking discovery in state court actions.²³

Certainly there may be reasons unrelated to attempts to avoid the provisions of the Reform Act that may make a state forum appear to be more desirable than a federal forum. For example, the Grundfest-Perino Study notes that the Supreme Court's recent decision in Matsushita v. Epstein²⁴ may have increased the incentives to file in state court because it clarified that state court settlements may resolve federal Rule 10b-5 cases that could not have been brought in state court.²⁵

Other commentators have noted additional reasons unrelated to the Reform Act that they suggest may lead attorneys to conclude that state court is a more advantageous forum, including the availability of non-unanimous juries, differing liability standards, broader jury pools, and differing procedural rules.²⁶ The difficulty in attributing the sudden rise in state securities filings to these procedural advantages is that they were readily available in state court prior to the passage of the Reform Act. Yet before 1996, anecdotal evidence suggests that securities fraud class action lawsuits were rarely filed in state court.²⁷ To test the accuracy of these anecdotal reports, I reviewed all new cases reported in SCAA during 1994. SCAA is the most readily accessible source of pre-Reform Act filing data. Data from 1994 was chosen to eliminate any effects associated with passage of the Reform Act in late 1995.

Table 2 suggests that there has been a real shift in state court shareholder litigation patterns since the Reform Act. Of the forty-one new cases identified in SCAA in 1994 as being filed in state court, only one case involved misrepresentations or omissions in the sale of publicly-traded securities. The remaining cases involved either breaches of fiduciary duty in connection with a merger or other corporate transaction (thirty-two cases) or allegedly fraudulent activity involving limited partnerships, bonds, or brokerage accounts (eight cases). By contrast, Table 3 demonstrates that the vast majority of state court class actions filed since the Reform Act (81.5%) involve securities that trade on national markets. These cases typically involve allegations that the price of the company's securities was inflated due to misrepresentations or omissions, precisely the kinds of claims that were most often filed in federal court prior to the Act. The timing of the sudden upswing in the appearance of these cases in state court strongly supports the inference that the shift in forum selection was driven by the passage of the Reform Act.

IV. Securities Class Action Litigation in 1997

What has happened to securities fraud filings in 1997? Has the movement to state court observed in 1996 continued?

Table 1 demonstrates that state court class actions continue to be filed in 1997, although the number of state filings has decreased from 1996 levels despite a significant increase in 1997 federal filings. From January 1, 1997 through June 30, 1997, at least twenty-two issuers were sued in state court securities fraud class actions, for a total of ninety-two issuers sued in state court since the passage of the Reform Act. Of the companies sued in 1997, fourteen were sued solely in state court.²⁸ Annualizing these data suggests a 30% drop in 1997 in the number of companies sued solely in state court proceedings. Only eight issuers were sued in parallel federal and state actions in 1997, although an additional nine issuers that were sued in 1997 federal actions had previously been sued in 1996 state court actions. Together, 20.5% of 1997 federal filings had parallel state actions, a decrease of 6.8% from 1996 levels. At the same time there has been a substantial increase in federal filings. Through June 30, 1997, eighty-three issuers were sued in federal class actions. Even adjusting for the one-time effects that appear to have decreased filings in the first quarter of 1996, the 1997 data still suggest a 33.9% increase in the number of companies sued in federal court.

These data suggest an annualized total of 194 issuers sued in state and federal court in 1997, a 29.3% increase from 1996 and a 10.2% increase from the average number of companies sued in federal court per year in the period from 1991 to 1995. The 1997 estimate, however, is still within the variance found in the pre-Reform Act filing data, thus tending to confirm the Grundfest-Perino Study's conclusion that the Reform Act has not caused a net decline in securities class action activity.

There is simply too little data to determine whether the decline in 1997 state court filings is a short-term aberration or whether there has been a real decline in the use of state court proceedings as a means to circumvent the Reform Act. There are a number of plausible explanations for the decrease in identified state filings in 1997. One possibility is that 1996 state court filings may have been inflated due in large part to certain one-time effects, such as the prospect that Proposition 211 would pass in California. Absent those effects state litigation may have returned to a more "normal" rate. Alternatively, plaintiffs' attorneys may have begun to determine that the costs associated with state court litigation outweigh its benefits. A number of state court actions filed in 1996 were dismissed.²⁹ Plaintiffs also may have had difficulties certifying nationwide classes under state law, which would tend to decrease the economic viability of a case. State court actions could still be useful if plaintiffs were able to obtain discovery that would be stayed in a federal proceeding or if discovery in the state action allowed plaintiffs to file a later federal proceeding. But several state courts were also willing to stay discovery in state class actions, thereby potentially decreasing the utility of state litigation.³⁰

These factors may suggest to some that state litigation may return to its pre-Reform Act levels. While such a permanent decrease is possible, other factors suggest that such a conclusion may be premature. Despite the observed declines in 1997 state litigation, the shift to state court litigation remains one of the most significant consequences of the Reform Act. Overall, approximately ninety-two of 238 post-Reform Act litigations (38.6%) involve at least some state component. The percentage in 1997 of state securities class actions that involve publicly-traded companies (72.7%) is still significantly greater than before the Reform Act (11.1%). As a result, it is possible that the current decline in state filings does not represent a long-term rejection of state courts. Rather, plaintiffs' attorneys may have strategically chosen not to pursue a significant number of state cases in order to decrease the apparent necessity for Congress to pass a federal preemption statute.

Moreover, the state case law remains undeveloped and inconsistent. Many courts have permitted significant discovery to proceed, even where a parallel federal action is pending.³¹ Very few cases have been subject to appellate review.³² Absent federal legislation, whether and to what extent filing state court proceedings will remain a viable strategy option for plaintiffs' attorneys will depend on whether state courts consistently resolve these issues against plaintiffs. Such consistency may never come about or may take a significant amount of time to develop.

V. Conclusion

It is too soon to draw any firm conclusions from these data with respect to the effect the Reform Act will have on the aggregate number of securities class action lawsuits filed per year. Eighteen months of data are simply insufficient to draw conclusions as to long-term trends associated with passage of the Act. This difficulty is compounded by the difficulties associated with obtaining a complete census of state court class action activity. Nonetheless, these data suggest that the Reform Act appears to have had a modest effect at best on aggregate securities litigation activity. The more significant effects associated with the Reform Act appear to be the substitution effect that has shifted the venue for much of this litigation from federal to state court and the newly created incentives to file parallel litigation in state court.