Troubling Transparency: The History and Future of Freedom of Information

THE FREEDOM OF INFORMATION ACT¹ creates legal rights. Specifically, it creates certain rights of access to certain federal government documents. But understanding FOIA requires understanding what kind of rights the act creates and what it is for there to be a “right to know,” as the ubiquitous phrase expresses it.² In this chapter, I explore the right to know from the perspective of various topics in the theory of rights more generally, especially, but not exclusively, with reference to the important distinction between positive and negative rights. Understanding this distinction facilitates situating FOIA within a host of broader and deeper legal and constitutional questions. Thus, when Justice Potter Stewart, himself a noteworthy supporter of strong and sometimes special First Amendment rights for the institutional press,³ observed that “[t]he Constitution itself is neither a Freedom of Information Act nor an Official Secrets Act,”⁴ he distinguished between the First Amendment, which creates (or recognizes) rights, and the Freedom of Information Act, which also creates rights. But did Justice Stewart mean to suggest that there is no freedom of information component to the First Amendment? If so, was he right? A look at FOIA through the lens of the theory of rights may help to answer these questions.

About a century ago, the legal theorist Wesley Hohfeld sought to clear up a cluster of confusions about the very idea of a “right.”5 Hohfeld believed that the same word—right—was used to label (at least) four different legal relationships, and his enduring analysis and taxonomy of what he called jural relations and the connections among them (he called them “opposites” and “correlatives”) is a useful starting point for understanding what rights are and how they work.

In Hohfeld’s taxonomy, some rights create or recognize⁶ “privileges,” which others have labeled “liberties.”⁷ These are rights to do something without interference, and most of the familiar U.S. constitutional rights are of this variety. The First Amendment right to the free exercise of religion, for example, is a right against state interference with (and restriction of) one’s religious beliefs, and sometimes with religious practices.⁸ Similarly, the Second Amendment right to keep and bear arms is a right not to be restricted in aspects of one’s gun ownership,⁹ and much of the First Amendment right to freedom of speech is a right against state interference with speaking or writing.¹⁰ Because such rights require nothing from the state except inaction, they are understood as “negative” rights, or liberties.¹¹

Negative rights need not be rights against the world. Under U.S. “state action” doctrine, for example, constitutional liberties are rights against state, but not non-state, restriction.¹² Students at government institutions such as the University of Michigan thus have (some) First Amendment rights against (some) restrictions on (some) speaking and writing, but students at a private institution, such as Columbia University, have no such rights at all. Although the New York Times has liberty rights protecting the Times from sanctions by the United States or the State of New York for what it prints, reporters for the Times have no such rights against the Times itself.

In contrast to negative rights, other rights, often described as “positive,”¹³ require the entity against which the rights are claimed to do something rather than simply refrain from doing something. A right against my landlord to have him provide heat,¹⁴ for example, is not merely the right to keep the landlord from interfering with my providing my own heat with my own resources but is also the right to have him take the positive act of providing that heat. In the constitutional context, the Sixth Amendment right to counsel is, since 1963,¹⁵ not simply the negative right against the state interfering with my right to be represented by a lawyer in a criminal case against me but also the positive right to have the state take the positive action of (sometimes) providing me with a lawyer at the state’s expense. But this right to counsel is a positive right only against the state, not against the world. For example, Clarence Earl Gideon had a right to have Florida provide him with a lawyer, but he did not have a right to have an individual Florida citizen do the same.

In addition to being only against some but not all entities, liberty rights are typically held by some people or entities but not by others. As the current controversy about corporate free speech and freedom of religion rights illustrates,16 the existence of a right does not mean that every person or entity will possess or be able to enforce it.¹⁷ Less controversial, infants are widely understood not to have free speech rights,¹⁸ and corporations have no rights against self-incrimination.¹⁹ An important and persistent question about both positive and negative rights is just who has and can enforce them.²⁰

Just as rights are neither possessed by everyone nor against everyone, nor are they rights to do everything. The right of conscientious objection to conscription, for example, is a right to an exemption on the basis of religious objection to war in general but not on the basis of objection to particular wars.²¹ Less controversial, the right to free speech is not a right to engage in non-speech conduct,²² although it is more controversial whether the right includes the right to spend money on political campaigns,²³ make misleading statements in securities offerings,²⁴ or engage in public nude dancing for paying customers.²⁵

In summary, it is important in discussing particular rights to specify (1) who has the right, (2) what the right is a right to do, (3) whom the right is a right against, and (4) what the entity against which the right is claimed is required to do or not do.

The rights provisions of the U.S. Constitution, at least as interpreted by the courts, largely protect negative and not positive rights, with the right to counsel one of very few exceptions. A woman has a constitutional right to obtain an abortion,²⁶ but under current doctrine she has no constitutional right to financial or other assistance from the government in obtaining it.²⁷ There is a right to freedom of speech but no right to the education that would make such speech effective.²⁸ There is a right to keep and bear arms but no right to government assistance for the citizen who cannot afford a gun, just as the right to the free exercise of religion does not create a right to government assistance in erecting houses of worship.

More generally, the U.S. Constitution has long been interpreted as not containing constitutional rights to welfare,29 health care, housing, a minimum wage, a pension, protection against violence or mistreatment,³⁰ or many of the other things most Americans would think, the Constitution aside, were components of a minimally satisfactory existence. Securing such positive social rights or social welfare rights, as they are often called,³¹ has generally (as a descriptive and historical matter) not been part of U.S. constitutional doctrine. Why this is so is a complex historical, philosophical, cultural, political, and economic question, but it is not irrelevant that the Constitution of the United States was drafted and gained its judicial and public acceptance long before the very notion of enforceable citizen entitlements against government, whether constitutional or otherwise, had much currency. In this respect, the U.S. Constitution differs from those constitutions whose origins coincided with or postdated twentieth-century understandings of the very idea of the welfare state.

In addition to the “exception” emerging out of Gideon v. Wainwright, other exceptions to the negative rights focus of U.S. constitutionalism occasionally surface. The First Amendment right of affirmative access to streets, parks, and sidewalks might be considered to be a positive right, especially insofar as it is based on the view that freedom of speech would be essentially meaningless were there no public places at which to exercise it.³² Some state constitutions do recognize the positive right to education (and occasionally other needs) that the federal Constitution, as interpreted, appears to disclaim.³³ But the U.S. Constitution, unlike these state constitutions,³⁴ and unlike many constitutions around the world,³⁵ remains largely a protector of negative rights and not a creator, recognizer, or enforcer of positive rights.

The largely negative character of the U.S. Constitution—far more “freedom from” than “freedom to” or “ability to”—has long been reflected in First Amendment doctrine. With respect to information (and, a fortiori, opinions), individuals and the press are largely free to publish and distribute the information they happen to have or happen to obtain, and, indeed, that freedom generally extends even to information originally obtained unlawfully.³⁶ But despite long efforts by the institutional press and its allies, there as yet exists no positive right to obtain that information in the first place. The press is free to publish The Pentagon Papers, but it had no constitutional right to demand that document from the government. More broadly, the publication of “leaked” documents and information is protected, but the First Amendment provides no assistance to a publisher who wishes to obtain government information without the benefits of a willing leaker.³⁷

Thus, when in a series of 1970s cases the press claimed a First Amendment right of positive access to various government facilities, these efforts were overwhelmingly unsuccessful.38 In the context of a claim to access to prisons, for example, Chief Justice Burger, writing for the plurality in Houchins v. KQED,³⁹ summarized the prevailing judicial conclusion: “[N]either the First Amendment nor the Fourteenth Amendment mandates a right of access to government information or sources of information within the government’s control.”⁴⁰ For complex doctrinal and historical reasons, access to criminal trials has been treated differently,⁴¹ and even here the Supreme Court has made clear that the First Amendment right emerges largely out of historical practice specific to the openness of criminal trials. But apart from this right of public access to criminal trials, a right enjoyed by members of the press as members of the public rather than because of any special status for reporters or the press, the doctrine that has explicitly prevailed for almost half a century (and implicitly for much longer than that) is that information controlled by the government need not, as a matter of constitutional law, be disclosed directly to interested or curious members of the public. There are many circumstances in which Congress may demand information from the other branches of government,⁴² information that Congress or its individual members may then reveal to the public. But from the perspective of the First Amendment specifically, and constitutionally law generally, a government agency that wishes to keep some or even all of its documents and proceedings from direct public access is constitutionally permitted to do so.⁴³

“Information” is a broad concept, but in practice the lack of a general First Amendment right of access to information has two principal manifestations. One is the lack of a constitutional right of access to government facilities and government meetings. The other is the equivalent absence of a constitutional right to obtain or examine government documents. Together these two aspects of the lack of a positive First Amendment right to access to information produce the conclusion that Chief Justice Burger accurately summarized, and, of course, helped to create. Although Justice Stewart and Chief Justice Burger disagreed about the results in many of the important cases involving press access or press privileges, Justice Stewart’s observation that the First Amendment is not “a Freedom of Information Act”44 is an accurate statement of the doctrine, and he might have added as well that it is not a “sunshine” law, insofar as we understand such state and federal laws to require public and press access to government meetings of countless varieties.⁴⁵ Thus the absence of a positive freedom of information component in existing First Amendment understandings—the absence of a positive constitutional right to access to information—reflects both a First Amendment and a constitutional philosophy more generally. The rights to freedom of speech and freedom of the press are negative rights to noninterference, but they are not positive rights to obtain the information that in many aspects is necessary to make the negative right effective.⁴⁶

In light of the foregoing, one way of understanding the Freedom of Information Act is as “remedying” the absence of a positive constitutional right to government information and thus fulfilling some sort of constitutional or quasi-constitutional function.⁴⁷ But “remedying” is placed in scare quotes here for two reasons. One is that the idea of a remedy suggests some attempt to repair or compensate for that which has already occurred. The 1966 Freedom of Information Act slightly predates the Supreme Court cases on access,⁴⁸ and thus it is more accurate to understand FOIA as being part of a more or less simultaneous two-pronged effort on the part of the press and various other transparency advocates to gain an enforceable right to government information. FOIA was largely successful, whereas the attempt to have the First Amendment interpreted as providing positive rights to government information was, for the most part, unsuccessful.

In addition, the language of “remedy” implies that the constitutional approach is primary and is to be preferred, with the statutory approach being the backup. But I want to challenge this implication. Assuming that substantial citizen and press access to government documents is a good thing,⁴⁹ as a normative policy matter, is it better, holding constant the amount of access, for that access to be achieved by statute or by judicial interpretation of the First Amendment?

The arguments for the constitutional approach are familiar. One of them is that giving a right the aura of constitutional protection will make the right seem more important and will thus inculcate for the public and for officials the genuine internalization of that value. There has never been a Supreme Court case involving restriction of a religious belief or practice just because of its religious provenance, for example. One reason for this might be that the impropriety of such a course of action is by now self-evident to lawyers, judges, political figures, and to the larger community. As a result, there is little more reason to believe that a state or the federal government would prohibit the religious practice of a particular religion than that a sitting president would run for a third term in violation of the Twenty-Second Amendment, or that federal prosecutors would attempt a treason prosecution with only one witness in violation of the two-witness requirement in Section 3 of Article III.50 One argument for constitutionalizing the positive right of access to some or most government information, therefore, is that constitutionalization, even if by judicial decision rather than by plain text as in the examples just mentioned, will have considerable value in shaping public opinion, a value not present, or less present, when the same goal is embodied in a “mere” statutory protection.⁵¹

Although this argument for the effect of constitutionalization on public and official attitudes has some surface and logical appeal, empirical evidence supporting the argument turns out to be thin. Experimental research indicates that people rarely choose legal or constitutional values over first-order substantive preferences when the two conflict,⁵² and the same holds for the actions of the public officials who are typically responsive to public opinion.⁵³ Moreover, there is some evidence, consistent with the foregoing, that people’s institutional preferences, if any, are largely determined or dominated by their first-order substantive preferences.⁵⁴ Other research suggests that Supreme Court opinions recognizing various values have little positive effect on the acceptance of those values⁵⁵ and may even, at times and on some issues, have the reverse effect.⁵⁶ Much of the evidence is complex and contested, and much of it does not precisely focus on the comparative effect of statutes when compared to judicial (or Supreme Court, specifically) constitutional decisions. At the very least, however, we can conclude that there is little evidence supporting the proposition that constitutionalizing a right of access to government information through judicial opinions will have greater public opinion shifting effects than simply embodying the same right in a statute.

Of course, it is, in theory, much easier to repeal or modify an “ordinary” statute than it is to amend the Constitution. Repealing a statute requires only a majority of the House of Representatives and of the Senate, followed by the president’s signature, but amending the Constitution requires the votes of two-thirds of both houses followed by the usually impossible ratification of three-quarters of the states. As a result, another argument for the constitutional approach to access to government information generally and government documents in particular is that statutes are vulnerable to the vagaries of politics in ways that the more stable and more entrenched Constitution and its interpretation are not. Embodying the positive right to government information in a series of constitutional judicial decisions⁵⁷ might thus be thought to entrench the right to access to government information in ways that ordinary legislation cannot.

This argument is entirely consistent with the fact of constitutional supremacy, as well as being consistent with the practical difficulty of amending the document itself. Yet although the argument might have seemed plausible in the 1960s, when FOIA was in its infancy, it is far from clear that the plausibility of the view that constitutional interpretations are more stable and entrenched than statutes persists today. Assisted by the fact that FOIA has the press as its very powerful advocate—“Never argue with the fellow who buys ink by the barrel,” as the old political adage has it—and assisted as well by FOIA’s increasing age, it seems no more likely that the act would be repealed or that its force would be substantially blunted by legislative action than that the same could happen to the 1890 Sherman Antitrust Act, to the Civil Rights Act of 1964, to the Social Security Act of 1935, or to the National Environmental Policy Act of 1970. None of these laws has constitutional status and all could thus be repealed without constitutional objection, but as a practical political and sociological matter, all of these laws, and others, are as deeply entrenched as any Supreme Court opinion, and arguably more than most.58

Other arguments also weigh in favor of the statutory approach. One such argument starts with the premise that freedom of information is not free. The compliance costs of FOIA are considerable, and although some of these costs might be worth bearing, or at least so we can assume here, it is nevertheless plausible to conclude that, in a world of limited resources, the allocation of those resources in a democracy ought to be made by a representative body able to evaluate the benefits that those costs bring compared to potential alternative uses of the same resources. Congress, or a parallel state legislative body, can decide how much access to information is worth, for example, how much health care, or how much police protection, or how much education, but it is not entirely obvious that courts can or should make these decisions.⁵⁹ By treating the costly process of access to information as a subject for judicial interpretation and enforcement, the constitutional approach may risk allocating scarce government resources on the basis of who gets to the courthouse first rather than on some more systematic foundation. When a court says to Congress that it must allocate some amount to access to information because the Constitution requires it, as would be the case were positive access to information a requirement of the First Amendment, and when Congress then finds itself with fewer resources for health care or environmental protection or national defense, for example, there is a risk of winding up with an allocation process even less rational than the one we have now.

Moreover, the very process of lawmaking by litigation is slow, inefficient, and suboptimal in other ways. Courts tend to decide cases one at a time,60 and thus they tend to decide issues one at a time.⁶¹ Had the Supreme Court decided Houchins v. KQED differently, for example, and held that the press had a First Amendment right of access to prisons for the purpose of interviewing prisoners, there would then inevitably have followed litigation about whether that decision extended to documents as well as to personal physical access, and thereafter litigation about who was entitled to access and who was entitled to information, and so on and so on. The law, in common law fashion, would have developed incrementally in the context of particular controversies, and it would have developed slowly and unsystematically. There may well be advantages to regulation by litigation, or rights creation by litigation, but the process is invariably more piecemeal, much slower, and much less systematic than an initial pronouncement and specification of a large range of applications.⁶² It is true that the 1966 Freedom of Information Act has been amended on numerous occasions, has been supplemented by other laws,⁶³ and has been explicated by literally thousands of court cases. But the right of access to federal government information, as we now know it, nevertheless is built on a framework set forth in 1966, a framework that might have taken decades to develop had the principal vehicle for developing the right been constitutional litigation rather than a single and at least somewhat, for all of its gaps and indeterminacies, comprehensive congressional enactment.

This volume is about FOIA and not about the idea of positive rights more generally. Nevertheless, many of the themes surrounding the positive right of access to government information have surfaced worldwide as various constitutional regimes deal with the question of positive social welfare rights and other positive constitutional rights.64 Many of the considerations raised here about a positive right of access to information, and especially one found in the Constitution and enforced by courts as a matter of constitutional interpretation, bear some resemblance to issues raised by decisions such as those of the South African Constitutional Court recognizing the positive right to housing⁶⁵ and the positive right to antiretroviral medication,⁶⁶ of decisions in Eastern Europe dealing with judicial determination of minimum levels of pension and salary payments⁶⁷ and many others, and of issues in Ireland and India as those countries negotiate the existence of positive constitutional rights that are designated by their respective constitutions as judicially unenforceable.⁶⁸ There is undoubtedly a place for positive legal rights, and undoubtedly a place for a positive right to access to government information, but experiences elsewhere suggest that the question whether that place is in a constitution and is enforced by the normal methods of constitutional adjudication is hardly one with an easy answer.

The question of a positive right of access to government information is merely one component of larger questions about positive rights generally, and even that latter question is merely one component of still larger questions about the nature and structure of rights. As it now exists, FOIA is a statutory grant of positive rights, granted to all persons,⁶⁹ and consisting of a right against most agencies of the federal government to obtain access to many of the documents used by those agencies. Virtually every word in the previous sentence is subject to explanation and controversy, but by understanding just what kind of rights the rights created by and embedded in FOIA are, we will have made a start in understanding what the act does and what it does not, and what it can do and what it cannot.

3. See Richmond Newspapers, Inc. v. Virginia, 448 U.S. 555, 598 (1980) (Stewart, J., concurring in the judgment); Zurcher v. Stanford Daily, 436 U.S. 547, 570 (1978) (Stewart, J., dissenting); Pittsburgh Press Co. v. Pittsburgh Commission on Human Relations, 413 U.S. 376, 400 (1973) (Stewart, J., dissenting); Branzburg v. Hayes, 408 U.S. 665, 725 (1971) (Stewart, J., dissenting); Potter Stewart, “ ‘Or of the Press,’ ” Hastings Law Journal 26 (1975): 631–37.

6. I say “recognize” to allow for the possibility that some constitutional or legal rights recognize and enforce preexisting moral or natural rights, as perhaps with the Eighth Amendment right to be free from “cruel and unusual punishments” and the First Amendment right to the “free exercise” of religion, whereas other legal or constitutional provisions create rights that have no prelegal natural or moral status, as with the (now uniquely American) Seventh Amendment right to jury trial in civil cases. On the distinction, see Ronald Dworkin, A Matter of Principle (Cambridge, Mass.: Harvard University Press, 1985), 375–80.

8. See, for example, Church of the Lukumi Babalu Aye, Inc. v. City of Hialeah, 508 U.S. 520 (1993); West Virginia State Board of Education v. Barnette, 319 U.S. 624 (1943). In practice, “pure” free exercise cases rarely arise in the United States, at least in the sense of laws singling out a religious practice for restriction because of its religious nature. Most free exercise cases involve either seemingly neutral restrictions that in reality are aimed at a particular religion, as with the prohibition on animal sacrifice in Church of the Lukumi Babalu Aye, or claims of exemption from laws of genuinely general application in intent and design that nevertheless incidentally turn out to restrict the religious practices of adherents of particular religions, as with the flag salute requirement in Barnette or the mandatory schooling requirement in Wisconsin v. Yoder, 406 U.S. 205 (1972), which the Supreme Court held could not constitutionally be applied to the Old Order Amish.

10. The “pure” examples include protections against government criminal sanctions against speaking or writing because of the content of the speech or writing, as in United States v. Alvarez, 567 U.S. 709 (2012), which invalidated a statute prohibiting falsely claiming to have won the Medal of Honor; Brandenburg v. Ohio, 395 U.S. 444 (1969), which recognized the right to advocate illegality short of explicit incitement to imminent violence; and Kingsley International Pictures Corp. v. Regents, 360 U.S. 684 (1960), which stated that the freedom to advocate even the propriety of adultery is “at the very heart of constitutionally protected liberty” (per Stewart, J.).

12. See Civil Rights Cases, 109 U.S. 3, 11 (1883), announcing that “individual invasion of individual rights is not the subject-matter of the [Fourteenth] amendment.” See also Jackson v. Metropolitan Edison Company, 419 U.S. 345 (1974); Moose Lodge v. Irvis, 407 U.S. 163 (1972). Numerous Supreme Court cases have dealt with the question of what kind and amount of connection between a non-state entity and the state is sufficient to treat the actions of the non-state entity as actions of the state for constitutional purposes. See, for example, Marsh v. Alabama, 326 U.S. 501 (1946), which treated a “private” company town as the state for First Amendment purposes; Burton v. Wilmington Parking Authority, 365 U.S. 715 (1961), which found sufficient “entanglement” between a private restaurant and the state as to render the restaurant’s discriminatory actions state action. But that area of constitutional doctrine is far from the principal concerns of this chapter.

16. Compare Citizens United v. Federal Election Commission, 588 U.S. 310 (2010), and First National Bank of Boston v. Bellotti, 435 U.S. 765 (1978), recognizing corporations as right-holders under the First Amendment’s speech and press clauses, with Robert C. Post, Citizens Divided: Campaign Finance Reform and the Constitution (Cambridge, Mass.: Harvard University Press, 2014), arguing that corporations have First Amendment free speech rights only as derivative from and for the benefit of natural persons. See also Frederick Schauer, “Constitutions of Hope and Fear,” Yale Law Journal 124 (2014): 528–62. With respect to the free exercise of religion (as embodied in a statute, the Religious Freedom Restoration Act of 1993, 42 U.S.C. § 2000 (2012), and not as a matter of constitutional law), see Burwell v. Hobby Lobby Stores, Inc., 134 S. Ct. 2751 (2014).

31. See, for example, Katharine G. Young, Constituting Economic and Social Rights (Oxford: Oxford University Press, 2012). Such rights are often referred to as “second-generation” rights. See Dieter Grimm, “The Protective Function of the State,” in European and US Constitutionalism, ed. Georg Nolte (Cambridge, UK: Cambridge University Press, 2005), 137–55, at 145–46. See also T. H. Marshall’s distinction among civil rights, political rights, and social rights, with social rights being both more recent and encompassing the kinds of positive rights often described as social welfare rights. T. H. Marshall, Citizenship and Social Class, and Other Essays (Cambridge, UK: Cambridge University Press, 1950).

36. The existing cases are typically ones in which the publisher has legally obtained information from someone who has obtained or, often, stolen it illegally. The prominent example in the Supreme Court is New York Times Co. v. United States (the Pentagon Papers Case), 403 U.S. 713 (1971), and others are Bartnicki v. Vopper, 532 U.S. 14 (2001), protecting the right to publish information originally obtained by an unlawful wiretap, and Landmark Communications, Inc. v. Virginia, 435 U.S. 829 (1978), upholding the right of a newspaper to publish confidential court records obtained unlawfully by a court employee. In Bartnicki and Landmark Communications, the Supreme Court was careful to emphasize that the publisher played no part in the original illegality. Whether a publisher can be punished for publishing information that the publisher has obtained unlawfully, or where the publisher has been more involved with the original illegality than was the case with respect to the Pentagon Papers or the material in Bartnicki and Landmark Communications, remains an open question.

37. See generally, Rahul Sagar, Secrets and Leaks: The Dilemma of State Secrecy (Princeton, N.J.: Princeton University Press, 2013); David E. Pozen, “The Leaky Leviathan: Why the Government Condemns and Condones Unlawful Disclosures of Information,” Harvard Law Review 127 (2013): 512–635. For suggestions that existing legal doctrine ought to be changed to provide First Amendment protection for the original leaker, see Heidi Kitrosser, “Leak Prosecutions and the First Amendment: New Developments and a Closer Look at the Feasibility of Protecting Leakers,” William and Mary Law Review 56 (2015): 1221–77; Mary-Rose Papandrea, “Leaker Traitor Whistleblower Spy: National Security Leaks and the First Amendment,” Boston University Law Review 94 (2014): 449–544.

41. Richmond Newspapers, Inc. v. Virginia, 448 U.S. 555 (1980). See also Globe Newspaper Co. v. Superior Court, 457 U.S. 596 (1982); Press-Enterprise Co. v. Superior Court, 464 U.S. 501 (1984); Press-Enterprise Co. v. Superior Court, 478 U.S. 1 (1986). An earlier case, Gannett Co. v. DePasquale, 443 U.S. 368 (1979), had held that a right of public access (in that case to a pretrial suppression hearing) was not required by the Sixth Amendment right to a public trial. Richmond Newspapers did recognize a constitutional right of public access and located that right in some combination of historical practice and a First Amendment right to receive information and ideas.

43. A government agency cannot, as the above cases and discussion make clear, enforce this secrecy by prohibiting the publication of such information, or by sanctioning a publisher who has lawfully obtained it. But an agency that wishes to guard its documents and information with strong walls, good locks, and competent security guards is, as far as the First Amendment is concerned, fully entitled to do so.

45. See Kathy Bradley, “Do You Feel the Sunshine? Government in the Sunshine Act: Its Objectives, Goals, and Effect on the FCC and You,” Federal Communications Law Journal 49 (1997): 473–89; “Report and Recommendations by the Special Committee to Review the Government in the Sunshine Act,” Administrative Law Review 49 (1997): 421–28.

47. See Seth F. Kreimer, “The Freedom of Information Act and the Ecology of Transparency,” University of Pennsylvania Journal of Constitutional Law 10 (2008): 1012–80, at 1013; David E. Pozen, “Deep Secrecy,” Stanford Law Review 62 (2010): 257–339, at 314 n. 204. The extent to which the idea of transparency or freedom of information is small-“c” constitutional depends on one’s views both of democracy and of a constitution, issues that are far too large to take on here. Still, it is worth noting the possibilities that not everything that facilitates openness in government is or ought to be considered a necessary component of democracy, and, less controversially, that not everything that is a component of democracy is or ought to be thought of as constitutional.

48. The Freedom of Information Act was itself preceded by the access to information provisions in Section 3 of the Administrative Procedure Act, Pub. L. No. 79–404, 60 Stat. 238 (1946), but those provisions were widely understood to be more concerned with ensuring secrecy than with providing increased access to information. See Norman Dorsen, Paul Bender, and Burt Neuborne, Political and Civil Rights in the United States, Volume I, 4th ed. (Boston, Mass.: Little, Brown, 1976), 203–04.

49. It is not quite self-evident that openness is an unqualified good. Not only do rights to information at times interfere with the right of informational privacy (as in, for example, Florida Star v. B.J.F., 491 U.S. 524 (1989), where, to oversimplify, the right to public information about a trial was deemed to outweigh the privacy interests of the juvenile rape victim), but these rights also may interfere with the efficient or otherwise optimal operation of government. The law of privileges, after all, is based on the view that immunity from compelled disclosure—secrecy, in a word—in various trial and nontrial contexts will, at times, produce more and not less candor and thus more and not less information. See Christopher B. Mueller and Laird C. Kirkpatrick, Evidence, 5th ed. (New York: Wolters Kluwer, 2012), § 5.1, at 294–95. Perhaps ironically to some, therefore, the quest by the press for more information is, in the press’s view, sometimes facilitated by laws ensuring access and sometimes by privileges—shield laws, most obviously—that limit access to information. Not only will transparency—access to information—sometimes produce less information, at times it also may produce less sound decisions. See Frederick Schauer, “Transparency in Three Dimensions,” University of Illinois Law Review (2011): 1339–57. For the view in another but related context that more information may produce worse judgments, see Omri Ben-Shahar and Carl E. Schneider, More Than You Wanted to Know: The Failure of Mandated Disclosure (Princeton, N.J.: Princeton University Press, 2014).

53. Schauer, The Force of Law, 75–92, 205–11. See also Frederick Schauer, “Official Obedience and the Politics of Defining ‘Law,’” Southern California Law Review 58 (2013): 1165–94; “The Political Risks (If Any) of Breaking the Law,” Journal of Legal Analysis 4 (2012): 83–101; “Is Legality Political?,” William and Mary Law Review 53 (2011): 481–506; and “How and When (If at All) Does Law Constrain Official Action?,” Georgia Law Review 44 (2010): 769–801.

55. This is the conclusion, drawing on and summarizing much of the primary research at the time, in Gerald N. Rosenberg, The Hollow Hope: Can Courts Bring About Social Change?, 2nd ed. (Chicago, Ill.: University of Chicago Press, 2008), 25–36. Rosenberg’s conclusions have been highly controversial, and there is evidence that in some areas judicial constitutional decisions may, even if through an indirect causal mechanism, produce attitudinal change. See, for example, Thomas M. Keck, “Beyond Backlash: Assessing the Impact of Judicial Decisions on LGBT Rights,” Law and Society Review 43 (2009): 151–86; Scott L. Cummings, “Empirical Studies of Law and Social Change: What Is the Field? What Are the Questions?,” Wisconsin Law Review (2013): 171–204. A good broader overview of research on law and attitudinal and behavioral change is Kenworthey Bilz and Janice Nadler, “Law, Moral Attitudes, and Behavioral Change,” in Oxford Handbook of Behavioral Law and Economics, ed. Eyal Zamir and Doron Teichman (New York: Oxford University Press, 2014), 241–67. Like much of the post-Rosenberg research in the field, however, it does not distinguish the attitudinal influence of constitutional judicial decisions from the attitudinal influence of law generally. But if the question is whether to achieve some goal by constitutional judicial decisions as opposed to statutes, this is precisely the distinction that is important.

58. See William N. Eskridge Jr. and John Ferejohn, A Republic of Statutes: The New American Constitution (New Haven, Conn.: Yale University Press, 2010); William N. Eskridge Jr. and John Ferejohn, “Super-Statutes,” Duke Law Journal 50 (2001): 1215–76; Kathryn E. Kovacs, “Superstatute Theory and Administrative Common Law,” Indiana Law Journal 90 (2015): 1207–60.

59. See Ronald Dworkin, Law’s Empire (Cambridge, Mass.: Harvard University Press, 1986), 178–84, arguing that “checkerboard” policies requiring allocation of resources across competing claims of policy (but not of principle) should be made by legislatures and not by courts. In the context of FOIA, however, there is little evidence that Congress actually carefully considered these allocative questions, however desirable it might have been had Congress done so. As a result, application of FOIA might at times be more arbitrary than is optimal, as is argued in David E. Pozen, “Freedom of Information Beyond the Freedom of Information Act,” University of Pennsylvania Law Review 165 (2017): 1097–1158. The issue is complex because at times the unintended consequences of congressional action may have useful dimensions. Consider, for example, the sometimes lamented (see Margaret B. Kwoka, “FOIA, Inc.,” Duke Law Journal 65 [2016]: 1361–1437) domination of FOIA requests by self-interested corporations as opposed to journalists and members of the public. Given that the vast majority of federal administrative regulation involves the regulation of corporations, FOIA, even as so used, might serve a valuable even if unintended due process function—giving access to information about regulation to those who are most directly affected by the regulation.

61. Some of us wish it were otherwise, and that courts, especially the Supreme Court, would acknowledge their law-making function and thus be more willing to set out general rules for the guidance of lower courts and primary political and policy actors. See Frederick Schauer, “Our Informationally Disabled Courts,” Daedalus 143 (2014): 105–14; “Abandoning the Guidance Function: Morse v. Frederick,” Supreme Court Review (2007): 316–48; and “Opinions as Rules,” University of Chicago Law Review 62 (1995): 1455–75.

67. See Julia Szalaï, “Fragmented Social Rights in Hungary’s Postcommunist Welfare State,” in Policy and Citizenship: The Changing Landscape, ed. Adalbert Evers and Anne-Marie Guillemard (Oxford: Oxford University Press, 2012), chap. 12; Andras Sajo, “Implementing Welfare in Eastern Europe After Communism,” in Economic, Social and Cultural Rights in Practice: The Role of Judges in Implementing Economic, Social and Cultural Rights, ed. Yash Ghai and Jill Cottrell (London: InterRights, 2004), 50–57.

69. In a comprehensive recent study, Kwoka, “FOIA, Inc.,” documents the extent to which the vast majority of claimants under FOIA at certain agencies are corporations and neither individual citizens nor individual journalists. Kwoka views this state of affairs with alarm, but it is not so clear why it should be thought of as alarming. It should be no surprise that the subjects of regulation would be disproportionately interested in obtaining the information relevant to the regulation of their activities. Indeed, we might think of FOIA as embodying, in part, something of a due process interest insofar as it provides the kind of information to regulatory subjects that the subjects might desire in order to comply with, resist, or seek to change the regulation.