Judging Facts like Law

Article excerpt

INTRODUCTION

The Supreme Court's review of legislative facts found by Congress can make all the difference between enjoying a constitutional right and losing it. The Court's recent decision in Gonzales v. Carhart (1) powerfully illustrates this point. There the Court in an opinion by Justice Kennedy held that women did not possess a right to a "partial birth" abortion, although six years previously in Stenberg v. Carhart (2) the Court held the opposite. According to Justice Kennedy, one of the two key differences between the current and prior case was that in Gonzales Congress had found that the practice was a "gruesome and inhuman procedure that is never medically necessary." (3)

The opinion, however, also underscores the peculiar and radically under-theorized nature of the treatment accorded congressional fact-finding. Justice Kennedy admitted that several of Congress' other findings about partial-birth abortions were factually incorrect. (4) The patent infirmity of many of Congress' fact assessments on the very subject under consideration raises obvious questions about why the Court should be influenced by any of Congress' other findings.

Second, the Court actually did not defer to the Congress' strong conclusion that partial-birth abortions were never medically necessary, but to the weaker implicit claim that it was at least uncertain whether partial-birth abortions were ever medically necessary. (5) The Court argued that this weak claim was all that was necessary to sustain the facial constitutionality of the statute because Congress should have the discretion to regulate abortion under conditions of medical uncertainty. (6) Thus, the Court reconstructed Congress' fact-finding before giving it weight.

Moreover, the opinion provides completely unconvincing support for its stance toward congressional fact-finding. Justice Kennedy cites precisely one case in his discussion of why the Court should give weight to congressional fact-findings but not simply accept them. The case is Crowell v. Benson, (7) a famous administrative law decision from seventy-seven years ago in which the Court considered the question of how to scrutinize facts for ascertaining that work had taken place in navigable waters--a premise that was necessary to establish that federal jurisdiction over the workmen's compensation claim was constitutional. (8) But in the relevant portion of that case the Crowell Court was reviewing an administrative agency's fact-finding, not Congress'. (9) Moreover, the Court held that the findings necessary to establish the constitutional authority to extend federal jurisdiction were subject to the federal judiciary's de novo review, not a review with any degree of deference. (10) The current Court's citation of such a manifestly inapposite case highlights that it still has no coherent theory of the judicial role in cases where Congress has found social facts to support to the constitutionality of its legislation.

This article responds to the Court's confusion about fact-finding in the partial-birth abortion case by offering a comprehensive view of how the Court should treat legislative views of social facts, such as the medical need for partial-birth abortions, which provide an essential foundation for the constitutionality of legislation. We reject the notion, which the Court often but inconsistently deploys, that the judiciary should treat legislative views of the facts more deferentially than legislative views of the law.

Displacing the judiciary from an independent, de novo fact-finding role must depend either on the argument that law is inherently different from fact or that Congress has a greater comparative advantage vis-a-vis the judiciary in finding social facts than in assessing the law. Neither is supportable. There is no analytic dichotomy between law and fact. Law is a social fact, just as are the data or statistical analysis that may be relevant to questions such as whether partial abortions are ever medically necessary.

Thus, the key question about whether the judiciary should defer to Congress' fact-finding or rely on its own assessment of social facts is functional. If Congress had a superior ability to find facts, deference might make sense. But we show that Congress' fact-finding abilities are less capacious and more biased than those in the judiciary. As an elected body, Congress is designed to respond to its constituents' subjective desires, not to the objective facts of the world. In contrast, the judiciary is insulated from the preferences of constituents and less subject to partisan bias. Its salient institutional structure is the adversarial proceeding where each side has incentives to scrutinize relentlessly the factual claims of its opponent. Accordingly, the judiciary would appear to be a superior fact-finder both because of its institutional capacity and because of its relative lack of bias.

In arguing for independent judicial evaluation of facts necessary to assess legislation's constitutionality, we do not suggest that the Court should sit in oversight of the record of Congress' fact-finding. It may well create tension with the separation of powers for the judiciary to sift through facts Congress has found. But it does not follow that a judiciary that declines to sit in judgment on the internal work product of a coordinate branch should not compile its own factual record to test the constitutionality of a statute.

Indeed, the separation of powers supports a de novo judicial role in fact-finding. It is now well established that each branch of government has the responsibility to measure the text of legislation against the Constitution. By similar logic, each branch has the obligation to determine whether facts support the constitutionality of legislation. For instance, if the constitutionality of partial-birth abortion turns on the question of whether partial birth abortions are ever medically necessary, each branch should make this determination for itself. Just as the departmental approach to constitutional interpretation safeguards the rights and structures of the Constitution by subjecting the constitutionality of legislation to multiple scrutiny in legal interpretation, so too does a departmental approach to constitutional fact-finding.

One reason the Court has been so inconsistent in taking responsibility for fact-finding is that it has never laid out procedures for doing so. Successful procedures require transparency and the right of all parties to be heard in an adversary process because these are the characteristics of judicial fact-finding that increase the capacity for accuracy and decrease the risk of bias.

Our article proceeds in four parts. Part I considers the Supreme Court's treatment of legislative or social facts that provide a necessary predicate to the constitutionality of legislation. We first show how these social facts differ from the adjudicative facts. Second, this section shows that the Court's implicit or explicit determination of social facts has been essential to many well known holdings. Third, we show that the Court has been inconsistent and result oriented in its approach to social fact-finding. Even within the same area of law, such as the First Amendment, the Court has sometimes deferred to Congress and sometimes has not. It has sometimes relied on a lower court record of facts entered into evidence, but sometimes it has taken "judicial notice" of controversial facts for itself.

In Part II, we review scholarly treatments of the judiciary's role in finding social facts. Unfortunately, most previous scholars have failed to provide a useful framework of analysis. First, many tend to complain about the Court's deference to congressional fact-finding or lack thereof in a particular area of law, like the recent federalism cases. In our view, this focus confuses the specific question of whether judicial scrutiny of the legal area is appropriate with the general question of whether judicial scrutiny of social facts is appropriate. Second, other scholars have recently attacked the Court for reviewing the record of fact-finding compiled by Congress. We agree with this criticism, but do not believe it shows that the Court should not be engaged in it own fact-finding. We also briefly distinguish our analysis from that of other scholars, such as Richard Posner, who argue that legal doctrine should depend more on analysis of social facts than formal rules. Our point is more limited: whenever legal doctrine requires the analysis of social facts, the judiciary should analyze the social facts for itself de novo.

In Part III, which is the heart of our analysis, we show that the legislature has no advantage in social fact-finding compared to the judiciary. First, we demonstrate that there is no sharp analytical difference between legal interpretation and social fact-finding. Second, we offer two kinds of analysis that suggest that the courts are at least as competent as Congress in finding facts. The first analysis, drawn from public choice theory, regards members of Congress as interested primarily in reelection. As a result, they will tend to focus more on the preferences of constituents rather than objective facts. We also broaden the traditional focus of public choice to consider the incentives of the public at large. We show these incentives will also undermine the accuracy of congressional fact-finding, because the public expects Congress to find the facts that support its preferences, rather than the actual facts. Our second analysis relaxes the public choice view and permits members of Congress a substantial independent interest in the accuracy of the facts. Nevertheless, because of cognitive biases, such as the confirmation bias and the availability heuristic, Congress remains less likely to find facts accurately than is the judiciary with its more disciplined, adversarial methods.

Finally, we show that the increasing consensus on departmentalism--the notion that each branch should independently interpret the Constitution--also strongly supports an independent role for judicial fact-finding. While departmentalism has recently emphasized that the legislature and executive are independently charged with legal interpretation and should not yield that responsibility to the Court, departmentalism similarly shows that that the Court has a de novo role in fact-finding which it should not yield to the legislature or the executive.

In Part IV, we argue that, regardless of whether Congress has found social facts, the judiciary must create a transparent and adversarial process of social fact-finding for itself, because it is those qualities that provide its comparative advantage in its sphere. Thus, the lower courts should determine what facts are relevant to the constitutionality of a statute and compile a record on those facts. In this way, parties will have notice and be able to bring to bear the best and most relevant evidence. The Supreme Court should limit its factual determinations to this record and not be swayed by last minute proffers of facts that have not been subjected to an adversarial process.

I. CONGRESSIONAL FACT-FINDING IN CONSTITUTIONAL LAW

In this section, we set the stage for our own analysis of how the Supreme Court should act in cases involving congressional fact-finding by briefly reviewing its past performance. The Court's treatment of congressional fact-finding has been marked by three kinds of substantive inconsistencies. First, it has been inconsistent in its standard of deference to those findings. Sometimes it defers and sometimes it does not, even in the same areas of the law, raising questions about whether it simply makes up its mind as to whether it agrees with Congress and then fabricates a standard of review. Second, the Court is inconsistent in deciding whether its review of facts is limited to facts found by Congress on a record or whether these facts can be supplemented by other evidence supporting the constitutionality of the statute that was not relied upon by Congress. Third, the Supreme Court is also inconsistent in sometimes relying on a factual record laid out by lower courts and yet at other times relying on facts offered by amici or of its own devise--facts that have never been subject to any adversarial process. The approach we suggest in Part IV would eliminate these inconsistencies by requiring the judiciary to do its own fact-finding and the Supreme Court to confine itself to the factual record created by lower courts. It would prevent ad hoc claims of deference from substituting for the Court's own reasoned analysis of the factual predicate of legislation.

To pinpoint the kind of fact-finding with which this article is concerned we also briefly describe the difference between adjudicative facts on the one hand and legislative or social facts on the other. We then discuss those cases where the Supreme Court's own doctrine makes the existence of social facts dispositive of the constitutionality of legislation. Our principal concern in this article is with a subset of those cases--those in which the legislature, generally Congress, has itself found that the relevant social facts support the constitutionality of the statute. (11)

A. LEGISLATIVE FACTS AND ADJUDICATIVE FACTS

Legislative facts are best understood by comparing them to adjudicative facts. Adjudicative facts cover "what the parties did, what the circumstances were, [and] what the background conditions were." (12) Legislative facts, in contrast, are facts that are less particular to the parties, but rather "are general facts which help the tribunal decide questions of law and policy and discretion." (13) While no bright line distinction exists between legislative and adjudicative facts, the former tend to "relate legislative policy to the purported constitutional authorization of a statute." (14) Legislative facts "transcend individual disputes and would likely recur in different cases involving similar subjects." (15)

Legislative facts are often salient to employing balancing tests, (16) strict scrutiny, (17) determining what constitutes an endorsement of religion, (18) whether there is a national consensus that the juvenile death penalty is cruel and unusual, (19) whether a six-member jury functions in the same way as a twelve-member jury, (20) and in the many other doctrines the Supreme Court utilizes to "say what the law is." (21) Indeed, since the Supreme Court takes cases in order to formulate law rather than solely to adjudicate particular cases, legislative facts often drive its opinions much more than adjudicative facts.

Many of the Court's judicial opinions reviewing congressional legislation do not address the issue of congressional fact-finding simply because Congress had made no such findings in connection with the law at issue. But if one believes that Congress takes the Constitution seriously, then Congress implicitly finds facts whenever a set of facts are necessary to the constitutionality of a law, since in passing the law, Congress is making a determination that the law is constitutional. When the Court strikes down a congressional law, it likewise makes an explicit or implicit determination that Congress got the facts wrong. To provide a well known example, "separate educational facilities are inherently unequal" was the law set forth in Brown v. Board of Education. (22) It was also a legislative fact found by the Court. On the same day as Brown, the Court applied this rule to the federal government via the Fifth Amendment in Boiling v. Sharpe. (23) A necessary implication of that opinion is that the Court did not consider Congress to be the superior fact-finder on the question of whether segregation in the Washington, D.C. schools generated such a feeling of inferiority in black students to the point of being inherently unequal. (24) One way of focusing our inquiry is to ask whether the Court should have acted any differently had Congress assembled a record of factual findings supporting the constitutionality of the segregated D.C. public schools.

B. INCONSISTENCIES IN STANDARD OF REVIEW

When the factual assumptions of Congress are more explicit the Court has been unable to formulate a consistent approach towards Congress' fact-finding. (25) Sometimes the Court explicitly defers to the facts found by Congress, sometimes it makes an independent judgment of the facts, and sometimes it engages in a combination of these approaches

Even on one specific issue, such as whether the danger of Communism is significant enough to warrant restrictions on First Amendment rights, the Court has been inconsistent on its deference to congressional findings. For example, in Dennis v. United States, the Court deferred to Congress' finding that members of the Communist Party of the United States posed a sufficient threat to national security to be convicted under the Smith Act. 26 After citing hearings before the House Committee on Un-American Activities, Justice Frankfurter in his concurrence concluded that it was not the Court's position to replace a judgment Congress had made with "due deliberation." (27)

Sixteen years later, however, in United States v. Robel, the Court struck down a provision of the Subversive Activities Control Act which barred any member of a communist organization from working at a defense facility. (28) The statute at issue was found to be an overbroad impediment to freedom of association rights. The Court found that "[a] number of complex motivations may impel an individual to align himself with a particular organization," so membership in a communist organization was "insufficient to impute to him the organization's illegal goals." (29)

Justices Harlan and White both dissented, maintaining that the Court had wrongly "arrogate[d] to itself an independent judgment of the requirements of national security." (30) The dissenters argued that Congress passed the statute after years of investigation into communism, which led to fact-findings of an international conspiracy with thousands of adherents in the United States intent on employing "treachery, deceit, espionage, and sabotage" to overthrow the government of the United States. (31) The Robel Court simply declined to provide Congress with the same deference that the Dennis Court provided on the politically volatile issue of communist infiltration. (32)

There are more recent examples of inconsistent standards of deference in analyzing social facts dispositive to individual rights. In Board of Ed. v. Mergens, the Court upheld the Equal Access Act, which prohibits secondary schools from denying religious student groups equal access to school facilities. (33) Before enacting the law, Congress found that secondary students would be able to distinguish student expression of religious speech from state endorsements of religion. (34) After citing the Senate reports on the topic, the Court added, "we do not lightly second-guess such legislative judgments, particularly where the judgments are based in part on empirical determinations." (35)

In other recent instances, however, the Court has ignored Congressional findings altogether. In Dickerson v. Unites States, the Court was faced with the question of whether the Miranda rule could be supplanted by congressional statute. (36) In Miranda v. Arizona the Court announced the warnings that are required if confessions are to be deemed voluntary, and therefore admissible into evidence. (37) The Miranda Court made extensive findings about the "nature and setting" of in-custody interrogation, including a survey of police manuals from around the country. (38) The Miranda opinion acknowledged that the appropriate way to enforce the privilege against self-incrimination was dependant on legislative facts, and even invited legislative alternatives to the newly announced rule. (39) Two years after Miranda, Congress enacted a rule that confessions would be admissible in federal prosecutions so long as they were voluntarily made, even if Miranda warnings had not been given. (40) Congress made extensive factual findings in the legislative record to support the constitutionality of this legislation. (41) When the law was eventually challenged in Dickerson, the Court received briefs emphasizing the factual findings that Congress relied on when it enacted the congressional statute, (42) but the Court did not even address, let alone defer, to these findings in its opinion. (43) It simply held that the remedies Congress offered were not sufficient to reduce the risk of coerced confession. (44) The Court's tendency to employ different standards of review towards congressional fact-finding not only undermines the candor of its opinions but the accuracy of its own fact-finding

Because the Court is able to simply defer to Congress when it agrees with the factual predicate for its statute, it does not have to provide a more formal accounting of what factual support the statute actually enjoys. Our framework of de novo review of the facts necessary to support the constitutionality of a statute would thus not only provide greater consistency but promote greater accuracy as the Court became more rigorous about social fact-finding. (45)

C. INCONSISTENCY BETWEEN "ON THE RECORD REVIEW" AND REVIEW INDEPENDENT OF THE CONGRESSIONAL RECORD

In most cases where Congress has found social facts, the Court does not treat these facts as the exclusive set relevant to determining the constitutionality of the statute. Regardless of whether it defers to these facts or does not, it considers other data that the litigants develop. (46) In cases involving Congress's power to enforce section 5 of the 14th Amendment, however, the Court has more recently reviewed the constitutionality of the statute on the fact-finding record that Congress has assembled to try to establish whether Congress's enforcement scheme is "congruent and proportional" to the rights Congress is trying to enforce. (47) In Kimel v. Board of Regents, the Court found the Age Discrimination in Employment Act (ADEA) could not abrogate state sovereign immunity because "[a] review of the ADEA's legislative record as a whole, ... reveals that Congress had virtually no reason to believe that state and local governments were unconstitutionally discriminating against their employees on the basis of age." (48) One year later, in Board of Trustees of the University of Alabama v. Garrett, an abrogation provision of the Americans with Disabilities Act was similarly invalidated. (49) The majority opinion concluded that Congress failed to identify a pattern of irrational state discrimination against the disabled. (50) In Nevada Department of Human Resources v. Hibbs (51) and Tennessee v. Lane (52) the Court upheld Congress abrogation of the Eleventh Amendment, but only after satisfying themselves that Congress had found sufficient facts on the record. (53)

This inconsistency in the Court's cases is also troubling First, it is not at all clear why the Constitution justifies on the record review in some cases and not others. Second, on the record review provokes tension with Congress, because the Court is delving into congressional work product. (54) In contrast, an independent review of facts would bring unity across constitutional doctrine and reduce direct conflicts with Congress

D. THE METHOD OF REVIEW: THE LOWER COURT RECORD AS A FACTUAL RECORD OR A FACT-FINDING FREE-FOR-ALL

The Court has never developed a consistent framework for receiving legislative facts into evidence. This inconsistency appears whether Congress has made factual findings or not. The Court does not seem to require a fully developed factual record from the lower courts before deciding a constitutional question Indeed, in many instances the Court has not confined its review to evidence introduced in the district court, instead relying on evidence submitted by amicus briefs at the Supreme Court level

The Court's lack of a firm methodology for constitutional fact-finding is exemplified by Waiters v. Nat'l Association of Radiation Survivors, in which the Court held constitutional the congressional limitation of a $10 maximum fee that a veteran may pay an attorney or other representative before the Veterans Administration when seeking benefits for a service related death or disability. (55) The Court reversed a district court's grant of a preliminary injunction, finding that the congressional statute did not violate due process, and explicitly gave deference to congressional findings on the appropriateness of the fee limitation. (56) Despite the fact that the due process inquiry was very much dependent on legislative facts and despite complaints from concurring and dissenting Justices that there was an inadequate factual record, (57) the Court declared that the statute was not per se unconstitutional based on the very limited lower court record from the preliminary injunction hearing. (58) The fact that the Court did not consider a remand warranted demonstrates an unwillingness to engage in a rigorous approach towards fact-finding.

Even in cases that do not involve congressional findings, the Court will often take judicial notice of factual contentions that were not introduced in the adversarial proceedings in the lower court. In Grutter v. Bollinger, for instance, which upheld the University of Michigan Law School's race-based admissions policy, the Court relied on factual assertions in the amicus briefs of educational associations, businesses and some retired generals in finding that diversity in education is a compelling state interest, although these claims had never been subject to cross examination or other procedural scrutiny. (59) In United States v. Virginia, (60) where the Court held that the Virginia Military Institute's male-only admission policy violated equal protection, the Court relied on amicus briefs and newspaper articles in making its finding that the justification for the policy was not "exceedingly persuasive." (61) Although this conclusion directly conflicted with the findings of the district court, (62) the Court did not find error in the lower court's findings but rather simply substituted its own new, ex parte evidence. (63)

This kind of procedural inconsistency undermines the virtues of judicial fact-finding, and accordingly in Part IV we offer a framework that helps assure that judicial social fact-finding follows traditional, adversarial legal procedures that promote fairness and accuracy. The Court is certainly capable of following such a framework. It has engaged in something close to the type of review we advocate in recent cases dealing with pornography under the First Amendment. For instance, in Reno v. ACLU, (64) where the Court found internet regulations aimed at protecting minors unconstitutionally overbroad, the opinion focused on the 410 factual findings made by the district court. (65) The Court based its own factual judgments on the findings in this record and did not rely on evidence drawn from amicus briefs. (66) The Court took this approach because there was no congressional record of fact-findings to rely on. (67) We will argue that the Court should not vary its fact-finding method based on the absence or presence of congressional fact-finding.

II. PAST ANALYSIS OF THE RELATION BETWEEN JUDICIAL AND CONGRESSIONAL FACT-FINDING

The question of the relation between judicial and congressional fact-finding is not a new one, and here we briefly review some of the prior literature on the subject. The early commentators correctly saw that a key determinant of whether the judiciary should make its own de novo findings of fact was the relative capacity of Congress and the Courts to engage in the finding of social facts. The principal problem with their analysis was that heir models of judicial and congressional behavior lacked sophistication. In contrast, modern academic treatments often skirt this key question and instead rely on two other lines of analysis. The first argues that the Court should defer to congressional fact-finding in a particular area of the law. The second contends that the Court should not review the record of congressional fact-finding as if Congress were an administrative agency and the Court were engaged in "arbitrary and capriciousness review."

We believe that neither of these critiques addresses the essential issues for determining the proper role of the court vis-a-vis Congress in social fact-finding. By focusing on congressional fact-finding in specific areas rather than comparing judicial to congressional fact-finding in general, the first set of commentators have tended to confuse the level of deference due to Congress in a particular area of law with the level of deference to congressional fact-finding. The second set of commentators criticize the practice of reviewing a factual record created by Congress, but do not posit how the Court should review the relevant legislative facts. None of the arguments made by these commentators undermines the judiciary's obligation to engage in independent fact-finding on its own record. (68)

We end this section by distinguishing our position from those who want to make social facts bulk larger in constitutional doctrine more generally. We do not seek to make constitutional doctrine depend more on the finding of social facts than implementation of formal rules. Instead we take the shape of constitutional doctrine as a given and suggest only that if social facts are dispositive under that doctrine, the Court should engage in de novo review of what those facts are.

A. EARLY ATTEMPTS TO REFINE THE JUDICIAL ROLE IN SOCIAL FACT-FINDING

Legal academics first began wrestling with the intersection of law and science at the turn of the twentieth century. Legal realism acknowledged that facts were often the driving force behind legal reasoning. (69) Oliver Wendell Holmes, commonly credited as a founder of this movement, argued that "[t]he life of the law has not been logic: it has been experience." (70) Roscoe Pound, (71) Louis Brandeis, (72) and Benjamin Cardozo (73) emphasized the importance of the judicial analysis of social facts in their academic writing.

A 1924 article by Henry Wolf Bikle in the Harvard Law Review was the first academic attempt to address the relation between the judiciary and the legislature in legislative fact-finding in constitutional decisions. (74) Bikle thought the Court had more legitimacy in determining 'legal' questions, like the meaning of "ex post facto law," as opposed to "factual" questions such as whether limiting the hours of a baker has a substantial relation to public health. (75) Nevertheless Bikle conceded that social fact-finding is a necessary aspect of judicial review. (76)

Around the time of the New Deal two students addressed the issue and came to diametrically opposed positions on the judicial treatment of a legislature's fact-finding. A Columbia law student was wary of deferring to legislatures in due process cases in an article published in 1930. (77) This author thought that the facts should be developed by the adversarial trial process. (78) In contrast, a 1936 Harvard student was critical of the Supreme Court's stringent review of legislation, arguing that a minimal legislative record of factual findings should suffice. (79) He thought that the judiciary should not overturn a law unless it appeared that "no reasonable man in possession of such facts could agree with the legislature's finding." (80)

These early notes feature arguments that remain salient today. On the one hand, if judges must defer to a legislature's findings of fact, the legislature can potentially avoid judicial review by shielding themselves with fact-finding that may not be true. (81) On the other hand, if legislatures are more in touch with the factual matters they regulate than judges, the Court should defer to their findings, merely considering whether a reasonable legislator could have found the facts in question to be true. (82) What these articles lack are more sophisticated analysis of legislative and judicial behavior that permit comparison of the advantages of judicial and congressional fact-finding. We try to fill that gap in section III.

B. THE MODERN LITERATURE'S MISTAKEN FOCUS ON SPECIFIC LEGAL AREAS

Unlike these early articles, the modern literature mostly fails to ask the fundamental question of how the judiciary should review legislative facts generally, and instead skips directly to criticisms of how the Court reviews congressional fact-finding in specific legal areas. For instance, some academics have responded to the Rehnquist Court's federalism decisions by criticizing the Court for its failure to defer to congressional fact-finding. For instance, a large number of academics suggest that the Court should defer to Congress's fact-finding under section 5 of the 14th Amendment. (83) Professors Colker and Brundey, Siegel and Post, and Christopher Banks advocate a view of Congress as a superior fact-finder that skillfully analyzes what is in the public interest. (84) As a representative body, they argue, Congress receives information both formally and informally, and in a way that is institutionally different than the Court. (85) Therefore the Court should defer to the facts Congress finds because Congress better reflects the understanding of the people than does the Court. (86)

Insofar as these arguments rest on general claims about Congress's fact-finding ability, these articles make relevant arguments, although in section III we show why they are mistaken. But we do not believe that for the most part these articles are making general comparisons about judicial and congressional fact-finding ability. None of the authors suggest that the Court should defer to Congress across the full range of constitutional doctrine or indeed in many areas other than federalism. For instance, none of these critics criticize the First Amendment cases for failing to defer to Congress' judgment about the practices needed to protect children against indecency. None criticize Dickerson (87) for second guessing the judgment of Congress as to what is the best means of preventing coerced confessions. In fact, Siegel and Post claim that "Garrett represents the opposite danger from Dickerson" because the former "vindicates the value of the rule of law in ways that threaten the value of self determination." (88) They do not elaborate on how Garrett and Dickerson are opposite. In both decisions the Court overruled the judgment of Congress on a fact-laden constitutional question.

Thus, we think these articles are best seen as part of the continuing debate about judicial intervention to protect federalism. Commentators have long made arguments that members of Congress, being elected from states, have the information and incentives to strike the right balance between federalism and other values. (89) Because the political process adequately protects federalism, they argue, there is no need for judicial intervention.

But these arguments apply just as well to Congress's legal interpretations as to congressional fact-finding, and in fact those who believe in the political safeguards argument tend to oppose judicial intervention in federalism cases whether based on law or facts. (90) But if the Court is correct to defer to Congress in federalism cases for these reasons, its deference is not dependent on a close analysis of its institutional fact-finding abilities more generally. And if the controversial claims that Congress is particularly suited to balance federalism against other values prove incorrect, (91) the rationale for deference in fact-finding about federalism dissolves no less than the rationale for limiting judicial review in federalism cases more generally." (92)

C. THE MODERN LITERATURE: CRITICISM OF "ON THE RECORD REVIEW"

Several commentators have also criticized the Court for treating Congress like an agency when it reviews congressional fact-finding. (93) Professor Buzbee and Schapiro offer several arguments why the justifications for record review of administrative agencies should not translate to judicial review of Congress in cases involving Section Five of the Fourteenth Amendment. (94) First, the separation of powers issues at stake in administrative and congressional review are different. (95) The Court reviews the agency's record at Congress' behest to assure legislative supremacy, but protection of legislative supremacy can hardly be a justification for reviewing the record of Congress. (96) Second, record review is too narrow a compass to measure Congress's fact-finding, because members of Congress, unlike agency officials, have access to sources of information too diffuse and wide-ranging to be captured by the record. (97) Third, record review bespeaks an unjustified suspicion of Congress's motives, because it is designed to see whether Congress's factual claims are pretextual. (98)

These three arguments seem solid to us, but they do not suggest that the Court should not compile its own record to review. It is true that independent judicial review of congressional fact-finding is not warranted to protect legislative supremacy. Nevertheless independent judicial fact-finding still may be warranted to protect the supremacy of the Constitution by assuring that a statute is supported by the factual predicate constitutionally required. Congress's diffuse sources of information do not negate an independent role for the Court if the Court has comparative or even distinctive strengths in fact-finding. Finally, judicial fact-finding on its own record does not impugn Congress' motives any more than does judicial interpretation of the Constitution, because in passing legislation, Congress must find that it is constitutionally warranted both in law and in fact.

Professors Frickey and Smith also argue that the Court should not oversee congressional fact-finding, but their grounds are very different: Congress is not a fact-finding body at all but rather a market for competing influences. (99) But this critique suggests only that the Court should not engage in fact-finding by reviewing a congressional record because that record is inherently unreliable. It does not suggest that the Court should abandon review of the facts. Indeed, because under Frickey/Smith's view Congress cannot be trusted to care about the facts relevant to a constitutional restraint, their analysis would appear to support an independent and de novo judicial review of the facts. (100)

Thus, while commentators have persuasively challenged "on the record review" of congressionally found facts, these arguments do not suggest, as we do, that the judiciary should engage in its own independent de novo review.

D. THE SCOPE FOR SOCIAL FACT-FINDING IN CONSTITUTIONAL DOCTRINE

Finally, our position differs from those who want to make constitutional doctrine more dependent on social facts and less on formal rules. For instance, Judge Posner's constitutional "pragmatism" is based on the belief that the constitution is so vague, and it has been interpreted so inconsistently, that the only way to resolve difficult questions is to look empirically at how different doctrines would help achieve widely shared policy objectives. (101) Rather than look to the "purpose" of the legislation (an impossible task, in his view) or to abstract principles of morality or legal reasoning, Posner thinks that judges need empirical knowledge for the "task of exploring the operation and con. sequences of constitutionalism." (102) We, in contrast, do not take a position on the extent to which constitutional doctrine should turn on social facts. Thus, we do not take a position on whether the medical necessity of partial-birth abortions should be relevant to the constitutional jurisprudence of abortion. Assuming that a particular constitutional doctrine does depend on a question of social facts, we instead analyze how the Court should act when Congress has claimed to have found the social facts necessary as a constitutional predicate.

III. JUDICIAL VS. CONGRESSIONAL FINDING OF SOCIAL FACTS

In this section, we dispel philosophical and pragmatic objections to independent judicial assessment of social facts. We first show that there are no ontological or analytic distinctions between interpretation of law and finding of fact. Legal meaning and factual findings are both social facts and subject to conventional methods of proof and disproof.

We then show the judiciary is at no comparative disadvantage in social fact-finding. We first consider a pure public choice perspective on Congress' fact-finding ability. In a public choice model, Congress is viewed as an institution populated by members who are predominantly interested in reelection. Such a model suggests that Congress would have little interest in finding the actual facts about social policy, because they would be primarily focused on satisfying the preferences of their constituents. Moreover, drawing on very recent public choice analysis, we show that these constituents would expect Congress to find social facts in line with their preferences rather than the actual facts. We then relax the stringency of the public choice model and credit members of Congress with an independent interest in getting the relevant facts right. Even under these assumptions, which we call the behavioral law and economics model, congressional fact-finding will often be deficient, because interest groups will wield substantial influence on shaping the consideration of facts before Congress and members will be subject to cognitive biases not easily dislodged by the process of congressional hearings. Thus, it is not only the motivations of Congress but also its structure that inhibit accurate finding of social facts.

In contrast, the federal judiciary is insulated from the pressures of constituents because judges do not face reelection. Moreover, interest groups yield relatively less influence in litigation than legislation. The adversarial process also disciplines cognitive bias and limits the effects of certain heuristics that make congressional fact-finding less accurate.

We then respond to arguments that the judiciary is nevertheless less competent than Congress at social fact-finding because Congress, like a jury, has a range of perspectives that the judiciary lacks. Social facts, however, do not resemble adjudicatory facts that can be resolved by the collective and diverse experience of individuals. Like the question of whether abortions are ever medically necessary, such social facts tend to depend on technical and statistical investigation that is beyond most people's experience. What the judiciary lacks in diversity it makes up for in technical competence. In any event, judicial insulation and lack of diversity is at least as much of a problem in judicial legal interpretation and yet judicial review of the law remains a cornerstone of our polity.

We move from social science analysis to structural analysis of the Constitution and show that departmentalism--the notion that each branch has an independent duty to interpret the Constitution--supports de novo judicial fact-finding even if judicial fact-finding is not better than congressional fact-finding, but simply has different strengths. We end by responding to those who have argued that the judiciary should defer to fact-finding in some areas of the law, such as separations of powers or campaign finance, because of the peculiar incentives at work there. We show that our analysis of legislative and congressional behavior still suggests that judicial fact-finding has distinctive strengths throughout the law and thus judges should exercise a de novo review of facts across the board.

A. THE COMMONALITY OF LAW AND FACT

There is no ontological or analytic difference between law and fact that makes social facts unsuitable for judicial investigation and proof. Indeed, like many scholars of evidence, we do not believe an ontological or epistemological distinction exists between what is 'law' and what is 'fact.' (103) Consider, for instance, the process by which a judge would discover the meaning of "recess" in the Recess Appointments Clause. (104) An originalist judge would weigh various bits of factual evidence from the time of the Framing. He might consult an eighteenth-century dictionary to pin down the conventional usage of the time. He might look at the ratification debates to gain a better understanding of the clause's meaning in the particular context. Thus, he would be investigating social facts about the world. A judge who embraced a consequentalist approach would employ a different legal method, but nevertheless also sift through facts about the world. For instance, she might focus on how the various branches would interact, depending on the meaning assigned "recess." She might consult political science texts that describe the relation between the President and Congress, and review the aftermath of past recess appointments.

Legal questions are simply "part of the more general category of factual questions." (105) As such, both are subject to dispute as to their content and need to be established by weighing conflicting evidence. (106) To be sure, the kind of evidence to be weighed may differ between pure legal interpretation and fact-finding, but, as our example above shows, different kinds of legal interpretation require the weighing of different kinds of evidence. Indeed, it is ultimately difficult to understand what it would mean to adhere to a metaphysical or epistemological distinction between legal interpretations and social facts since law itself is a social fact. (107)

As a result, the real questions about whether the judiciary should wield an independent, de novo role in social fact-finding are pragmatic, comparative, and structural. In our next sections, we thus turn to two questions. First, does Congress have an advantage over the judiciary in finding social facts? Second, even if one were to agree that Congress has some general advantage, does the judiciary nevertheless bring some distinctive strengths to social fact-finding?

B. THE RELIABILITY OF CONGRESSIONAL FACT-FINDING

The prevailing view has largely been that legislative bodies (and particularly Congress) are better than courts at making broad factual determinations. (108) This belief is based in part on Congress' size and diverse membership, its responsiveness to the public needs and its tendency to look at the "big picture" rather than an individual case in which narrower issues are decided. But the assumption that Congress is a superior fact-finding body overlooks the incentives of the individuals who compose it. Congress certainly can assemble information designed to support its legislative objectives. But Congress' political nature makes it unlikely that this process will resemble a fair and unbiased attempt to evaluate all the facts relevant to the legislation and its constitutionality.

First, we consider how public choice analysis, which applies "economic models of human behavior to the political process," (109) would model the fact-finding abilities of legislatures. Once one recognizes that well organized interest groups have an incentive to use their power to the detriment of others, (110) and elected politicians have an incentive to seek rewards in votes and money, the reliability of congressional fact-finding should be immediately called into question. Here we add an important but never before discussed confirming factor: the public will expect Congress to find the facts that support its preferences, not the actual facts. Second, we apply a more mixed model rooted in

behavioral economics and assume that members of Congress also pursue some general sense of the public interest in their fact-finding. We will show that even if this is the case, Congress' fact-finding has systemic defects because its structure for fact-finding cannot cabin the biases created by widely shared behavioral traits.

1. The Public Choice Analysis

a. Incentives of Legislatures Under a Public Choice Analysis

The basic assumptions of public choice theory suggest that deference to the factual findings of Congress is undesirable. If members of Congress care only or even significantly about reelection and the material benefits that can be provided by interest groups, they cannot be trusted to accurately find the facts that determine a statute's constitutionality. Indeed, the facts about which they would predominantly care are the demands of their influential constituents and the necessary means of fulfilling these demands. (111) Under this view, Congress might nevertheless conduct committee hearings to investigate facts for a variety of reasons. Hearings may help them learn the demands of constituents. They may create a forum for bargaining with other members for votes. They may also provide a facade to mask what is really driving the content of legislation. For instance, if a powerful company is asking for anticompetitive regulations, the committee may create a focus on consumer complaints in the area. In fact, public choice predicts that members of Congress will try to create information to confuse the opposition while pleasing concentrated interest groups. (112)

Most relevantly for our purposes, members of Congress may generate a record supporting their legislation if challenged in court. But this purpose hardly suggests that the members of Congress will be interested in the objective facts. Instead, they will be focused on creating a legislative record that will put the legislation in the most favorable light. (113) They will in fact be less careful and objective insofar as the judiciary gives the findings more deference, because their assembly and slanting of the facts will escape scrutiny. The Court's decision in Carhart v. Gonzales to defer to Congress' fact-findings even when other facts in the record were found false would hardly encourage Congress to be meticulous or objective. (114) In fact, one of the findings Congress made when enacting the Partial Birth Abortion Act is that its legislative findings would be accorded deference by the Supreme Court. (115) Thus, there is a dynamic relationship between the Court's deferential treatment and the brazenness of Congress.

Therefore, under the public choice model, the Court's current approach of examining the facts found in the congressional record as an aid to determining a law's constitutionality is inherently unreliable. It is comparable to reviewing the facts found by a litigant in an ex parte proceeding. It encourages members of Congress to manipulate the facts that enter the legislative record in order to secure delivery of the goods demanded of them.

Nor is Congress's fact-finding likely to be improved by the structure of its processes, because its committee hearings may not even elicit relevant information, and in any event the information elicited will not substantially constrain the facts Congress finds in its committee reports. The structure of committee hearings is not well-designed to gather the best information representing a perspective. (116) First, the committee in general and its chairman in particular control the witness list, and thus only witnesses from the perspectives chosen by the committee will be heard. (117) It is true that both the majority and the minority in practice can call witnesses, even if the majority calls the lion's share. But this does not assure that contending perspectives on legislation will necessarily be heard. The majority and minority may essentially agree on legislative objectives when popular opinion is on one side of an issue. Even when popular opinion is divided, the committee may follow the views of interest groups united in favor of the legislation.

In any event, no matter who testifies, the majority members of the committee can determine which witnesses to credit. Thus, the committee report will reflect the determinations of its majority members, and it is the findings in that report on which the legislation will be based to which the judiciary will defer. (118)

b. Incentives of the Public Under a Public Choice Analysis

On many, if not most, issues the general public will exercise little influence on congressional fact-finding and thus pose no constraint on congressional inaccuracy. Citizens have very little knowledge of legislation in general, let alone more technical political information like the content of a factual record compiled by Congress. The ignorance of citizens in this regard is "rational" and thus intractable. First, since the collective decisions of legislators will generally have less effect on the individual citizen than will his own private decisions, he will rationally invest his scarce time and resources in gathering information about his private enterprises rather than information about the common enterprise of government. (119) Second, since any citizen is unlikely in the extreme to provide the decisive vote in any legislative election, he will stint on acquiring information about public policy for this reason as well. (120) Studies confirm that citizens have very low knowledge about the basic facts of government--let alone the specific factual claims by which Congress supports specific legislation. (121))

Second, insofar as the public exerts influence on congressional fact-finding, it is likely to undermine Congress' accuracy. The public is "rationally irrational" as well as "rationally ignorant," and thus believes many things that are not true when those beliefs coincide with substantive preferences and biases. (122) Citizens are rationally irrational for reasons of economic logic similar to those that encourage rational ignorance. Citizens will prefer to believe factual claims that cohere with their preferences or otherwise make them feel happy or virtuous, regardless of truth or falsity, unless falsity exposes them to harm. (123)

As a result, individuals face a truth versus falsity calculus in their beliefs about their private life that is very different from that about public policy. For instance, believing that a poisonous household substance is harmless may be very costly indeed. But holding a false belief about a public rather than private matter may impose negligible costs. The chance that one's false belief will be decisive in influencing the public policy at issue is no greater than the chance that one's vote will decisive. (124) And only if the false belief is decisive will the individual's adherence to the belief bring about the public policy and thus inflict harm on the individual. As a result the public has little incentive to embrace true factual beliefs in public policy, and in areas as diverse as economics and toxicology it has been shown that the public's factual beliefs about public policy are distorted by biases and in many respects are simply mistaken. (125) The public's factual beliefs are sometimes forms of self-expression rather than products of factual inquiry. (126)

Thus, it seems likely that where the public has policy preferences, they will be biased to believe in facts congenial to those policies, and expect their legislature to find these congenial facts, including those facts that would permit those policies to be upheld by the Supreme Court. As a result, when an issue becomes so salient as to attract the public's attention, citizens will not only fail to be a force for accuracy in legislative fact-finding, they may well become catalysts of inaccuracy.

Take an issue as politically salient as partial-birth abortion. Large majorities believe that partial-birth abortion should be prohibited. (127) Given "rational irrationality" it seems likely that these same majorities would be primed to expect that partial-birth abortion is never medically necessary, because that factual belief would most easily cohere with their preferences. As a result, insofar as they did pay attention to congressional fact-finding on this matter, most would expect Congress to find abortions never medically necessary, thus creating some pressure for such factual findings regardless of truth.

2. Behavorial Law and Economics Analysis

In this section we consider a model of congressional behavior in which the motives of members of Congress are more mixed. While members of Congress retain an interest in shaping the fact-finding that enhances their electoral prospects, they also have an interest in the accuracy of their fact-finding. Nevertheless, because of heuristics and biases that members of Congress share with people in general, congressional fact-finding will still have systematic defects. If our first model relies on the latest insights of public choice, our second draws on recent findings from behavioral economics. Neither model captures the whole truth about congressional behavior in legislative fact-finding, but the similarity of their conclusions provides far more powerful evidence about congressional fact-finding's reliability than a model based on a single view of the world.

Many observers believe that members of Congress care about matters other than reelection, such as ideology, the conscientious carrying out of their duties, and the respect of their colleagues. (128) Some of these objectives, such as conscientiousness, should translate into concern about the accuracy of fact-finding. Moreover, at least in some instances a long-term self-interest of members of Congress coincides with finding facts accurately, because in the long run the way the world actually works may change the preferences of citizens. (129) Even under these more generous assumptions, Congress has important systemic defects as a fact-finder.

One problem arises from the selective flow of information to Congress. It is widely recognized that interest groups, both economically and ideologically based, are the most effective conduits of information to the legislature. (130) There are an infinite number of facts Congress could be aware of, but even a member of Congress who wants to know as many facts as possible is limited by the demands of his office. (131) Therefore it requires the coordination and investment of interest groups to bring facts to a member's attention, particularly if they know that a statute's constitutionality might be based on favorable facts in the congressional record. (132) Thus, even the conscientious member of Congress is in some sense a prisoner of the information he receives. In the unstructured legislative fact-finding process, interest groups will shape the information stream that surrounds him.

Second, even when a member of Congress bases his fact-finding on more general sources of information not directly manufactured by special interests, like the facts contained in newspapers or on television, he faces the bias of the "availability heuristic." The availability heuristic shows "individuals will overestimate the likelihood of events that are more cognitively salient, or available to them, ignoring or undervaluing the relevant base rates of outcomes." (133) This heuristic is well documented in numerous psychology experiments. (134) For instance, when trying to establish the frequency of a kind of event or in weighing the cost and benefits of proposed legislation, a legislator will have a tendency to put disproportionate weight on the facts that are readily accessible or salient. Particularly given the expanding number of complex policy issues with which Congress deals, (135) even a well-meaning member will also be prompted to take this kind of mental shortcut.

The availability heuristic of course also exists in ordinary citizens and thus can affect Congress indirectly as well as directly, since in this mixed model Congress responds in part to citizens' views of the facts. (136) Voters often respond to cognitively salient news items that are inflated by short bursts of intense media coverage. (137) For example, a federal carjacking law was enacted after a spurt of carjackings received national attention, despite the fact that carjackings were already being prosecuted at the state level and the rate of carjackings was on the decline. (138) This is quite probably an instance in which the availability heuristic was at work, since carjackings on the news are much more available and vivid than lectures on the benefits of local crime being prosecuted at the local level in a federal system. (139) To take an example from Carhart, citizens could well believe that one particularly gruesome example of a partial-birth abortion that appeared medically unnecessary was representative of all partial-birth abortions.

Another important cognitive tendency affecting congressional fact-finding is the confirmation bias. When a person has an initial hypothesis, he will be "prone to search for information that confirms the hypothesis, as well as to interpret information that he has as confirming the hypothesis." (140) On the assumption that members of Congress are generally motivated by ideology as well as an interest in reelection, they possess a strong set of beliefs about the way the world works. In fact, in a party system in particular, a strong set of priors is needed to get elected. (141) Congress' committee system might encourage specialization, (142) but it also exacerbates the confirmation bias. (143) Members of Congress will bid for assignments on committees which govern policy they have particularly strong beliefs about, or that affect special interests influential in a congressman's home state. (144) Therefore the factual findings made at the committee level will be particularly subject to the biases of members of Congress. (145)

Accordingly, even when members of Congress focus on objectives other than reelection in their fact-finding, their psychology interacts with the legislative fact-finding process in a way that undermines its reliability.

C. SOCIAL FACT-FINDING BY THE JUDICIARY

Of course it is not enough to show that Congress is an unreliable fact-finder on constitutional issues. We must also compare the reliability of judicial fact-finding with that of Congress. In this section, we show first that the structure of the judiciary systematically reduces the problems that beset congressional fact-finding under either a public choice or behavioral economics model. Second, we respond to arguments that the judiciary has its own inherent defects that may nonetheless make it a worse finder of social facts than is Congress.

Members of the judiciary have life tenure and therefore do not share the incentives of members of Congress to shape the relevant facts to enhance their electoral prospects. Moreover, judges' indifference to electoral pressures will make the judiciary less concerned about public opinion. As a result, the public's biases about facts--which, as we have shown above, may in some cases distort congressional fact-finding--are unlikely to have much effect. Accordingly, the pivotal influences in the public choice model that undermine the accuracy of fact-finding will be less present in the judicial than in the legislative context.

Similarly, the defects of the fact-finding process underscored by our mixed model do not beset judicial fact-finding. Here the adversarial and disciplined nature of the litigation process rather than life tenure is responsible for superior judicial performance. First, interest groups will wield less influence with the judiciary. (146) Although interest groups do influence litigation through test cases and amicus briefs, the adversarial system will produce a more balanced view of the facts than the political processes of Congress. In litigation, each side has the incentive to get its facts into the record and discredit the facts put forward by opposing counsel. A congressman is usually provided only with facts that affect well coordinated interests groups; in litigation, however, every relevant fact is at least marginally more beneficial to one side of the dispute rather than the other. (147) Concerns that would not be raised in the legislature--because they involve facts of which voters are rationally ignorant or interests that are diffusely spread across the population--would be ferreted out by the parties.

We recognize that interest groups might have more resources at their disposal in litigation as well, but they cannot effectively out-lobby the opposition as they can in the legislative process. (148) An interest group can hire a top law firm only once, and even though they may outspend the other side, they face the same page limit on their briefs and the same amount of time for oral argument. (149) In economic terms, the marginal effectiveness of additional expenditures by interest groups is lower with the judiciary than with the legislature, because courts operate under tighter constraints on their decision-making process and with more barriers to interest group influence. (150)

The adversarial system also constrains the availability heuristic and confirmation bias and makes such cognitive biases less problematic than they are in Congress. The adversarial system assures that evidence on both sides of a disputed factual issue will be heard, thus helping to diminish the power of the availability heuristic. Because of the structure of litigation, judges will also be less subject to the confirmation bias than legislators, as they cannot seek out the facts themselves, but must hear the facts presented by both sides before deciding an issue. Furthermore, judges are less prone to the confirmation bias because they are forced to address contrary facts in their opinion and explain why they are not relevant. If a judge ignores facts in the majority opinion, he will suffer the embarrassment of a strong dissenter (a factual "whistleblower") who points out an opinion's factual flaws. (151) As many commentators have observed, concern about reputation and status among peers is a factor likely to affect judicial behavior. (152) Peer pressure and professional norms are more powerful in the judiciary than in Congress, because judges make factual determinations in smaller groups and, as elite lawyers, share a greater core of professional experience. (153)

Against these substantial advantages in judicial fact-finding, commentators have suggested the judiciary is liable to several disadvantages. The most important argument is that the judiciary's insularity and lack of geographic and socioeconomic diversity makes its social fact-finding less accurate than a more geographically dispersed and representative body like Congress. (154) This kind of argument against a judicial finding of social facts faces several difficulties.

The argument that the greater diversity of experience and background of members of Congress make them a fact-finder superior to the judiciary tracks the familiar argument that the greater diversity of experience and background of members of the jury make them a fact-finder superior to the judiciary. (155) But the jury is generally charged with finding adjudicative facts, whereas the legislature must find more general social facts. Adjudicative facts more often call for the application of the common sense of the community, as when a jury must infer individual intent from a welter of facts about an event or series of occurrences or must evaluate the credibility of witnesses from all walks of life. (156) For such evaluations, it is quite clear that diversity in experience and background will prove very useful.

In contrast, social facts of the kind that support the constitutionality of legislation more often comprehend impersonal and general data where inferences depend on techniques of objective analysis, including statistical inference, rather than evaluations of personal credibility. Here it is much less clear that diversity of experience and background are more useful than training in the specific experience of evaluating this kind of evidence. (157) Moreover, the analysis needed to find social facts is becoming more complex and sophisticated. The exponentially increasing power of computers permits ever more comprehensive data collection and ever more intricate dissection of these data by mathematical techniques, like regression analysis. (158) Even if Congress were thought at some point to have had a comparative advantage akin to the jury in social fact-finding, the claim is less and less plausible in the modern era. (159)

In any event, the central question about judicial social fact-finding cannot simply be whether the relative insularity and socioeconomically uniform character of the judiciary impedes its fact-finding, but whether it does so to a greater degree than in ordinary legal interpretation. After all, it is frequently noted that the judiciary's composition affects its interpretation of legal texts. Indeed, this criticism of judicial review has been heard since the early republic and remains a frequent staple of those who want to cast doubt on the wisdom of a particular judicial decision. (160) Yet, despite these complaints, our system provides for rigorous judicial review, and permits the Supreme Court to invalidate legislation of a branch with greater geographic and socioeconomic diversity. If the judiciary, despite its insularity and uniformity, reviews Congress's reading of the Constitution's text in order to protect the Constitution's structure or the rights of minorities (or for whatever other reason judicial review is justified), one must show why a similar review of the social facts supporting the constitutionality of legislation is not similarly necessary to achieve those same goals.

Indeed, social fact-finding would appear to be less influenced by the characteristic judicial biases of insularity and socioeconomic uniformity than legal interpretation. For instance, fundamental rights jurisprudence permits the judiciary substantial discretion to determine what rights should be protected. (161) The Court claims it looks at traditions in making those determinations, but such assessments involve large measures of discretion and thus can be seen as subject to bias. When socioeconomic elites favored laissez-faire, the Supreme Court ruled in favor of economic liberties, (162) and now when elites favor sexual autonomy, the Court rules in favor of rights to abortion and same-sex conduct. (163) Compared to these kinds of determinations, the evaluation of data necessary to establish social facts appears to involve less, not more discretion, and thus is less subject to the biases of greater social uniformity. (164) Thus, when such factual determinations are as necessary to evaluating the questions of a statute's constitutionality as are other legal issues, the case for an independent judicial role is no more weakened by the judiciary's insularity and socioeconomic uniformity than are other aspects of judicial review.

Other typical attacks on the impartiality of the judiciary are similarly less persuasive in the area of social fact-finding. For example, some might argue that the Supreme Court has become so "politicized" that the Justices cannot be relied on to be any less biased than Congress when it comes to finding facts. But any real or perceived political biases manifest themselves as much in traditional legal interpretation. For instance, many commentators argued that the Court's change in composition between the Stenberg decision, which struck down Nebraska's ban on partial-birth abortion, and the Carhart decision which upheld Congress's ban, drove the difference in result. (165) Even if this is the case, it does not appear to be the result of biased fact-finding. In Carhart both the majority and the dissenters seemed to agree that some of the Congress's factual finding were incorrect. (166) The majority did not take the position that partial-birth abortions were never medically necessary, but rather found that the answer to the question was uncertain. (167) The majority concluded that Congress should be accorded deference on a matter of factual uncertainty, while the dissenters argued that this uncertainty should weigh in favor of protecting the health of the mother, which was the majority's view in Stenberg. (168) Thus it was the Justices' views on the legal standard to be applied in reviewing Congress, rather than their interpretation of the statistical evidence, that resulted in the departure from Stenberg. Even if one thinks that the majority opinion was result oriented, the Justices in the majority were more intellectually honest than Congress about the legislative facts.

Other scholars have focused less on the biases of the judiciary and more on the judicial process itself as a constraint on sound decision-making. For instance, it has been noted that the judicial process has trouble capturing the multiple perspectives that best map reality, because in some cases there may be more plausible positions than there are litigants, and yet the adversarial process highlights only the positions of the plaintiff and defendant. (169) But, as we noted above, the committees of Congress function as gatekeepers that may often neglect or limit the exposition of a salient contending position. In any event, the judiciary permits amicus briefs to represent positions different from those advanced by the plaintiff or defendant. (170) In important constitutional litigation, the filing of amicus briefs has been on the rise. (171)

It has also been argued the adversarial process encourages courts to hear only the facts that are relevant to the parties at hand, leaving them under-informed about "the systematic and prospective consequences of their decisions." (172) But this assumes that courts ignore the consequences of their decisions. It is questionable whether this is the case, particularly in constitutional litigation. When a constitutional doctrine requires the Court to find and weigh social facts, such as whether partial-birth abortion is ever necessary, the doctrine itself factors in the consequences of the Court's decision. (173) Moreover, important long term consequences of constitutional decisions, even if not immediately connected to the parties, will support one side of the dispute, and thus are likely to be raised in argument.

D. DEPARTMENTALISM AND CONGRESSIONAL FACT-FINDING

In the previous two sections, we have argued that judicial finding of social facts is likely to be better than legislative fact-finding. The evidence we have adduced also supports the weaker claim that the judicial finding of social facts has peculiar strengths that legislative fact-finding lacks. Here we show that departmentalism turns even this weaker claim into a compelling argument that the Constitution itself contemplates independent, de novo judicial finding of social facts.

Departmentalism is the view that each branch of the government has coordinate responsibilities to assess the constitutionality of legislation. Accordingly, a member of the legislature should determine whether the legislation for which he votes is constitutional. (174) Likewise, the President should decide whether legislation is constitutional in deciding whether to affix his signature to a bill or to veto it. (175) In this regard, judicial review is not a special power vouchsafed by the United States Constitution to the judiciary, but, as Marbury v. Madison itself suggests, a preexisting requirement for judges to determine which is the controlling law in the context of a case. (176) The introduction of the Constitution simply means that the Constitution now provides the body of law controlling over any statute, and Congress and the President as well as the judiciary thus measure legislation against the Constitution in the course of their duties. (177)

In the past decade, departmentalism has gained substantial acceptance among academics. (178) The reasons are both formal and pragmatic. Formally, no express clause of the Constitution singles out one branch or the other for exclusive responsibility of constitutional assessment. Indeed, members of all branches take an oath to uphold the Constitution. (179) Moreover, the Framers themselves appeared to believe that the three branches had coordinate duties in constitutional interpretation. (180) These arguments apply to judicial review in social fact-finding: if social facts are relevant to constitutionality, nothing in the Constitution deprives the judiciary of the obligation to treat the Constitution as paramount by ascertaining for itself that relevant facts support the constitutionality of the legislation. (181)

Departmentalism also has some pragmatic advantages. Because each branch should assess for itself whether legislation is constitutional, the rights of the people have multiple protections. Moreover, the disagreement among the branches raises the salience of the issue to the people and helps bring the debate over constitutional principle to their attention. As Justice James Wilson stated, "[t]here is not in the whole science of politicks a more solid or a more important maxim than this--that of all governments, those are the best, which, by the natural effect of their constitutions, are frequently renewed or drawn back to their first principles." (182) In this way, departmentalism in fact-finding would be an additional security for rights and structures and provide a mechanism for involving the people themselves in constitutional discourse about relevant social facts when the branches disagree. (183)

Departmentalism possesses these advantages even when all the branches have similar institutional structures for constitutional assessment. But if the branches have different capacities and thus different strengths and weaknesses, departmentalism has the additional advantage of bringing to bear different perspectives that would be lost in a unitary system of interpretation. Some commentators have noted that the political branches have complementary capacities to the judicial branch in legal interpretation. (184) For instance, the judiciary excels in formal reasoning, whereas the political branches with their greater popular input may be more sensitive to analogous reasoning. (185)

Whatever one's view of whether the legislature or the federal judiciary is the better finder of social fact, it is clear from the discussion above that the judiciary at least possesses certain advantages. Its insulation and disciplined adversarial process focus a searchlight along wavelengths that Congress or state legislators cannot hope to replicate. Its distinctive strengths provide yet another reason that the departmental perspective makes as much sense in fact-finding as in pure legal interpretation. (186)

Departmentalism also shows why a legislature's decision expressly to find social facts in support of a statute's constitutionality should not displace the judiciary's independent role any more than the judiciary's role would be displaced by Congress's legal interpretations. Congress is to be commended if its members offer a serious defense of their legal interpretation of the Constitution to show that their legislation is constitutional. In subsequent litigation, the government would surely draw such legislative opinions to the judiciary's attention so the judges could adopt any of the legislature's legal arguments that they believed sound. In this context, few would suggest that a legislative legal opinion should diminish the judiciary's independent legal interpretive role. Similarly, Congress's decision to find facts uncovers a quarry from which the government may draw the facts necessary to construct a defense of the statute, but hardly suggests the judiciary should take its own fact-finding responsibilities less seriously.

Another structural reason that the judiciary should engage in de novo review is that it is the branch to which the other branches can most easily respond if it systematically abuses its power or makes inaccurate findings. Article III of the Constitution provides the Supreme Court with appellate jurisdiction "both as to Law and to Fact" but permits Congress to make "exceptions" to that jurisdiction. (187) Congress can make its displeasure known by using the exceptions clause to strip the Supreme Court of its jurisdiction to find facts. (188) It also has the power of the purse which it can use to cut off funds for desired judicial projects. (189)

In contrast, the Court has no way to check errant congressional fact-finding other than to engage in its own fact-finding to see if Congress is getting the facts right. Deferential review will not accommodate this goal as Congress would have substantial discretion to be inaccurate so long as it was not patently or grossly so. Thus, the potential error correction implicit in the Constitution's system of checks and balances also supports independent judicial review, because in the long run there is a lesser risk of unchecked error in independent review than in judicial deference. Thus, overall accountability will be strengthened by de novo judicial review of social fact-finding. (190)

E. AGAINST VARYING THE TREATMENT OF CONGRESSIONAL FACT-FINDING

A few commentators suggest that the Court should vary its deference to congressional fact-finding depending either on the expertise of Congress in an area of law or on the incentives it has to get the facts right. An example of the first kind of argument comes from recent commentary on the Court's 2006-2007 term, where Professor Kermit Roosevelt complained in passing that the Court deferred to Congress in the partial-birth abortion case where Congress had no medical expertise but did not defer to Congress on campaign finance reform where it has political expertise. (191) Professor Neal Devins has argued that the Court should defer to Congress in certain separation of powers cases where Congress has incentives to get the facts rights because it does not face pressures for inaccuracy from special interest groups. (192)

We are skeptical of such arguments. First, varying the level of deference depending on a list of factors complicates constitutional doctrine and is an invitation to ad hoc, result-oriented jurisprudence. Second, as we have noted, under a departmentalist view of the Constitution the virtues of one branch's mode of constitutional interpretation do not suggest that the other branches should abandon their own modes if those have distinctive virtues. (193) But most importantly, the models of congressional behavior that we have offered show that the distinctions between subject matter areas have little basis.

First, the institutional structures on which Congress relies, such as committee hearings, do not depend on the experience of individual members of Congress. Committee hearings can take evidence from experts in medicine and campaign finance alike. It is true that members of Congress will have more experience in the area of campaign finance than in medicine, because all have run for election and few are doctors. But greater familiarity with an issue is as likely to lead to worse, not better, accuracy in fact-finding. First, it exacerbates the availability heuristic: members of Congress will give greater weight to experiences that have happened to them, finding facts by personal anecdote rather than objective evidence. (194) Second, greater personal experience tends to translate into greater personal stakes and the potential for greater bias. Campaign finance is the best possible example of this: members of Congress as incumbents have important stakes in the outcome and obvious biases in finding facts that support legislative initiatives to protect Incumbency. (195)

Professor Devins' arguments about Congress' incentives in separation of powers cases are more sophisticated but they also do not persuade that judges should sometimes defer to congressional fact-finding. Professor Devins's prime example is Congress's line item veto. (196) It is not entirely clear what facts Congress found to serve as a predicate for the constitutionality of that statute, but Professor Devins argues that Congress had strong incentives to get the facts rights at least in the long run, because they would have institutional incentives to react to the President's use of the line item statute and modify the legislation in response. (197)

One possibility is that this is again not an argument about deference to fact-finding, but about deference due in a particular area of law. Given that the institutional clash between the President and Congress in the separation of powers area may lead to a plausible equilibrium, perhaps the Court should be loath to intervene on the facts or the law. Like those relying on the political safeguards to argue against intervention in federalism cases, other commentators have advanced this argument. (198) But if the view is that there something distinctive about congressional fact-finding in the separation of powers that call for deference, our analysis would suggest otherwise.

For instance, even if no special interest is involved in the line-item veto, the public's mistaken view of federal spending constrains congressional action. For instance, the public believes that foreign aid is twenty-four percent of the budget when it is in reality only one percent. (199) Entitlement spending in fact makes up the bulk of the federal budget, but it would not have been subject to the line item veto. (200) Such mistaken views of the social facts may make the public incorrectly perceive the balance between the utility of the line item veto in cutting spending and the additional power it gives to the President to pressure members of Congress on spending and nonspending matters. Similarly, the availability heuristic gives the public a mistaken idea about the importance of pork barrel spending. The "bridge to nowhere" attracted more attention than sober analyses of the need to control entitlements. (201) These kinds of influences on the congressional fact-finding process will undermine accuracy and seem unlikely to change over time.

As we have shown in this section, there are a number of powerful arguments for an independent, de novo review of social or legislative facts in constitutional cases. We now make a preliminary inquiry into how the judiciary should go about finding those facts.

IV. HOW SHOULD THE COURT FIND LEGISLATIVE FACTS IN CONSTITUTIONAL CASES?

In his famous 1923 treatise on evidence, John Henry Wigmore wrote:

   When a legislative act is argued to be unconstitutional, and this
   is to depend upon the reasonableness, or the lack of possible
   reasonableness, of the law in its purpose or operation, and thus
   the external facts furnishing the possible legislative motive or
   the possible actual effect must be considered, this incidental
   question is not for the jury but for the court.... But by what
   theory or method shall the Court.... receive information of the
   alleged facts? That is an interesting inquiry, hitherto not
   carefully worked out by the courts. (202)

This "interesting inquiry" has still not been carefully worked out by the courts. (203) Indeed, as we discussed in Part I, the Supreme Court has been inconsistent in its finding of social facts and certainly has not set out a procedural framework for such judicial fact-finding. In this section, we offer such a framework, drawing on our analysis in previous sections and on established rules of judicial procedure. Given our view that the Court should find social facts in the same manner regardless of congressional fact-finding, this framework should apply whether or not Congress has expressly found facts.

Our framework is guided by two considerations. First, we seek procedures that maximize the advantages of judicial fact-finding discussed above--the transparency, discipline, and adversarial nature that contributes to the fairness of the process and the accuracy of the results. Second, we must nevertheless assure that these procedures take account of the reality that establishing social facts, unlike establishing adjudicative facts, creates precedents and affects many beyond those who are party to the case.

First, we consider the burden of proof in social fact-finding. We argue that where doctrine makes a social fact relevant to the constitutionality of the statute, the government always should shoulder the burden of proof. This requirement should apply even in cases where the constitutional standard of scrutiny is very lenient, such as those reviewing economic legislation. The government has access to facts and it is too difficult to require a plaintiff to prove a negative. But in cases with a lenient standard of review, that standard will allow the government to prevail with very little evidence.

Second, we move from the relation of the government to the judiciary in social fact-finding to the relation of the Supreme Court to the lower courts. Lower courts possess an advantage over the Supreme Court in fact-finding. They are well-suited to preside over the assembly of a factual record and the examination and cross examination of relevant witnesses. Accordingly, lower courts should identify factual matters necessary to resolving constitutional disputes and assure that its factual determinations on these matters are made only after giving the parties the opportunity to elicit all relevant evidence.

On the other hand, unlike the determination of adjudicatory facts, the determination of social facts potentially establishes broad rules for the future. Social facts represent social authority no less than legal interpretations that become precedent. As a result, first appellate courts and then the Supreme Court must be able to review the facts de novo to protect their authority to establish precedent vis-a-vis lower courts in the judicial hierarchy. Nevertheless, appellate courts should confine themselves to the records made below (as supplemented by remands) to provide fairness to the parties. Such a process ensures greater accuracy by eschewing evidence that has not been tested in an adversarial process.

Third, we end by considering the relation of the social fact-finding of a current Supreme Court and the social facts found in its past decisions. Because social facts constitute judgments establishing ongoing social authority, they can have the effect of precedent. But the world changes and as a result the social facts underlying constitutional judgment themselves change. As a general matter, we believe that parties challenging social facts previously embedded in Supreme Court judgments bear the burden of showing that the factual findings are no longer true before these social judgments can be upended.

A. BURDEN OF PROOF

The Court has not made clear the burden of proof with regard to social facts that are necessary to decide constitutional cases. Kenneth Culp Davis once dramatically illustrated the Court's inconsistency in assigning the burden of proof. In Paris Adult Theater v. Slaton, the Court put the burden of proof on the petitioner to disprove that pornography had adverse secondary effects. (204) The Court cited a minority report (that only two members of a nineteen member commission signed) to support the legislative finding that there was "a connection between antisocial behavior and obscene material." (205) In Carey v. Population Services International, the Court put the burden on the government to show that "limiting access to contraceptives will in fact substantially discourage early sexual behavior." (206) In its decision striking down the law, the Court cited six studies indicating that there was no deterrent effect. (207) As Professor Davis observed, Paris and Carey were decided without "relating the second case to the first, and without mention in either opinion of any case law about allocation of the burden." (208)

Even apart from such gross inconsistencies, the burden of proof for social facts is a constitutional crazy quilt. In strict scrutiny cases, the burden of proof is generally placed on the government. In some intermediate scrutiny cases involving gender the Court places the burden on the government. (209) But in First Amendment intermediate scrutiny cases the Court often takes the government at its word. (210) In rational basis cases, however, it is always clear that the burden is not on the government but the challenger of the law, (211) and the Court can even fabricate a rational basis on the government's behalf. (212)

The Court can avoid such inconsistencies in the future by placing the burden of proof for social fact-finding on the government as an explicit element of its constitutional doctrine. (213) It will prevent citizens from having the often insurmountable burden of proving a negative by showing that some government interest does not exist. It will provide greater certainty in law and procedure, and greater transparency in the Court's decision making. In combination with the other recommendations made in the section, it will help bring judiciary out of the "black box" that hides its weighing of legislative facts. (214)

The effect of the burden of proof will vary, depending on what the government is required to prove. For instance, in challenges to economic regulations, the Court has made clear that the right to due process means only that the legislature must have a 'rational basis' for its legislation. (215) Even if the burden of proof is on the government, its burden is only to show that there is a rational basis for the policy. Thus, the government can meet this burden even in the face of very substantial and even predominant evidence that this legislation will not meet its objective. Accordingly, challenges to this kind of legislation will generally be defeated in early stages as soon as the government gives any rational justification for the law. (216)

In cases requiring a more searching scrutiny, the government should show that the requisite factual predicate for the legislation is more likely than not to be the case. For instance, if keeping children from seeing indecency on the Internet is a compelling justification for imposing restrictions on First Amendment rights, the government must shoulder the burden of proving that those restrictions keep children from seeing indecency on the Internet. (217)

Accordingly, it is important to carefully describe the social fact that the constitutional doctrine at issue requires to be proven, because the burden of proof will have more or less bite depending on what is required to be proved. For instance, in Gonzales v. Carhart, the courts below appeared to believe the Court's previous decision on the same issue in Stenberg required the government to prove that partial-birth abortions were never medically necessary in order for the prohibition be constitutional. (218) But the Supreme Court required only that it be uncertain whether partial-birth abortion is ever medically necessary for the government to prevail. (219) Even if the government must shoulder the burden of proof in both instances, it is obviously much easier to prove that it is uncertain that partial-birth abortions are ever medically necessary than it is to prove that they are never medically necessary. Clarifying the burden of proof may thus prompt clarifications in doctrine by requiring the judiciary to focus more clearly on what social facts must be proved.

B. PROCEDURES FOR FINDING SOCIAL FACTS

The Supreme Court has never made clear the appropriate procedures for finding social facts in the federal judicial system. In this subsection, we offer a sketch of rules that would maximize the judiciary's advantages in finding social facts and yet recognize their distinctive role as social authority with precedential effect. To that end, the lower courts should identify the social facts necessary to resolving constitutional litigation, compile a record for establishing those facts, and make their factual determinations on that record. Appellate courts, including the Supreme Court, should then review de novo the factual determinations of the lower courts to preserve the precedential authority of these facts. But they should not go beyond the record that lower courts have established. Remands are always available for the creation of a record necessary to establish facts that the appellate courts deem necessary to resolve the constitutional dispute at issue.

1. The essential role of lower courts.

In our judicial system, trial courts take the lead in finding adjudicative facts, because they possess a greater capacity to identify the potentially many facts that are legally relevant, to hold evidentiary hearings to establish those facts, and to develop evidence and findings for appellate review. (220) Trial courts have similar advantages in finding social facts but only if they establish these facts by gathering evidence and not by judicial notice. (221)

Currently, Section 201 of the Federal Rules of Evidence sharply limits judicial notice of adjudicative facts to those "(1) generally known within the territorial jurisdiction of the trial court or (2) capable of accurate and ready determination by resort to sources whose accuracy cannot reasonably be questioned." (222) When the court takes judicial notice, a party must be given an opportunity to be heard "as to the propriety of taking judicial notice and the tenor of the matter noticed." (223) While the Advisory Committee explicitly excluded legislative facts from these rules, (224) there is no reason that the Supreme Court cannot use them as guidelines for when evidentiary hearings are required to resolve disputes about social facts relevant to litigation at hand. Accordingly, when a federal district court first entertains constitutional litigation, it should work with the parties to identify the salient social facts that are necessary to establish the constitutionality of the legislation being litigated. (225) If these facts can be reasonably contested, the court should hold evidentiary hearings and provide each party a right to call witnesses and be heard on the factual issue. It should take judicial notice only of social facts that cannot be reasonably contested. Even for social facts for which judicial notice is taken, parties should be given an opportunity to be heard as to the propriety of taking judicial notice.

Two phenomena of special relevance to evidence about social facts should be noted. While statistical evidence only comprises a portion of the objective evidence required for the analysis of social facts, it may comprise an increasing portion of that evidence. Because of ever more powerful computers, social scientists are able to gather more data and analyze it more deeply than ever before. (226) Courts should be able to take advantage of this trend, because adversaries on both sides of a salient factual question will have every incentive to produce the best experts and the latest statistical data. It is true that lower courts can make best use of such evidence only as they become sophisticated consumers of statistics, including the regression analyses that dissect social science evidence. But just as judges in antitrust case have become more sophisticated over time in dealing with complex economic arguments, (227) there is no reason that judges cannot attain a commensurate level of sophistication with statistics. The alternative is to find social facts implicitly through intuition and hunches, because as we have shown, in many cases social fact-finding is called for by constitutional doctrine. The latter course reduces the accuracy of social fact-finding and the legitimacy of constitutional decisions.

Second, the district court should not directly rely on testimony heard by Congress, but require that such evidence be introduced according to the usual judicial procedures. While it may seem duplicative for the same experts who testified in Congress to testify again in court about the same legislation, judicial procedures are different from those used in Congress, and, as discussed above, often make for greater accuracy. For instance, judicial procedures allow for cross examination, something that might have been useful in cases like U.S.R.R. Retirement Board v. Fritz, where Congress was essentially deceived by the Railroad Retirement Board's expert. (228) Requiring the application of judicial procedures and standards to evidence originally gathered in Congress will avoid 'on the record' review and become a hallmark of the independence of judicial fact-finding.

2. De Novo Review by Appellate Courts

Appellate courts, including the Supreme Court, generally defer to lower courts' findings of adjudicative fact in all constitutional cases except defamation cases covered by the First Amendment. (229) Appellate courts, however, generally and the Supreme Court in particular have been less express about their stance toward what we have called the social fact-finding of lower courts. They generally do not defer but sometimes suggest that such findings deserve deference. (230)

While appellate courts should defer to findings of adjudicative facts by lower courts, they should review social facts de novo, because findings of social facts, unlike findings of adjudicative facts, create precedents for future cases no less than legal interpretations. (231) Thus, if social fact determinations by lower courts were insulated from full review, lower courts rather than appellate courts would have the final say in establishing the law in a wide range of cases. Moreover, the evidence probative of social facts more likely involves statistical and other general data rather than the kind of personal testimony for which determinations of credibility are essential. As a result, lower courts possess less of a comparative advantage from observation of witnesses.

Of course, facts do not come labeled "adjudicative" on the one hand or "social" or "legislative" on the other. Courts must look to the function those facts play in the legal system to classify them properly. If the facts at issue are specific to the parties in the case, the appellate courts should review them under the deferential standard applicable to adjudicative fact-finding. (232) If they govern factual issues of more general scope, such as the question of whether a specific protocol will safeguard children from indecency on the Internet or whether partial-birth abortions are ever medically necessary, they function as social authority. As a result, appellate courts should review such determinations de novo in the same way as they review without deference the legal interpretations of lower courts. (233) The opposite conclusion would lead to the absurd result that a single district court judge's findings of legislative fact would constrain the Supreme Court in a constitutional determination. (234)

We do have a model that helps to make clear that appellate review of a factual record made by others can work. The Administrative Procedure Act contemplates a process for the creation of agency rules through formal rather than informal review. (235) In formal review the agency assembles a factual record though trial type procedures, including the application of evidentiary rules and cross-examination. (236) The courts then review the whole record to see if the rule has factual support. (237) It is true that courts do so on a "substantial evidence standard," (238) but changing that standard to de novo review would not change the basic division of labor between the record creating and record reviewing institutions.

To be sure, creating a record and then subjecting it to de novo review will create substantially more work for courts. But the alternative to more labor is either more inaccuracy (if the lower courts do not compile a record with trial-type procedures) or more arbitrariness (if appellate courts do not engage in de novo review). The additional work does have another advantage: it will encourage the Supreme Court to think seriously about the extent it wants constitutional jurisprudence to depend on social facts rather than formal rules.

3. Remand for further factual findings

While appellate courts should review the record de novo in determining social facts, they should endeavor not to use evidence outside the record made by the lower courts. Relying on evidence outside the record is unfair to parties who have no chance to rebut the evidence that the appellate court makes dispositive. For the same reason, relying on evidence outside the record undermines the accuracy of the social fact-finding which it establishes. Thus, appellate courts' reliance on the factual record of their own creation erodes the distinctive advantages that the judiciary has vis-a-vis Congress in social fact-finding.

The straightforward way to avoid these costs is for the appellate court to remand the case to the lower court for consideration of a dispositive factual issue that the appellate court believes was missed or for reconsideration of an issue which, in the view of the appellate court, needs further evidentiary vetting.

Remands for further evidentiary hearings would not deprive appellate courts of their ultimate power to find the relevant social facts. Parties dissatisfied with social facts found by the court below would always be able to appeal the case again to the higher courts. Of course, remands for additional fact-finding will cause delay in resolving a case. But that is true for remands on any issue. Delay is simply the price of more accurate and fair fact-finding, and given that social facts constitute precedents no less than binding legal interpretations, the additional cost of delay to the particular parties in the case are likely to be outweighed by the advantages to society as a whole. Moreover, more frequent remands for social fact-finding will no doubt encourage parties and lower court judges to make a full record of all the evidence substantially relevant to dispositive facts for which the constitutional issue calls. Amici will also have incentives to intervene at the trial court stage rather than wait for a Supreme Court case, because this new framework will prevent them from prevailing in the appellate courts by making evidentiary claims in their briefs that their opponents have no opportunity to rebut. (239)

The Supreme Court in Gonzales v. Carhart should have considered remanding the case to the lower court when it was determined that the government could prevail simply by showing that there was uncertainty about whether partial-birth abortions were ever medically necessary. (240) The lower courts had focused on the related but distinct claim of whether partial-birth abortions were ever medically necessary. (241) The Supreme Court resolved the former issue in part by deferring to Congress's judgment and in part by assuming that there was genuine uncertainty because of conflicting evidence on whether abortions were ever medically necessary. (242) But, in our view, this was error. The Court should not defer to Congress's fact-finding, and it simply does not follow that there is genuine uncertainty about an issue because there is some conflicting evidence in the record. The trial court could have been asked to resolve whether there was in fact genuine uncertainty.

C. CHALLENGING SOCIAL FACTS PREVIOUSLY FOUND

Tying judicial decisions to express findings of social facts opens those decisions to challenge as the world changes and social scientific knowledge progresses. Both factual changes and changes in information about the facts can cast doubt on social facts previously found. Some might consider such potential challenges a large cost of greater candor about the factual underpinnings of constitutional holdings because those changes may expose deficiencies in previous precedent, thereby undermining the stability of the constitutional order. We believe, to the contrary, that when constitutional doctrine is dependent on social facts, precedent should be open to reconsideration when those facts are seen to change. Courts have the doctrinal resources to prevent such reconsideration from being taken lightly or from undermining important social interests in reliance and stability.

First, the alternative to reconsidering facts is permitting constitutional law to be built on falsehoods. This has very substantial costs in and of itself. Policies built on false factual premises tend not to work well, and as the falsity of those premises becomes widely known, those decisions tend to discredit the institutions that adhere to them. A jurisprudence that fails to take account of changing facts is likely particularly damaging to an institution, such as the judiciary, whose power depends in large measure on its reasoning. It is subject to the devastating response that John Maynard Keynes made to complaints about his changing his mind: "When the facts change, I change my opinions. Pray, sir, what do you do.?" (243)

Second, a measure of stability can be maintained by placing the burden of proof on the party that wishes to challenge the social facts previously found in precedents. The Court itself has recognized that its precedent is open to challenge if one of the parties can show the relevant facts have changed. In Planned Parenthood of Southeastern Pennsylvania v. Casey, the Court said that "West Coast Hotel [v. Parrish, 300 U.S. 379 (1937)] and Brown each rested on facts, or an understanding of facts, changed from those which furnished the claimed justifications for the earlier constitutional resolutions." (244) Although the Court's choice of West Coast Hotel as an example was somewhat problematic, (245) the general point was clear: while the burden of proof was on those who wanted to challenge the factual underpinning of those precedents, they could be challenged in an appropriate case.

Finally, the Court always retains the option to retain a precedent whose factual underpinnings are discredited, so long as it can provide other reasons for its retention. For instance, reliance interests in some kinds of cases may overwhelm the concerns about changing circumstances: it may be more important that the rule is irrevocably settled then that it be settled in accordance with the current facts. (246) But our framework for social fact-finding will force greater judicial candor about the reasons for retaining precedent. As we have stressed throughout this article, express attention to social fact-finding and its place in constitutional doctrine will make for analytically cleaner judicial reasoning and doctrine.

CONCLUSION

Independent de novo review of the social facts necessary to determine the constitutionality of legislation is the judicial stance that best coheres with the rest of the law. The judiciary takes an independent, de novo view of the law, regardless of whether Congress informally or formally offers its own opinion on a statute's constitutionality. We have shown that there are no analytical or functional differences between law and fact that militate against the judiciary undertaking a similarly independent view of social facts. Indeed, it is arguable that because of its life tenure and adversarial process the federal judiciary has even greater comparative advantages in the accuracy of its social fact-finding than in the accuracy of its legal interpretations. And even if the judiciary had only distinctive strengths in social fact-finding, the departmental structure of the Constitution would argue for an independent role in this regard no less than it does for an independent executive and legislative role in legal interpretation.

Independent de novo review of social facts also is a simplifying doctrine that promotes the rule of law. We showed that claims that the judiciary should defer to congressional fact-finding in a particular area are ad hoc and often result-oriented. Independent review, in contrast, is a principled stance applicable across the full range of law. To be sure, the evidence needed to support the constitutionality of legislation may vary depending on the area of the law in question. But the standard should be varied based on the legal doctrine that applies to the issue being decided, such as the relevant balancing test or level of scrutiny. Thus, if legislation in the economic area requires only a rational basis, it will take a relatively small amount of evidence to show that that policy is rational. But the standard of review will not be chosen because of some implausible claim about judicial fact-finding in economics but instead will derive from settled principles about the appropriate scope of judicial review for economic matters.

Finally, a consistent stance of independent review should help the judiciary improve its own findings of social fact, regardless of whether Congress has expressly found facts. The judiciary independently finds adjudicative facts and has a well-developed framework for doing so. That framework can be applied with a few adjustments to the finding of social facts. The lower courts should be responsible for creating a factual record through an adversarial process, and the appellate courts, including the Supreme Court, should be confined to that record (and any further record created on remand) in its fact-finding. This framework will greatly improve the accuracy of judicial finding of social facts.

Because of ever more powerful computers, social scientists are able to gather more data and analyze it more deeply than ever before. It is thus more important than ever before that the Court both take seriously its responsibility for assuring that legislation has whatever factual predicate is constitutionally required and shape its procedures so that that factual predicate can be best assessed.

(1.) 127 S. Ct. 1610 (2007).

(2.) 530 U.S. 914 (2000).

(3.) Carhart, 127 S. Ct. at 1624.

(4.) Id. at 1637-38.

(5.) Id. at 1636.

(6.) Id. at 1636-37.

(7.) Id. at 1637 (citing Crowell v. Benson, 285 U.S. 22 (1932)).

(8.) Crowell, 285 U.S. at 53-55.

(9.) Id. at 60-61.

(10.) Id. at 45-46, 64-65. The Court stated that "we review congressional factfinding under a deferential standard" but then quoted Crowell for the proposition that "[i]n cases brought to enforce constitutional rights, the judicial power of the United States necessarily extends to the independent determination of all questions, both of fact and law, necessary to the performance of th[e] supreme function [of enforcing constitutional rights]." Carhart, 127 S. Ct. at 1637 (quoting Crowell, 285 U.S. at 60).

(11.) While this is our principal concern, our recommendations for procedural rules for judicial finding of social facts are applicable to all judicial fact-finding, regardless of whether Congress has expressly found facts.

(12.) Kenneth Culp Davis, An Approach to Problems of Evidence in the Administrative Process, 55 HARV. L. REV. 364, 402 (1942).

(13.) David Faigman, Fact-Finding in Constitutional Cases, in HOW LAW KNOWS 162 (Austin Sarat et al. eds., 2007) (quoting KENNETH CULP DAVIS, ADMINISTRATIVE LAW TEXT [section] 7.03, at 160 (3rd ed. 1972)).

(14.) PAUL ROSEN, THE SUPREME COURT AND SOCIAL SCIENCE 53 (1972); see also Ann Woolhandler, Rethinking the Judicial Reception of Legislative Facts, 41 WAND. L. REV. 111, 114 (1988) ("The key difference between adjudicative and legislative facts is not the characteristics of particular versus general facts, but rather, evidence whose proof has a more established place and more predictable effect within a framework of established legal rules as distinct from evidence that is more manifestly designed to create the rules.").

(15.) Faigman, supra note 13, at 162.

(16.) See T. Alexander Aleinikoff, Constitutional Law in the Age of Balancing, 96 YALE L.J. 943 (1987).

(17.) Adarand Constructors, Inc. v. Pena, 515 U.S. 200, 236 (1995) (remanding to lower court to apply strict scrutiny to race-based classifications in federal statute).

(18.) Board of Ed. v. Mergens, 496 U.S. 226, 250-51 (1990) (declining to second guess congressional judgment that "secondary school students are mature enough and are likely to understand that a school does not endorse or support student speech that it merely permits on a nondiscriminatory basis").

(19.) Roper v. Simmons, 543 U.S. 551 (2005).

(20.) Williams v. Florida, 399 U.S. 78 (1970); Colgrove v. Battin, 413 U.S. 149 (1973).

(21.) Marbury v. Madison, 5 U.S. (1 Cranch.) 137, 177 (1803).

(22.) 347 U.S. 483, 495 (1954).

(23.) 347 U.S. 497 (1954).

(24.) The losing side of Brown took this position, arguing: "Whatever the empirical truth of the matter ... the courts were not the proper forums for determining these facts. These matters were more properly decided by state legislatures and educational authorities." DAVID L. FAIGMAN, LABORATORY OF JUSTICE: THE SUPREME COURT'S 200-YEAR STRUGGLE TO INTEGRATE SCIENCE AND THE LAW 189 (2004).

(25.) The Court's uncertainty on how it reviews Congress is exemplified by Justice Steven's question to Solicitor General Clement in the oral argument in Gonazales v. Carhart. Justice Stevens asked "is there a different standard of review of what the district court found as opposed to what Congress found?" Oral Argument for Gonzales v. Carhart at 11 (No. 05-380) (Nov. 8, 2006). The Solicitor General responded that "if you have situations, which you have in this case, where the district court heard some of the same witnesses who testified before Congress and before the district court, and the district court makes a different credibility finding than the Congress, I don't think that's a basis for the district court to overcome the contrary findings of Congress." Id.

(26.) 341 U.S. 494 (1951)

(27.) Id. at 547 (Frankfurter, concurring)

(28.) 389 U.S. 258 (1967)

(29.) Id. at 266n.16

(30.) Id. at 289

(31.) Id. at 285-86

(32.) See Richard A. Posner, Pragmatism Versus Purposivism in First Amendment Analysis, 54 STAN. L. REV. 737, 741 (2002) ("Hence, when the country feels very safe the Justices of the Supreme Court can without paying a large political cost plume themselves on their fearless devotion to freedom of speech"). For another example of the Court's inconsistency in the First Amendment area, see Archibald Cox, Foreward: Freedom of Expression in the Burger Court, 94 HARV. L. REV. 1, 31 (1980) (contrasting Landmark Communications v. Virginia, 435 U.S. 829, 843 (1978), which held that "deference to a legislative finding cannot limit judicial inquiry when First Amendment rights are at stake" with Columbia Broadcasting Sys., Inc. v. Democratic Nat'l Comm., 412 U.S. 94 (1973), which relied heavily on facts found by Congress in permitting a broadcast licensee to adopt a blanket rule refusing to sell time for spots on public issues)

(33.) Mergens. 496 U.S. 226 (1990)

(34.) Id. at 250-51

(35.) Id. at 251. Similarly, in United States v. Lee, 455 U.S. 252 (1982), the Court upheld the government's imposition of social security taxes against an Amish litigant who objected on religious grounds. The Court deferred to Congress' judgment that "mandatory participation is indispensable to the fiscal vitality of the social security system." Id at 258

(36.) 530 U.S. 428 (2000)

(37.) 384 U.S. 436 (1966)

(38.) Id. at 445-58

(39.) Id. at 467 ("It is impossible for us to foresee the potential alternatives for protecting the privilege which might be devised by Congress or the States in the exercise of their creative rule-making capacities. Therefore we cannot say that the Constitution necessarily requires adherence to any particular solution for the inherent compulsions of the interrogation process.... We encourage Congress and the States to continue their laudable search for increasingly effective ways of protecting the rights of the individual while promoting efficient enforcement of our criminal laws")

(40.) 18 U.S.C. [section] 3501. The standard set out by this statute was similar to the "totality of the circumstances" test the Court had created before Miranda to determine whether confessions were voluntarily made. Dickerson, 530 U.S. at 434, 442-43

(41.) See S. REP. NO. 90-1097 (1968) reprinted in 1968 U.S.C.C.A.N. 2112, at 2127-38

(42.) See Brief of Court-Appointed Amicus Curiae Urging Affirmance of the Judgment Below, at 26-27, Dickerson v. United States, 530 U.S. 428 (2000) (No. 99-5525) (citing Senate Report finding that "the rigid and inflexible requirements of the majority opinion in the Miranda case are ... extremely harmful to law enforcement."). Other briefs emphasized the congressional findings, and the deference these findings were owed under Supreme Court precedent. See Amicus Curaie Brief of the Bipartisan Legal Advisory Group of the United States House of Representatives in Support of Affirmance, at 19-20 Dickerson v. United States, 530 U.S. 428 (2000) (No. 99-5525) (stating that Congress' fact-findings were entitled to deference); Brief Amicus Curaie of Senators Orrin G. Hatch, et al., Urging Affirmance, at 13. Dickerson v. United States, 530 U.S. 428 (2000) (No. 99-5525) (stating that "according to congressional findings, the basis for Miranda's conclusive presumption is incorrect as an empirical matter")

(43.) The Court briefly touched on the factual issue, but not the findings themselves, by stating: "we agree with the amicus' contention that there are more remedies available for abusive police conduct than there were at the time Miranda was decided ... but we do not agree that these additional measures ... meet the constitutional minimum." Id. at 442

(44.) 530 U.S. at 438-41. The opinion focused on whether the Court's precedents indicated that the Miranda rule was "constitutionally based." Id. Dickerson is a good example of how the Court sometimes avoids factual questions by focusing on more traditionally "legal" authority, such as the Court's own precedent

(45.) We recognize that one might argue that the Court is simply changing its standard of legal review but cloaking this move under the guise of more or less deference to congressional fact-finding. A realist would not see much difference between the standard of legal review and the deference affording congressional findings. But if one believes as we do in the discipline of law, there is an advantage to candor and clarity in the Court's legal doctrine. A more consistent approach to fact-finding would require the Court in turn to acknowledge and justify the change in the standard of review that is really driving its opinion

(46.) See, e.g., Eldred v. Ashcroft, 537 U.S. 186, 204-07 (2000) (citing the congressional record as well as scholarly journals to conclude that Congress's extension of copyright protection to the author's life plus seventy years was "a rational exercise of the legislative authority conferred by the copyright clause"); Turner Broadcasting System, Inc v. FCC, 512 U.S. 622, 669 (1994) (Blackmun, J., concurring) (stating that in addition to congressional findings, "[t]he record before the District Court no doubt will benefit from any additional evidence the Government and the other parties now see fit to present")

(47.) City of Boerne v. Flores, 521 U.S. 507 (1997)

(48.) 528 U.S. 62, 91 (2000). The evidence Congress had compiled consisted of isolated sentences clipped from floor debates and legislative reports, as well as a 1966 report prepared by the state of California. Id. at 69

(49.) 531 U.S. 356 (2001)

(50.) Id. at 369

(51.) 538 U.S. 721 (20(13)

(52.) 541 U.S. 509 (2004)

(53.) Lane, 541 U.S. at 527 (noting a report before Congress showed that 76% of public services and programs housed in state owned building were inaccessible to and unusable by persons with disabilities); Hibbs, 538 U.S. at 730-31 (noting that the legislative record included a Bureau of Labor Statistics survey indicating that 37% of private sector employees had maternity leave policies, while only 18% were covered by paternity leave policies, and that Congress heard testimony that fathers faced discriminatory treatment with respect to child-care leave policies in both the private and public sectors)

(54.) Cf William W. Buzbee & Robert A. Schapiro, Legislative Record Review, 54 STAN. L. REV. 87, 136-43, 160-61 (2001); Ruth Colker & James J. Brudney, Dissing Congress, 100 MICH. L. REV. 80, 83-86, 144 (2001). See generally Senator Dewine's inquiries during Justice Roberts confirmation hearings regarding fact-finding, including questions such as: "In your opinion, what role should a judge play when reviewing congressional fact-findings? In your view, how much deference do congressional fact-findings deserve?" Available at http://www.washingtonpost.com/wp-dyn/content/article/ 2005/09/13/AR2005091301469.html; see also Senator Dewine's similar inquiries during Justice Alito's confirmation hearing. Available at http://www.washingtonpost.com/wpdyn/content/article/2006/01/10/ AR2006011001087.html

(55.) 473 U.S. 305 (1985).

(56.) Id. at 330 n.12 ("When Congress makes findings on essentially factual issues ..., these findings are of course entitled to a great deal of deference.").

(57.) Id. at 338 (O'Connor, J., concurring) (noting that "'[t]he merits of these claims are difficult to evaluate ... at the preliminary injunction stage."); id. at 356-57 (Brennan, J., dissenting) ("The Court rummages through the partially developed record and seizes upon scattered evidence introduced by the Government on the eve of the preliminary-injunction hearing--evidence that never has been tested in a trial on the merits--and pronounces that evidence 'reliable" and compelling.").

(58.) Id. at 334.

(59.) 539 U.S. 306, 330-32 (2000).

(60.) 518 U.S. 515 (1996).

(61.) Id. at 544-45 & n. 13-15.

(62.) 852 F. Supp. 471 (W.D. Va. 1994).

(63.) See United States v. Virginia, 518 U.S. at 585-86 (Scalia, J., dissenting).

(64.) 521 U.S. 844 (1997).

(65.) Id. at 849 n.2 (noting that the lower court's findings included "356 paragraphs of the parties' stipulation and 54 findings based on evidence received in open court").

(66.) See also United States v. Playboy Entertainment Group, Inc., 529 U.S. 803,818-22 (2000) (focusing on district court record on the question of whether "signal bleed" was a pervasive enough problem that Congress could constitutionally require cable operators to block or scramble such programming).

(67.) Reno, 521 U.S. at 876: see also id. at 858 n.24.

(68.) In our view, the best modern article about congressional fact-finding is Neal Devins, Congressional Factfinding and the Scope of Judicial Review: A Preliminary Analysis, 50 DUKE L.J. 1169 (2001). While we critique the basic conclusions of this article, see notes 200-201 and accompanying text, and believe that his comparison of the Court and Congress in social fact-finding is not sufficiently comprehensive, it begins to ask the right questions.

(69.) Roscoe Pound, Liberty of Contract, 18 YALE L.J. 454, 469 (1909) ("In the ordinary ease involving constitutionality, the court has no machinery for getting at the facts. It must decide on the basis of general knowledge and on accepted principles of uniform application.").

(70.) OLIVER WENDELL HOLMES, THE COMMON LAW 5 (Mark D. Howe ed., 1963). But as cases such as Lochner show, Holmes was also a realist about the role a judge could effectively play in society, and refused to look at factual matters when his authority did not mandate it. See Lochner v. New York, 198 U.S. 45, 65 (1905) (Holmes, J., dissenting).

(71.) Roscoe Pound, Mechanical Jurisprudence, 8 COLUM. L. REV. 605 (1908).

(72.) Louis Brandeis, The Living Law, 10 ILL. L. REV. 461, 467 (1916) ("[N]o law, written or unwritten, can be understood without a full knowledge of the facts out of which it arises, and to which it is to be applied.").

(73.) BENJAMIN CARDOZO, THE NATURE OF THE JUDICIAL PROCESS 113 (1921) (judges should "get [their] knowledge just as the legislator gets it, from experiences and study and reflection; in brief, from life itself"); Benjamin Cardozo, A Ministry of Justice, 35 HARV. L. REV. 113 (1921) (advocating a board that would consolidate all facts necessary for the judicial making process).

(74.) See Henry Wolf Bikle, Judicial Determination of Questions of Fact Affecting the Constitutional Validity of Legislative Action, 38 HARV. L. REV. 6, 12-13 (1924).

(75.) Id.

(76.) Id. at 19 ("It is clear that the legislative finding as to the fact upon which the validity of the legislation depends cannot be allowed to be binding upon the courts, since this would furnish a simple means of preventing judicial review of such legislation in this class of cases.").

(77.) Note, The Consideration of Facts in "Due Process" Cases, 30 COLUM. L. REV. 360, 362 (1930) ("The fact that an enactment has been passed by those who are elected to give expression to popular wishes and who have knowledge of local conditions is always a factor of importance in its favor. But the facts which are important for the consideration of the legislature are none the less material to a determination of the judicial question.").

(78.) "The deference owed to the legislature's conclusion should require that the facts on which a statute is declared invalid appear on the record. They should be accepted only when their accuracy is carefully checked, and the witnesses presenting them subject to cross examination." Id. at 367.

(79.) Note, The Presentation of Facts Underlying the Constitutionality of Statutes, 49 HARV. L. REV. 631, 633 (1936) ("Extensive legislative findings based upon detailed reports would, perhaps, be the most satisfactory method of establishing constitutional facts. The ability to see, hear and investigate actual conditions is undoubtedly possessed in a higher degree by legislative fact-finding bodies than by their judicial counterparts.").

(80.) Id. at 636.

(81.) See, e.g., Daniel J. Solove, The Darkest Domain: Deference, Judicial Review, and the Bill of Rights, 84 IOWA L. REV. 941, 958-59 (1999) (arguing that deference can undermine scrutiny and affect judicial outcomes): Lamprecht v. FCC, 958 F.2d 382-92 n.2 (D.C. Cir. 1992) ("If a legislature could make a statute constitutional simply by 'finding' that black is white, or freedom, slavery, judicial review could be an elaborate farce."); cf. Spencer Overton, Restraint and Responsibility: Judicial Review of Campaign Reform, 61 WASH. &, LEE L. REV. 663, 720 (2004).

(82.) See William D. Araiza, The Section 5 Power and the Rational Basis Standard of Equal Protection, 79 TUL. L. REV. 519, 543, 546, 557 (2005) (arguing that Congress' mandate from and similarity to the populace makes Congress better-suited to finding and acting on social fact, and that the Court's rational basis review illustrates the difficulty of second-guessing legislative findings except in cases of clear animus or irrationality).

(83.) See generally Joseph M. Pellicciotti, Redefining the Relationship Between the States and the Federal Government: A Focus on the Supreme Court's Expansion of the Principle of State Sovereign Immunity, 11 B.U. PUB. INT. L.J. 1 (2001).

(84.) Ruth Colker & James J. Brudney, Dissing Congress, 100 MICH. L. REV. 80, 118 (2001) (arguing that the Court's approach fails to appreciate the skill and sophistication that Congress brings to the information gathering process.); Robert C. Post & Reva B. Siegel, Protecting the Constitution from the People: Juricentric Restrictions on Section Five Power, 78 IND. L.J. 1, 14 (2003) ("Congress enacts Section 5 legislation in order to vindicate public understandings about the nation's needs and obligations under the Fourteenth Amendment, not to single out wrongdoers for blame and punishment."); Robert C. Post & Reva B. Siegel, Legislative Constitutionalism and Section Five Power: Policentric Interpretation of the Family and Medical Leave Act, 112 YALE L.J. 1943, 2025 (2003) ("[W]e vigorously contest the Court's premise that politics is a sphere so debased that it can only corrupt constitutional deliberation."); Christopher P. Banks, The Constitutional Politics of Interpreting Section 5 of the Fourteenth Amendment, 36 AKRON L. REV. 425, 469 (2003) ("the legislature has the institutional competence to study complex problems of policy through the open deliberative process of many (instead of the opinion writing proclivities of the few in closed chambers)").

(85.) Colker & Brudney, supra note 84, at 118-20, 140; Post & Siegel, Protecting the Constitution, supra note 84, at 16 ("Legislative hearing ought not to stimulate adversaries; their primary purpose is to inspire new forms of collective commitment.").

(86.) Colker & Brudney, supra note 84, at 118.

(87.) Dickerson v. United States, 530 U.S. 428 (2000).

(88.) Post & Siegel, Protecting the Constitution, supra note 84, at 22.

(89.) Herbert Wechsler, The Political Safeguards of Federalism: The Role of the States in the Composition and Selection of the National Government, 54 COLUM. L. REV. 543 (1954); JESSE CHOPER, JUDICIAL REVIEW AND THE NATIONAL POLITICAL PROCESS 176-81 (1980) (arguing that the structure of congressional committees and leadership system protects state interests); Larry Kramer, Putting the Politics Back into the Safeguards of Federalism, 100 COLUM. L. REV. 215, 278-79 (2000) (arguing that political parties will operate in Congress to protect state interests).

(90.) See, e.g., Choper, supra note 89, at 169-70, 314-15 (concluding that federalism issues should be treated like political questions).

(91.) See John O. McGinnis & Ilya Somin, Federalism vs. States' Rights: A Defense of Judicial Review in a Federal System, 99 NW. U. L. REV. 89 (2004) (arguing that judicial review is needed to protect federalist system).

(92.) We thus distinguish complaints that federalism cases are wrongly decided because Congress is in fact a superior fact-fact finder from those that suggest the Court has wrongly intervened by converting questions of fact into questions of law, and should take a deferential approach to both facts and law in federalism cases. This has been Justice Souter's critique of the Commerce Clause cases. See United States v. Morrison, 529 U.S. 598, 638 (2000) (Souter, J. dissenting) (faulting the majority for its "new characterization" of commerce clause doctrine which "suggest[s] that 'substantial effects' analysis is not a factual inquiry for Congress in the first instance with subsequent judicial review looking only to the rationality of the congressional conclusion, but one of a rather different sort, dependent on a uniquely judicial competence").

(93.) See, e.g., Harold J. Krent, Turning Congress into an Agency: The Propriety of Requiring Legislative Findings, 46 CASE W. RES. L. REV. 731, 737 (1996) (arguing that "requiring the legislature to make findings appears to denigrate the respect due a coordinate branch of government").

(94.) See Buzbee & Schapiro, supra note 54.

(95.) Id.; see also A. Christopher Bryant & Timothy J. Simeone, Remanding to Congress: The Supreme Court's New "On the Record" Constitutional Review of Statutes, 86 CORNELL L. REV. 328, 369-83 (2001) (arguing that on the record review is in tension with separation of powers because it involves improper scrutiny of Congress' s internal processes).

(96.) Buzbee & Schapiro, supra note 54, at 124-25.

(97.) Id. at 134-35.

(98.) Id. at 139.

(99.) Philip P. Frickey & Steven S. Smith, Judicial Review, The Congressional Process, and the Federalism Cases: An Interdisciplinary Critique, 111 YALE L.J. 1707, 1745 (2002) (When the court is expecting Congress to use rational decision-making based on evidence before it in federalism cases, "[i]t is demanding that the legislative process become something it is not and cannot be in a system of competitive parties operating through several institutions with shared policymaking responsibilities.").

(100.) In certain instances, such as equal protection and First Amendment cases, the Court looks to the "motive" of the legislature that passed the law in order to determine its constitutionality. See, e.g., United States v. Eichman, 496 U.S. 310 (1990) (invalidating congressional flag burning ban and noting that it was passed in response to the Court's Texas v. Johnson decision striking Texas's ban); United States Dep't. of Agriculture v. Moreno, 413 U.S. 528, 534 (1973) ("if the constitutional conception of 'equal protection of the laws' means anything, it must at the very least mean that a bare congressional desire to harm a politically unpopular group cannot constitute a legitimate governmental interest"). In some instances, the Court's analysis of the facts will be an implicit means of "smoking out" unconstitutional motives, See, e.g, Gil Seinfeld, The Possibility of Pretext Analysis in Commerce Clause Adjudication, 78 NOTRE DAME L. REV. 1251, 1324-27 (2003) (arguing that the absence of congressional findings can indicate a pretext): see also Ashutosh Bhagwat, Purpose Scrutiny in Constitutional Analysis, 85 CAL. L. REV. 297, 312-13 (1997). In cases in which the legislature's motive is relevant, the evidence Congress amasses to explain the need for its actions would bear on its motive. But this does not undermine our view that in social fact-finding the Court should not confine itself to the congressional record. First, if motive is the focus on constitutional doctrine, the question is not the objective truth of social facts, but what Congress believed were the social facts. The congressional record is directly probative of Congress's belief. Second, even in these circumstances, the record may not be the only evidence of Congress' motive.

(101.) RICHARD POSNER, OVERCOMING LAW 207 (1995) ("The interpretative question is ultimately a political, economic, or social one to which social science may have more to contribute than law.").

(102.) Richard A. Posner, Against Constitutional Theory, 73 N.Y.U.L. REV. 1, 11 (1998); see also Richard A. Posner, The Problematics of Moral and Legal Theory, 111 HARV. L. REV. 1637, 1701-02 (1998) (arguing that technical factual information is more easily ascertainable than esoteric legal or philosophical arguments).

(103). Ronald J. Allen & Michael S. Pardo, The Myth of the Law-Fact Distinction, 97 NW. U. L. REV. 1769 (2003).

(104). U.S. CONST. Art. II., sec. 2, cl. 3; see also Michael B. Rappaport, The Original Meaning of the Recess Appointment Clause, 52 UCLA L. REV. 1487 (2005) (describing how to discover the original meaning of the clause).

(105). Allan & Pardo, supra note 103, at 1770.

(106). One analytically sound distinction can be made between law creation and fact-finding, but this difference is only temporal--once law is created it is a social fact. Id. at 1804.

(107.) H.L.A. HART. THE CONCEPT OF LAW (1961).

(108.) For a good summary of these arguments, see Devins, supra note 68, at 1178-82; see also Buzbee & Schapiro, supra note 54, at 143 ("[Courts] are not well-suited to gather the evidence necessary to assess the magnitude of complex social practices, such as invidious discrimination against the disabled."): Louis Fisher, Constitutional Interpretation by Members of Congress, 63 N.C.L. REV. 707, 722 (1985).

(109.) Craig S. Lerner, Legislators as the "'American Criminal Class": Why Congress (Sometimes) Protects the Rights of Defendants, 2004 U. ILL. L. REV. 599, 608.

(110.) See generally MANCUR OLSEN, THE LOGIC OF COLLECTIVE ACTION (1965).

(111.) Devins, supra note 68, at 1182; see also Saul M. Pilchen, Politics v. the Cloister: Deciding When the Supreme Court Should Defer to Congressional Fact-finding Under the Post-Civil War Amendments, 59 NOTRE DAME L. REV. 337, 366-67 (1984) ("The tool of investigation may be wielded not to provide legislators with a sound basis for selecting one legislative program over another, but rather to mollify constituents' demands that some action be taken on particularly controversial issues.").

(112.) See John O. McGinnis, The Bar Against Challenges to Employment Consent Decrees: A Public Choice Perspective, 54 LA. L. REV. 1507, 1530-31 (1994) (discussing the way political actors try to raise information costs of their opponents by disguising their objectives).

(113.) Wendy M. Rogovin, The Politics of Facts: The Illusion of Certainty, 46 HASTINGS L.J. 1723, 1750 (1995) (arguing that a congressional hearing is a legal fiction; the facts are found by the members of Congress before the hearing takes place).

(114.) Carhart v. Gonzales, 127 S. Ct. at 1638. Justice Ginsburg's dissent gives a tour de force of Congress's false "findings." Id. at 1643-44.

(115.) 18 U.S.C.A. [section] 1531 History and Statutory Notes, Congressional Finding 8 (2005).

(116.) Julius Cohen, Hearing on a Bill: Legislative Folklore?, 37 MINN. L. REV. 34, 42 (1952) ("Why complicate unnecessarily the harassed life of the average Congressman and Senator by endeavoring to foist on him the duties for which his talents are not especially equipped, and for which he scarcely has any time, considering the tremendous demands upon his office. For the task of checking the immediate fact situation which a legislative proposal seeks to correct, something more is necessary than mere testimonials from witnesses with axes to grind.").

(117.) See Pilchen, supra note 111, at 367-68 (discussing committee chairpersons' power and arguing that "the structure and composition of specialized committees may lead to fact-finding that merely ratifies prejudices or other legislative proclivities").

(118.) We also recognize that at times Congress gets some of its information from administrative agencies and commissions, but administrative fact-finding is influenced by special interest capture in the agencies themselves. Even if agencies are not captured, they depend on Congress for their authority and budget, and tend to provide the facts Congress wants to hear. See Jeffrey J. Rachlinski & Cynthia R. Farina, Cognitive Psychology and Optimal Government Design, 87 CORNELL L. REV. 549, 569 (2002). Finally, on many important issues, like partial-birth abortions, agencies simply do not testify. See e.g. Partial-Birth Abortion Ban Act of 2003: Hearing on H.R. 760 Before the Subcomm. on the Const. of the H. Comm. on the Judiciary, 108th Cong. 6-35 (2003).

(119.) See John O. McGinnis, The Once and Future Property Based Vision of the First Amendment, 63 U. CHI. L. REV. 49, 125 (1996) (describing this motivation for indifference to public policy information).

(120.) See Frank H. Easterbrook, The State of Madison's Vision of the State: A Public Choice Perspective, 107 HARV. L. REV. 1328, 1331 (1994) (discussing this contribution to rational ignorance).

(121.) See, e.g., Ilya Somin, Political Ignorance and the Countermajoritarian Difficulty, 89 IOWA L. REV. 1287, 1290-91 (2004) (summarizing evidence and research on the subject). For the most thorough political science analysis of voter ignorance, see MICHAEL X. DELLI CARPINI & SCOTT KEETER, WHAT AMERICANS KNOW ABOUT POLITICS AND WHY IT MATTERS (1996) (documenting widespread voter ignorance and explaining the importance of political knowledge to the democratic process). See also Ilya Somin, Voter Ignorance and the Democratic Ideal, 12 CRITICAL REV. 413 (1998).

(122.) For a discussion of "rational irrationality," see BYRAN CAPLAN, THE MYTH OF THE RATIONAL VOTER: WHY DEMOCRACIES CHOOSE BAD POLICIES 17-18 (2007).

(123.) Id.

(124.) Id. at 131.

(125.) Id. at 51-92 (presenting evidence of public's mistaken views on economics) & id. at 160-62 (presenting evidence of the public's mistaken views on toxicology).

(126.) See Dan M. Kahan & Donald Braman, Cultural Cognition and Public Policy, 24 YALE L. & POL'Y REV. 149, 171 (2006) ("The phenomenon of cultural cognition refers to a series of interlocking social and psychological mechanisms that induce individuals to conform their factual beliefs about contested policies to their cultural evaluations of the activities subject to regulation.").

(127.) In 2003, 68% of adults nationwide favored a ban on partial-birth abortion in the last few months of pregnancy. See CNN/USA Today/Gallup Poll, Oct. 24-26, http://www.pollingreport.com/abortion.htm. When asked to consider the procedure over the last six months of pregnancy, 57% of adults nationwide supported a ban. See Los Angelos Times Poll, Jan. 30-Feb 2, 2003, http://www.pollingreport.com/abortion.htm. In one poll, only 41% of adults nationwide thought partial-birth abortion should be legal, even if it would prevent a serious threat to the woman's health. See ABC News Poll, July 16-20 2003. http://www.pollingreport.com/abortion.htm.

(128.) Edward L. Rubin, Public Choice, Phenomenology, and the Meaning of the Modern State: Keep the Bathwater, But Throw Out That Baby, 87 CORNELL L. REV. 309, 322 (2002) ("[I]deology, respect from colleagues, and the desire to act conscientiously have all been empirically confirmed as determinants of political behavior."); see also Lerner, supra note 109, at 629 (arguing that public choice theory is not incompatible with the idea of members of Congress envisioning themselves as members of the "public good").

(129.) See John O. McGinnis & Michael B. Rappaport, Majority and Supermajority Rule: Three Views of the Capitol, 85 TEX. L. REV. 1115, 1160 (2007).

(130.) See, e.g., Jonathan R. Macey, Promoting Public-Regarding Legislation Through Statutory Interpretation: an Interest Group Model, 86 COLUM. L. REV. 223, 230-31 (1986) (arguing that interest groups often control the flow of information to possibly unsuspecting legislators). Legislative reforms also focus on the role of interest groups in political support and legislative fact-finding. See e.g. Michael A. Fitts, The Vices of Virtue: A Political Party Perspective on Civic Virtue Reforms of the Legislative Process, 136 U. PA. L. REV. 1567, 1626 (1988).

(131.) Mancur Olson, Rational Ignorance, Professional Research, and Politicians' Dilemmas, in KNOWLEDGE, POWER AND THE CONGRESS 153 (William H. Robinson & Clay H. Wellborn eds., 1991).

(132.) Devins, supra note 68, at 1184 ("[L]obbyists (as well as senior staffers) understand that courts sometimes look towards legislative history and, consequently, that it is useful to pad the legislative history in ways that support their objectives.").

(133.) Russell B. Korobkin, Behavioral Analysis and Legal Reform: Rules vs. Standards Revisited, 79 OR. L. REV. 23, 49-50 (2000); see also Timur Kuran & Cass R. Sunstein, Availability Cascades and Risk Regulation, 51 STAN. L. REV. 683, 685 (1999) (defining the availability heuristic as "a pervasive mental shortcut whereby the perceived likelihood of any given event is tied to the ease with which its occurrence can be brought to mind").

(134.) See, e.g., Amos Tversky & Daniel Kahneman, Availability: A Heuristic for Judging Frequency and Probability, in JUDGMENT UNDER UNCERTAINTY: HEURISTICS AND BIASES (Daniel Kahneman, Paul Slovic, & Amos Tversky eds., 1982).

(135.) Abner Mikva, How Well Does Congress Support and Defend the Constitution?, 61 N.C.L. REV. 587, 609 (1983) ("Because of the volume of legislation, the time spent with constituents, and the technical knowledge required to understand the background of every piece of legislation, it is infrequent that a member considers the individual merits of a particular bill."); see also Elizabeth Garrett & Adrian Vermeule, Institutional Design of a Thayerian Congress, 5 DUKE L.J. 1277, 1290 (2001) ("Despite the enormous growth of congressional staff and the refined specialization of its internal structure, Congress faces tight deliberative constraints of time and information.").

(136.) Allen Schick, Informed Legislation: Policy Research versus Ordinary Knowledge, in KNOWLEDGE, POWER AND THE CONGRESS 102 (1991) ("Contradictions between ordinary and research knowledge are not uncommon. Researchers often boast that their findings are counterintuitive.... In legislating, committees and members of Congress cannot ignore ordinary thinking because their election success depends on garnering the votes of citizens who have no special expertise in policy matters.")

(137.) Cass R. Sunstein, Probability Neglect: Emotions, Worst Cases, and Law, 112 YALE L.J. 61, 86 (2002) ("When the media emphasizes particular incidents, those incidents will become cognitively available, and hence they might seem to be far more probable than they are in fact.").

(138.) Rogovin, supra note 116, at 172-73.

(139.) In contrast, voters are likely to ignore events if the effects of these events are diffused among the entire population and are thus individually small even if cumulatively large.

(140.) Mark Seidenfeld, Cognitive Loafing, Social Conformity, and Judicial Review of Agency Rulemaking, 87 CORNELL L. REV. 486, 504 (2002).

(141.) Closely related to the confirmation bias is the phenomenon of biased assimilation. which "refers to the tendency of individuals to condition their acceptance of new information as reliable based on its conformity to their prior beliefs." Kahan & Braman, supra note 130.

(142.) See Rachlinski & Farina, supra note 118, at 573-74 (arguing that the committee system makes it institutionally possible for Congress to overcome the limitations of lay membership in making decisions).

(143.) Devins, supra note 68, at 1183 (noting that "committee chairs can pack their committees with like minded thinkers, can determine when and what the committee investigates, and can arrange hearings in a way that frustrate the search for truth").

(144.) Barry R. Weingast & William J. Marshall, The Industrial Organization of Congress; or, Why Legislatures, Like Firms, Are not Organized as Markets. 96 J. POL. ECON. 132 (1988) ("committees are composed of members who are significantly above-average supporters of the relevant interest group").

(145.) William H. Riker & Barry Weingast, Constitutional Regulation of Legislative Choice: The Consequences of Judicial Deference to Legislatures, 74 VA. L. REV. 373, 387 (1988) (noting that changes in committee agenda power produce dramatic swings in agency policy).

(146.) Richard A. Epstein, The Independence of Judges: The Uses and Limitations of Public Choice Theory, 1990 BYU L. REV. 827, 833 ("Judges may not talk about pending business with any outsider: nor reveal privately the decision in any case prior to its disclosure to the immediate parties and the public: nor allow connection or money to (even appear to) influence the outcome of a case." (citing Code of Judicial Conduct (1990))).

(147.) This point tends to be ignored by critics who claim that the judiciary takes an undesirably narrow look at the facts because they are only deciding the narrow issue before them. See DONALD HOROWITZ, THE COURTS AND SOCIAL POLICY 45-56 (1977).

(148.) Thomas W. Merrill, Does Public Choice Theory Justify Judicial Activism After All?, 21 HARV. J.L. & PUB. POL'Y 219, 228 (1997) ("the judicial supply curve becomes inelastic at much lower prices than the legislative supply curve"); A.C. Pritchard & Todd J. Zywicki, Finding the Constitution: An Economic Analysis of Tradition's Role in Constitutional Interpretation, 77 N.C.L. REV. 409, 499 (1999) ("money can be used to purchase high quality legal representation, but the structure of appellate litigation ... places important limits on the marginal benefit of money to an interest group.").

(149.) Merrill, supra note 148, at 228.

(150.) See John O. McGinnis & Mark L. Movsesian, The World Trade Constitution, 114 HARV. L. REV. 511,567 (2000) (describing the effects of different processes on the incentives of lobbying by interest groups).

(151.) Frank B. Cross & Emerson H. Tiller, Judicial Partisanship and Obedience to Legal Doctrine: Whistleblowing on the Federal Courts of Appeals, 107 YALE L.J. 2155 (1998).

(152.) Pritchard & Zywicki supra note 148, at 495; Richard A. Posner, What Do Judges and Justices Maximize? (The Same Thing Everybody Else Does), 3 SUP. CT. ECON. REV. 1, 15 (1993) (noting that "a potentially significant element in the judicial utility function is reputation, both with other judges, especially one on the same court ... and with the legal profession at large.").

(153.) Jonathan T. Molot, Reexamining Marbury in the Administrative State: A Structural and Institutional Defense of Judicial Powers over Statutory Interpretation, 96 NW. U. L. REV. 1239, 1304-06 (2002) (arguing that judges "are influenced by strong institutional norms that lead them to value consistency and stability more than officials in the political branches of the government," and discussing the role of shared professional experience).

(154.) Archibald Cox, The Role of Congress in Constitutional Determinations, 40 U. CIN. L. REV. 199, 209-10 (1971); see also Devins, supra note 68, at 1179 (citing Cox and MARK TUSHNET, TAKING THE CONSTITUTION AWAY FROM THE COURTS 68 (1999)).

(155.) Cf. Andrew E. Taslitz, Temporal Adversarialism, Criminal Justice, and the Rehnquist Court." The Sluggish Life of Political Fact-finding, 94 GEO. L.J. 1589, 1607 (2006) ("Diverse groups also bring more cultural information relevant to judging credibility and are more likely to avoid error because every argument must survive group challenge. A judge does not face the same intellectual gauntlet in making his own fact-findings.").

(156.) Id. at 1607-09.

(157.) See Schick, supra note 136; see also Douglas J. Sylvester & Sharon Lohr, Counting on Confidentiality: Legal and Statistical Approaches to Federal Privacy Law After the USA Patriot Act, 2005 WIS. L. REV. 1033, 1101 (arguing that congressional representatives lack the training and expertise to avoid cognitive traps when dealing with statistics and certain other data).

(158.) See John O. McGinnis, Age of the Empirical, 137 POLICY REVIEW (2006) (describing how computers are making empirical work increasingly important to law and social science).

(159.) We recognize that juries evaluate complex statistical evidence in some cases, like those involving mass torts or monopolies. But it is in precisely these cases that commentators have doubted that the jury is competent, and many in fact have suggested that such cases be tried by judges alone. See generally, e.g. Note, The Jury's Capacity to Decide Complex Civil Cases, 110 HARV. L. REV. 1489 (1997) (discussing a number of possible alterations to the trial fact-finding process). And these cases would still involve evaluating witness credibility about specific events--valuations largely absent in social fact-finding.

(160.) See Anthony Champagne, The Politics of Criticizing Judges, 39 LOY. L.A.L. REV. 839 (2006) (arguing that recent attacks on the judiciary frequently pit religion against liberal ideology).

(161.) See Nelson Lund & John O. McGinnis, Lawrence v. Texas and Judicial Hubris, 102 MICH. L. REV. 1555, 1575 (2004).

(162.) See Stephen M. Feldman, Unenumerated Rights in Different Democratic Regimes, 9 U. PA. J. CONST. L. 47, 86 (2006) (observing that elites during the Lochner era used the judicial system to protect their economic interests).

(163.) Id. at 86-87 (after Lochner, "[elites] still sought to protect their interests and values through judicial enforcement of non-economic rights").

(164.) David L. Faigman, "Normative Constitutional Fact-Finding": Exploring the Empirical Component of Constitutional Interpretation, 139 U. PA. L. REV. 541, 6(16 (1991) ("Science has one main advantage over the other sources of interpretation: replicability. The validity of hypotheses about the world of constitutional facts does not depend on 'plain meaning,' the materials discovered in historical archives, logical argument, or moral persuasion. The validity of a factual statement depends on its amenability to testing.").

(165.) See, e.g., Erwin Chemerinsky, Turning Sharply to the Right. 10 GREEN BAG 2d 423, 425 (2007) ("The key to the [Carhart decision] was not in the difference in wording between the federal law and the Nebraska act: it was Justice Alito having replaced Justice O'Connor.").

(166.) Carhart. 127 S. Ct. at 1637-38; id. at 1643-1644 (Ginsburg, J., dissenting).

(167.) Carhart, 127 S. Ct. at 1636.

(168.) Stenberg, 530 U.S. at 937 ("[T]he division of medical opinion about the matter at most means uncertainty, a factor that signals the presence of risk, not its absence.... the uncertainty means a significant likelihood that those who believe that D & X is a safer abortion method in certain circumstances may turn out to be right. If so, then the absence of a health exception will place women at an unnecessary risk of tragic health consequences."); Carhart, 127 S. Ct. at 1643, 1647 (Ginsburg, J., dissenting) (reiterating this aspect of Stenberg and arguing that "the procedures [the Court now] deem[s] acceptable might put a woman's health at greater risk").

(169.) Einer Elhauge, Does Interest Group Theory Justify More Intrusive Judicial Review?, 101 YALE L.J. 31, 78 (1990).

(170.) The amicus brief has long been permitted in order to preserve the interests of affected third parties. See Nancy Bage Sorenson, The Ethical Implications of Amicus Briefs: A Proposal for Reforming Rule 11 of the Texas Rules of Appellate Procedure, 30 ST. MARY'S L.J. 1219, 1224-28 (1999) (describing the history and purpose of the amicus brief from Roman through modern times).

(171.) Joseph D. Kearney & Thomas W. Merrill, The Influence of Amicus Curiae Briefs on the Supreme Court, 148 U. PA. L. REV. 743,749-50 (2000). ("While the number of cases the Court has disposed of on the merits has not increased appreciably during [the last 50 years], ... the number of amicus filings has increased by more than 800%."). As we note in Section 4(B)(1), infra, amicus briefs must be presented at the district court level to the extent that they introduce legislative facts into evidence, so that parties will have an opportunity to rebut the facts submitted by the opposition.

(172.) Elhauge, supra note 169, at 78

(173.) Another objection may be that courts do not have the specialization found in congressional committees. But as we noted above, committee specialization may have more negative than positive effects.

(174.) See Saikrishna Prakash & John Yoo, Against Interpretative Supremacy, 103 MICH. L. REV. 1539, 1556 (2005) (noting responsibility of member of Congress to assess constitutionality of legislation).

(175.) See Michael B. Rappaport, The Unconstitutionality of Signing and Not Enforcing, 16 WM. & MY. BILL RTS. J. 113 (2007).

(176.) Marbury, 5 U.S. at 177-78 (1803).

(177.) See Nelson Lund, Presidential Signing Statements in Perspective, 16 WM. & MARY BILL OF RIGHTS J. 95 (2007).

(178.) See Brian Galle, The Justice of Administration: Judicial Responses to Executive Claims of Independent Authority to Interpret the Constitution, 33 FLA. ST. U. L. REV. 157, 160 (2005) (outlining the consensus and describing academic opponents of departmentalism as "few" in number).

(179.) See U.S. CONST. art VI ("The Senators and Representatives before mentioned ... and all the executive officers and judicial officers ... of the United States shall be bound by Oath or Affirmation to support this Constitution"). The President must take an "oath to preserve, protect and defend the Constitution of the United States", see U.S. CONST. art. II, sec. 1. If obligation were measured by an oath's definitiveness, the President's obligation to enforce the Constitution would be the most substantial.

(180.) See Michael Stokes Paulsen, The Merryman Power and the Dilemma of Autonomous Executive Branch Interpretation, 15 CARDOZO L. REV. 81, 84 (1993) (quoting historical evidence).

(181.) To be sure, the contours of departmentalism are not well-defined, because there is disagreement about the scope of the duties of the respective branches in constitutional interpretation. Some scholars, for instance, argue that the President has a duty not to follow even orders of the Supreme Court itself, when he believes that those orders are not constitutionally grounded. See Michael Stokes Paulsen, The Most Dangerous Branch: The Executive Power to Say What the Law Is, 83 GEO. L.J. 217, 343 (1994). Many disagree. See, e.g., Thomas W. Merrill, Judicial Opinions as Binding Law and as Explanations for Judgments, 15 CARDOZO L. REV. 43, 56 (1993). But no one disputes the content of social facts are relevant, indeed essential, to determining the constitutionality of some kinds of legislation. Nor does anyone dispute that deciding a case--the context in which the judiciary will engage in independent fact-finding--is within the core of judicial review.

(182.) JAMES WILSON, 1 WORKS OF JAMES WILSON 290-91 (Robert Green McCloskey ed., 1967).

(183.) Although we have observed that the general public has little incentive to ascertain legislative facts because of rational ignorance and rational irrationality, to the extent that inter-branch disagreement brings disputed facts to public attention that, of course, is a desirable result.

(184.) See Jerry L. Mashaw, Norms, Practices and the Paradox of Deference," A Preliminary Investigation into Agency Statutory Interpretation, 57 ADMIN. L. REV. 501,52425 (2005) (observing that administrative agencies interpret legislation for different purposes and using different methods than does the judiciary). See also John O. McGinnis, Models of the Opinion Function of the Attorney General: A Normative, Descriptive, and Historical Prolegomenon, 15 CARDOZO L. REV. 375, 398-99 (1993) (suggesting executive branch interpretation has methodology in some measure distinct from judiciary).

(185.) McGinnis, supra note 184, at 398-99.

(186.) It is true as well that most scholars believe that departmentalism permits one branch of the government to follow the precedent minted by other branches. Thus, in the vetoing a bill, the President can consider Supreme Court case law in deciding whether it is constitutional, and need not rely on his own view as if he were writing on a clean slate. See Lund, supra note 177, at 105. The issue for independent judicial fact-finding, at least in the first instance, however, does not concern precedent. The question is whether the judiciary should take an independent view of social facts relating to a statute whose constitutionality has not been previously determined.

(187.) See U.S. CONST. art. III [section] 2 ("The Supreme Court shall have appellate jurisdiction both as to Law and Fact, with such exceptions, and under such Regulations as Congress shall make.")

(188.) It is true that the exercise of power by Congress would continue to allow lower courts to find facts, but the use of the exception clause would send a strong message to the entire judiciary. For a discussion of the positive impact of public criticism (including jurisdiction-stripping threats) on judicial restraint and independence, see Viet D. Dinh, Threats to Judicial Independence, Real and Imagined. 95 GEO. L.J. 929 (2007).

(189.) See Frank B. Cross & Blake J. Nelson, Strategic Institutional Effects on Supreme Court Decisionmaking, 95 NW. U. L. REV. 1437, 1466-67 (2001) (discussing Congress' power of the purse and the use of that power to punish the Warren Court).

(190.) We have already discussed the reason that the public at large cannot be counted upon to hold Congress accountable for inaccurate fact-finding. See notes 119-127 and accompanying text.

(191.) See Kermit Roosevelt. How to Judge the Robert's Court, CHRISTIAN SCIENCE MONITOR, July 6, 2007, at 3.

(192.) See Devins, supra note 68, at 1189-92: see also David A. Strauss, Miranda, The Constitution, and Congress, 99 MICH. L. REV. 958, 973 (2001) (arguing that Congress should be given varying deference on constitutional questions based on an "assessment of Congress's own capacities and propensities.").

(193.) See notes 174-189 and accompanying text.

(194.) For discussion of the availability heuristic, see supra notes 137-139, and accompanying text.

(195.) See, e.g., Frank B. Cross, The Judiciary and Public Choice, 50 HASTINGS L.J. 355, 356-57 (1999); cf. Schick supra note 136.

(196.) Devins, supra note 68, at 1190.

(197.) Id.

(198.) See, e.g., Choper, supra note 89, at 169-70, 314-15 (concluding the Court should, as in federalism cases, leave separation of powers to the superior political process).

(199.) See Caplan, supra note 122, at 79-80.

(200.) The Line Item Veto Act of 1996 empowered the President to cancel new "direct spending," or entitlements, but did not extend that power to existing entitlement spending. See Pub. L. No. 104-130, 110 Star. 1200, [section][section] 1021(a), 1026, 2 U.S.C.A. [section] 691 (West 2006).

(201.) Alaska's "bridge to nowhere" commanded a prominent position in George Stephanopoulos's Feb. 10, 2006 20/20 presentation; available at http://abcnews.go.com/ Video/playerIndex?id=1604124; Jerry Bowen climbed that stump for 60 Minutes by August 18th that year. Available at http://www.cbsnews.com/sections/i_video/main500251. shtml?channel=60Sunday (search "bridge to nowhere"). The print media did not miss the opportunity to comment. See, e.g, William Satire, Bridge to Nowhere, N.Y. TIMES, Oct. 8, 2006; Nick Jans, Alaska Thanks You, USA TODAY, May 17, 2005; Shailagh Murray, For Senate Foe of Pork Barrel Spending, Two Bridges Too Far, THE WASHINGTON POST, Oct. 21, 2005 (and circulated in newspapers nationwide as an Associated Press article).

(202.) JOHN HENRY WIGMORE, 5 WIGMORE ON EVIDENCE [section] 2555(d) (2d ed. 1923) (emphasis in the original).

(203.) Jeffrey M. Shaman, Constitutional Fact: The Perception of Reality by the Supreme Court, 35 U. FLA. L. REV. 236, 237 (1983) ("Throughout its history, the Court has devoted little attention to developing proper methodology to deal with constitutional facts.").

(204.) 413 U.S. 49 (1973).

(205.) Id. at 60-61.

(206.) 431 U.S. 678, 695 (1977).

(207.) Id. at 695-96 n.19. In the same footnote, however, the Court upheld its tradition of citing evidence and then declaring it irrelevant, stating that "the studies ... play no part in our decision." ld.

(208.) Kenneth Culp Davis, Facts in Lawmaking, 80 COLUM. L. REV. 931,939 (1980).

(209.) See, e.g., Mississippi Univ. for Women v. Hogan, 458 U.S. 718, 724 (1982) ("[T]he party seeking to uphold a statute that classifies individuals on the basis of their gender must carry the burden of showing an 'exceedingly persuasive' justification for the classification").

(210.) City of Renton v. Playtime Theatres, Inc., 475 U.S. 41, 51 (1986) ("The First Amendment does not require a city before enacting such an ordinance, to conduct new studies or produce evidence independent of that already generated by other cities, so long as whatever evidence the city relies upon is reasonably believed to be relevant to the problem that the city addresses."): Clark v. Comty. for Creative Non-Violence, 468 U.S. 288 (1984) (granting the government's contention that administrative difficulties would result if protest groups were allowed to camp overnight on the national mall).

(211.) See, e.g., Vance v. Bradley, 440 U.S. 93, 111 (1979) ("In an equal protection case of this type [involving age discrimination], ... those challenging the legislative judgment must convince the court that the legislative facts on which the classification is apparently based could not reasonably be conceived to be true by the governmental decision-maker.")

(212.) Williamson v. Lee Optical, 348 U.S. 483, 487 (1955) ("The legislature might have concluded that the frequency of occasions when a prescription is necessary was sufficient to justify this regulation of the fitting of eyeglasses.").

(213.) The single exception to this rule occurs when plaintiffs challenge a social fact found in a case that is controlling precedent: as we discuss below, the party challenging the factual underpinning in the previous case has the burden of proof.

(214.) Aleinikoff, supra note 16, at 976.

(215.) See Lee Optical, 348 U.S. at 487-88 (1955); see also Kelo v. City of New London, Conn., 545 U.S. 469, 490 (2005) (Justice Kennedy concurring and citing the doctrine of rational basis scrutiny for economic regulation).

(216.) Even if every law passes this test, the ones that are truly absurd will be exposed as such by the government's strained attempts to plead a rational basis. For instance, when the Filled Milk Statute upheld in United States v. Carolene Products, 304 U.S. 144 (1938), was later struck by a district court in 1972, the government declined to appeal the case, perhaps out of the belief that the law was not worth preserving. See Milnot Co. v. Richardson, 350 F. Supp. 221,225 (N.D. Ill. 1972) (striking the Filled Milk Statute and declaring it "devoid of rationality").

(217.) Reno, 521 U.S. at 876; see also Federal Elections Comm. v. Wisconsin Right to Life, Inc., 127 S. Ct. 2652, 2669-70 (2007) (government bears the burden of showing that the law targets candidate endorsement rather than legitimate discussion of issues).

(218.) See Carhart v. Gonzalez, 413 F. 3d 791,796 (8th Cir 2005) ("Thus, when 'substantial medical authority' supports the medical necessity of a procedure in some instances, a health exception is constitutionally required.").

(219.) Carhart, 127 S. Ct. at 1638 ("The Act is not invalid on its face where there is uncertainty over whether the barred procedure is ever necessary to preserve a woman's health.").

(220.) Mark A. Bross, The Impact of Ornelas v. U.S. on the Appellate Standard of Review Under the Fourth Amendment, 9 U. PA. J. CONST. L. 871, 876-77 (2007) (arguing that the trial court in general is in a superior position to evaluate fact and implying that appellate resources are not sufficient to the task). Alternately, our judicial system assigns trial courts superior capacities to find adjudicative facts to achieve an efficient division of labor. See George G. Nelson, The Standard of Review in Title VII Disparate-Treatment Actions, 50 U. CHI. L. REV. 1481, 1485-86 (1983).

(221.) Cf. Christopher Onstott, Judicial Notice and the Law's "Scientific" Search for Truth, 40 AKRON L. REV. 465 (2007) (discussing various problems with taking judicial notice without evidence).

(222.) Fed. R. Evid. 201(b).

(223.) Fed. R. Evid. 201(e).

(224.) The Advisory Committee noted the disagreement in the legal literature on whether a party should be able to rebut legislative facts of which the Court takes judicial notice, but declined to take a position given the wide discretion judges have in the area of legislative facts. Fed. R. Evid. 201 Advisory Committee Note, Subdivision (g) ("Ample protection and flexibility are afforded by the broad provision for opportunity to be heard on request ...").

(225.) The lower court followed this approach in Reno by having the parties stipulate to undisputed facts concerning the internet so that the court could focus on the disputed legislative facts at issue. See id. at 849 n.2 (noting that the lower court's findings included "356 paragraphs of the parties' stipulation and 54 findings based on evidence received in open court").

(226.) McGinnis, supra note 158.

(227.) See Peter J. Hammer, Antitrust Beyond Competition: Market Failures, Total Welfare, and the Challenge of Intramarket Second-Best Tradeoffs, 98 MICH. L. REV. 849, 917-18 (2000) (observing that the sophistication of analysis and receptivity of courts to economic arguments has increased).

(228.) See United States R.R. Retirement Bd. v. Fritz, 449 U.S. 166, 192-93 (1980) (Justice Brennan dissenting and arguing that the record indicates Congress was deceived into thinking employees would not lose benefits); see also Dennis Caraher, Constitutional Law--The Rational Basis Test Becomes Less Rational--U.S. R.R. Retirement Bd. v. Fritz & Schweiker v. Wilson, 4 W. NEW ENG. L. REV. 429, 436-37 (1981-1982) (discussing how legislative history shows that Congress was deceived).

(229.) Bose Corp. v. Consumers Union of United States, 466 U.S. 485 (1984).

(230.) Compare Kassel v. Consolidated Freightways Corp., 450 U.S. 662, 674 & n.16 (1981) (dormant commerce clause case appearing to defer to findings of district court and faulting the dissent for "ignor[ing] the findings of the courts below and rel[ying] on largely discredited statistical evidence."), with Dunagin v. City of Oxford, 718 F. 2d 738, 748 n.8 (5th Cir. 1983) (reviewing Supreme Court precedents on legislative facts and finding that district court's conclusions should not receive deference).

(231.) John Monahan & Laurens Walker, Social Authority: Obtaining, Evaluating, and Establishing Social Science in Law, 134 U. PA. L. REV. 477,495 (1986).

(232.) Id.

(233.) Id. at 491.

(234.) See Hope Clinic v. Ryan, 195 F. 3d 857, 883-85 (7th Cir. 1999) (Posner, J., dissenting) ("Consistent deference in this pair of cases would lead to an inconsistent result--the upholding of one statue and the condemnation of its sister.... lain issue of legislative fact.., is not to be cabined by facts determined in an adjudicative hearing.").

(235.) 5 U.S.C. [section][section] 556-57 (1987).

(236.) See Richard B. Stewart, U.S. Administrative Law: A Model for Global Administrative Law, 68 L. & CONTEMP. PROBS. 63, 74 (2005) (describing the trial type model of formal review).

(237.) 5 U.S.C. [section] 706 (2) (A)

(238.) Id.

(239.) One interesting question is whether amicus participation alone at the district court stage will provide the optimal kind of intervention to improve the lower court factual record. Permitting interested parties to introduce evidence and rebut the evidence introduced by others may improve the factual accuracy of lower court determinations. If so, the Advisory Committee on the Rules of Civil Procedure might considering giving judges the discretion to permit a stronger form of intervention than amicus participation for factual development. The same reforms should be encouraged at the state level, where lower courts often decide constitutional issues of legislative fact that are subsequently reviewed by the Supreme Court. See, e.g., Simons v. Roper, 112 S.W.3d 397 (Mo. 2003) (finding that there was a national and international consensus against the juvenile death penalty and that the youth of juveniles make them less culpable for their crimes and more likely to be falsely convicted).

(240.) See Gonzales v. Carhart, 127 S. Ct. at 1610.

(241.) See Carhart v. Gonzales, 413 F. 3d at 791.

(242.) See Carhart, 127 S. Ct. at 1638 (observing that "zero tolerance" adjudication would inhibit Congress's ability to act when differences of opinion exist and relying on a finding that "there is uncertainty").

(243.) Quotations of John Maynard Keynes, printed at http://www.history.mcs.standrews.ac.uk/history//Quotations/Keynes.html

(244.) 505 U.S. 833, 863 (1992) (citing West Coast Hotel v. Parrish, 300 U.S. 379 (1937) and Brown v. Bd. of Educ., 347 U.S. 497,483 (1954)).

(245.) The Court stated that "the Depression had come, and with it, the lesson that seemed unmistakable to most people in 1937, that the interpretation of contractual freedom protected in Adkins [v. Children's Hospital, 261 U.S. 525 (1923)], rested on fundamentally false factual assumptions about the capacity of a relatively unregulated market to satisfy minimal levels of human welfare." Nothing in the West Coast Hotel opinion itself nor in the historical consensus about the Depression suggests that freedom of contract jurisprudence led to the Depression, or establishes that a relatively unregulated market cannot satisfy minimal levels of human welfare. See Alan J. Meese, Will, Judgment and Economic Liberty: Mr. Justice Souter's Mistranslation of the Due Process Clause, 41 WM. & MARY L. REV. 3, 44-52 (1999).

(246.) See Arkwright-Boston Mrfs. Mut. Ins. Co. v. Calvert Fire Ins. Co., 887 F. 2d 437, 442 (2d Cir. 1989) (observing that certainty may reduce legal costs); Planned Parenthood of Southeastern Pa. v. Casey, 505 U.S. 833,854-55 (1992) (stating that the Court considers reliance on a legal rule and its continuing effectiveness rather than simply any changed factual situation). The Court seems to take this approach in Dickerson when it stated that "Miranda has become embedded in routine police practice to the point where the warnings have become part of our national culture." 530 U.S. at 443.

John O. McGinnis, Professor of Law, Northwestern University.

Charles W. Mulaney, Associate, Perkins Coie LLP. Thanks to Professors Steven Calabresi, Neal Devins, Lee Epstein, Heidi Kitrosser, Nelson Lund, Mark Movsesian, Marty Redish, and Jim Pfander for helpful comments. We are also grateful for suggestions from the audience at faculty scholarship day at Northwestern Law and in the public law and political science workshop at William and Mary Law School.