This danger can be avoided, however, if Twombly and Iqbal are read carefully and with due regard for authoritative pre-Twombly sources such as the text of the Federal Rules, their Forms, and Supreme Court decisions that remain good law. Doing so reveals an approach that is not necessarily inconsistent with the notice-pleading framework that most attorneys, judges, and professors alive today learned when they were in law school. To urge this approach is not to praise Twombly and Iqbal. At best, they appear to be result-oriented decisions designed to terminate lawsuits that struck the majorities as undesirable. And incorporating the “plausibility” concept into pleading doctrine has proven troublesome (to say the least). Courts should not, however, compound these problems by misreading Twombly and Iqbal to drastically change federal pleading standards going forward.

I. Misreading Twombly and Iqbal

Many critiques of Twombly and Iqbal presume that these decisions have (1) abrogated core pre-Twombly precedents on pleading standards and (2) made “plausibility” the central inquiry at the pleadings phase. A careful reading of Twombly and Iqbal, however, refutes these presumptions.

First, the significant pre-Twombly authorities on pleading remain good law, even after Twombly and Iqbal. There is only a single instance where either Twombly or Iqbal explicitly abrogates earlier precedent; Twombly put into “retirement” the statement from Conley v. Gibson that “a complaint should not be dismissed for failure to state a claim unless it appears beyond doubt that the plaintiff can prove no set of facts in support of his claim which would entitle him to relief.” The Twombly majority read this “beyond doubt . . . no set of facts” language as precluding dismissal “whenever the pleadings left open the possibility that a plaintiff might later establish some set of undisclosed facts to support recovery.”  Read this way, Conley’s phrase is indeed problematic—a complaint that alleged nothing more than “The planet Earth is round” would survive, because any number of actionable facts might be consistent with the Earth being round.  That Twombly “retire[d]” this view should not be cause for concern.

To be fair to Justice Black and his Conley opinion, this now-discredited phrase was subject to a far more sensible reading than the straw man that Twombly struck down. Its true meaning was simply that speculation about the provability of a claim is typically not a proper inquiry at the pleadings phase; provability is relevant only when it appears “beyond doubt” that the plaintiff cannot prove her claim. As to this point, Twombly fully agrees: “a well-pleaded complaint may proceed even if it strikes a savvy judge that actual proof of those facts is improbable, and that a recovery is very remote and unlikely.” The dispositive question is—and always has been—what makes a complaint “well-pleaded”?  Again, Conley and Twombly provide precisely the same answer: “[A]ll the Rules require is ‘a short and plain statement of the claim’ that will give the defendant fair notice of what the plaintiff’s claim is and the grounds upon which it rests.”  Twombly not only endorsed this crucial “fair notice” language from Conley; it also relied on many of the Court’s other pre-Twombly cases, including the unanimous Swierkiewicz v. Sorema decision from just five years earlier.

Admittedly, Twombly/Iqbal’s “plausibility” inquiry is a concept that was absent from the pre-Twombly pleading framework. But under Twombly and Iqbal, plausibility is not a basis for disregarding allegations in a complaint. Rather, plausibility is the second step in a two-part framework that actually makes the framework more lenient than it would be without it. Here are the two steps:

(1) Identify allegations that are mere legal conclusions, and disregard them for purposes of determining whether the complaint states a claim for relief.

(2) Determine whether the remaining non-conclusory allegations, accepted as true, plausibly suggest an entitlement to relief.

It follows that when a complaint contains non-conclusory allegations on every element of a claim for relief, the plausibility issue vanishes completely. A court must accept non-conclusory allegations as true. If such allegations address each element that would be needed to ultimately prove the plaintiff’s claim, then they do more than make an entitlement to relief “plausible”—they confirm an entitlement to relief, at least for purposes of the pleadings phase. Under this framework, only conclusoriness is grounds for refusing to accept an allegation as true (step one). The plausibility inquiry (step two) is not destructive, but generative. It allows a plaintiff to overcome the fact that a key element was alleged in a merely conclusory manner, provided the entitlement to relief is plausibly suggested by the valid, non-conclusory allegations. Properly understood, then, the true impact of Twombly and Iqbal will depend on how courts decide what allegations are “mere legal conclusions” that can be disregarded at step one of the new framework.

II. What Conclusory Means

Neither Twombly nor Iqbal provide much guidance on how to define “conclusory” for purposes of this new approach to pleading. It would be a serious mistake, however, to define conclusory in evidentiary terms—that is, to allow courts to disregard an allegation merely because its truth is not suggested by some other allegation in the complaint. While this is a common understanding of the term (judges and litigants often describe assertions that lack evidentiary support as “conclusory”) it should not be transplanted to the pleading context for several reasons.

An evidentiary approach would be flatly inconsistent with pre-Twombly authority that remains good law. Form 11’s model negligence complaint deems it sufficient to allege merely: “On <Date>, at <Place>, the defendant negligently drove a motor vehicle against the plaintiff.” It does not require allegations indicating that the defendant was, in fact, driving negligently. Form 18’s complaint for patent infringement, using the example of electric motors, deems it sufficient to allege: “The defendant has infringed and is still infringing the Letters Patent by making, selling, and using electric motors that embody the patented invention.” Form 18 does not require allegations suggesting that the defendant’s electric motors, in fact, embody the plaintiff’s invention. And the Supreme Court’s unanimous 2002 decision in Swierkiewicz—an employment discrimination case—held that federal courts “must accept as true all of the factual allegations contained in the complaint,” including Swierkiewicz’s allegation of his employer’s discriminatory intent.³ Swierkiewicz squarely rejected the idea that the complaint must indicate the availability of supporting evidence or facts suggesting that the allegations might be proven indirectly. The Court noted that the discovery process might reveal evidence of discrimination that was not yet known. It therefore found it “incongruous to require a plaintiff, in order to survive a motion to dismiss, to plead more facts than he may ultimately need to prove to succeed on the merits if direct evidence of discrimination is discovered” even where the plaintiff is “without direct evidence of discrimination at the time of his complaint.” Moreover, an evidentiary approach would be conceptually unworkable, because each new allegation offered to support an earlier allegation would itself require support. If taken to its logical extent, an evidentiary approach imposes on courts an endless cascade of inquiry that can never be satisfied.

A better approach is to define conclusory in transactional terms: an allegation is conclusory only when it fails to identify adequately the acts or events that entitle the plaintiff to relief from the defendant. What made the crucial allegations in Iqbal and Twombly impermissibly “conclusory” were legitimate (though certainly debatable) questions about whether those allegations were grounded in a series of real-world events. In Twombly, the Court was unsure whether the complaint contained an “independent allegation of actual agreement” or merely presumed that parallel conduct alone constituted an agreement for purposes of the Sherman Act. In Iqbal, it was unclear from the complaint what acts by Ashcroft and Mueller themselves were undertaken with discriminatory intent. These are not concerns about provability, but rather concerns about whether the complaints had identified a tangible set of events that would entitle the plaintiffs to relief.

Reasonable minds can differ over whether the crucial allegations in Twombly and Iqbal should have been deemed conclusory, and the Court’s spotty reasoning on that issue in both opinions is troubling. That said, the pre-Twombly examples discussed above are qualitatively different from a transactional standpoint. As cursory as Form 11 is, it identifies in “short and plain” terms the liability-generating event: the defendant negligently driving his car against the plaintiff. Form 18 does the same, identifying the plaintiff’s receipt of a patent for electric motors and alleging that the defendant is “making, selling, and using electric motors that embody the patented invention.” Swierkiewicz also provides a straightforward transactional narrative: the plaintiff was employed by the defendant and he was fired because of his age (fifty-three) and national origin (Hungarian). To be sure, these complaints characterize the underlying events with some conclusory language (e.g., that the Form 11 defendant drove “negligently”); but the liability-generating events themselves are identified in a fairly concrete way. This was arguably not so in Twombly and Iqbal, given the uncertainties over whether the Twombly plaintiffs had made an “independent allegation of actual agreement” and which (if any) decisions by Ashcroft and Mueller themselves were motivated by discriminatory animus. Again, one can legitimately question the Supreme Court’s reading of the Twombly and Iqbal complaints. But such concerns should incline courts to proceed cautiously in applying Twombly and Iqbal, rather than overreading them to compel a drastically more strict approach to pleading.

III. Closing Thoughts

The lower federal courts are paying a lot of attention to Twombly and Iqbal. Twombly is already the seventh-most-cited decision of all time, with about 24,000 federal-court citations in less than three years on the books. Iqbal is just getting started, but during its first year it has been cited at a remarkable rate of over 500 decisions per month. Of course, an opinion’s citation frequency alone does not reveal what courts are actually doing with that opinion. There are certainly federal courts who have used Iqbal to impose a much stricter pleading standard. Some lower federal courts, however, are on the right track.

One example is the Third Circuit’s decision in Fowler v. UPMC Shadyside,⁴ a disability discrimination case. Interestingly, the opinion starts by stating that Iqbal effectively overruled Swierkiewicz. But what follows in Fowler is vintage notice-pleading, entirely consistent with Swierkiewicz. The Third Circuit accepted Fowler’s allegations as true, without regard to whether they were “plausible,” and chastised the district court for taking a stricter approach. As to the complaint’s allegation that Fowler was disabled, the Third Circuit wrote:

At this stage of the litigation, the District Court should have focused on the appropriate threshold question — namely whether Fowler pleaded she is an individual with a disability. The District Court and UPMC instead focused on what Fowler can “prove,” apparently maintaining that since she cannot prove she is disabled she cannot sustain a prima facie failure-to-transfer claim. A determination whether a prima facie case has been made, however, is an evidentiary inquiry . . . . 

The Third Circuit also accepted the plaintiff’s allegations that she was “terminated because she was disabled” and that UPMC discriminated against her by failing to “transfer or otherwise obtain vacant and funded job positions” for her. It held that these were “‘more than labels and conclusions’ or ‘a formulaic recitation of the elements of a cause of action’” and concluded: “We have no trouble finding that Fowler has adequately pleaded a claim for relief under the standards announced in Twombly and Iqbal.”

There are also post-Iqbal decisions that affirm dismissals but do so in a way that is faithful to pre-Twombly principles. In Brooks v. Ross,⁵ the Seventh Circuit found the following allegation to be insufficient:

Plaintiff is informed, believes and alleges that the Defendants while acting in concert with other State of Illinois officials and employees of the Attorney General’s Office, Department of Corrections and Prisoner Review Board did knowingly, intentionally and maliciously prosecute Plaintiff and Ronald Matrisciano in retaliation for Plaintiff and the said Ronald Matrisciano exercising rights and privileges under the Constitutions and laws of the United States and State of Illinois.  

The Seventh Circuit’s reasoning was founded on notice principles, not the newfound plausibility inquiry. It wrote: “[T]his paragraph fails under Iqbal, because it is merely a formulaic recitation of the cause of action and nothing more. It therefore does not put the defendants on notice of what exactly they might have done to violate Brooks’s rights under the Constitution, federal law, or state law.” The Seventh Circuit explicitly rejected the argument “that Twombly had repudiated the general notice-pleading regime of Rule 8” and concluded: “This court took Twombly . . . to mean that at some point the factual detail in a complaint may be so sketchy that the complaint does not provide the type of notice of the claim to which the defendant is entitled under Rule 8. This continues to be the case after Iqbal.”

As in Twombly and Iqbal, one can certainly question whether the key allegation in Brooks should have been disregarded as conclusory. But such questions were present even under a notice-pleading regime. Notice pleading was never a blank check. No court would have been forced to accept as true an allegation that “the defendant violated the plaintiff’s legal rights in a way that entitles the plaintiff to relief.” So there always lurked the question, what “notice” was required? As Judge, Dean & Federal Rules architect Charles Clark observed, a notice-pleading framework is not inherently a lenient one. It depends on what “content [is] given to the word ‘notice.’”

Likewise, a pleading standard that allows courts to disregard conclusory allegations is not inherently a strict one. It depends on how “conclusory” is defined. Twombly/Iqbal’s recognition that conclusory allegations need not be accepted as true does not necessarily mean the end of notice pleading. It merely cloaks the inquiry in different doctrinal garb. The silver lining might be this: Before Twombly and Iqbal, lower courts’ handling of this issue was often able to fly under the radar. Recent empirical studies on the effect of Twombly and Iqbal reveal remarkably high dismissal rates even before Twombly. After Twombly and Iqbal, however, the world is watching. And many judges, commentators, legislators, and rulemakers are giving more thought than ever to what pleading standards ought to look like.

Acknowledgments:

Copyright © 2010 Stanford Law Review.

Adam Steinman is a Professor of Law at the University of Cincinnati College of Law. Effective July 2010, he will be Professor of Law and Michael J. Zimmer Fellow at Seton Hall University School of Law.

This Legal Workshop Editorial is based on the following Law Review Article: Adam Steinman, The Pleading Problem, 62 STAN. L. REV. 1293 (2010).

Compelling anecdotes tend to attract rigorous analysis, and indeed, in the business context, the economics literature has long grappled with precisely how and why the timing of information release affects market response. The legal literature, however, has been comparatively devoid of either theoretical or empirical analysis of the policy context. Few scholars have sought a theoretical account of when agencies act, as opposed to how agencies act (what procedures are used) or what agencies say (what substance is promulgated). Our Article seeks to remedy this oversight by constructing a theoretical and empirical analysis of the timing of agency action. Simply put, once an agency has made a policy decision internally, when will that decision be announced to the public and why does it matter?

To the extent that there is conventional wisdom about the use of timing by agencies, it is that the visibility of agency actions can be significantly reduced or eliminated if actions are announced during a holiday or weekend news cycle. The micro-foundation for this view is generally left unspecified, and, on close examination, it is somewhat inconsistent. Is it that journalists simply do not register government actions taken on Friday afternoons? Do representatives of interest groups whose members potentially have millions of dollars at stake in agency decisions head out early for their vacation homes and never bother to check what happened the week before?

We suggest that the conventional anecdotal wisdom about the bureaucracy burying bad news is likely simply wrong or incomplete, at least in its most typical form. This critical claim is part conceptual and part empirical. A primary conceptual difficulty with the conventional account is its assumption that issuing policy during low-visibility time periods makes monitoring costs essentially so high such that no one will observe the hidden policy and hence that it will be all but impossible to mobilize political opposition. A problem for this view has to do with short-term versus long-term equilibria. Even if this were a successful short-term strategy, it is difficult to construct a long-term equilibrium in which such behavior is rational. If the media and interest groups do not pay attention to agency decisions announced on Friday afternoons or holidays, then surely it is in an agency’s interest to announce certain decisions at those times. But once interest groups or reporters get wind of the practice, they should pay extra-special attention to an agency’s Friday and holiday announcements. And if the level of public attention is no less intense on those days, facilitated by a now 24-hour news cycle, then it will no longer be a best response for the agency to announce controversial policies at such times. The conventional wisdom describes a set of strategies that is off the equilibrium path.

A related reason the conventional account falters is that administrative agencies in the United States are some of the most extensively monitored government actors in the world, in large part because of legal constraints imposed on agencies by statutes like the Administrative Procedure Act. Almost all policy decisions an agency makes must be published in the Federal Register for all to see. Even informal policies that are not legally binding typically are publicly available. More specifically, most legally binding agency rules require notice and an opportunity for public comment by any affected interests—comments to which the agency must adequately respond. A Notice of Proposed Rulemaking (NPRM) is typically open for at least sixty days so that comments can be taken.  A final rule generally does not go into effect for at least thirty days to allow for notice and legal challenges to take place prior to implementation; major rules, in particular, cannot take effect for at least sixty days. With some notable exceptions, final policy decisions by federal agencies in the United States are stunningly visible, even if the internal decisionmaking process of agencies is not entirely transparent. The idea of agencies hiding controversial policy actions by announcing them on Friday afternoons and running for the door is about as silly as Gulliver hiding among the Lilliputians by covering his eyes. The actions may not produce leading newspaper headlines by the time Monday rolls around, but that does not mean there is no one watching.

In addition to these conceptual concerns with the conventional account, an empirical analysis of actual agency rulemaking actions over twenty-five years reveals mixed evidence of timing manipulation by agencies. Although important agency rulemaking decisions are slightly more likely to be issued on Fridays or weekends, this form of strategic timing manipulation is not correlated with political or institutional conditions commonly thought to drive agency desires to reduce the visibility of decisions. The manipulation of timing may be less tied to days of the week than times of the year, however. Rules producing an impact on state government, for example, appear more likely to be issued during a congressional recess than during other periods. In short, the empirical analysis suggests a more complex approach than that advanced by the conventional account.

The constructive part of our Article is to develop a more nuanced or complex approach. The average agency is monitored by a diverse mix of public actors and private interest groups. The Article’s theoretical innovation is to emphasize the relationship between timing, monitoring costs, and selective participation by interest groups in the agency policymaking process. Instead of using timing to hide decisions, we argue that agencies can make strategic timing decisions to affect the monitoring costs of Congress, the White House, interest groups, the media, and the general public. Again, this does not, as is commonly asserted, block the visibility of agency actions. Rather, it drives a shift in the population of potential monitors for any given agency action. Some of this monitoring is formal. The White House, for example, reviews agency rules and significant guidance documents before they are issued. Congress creates and funds agencies, prescribes specific responsibilities, and often supervises work using information requests, committee hearings, and other oversight tools. Courts review the procedure and substance of agency actions relying on an extensive body of statutory and doctrinal tools. Less formally, interest groups and members of the public often track agency actions and may petition other political actors as well as the agencies themselves to shift regulatory outcomes. The media brings agency deeds and misdeeds to light too.

Each potential monitor naturally has preferences about the substance of agency policy, usually preferring that an agency’s final decisions be as close to its preferences as possible. Monitoring the bureaucracy, however, is not costless. Interest groups monitoring agency action must balance the benefits of monitoring agency behavior with its costs. As a result, it is generally only groups with something at stake that are willing to bear the costs of monitoring. Suppose that each of these interest groups has a different expected benefit from monitoring agency decisions, perhaps because they have different concerns or because they have different abilities to respond to decisions. If so there will always be some group for whom the existing marginal cost of monitoring is near equal to the marginal return from monitoring. Any factor that increases the costs of participation will make the expected returns from monitoring negative. These interest groups, for whom it was just barely worth participating given the existing costs and benefits, will cease monitoring when monitoring costs increase.

When monitoring costs increase, the groups with the most at stake will continue to monitor because the marginal cost is still much less than the marginal benefit. The composition of interest groups monitoring agency decisions is now different, however. Because the remaining groups have preferences, in aggregate, different from the exiting group, the response to agency action taken on a low-visibility day may be different than the response the agency would have received had the policy been announced on a high-visibility day. It is not that no one is paying attention or that the agency has succeeded in hiding its actions. Rather, the composition of actors who are paying attention has changed. Given that the agency itself decided when to announce its decision, it stands to good reason that the new group of monitoring interest groups will produce a public reaction more in keeping with the agency’s underlying preferences.

Importantly, monitoring costs are a partial function of the timing of agency decisions. In many cases, timing will be a trivial share of overall monitoring costs, and for some interest groups the change in cost structure will be unimportant; that is, it will result in no observable behavioral change. However, so long as there is some actor who was just willing to pay the monitoring costs before an increase, there will be some actor who will cease to do so when monitoring costs increase at all.  The conventional account suggests that policies announced on a Friday afternoon are forever lost in the news cycle. It is as though these policies are subsequently implemented behind closed doors, forever locked away. More plausible is simply to say that announcing policies on Christmas Eve or when Congress is out of session forces monitors to exert more effort to observe the policy decision. It requires business associations or nonprofits to pay someone to be on call or in the office. Moreover, and likely more importantly, it also increases the costs of publicizing the objectionable action and mobilizing a political response. Almost no monitor is able to simply stop the agency from moving forward alone. Changing the policy requires notifying and organizing other actors in the political process. Simplistically, but accurately, the costs of doing so increase after hours, on weekends, when Congress is out of session, or on major holidays. Put in the colloquial language of the political science, the timing of agency action affects the costs of both police patrols and fire alarms.

The effectiveness of this strategy, however, will vary depending on the type of underlying action. Existing procedural restrictions in the law ensure that this strategy may be exceptional rather than typical. The strategy may work sometimes for final rules, but it should work much more effectively for a subset of less prominent agency actions. Where delay, transparency, and judicial scrutiny are not built into the administrative process, the prospect of strategically manipulating the timing of decisions is more sensible.

The promulgation of final rules, for example, is typically associated with a delay before implementation and an extensive set of possible grounds for challenging the decision in court. The delay rule facilitates monitoring and makes strategic timing a more difficult strategy to use effectively. Part of what makes the conventional timing story less than wholly compelling for final rules is that such rules are usually proposed, considered for many months with extensive public comments, announced, and then implemented only after affected parties have considered whether to challenge the decision in court. Both NPRMs and Notices of Inquiries—typically mandatory before issuing new policy—are explicitly designed to generate public attention and allow for interested parties to participate in the regulatory process. Such decisions are “running public performances”: they are on display for all to see, and while they often do not last forever, there is no meaningful sense in which the performance can be hidden from view.

For the subset of once-in-time decisions, however, timing could play a much greater role. Withdrawals of previously proposed rules or the abandonment of existing agency process, for example, occur without prior notice and comment or an ex post lag. For reasons we discuss at some length, it can be more difficult to challenge such withdrawals in court. Immediate scrutiny and a nonjudicial political reaction will be more important. The empirical analysis of twenty-five years of rulemaking actions suggests that the manipulation of timing is more common for the withdrawal of proposed rules than for NPRMs or final rules. Especially after shifts in administration, many of the most controversial agency decisions will be whether to implement or withdraw rules started by prior administrations.

If our critique and reformulation of the timing of agency action is persuasive, more attention should be paid to rulemaking withdrawals as a class of administrative actions. At present, the abandonment of proposed rulemakings is largely an absentee category in administrative law. Courts tend to treat withdrawals differently than other forms of agency decisions, though explicit discussion of such agency action is sparse. There is consensus that courts can review them if specific statutory schemes explicitly contemplate the abandonment of proposed action or if the agency faces a mandatory duty to regulate. Conflict currently exists, however, among courts as to whether review of other (typically discretionary) withdrawals is permissible. Although there are good reasons for distinguishing withdrawals of unfinished rulemakings from the enactment of new rules and the rescission of old rules, the differential treatment in the law makes timing more important for withdrawals than for other agency decisions. Given the spike in rulemaking withdrawals after a presidential transition (typically the abandonment of rulemakings that were started but not completed under the previous administration), rule withdrawals should occupy a more central role in administrative law scholarship. Yet, because withdrawals combine features of both agency inaction and agency policymaking, balancing the competing doctrinal imperatives to protect agency discretion and to keep agencies accountable presents serious challenges for administrative law.

* * *

Politicians and journalists may like to swap anecdotes and innuendo of administrative agencies using timing to avoid public scrutiny. Much of this anecdotal behavior, if systemic, would seem perplexing, as a theoretical matter—off the so-called equilibrium path—and insensitive to the reality of regulatory politics. We have tried to bring both conceptual clarity and empirical rigor to these stories. Our view is not that timing is unimportant, but that timing influences regulatory politics in a somewhat different way than common intuition suggests. Timing decisions are best understood as strategic decisions by agencies that can make it more difficult, other factors being equal, for watchers to interfere with their policy implementation. The timing of action makes effective monitoring of agency action more costly, which in turn should change the universe of interests participating in agency process. While agencies may prefer to reduce the visibility of their actions, for most actions, strategic timing of their issuance does not prevent visibility, but instead simply increases the difficulty of generating political opposition in Congress.

Our empirical study also shows a more complex picture of agency timing decisions. Although significant rules are more likely to be announced on Fridays and during congressional recesses, other political or institutional variables that one might expect to be associated with the timing of such announcements are not. Political conditions seem to matter more for rule withdrawals, a subset of agency actions less subject to ex ante viewing or ex post challenge. Because rule withdrawals are more difficult to challenge in court, rule withdrawal is one of the few types of agency policies for which announcing in a lower visibility environment does in fact raise monitoring costs substantially.

Our aspiration is that our conceptual, empirical, and doctrinal work provides an analytic framework for courts and commentators to pursue renewed work on questions of timing in the administrative state. In addition, our analysis of timing reveals the importance of the abandonment of rulemaking proceedings. There is administrative law doctrine on such withdrawals, but it is underdeveloped. While we have sketched some preliminaries on this front, there is much work to be done. And that work is particularly relevant now as control of the White House has recently shifted and critical regulatory decisions are thus likely to be made in the withdrawals domain.

Acknowledgments:

Copyright © 2010 University of Chicago Law Review.

Jacob E. Gersen is a Professor of Law at the University of Chicago Law School.

Anne Joseph O’Connell is a Professor of Law at the University of California, Berkeley School of Law.

This Legal Workshop article is based on Jacob E. Gersen and Anne Joseph O’Connell, Hiding in Plain Sight? Timing and Transparency in the Administrative State, 76 U Chi L Rev 1157 (2009).

Very useful comments were provided by Ken Bamberger, Eric Biber, Tino Cuellar, Dan Farber, Jesse Shapiro, Matthew Stephenson, Adrian Vermeule, and John Yoo. Financial support has been provided by the Hellman Family Faculty Fund, the Boalt Hall Fund, the University of California, Berkeley, Committee on Research, and the Jerome Kutak Fund at The University of Chicago Law School. Thanks to Tess Hand-Bender, Roman Giverts, Monica Groat, Edna Lewis, Harry Moren, Stacey Nathan, and John Yow for research assistance.

For many years, the traditional law and economics literature advocated strongly against legal transition relief. Led most prominently by Louis Kaplow, scholars argued that we should treat legal transitions no differently from other types of transitions faced by societal actors, for which the government does not provide relief. Recent commentary, however, questions the scope of Kaplow’s claim. Scholars have pointed out that considerations of efficiency, incentives for socially desirable investments, governmental legitimacy, and fairness might justify legal transition relief.

Assuming then that transition relief is appropriate under at least some circumstances, we identify two centrally important questions for which scholars have yet to find satisfactory answers. First, while societal actors often hedge against transitions in technology and the economy by obtaining insurance in the private market, such a market does not exist with respect to legal transitions. We attempt to explain its puzzling absence. Second, commentators who advocate transition relief in limited circumstances do not confront the critical question of what institutional structure is best designed to ensure that transition relief is meted out only where justified and in an appropriately limited form. We suggest a framework for allocating authority over transition relief within the government.

First, there exists no meaningful market for regulatory insurance in the United States—not even a market for insurance against government takings (which would appear to be a much simpler endeavor). In attempting to explain this gap in the market, scholars have pinned the blame on a variety of the usual economic culprits: moral hazard, adverse selection, and the difficulty of finding uncorrelated risks. But we do not believe that these effects, or any combination of them, can explain the lack of a private market.

Moral hazard problems exist when an insured party engages in behavior—particularly as a consequence of purchasing insurance—that the insurer has not priced into the contract. So long as coverage is based upon a firm’s current business rather than hypothetical developments and new product lines, the insurer should be able to accurately price the firm’s behavior. A fixed-payment insurance contract that provided a lump-sum payout in the event of regulation, rather than one that protected a firm against its full losses, would protect the insurer against threats of moral hazard. Similarly, insurers would need to protect themselves against lobbying that increased the likelihood of regulation. Insurers should again be able to cure these moral hazard problems through contract. The parties simply could write regulatory insurance contracts to ban any lobbying activities by insured firms (and to force them to take no public position on relevant regulatory action). And it should not be difficult for insurers to monitor this type of activity. Thus, it seems unlikely that private insurance markets have failed to arise due to unavoidable moral hazard problems.

Several scholars have suggested that adverse selection problems are likely to plague systems designed to insure against takings or regulation. For instance, homeowners who know that they are more likely to be subject to takings will opt into insurance plans at higher rates. Like threats of moral hazard, however, adverse selection problems depend at their core on information asymmetries. If all relevant information is public, insurers can price contracts accurately, and higher-risk private parties who wish to opt in will be able to do so only at elevated rates. Information asymmetries may be present in the context of takings of real property, but they are unlikely to plague more general regulatory insurance. The key to the adverse selection problem for takings insurance is that both the vast majority of the relevant governmental action and the potential insured parties are local, while the principal insurers are not. It is this geographic and political divide that gives rise to the necessary informational asymmetries. The types of economic regulation that concern us here, by contrast, are rarely local; they are almost always created by state and federal governmental entities. There is negligible private information about these types of regulation (except the information held by the government actors themselves), and so regulated firms possess essentially no informational advantage over their putative insurers. Without such an asymmetry, there can be no problem of adverse selection.

Finally, state and national regulation can have potent and widespread effects, particularly if it comes from a populous, highly industrialized state such as California or New York. The difficulty in assembling a portfolio of truly uncorrelated risk positions in the face of such widespread single-event threats might be preventing a robust market for regulatory insurance from forming. We do not, however, believe that this is the case. Well-conceived regulatory insurance would cover only one (or a finite number) of the potential business risks to a firm. A potential insurer could select which of these many available risks it is willing to assume, knowing that any individual regulation would lead only to a partial decrease in firm value—not the complete destruction of the firm. In a competitive marketplace of multiple insurers, any given firm should be able to find one or more insurers willing to take on some slice of risk. Thus, a fear of correlated risks cannot account for the complete absence of a market for regulatory insurance.

Rather, we believe that the major impediment to a private market for legal transition insurance is the chore of pricing. In comparison to typical accidents, significant regulatory acts occur extremely infrequently, usually numbering just below one thousand per year nationally. Even this description overstates their quantity in the same way that a reporting of all fires, floods, automobile accidents, and illnesses would overstate the effective number of insurance claims (and thus the number of useful data points) in a given year. Each federal agency issues no more than a handful of regulations each year, and thus any given regulatory field is altered only rarely. Without a broader pool of data to draw upon, an insurance firm cannot reliably estimate the hazards presented by any given regulation. In addition, unlike traffic accidents or house fires, regulatory acts are effectively one-off, nonstochastic events. An individual fitting a given demographic profile in 2005 is, for the most part, equally likely to have an automobile accident as a similarly situated individual in 2006; what variation exists is captured by the easily obtainable demographic information that insurers collect. Accordingly, data from 2005 are useful in predicting 2006 outcomes, data from 2004 are useful in predicting both 2005 and 2006, and so on. The likelihood of a particular regulation, on the other hand, depends upon a wide variety of factors, the impact of which is often unobservable or unpredictable. A shift in agency leadership or political priorities, a transfer of governmental power, a change in membership or chairmanship on a key committee, or even new developments in science or technology (or even culture) can affect the probability of any given regulation in any particular field in unforeseeable ways.

Worse still, regulation in one period is not necessarily a good proxy for regulation in another period. The fact that the EPA has acted once to regulate the level of arsenic in drinking water has ambiguous effects on the likelihood that the agency will act again, either to raise or lower allowable levels. It may indicate that a similarly situated EPA will tighten the arsenic standard; it may lead the EPA to learn that the current level of protection is needlessly high and prompt a relaxation of those limits; or it may simply indicate that the EPA already has selected a near-optimal level of regulation and that the status quo is likely to persist. Based on available quantitative data alone, an outside observer has almost no capacity to select among these possibilities. Even the meaning of potential explanatory variables can change over time, and often rapidly. Democrats in Georgia in 1972 were very different than Democrats in Georgia in 1992, who were in turn very different than Democrats in Georgia in 2006.

These pricing difficulties imply that government efforts to foster a private market in insurance will be ultimately unsuccessful; the informational difficulties are too great, and the government lacks a means of surmounting them. In addition, even more exotic options such as information markets and regulatory derivatives will not serve as workable substitutes. The same pricing problems that inhibit private insurance will prevent firms from investing in regulatory derivatives in quantities necessary to make them a useful hedge.

In the absence of a private market for transition relief, government-provided relief remains a viable option. The key to our solution is the disaggregation of transition relief into various steps and the allocation of individual duties based on institutional competency. We proceed in three stages. First, we unbundle the various steps that compose transition relief, and we explain how the decisions or decisionmaking involved in some of those steps differ from those involved in other steps (and from decisions and decisionmaking in the ordinary regulatory context). Some decisions regarding transition relief are more akin to plenary lawmaking. These decisions affect numerous societal actors and draw their resolution from broad societal values. Here, one might think of the broad decision of whether transition relief is warranted in the first instance. Other decisions are more in the nature of applications of an existing legal structure to particular private actors. Numerous issues that arise in transition relief settings are highly technocratic (as opposed to value-laden) in this sense: for instance, whether a modification of an existing building should subject the structure to regulation as if it were new construction; whether a transaction was consummated before or after the advent of a new legal regime; and how to allocate limited funds or grandfathering rights.

Second, consider the role of expertise in making these narrow, technocratic decisions. To be sure, expertise in the particular area of law at issue is of some value. But that kind of expertise is often exceeded in value by more general expertise in meting out transition relief. Questions that arise in these decisions transcend particular areas of law. The question of whether a modification should be treated as a new construction arises in environmental law, land use law, and disabilities law, to name just a few areas. The question of whether a transaction should be deemed consummated before or after a legal change takes effect arises in tax law and bankruptcy law. And the question of how to distribute limited funds or grandfathering rights arises in environmental law and natural resources law. This strongly suggests that a single government agency could accumulate considerable relevant expertise were it charged with handling such transition relief decisions across various legal specialties.

Finally, we observe that private actors will naturally be willing to invest money and time to obtain transition relief, and government actors will face an incentive to mete it out in return for private benefits. A government actor that is charged with distributing transition relief—even in accordance with some set legal scheme—likely will enjoy some discretion in making those decisions. The less that a government body is subject to lobbying by outside influences, the less it will fall prey to private rent-seeking in the allocation of transition relief. Accordingly, we conclude that an independent agency might be best situated to make some decisions related to the provision and application of transition relief.

Acknowledgments:

Copyright © 2010 NYU Law Review.

Jonathan Masur is a professor at the University of Chicago Law School.

Jonathan Nash is a professor at Emory University Law School.

The arguments advanced in this Article are narrowly aimed at functional brain imaging evidence—both images and expert testimony—addressed to issues regarding criminal mens rea. In narrowing our argument in this way, we are explicitly not analyzing the probative value of structural images such as those produced by CT scans or regular MRIs. We are also not critiquing the use of medically-accepted brain scans introduced to prove relevant medical diagnoses (such as brain trauma) when the presence and extent of the medical diagnosis is a key issue of causation, such as in a tort claim. The link between the brain’s structural and functional integrity and its complex mental functions is at the crux of our critique; at present, this link is a matter of scientific research and speculation, rather than diagnostic certainty. Some of our arguments may generalize to other courtroom uses of functional brain imaging; certainly, we expect the scientific background to be useful in many cases that attempt to use functional neuroimaging to make claims about complex mental processes such as lying, experiencing pain, or recalling memories. However, as the fMRI image construction is incredibly complex and its admissibility is specific to the purpose for which it is being offered, we sought to restrict our analysis to the admissibility of fMRI when used by a criminal defendant to argue that he did not have the requisite intent to be held culpable for the charged offense.

Functional brain images are specialized scientific evidence, but they are also just evidence. In order to be admissible, images must be relevant, authenticated, and more probative than unfairly prejudicial; cannot be impermissible hearsay; and, if introduced by an expert, must satisfy the tests for expert scientific testimony in that jurisdiction. This Article focuses on the first three requirements, but does not by exclusion suggest that other rules pertaining to scientific evidence might be incapable of properly evaluating and excluding fMRI images. In particular, the analysis of the minimal probative value of fMRI images could similarly be argued in a Daubert or Frye analysis about the data’s reliability, validity, and general acceptance in the relevant scientific field.

The weak probative value of fMRI evidence with respect to past mental states is best made clear by understanding the process through which an fMRI image is constructed. While a brief summary is provided below, this complex process is covered in greater detail in our full Article.

fMRI is a relatively safe and non-invasive medical technique that indirectly measures the brain’s activity. The emphasis on indirectly cannot be stressed enough. The brain is made up of millions of electrically-active cells called neurons, which use tiny electrical pulses and chemicals to communicate with each other via networks that enable sensation, perception, movement, cognition, emotions, and consciousness. fMRI does not directly measure this neuronal activity, but rather uses a physiological correlate to approximate areas of activity. Because neurons do not have internal reserves of energy, when they send electrical pulses in response to some activity, oxygen-carrying blood must be transported to the neurons to replenish the cell’s metabolic reserves. This is called the “hemodynamic response.” Blood that is carrying oxygen behaves differently in magnetic fields than deoxygenated blood. The difference in the magnetic properties of oxygenated blood allows fMRI to detect changes in blood flow related to activity. This is called the Blood Oxygen Level Dependent (BOLD) response. In simple terms, when a region of the brain is “activated” in response to a perception or to enable a behavior, that region receives more oxygenated blood. Because oxygenated blood behaves differently in a magnetic field, the large magnet in the fMRI device can measure this influx. If the local oxygen use is more than adequately supplied by the influx of blood, then a positive BOLD response will be observed. If the local demand for oxygen exceeds that provided by the regional blood flow, then a negative BOLD response will be observed. The change in the blood oxygenation level in a spatial volume called a voxel (which can be conceptualized as a three-dimensional pixel), does not directly capture the activity of neurons; thus, fMRI does not yet provide detailed physiological information about specific neural mechanisms underlying any specific mental state. Each new finding about the nature of the BOLD response suggests that our understanding of its mechanistic basis and correlation to neuronal activity is just scratching the surface. Understanding this phenomenon at the level of the neuron is critical to understanding if it is capturing little, some, or most of the brain’s actual neuronal activity in the creation of a mental event. In turn, this knowledge is necessary to bridging the gap between a particular cognition or behavior and the neural mechanism underlying it.

Without more confirmed data on what the BOLD response is capturing, inferences about individuals’ past mental states that are drawn from fMRI evidence will be quite useless in a legal setting that demands a strong, or at least probable, causal story. The task-dependency of fMRI may be obscure to the fact finder and, in any event, should severely limit the probative value of the resulting image. This is particularly true given the lack of ecologically valid behavioral tasks we have at our disposal that subjects could conduct in an fMRI scanner. What, exactly, would an assessment of an intentional, knowing, reckless or remorseful mental state be comprised of? Even well-validated behavioral tasks, such as those used to assess executive functions (e.g., judgment, decision making, attentional focus and switching), do not directly map onto Model Penal Code psychological mental state constructs such as purpose, knowledge, recklessness, and negligence. At best, deficiencies in task performance and correlated abnormalities in brain activity are several inferential steps away from a legal lack of intent or lack of capacity—steps so large they are almost certainly fatal to a legally relevant causal story.

In evaluating the weak probative value of fMRI addressed to past mental states, this Article analyzes cross-disciplinary expertise from neuroscience, philosophy, and legal literature. Such concepts as neuroessentialism, ecological validity, the “psycho-legal error,” the “epistemic mismatch” between what the images can say and what they appear to say, the appearance of objectivity and reductionism, and the allure of the colorful image of the brain itself, are explained and evaluated in order to illustrate why the probative value of fMRI evidence is weaker than it might appear at first blush. Further technical details about how fMRI images are constructed expose the many points at which an opportunistic expert could “dial a defect.” Statistical analysis is a critical and often obscure part of all elements of fMRI, from behavioral assessment to filtering functions that attempt to find meaningful “signals” against biological background “noise.” Examples of “dialing a defect” include fudging the statistical threshold, sampling various software packages until you find one that reveals a defect, manipulating base rates, or drawing inferences based on a peculiar reference class of normals. A recent satirical publication showed how researchers were able to demonstrate neural activation in a dead fish’s brain just by throwing enough analytical darts and eventually having something stick—a result obviously attributable to statistical chance. This is possible because of the high level of noise in the blood flow signal on which fMRI relies, and we are also careful to point out this possibility of finding chance activation in a forensic context.

An fMRI image is not a snapshot of an individual’s brain. While thoughtful authentication of the image itself might serve to counteract this “epistemic mismatch,” this is only one hurdle that the evidence needs to overcome. One of these hurdles is a problem of time travel. fMRI can only measure a defendant’s present mental state, which can be linked to his past (and criminally relevant) mental state only by way of inference and supposition. Others have argued that adversarial cross-examination is a powerful tool that can counteract the potential for prejudice and deal with such inferential leaps. But without any professional standards related to the software package, template brains, behavioral tasks or statistical analysis that is appropriate to use in any given setting, such courtroom conversation would waste the court’s time and mislead the jury. While fMRI methodologies are constantly evolving, the purpose of this Article is to provide thoughtful context on how the methodologies would need to improve in order to have the kind of probative value that would warrant their admission as evidence in a court of law. As the methodologies employing fMRI may one day improve such that its probative value is no longer eclipsed by the potential for unfair prejudice, we offer at the end of our Article a non-exhaustive checklist that judges or counsel can use to authenticate the functional brain images and assess the weight these images should be accorded by fact finders.

At this stage, we can only speak to the potential for unfair prejudice, as data is currently being gathered as to just how prejudicial functional brain images are on mock jurors. Even so, a few preliminary findings suggest that the mere presentation of brain images can make a bad argument appear more sensible. We address other forms of unfair prejudice as well, such as fMRI’s potential to mislead jurors, be cumulative of other cheaper psychological evidence, and waste the court’s resources and time.

The familiar story is one of weak circumstantial evidence, unsympathetic defendants, and impressive scientific reports and images. When you mix these three together, you have a prescription for potential injustice that is so compelling it could sway even the most skeptical jurors and convince them that the elements of a weak case are proved beyond a reasonable doubt. If, on the other hand, it is the defendant that catches the court’s sympathies, then the science may swing in the opposite direction and fMRI could be used to make a weak defense appear stronger. This tale has been told before with phrenology, the polygraph, and countless other forensic technologies that have since been discredited. Improper reliance on each of these untested and unreliable technologies has led to unjust outcomes.

Science can appear to be beyond the reach of human distortion. As a result, the more the scientific evidence relies on complex technologies like computers or imaging devices, the greater the risk that it may be endowed with powers to solve hard legal problems. Litigants have used this fact to their advantage, and throughout history have stretched scientific findings to fit them to pre-existing legal conclusions. This may be what is happening with fMRI. The device is not yet capable of capturing past mental states, but because the criminal law may sometimes be desperate to prove the unprovable, we anticipate an increase in proffered evidence and testimony based on this new technology.

Acknowledgments:

Thanks to the MacArthur Funded Law and Neuroscience Project, Jeff Cooper, David Faigman, Jaime King, Michael Saks, Hank Greely, Stephen Morse, Kathryn Abrams, Walter Sinnott-Armstrong, Ryan Calo, R. Duncan Luce, William Uttal, and Sean Mackey for their thoughtful comments on earlier drafts of this paper. Special thanks to Nita Farahany and Adina Roskies for their in-depth feedback and guidance.

Copyright © 2010 Stanford Law Review.

Teneille R. Brown, JD, is an Associate Professor of Law at the University of Utah, S.J. Quinney College of Law. While writing this Article she was a fellow with the Stanford Center for Biomedical Ethics, a fellow at the Stanford Law School Center for Law & the Biosciences, and a research fellow with the MacArthur Foundation Law and Neuroscience Project.  Emily R. Murphy, PhD, is a JD candidate in the Stanford Law School class of 2012.  While writing this article she was a fellow in the Stanford Law School Center for Law and Biosciences and research fellow on the MacArthur Foundation Law and Neuroscience Project.

This Legal Workshop Editorial is based on the following Law Review Article: Teneille Brown & Emily Murphy, Through a Scanner Darkly: Functional Neuroimaging as Evidence of a Criminal Defendant’s Past Mental States, 62 STAN. L. REV. 1119 (2010).

The doctrine through which the Supreme Court accomplished these results remains relevant today. For example, total invalidity is exactly the outcome sought by the Virginia Attorney General in challenging the individual mandate component of the recently enacted health care reform bill. Like others who have sought to invalidate wide-ranging legislation in the past, the Virginia Attorney General’s lawsuit seeks to use severability doctrine to leverage the alleged unconstitutionality of one statutory section to accomplish complete invalidation of a statute that is almost one thousand pages long. And, if the individual mandate is unconstitutional (admittedly a big “if” under current law), severability doctrine could very well enable Virginia’s slingshot lawsuit to take down the entire legislative Goliath of federal health care reform.

When a law is partially unconstitutional, severability doctrine governs whether a court may first separate out or “sever” the unconstitutional provisions or applications of a law, and then subtract or “excise” them so that the constitutional remainder can be enforced going forward. The lodestar for the severability determination is legislative intent. A court considering whether to sever asks whether the legislature would have preferred what is left of its statute after severance to no statute at all, and whether the statute can function without the unconstitutional part. If the answer to either question is “no,” then the court does not sever but instead declares the statute entirely invalid and enjoins its enforcement in all applications.

A legislature may try to control the severability determination with a severability or inseverability clause, but under current doctrine such clauses are neither necessary nor sufficient to control the judicial determination of severability. Additionally, the inquiry into legislative intent required by severability doctrine is unlike many other interpretive inquiries, in that it explicitly requires a determination of what the legislature would have done, not what the legislature actually did. The doctrine thereby leaves courts with significant discretion in deciding whether and how to sever and excise statutes.

The existence of this judicial discretion is troubling given the potency of the doctrine—severability allows the invalidation of an entire statute based on a problem with just one part. The doctrine can also enable invalidity to expand beyond unconstitutionality even when it yields the conclusion that particular statutory provisions are severable. For example, in the Supreme Court’s deployment of the doctrine in Booker v. United States,¹
the Court “severed and excised” two statutory provisions—one that made the United States Sentencing Guidelines binding and another that governed appellate review of federal sentencing. As a consequence of the Court’s supposition about what the enacting Congress would have intended about a problem that Congress neither anticipated nor addressed when legislating, the Federal Sentencing Guidelines are now nonbinding even in the thousands upon thousands of cases in which binding Guidelines pose no constitutional problem.

If severability doctrine is threatening because of the significant discretion it vests in courts to invalidate a statute beyond just those uses and provisions that are unconstitutional, why do we have it? Mostly, it seems, because of the belief that we cannot do without it. Severability doctrine is thought by many to be the only way of dealing with partial unconstitutionality under the law as it now stands.

The problem is a big one because, while many laws are unconstitutional, few are entirely so. As a result, partial unconstitutionality is pervasive. Sometimes a statute’s unconstitutionality resides in a discrete textual provision, such as a legislative veto in a complex immigration statute or a private cause of action for gender-motivated violence in a massive crime bill. More commonly, a statute’s unconstitutionality inheres in some applications of the statute but not in others, such as a speech restriction that is unconstitutional as applied to speech on public sidewalks but not elsewhere, or heightened protection for religious liberty that is unconstitutional as applied to states and localities but constitutional as applied to the federal government.

The problem of what to do with partially unconstitutional laws is not only pervasive but persistent. ² for instance, resulted in a holding of partial unconstitutionality. And, though it is often unnoticed, almost every instance of judicial review in present times involves an issue of partial unconstitutionality.

Operating on the assumption that severability is inevitable in one form or another, scholars have leveled many criticisms at its doctrinal particulars and offered proposals for reform. But nothing has changed. Modern severability doctrine exists in much the same form today as when it first emerged.

This article attempts to change the terms of the debate by demonstrating that, contrary to the assumption underlying modern scholarship, the counterfactual speculation required by modern severability doctrine is not necessary to a workable system of judicial review. In fact, legislative intent–based severability analysis did not emerge until the 1850s. Before that time, there was a way to specify the scope of invalidity flowing from a holding of unconstitutionality that did not require suppositions about the enacting legislature’s intent. An examination of judicial practice in the first several decades of judicial review following Marbury reveals this method and disproves the prevailing notion that judicial review without modern severability doctrine is unworkable—a widely accepted notion that rests on an anachronistic attribution of modern severability doctrine to Chief Justice Marshall, his immediate successors on the federal bench, and their state-court contemporaries. Whatever else may be said about the first several decades of post-Marbury judicial review in the United States, it was workable. And it worked without modern hypothetical intent–based severability doctrine.

Judicial review can work that way again today. Most fundamentally, it is necessary to change the foundational metaphor that structures our thinking about judicial review—namely, a shift from excision to displacement. The familiar excision-based approach to judicial review implies that a court has the power to eliminate unconstitutional provisions by a process of subtraction. In contrast, under a displacement-based approach, a court does not excise anything from a statute but instead determines the extent to which superior law displaces inferior law in resolving the particular case before it.

After changing the foundational metaphor, a question remains about the scope of displacement. First principles dictate that superior law can displace no less of the inferior law than is repugnant to it, but what about more? The proposed approach instructs courts not to infer invalidity beyond unconstitutionality. The goal is for the enforceable law resulting from an exercise of judicial review to be traceable to a combination of constitutional command and legislative provision, rather than judicial supposition about what the legislature would have wanted.

This approach is grounded on four principles drawn from the earlier approach to partial unconstitutionality. First, the function of judicial review is to determine what law governs in the course of resolving a particular case. More specifically, judicial review is to be understood as the process of determining whether, and to what extent, the Constitution supersedes otherwise operative law in resolving the particular case before the court. Second, upon recognition of a conflict, superior law must displace inferior law, which is thereby deemed void. Third, a law may be void in part and valid in part, and the voidness of one part has no effect on the validity of the other parts. Fourth, a law is void to the extent of its conflict with the Constitution, but no further.

In the proposed approach, there is no “next step” that looks to hypothetical legislative intent. Certainly, if the partially unconstitutional statute contains an explicit fallback provision or some type of inseverability clause, that provision or clause may be given effect (as long as it is not itself unconstitutional). But for most statutes, which contain no such provision or clause, the scope of invalidity will be coextensive with the scope of unconstitutionality.

The recovery of the earlier approach to partial unconstitutionality enables a before-and-after comparison that upends the dominant view that severability doctrine saves partially unconstitutional statutes from total invalidity. Before severance doctrine emerged, displacement-based partial unconstitutionality doctrine was already functioning in this manner: Courts routinely held an unconstitutional law void to the extent of repugnancy but no further; there was no “next step” in which courts inquired into whether the legislature would have preferred no law at all to the constitutional remainder. After severance emerged, partial unconstitutionality (now seen as resulting from excision and not displacement) became dependent on the satisfaction of a legislative-intent test. The “next step” of a legislative-intent analysis injected into judicial review the possibility that invalidity extends beyond unconstitutionality. That is, whenever severability doctrine resulted in a holding of inseverability, modern severability doctrine enables total invalidity to flow from partial unconstitutionality. The doctrine does not save; it destroys.

Even in circumstances in which severability doctrine does not expand invalidity beyond unconstitutionality, its presence within an excision-based understanding of judicial review blurs the boundary between the judicial and legislative powers. Within an excision-based approach, the judicial operation of severance is understood to create a new law that consists of the old law “minus” its unconstitutional provisions or applications. Scholars steeped in this understanding comfortably refer to severability doctrine as resulting in judicial rewriting. As a description of what courts do to partially unconstitutional legislation, the imagery of severance and excision thereby underwrites a conception of the judicial role that is avowedly legislative, cementing in place a pernicious understanding of the judicial role and effacing an understanding of how courts can deal with partial unconstitutionality in a manner properly judicial.

For too long, judicial rewriting through severability doctrine has been accepted as inevitable. This perceived inevitability has kept the debate centered on how courts should rewrite statutes, bypassing entirely the antecedent question of whether they should do so. Now is the time to restore review without rewriting.

Acknowledgments:

Copyright © 2010 NYU Law Review.

Kevin C. Walsh is a professor at the University of Richmond School of Law.

What to do with these lost neighborhoods?  It is the late dawn of the twenty-first century, when integration is stronger and civil rights laws are weaker, when local government budgets are dwarfed by demands.  Suing local governments or lobbying them, two of the most important strategies of twentieth-century advocacy for social justice, have been undermined by legal developments that drained city revenues and withdrew federal remedial power to address racial segregation. 

Yet state and local government laws retain malleability.  Laws governing the allocation of power among local agencies exert significant influence over unincorporated urban areas specifically and spatial polarization by race and class more generally.  This Editorial argues for a new priority in metropolitan law and policy: state legislative reforms to empower and reshape county governments to represent regional interests and regional logic in intergovernmental negotiations, including negotiations over annexation.  Strengthening counties to bargain with other local agencies over matters with redistributive consequences, like annexation, can bring a metropolitan perspective to critical local decisionmaking and create a promising corridor for addressing contemporary issues of urban inequality. 

Part I introduces the unincorporated urban areas problem, describing these neighborhoods and the pattern of selective annexation (known as municipal underbounding) that underlies their formation.  Part II explores the two traditional tools of change and their current constraints: first, organizing efforts to lobby for inclusion and redistribution within existing local politics, which are hampered by cities’ finance-driven rules for growth management and annexation; and second, antidiscrimination litigation to remedy a pattern of discriminatory annexation, which has been confined by courts’ reluctance to mandate the movement of a local border.  Part III harnesses the flexibility created by a strong political and judicial doctrine of local autonomy by proposing reforms that empower county governments to exert greater influence over (without dictating) annexations. 

Part problem-solving for the unincorporated urban areas problem in particular, and part exploration of solutions to metropolitan inequality in general, this Editorial offers a new direction for state and local government law to seek progress on economic and racial polarization in America’s cities.  It brings counties—our most neglected, undertheorized layer of state government—into sharper relief and conceives for them a regional, redistributive purpose. 

 I. Unincorporated Urban Areas, Municipal Underbounding, and the Annexation Solution

Discriminatory annexation is known to urban geographers as municipal underbounding: annexation policies and practices in which municipalities grow around or away from low-income minority communities, thus excluding them from voting rights in city elections and municipal services.   Excluded from city boundaries (and too poor, if not too small, to form their own city), underbounded neighborhoods remain unincorporated and dependent on counties as their most proximate layer of general-purpose local government.  In past work, I have called these neighborhoods unincorporated urban areas, and I’ve found that they are characterized by one or both of two main challenges: a lack of basic infrastructure and services like wastewater treatment and street lights and/or an overconcentration of undesirable land uses like freeways and municipal utility plants.  Each such neighborhood identified to date is predominantly Latino and African-American, often with a history of settlement under de jure and de facto segregation.  

Where are these areas, and how many towns and cities exhibit this problem?  We have no consolidated national research on these questions, but we can triangulate a general picture of the problem as prevalent in the South and the Southwest.  Qualitative research has investigated dozens of high-poverty unincorporated urban areas in Texas, California, North Carolina, Florida, and Mississippi.  A study of census data from 1990 and 2000 found that African-American communities adjacent to non-metropolitan towns in the South are more likely to be bypassed for annexation than similar white communities.  Studies of poverty in unincorporated areas more generally (not necessarily unincorporated urban areas near a city border), have identified thousands of low-income unincorporated subdivisions in the Southwest—including at least 1,800 in Texas alone.   

Heretofore, municipal underbounding and the unincorporated urban areas it creates have been understudied, if not invisible.  In the annexation context, municipal underbounding has been hidden in the shadows of academic and policy assumptions that increasing city power and autonomy over annexations will increase the number and extent of annexations, and that such change will be against the interests of unincorporated areas.  In the spatial inequality context, the problem of unincorporated poverty has been overshadowed by other forms of interjurisdictional segregation and fragmentation, particularly the economic and racial polarization of core cities and their independently incorporated suburbs.  A wave of awareness and advocacy on unincorporated urban areas, however, is beginning to crest. 

Meeting the needs of unincorporated urban areas can take at least four directions: relocate households individually, improve county government, form an independent city, and seek annexation to an existing city.  This Editorial focuses on annexation.  This avenue is not offered as a one-size-fits-all solution—it has risks and downsides for some communities.  Yet it is a realistic and desirable prescription in many cases for three reasons.  First, annexation leads to service improvements and extensions by bringing unincorporated urban areas into a city that already provides such urban services and triggering existing legal requirements that cities bring underserved areas up to municipal health and safety standards.  Second, by providing city voting rights (giving residents of unincorporated urban areas the same two levels of local government that city dwellers usually enjoy), annexation also increases the proximity and, potentially, the responsiveness of political representation.  Finally, annexation leaves historically rooted communities intact. It moves borders—not homes or people—in order to “relocate” a neighborhood from one jurisdiction to another.   

II. Traditional Tools: Community Organizing and Civil Rights Litigation

In the tradition of their times, residents of unincorporated urban areas and their advocates have deployed conventional tools of change: they have organized locally and they have sued.  Activism in (and patience with) the local political and market economy is a commonly recommended antidote to problems of spatial inequality in a post-civil rights era.  Where local politics are contaminated by racial discrimination, constitutional equal protection and statutory voting rights protections are assumed to reach the problem.  Yet changes in state and federal law have enfeebled both tactics. 

Applied to the problem of unincorporated urban areas, a community organizing strategy predicts that lobbying city and county governments can lead to service improvements and annexations.  Yet the problems of unincorporated urban areas, including municipal underbounding, are ill-suited to this approach, if one adheres to the premise, as the Article does, that avoiding the involuntary displacement of existing residents is a central priority.  Under the current system of local finance, cities enjoy the legal right, if not the fiduciary duty, to engage in class discrimination when making annexation choices.  Borders are, in essence, saleable goods in the contemporary political economy.  Local governments now effectively set a price for (1) entry (a landowner or group of landowners seeking to relocate a local border to encompass their parcels through annexation), and (2) residence (services provided within a jurisdiction).  State laws widely mandate fiscal impact assessments of boundary changes, and they currently provide few incentives or benefits to offset fiscal losses from the annexation of a poor neighborhood.  As a practical matter, capital investment and increased tax revenue in unincorporated urban areas are currently prerequisites to annexation within local political economies. 

When politics fails, municipal underbounding seems ripe for a litigation solution reflective of the problem’s history of racial segregation and racially ordered provision of municipal services and voting rights.  Yet addressing any problem of spatial inequality—be it racial segregation, disparities in neighborhood services, or discriminatory annexation, to name a few—through a civil rights lawsuit faces formidable, well-known doctrinal barriers.  And in the context of municipal underbounding, an unincorporated urban area also faces an additional barrier: Federal courts are reluctant, or perhaps even unwilling, to move a local border. Only a narrow band of factual scenarios can be redressed with existing antidiscrimination protections, and even in those cases, local autonomy to establish and move local borders has come to serve not only as a license to behave in any way consistent with state law, but also as a quasi-affirmative defense to claims that racial discrimination was a motivating force behind service or annexation decisions.  My full Article discusses this particular barrier to using civil rights laws as a mode of redress for municipal underbounding, re-reading voting rights and school desegregation cases to investigate federal authority to mandate a shift in local borders. 

III. Overcoming County Powerlessness in Annexation Law

This new landscape of hardened American municipal borders—subject to a marketplace for entry and residence and insulated from federal antidiscrimination law—impedes solutions to the unincorporated urban areas issue that rely on the tools of local activism or civil rights litigation.  Yet the local autonomy fostered over the twentieth century suggests an alternative strategy: to preserve local control as against state or federal power, but to reshuffle the metropolitan agenda by giving county governments a stronger and more regional role to play. 

Academic and public debates about American annexation law have pivoted around a single story with two actors (the revenue-hungry city and the property-rich suburb, whether existing or planned) and one scenario (an aggressor city seeks to annex the suburb against its landowners’ wishes).  Municipal underbounding requires a place for two new actors (the service-poor county and the low-income suburb) and one additional scenario (a city refuses to annex a supplicant suburb).  Accommodating this more complex picture of annexations requires that we give county governments some influence over annexation decisions and ease the procedures by which outsiders to a city can seek annexation. 

Changes in pursuit of these objectives are important, but they must be subject to a key counterbalance: the positive attributes of city control and veto power over some annexations.  City power and discretion over matters of annexation enables cities to capture population growth and new revenues at their fringes, which is a critical element of the elasticity widely believed essential for urban fiscal health.   Sound urban growth management requires that cities can check counties’ incentive to permit (or simply fail to control) new land uses outside of city lines that facilitate uncoordinated and inadequately served urban sprawl that can later be handed over to a city’s balance sheet.  For these reasons, advocates for giving cities stronger annexation powers—including the “involuntary annexation” power to annex areas against residents’ will—ground their views in the sensible and worthwhile goal that city boundaries encompass city growth.   They argue that annexation law should prevent wealthy unincorporated suburbs from freeriding on city taxpayers by enjoying the advantages of adjacency to city life without paying city taxes for city services. 

City boundaries should indeed encompass urban land, but that principle should apply across the board—whether the urban land at issue is rich or poor.  To find that consistency, balancing freerider effects against inclusionary goals, cities cannot be the sole governmental negotiating interest in annexations.  Thus, rather than substituting strong counties for strong cities (simply empowering a new self-interested local government in place of an old one) the proposals advanced in my Article do not advocate replacing city power over annexations, but rather to introduce counties as a stronger negotiating partner to guide annexation outcomes.  Interlocal negotiations can balance cities’ legitimate need for urban growth control (and anti-sprawl objectives in general) against state and regional needs for adequate and efficient urban services in all urbanized areas. 

Indeed, certain existing features of county government make it an attractive negotiating partner over annexation, even from the perspective of city interests.  The same rule of county government that makes it a diffuse and distant local government (thus increasing the attraction of annexation or incorporation solutions for urban areas) makes it a ready-made regional government tied to city interests: Residents of incorporated areas enjoy the same voting rights in county government as are held by unincorporated area residents.  Political accountability to city interests is thus built in to county governance.  When it comes to interlocal negotiations of the kind described here, stronger counties do not necessarily mean weaker cities. 

Readers interested in the specific portfolio of reforms proposed are referred to my Article, but suffice it to summarize here that I propose two categories of state legal reforms to bring counties into annexation decisions: empowering counties and their residents to initiate, facilitate, and consent to annexations; and protecting counties, where appropriate, from annexations that will cherrypick unincorporated land to leave behind residual territory that is underserved, inefficient to govern, or too costly to serve at habitable standards. 

If state law facilitates annexations, the question remains: who will pay for the necessary infrastructure upgrades to bring unincorporated urban areas within city service networks and up to city standards?   Infrastructure financing is messy and scarce.  My Article looks to historical models of infrastructure funding to argue for a modernized version of nineteenth-century extraterritorial service provision (an adaptation that is peculiarly well suited to recessions with thin local budgets): conscript cities to compete for federal capital funds (including economic stimulus dollars in our current economy), that can retrofit affordable housing at the urban fringe with urban services and infrastructure.  Just as turn-of-the-twentieth-century cities in the United States served as brokers and leaders of fringe infrastructure, contemporary cities in many states are in the best position to plan and compete for, not to mention spend, federal dollars on urban infrastructure.  Cities’ higher levels of professional staffing, greater experience with competitive intergovernmental grant programs, smaller constituencies, and control over urban service providers make them better equipped to compete in a world of costly, highly engineered proposals.  The duty to fight for funds on behalf of unincorporated urban areas is particularly justified in states that give cities extraterritorial land use authority, because such power should come with corresponding obligations to pursue habitability improvements aggressively in urban fringe areas. 

IV. Conclusion

The problem of high-poverty, urbanized areas mapped out of city borders by annexation decisions is among America’s most severe problems of spatial inequality.  The lobby and the lawsuit, two dominant strategies for achieving social justice in the twentieth century, have proven to be blunt instruments in the fight to improve services and political rights in these areas.  Yet the very local autonomy fostered by state legislatures and federal courts suggests an alternative strategy: to preserve local control as against state or federal power, but to reshuffle the metropolitan agenda by giving county governments a stronger and more regional role to play.  Applied to the problem of municipal underbounding, county governments need a stronger influence over in annexation decisions.

In this reconception of county power over annexations lies the seed of future potential, and indeed this author’s own future research: Might counties offer untapped potential over other matters of interjurisdictional segregation and spatial inequality?  In particular, can we adjust county power (through a process of both adding and subtracting measures of their authority) to alleviate economic and racial polarization among their constituent cities?  The characteristic that makes counties poor substitutes for a first-tier city government—their equal obligations to the residents of all cities and unincorporated places in their territory—suggests an untapped potential to coordinate and rationalize metropolitan regions.  Many issues remain for future inquiry, not the least of which will be counties’ potential for political capture by their largest cities, their thin record on the pursuit of redistributive justice, and their poor land use planning.  For now, suffice it to observe that over matters of annexation, counties’ weak hand has enabled the unchecked practice of municipal underbounding and stranded unincorporated islands and fringes in material stasis and decay.  Counties have no vested interest in the status quo (quite the opposite is true), and there is every reason to expect them to seek annexations for their high poverty pockets.  On this issue at least, reform of county power provides a footpath forward.       

Like fossils, unincorporated urban areas reveal a history of twentieth-century urban change in the control of local boundaries.  Like laboratories, they offer a setting in which to test twenty-first-century strategies for resolving questions of social injustice.  By investigating a contemporary problem of pre-civil rights vintage, we find old wounds and new possibilities, both hidden under history’s crust.

Acknowledgments:

Copyright © 2010 Stanford Law Review.

Michelle Wilde Anderson is an Assistant Professor of Law at the University of California, Berkeley School of Law. 

This Legal Workshop Editorial is based on the following Law Review Article: Michelle Wilde Anderson, Mapped Out of Local Democracy, 62 STAN. L. REV. 931 (2010).

The Existing Debate

Because courts in making factual determinations rely on witness accounts rather than on the direct investigation common in so many other fields, witness credibility has long been a central concern of the law.  These concerns are often about the misperception, misrecollection, hedging, fudging, bending, and slanting rampant in a system dominated by testimony from self-interested parties , but the law also worries about flat-out lying.   Because we have lost much of our faith in the oath to assure veracity, and because cross-examination is more effective in exposing liars on television than in real courtrooms, the legal system is constantly seeking better ways of determining who is lying and who is not.  Legal debates about lie-detection technology have existed since the 1920s, and, indeed, the long-enduring Frye test¹ for determining the admissibility of expert testimony arose in the context of a rudimentary lie-detection device.  Although lie-detection technology has improved substantially over what it was when the D.C. Circuitdecided Frye in 1923, with few exceptions the law still prohibits the use of polygraphs, electroencephalography, periorbital spectrography, analysis of facial micro-expressions, and various other technologies, relying instead on the traditionally alleged ability of the technologically unaided judge and jury to determine when witnesses are lying and when they are telling the truth.

The terrain has changed considerably as a result of relatively recent claims that the techniques of modern neuroscience—especially fMRI—can identify deception more accurately than previous techniques.  Those who have made these claims support them with reference to a considerable quantum of experimental studies and published research.  But numerous neuroscientists have challenged these claims, arguing that the differences between experimental subjects and those offering actual evidence in court are so great as to create fatal problems of external validity.  Even more troubling, it is said, is that major problems of construct validity arise because an instructed lie in the laboratory is simply not a real lie at all.  In addition, much of the research has not been published in peer-reviewed journals, has been produced by financially interested scientists connected with commercial “truth verification” companies such as CEPHOS or No Lie MRI, and obscures significantly lower rates of reliability than the proponents of the technology have alleged.

For such reasons, the overwhelming scientific opinion is that fMRI lie-detection is a “research topic and not a legal tool,”² a conclusion emphasized by worries about how the research might be used.  It would be wrong, some neuroscientists have argued, to use brain scans to “send [someone] to prison,” ⁴  And thus one widely discussed article has urged a moratorium on the law’s use of neuroscience-based lie-detection until sponsors of the technology can establish its reliability to the satisfaction of a federal regulatory agency.⁵

Critiquing the Critics

At one level, these criticisms appear sound.  External and construct validity really are problematic, the claims of reliability are genuinely exaggerated, the lack of peer review is a fact, and some of the most prominent researchers are indeed connected with interested commercial entities.  But at another level the criticisms are based on three flawed premises: the first is that most legal decisions are about sending people to prison; the second is that the standards of science should be the standards of law; and the third is that the legal system’s current approach to determining veracity is acceptable.  I will discuss each of these in turn.

Although the legal system does send people to prison, it can do so only when prosecutors have proved the case against the defendant beyond a reasonable doubt.  So even if we were to imagine that prosecutors, in the face of existing Fifth Amendment self-incrimination law, could compel an involuntary brain scan of a defendant, there is no question that the reliability of fMRI-based lie-detection is nowhere near high enough to support a conviction under the beyond-a-reasonable-doubt standard.  But the converse of the prosecution’s heavy burden is the defendant’s ability to defeat conviction by showing only a reasonable possibility—not even a probability—of innocence.  Even if neuroimaging lie-detection reliability were only, say, .60, a level grossly insufficient to support conviction, that level, if used to support a defendant’s alibi or to undercut an arresting officer or eyewitness’s testimony, could easily raise a reasonable doubt as to guilt.

That slight, but somewhat plausible, confidence levels in the evidence indicating innocence are sufficient to acquit, even if nowhere near sufficient to convict, is just one example of legal decisions made on the basis of confidence levels far below those necessary for scientific publication or other scientific uses within the scientific community.  Judgments in civil cases, for example, demand only a “preponderance of the evidence,” and other decisions require “reasonable suspicion,” “probable cause,” or occasionally merely a “scintilla of evidence.”  Most importantly, the confidence necessary to admit an individual item of evidence is much lower than that necessary to justify a verdict—under Rule 401 of the Federal Rules of Evidence, for example, admission into evidence requires only that a proposition be more likely with the evidence than without, a standard dramatically lower than proof beyond a reasonable doubt and even than a preponderance of the evidence.  To hold that evidence must in some way be very highly reliable just to be admitted is to urge a conclusion dramatically at odds with fundamental premises of the trial process and the law of evidence.

The foregoing is an explanation of why the levels of reliability commonly (and properly) demanded by scientists for scientific purposes are not and should not be the levels for determining the admissibility of single items of evidence, but the same considerations apply to degrees of external validity.  Often-criticized psychological experiments on college undergraduates, for example, can justify conclusions about the behavior of people in non-laboratory settings when and because we know from other research that the behavior of subjects correlates with that of non-subjects.  Similarly, although deception in the laboratory differs from deception in the non-laboratory real world, the existence of any positive correlation between laboratory and non-laboratory deception would mean that the laboratory results provide some evidence justifying non-laboratory conclusions, and determining whether that “some” evidence was sufficient requires the use of legal and not scientific standards.

So too for questions of construct validity.   With one prominent exception, ⁶ the research on neuroscience-based lie-detection has been conducted largely with instructed lies, which are not lies at all, but simply examples of subjects following the instructions of the researchers.  But if the ease of telling an instructed lie in the laboratory correlates with the ease of telling a real lie outside the laboratory, then the research on instructed lies is no longer irrelevant to detecting real lies.  With any positive correlation between instructed and real lies, experiments on the former will tell us something about the latter, and whether that “something” is enough again depends on the uses for which the research is employed.  That which is inadequate for scientific publication or criminal prosecution may be sufficient, for example, for a defendant seeking only to raise the possibility of a reasonable doubt.

In law as well as in science, “compared to what?” is often the important question. Accordingly, evaluating fMRI lie-detection requires knowing what it would supplement or replace.  As it turns out, the methods that are now used to determine witness veracity are substantially worse than the lie-detection science that remains routinely excluded.  Currently, the jury or judge (when there is no jury and the judge is serving as trier-of-fact) is charged with determining if witnesses are telling the truth.  And, thus, when cross-examination provides little assistance, as it rarely does, courts instruct juries to evaluate the “demeanor” of a witness to determine veracity.  In doing this, however, jurors (and, we suspect, judges as well) rely on numerous myths, urban legends, and pop psychology with little reliability.  They distrust witnesses who perspire, fidget, and fail to make eye contact, and trust those who speak confidently while looking directly at them.  Research has shown that ordinary people’s ability to distinguish truth from lies rarely rises much above random, and juries and judges are unlikely to do better.  The question is thus not whether neuroscience-based lie-detection is good in the abstract, but instead whether it should be rejected in favor of methods—methods that go to the heart of the jury system, have been in place for centuries, and are extremely unlikely to be changed substantially—that are demonstrably unreliable.  The answer to this question will depend on a comparison for which there is yet little evidence, but the point is only that the admissibility of neural lie-detection evidence must be based on an evaluation of the realistic alternatives within the legal system and not on a non-comparative assessment of whether neural lie-detection meets the standards that scientists use for scientific purposes.

But might jurors and judges interpret a brain image as having more evidentiary value than it actually possesses?  If a brain scan supporting a defendant’s alibi, for example, had a reliability rate of .72, might a juror, seeing a “picture” of a brain in vivid color, ignore the .28 chance of error and assume absolute accuracy?  Might a juror take a scan supporting only one aspect of a defendant’s story as “proving” that the defendant was innocent?  These risks are real, but the little research that exists on over-valuation of brain scans tends to fail to distinguish seeing a brain image from seeing an equally realistic color photograph not of a brain, or even any pseudoscientific explanation unrelated to brain imaging.  One part of one study, for example, compared perception of a brain image with straight text,  with a two-color bar graph, and with a complex color diagram, but such comparisons alone would be insufficient to conclude that the brain image produced more unjustified reliance than the color photographs that are a routine component of the typical trial.  On the existing research, we cannot say that the degree of unjustified reliance, therefore, is a function of a brain scan appearing photographic or instead a function of it appearing to be a photograph of a brain.  Studies that do not control for the potential confounds of the non-brain attributes of a brain image thus cannot tell us whether it is a brain image that produces unwarranted attribution of content, or whether the effect comes simply from the kinds of multi-color or photographic images routinely used for forensic purposes.  Although overvaluation is a legitimate worry, the degree of legitimacy is a function of how much, if at all, a brain image produces perceptual distortions beyond those that are already endemic to litigation.

The Standards for Legal Use of Science Cannot be Derived from Science Itself

Once we comprehend the range of standards the legal system now uses, the scalar rather than binary character of reliability and validity, and the legal system’s venerable reliance on techniques for identifying deception that are worse than even the most modest possibilities for neural lie-detection, the case against fMRI becomes substantially less compelling.  Still, the use of neural lie-detection now is probably unwarranted.  But, whether it is, and if not now then when, is a determination that the legal system cannot make solely on the basis of scientific standards.  The goals of the legal system differ from those of science, and law and science make different kinds of decisions for different purposes.  Consequently, what is good enough for science may sometimes not be good enough for law.  Conversely, and more directly relevant here, what is not good enough for science may sometimes be good enough for law.  Science must inform the legal system about reliability rates and degrees of validity, but whether some rate or degree is good enough for some legal purpose is, in the final analysis, a question of law and not of science.

Is Daubert the Problem?

Some of the foregoing conclusions may well be in some tension with the Supreme Court’s conclusions in Daubert v. Merrill-Dow,⁷ Kumho Tire v. Carmichael,⁸ and General Electric v. Joiner,⁹ To the extent that that is so, however, then perhaps certain aspects of the Daubert revolution need to be rethought.  Daubert’s concern about junk science is legitimate, but more vigorous application of summary judgment and related forms of pre-trial dismissal may better address that concern than does distorting longstanding evidentiary principles.  And, even if Daubert is correct in using the law of evidence to deal with the junk science problem, Daubert’s often-criticized acceptance of scientific criteria for determining validity and reliability may have started us on a path that is in need of correction.  If my conclusions about neuroscience-based lie-detection are in some tension with Daubert, and it is not entirely clear that they are, then it may be that it is Daubert and not those conclusions that is in need of adjustment.

Acknowledgments:

Copyright © 2010 Cornell Law Review.

Frederick Schauer is David and Mary Harrison Distinguished Professor of Law at University of Virginia Law School.

The Article begins by explaining how the Fourth Amendment applies to physical space. It makes two main points. First, in physical space the basic dividing line of Fourth Amendment protection is between inside spaces and outside spaces. Outside spaces are not regulated by the Fourth Amendment: the government can go anywhere outside without triggering Fourth Amendment protections. However, the Fourth Amendment does ordinarily protect inside spaces. The government ordinarily needs a warrant to enter an enclosed space unless there are special circumstances such as consent. Second, once the Fourth Amendment protects a space, its rules are based on a physical conception of how broad a search can be. In particular, the Fourth Amendment allows a search warrant for physical spaces using the basic size constraints of the physical world. For the most part, a search warrant uses the basic unit of a home to limit the scope of warrants.

Applying the Fourth Amendment to the Internet is challenging and interesting because these two principles no longer work. The Internet does not rely on a distinction between inside and outside, and there are no clear notions of space that would limit the scope of space over which a search can occur. The challenge of applying the Fourth Amendment to the Internet is finding new principles that serve the same function as the old principles in the new network space. New rules are needed. What should they be?

First, the distinction between inside and outside in the physical world should be replaced in the online setting by the distinction between content and non-content information. In the online setting, courts should treat non-content information relating to communications as if it were functionally “outside” and content information as if it were functionally “inside.” Internet surveillance of non-content information should not trigger the Fourth Amendment just like surveillance of public spaces does not trigger the Fourth Amendment, and surveillance of content should presumptively trigger the Fourth Amendment in the Internet setting just like surveillance of inside spaces presumptively triggers the Fourth Amendment in the physical world.

The core reason to use the content/non-content distinction is because it captures the basic function of the inside/outside distinction. Outside surveillance is usually surveillance relating to identity, location, and time. By watching a person in public, the police normally can learn where he was at a particular time and where he was going. In contrast, inside surveillance is much more likely to be surveillance relating to private thoughts. By breaking into a person’s private spaces, the police can obtain insights into the contents of the person’s mind that he normally keeps to himself or only shares with a trusted few. That distinction correlates reasonably accurately to the online distinction between content and non-content surveillance. Online, non-content surveillance is usually surveillance related to identity, location, and time; content surveillance is surveillance of private thoughts and speech.

The basic function of communications networks explains the correlation of the two principles. Communications networks are tools for delivering content that would otherwise have to be delivered in person. The network carries the communication to its destination so that the user does not need to travel there. The distinction between content and non-content information over the network reflects the distinction between the private inside places where communications are typically sent and received and the public outside places over which the communications must travel. The non-content information is the information generated by the network needed to deliver the communication: that information substitutes for what would have been observable in public if the user had delivered the communication in person. In contrast, the content information is the private message that the user seeks to deliver. It substitutes for the contents of the package or letter that would be delivered.

Importantly, this basic insight is only a first step towards applying the Fourth Amendment to the Internet. The line between content and non-content information can be difficult in some cases involving person-to-computer applications. Further, a presumption that contents of communications are protected by the Fourth Amendment is just a presumption: there will likely be many exceptions to this rule, just as there are exceptions to the presumption that inside surveillance is protected. However, it is beyond the scope of this Article to identify in exactly which circumstances content surveillance should be allowed. While this is only a first step, it is an important one. Recent court decisions have pointed somewhat hesitatingly in this direction, and this section shows why this direction is correct and why courts should follow this distinction in the future.

With the basic distinction between content and non-content communications in place, the next issue is how much protection the Fourth Amendment should extend to the content of communications. First, if the Fourth Amendment protects the content of communications, exactly what kind of protection does it offer? Does the Fourth Amendment require a warrant, or do Internet applications justify different treatment, like a probable cause requirement without a warrant? And second, if a warrant is required, how much authority should a warrant provide? Should a warrant allow a narrow search or a broad one? Should a warrant for Internet communications be limited to a single account hosted by a single provider, or should it allow for searches through multiple accounts, and if so, how many?

This part of the Article contends that the Fourth Amendment ordinarily requires a warrant for the collection of the content of Internet communications. The Internet should not trigger a lower standard such as are found with the automobile exception. Content stored in and transferred through Internet accounts should be protected with the same default warrant requirement that are required for access to homes, telephone calls, and postal letters. Further, the Fourth Amendment should require a narrow exception permitting for the warrantless copying of data pending a warrant. Such a power would mirror similar authorities to temporarily detain packages pending a warrant to open a package, and would allow for a warrant requirement for access to content of communications. Under my proposed framework, the federal privacy statute that permits the government to compel access to stored files without a warrant is unconstitutional in many of its applications.

Finally, the particularity requirement should allow searches of multiple accounts with multiple providers used by the same criminal suspect instead of requiring separate warrants with separate probable cause for every individual account. The multiplicity of services in the Internet setting should lead the particularity requirement to allow one warrant for multiple accounts, much like the statutory roving wiretap authority allows one warrant for multiple telephones in the traditional telephone setting. Put another way, the particularity requirement of Fourth Amendment law should be applied so the basic building block of particularity in the online environment is a specific Internet user, not a specific account or physical device.

Acknowledgments:

Copyright © 2010 Stanford Law Review.

Orin S. Kerr is a Professor of Law at The George Washington University Law School.

This Legal Workshop Editorial is based on the following Law Review Article: Orin S. Kerr, Applying the Fourth Amendment to the Internet: A General Approach, 62 STAN. L. REV. 1005 (2010).

Much of the literature related to large settlements seeks to improve how judges and private actors serve the large groups of people impacted by a massive settlement.¹  Settlement funds have reformed, in turn, by giving people more choices, such as more filing opportunities, different settlement outcomes, and extended deadlines. More opportunities to opt out of a large settlement theoretically assure that the fund’s administrators represent the interests of those who do not opt out. More choices of settlement awards means that more claimants can elect awards that fit their individual needs and circumstances. And more time to decide helps claimants to arrive at decisions that reflect their divergent interests. Few commentators have considered, however, how claimants to a large settlement fund make those choices.² Modern reform efforts, rather, assume that claimants make rational decisions about their options based on their own stable values and preferences.³

But is that correct? Studies have long shown that because of cognitive bias, people may buy things they do not want, save too little for retirement, or make risky choices about their health—based on their point of reference, the timing of the decision, and the presence of seemingly irrelevant choices.⁴ Behavioral economists have examined these ostensibly irrational decisions in many other legal contexts, but few commentators have explored these effects in the context of a large group settlement. Because claimants to large settlements are generally unassisted laypersons, large settlement funds may be particularly compelling settings to examine the adverse impact of cognitive bias.

To that end, I make three claims in this Essay. First, people may make irrational decisions about their settlement options in a large settlement fund because of cognitive bias. Second, cognitive bias may undermine some of the stated purposes of public and private settlement funds—to provide claimants with more access, efficiency, and fairness than in traditional litigation. Third, “fund designers”—judges, lawmakers, and agencies—should identify and, in some cases, capitalize on claimants’ cognitive bias by altering the context, timing, and sequence of settlement options. Fund designers, however, should avoid reforms that unduly eliminate settlement options or impose excessive administrative costs. Rather, the benefits of any reform—preventing avoidable harm to irrational claimants—must outweigh the potential costs, including the value of client autonomy, the chance of error, and the burden on the courts and public administrators.

I examine these three claims by describing how three cognitive biases are likely to affect claimants in large settlement funds. These biases are: (1) status quo bias, (2) contrast bias, and (3) time-inconsistency bias.

I.

Status Quo Bias

Status quo bias refers to a person’s tendency to stick to the status quo even when other options would increase well-being. In principle, a completely rational person will choose between alternatives based on his or her preferences and the potential costs of making an informed decision. In practice, however, simply characterizing an option as the status quo significantly increases the chances that a person will choose that option. Some speculate that this preference for the status quo derives from a general aversion to risks caused by one’s own actions, even when there are greater risks associated with inaction.⁵

For example, alternative investment and saving options are significantly more popular among college professors when designated as the status quo or the default choice.⁶ Because of the status quo effect, some commentators like Cass Sunstein and Richard Thaler have advocated “libertarian paternalistic” ways to encourage saving.⁷ They advocate changing the default rules to promote particular outcomes—like an employee’s decision to enroll in a 401(k) retirement plan—without limiting the employee’s opportunity to opt out of the plan at a later time.

Of course, switching costs also might explain adherence to default rules. A decisionmaker may rationally determine that it is not worth the time, money, or potential opportunity cost to deviate from the status quo. Moreover, people may be rationally indifferent to certain choices. Such explanations, however, do not fully account for how people make decisions. Although the subject of some criticism, many studies show that people irrationally overvalue either the default option or the costs associated with departing from the default option.⁸

In class action settlements and public settlement funds, status quo bias may be unavoidable. After all, there must be a default rule that asks people either to affirmatively join or affirmatively withdraw from a large settlement fund. But status quo effects complicate the long-held belief that opt-out rights (1) ensure fairer settlements and (2) provide an adequate opportunity to claim or reject awards through the fund. When few people affirmatively opt out or object to a settlement, courts and administrators have assumed that the fund successfully represents what claimants rationally want and therefore ensures a “fair, reasonable, and adequate” settlement.⁹ Status quo bias, however, provides a reason to be skeptical of these assumptions and the policies based on them, even when very large payouts are involved. Many people will join a large fund not because the overall settlement reflects their values and interests but simply because the default rule requires parties to affirmatively opt out of the fund.

The status quo bias also contributes to the phenomenon of underclaiming, in which parties refuse to opt out of a settlement but never claim an award. Many public and private settlements require parties to complete a new form to claim an award, to choose among substantive settlement options, or to select a settlement process. Commentators studying claim rates in class action settlements have found that the fraction of funds actually disbursed was very modest in these so-called claims-made settlements.¹⁰ This includes cases in which claimants were otherwise entitled to substantial awards.

Accordingly, settlement funds could automatically process claims, not unlike automatic 401(k) plan enrollment. Under such a system, a settlement fund would automatically distribute presumed awards to claimants who join the fund. Such a policy, however, would come at a cost. Among other things, funds would bear the administrative cost of precisely identifying eligible claimants in advance of payment.

In light of potential costs, automatic processing would be more justified in certain funds. In large-value cases, for example, automatically processing claims would not be worth the administrative cost, the burden on the courts, and the potential for error or fraud. Thus, automatic processing may be warranted in welfare benefit settlements or shareholder class action funds, in which fund designers typically have a great deal of information about claimants, the awards are modest, and claimants generally do not choose among multiple settlement options.¹¹ Such policies would be more problematic in large mass tort settlements, in which settlement trusts or public settlement funds have less information about potential claimants, the awards are large, and claimants may be offered various procedural and substantive options in the settlement.

II.

Contrast Bias

Contrast bias is the irrational tendency to weigh an option more or less favorably depending on the presence of other options. Theoretically, a rational decisionmaker should not rank options differently simply because the options are described in a particular way. Moreover, the introduction of an additional choice should not alter a decisionmaker’s relative valuation of the original options. But this is not always the case.

Take, for example, the uncanny effects of decoy options—options that no one ever chooses but that make another alternative more appealing—on physical attraction. In a survey of six hundred students, a behavioral economist asked subjects to rate the looks of two men.¹² When asked to choose between the photographs of two equally attractive candidates—call one “George Clooney” and the other “Brad Pitt”—subjects were equally divided. When another group of subjects was asked to choose between the two initial candidates and a third candidate, a photoshopped and deformed version of George Clooney, however, 75 percent chose the unspoiled version of George Clooney and 25 percent chose Brad Pitt. Although no one selected the third option, the seemingly irrelevant introduction of an ugly version of George Clooney led 50 percent more students to believe that the original George Clooney was better looking than Brad Pitt.

Psychologists and behavioral economists have found that contrast effects directly impact a wide array of decisions, including consumer purchases, employment decisions, elective medical procedures, and even presidential elections.¹³ There are many explanations for contrast bias. Some suggest that it is simply easier to compare similar options among a set of choices than to give an absolute or innate value to any particular option. ¹⁴

The presence of contrast bias may be relevant to laws designed to improve the oversight of large settlements. Many settlement funds ask claimants to choose from an array of options after joining a settlement, in part to maximize the benefit to claimants with different interests in settlement. But the interrelationship of various settlement options may unwittingly impact a choice between cash and nonpecuniary awards, like coupons and warranties. In my own classes, I distribute an altered Apple iPod Settlement Notice as an illustration. After litigation over reported battery problems in old Apple iPods models, Apple settled and offered customers a choice of a $50 store credit or $25 in cash. Half of my students receive the original version of the iPod settlement notice. The other half receives a modified notice that contains the same two options—a $50 store credit and $25 in cash—and a decoy option, a $35 store credit. The results show the addition of a seemingly irrelevant coupon dramatically affects the students’ willingness to take the coupon settlement. The original group chooses cash more than 61 percent of the time; the decoy group chooses the cash only 40 percent of the time.

Contrast bias has implications for laws like the Class Action Fairness Act (“CAFA”), which expressly requires courts to conduct “fairness hearings” in coupon-only settlements and to postpone decisions about the amount of attorneys’ fees until after the coupons have been redeemed.¹⁵ CAFA’s purpose is to ensure that the attorneys’ fees are closely connected to the actual value of the settlement to the class. But CAFA does not impose a similar requirement for settlements involving both coupons and other options. Rather, courts may award attorneys’ fees upfront, based on an estimate of the cash value of the settlement apart from the portion of the settlement involving coupons. Due to contrast bias, courts may also have reason to wait for claimants to redeem these kinds of settlement awards: the coupon may encourage claimants to accept another settlement option that, by comparison, seems to offer a better value or greater liquidity. Such delay imposes costs. Class action litigation is risky business, and delaying even a portion of attorneys’ fees may dampen some attorneys’ willingness to file in the first place. But delay may be justified if it ensures that the attorneys’ fees better reflect the actual value that class members derive from the settlement.

III.

Time-Inconsistency Bias

Rational models of choice assume that people have time-consistent preferences. That is, a person’s relative preference for gratification will be the same no matter when he or she is asked. Substantial evidence, however, demonstrates that people have time-inconsistent or present-biased preferences. Ask whether a person prefers to rent Schindler’s List or So I Married an Axe Murderer, and the answer should not depend upon whether the decisionmaker plans to watch the movie today or later next week. But the proportion of people who elect to watch Schlindler’s List in the near future may be thirteen times higher than those willing to watch it on the same day they are asked.¹⁶ Present-biased preferences explain the systematic tendency to seek out more immediately gratifying benefits today than the long term benefits called for by earlier plans.¹⁷ More often than not, people choose a bird in the hand—be it dessert, a little extra cash, or a silly movie¹⁸—over three or four in the bush.

Time-inconsistency is compounded by nonintegrated decisionmaking. Nonintegrated decisions are rational decisions about costs and benefits in irrationally short periods of time. If a person had to choose whether to spend the next five minutes writing a paper or watching a YouTube video, she would rationally choose YouTube, the more pleasurable activity. After five minutes, she would rationally make the same decision again. But when the decision is viewed under a more integrated time horizon—four hours of paper writing versus four hours watching YouTube—she would rationally choose to write her paper. Because people are susceptible to nonintegrated decisionmaking, even small tastes for immediate gratification, or small costs associated with a task, may cause a naïve person to continuously postpone making decisions.

The converse of nonintegrated decisionmaking is that procrastinators will be highly sensitive to very small short-term incentives or penalties. Policies that make the cost of a short delay loom larger thus make procrastination less likely.¹⁹

Time-inconsistency bias may prove costly to claimants filing with a fund and to the administrative operation of the fund. Although some settlement funds fix relatively short deadlines, requiring filing within three to six months of settlement, other more complicated mass tort funds may allow one to two years to file. In many cases, there is no overt penalty for failing to file at an earlier time. There is a very powerful hidden penalty, however, to claimants—the time value of money and potential lost interest. For example, as illustrated in the graph below, more than half of the families affected by the September 11 attacks waited two years to file with the September 11 Victim Compensation Fund; as a result, each gave up, on average, over $100,000 in lost interest per year.

September 11 Victim Compensation Fund Claims Filed²⁰

Note the large spike in claims that appears just before the filing deadline on December 22, 2003. Rational considerations certainly explain some of the late filings. Claimants may choose to gather more information before filing with a large settlement fund. Or, particularly in large funds involving personal injuries, parties may need additional psychological distance from the event that gave rise to the claim. These explanations, however, are insufficient to account for the concentration of claims that appear just at the filing deadline of many large settlement funds. It is more likely that many claim filings represent present-biased preferences.

One solution is that large settlements could adopt rolling deadlines to encourage earlier filings. Parties could be required to file in the first week of each month until the final deadline. Human resource departments often use such rolling window systems to encourage employees to enroll in benefit programs, but these systems have never been applied in large public or private settlement funds. Cognitive science, however, suggests that such short-term incentives will encourage claimants to file more often over the duration of the fund, saving both opportunity costs to claimants and administrative costs to the fund.

Any such solution must take costs into account. Undoubtedly, rolling deadlines impose a cost on individual actors, who would suffer the inconvenience of filing at the beginning of the month, as well as on the fund, which would have to expend additional resources making such a filing system easy and transparent. But it would impose comparably small costs to claimants’ autonomy. A party unable to file at the beginning of the first month would always retain the ability to file the following month. A party that wants to wait for other strategic, information-driven, or psychological reasons would still retain that right.

Because of such costs, rolling deadlines may be more justified in funds that award high-value claims, like mass torts and some antitrust settlements, but not necessarily low-value claims, like consumer class actions. For high-value claims, the additional savings to the individual and the fund justify taking measures to encourage parties who might otherwise suboptimally delay filing.²¹

***

By “funding irrationality,” I do not challenge efforts to increase choices and opportunities for claimants to large funds. I only question whether such efforts, by themselves, are enough to accomplish their objectives of greater fairness, efficiency, and equity. Although such measures help rational participants to monitor, object, and exclude themselves from such funds, few measures exist to protect claimants who will make decisions based upon cognitive error. As this Essay demonstrates, there will be cases in which, on balance, many subjects will make poor decisions—both for themselves and for the fund as a whole—when available settlement options are not adjusted to account for cognitive biases. This is, in part, because in many large funds parties lack individual access to third-party expertise, like lawyers. Given the tremendous economic, social, and institutional resources devoted to operating large funds, it is worth asking: Are there better ways to design large funds? Can their design accommodate both rational and irrational decisionmaking?

I answer both questions with a qualified “yes” by recommending policies that benefit those prone to make cognitive errors but impose minimal costs on those who otherwise choose rationally. In so doing, I recommend accounting for and sometimes exploiting the timing, structure, and combination of options in large settlements to increase the welfare of all potential participants.

But such solutions raise fundamental questions of fairness and efficiency themselves: Will fund designers suffer from their own biases? Will procedures that fund irrationality unfairly limit claimants’ rights to control their own litigation? Will funding irrationality risk replacing one set of claimant biases with new biases that lead to even less desirable outcomes? These are all valid concerns. The compensatory goals of large funds require, however, that fund designers understand how claimants make choices and, when possible, adjust rules so that funds better serve them.

Acknowledgments:

Copyright © 2010 Duke Law Journal.

Adam S. Zimmerman is an Acting Assistant Professor at the New York University School of Law.

This Legal Workshop Editorial is based on the following article: Adam S. Zimmerman, Funding Irrationality, 59 DUKE L.J. 1105 (2010)

The false confessions in DNA cases all involved very long interrogations, and most involved vulnerable juveniles or mentally disabled individuals. In all cases but one (95% or 36 of the 38 exonerees for whom trial or pre-trial records could be obtained) police reported that suspects confessed to a series of specific details concerning how the crime occurred. As the prosecutor in Robert Miller’s case briefly put it, “He supplied detail after detail after detail after detail. And details that only but the killer could have known.”³ The non-public facts contained in confession statements then became the centerpiece of the State’s case. Although defense counsel moved to exclude almost all of these confessions from the trial, courts found each to be voluntary and admissible, often citing to the apparent reliability of the confessions. The facts were typically the focus of the State’s closing arguments to the jury. Even after DNA testing excluded these people, courts sometimes initially denied relief, citing to the seeming reliability of these confessions. For example, Nathaniel Hatchett had been excluded by DNA testing at the time of trial, but postconviction, the court concluded that “the prosecution presented overwhelming evidence” where the “defendant’s statement included information that only the perpetrator of the crimes would know.”⁴ The ironic result is that the public learned about these false confessions in part because of the contaminated facts. These false confessions were so persuasive, detailed, and believable that they resulted in convictions upheld during appeals and habeas review. After years passed, and appeals and habeas petitions were all denied and dismissed, these convicts had no option left but to seek the DNA testing that finally proved their confessions false.

False confessions uncovered by DNA testing are almost certainly not representative of other false confessions, much less confessions more generally. The 40 cases examined, consisting chiefly of confessions to 1980’s rape-murders, cannot speak to how often people confess falsely. While unusual, only in such examples of known false confessions can one assess whether detailed or supposedly non-public facts contaminated a confession. These data provide a set of examples of a very troubling problem that deserves further study.

One reason why these examples are important is that courts, lawmakers, and executives, while perhaps increasingly aware that false confessions can occur, continue to ignore that there is even a possibility that false confessions may appear uncannily reliable precisely because of the insidious problem of confession contamination.  Justice Alito, concurring in U.S. Supreme Court in District Attorney’s Association v. Osborne, noted that “[a]fter conviction, in an unsuccessful attempt to obtain parole, respondent confessed in detail to the crime.”⁵ Of course, if it is true that Osborne is innocent, he certainly could have known the details of his case. His “confession” to the parole board was by definition contaminated because he could very easily admit guilt “in detail” based on the victim’s account of the crime that he heard at his own trial. In one additional recent example, Virginia’s Governor granted a conditional but not full pardon in the “Norfolk Four” case, despite DNA exclusions, citing to the “cumulative power” of confession statements, but never acknowledging that the statements could have been contaminated.⁷ The problem of contamination is thus not limited to the accuracy of the crime narrative, but also extends to the interrogation room shaping of mens rea and of defenses that the suspect might otherwise rely upon. We should look more carefully at the substance of scripts used by police interrogators and the substance of statements by suspects regarding defenses and mens rea, and not just crime details.

Perhaps the prolonged involvement of law enforcement in generating a confession, true or false, should itself raise concerns about possible contamination. Barry Feld’s work reviewing recorded juvenile interrogations in Minnesota suggests that the typical interrogation of a suspect lasts much less than an hour, and not hours, and without prolonged questioning, elaborate ruses, or schemes designed to produce coercion. Rather, in those interrogations of juveniles, many of which produced confession statements, police used open-ended questioning designed to elicit a narrative, together with questions designed to directly confront them with an accusation that they committed the crime or accusing them of lying. Other studies suggest that the vast majority of interrogations, even in more serious cases, are fairly short and benign. Feld asks us to consider, “Are confronting suspects with false evidence and lying during interrogation necessary evils?”⁸ The over-the-top and elaborate psychological techniques used in lengthy interrogations described in police manuals and used in many of these DNA exonerees’ harrowing interrogations may not be the norm. Perhaps such lengthy interrogations are needed in certain difficult and serious cases. However, scholars are right to increasingly scrutinize lengthy interrogations using complex confession-making tactics—particularly if the ornate storytelling, narrative formation, fabrication, deception, threats, psychological manipulation, and the like occurs in darkness, absent any electronic recording.

A series of reforms could reorient our criminal system towards the substance of confessions. First, although constitutional criminal procedure could regulate reliability, such constitutional change is unlikely. However, an understanding of the vulnerability of confessions to contamination can also inform courts reviewing trials postconviction, particularly in cases involving persons vulnerable to suggestion, such as juveniles and mentally disabled individuals. Second, unless interrogations are recorded in their entirety, courts may not be aware of what transpired during the confession-making process, much less detect contamination of facts, especially when no DNA testing can be performed. In recent years there has been a concerted movement towards reform. In response to some of these false confessions, state legislatures, police departments, and courts have increasingly required videotaping of entire interrogations. Third, additional police procedures can safeguard reliability, such as procedures intended to assure against contamination, to assess suggestibility of vulnerable suspects, and to avoid coercion postadmission.

Contamination of confessions is to a great degree preventable, and the remedy involves making interrogations less secretive. The remedy—electronic recording combined with reliability review—may produce benefits to law enforcement and to society beyond detecting errors. The Supreme Court complained in Miranda, “Interrogation still takes place in privacy. Privacy results in secrecy and this in turn results in a gap in our knowledge as to what in fact goes on in the interrogation rooms.”⁹ Recording will help to close that gap in our knowledge about interrogation practices. In most cases, police will document professional, brief, accurate, and unprompted confessions. However, recording and review will also help to identify the hopefully dwindling number of contaminated false confessions, which before the advent of DNA technology had remained shrouded in secrecy and darkness. Sunlight in the interrogation room will bring interrogation practices out into the open, so that we can better understand and evaluate them. That shift towards reviewing the substance of confessions has the potential to enhance confidence in both the accuracy and legitimacy of the criminal process.

Acknowledgments:

Copyright © 2010 Stanford Law Review.

Brandon L. Garrett is an Associate Professor of Law at the University of Virginia School of Law.

This Legal Workshop Editorial is based on the following Law Review Article: Brandon L. Garrett, The Substance of False Confessions, 62 STAN. L. REV. 1051 (2010).